To begin searching the abstracts, please use the search feature above.
| 1818 | Computer 1
|
Self-supervised contrastive learning for motion artifact detection in whole-body MRI: Quality assessment across multiple cohorts |
| Thomas Küstner1, Jan Borst1, Dominik Nickel2, Fabian Bamberg3, Marcel Früh1, and Sergios Gatidis1,4 | ||
1Medical Image and Data Analysis (MIDAS.lab), Department of Diagnostic and Interventional Radiology, University Hospital of Tuebingen, Tuebingen, Germany, 2Siemens Healthineers, Erlangen, Germany, 3Department of Diagnostic and Interventional Radiology, University of Freiburg, Freiburg, Germany, 4Max Planck Institute for Intelligent Systems, Tuebingen, Germany |
||
Keywords: Machine Learning/Artificial Intelligence, Motion Correction, Motion Detection, Self-Supervised Learning Motion is still one of the major extrinsic sources for imaging artifacts in MRI that can strongly deteriorate image quality. Any impairment by motion artifacts can reduce the reliability and precision of the diagnosis and a motion‐free reacquisition can become time‐ and cost‐intensive. Furthermore, in large-scale epidemiological cohorts, manual quality screening becomes impracticable. An automated quality assessment is thus of interest. Reliable motion estimation in varying domains (imaging sequences, multiple scanners, sites) is however challenging. In this work, we propose an attention-based transformer that can detect motion in various MR imaging scenarios. |
| 1819 | Computer 2
|
Effectiveness and Reliability Assessment on Head Motion Capturing and Correction (MoCAP) |
| Zhuoyang Gu1,2, Lianghu Guo1, Qing Yang1, Xinyi Cai1, Tianli Tao1, Sifan He1, Hua Jiang1, Haifeng Tang1, Qian Wang1, Xiaopeng Zong1, Dinggang Shen1, Qiang He2, and Han Zhang1 | ||
1School of Biomedical Engineering, ShanghaiTech University, Shanghai, China, 2United Imaging Healthcare Co., Ltd., Shanghai, China |
||
Keywords: Motion Correction, Brain, MRI acquisition Head motion monitoring and compensation during MRI is essential to imaging quality and success rates of acquisition. MoCAP is a novel real-time head motion monitoring and correction technique utilizing structured light to perform prospective motion correction by adjusting MR gradient. We perform effectiveness and reliability assessment on MoCAP in structural and functional MRI. MoCAP can significantly improve quality of the two modalities, reducing motion artifacts and enhancing validity and reliability of post-processing results. MoCAP is promising in the MRI field for special populations like children and patients with difficulty in keeping still during scan. |
| 1820 | Computer 3
|
Prospective Motion Correction for 3D-EPI fMRI using Orbital Navigators and Linear Regression |
| Thomas Ulrich1, Malte Riedel1, and Klaas Prüssmann1 | ||
1Institute for Biomedical Engineering, ETH Zurich and University of Zurich, Zürich, Switzerland |
||
Keywords: Motion Correction, fMRI (resting state) We propose to apply a navigator-based prospective motion correction method to 3D-EPI fMRI scans. The method is investigated during in-vivo volunteer experiments with and without intentional motion. Post-processing with SPM12 shows that head motion is largely compensated by the navigator method. We show that the technique is sensitive to breathing motion and cardioballistic motion by comparing to physiological measurements from a breathing belt and pulse oximeter, and we show that it strongly reduces inter-volume displacements. |
| 1821 | Computer 4
|
Effective removal of the residual head motion artifact after motion correction in fMRI data. |
| Wanyong Shin1 and Mark J Lowe1 | ||
1Radiology, Cleveland Clinic, Cleveland, OH, United States |
||
Keywords: Motion Correction, Brain, Head motion, fMRI We investigate the source of residual motion artifact after volumetric motion correction using a custom MR sequence acquisition with prospectively injected motion (SIMPACE). We injected various patterns of motion during scanning ex-vivo brain phantoms at 3T to synthesize head motion in an fMRI dataset. We propose voxelwise retrospective motion regressors in addition to 6 rigid body motion parameters, then compare it with the “standard” 6 motion parameters and their derivatives regressor models. The proposed model improves tSNR by 40% and 11% in linear drift of motion and the realistic motion case, respectively, compared to 6 motion parameter model. |
| 1822 | Computer 5
|
Combined motion and B0 correction of susceptibility weighted imaging with jointly acquired FID and spherical navigators |
| Miriam Hewlett1,2, Junmin Liu2, and Maria Drangova1,2 | ||
1Medical Biophysics, Western University, London, ON, Canada, 2Robarts Research Institute, London, ON, Canada |
||
Keywords: Motion Correction, Motion Correction Jointly acquired FID and spherical navigators were applied for prospective correction of motion and retrospective correction of zeroth order field offsets in a susceptibility weighted imaging protocol. Initial results showed a significant reduction in motion artifacts with prospective motion correction, quantified using structural similarity index. Retrospective B0 correction provided an additional improvement in image quality (also statistically significant). Future work will investigate methods to reduce residual artifacts, as well as incorporating the navigator within deadtime in the sequence to reduce scan time. |
| 1823 | Computer 6
|
Velocity-Encoding Navigator for First-Order Motion Compensation in Diffusion MRI |
| Bo Li1 and Thomas Ernst1 | ||
1Department of Diagnostic Radiology and Nuclear Medicine, University of Maryland, Baltimore, Baltimore, MD, United States |
||
Keywords: Motion Correction, Diffusion/other diffusion imaging techniques, velocity navigator Uncorrected head rotations during diffusion weighted imaging (DWI) can induce gradient imbalances that may shift the echo outside of the k-space window and cause signal dropouts. Therefore, we developed a rotational velocity navigator (~10ms duration) that is acquired immediately after each excitation and estimates the rotational velocities perpendicular to the diffusion gradient. The accuracy of estimated velocities was 4.1°/s relative to an optical tracking system (“gold standard”). Ultimately, the rotational velocities can predict the gradient moment error in first order, and a gradient blip can be applied to recenter the echo in k-space. |
| 1824 | Computer 7
|
Fast, model-based, and navigator-free retrospective motion correction for non-Cartesian fMRI |
| Guanhua Wang1, Shouchang Guo2, Jeffrey A. Fessler2, and Douglas C. Noll1 | ||
1Biomedical Engineering, University of Michigan, Ann Arbor, MI, United States, 2EECS, University of Michigan, Ann Arbor, MI, United States |
||
Keywords: Motion Correction, Motion Correction This abstract presents a retrospective motion correction method for fMRI. The method alternates between motion estimation and motion-informed model-based reconstruction. Compared to registration-based correction, this approach resolves intra-frame, inter-shot motion without additional navigators. The open-source GPU-based implementation enables efficient correction/reconstruction for large-scale non-Cartesian fMRI data. With prospective experiments, we demonstrate that our approach outperformed retrospective registration by providing higher-resolution images with reduced false positives in activation maps. |
| 1825 | Computer 8
|
Inline Retrospective Motion Correction and Dynamical Alignment Using an Optical Markerless Motion Tracker |
| Ulrich Lindberg1, Stefan Glimberg2, Gerard Crelier3, Martin Buehrer3, and Henrik Bo Wiberg Larsson1 | ||
1Functional Imaging Unit, Department of Clinical Physiology and Nuclear Medicine, Rigshospitalet, Copenhagen, Denmark, 2TracInnovations, Ballerup, Denmark, 3GyroTools, Zurich, Switzerland |
||
Keywords: Image Reconstruction, Motion Correction Using an external motion tracking system for recording head movement during MR-acquisition and a modified version of the Philips reconstruction software (Recon2.0) that enables phase re-alignment and re-gridding of acquisition profiles during the standard immediate reconstruction process, we present a fully integrated retrospective motion correction solution that deliver both corrected and non-corrected images to the user and allows for the immediate visual assessment of the correction quality upon acquisition. |
| 1826 | Computer 9
|
Severe MR Motion Artefact Correction with 2 step Deep Learning-based guidance |
| Julian Hossbach1,2, Daniel Nicolas Splitthoff2, Bryan Clifford3, Daniel Polak4,5, Stephan Cauley5, and Andreas Maier1 | ||
1Pattern Recognition Lab, Friedrich-Alexander-University Erlangen-Nuremberg, Erlangen, Germany, 2Siemens Healthcare GmbH, Erlangen, Germany, 3Siemens Medical Solutions, Boston, MA, United States, 4Siemens Healthcare Gmbh, Erlangen, Germany, 5Department of Radiology, Athinoula A. Martinos Center for Biomedical Imaging, Boston, MA, United States |
||
Keywords: Machine Learning/Artificial Intelligence, Motion Correction Motion artifacts can pose a difficult challenge in the clinical workflow. For addressing this issue, we here investigate the performance of two Deep Learning based motion mitigation strategies, MoPED and NAMER, and demonstrate that both approaches can readily be combined. This allows for the correction of severely corrupted images. |
| 1827 | Computer 10
|
Self-navigated free-breathing ZTE lung imaging |
| Jose de Arcos1, Ana Beatriz Solana2, Jonathan Weir-McCall3,4, Emil Ljungberg5,6, Joshua D Kaggie3, and Florian Wiesinger2 | ||
1GE HealthCare, Little Chalfont, Amersham, United Kingdom, 2ASL Europe, GE HealthCare, Munich, Germany, 3Department of Radiology, University of Cambridge, Cambridge, United Kingdom, 4University of Cambridge School of Clinical Medicine, Royal Papworth Hospital, Cambridge, United Kingdom, 5Medical Radiation Physics, Lund University, Lund, Sweden, 6Neuroimaging, Institute of Psychiatry, Psychology, and Neuroscience, King's College London, London, United Kingdom |
||
Keywords: Motion Correction, Motion Correction, Self-navigation Here we propose a self-navigated technique, based on the extraction of respiratory motion estimates using interleaved spiral phyllotaxis trajectories in 3D radial zero echo time (ZTE) acquisitions. These are particularly well-suited to capture the short T2* signal, characteristic of lung parenchyma, and have high acquisition efficiency allowing to generate fast temporal resolution navigators. The self-navigation technique worked robustly for different respiratory patterns and on 1.5T and 3.0T fields strengths, obtaining high quality images comparable to the ones obtained with bellows-gating. |
| 1828 | Computer 11
|
Optical camera calibration to implement various marker-based motion correction techniques in open geometry 0.5 T upright scanner |
| Laura Bortolotti1, Olivier Mougin1, Paul Glover1, Richard Bowtell1, and Penny Gowland1 | ||
1SPMIC, University of Nottingham, Nottingham, United Kingdom |
||
Keywords: Motion Correction, Motion Correction, Marker-based, calibration Open MRI scanner improves patient comfort but allows movement during scanning. Hence, implementation of MoCo techniques is crucial to maximize image quality. Here, we established a method to use a standard commercial optical motion tracking set-up in an Open MRI scanner. The optical system has been calibrated to provide tracking measurements in the image reference frame. Then, a simple phantom movement was tested and corrected using NUFFT algorithm in BART. |
| 1829 | Computer 12
|
Pilot Tone-navigated motion estimation for liver DW-MRI |
| Cemre Ariyurek1, Serge Vasylechko1, Xiaodong Zhong2, Vibhas Deshpande3, Michael Bush4, Stephan Voss1, Onur Afacan1, and Sila Kurugol1 | ||
1Radiology, Boston Children's Hospital and Harvard Medical School, Boston, MA, United States, 2MR R&D Collaborations, Siemens Medical Solutions USA, Inc., Los Angeles, CA, United States, 3MR R&D Collaborations, Siemens Medical Solutions USA, Inc., Austin, TX, United States, 4MR R&D Collaborations, Siemens Medical Solutions USA, Inc., New York, NY, United States |
||
Keywords: Motion Correction, Diffusion/other diffusion imaging techniques Diffusion-weighted MRI (DW-MRI) is capable of detecting and characterizing liver tumors and following-up treatments. Unfortunately, respiratory motion during DW-MRI scan causes misalignments between slices and reduces image quality. 3D slice-to-volume registration (SVR) can be employed to correct for motion. However, motion estimates may be inaccurate for high b-value images where SNR decreases. In this work, we propose to use PT estimated motion correction, which is calibrated on 3D SVR motion parameters obtained from low b-value images. We showed misalignments between slices are reduced by the proposed PT-based motion correction compared to SVR-based motion correction and no correction. |
| 1830 | Computer 13
|
Multislice-to-volume Prospective Motion Correction for Functional MRI Protocols at 7T |
| Steven Winata1, Daniel Christopher Hoinkiss2, Graeme Alexander Keith1, Salim Mohammed al-Wasity1, and David Andrew Porter1 | ||
1Imaging Centre of Excellence, University of Glasgow, Glasgow, Scotland, 2Fraunhofer Institute for Digital Medicine MEVIS, Bremen, Germany |
||
Keywords: Motion Correction, fMRI, Prospective motion correction; Markerless; Real-time 7 Tesla MRI provides higher signal-to-noise ratio (SNR) and spatial resolution, but is also more sensitive to motion-induced artefacts. The restricted environment within 7T scanners make non-hardware options for motion correction attractive. In this abstract, we investigate the use of the markerless, image-based, real-time Multislice Prospective Acquisition Correction (MS-PACE) technique for functional MRI protocols at 7T. This multislice scheme allows for a sub-TR motion detection and correction. It is demonstrated that the technique is able to correct longer-term motion components occurring during the acquisition. |
| 1831 | Computer 14
|
Motion compensated multi-contrast MRI using deep factor model |
| Yan Chen1, James H. Holmes1, Curtis A. Corum2, Vincent Magnotta1, and Mathews Jacob1 | ||
1University of Iowa, Iowa City, IA, United States, 2Champaign Imaging, LLC, Minneapolis, MN, United States |
||
Keywords: Motion Correction, Multi-Contrast Recent quantitative parameter mapping methods including MR fingerprinting collect a time series of images that capture the evolution of magnetization. The focus of this work is to introduce a novel approach termed as deep factor model, which offers an efficient representation of the multi-contrast image time series. The higher efficiency of the representation enables the acquisition of the images in a highly undersampled fashion, which translates to reduced scan time in 3D high-resolution multi-contrast applications. The approach integrates motion estimation and compensation, making the approach robust to subject motion during the scan. |
| 1832 | Computer 15
|
A Dynamic 2D TSE Acquisition Strategy for Robust SAMER Motion Mitigation |
| Daniel Polak1, Daniel Nicolas Splitthoff1, Bryan Clifford2, Wei-Ching Lo2, Yantu Huang3, Susie Huang4, John Conklin4, Lawrence L. Wald5, and Stephen F. Cauley5 | ||
1Siemens Healthcare GmbH, Erlangen, Germany, 2Siemens Medical Solutions, Boston, MA, United States, 3Shenzhen Magnetic Resonance, Shenzhen, China, 4Massachusetts General Hospital, Boston, MA, United States, 5A. A. Martinos Center for Biomedical Imaging, Charlestown, MA, United States |
||
Keywords: Motion Correction, Brain, Value, Clinical Application Retrospective motion correction for 2D TSE/FSE is challenging due to interpolation through slices with gaps, interleaved slice orderings, and spin history effects. Optimized Cartesian sampling trajectories provide decreased motion sensitivity in specific situations but can also exacerbate motion sensitivity under typical patient motion. In this work, we introduce a dynamic acquisition strategy that determines the k-space lines acquired in the next shot based on the prior patient motion. Specifically, each TR dynamically encodes available lines which minimize motion variance. We show that this dynamic acquisition strategy results in improved reconstruction robustness under typical clinical motion scenarios. |
| 1833 | Computer 16
|
Patient-specific respiratory liver motion analysis for individual motion-resolved reconstruction |
| Veronika J Spieker1,2, Jonathan K Stelter3, Veronika A Zimmer2, Kilian Weiss4, Rickmer F Braren3, Dimitrios Karampinos3, and Julia A Schnabel1,2,5 | ||
1Helmholtz Center Munich, Munich, Germany, 2School of Computation, Information and Technology, Technical University of Munich, Munich, Germany, 3Department of Diagnostic and Interventional Radiology, Technical University of Munich, Munich, Germany, 4Philips GmbH Market DACH, Hamburg, Germany, 5School of Biomedical Imaging and Imaging Sciences, King's College, London, United Kingdom |
||
Keywords: Motion Correction, Liver Respiratory motion is a major source of artefacts in abdominal MRI and varies considerably across subjects. Motion-resolved strategies utilize the periodic nature of respiratory motion, but do not always consider absolute patient-specific motion information. The purpose of the present work is to develop a framework incorporating the patient-specific absolute body motion into a motion-resolved reconstruction (XD-GRASP), while adapting different motion binning strategies. To explore the potential of respiratory motion knowledge for MR reconstruction, individual absolute body motion, i.e. of the liver, is identified. |
| 1834 | Computer 17
|
Combining FID navigators with field probe monitoring for improved head motion tracking and prospective correction at 7T |
| Matthias Serger1, Ruediger Stirnberg1, Philipp Ehses1, and Tony Stoecker1,2 | ||
1German Center for Neurodegenerative Diseases (DZNE), Bonn, Germany, 2Department of Physics and Astronomy, University of Bonn, Bonn, Germany |
||
Keywords: Motion Correction, Motion Correction FID navigators and field probes can measure local motion-induced magnetic field changes at different positions. Both methods are suited for marker-less prospective motion correction but with limited accuracy. A small 7T study with five subjects was performed to compare their performance for small involuntary and large voluntary motion. It is demonstrated that errors of regression models, calibrated between magnetic field measurements and motion parameters, are consistently reduced by combining both techniques. Furthermore, a proof-of-concept prospective motion correction experiment based on FID navigators is presented, which will be extended to include field probe measurements in the near future. |
| 1835 | Computer 18
|
Multi-Echo MRI Motion Artifact Reduction via Knowledge Interaction Learning for Better SWI Enhancement |
| Mohammed A. Al-masni1, Seul Lee2, Sewook Kim2, Sung-Min Gho3, Young Hun Choi4, and Dong-Hyun Kim2 | ||
1Department of Artificial Intelligence, Sejong University, Seoul, Korea, Republic of, 2Department of Electrical and Electronic Engineering, Yonsei University, Seoul, Korea, Republic of, 3GE Healthcare, Korea, Seoul, Korea, Republic of, 4Department of Radiology, Seoul National University Hospital, Seoul, Korea, Republic of |
||
Keywords: Motion Correction, Artifacts Patient movement during MRI scan can cause severe degradation of image quality. In Susceptibility-Weighted Imaging (SWI), several echoes are measured during a single repetition period, where the earliest echoes show less contrast between various tissues, while the higher echoes are more susceptible to artifacts and signal dropout. This paper proposes a data-driven retrospective deep learning method by taking the advantage of interactively learning multiple echoes together through sharing their knowledge using unified training parameters. The proposed method allows to share information and gain an understanding of the correlations between multiple echoes towards generating high-resolution susceptibility enhanced contrast images. |
| 1836 | Computer 19
|
Highly Accelerated 3D MPRAGE using Robust SAMER Motion Mitigation |
| Daniel Nicolas Splitthoff1, Stephen Cauley2, Tobias Kober3, Julian Hossbach1, Bryan Clifford4, Wei-Ching Lo4, Yan Tu Huang5, Susie Y. Huang6, John Conklin6, Lawrence L. Wald2, and Daniel Polak1 | ||
1Siemens Healthcare GmbH, Erlangen, Germany, 2A. A. Martinos Center for Biomedical Imaging, Charlestown, MA, United States, 3Siemens Healthineers International AG, Lausanne, Switzerland, 4Siemens Medical Solutions, Boston, MA, United States, 5Shenzhen Magnetic Resonance Ltd.,, Shenzhen, China, 6Massachusetts General Hospital, Charlestown, MA, United States |
||
Keywords: Motion Correction, Data Acquisition, Fast imaging, Compressed Sensing Long clinical acquisitions can face the challenge of motion during the scan. One approach to address this problem is the recently introduced SAMER retrospective motion correction technique for 3D MPRAGE. As an alternative, it has been suggested to shorten the scan time by acceleration techniques, e.g., using Compressed Sensing, and thus reduce the likelihood of patient discomfort and severe motion. We here combine SAMER with Compressed Sensing for high acceleration factors (R=6). |
| 1837 | Computer 20
|
Evaluation of body motion at various patient position in a 0.5 T Upright scanner |
| Laura Bortolotti1, Isabel Clennell 1, Amy Bradbury1, Olivier Mougin1, Paul Glover1, Richard Bowtell1, and Penny Gowland1 | ||
1SPMIC, University of Nottingham, Nottingham, United Kingdom |
||
Keywords: Motion Correction, Motion Correction, Marker based tracking Body motion at nine typical upright MRI poses (standing, seated, supine), have been characterised for several body positions (head, shoulder, sternum, hip) for 20 subjects over 30 s (free breathing); and 20 s (breath hold) and for 9 subjects over 10 minutes. Free-standing caused most motion; lying supine caused least. The use of a body support and breath-holding reduced motion. Net motion was similar over 50-1000 ms sampling periods. Respiratory-related motion could be removed from tracking data. The degree of motion identified in this work will inform future respiratory triggering and motion correction work. |
| 1997 | Computer 1
|
Changes in the position of the eyeball in TAO patients with unilateral upper eyelid retraction by T2-weighted SPIR imaging |
| Xinyi Gou1, Yi Wang2, Lingli Zhou3, Jianxiu Lian4, Zilong Chen4, Xiuying Zhang3, Jingyi Cheng1, Lei Chen1, Nan Hong1, and Jin Cheng1 | ||
1Department of Radiology, Peking University People’s Hospital, Beijing, China, 2Department of Ophthalmology, Peking University Third Hospital, Beijing, China, 3Department of Endocrinology, Peking University People’s Hospital, Beijing, China, 4Philips Healthcare, Beijing, China |
||
Keywords: Image Reconstruction, Software Tools For patients with thyroid-associated orbitopathy (TAO) presenting unilateral upper eyelid retraction, the phenomenon that the impaired eye is lower than the healthy eye (“eyeball descending”) has been noticed. This study investigated the eyeballs' position changes by 3D reconstitution of magnetic resonance (MR) images. With reference to the central plane of the healthy side, 70.37% impaired eyeballs (19/27) were found in significantly lower positions, and with significant positive correlations of increased thicknesses of levator palpebrae superioris (LPS), superior rectus (SR), and LPS-SR complex volume. This study provided objective evidence of “eyeball descending” in unilateral upper eyelid retraction TAO patients. |
| 1998 | Computer 2
|
A combined model-based and ML-based approach performs better than model-based or ML-based applied individually |
| Srikant Kamesh Iyer1, Hassan Haji-Valizadeh1, and Samir Sharma1 | ||
1Canon Medical Research USA, Inc., Mayfield, OH, United States |
||
Keywords: Motion Correction, Motion Correction A motion correction framework was developed to suppress artifacts from rigid and non-rigid motion using a combination of model-based and ML-based approaches. The performance of this framework was compared with model-based only and ML-based only motion correction approaches on motion-simulated data and motion-corrupted in-vivo data using visual inspection and image quality metrics. The combined approach showed superior image quality than model-based and ML-based approaches applied individually. |
| 1999 | Computer 3
|
Optical position tracking fiducial marker for rigid body motion parameter estimation with high performance |
| Marina Silic1,2, Aravinthan Jegatheesan1,2, Fred Tam2, and Simon J Graham1,2 | ||
1Department of Medical Biophysics, University of Toronto, Toronto, ON, Canada, 2Sunnybrook Research Institute, Toronto, ON, Canada |
||
Keywords: Data Acquisition, Motion Correction, Optical Position Tracking Optical position tracking (OPT) using fiducial markers is advantageous for data acquisition of rigid body head motion parameters and motion correction in magnetic resonance imaging (MRI). Many opportunities still remain to improve OPT through the development of new devices. A promising prototype OPT marker and analysis pathway are described. The marker through-plane degree of freedom (DOF) precision was enhanced via moiré patterns and stereovision. Precision was assessed using positional and rotational stages in all 6DOF. Initial results strongly suggest that with minimal additional work, the OPT marker will provide excellent performance in a 3 T MRI system. |
| 2000 | Computer 4
|
The noise suppression of resting fMRI data based on eigenvector correction |
| Wei Zhao1, Huanjie Li 1, Yunge Zhang1, Dongyue Zhou1, and Fengyu Cong1,2 | ||
1Dalian University of Technology, Dalian, China, 2University of Jyvaskyla, Jyvaskyla, Finland |
||
Keywords: Motion Correction, Artifacts The fMRI signal has been very noisy for artifacts induced various reasons, and yet the head motion and non-neuronal contributions are always the most tickle one. And too severe of motion corruption usually will lead to abandonment of the participant’s data. In this study, proposed method manages to effectively control these noises meanwhile without losing valuable signals. Proposed method exceeds standard pipeline in both quantitative and qualitative metrics. |
| 2001 | Computer 5
|
The development of Inter-/intra-volume motion correction algorithm for fMRI using a custom MRI acquisition with prospectively injected motion. |
| Wanyong Shin1 and Mark J Lowe1 | ||
1Radiology, Cleveland Clinic, Cleveland, OH, United States |
||
Keywords: Motion Correction, fMRI In this study, we propose the new inter-/intra-volume motion correction algorithm. To test the proposed method, we modify simulated Prospective Acquisition CorrEcted (SIMPACE) EPI sequence and inject both inter- and intra-volume motion during actual EPI acquisition. Using ex-vivo brain phantom, we synthesize inter-/intra volume motion corrupted MR data. We evaluate the proposed method, compared to volume motion correction method. We find that the proposed inter-/intra-volume method outperforms volume motion correction method. |
| 2002 | Computer 6
|
Linear SAMER: linear motion optimization for scout accelerated retrospective motion estimation and reduction with linear+ reordering in MPRAGE |
| Yantu Huang1, Daniel Nicolas Splitthoff2, Bryan Clifford3, Daniel Polak2,4, Stephen Cauley4,5, Wei-Ching Lo3, Nan Xiao1, and Huixin Tan1 | ||
1Siemens Shenzhen Magnetic Resonance Ltd., Shenzhen, China, 2Siemens Healthcare GmbH, Erlangen, Germany, 3Siemens Medical Solutions, Boston, MA, United States, 4Massachusetts General Hospital, Charlestown, MA, United States, 5Harvard Medical School, Boston, MA, United States |
||
Keywords: Motion Correction, Head & Neck/ENT SAMER uses a fast reference scan (“scout”) and short additional guidance lines in each shot of an MPRAGE sequence (“linear+ reordering”) to calculate motion parameters. Optimization of rigid body motion parameters are typically nonlinear. In this work we exploit local linearity with SAMER to improve the performance. During motion estimation, the initial guess of motion parameters for a given shot is taken from the previous most similar shot. Similarity is determined by correlations of the linear+ guidance lines. Retrospective reconstructions of volunteer data show the proposed method has very good computation performance and can handle large motion well. |
| 2003 | Computer 7
|
Detection and remediation of ghost artifacts in low b-value diffusion MRI with ghost re-synthetization and 4-way phase-encoded data |
| Anh S Thai1,2, Carlo Pierpaoli2, Lin-Ching Chang1, and M. Okan Irfanoglu2 | ||
1Catholic University of America, Washington, DC, United States, 2QMI, NIBIB/National Institutes of Health, Bethesda, MD, United States |
||
Keywords: Artifacts, Artifacts, diffusion mri We propose a novel post-processing method for ghost artifact correction in low-b diffusion weighted images (DWIs) using ghost synthetization and four different phase-encoding direction (PED) images. The results from both simulations and real data show that our method can robustly detect and correct ghost artifacts on b=0 s/mm2 images for artifact-free image. The proposed method can be used further to generate ground truth training images for a machine-leaning method to remedy ghost artifacts in single PED acquisition. |
| 2004 | Computer 8
|
Phase-Constrained Reconstruction for Enhancing PROPELLER SNR |
| Yuchou Chang1, Gulfam Ahmed Saju1, Jasina Yu1, Reza Abiri2, Tianming Liu3, and Zhiqiang Li4 | ||
1Computer and Information Science, University of Massachusetts Dartmouth, North Dartmouth, MA, United States, 2Electrical, Computer and Biomedical Engineering, University of Rhode Island, Kingston, RI, United States, 3Computer Science, University of Georgia, Athens, GA, United States, 4Neuroradiology, Barrow Neurological Institute, Phoenix, AZ, United States |
||
Keywords: Motion Correction, Parallel Imaging PROPELLER blade acquisition has been accelerated by undersampling blade k-space data. Missing data on each blade can be reconstructed by parallel MRI reconstruction methods. However, noise deteriorates the blade and the overall image quality. To enhance the signal-to-noise ratio (SNR), phase-constrained reconstruction is studied for improving the SNR of PROPELLER imaging. Phase-constrained reconstruction can improve PROPELLER imaging SNR when the acceleration of data acquisition is used. Optimal selection of the ACS lines and the outer reduction factors is expected to achieve a better SNR and accelerate the PROPELLER imaging speed simultaneously. |
| 2005 | Computer 9
|
Fast Prospective Motion Correction using Directional Couplers as Motion Sensors |
| Jason Daniel van Schoor1,2, Mark Gosselink3, Dennis Klomp3, Giel Mens3, Hans Hoogduin3, Wim Prins4, and Tijl van der Velden3 | ||
1High Field MRI, UMC Utrecht, Utrecht, Netherlands, 2Utrecht University, Utrecht, Netherlands, 3UMC Utrecht, Utrecht, Netherlands, 4Philips, Best, Netherlands |
||
Keywords: Motion Correction, Motion Correction, Directional coupler Motion is a prevalent issue in MRI. Here, we present a fast prospective motion correction (PMC) algorithm using directional couplers as head-motion sensors. The mean reflectances (per volume) of an 8 channel pTx head coil are modelled to head-position recorded by the systems built-in PMC during a calibration scan. Thereafter, the model is used for motion correction at a per RF pulse temporal resolution. Results indicate partial motion correction with error in correcting fast motions due to the necessary presence of filters to remove unidentified low-frequency noise. Mitigating low-frequency noise contributions would allow for less filtering thereby improving dynamic performance. |
| 2006 | Computer 10
|
Spectrally-encoded multi-spectral imaging (SEMSI) at 0.55T provides improved imaging adjacent to metallic implants |
| Bochao Li1, Kübra Keskin2, Daehyun Yoon3, Nam Gyun Lee1, Brian Hargreaves3, and Krishna Shrinivas Nayak2 | ||
1Department of Biomedical Engineering, University of Southern California, Los Angeles, CA, United States, 2Department of Electrical and Computer Engineering, University of Southern California, Los Angeles, CA, United States, 3Stanford University, Stanford, CA, United States |
||
Keywords: Artifacts, Low-Field MRI, Metal Spectrally-encoded multi-spectral imaging (SEMSI) enables imaging adjacent to metal implants without blurring due to off-resonance, and has been demonstrated at 3T. Here, we demonstrate application of this technique at 0.55T with appropriate adjustments to the spectral resolution and RF pulse bandwidth. A total hip arthroplasty phantom was imaged with SEMSI, turbo spin echo (TSE), and Slice Encoding for Metal Artifact Correction (SEMAC). Experiments demonstrate the feasibility of SEMSI at 0.55T, and examine ΔTE to support multi-echo acquisitions in a single TR. |
| 2007 | Computer 11
|
Iterative Motion-Compensated Reconstruction with Convolutional Neural Network (iMoCo-Net) for Ultrashort Echo Time (UTE) Proton Lung MRI |
| Fei Tan1, Ke Wang2, Michael Lustig1,2, and Peder E. Z. Larson1,3 | ||
1UC Berkeley-UCSF Graduate Program in Bioengineering, University of California Berkeley and University of California, San Francisco, San Francisco, CA, United States, 2Electrical Engineering and Computer Sciences, University of California, Berkeley, Berkeley, CA, United States, 3Department of Radiology and Biomedical Imaging, University of California, San Francisco, San Francisco, CA, United States |
||
Keywords: Motion Correction, Motion Correction This abstract explores the feasibility of machine learning-based motion-compensated reconstruction for free-breathing UTE lung MRI. Specifically, we used respiratory motion-resolved Non-uniform Fourier Transform (NuFFT) reconstruction as input, iterative motion-compensated (iMoCo) reconstruction as target, and a 2D U-Net convolutional neural network. Test results demonstrate a sharper diaphragm and a higher apparent SNR compared to the averaged input. In conclusion, iMoCo-Net accelerates the reconstruction of 3D radial UTE data substantially, shortening the required time from hours to minutes. |
| 2008 | Computer 12
|
In-plane B0 inhomogeneity effects in T2* mapping |
| Glen Morrell1 | ||
1Radiology and Imaging Sciences, University of Utah, Salt Lake City, UT, United States |
||
Keywords: Artifacts, Relaxometry Main field B0 inhomogeneity may adversely affect the accuracy of T2* mapping in the abdomen even at locations spatially distant from the B0 variation. We show this effect in both simulation and phantom studies, and show that it can be ameliorated with the use of two-dimensional spatially selective excitation. These findings have important ramifications for applications of BOLD imaging in the abdomen such as renal BOLD or liver T2* mapping, where large B0 inhomogeneity is often present in areas of the image volume spatially remote from the organ of interest. |
| 2009 | Computer 13
|
SPAMM Tagged EPI Acquisition for Assessment of Geometric Distortion Caused by an Endorectal Coil |
| Ken-Pin Hwang1, R. Jason Stafford1, and Tharakeswara K. Bathala2 | ||
1Department of Imaging Physics, The University of Texas M.D. Anderson Cancer Center, Houston, TX, United States, 2Department of Abdominal Imaging, The University of Texas M.D. Anderson Cancer Center, Houston, TX, United States |
||
Keywords: Artifacts, System Imperfections: Measurement & Correction Diffusion weighted imaging is an essential sequence for diagnosis of prostate cancer but is often impacted by distortion. Current existing phantoms designed for assessing geometric accuracy do not sufficiently simulate the environmental factors that affect distortion in diffusion weighted EPI. In this work we apply a rectilinear tagging technique to visualize and understand patterns of distortion caused by an endorectal coil immersed in water when imaged with a PI-RADS compliant protocol. Maximum spatial distortion including regions of spatial collapse was observed near the coil element. Observed distortion pattern supports proper centering of the coil to avoid distortion. |
| 2010 | Computer 14
|
Reduction of motion artifacts in diffusion weighted brain images using first order motion-compensated diffusion gradients |
| Danielle Kara1, Mark J. Lowe2, Christopher T Nguyen1, and Ken E. Sakaie2 | ||
1Cardiovascular Innovation Research Center, Cleveland Clinic, Cleveland, OH, United States, 2Imaging Institute, Cleveland Clinic, Cleveland, OH, United States |
||
Keywords: Motion Correction, Diffusion/other diffusion imaging techniques, Diffusion Tensor Imaging, Motion Correction, Neuro Patient motion during acquisition of diffusion weighted (DW) images causes significant signal dropout. To perform motion correction, DW images of the brain were acquired during gross head motion using a first order motion-compensated diffusion gradient scheme and compared to a traditional DW-EPI sequence. Of the images acquired with the motion-compensated sequence, only 1.5% contained significant signal dropout compared to 36% of the images acquired with the traditional sequence. Overall, the motion-compensated sequence reduced signal dropout by 96% compared to the traditional sequence without increasing scan time or introducing new post-processing protocols. |
| 2011 | Computer 15
|
Respiratory binning with PilotTone Navigator For Motion Compensated Liver DW-MRI |
| SERGE DIDENKO VASYLECHKO1, CEMRE ARIYUREK2, ONUR AFACAN2, and SILA KURUGOL2 | ||
1HARVARD MEDICAL SCHOOL, BOSTON, MA, United States, 2RADIOLOGY, BOSTON CHILDREN'S HOSPITAL, HARVARD MEDICAL SCHOOL, BOSTON, MA, United States |
||
Keywords: Motion Correction, Body Respiratory motion substantially affects the accuracy of quantitative DWI-MR techniques in the upper abdomen. A conventional approach for motion correction in abdominal MRI uses a respiratory belt or other navigators for prospective triggering. This study explores use of a new approach - PilotTone (PT) navigator for binning. |
| 2012 | Computer 16
|
Artifacts Map-guided Nonlocal Mean denoising for abdominal 4D-MR images acquired by 3D view-sharing sequence |
| Yat Lam Wong1, Hing Chiu Charles Chang2, Weiwei Liu3, Weihu Wang3, Yibao Zhang3, Hao Wu3, Victor Ho Fun LEE4, Lai Yin Andy Cheung5, Tian Li1, and Jing Cai1 | ||
1Department of Health Technology and Informatics, The Hong Kong Polytechnic University, Hong Kong, Hong Kong, 2Department of Biomedical Engineering, The Chinese University of Hong Kong, Hong Kong, Hong Kong, 3Peking University Cancer Hospital & Institute, Beijing, China, 4Department of Clinical Oncology, The University of Hong Kong, Hong Kong, Hong Kong, 5Department of Clinical Oncology, Queen Mary Hospital, Hong Kong, Hong Kong |
||
Keywords: Data Processing, Data Processing, 4D-MRI 3D view-sharing sequences have been demonstrated its great promise for abdominal 4D-MRI due to its high temporal resolution and availability in clinical scanners. However, the sub-optimal sampling methods in these sequences under free-breathing condition deteriorates the image quality by ghost artifacts. Here, we propose a novel technique, Artifacts Map-guided Nonlocal Mean (AM-NLM), to suppress motion artifacts and to increase image quality of the 4D-MR images of liver cancer patients acquired by a 3D view-sharing sequence, TRICKS, on a 1.5T scanner. |
| 2013 | Computer 17
|
Motion robust multi-shot EPI of the abdomen using VFA-FLEET |
| Mustafa Utkur1,2, Tess E Wallace1,2, Sila Kurugol1,2, and Onur Afacan1,2 | ||
1Radiology, Harvard Medical School, Boston, MA, United States, 2Computational Radiology Laboratory, Boston Children's Hospital, Boston, MA, United States |
||
Keywords: Motion Correction, Body Multi-shot EPI can be a promising technique to improve the resolution limits of the single shot EPI. However, its application has been so far limited in the abdominal MRI, as phase errors between shots due to unavoidable breathing motion creates artifacts in the images. Here, we tested a Variable-Flip-Angle Fast Low-angle Excitation Echo planar Technique (VFA-FLEET) pulse sequence and showed that it can be used to create high-resolution motion-robust images with high geometric fidelity without breath holding. |
| 2014 | Computer 18
|
Pilot Tone respiratory signal processing with RF interference suppression and validation against image navigator |
| Yantu Huang1, Huixin Tan1, Qiuyi Zhang1, and Peter Speier2 | ||
1Siemens Shenzhen Magnetic Resonance Ltd., Shenzhen, China, 2Siemens Healthcare GmbH, Erlangen, Germany |
||
Keywords: Data Processing, Data Processing, physiology signal Pilot Tone (PT) has been demonstrated to extract respiratory signal successfully. However, RF interferences are not considered. We propose a method to create a high-quality PT respiratory signal, while suppressing RF interference. At beginning of a measurement, initial RF suppression matrix and respiratory combination vector are learned. Then a PT respiratory signal can be generated to trigger measurement. Afterwards, RF suppression matrix is updated during each RF train. We validate our method in a moving phantom and in 39 volunteers against image navigator. Results show that our method can suppress RF interference effectively- highly correlated with image navigator. |
| 2015 | Computer 19
|
Consolidation of expert ratings of motion artifacts using hierarchical label fusion |
| Yael Balbastre1,2, Robert Frost1,2, Khushi Morparia1,3, Richard L Carrington III1, Yuh-Shin Chang1,4,5, Brooks P Applewhite1,4, Marcio Aloisio Bezerra Cavalcanti Rockenbach6, and Bruce Fischl1,2,7,8 | ||
1Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Boston, MA, United States, 2Department of Radiology, Harvard Medical School, Boston, MA, United States, 3Northeastern University, Boston, MA, United States, 4Department of Radiology, Massachusetts General Hospital, Boston, MA, United States, 5Department of Radiology, Massachusetts Eye and Ear Infirmary, Boston, MA, United States, 6Data Science Office, Mass General Brigham, Somerville, MA, United States, 7Computer Science and Artificial Intelligence Laboratory, Massachusetts Institute of Technology,, Cambridge, MA, United States, 8Harvard-MIT Divison of Health Sciences and Technology, Massachusetts Institute of Technology, Cambridge, MA, United States |
||
Keywords: Data Analysis, Motion Correction Intra-scan motion costs tens of thousands of dollars per scanner annually due to the need to repeat non-diagnostic scans1. When assessing the scale of the problem and potential solutions, radiologists’ ratings of artifacts are considered the gold standard. However, inconsistent and conflicting ratings must be consolidated into a single gold-standard. We introduce a hierarchical label fusion algorithm that infers each rater's performance and promotes consistency across slices from a volume. This algorithm reduces label noise compared to majority votes, and allows non-expert ratings to be calibrated and included as additional silver-standards. |
| 2016 | Computer 20
|
Recovering slice location for unconventional acquisition plane with deep learning algorithm: cardiac magnetic resonance case |
| Habib Rebbah1 and Timothé Boutelier1 | ||
1Research & Innovation, Olea Medical, La Ciotat, France |
||
Keywords: Machine Learning/Artificial Intelligence, Data Analysis In cardiac MR field, the slice location along the long axis is a key parameter for the standardization proposed by the American heart association. We explore here the ability of a CNN to estimate it. |
| 2017 | Computer 21
|
The effect of imaging parameters, aging, and circadian rhythm on Freesurfer's estimates: A single subject study at 7T over 7 years |
| Hendrik Mattern1,2, Falk Lüsebrink1,3,4, and Oliver Speck1,2,5,6 | ||
1Biomedical Magnetic Resonance, Otto von Guericke University, Magdeburg, Germany, 2German Center for Neurodegenerative Disease, Magdeburg, Germany, 3Institute of Cognitive Neurology and Dementia Research, Otto von Guericke University, Magdeburg, Germany, 4Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig, Germany, 5Center for Behavioral Brain Sciences, Magdeburg, Germany, 6Leibniz Institute for Neurobiology, Magdeburg, Germany |
||
Keywords: Data Analysis, Aging In this explorative study, the effect of imaging parameters, image quality, circadian rhythm, and aging on FreeSurfer’s estimates is investigated. To that end, the human phantom, an openly available data of a single subject scanned at 7T over 7 years with various MR protocols, is used. |
| 2018 | Computer 22
|
Distortion-free diffusion imaging using single-shot diffusion-prepared TSE sequence with spiral-ring readouts and magnitude stabilizers |
| Zhixing Wang1, Xiaozhi Cao2, Kun Qing3, Xue Feng1, John P. Mugler4, and Craig H. Meyer1,4 | ||
1Biomedical Engineering, University of Virginia, Charlottesville, VA, United States, 2Department of Radiology, Stanford University, Stanford, CA, United States, 3Radiation Oncology, City of Hope, Duarte, CA, United States, 4Radiology & Medical Imaging, University of Virginia, Charlottesville, VA, United States |
||
Keywords: Data Acquisition, Data Acquisition This study provides an alternative approach to conventional diffusion-weighted (DW) EPI-based sequences. A 2D single-shot (SS) diffusion-prepared (DP) turbo-spin-echo (TSE) sequence, combined with spiral-ring trajectories and magnitude stabilizers, dubbed “SS-DP-SPRING TSE”, was developed for distortion- and motion artifact-free diffusion imaging. Compared to a SS-DW-EPI sequence, this method is less sensitive to B0-inhomogeneity and thus provides DW-images with improved geometric fidelity. |
| 2019 | Computer 23
|
Pulseq-gSlider: Scanner-independent high-isotropic-resolution dMRI on an open-source platform |
| Qiang Liu1,2, Lipeng Ning1, Congyu Liao3,4, Shaik Imam5, Scott Peltier6, Borjan Gagoski7, Berkin Bilgic8,9,10, William A. Grissom5, Maxim Zaitsev11, Jon-Fredrik Nielsen12, and Yogesh Rathi1,13 | ||
1Department of Psychiatry, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, United States, 2School of Biomedical Engineering, Southern Medical University, Guangzhou, China, 3Department of Radiology, Stanford University, Stanford, CA, United States, 4Department of Electrical Engineering, Stanford University, Stanford, CA, United States, 5Institute of Imaging Science, Department of Biomedical Engineering, Department of Electrical Engineering, Vanderbilt University, Nashville, TN, United States, 6Functional MRI Laboratory, University of Michigan, Ann Arbor, MI, United States, 7Fetal-Neonatal Neuroimaging and Developmental Science Center, Boston Children’s Hospital, Harvard Medical School, Boston, MA, United States, 8Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Charlestown, MA, United States, 9Department of Radiology, Harvard Medical School, Cambridge, MA, United States, 10Harvard-MIT Division of Health Sciences and Technology, Massachusetts Institute of Technology, Cambridge, MA, United States, 11Department of Radiology, Medical Physics, University Medical Center Freiburg, Freiburg, Germany, 12Department of Biomedical Engineering, University of Michigan, Ann Arbor, MI, United States, 13Department of Radiology, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, United States |
||
Keywords: Data Acquisition, Software Tools Using on the open-source pulse sequence programming platform Pulseq, we successfully developed and implemented the gSlider sequence on a 3T scanner without the vendor-specific sequence development environment. One-millimeter whole-brain isotropic dMRI data was acquired using the gSlider implementation from Pulseq and Siemens and test-retest variability was evaluated. The test-retest reliability in fractional anisotropy improved dramatically for the Pulseq implementation while it was comparable for both implementations for mean diffusivity. We demonstrated the feasibility of implementing advanced sequences using the open-source Pulseq platform and expect that it will be easily transferred to and executed on different vendors. |
| 2020 | Computer 24
|
Validation of respiratory correlated 4D-MRI for radiotherapy planning, using a motion phantom and comparing to 4D-CT. |
| Joan Chick1, Evanthia Kousi1, Andreas Wetscherek1, Julie Hughes2, Georgina Hopkinson2, Jessica Gough2,3, Rosalyne Westley2,3, Radhouene Neji4, Alex Dunlop1, Simeon Nill1, Henry Mandeville2,3, Katharine Aitken2,3, Dow-Mu Koh2,3, and Uwe Oelfke1 | ||
1Joint Department of Physics, The Institute of Cancer Research and Royal Marsden Hospital, London, United Kingdom, 2Royal Marsden Hospital, London, United Kingdom, 3The Institute of Cancer Research, London, United Kingdom, 4MR Research Collaborations, Siemens Healthcare, Frimley, United Kingdom |
||
Keywords: Data Acquisition, Radiotherapy, Respiratory Correlated MRI Respiratory correlated 4D-MRI has the potential to be a valuable tool in radiotherapy. This work aims to optimise and validate a 4D-MRI golden angle stack of stars radial sequence for measuring respiratory motion of abdominal organs. A 4D motion phantom is used to compare 4D-MRI against 4D-CT. For craniocaudal movement, motion is underestimated in 4D-MRI but motion estimation improves for non-axial acquisitions. This could be due to the alignment of the motion direction with the cartesian phase encoding direction for axial stack of stars acquisitions. |
| 2021 | Computer 25
|
Visualization of Perivascular Spaces using an Optimized 3D-TSE Sequence with Reduced Flip Angle at 7T |
| Gael Saib1, Zeynep Demir1, Paul Taylor2, S. Lalith Talagala3, and Alan P. Koretsky1 | ||
1NINDS/LFMI, National Institutes of Health, Bethesda, MD, United States, 2NIMH/SSCC, National Institutes of Health, Bethesda, MD, United States, 3NINDS/NMRF, National Institutes of Health, Bethesda, MD, United States |
||
Keywords: Visualization, Neurofluids, CSF Perivascular spaces (PVS) dilatation has been recently linked to aging and neurodegenerative diseases. T2-weighted ultra-high field MRI allows to image the PVS burden at high contrast and spatial resolution non-invasively. However, most of sub-voxel sized PVS cannot be resolved due to partial volume effects and remaining signal from blood and tissue. High-resolution 3D-TSE with reduced flip angle allowed the visualization of PVS with a very high CSF-to-tissue ratio of ~37:1 at 7T. |
| 2022 | Computer 26
|
XTE: The All-in-One Short-TE Imaging Sequence |
| Serhat Ilbey1, Michael Bock1, and Ali Caglar Özen1 | ||
1Division of Medical Physics, Department of Radiology, University Medical Center Freiburg, Freiburg, Germany |
||
Keywords: Pulse Sequence Design, New Trajectories & Spatial Encoding Methods XTE is a general implementation of various sequences, namely UTE, ZTE, SPI, PETRA, and gmPETRA. Each sequence type and their combinations can be realized within XTE by adjusting only a few parameters. We provide phantom and in vivo images acquired with XTE for different sequence settings. |
| 2023 | Computer 27
|
Swin golden angle: a radial profile order for golden ratio sampling with navigator compatibility and eddy-current suppression |
| Zhongsen Li1, Aiqi Sun2, Shuai Wang1, Chuyu Liu1, Sirui Wu1, Haozhong Sun1, Rui Guo3, Haiyan Ding1, and Rui Li1 | ||
1Center for Biomedical Imaging Research, Tsinghua University, Beijing, China, 2Institute of Science and Technology for Brain-Inspired Intelligence, Fudan University, Shanghai, China, 3School of Medical Technology, Beijing Institute of Technology, Beijing, China |
||
Keywords: Data Acquisition, Cardiovascular, golden angle radial, eddy-current suppresion, profile order Golden angle radial trajectory suffers from eddy-current when combined with trueFISP sequence. If additional navigators are required, how to keep golden ratio sampling while maintain stable steady-state is still a problem to be solved. In this work, we propose a novel radial profile order, named "swin golden angle"(swinGA), which is able to achieve golden ratio sampling, stable trueFISP signal, and navigator acquisition simultaneously. We validated the proposed profile in static phantom study and in-vivo study for free-running cardiac imaging. The results show that the proposed swinGA significantly outperforms other sampling profiles and achieves good image quality. |
| 2024 | Computer 28
|
Initial observations and potential pitfalls when validating respiratory correlated MRI for radiotherapy planning. |
| Evanthia Kousi1, Joan Chick1, Andreas Wetscherek1, Julie Hughes2, Georgina Hopkinson2, Jessica Gough2,3, Rosalyne Westley2,3, Radhouene Neji4, Simeon Nill1, Dow-Mu Koh2,3, and Uwe Oelfke1 | ||
1Joint Department of Physics, The Institute of Cancer Research and The Royal Marsden NHS Foundation Trust, London, United Kingdom, 2The Royal Marsden NHS Foundation Trust, London, United Kingdom, 3The Institute of Cancer Research, London, United Kingdom, 4MR Research Collaborations, Siemens Healthcare, Frimley, United Kingdom |
||
Keywords: Data Analysis, Radiotherapy, 4D-MRI, Respiratory correlated MRI, radiotherapy planning In this abstract we present our initial results from the validation of a respiratory correlated 4D MRI golden angle stack of stars radial sequence prior to its clinical implementation and highlight potential experimental pitfalls to avoid that may impact motion range estimations. Correct phantom set up reduced motion underestimations by 3%-18% for the simulated respiratory waveforms considered. Larger motion discrepancies observed for the low amplitude motion which improved by increasing the number of bins. |
| 2025 | Computer 29
|
A robust method for estimating CVR dynamics from breath-hold BOLD data without end-tidal carbon dioxide recordings |
| Nuwan D. Nanayakkara1, Liesel-Ann Meusel1, Nicole D. Anderson1,2, and J. Jean Chen1,3,4 | ||
1Rotman Research Institute, Baycrest Health Sciences, Toronto, ON, Canada, 2Departments of Psychology and Psychiatry, University of Toronto, Toronto, ON, Canada, 3Department of Medical Biophysics, University of Toronto, Toronto, ON, Canada, 4Department of Biomedical Engineering, University of Toronto, Toronto, ON, Canada |
||
Keywords: Data Analysis, fMRI (task based), Cerebrovascular reactivity, CVR Cerebrovascular reactivity (CVR) can be estimated from the BOLD response to a vasoactive stimulus such as a breath-holding (BH) approximated by a sinusoidal regressor. We proposed a robust approach for estimating CVR using the frequency spectrum of BOLD data itself, without regressions or correlations with external recordings. We also estimated regional CVR time delay patterns using the BOLD signal phase relative to a non-brain reference. We demonstrate our pipeline in 18 healthy adults. |
| 2026 | Computer 30
|
A nomogram based on DKI, molecular and clinical risk factors for survival prediction and treatment decision-making in astrocytomas |
| Yan Tan1, Zeliang Liu2, Xiaochun Wang1, and Hui Zhang1 | ||
1First Hospital of Shanxi Medical University, Taiyuan, Shanxi Province, China, China, 2LinFen People's Hospital, Linfen, Shanxi Province, China, China |
||
Keywords: Data Analysis, Tumor, diffusion kurtosis imaging, nomogram, astrocytoma, overall survival, treatment To establish a nomogram by integrating DKI, molecular and clinical risk factors for personalized OS estimation in astrocytomas, and explore the nomogram-based treatment benefits. Astrocytomas are the most common brain tumors with poor prognoses. The detection of risk factors may aid in individualizing therapeutic plans and improve survival. It is concluded the nomogram based on DKI, molecular and clinical risk factors could achieve the individualized OS estimation of astrocytomas with excellent performance. For high-risk patients, surgery plus chemo and/or radiotherapy were recommended; while for low-risk patients, additional chemo and/or radiotherapy did not increase survival benefits. |
| 2027 | Computer 31
|
Dual-echo 3D Spiral Navigators for the Detection of Temporal Motions and B0 Shifts in Gradient-echo Imaging at 3.0 T |
| Yuguang Meng1, James J. Lah2, Jason W. Allen1,2, and Deqiang Qiu1 | ||
1Department of Radiology and Imaging Sciences, Emory University, Atlanta, GA, United States, 2Department of Neurology, Emory University, Atlanta, GA, United States |
||
Keywords: Data Acquisition, Artifacts In gradient-echo imaging, the motion effects on the reconstructed images could be exacerbated by the temporal B0 field changes due to respiration and the subjects’ unintentional position/posture changes during scanning. An efficient 3D spiral navigator independent of the multi-echo gradient-echo (mGRE) acquisition train was designed and implemented in a 3D mGRE sequence. The results showed that although there were unintentional movements during acquisitions, the temporal B0 changes could be significant at 3.0 T. The design provides flexible TE options for mGRE in obtaining simultaneous T1-weighted and T2*-weighted contrasts, quantitative T2* and/or quantitative susceptibility mapping. |
| 2028 | Computer 32
|
Assessing the robustness of the correlation between intra-axonal T2 and axon diameter across participants |
| Veronica P Dell'Acqua1, Chantal M W Tax1,2, Malwina Molendowska1, Greg D Parker1, Derek K Jones1,3, Muhamed Barakovic1,4,5,6,7, and Erick Jorge Canales-Rodriguez6 | ||
1Cardiff University Brain Research Imaging Centre, Cardiff University, Cardiff, United Kingdom, 2University Medical Center Utrecht, Image Sciences Institute, Utrecht, Netherlands, 3Mary MacKillop Institute for Health Research, Australian Catholic University, Melbourne, Australia, 4Translational Imaging in Neurology (ThINk) Basel, University Hospital Basel and University of Basel, Basel, Switzerland, 5MS Center and Research Center for Clinical Neuroimmunology and Neuroscience Basel, University Hospital Basel and University of Basel, Basel, Switzerland, 6Signal Processing Laboratory 5 (LTS5), Ecole Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland, 7Roche Pharma Research & Early Development, Neuroscience and Rare Diseases, Roche Innovation Center, Basel, Switzerland |
||
Keywords: Data Analysis, Relaxometry, Validation, Axon Diameter, Diffusion-Relaxation In-vivo quantification of axon diameter is an attractive and debated topic in the MRI community. The possibility to resolve submicrometric axon diameters non-invasively yields the potential to push further the boundaries in research and clinics but yet, further work is needed to better explore and validate the existing approaches to estimate the inner axon diameter. Recently, the feasibility of estimating the axon diameter from the intra-axonal transverse relaxation time has been investigated combining a diffusion-relaxation protocol and histological data. In the present study, we apply this approach in a larger in vivo population to assess variability across participants. |
| 2029 | Computer 33
|
Optimized Fast Gray Matter Acquisition T1 Inversion Recovery MRI for Localizing Mammillothalamic Tract in Deep Brain Stimulation Targeting |
| Kuan Zhang1, Daehun Kang1, Maria A Halverson1, Shengzhen Tao2, Steven A Messina1, MyungHo In1, Joshua D Trzasko1, John III Huston1, Matthew A Bernstein1, and Yunhong Shu1 | ||
1Radiology, Mayo Clinic, Rochester, MN, United States, 2Radiology, Mayo Clinic, Jacksonville, FL, United States |
||
Keywords: Data Acquisition, Data Acquisition, FGATIR, Deep brain stimulation, Compact 3T, MRI sequence optimization Deep brain stimulation (DBS) is used for neurological disorder treatment but requires precise image-based targeting. The fast gray matter acquisition T1 inversion recovery (FGATIR) sequence was developed to visualize DBS targets, such as the mammillothalamic tract (MTT). We investigated the imaging protocol of FGATIR through simulation, phantom experiment, and volunteer scanning to improve the contrast-to-noise ratio. Adjustments to inversion time, flip angle, and receive bandwidth were made to achieve the highest MTT-thalamus contrast. A compact 3T scanner was also leveraged to further maximize the contrast as the high-performance gradients reduce repetition time, thereby decreased the acquisition window of FGATIR. |
| 2030 | Computer 34
|
Nonsusceptibility frequency contributions and brain size affect quantitative susceptibility mapping in a region-dependent manner |
| Thomas Jochmann1,2, Fahad Salman2, Ademola Adegbemigun2, Jens Haueisen1, and Ferdinand Schweser2,3 | ||
1Department of Computer Science and Automation, Technische Universität Ilmenau, Ilmenau, Germany, 2Buffalo Neuroimaging Analysis Center, Department of Neurology at the Jacobs School of Medicine and Biomedical Sciences, University at Buffalo, The State University of New York, Buffalo, NY, United States, 3Center for Biomedical Imaging, Clinical and Translational Science Institute, University at Buffalo, The State University of New York, Buffalo, NY, United States |
||
Keywords: Image Reconstruction, Quantitative Susceptibility mapping We studied the dependency of quantitative susceptibility mapping to scaling the resolution and amplitude in a brain phantom with and without tissue microstructure. We found that nonsusceptibility frequency contributions cause a resolution- and region-dependent bias to quantitative susceptibility mapping. |
| 2031 | Computer 35
|
Suppressing MRI Background Noise via Modeling Phase Variations |
| Yuchou Chang1, Gulfam Ahmed Saju1, Jasina Yu1, Reza Abiri2, Zhiqiang Li3, and Tianming Liu4 | ||
1Computer and Information Science, University of Massachusetts Dartmouth, North Dartmouth, MA, United States, 2Electrical, Computer and Biomedical Engineering, University of Rhode Island, Kingston, RI, United States, 3Neuroradiology, Barrow Neurological Institute, Phoenix, AZ, United States, 4Computer Science, University of Georgia, Athens, GA, United States |
||
Keywords: Image Reconstruction, Parallel Imaging The background phase variations exist in coil sensitivities. Optimized phase distribution can minimize the noise of the parallel MRI reconstruction. However, phase variation may be originated from different factors, and it is difficult to be modeled. A random feature method is proposed to model phase variations in coil sensitivities. Through a linear reconstruction using the random phase feature, background noise can be suppressed. Augmented phase features make the linear reconstruction better remove background noise. |
| 2032 | Computer 36
|
Model-based image reconstruction for highly accelerated point spread function encoded echo planar imaging |
| Nolan Meyer1,2, Myung-Ho In1, Maria Halverson1, John Huston III1, Matt Bernstein1, and Joshua Trzasko1 | ||
1Radiology, Mayo Clinic, Rochester, MN, United States, 2Mayo Clinic Graduate School of Biomedical Sciences, Rochester, MN, United States |
||
Keywords: Image Reconstruction, Signal Representations A model-based image reconstruction (MBIR) framework is developed for highly accelerated point spread function (PSF) encoded echo planar imaging (EPI). The reconstruction accounts for known physics, and utilizes a subspace representation with local low-rank (LLR) regularization along with a variable splitting solution. Comparisons with images obtained through standard methods demonstrate considerable artifact reduction and sharpness preservation through MBIR. MBIR accommodates arbitrary sampling trajectories along the PSF-encoding dimension, enabling reconstruction of quality images from less than one minute of scan time; and we demonstrate performance boosts with nonstandard trajectories precluded by conventional reconstruction methods. |
| 2033 | Computer 37
|
Automated Cardiac Quantification and Cardiomegaly Stratification: Feasibility of an AI-based Approach on Large-scale Non-Cardiac-Gated MRI |
| Thanh-Duc Nguyen1, Saurabh Garg1, Nasrin Akbari1, Saqib Basar1, Sean London2, Yosef Chodakiewitz2, Rajpaul Attariwala1,2, and Sam Hashemi1,2 | ||
1Voxelwise Imaging Technology Inc., Vancouver, BC, Canada, 2Prenuvo, Vancouver, BC, Canada |
||
Keywords: Machine Learning/Artificial Intelligence, Heart, Deep learning segmentation, cardiomegaly Despite its ubiquity, cardiac assessment at non-cardiac-gated MRI has yet to be standardized, leading to misdiagnosis. This paper examines the feasibility of AI for automatic 3D volumetric cardiac quantification used to detect and stratify cardiomegaly on non-cardiac-gated torso MRI. Using AI, we automatically measured the cardiothoracic ratio and 3D cardiac volumetric features as indicators for detections of cardiomegaly conditions. Large-scale results on 3485 normal-heart individuals revealed the effect of aging on cardiac features in both men and women. AI findings on non-cardiac-gated imaging offer opportunistic useful information, increasing the diagnostic precision, and allowing potentially useful imaging monitoring for cardiomegaly-associated conditions. |
| 2034 | Computer 38
|
An investigation of the performance of T2-weighted MR imaging with AI-assisted compressed sensing in routine clinical settings |
| Adiraju Karthik1, Apoorwa Devappa2, Aakaar Kapoor3, Dharmesh Singh4, and Dileep Kumar4 | ||
1Department of Radiology, Sprint Diagnostics, Jubilee Hills, Hyderabad, India, 2Department of Radiology, Mahadevappa Rampure Medical College, Kalaburagi, India, 3Department of Radiology, City X-Rays Scan & Clinical Private Limited, New Delhi, India, 4Central Research Institute, Global Scientific Collaborations, United Imaging Healthcare, New Delhi, India |
||
Keywords: Data Acquisition, Body, AI-assisted compressed sensing, T2-weighted Imaging T2-weighted imaging (T2WI) is an essential diagnostic tool for several diseases. However, one of the challenges faced by patients and radiology departments is the longer scanning time of MR examinations. Recent advancements in artificial intelligence (AI) and deep learning techniques have made it able to acquire images quickly while preserving high-quality image resolution. In this study, the efficacy of a deep learning-based reconstruction technique termed AI-Assisted Compressed Sensing (ACS) was evaluated qualitatively and quantitatively using T2WI in routine clinical settings for brain, spine, knee, kidney and liver. |
| 2035 | Computer 39
|
Practical Correction of Gradient Nonlinearities in Diffusion-Weighted Imaging |
| Praitayini Kanakaraj1, Leon Y Cai2, Francois Rheault3, Baxter P Rogers4, Adam Anderson2,4, Kurt G Schilling4, and Bennett A Landman1,2,4,5 | ||
1Department of Computer Science, Vanderbilt University, Nashville, TN, United States, 2Department of Biomedical Engineering, Vanderbilt University, Nashville, TN, United States, 3Department of Computer Science, Université de Sherbrooke, Sherbrooke, QC, Canada, 4Department of Radiology and Radiological Sciences, Vanderbilt University Medical Center, Nashville, TN, United States, 5Department of Electrical and Computer Engineering, Vanderbilt University, Nashville, TN, United States |
||
Keywords: Data Processing, Diffusion/other diffusion imaging techniques Gradient nonlinearity correction is well-established but not straightforward to implement in existing diffusion software packages due to it producing gradient tables that vary by voxel. We propose a simple, practical approach that approximates full correction by: (1) scaling the diffusion signal and (2) resampling the gradient orientations. Our approach results in uniform gradients across the corrected image and provides the key advantage of seamless integration into current diffusion pipelines. The proposed method resulted in negligible differences in multi- compartment indices from the standard voxel-wise empirical correction. |
| 2175 | Computer 1
|
A New Framework for 3D MR Fingerprinting with Efficient Subspace Reconstruction and Joint Posterior Distribution Estimation |
| Jiaren Zou1,2, Yuchi Liu3, Jesse Hamilton2,3, Yun Jiang2,3, Nicole Seiberlich2,3, and Yue Cao1,2,3 | ||
1Department of Radiation Oncology, University of Michigan, Ann Arbor, MI, United States, 2Department of Biomedical Engineering, University of Michigan, Ann Arbor, MI, United States, 3Department of Radiology, University of Michigan, Ann Arbor, MI, United States |
||
Keywords: MR Fingerprinting/Synthetic MR, MR Fingerprinting Iterative image reconstruction of highly undersampled high-resolution 3D MR fingerprinting (MRF) is time-consuming and has high memory requirements. In this work, we propose to use stochastic gradient descent to accelerate the reconstruction and reduce the memory footprint. In addition, a conditional invertible neural network is used as a fast and flexible tool for estimating the posterior distribution of tissue properties from MRF. In a simulation study, we achieved an 11-fold and 45.5GB reduction in reconstruction time and memory requirement, respectively, compared with a conventional iterative method. Uncertainty maps of tissue properties derived from the estimated posterior distributions correlate well with reconstruction errors. |
| 2176 | Computer 2
|
Multi-parametric quantitative MRI without inversion pulses by optimized RF phase modulation |
| Miha Fuderer1, Hongyan Liu1, Oscar van der Heide1, Cornelis A.T. van den Berg1, and Alessandro Sbrizzi1 | ||
1Division Imaging and Oncology, University Medical Center Utrecht, Utrecht, Netherlands |
||
Keywords: MR Fingerprinting/Synthetic MR, Relaxometry We applied optimized phase modulation to the RF sequence in MR-STAT. We observe that applying optimized phase is beneficial towards the accuracy of T1 and T2 maps, particularly under constrained conditions. In particular, when comparing sequences with and without an initial inversion pulse we observe that the difference vanishes when applying optimized RF phase modulation. Thus optimized phase modulation allows to omit the initial inversion pulse. We hypothesize that the same holds for other multi-parametric qMRI techniques such as MRF. |
| 2177 | Computer 3
|
Enhancing the precision of multi-phase parametric maps in 4D-MRF by optimization of local T1 and T2 sensitivities |
| Yat Lam Wong1, Tian Li1, Chenyang Liu1, Victor Ho Fun Lee2, Lai Yin Andy Cheung3, Peng Cao4, Edward Sai Kam Hui5, and Jing Cai1 | ||
1Department of Health Technology and Informatics, The Hong Kong Polytechnic University, Hong Kong, Hong Kong, 2Department of Clinical Oncology, The University of Hong Kong, Hong Kong, Hong Kong, 3Department of Clinical Oncology, Queen Mary Hospital, Hong Kong, Hong Kong, 4Department of Diagnostic Radiology, The University of Hong Kong, Hong Kong, Hong Kong, 5The Chinese University of Hong Kong, Hong Kong, Hong Kong |
||
Keywords: MR Fingerprinting/Synthetic MR, MR Fingerprinting Time-resolved magnetic resonance fingerprinting (MRF), or 4D-MRF, has demonstrated its feasibility in motion management in radiotherapy (RT). However, the prohibitive long acquisition time is one of challenges of the clinical implementation of 4D-MRF. The shortening of acquisition time may compromise the accuracies of the predicted tissues’ properties. Here, we propose a novel technique to enhance the accuracies of 4D-MRF with shortened acquisition by optimizing T1 and T2 sensitivities through inter-phase data sharing. |
| 2178 | Computer 4
|
Abdominal MR Fingerprinting with In-bore Breathing Guidance |
| Yoo Jin Lee1, Peter Koken1, Kristina Sonnabend2,3,4, Grischa Bratke3,4, Annerieke Heuvelink-Marck5, and Mariya Doneva1 | ||
1Philips Research, Hamburg, Germany, 2Philips GmbH Market DACH, Hamburg, Germany, 3Faculty of Medicine, University of Cologne, Cologne, Germany, 4Institute for Diagnostic and Interventional Radiology, University Hospital Cologne, Cologne, Germany, 5Philips Research, Eindhoven, Netherlands |
||
Keywords: MR Fingerprinting/Synthetic MR, Body, Respiratory motion MR fingerprinting (MRF) in the abdomen is challenging due to respiratory motion. In this study, as an alternative to a series of extended breath-holds which could be challenging especially for patients, visual and auditive in-bore breathing guidance was used during the MRF scans for better patient comfort. MRF sequences were designed based on predefined breathing patterns such that data are acquired only during exhale holds. The corresponding MRF dictionaries were calculated beforehand to enable real-time matching of quantitative maps. We demonstrate that MRF scans with guided breathing provide comparable abdominal T1 and T2 maps as those with breath-holds. |
| 2179 | Computer 5
|
Incorporating fat and spatial saturation in MR Fingerprinting |
| Christopher George Trimble1, Kaia I. Sørland1, Tone F. Bathen1,2, Mattijs Elschot1,2, and Martijn A. Cloos3 | ||
1Department of Circulation and Medical Imaging, Norwegian University of Science and Technology (NTNU), Trondheim, Norway, 2Department of Radiology and Nuclear Medicine, St Olavs Hospital, Trondheim, Norway, 3Centre for Advanced Imaging, University of Queensland, Brisbane, Australia |
||
Keywords: MR Fingerprinting/Synthetic MR, Prostate Magnetic Resonance Fingerprinting (MRF) enables fast quantitative MR imaging, and represents an opportunity to improve upon the current standard of multi-parametric MRI (mp-MRI). In this work we demonstrate how the MRF pulse sequence can be adapted to include different saturation techniques, whilst retaining accurate T1 and T2 quantification. Extended Phase Graphs (EPGs) of the sequence illustrate the effective spoiling of chemical elements, such as fat, and spatial saturation. Phantom imaging shows that T1 quantification is not significantly affected and T2 quantification remains accurate up to 100ms. As an example, the framework is tested in-vivo in the context of prostate imaging. |
| 2180 | Computer 6
|
Magnetisation Transfer effects on T1 and T2 values in MR Fingerprinting |
| Simran Kukran1,2, Iulius Dragonu3, Ben Statton1,4, Jack Allen5, Pete Lally1,6, Rebecca Quest1,7, Neal Bangerter1, Dow-Mu Koh2, Matthew Orton2, and Matthew Grech-Sollars8 | ||
1Imperial College London, London, United Kingdom, 2Institute of Cancer Research, London, United Kingdom, 3Siemens Healthcare Ltd, Frimley, United Kingdom, 4London Institute of Medical Sciences, Medical Research Council, London, United Kingdom, 5Independent Researcher, Norwich, United Kingdom, 6UK Dementia Research Institute Centre for Care Research and Technology, London, United Kingdom, 7Imperial College Healthcare NHS Trust, London, United Kingdom, 8University College London, London, United Kingdom |
||
Keywords: MR Fingerprinting/Synthetic MR, Magnetization transfer A fingerprinting sequence was developed with and without off resonance pulses before every TR to investigate MT effects on T1 and T2 values. Off-resonance pulses suppress MT effects by saturating signal from free protons that exchange with bound protons. Observed T2 values increased with off resonance pulses as expected. Observed T1 values decreased, this was an opposite effect to what was observed previously with a different pulse sequence. |
| 2181 | Computer 7
|
A steady-state MRF sequence optimization framework for 3D simultaneous water T1 and fat fraction mapping |
| Constantin Slioussarenko1, Pierre-Yves Baudin1, and Benjamin Marty1 | ||
1NMR Laboratory, Neuromuscular Investigation Center, Institute of Myology, Paris, France |
||
Keywords: MR Fingerprinting/Synthetic MR, MR Fingerprinting We introduce an optimization framework using the longitudinal steady-state equilibrium for FLASH MRF sequences, and taking into account signal noise by using a Monte Carlo method. We use this framework to optimize echo times, flip angles, timings and recovery time for T1H2O and FF mapping while shortening our original MRF T1-FF sequence. The resulting Fast MRF T1-FF sequence yields comparable estimation results to the original sequence with an almost twice reduced acquisition time. |
| 2182 | Computer 8
|
Quantifying 3D-MRF Reproducibility Across Subjects, Sessions, and Scanners Automatically Using MNI Atlases |
| Andrew Dupuis1, Yong Chen1, Michael Hansen2, Kelvin Chow3, Dan Ma1, Mark Griswold1, and Rasim Boyacioglu1 | ||
1Biomedical Engineering, Case Western Reserve University, Cleveland, OH, United States, 2Microsoft Research, Redmond, WA, United States, 3Siemens Medical Solutions USA, Inc., Chicago, IL, United States |
||
Keywords: MR Fingerprinting/Synthetic MR, Quantitative Imaging, Reproducibility, MNI, Automated, Gadgetron, FIRE We evaluate the reproducibility of 3D-MRF versus clinical-standard MPRAGE and TSE acquisitions for ten subjects across three acquisition sets on two scanner via a fully-automated registration and regional analysis framework. T1 and T2 quantitative maps from 3D-MRF were found to be highly reproducible (T1+/-4.6%, T2+/-6.3%) across scanners and sessions, with no significant difference between repeating a scan immediately on the same scanner, repeating a scan after repositioning the subject and reshimming, and repeating a scan on a different scanner entirely. The same is not true for MPRAGE (+/-12.4%) or TSE (+/-28.8%) acquisitions. |
| 2183 | Computer 9
|
Deep learning optimization of CEST MR Fingerprinting (CEST-MRF) for Quantitative Human Brain Mapping |
| Ouri Cohen1 and Ricardo Otazo1 | ||
1Medical Physics, Memorial Sloan Kettering, New York, NY, United States |
||
Keywords: MR Fingerprinting/Synthetic MR, CEST & MT CEST MR fingerprinting (CEST-MRF) enables fast quantitative relaxation and exchange mapping. The CEST-MRF signal depends on multiple acquisition and tissue parameters which makes optimization of the acquisition schedule challenging. The goal of this work is to develop a deep learning approach that uses a quantification network and a surrogate network to optimize the acquisition schedule for in vivo scans. Numerical simulations are used to characterize the optimized schedule and the benefits of optimization are demonstrated in vivo in a healthy subject. The optimized schedule can reduce scan time by 12% and provide better image quality than a randomly generated schedule. |
| 2184 | Computer 10
|
Five clinical contrasts from 1 minute whole brain MRF with B0 correction |
| Sophie Schauman1,2, Siddharth Iyer3,4, Xiaozhi Cao1,2, Quan Chen1,2, Mahmut Yurt1,2, Natthanan Ruengchaijatuporn5, Congyu Liao1,2, Greg Zaharchuk1, and Kawin Setsompop1,2 | ||
1Department of Radiology, Stanford University, Stanford, CA, United States, 2Department of Electrical Engineering, Stanford University, Stanford, CA, United States, 3Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Charlestown, MA, United States, 4Department of Electrical Engineering and Computer Science, MIT, Cambridge, MA, United States, 5Center of Excellence in Computational Molecular Biology, Chulalongkorn University, Bangkok, Thailand |
||
Keywords: MR Fingerprinting/Synthetic MR, MR Fingerprinting Clinical contrasts can be generated from a 1 min 1 mm isotropic whole brain MRF scan and B0 correction improves it further. We tested this method in healthy volunteers as well as clinical populations. |
| 2185 | Computer 11
|
Neural Network-Based Removal of Colored Noise in MR Fingerprinting |
| Jakob Meineke1, Christian Wülker1, Jean Tkach2, Usha Nagaraj2, Mariya Doneva1, and Tim Nielsen1 | ||
1Philips Reseach, Hamburg, Germany, 2Department of Radiology, Cincinnati Children's Hospital Medical Center, Cincinnati, OH, United States |
||
Keywords: MR Fingerprinting/Synthetic MR, MR Fingerprinting A neural network for the estimation and removal of colored noise in non-Cartesian MRI is trained and subsequently applied for denoising coefficient images in MR Fingerprinting. It is shown that the neural network does not introduce bias in the quantitative parameter maps and improves their precision. Invivo T1- and T2-maps are demonstrated to be visually improved. |
| 2186 | Computer 12
|
Iterative denoising of undersampled 3D MRF with exact line search and spatial regularization |
| Constantin Slioussarenko1 and Benjamin Marty1 | ||
1NMR Laboratory, Neuromuscular Investigation Center, Institute of Myology, Paris, France |
||
Keywords: MR Fingerprinting/Synthetic MR, Sparse & Low-Rank Models We propose an iterative algorithm to remove remaining artefacts of Magnetic Resonance Fingerprinting maps, using projected gradient descent with exact line search and spatial regularization. We use this framework to denoise 3D MRF T1-FF acquisitions undersampled in the partition direction and show that this allows to reduce undersampling artefacts for the T1H2O and FF maps after a few iterations. |
| 2187 | Computer 13
|
The influence of diffusion in fast multi-parametric relaxometry |
| Miha Fuderer1, Oscar van der Heide1, C.A.T. van den Berg1, and Alessandro Sbrizzi1 | ||
1Division Imaging and Oncology, University Medical Center Utrecht, Utrecht, Netherlands |
||
Keywords: MR Fingerprinting/Synthetic MR, Relaxometry In MRF, it has been established – but often ignored – that the presence of diffusion causes bias in the estimated T2. We measure this effect in MR-STAT. Thereby we confirm the theoretical model that the bias is proportional to the square of T2. We also find that the bias level strongly depends on the RF sequence used. This opens the prospect of optimizing sequences on minimal diffusion-induced bias in T2. |
| 2188 | Computer 14
|
T1ρ Dispersion Characterization by Magnetic Resonance Fingerprinting |
| Brendan Eck1,2, Jeehun Kim2,3, Mingrui Yang2, and Xiaojuan Li2 | ||
1Imaging Institute, Cleveland Clinic, Cleveland, OH, United States, 2Department of Biomedical Engineering, Program of Advanced Musculoskeletal Imaging (PAMI), Cleveland Clinic, Cleveland, OH, United States, 3Department of Electrical, Computer, and Systems Engineering, Case Western Reserve University, Cleveland, OH, United States |
||
Keywords: MR Fingerprinting/Synthetic MR, Quantitative Imaging, T1rho Dispersion T1ρ dispersion is a potential biomarker that can characterize chemical exchange or diffusive exchange, complementary to other MR relaxometry parameters (e.g. T1, T2). However, T1ρ dispersion conventionally involves quantification of T1ρ at multiple spin-lock frequencies, which can require prohibitively long scan times and high radiofrequency pulse energy deposition. MR Fingerprinting (MRF) has been reported for quantification of T1ρ at a single spin-lock frequency. We propose a framework to enable T1ρ dispersion MRF data acquisition and reconstruction. We report initial simulation and real-world phantom results that demonstrate the feasibility of this MRF-based T1ρ dispersion characterization method. |
| 2189 | Computer 15
|
Fast Deep Learning models for Magnetic Resonance Fingerprinting |
| Raffaella Fiamma Cabini1,2, Davide Cicolari2,3, Leonardo Barzaghi1,4, Paolo Arosio5,6, Stefano Carrazza5,6, Silvia Figini2,7, Marta Filibian2,8, Marco Peviani9, Anna Pichiecchio4,10, and Alessandro Lascialfari2,3 | ||
1Mathematics, University of Pavia, Pavia, Italy, 2INFN, Istituto Nazionale di Fisica Nucleare, Pavia, Italy, 3Physics, University of Pavia, Pavia, Italy, 4Advanced Imaging and Radiomics, Department of Neuroradiology, IRCCS Mondino Foundation, Pavia, Italy, 5Physics, University of Milano, Milano, Italy, 6INFN, Istituto Nazionale di Fisica Nucleare, Milano, Italy, 7Department of Social and Political Science, University of Pavia, Pavia, Italy, 8Centro Grandi Strumenti, University of Pavia, Pavia, Italy, 9University of Pavia, Pavia, Italy, 10Department of Brain and Behavioural Sciences, University of Pavia, Pavia, Italy |
||
Keywords: MR Fingerprinting/Synthetic MR, Data Processing We proposed a DL method and an automatic hyperparameters optimization strategy to reconstruct T1 and T2 maps acquired with two Magnetic Resonance Fingerprinting (MRF) sequences. The model was trained and validated on a preclinical MRF dataset and tested on an independent test set. Through a lower number of MRF images and a lower k-space sampling percentage than the standard post-processing, the DL-based method and the automatic hyperparameters optimization strategy deliver parametric maps with similar accuracy as the dictionary-based methodology. |
| 2190 | Computer 16
|
Simultaneous T1 T2 T2* quantification using 2D EPI-MRF by shuffled sampling and compressed time-resolved reconstruction with B0 B1+ correction |
| Di Cui1, Xiaoxi Liu1, Peder E.Z. Larson1, and Duan Xu1 | ||
1University of California, San Francisco, San Francisco, CA, United States |
||
Keywords: MR Fingerprinting/Synthetic MR, MR Fingerprinting A 2D EPI based MRF acquisition and reconstruction method was developed in this study. A shuffled acquisition order and pseudo-randomized sampling pattern was designed for 2D k-space and compressed time-resolved reconstruction was utilized for faster reconstruction. T1 T2 T2* were simultaneously quantified, B0, B1+ and proton density maps are generated. |
| 2191 | Computer 17
|
Correction of B0 Eddy Current Effects in Magnetic Resonance Fingerprinting at 5.0T Whole Body Scanner |
| Yifan Guo1, Lixian Zou2, Congcong Liu2, Fanshi Li2, Yanjie Zhu2, Xin Liu2, Hairong Zheng2, Dong Liang2, and Haifeng Wang2 | ||
1Research Center for Medical AI, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China, 2Paul C. Lauterbur Research Center for Biomedical Imaging, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China |
||
Keywords: MR Fingerprinting/Synthetic MR, MR Fingerprinting Spiral-based MR fingerprinting (MRF) is an efficient T1 and T2 quantification technique for simultaneous measurement of tissue properties. Affected by the artifacts caused by the B0 eddy currents effects at 5T, a system response eddy current correction strategy is proposed to mitigate the blurring and artifacts of the MRF Spiral. The results of in vivo experiments on the head and abdomen showed that our approach to correct the artifacts caused by the B0 eddy currents was effective. |
| 2192 | Computer 18
|
Super-resolution GAN network for fast quantification of proton density mapping from highly accelerated synthetic magnetic resonance imaging |
| Yawen Liu1, Pengling Ren2, Hongxia Yin3, Yi Zhu4, Rong Wei5, Tingting Zhang2, Zuofeng Zheng6, and Zhenchang Wang1,2 | ||
1School of Biological Science and Medical Engineering, Beihang University, Beijing, China, 2Department of Radiology, Beijing Friendship Hospital, Capital Medical University, Beijing, China, 3Department of Medical Engineering, Beijing Friendship Hospital, Capital Medical University, Beijing, China, 4Philips Healthcare, Beijing, China, 5Peking university Academy for Advanced Interdisciplinary Studies, Beijing, China, 6Beijing ChuiYangLiu Hospital, Beijing, China |
||
Keywords: MR Fingerprinting/Synthetic MR, Image Reconstruction Quantitative magnetic resonance imaging (qMRI) can reflect the inherent characteristics of human tissue of relaxation time and proton density, and has important value for clinical diagnostic. However, long scan times limit the use of qMRI. We propose a method to optimize fast qMRI using a super-resolution generative adversarial network, thereby reducing scan time and obtaining accurate quantitative values. The results showed that this method was able to improve the image quality of qMRI, and the quantitative values were not significantly different from those obtained in conventional acquisitions. |
| 2193 | Computer 19
|
The diagnostic value of synthetic MRI in predicting the activity of Graves ophthalmopathy |
| Luxi Yang1, Wenhua Wang1, Yunwen Chen1, Miaoqi Zhang2, Lisha Nie 2, Xiao Zhang1, and Zhihua Pan1 | ||
1416 Hospital of Nuclear Industry, Chengdu, China, 2GE Healthcare, MR Research, Beijing, China |
||
Keywords: Quantitative Imaging, Head & Neck/ENT In this study, we aim to investigate the performance of quantitative measurements from synthetic MRI in predicting the disease activity of Graves’ ophthalmopathy (GO). 40 patients with different Clinical Activity Scores are investigated. We found a trend that the muscles’ T1, T2 and PD values increased with increasing CAS level, spatially for T2 values. Furthermore, fact that T2 escalation in ‘inactive’ CAS groups compared with healthy controls, indicate CAS staging may underestimate disease activity. We conclude that synthetic MRI have potential to predict clinical activity of GO, and T2 value has the highest diagnostic efficiency. |
| 2194 | Computer 20
|
Synthetic MRI derived Quantitative Maps in Diagnosis of Endometrial Carcinoma |
| Hailei Gu1, Weiqiang Dou2, and Wenwei Tang1 | ||
1Women's Hospital of Nanjing Medical University, Nanjing Maternity and Child Health Care Hospital, Nanjing, China, 2GE Healthcare, MR Research China, Beijing, China |
||
Keywords: Data Analysis, Cancer This study aimed to investigate whether synthetic-MRI can diagnose endometrial carcinoma (EC). 116 patients were recruited, including submucous myomas(SM),endometrial hyperplasia and endometrial polyps(EH&EP) and EC. Synthetic MRI derived T1, T2 and proton density(PD) mapping were obtained for each patient. Significantly different T1 and T2 were shown between EC and SM, and the optimal diagnostic efficacy was for T1+T2 with AUC of 0.865 . T2 and PD were also different between EC and EH&EP. The corresponding optimal diagnostic efficacy was for T2+PD with AUC of 0.690. With these findings, synthetic-MRI may thus be effective in differentiating EC from other intrauterine diseases. |
| 2195 | Computer 21
|
Comparison of k-space Sampling Patterns for MR Fingerprinting |
| Felix Horger1,2,3, Raphael Tomi-Tricot1,2,3,4, Pierluigi Di Cio1,3, Joseph Hajnal1,2,3, and Shaihan Malik1,2,3 | ||
1Biomedical Engineering Department, School of Biomedical Engineering & Imaging Sciences, King's College London, London, United Kingdom, 2Centre for the Developing Brain, School of Biomedical Engineering & Imaging Sciences, King's College London, London, United Kingdom, 3London Collaborative Ultra high field System (LoCUS), London, United Kingdom, 4MR Research Collaborations, Siemens Healthcare Limited, Frimley, United Kingdom |
||
Keywords: MR Fingerprinting/Synthetic MR, In Silico The temporal low-rank property of signals produced by MR Fingerprinting can be used to circumvent reconstruction of individual time frames. Instead, temporally compressed components are estimated. Since every acquired sample then contributes to every compressed component, it is unclear which k-space sampling pattern is optimal. This work collects evidence for higher robustness of radial sampling towards measurement errors and its superior ability to resolve rapidly varying temporal signals, compared to a Cartesian pattern, simulated in a realistic scenario and supported by an in vivo observation. The aim is making informed decisions about optimal sampling patterns for MRF. |
| 2196 | Computer 22
|
Snapshot CEST at 3T using 3D True FISP |
| Yupeng Wu1, Zhichao Wang2, Qifan Pang1, Gaiying Li1, Mengying Chen1, Xu Yan3, Caixia Fu4, Yang Song3, and Jianqi Li1 | ||
1Shanghai key lab of magnetic resonance, East China Normal University, Shanghai, China, 2Zhejiang Lab, Hangzhou, China, 3MR Scientific Marketing, Siemens Healthineers, Shanghai, China, 4MR Application Development, Siemens Shenzhen Magnetic Resonance Ltd, Shenzhen, China |
||
Keywords: Pulse Sequence Design, CEST & MT, GRE FLASH TrueFISP For clinical implementation, chemical exchange saturation transfer (CEST) imaging must be fast with high signal‐to‐noise ratio (SNR), 3D coverage, and produce robust contrast. In general, CEST imaging with fast low angle shot GRE sequence (FLASH) can achieve fast acquisition, but may yield low SNR efficiency. This study evaluated APT and NOE maps for single saturation 3D CEST imaging with True FISP and FLASH sequences, respectively. We found that True FISP could achieve better SNR of Z-spectrum and obtain more reliable CEST effects contrast. 3D True FISP sequence is expected to become a more promising 3D CEST-GRE sequence for clinical application. |
| 2197 | Computer 23
|
Cartesian MR-STAT vs spiral MR Fingerprinting: a comparison |
| Oscar van der Heide1,2, Mariya Doneva3, Peter Koken3, Jakob Meineke3, Miha Fuderer1,2, Cornelis A.T. van den Berg1,2, and Alessandro Sbrizzi1,2 | ||
1Computational Imaging Group for MR Diagnostics and Therapy, Center for Image Sciences, University Medical Center Utrecht, Utrecht, Netherlands, 2Department of Radiology, Division of Imaging and Oncology, University Medical Center Utrecht, Utrecht, Netherlands, 3Philips Research, Hamburg, Germany |
||
Keywords: MR Fingerprinting/Synthetic MR, Quantitative Imaging MR Fingerprinting (“MRF”) and MR Spin Tomography in Time-domain (“MR-STAT”) are quantitative MRI techniques that allow multiple quantitative tissue parameter maps (e.g. T1, T2 and proton density) to be estimated from a single short scan. In this work, we compare the accuracy and precision of Cartesian MR-STAT and spiral MRF on gel phantoms and in-vivo. On gel phantoms we get excellent agreement with reference measurements. In-vivo we observe differences in reconstructed T1 values. As for the precision, a more in-depth study is required to take into account differences in noise-suppression mechanisms (i.e. k-space apodization and spiral sampling window).
|
| 2198 | Computer 24
|
3D Extension of the multi parametric acquisition Multi-Phase balanced non-Steady State Free Precession |
| Riwaj Byanju1, Gyula Kotek1, Mika W. Vogel2, Juan A Hernandez-Tamames1, and Dirk H. J. Poot1 | ||
1Erasmus MC, Rotterdam, Netherlands, 2GE Healthcare, Hoevelaken, Netherlands |
||
Keywords: Pulse Sequence Design, Quantitative Imaging We propose a 3D extension of the novel MP-b-nSSFP sequence, which interleaves RF pulse types for Multiparametric mapping from the transient response. Comparing selective versus non-selective refocusing pulses we observe lower bias with non-selective pulses, despite modelling the spatially varying effect of the pulses in the fitting process. Phantom and in-vivo comparison to QRAPTEST (MAGIC) are performed. |
| 2199 | Computer 25
|
Comparison of UTE-T1ρ vs. MAPSS-T1ρ Sequences in In-Vivo Knees |
| Sara E Sacher1, Michael Carl2, Hollis G Potter1, and Matthew F Koff1 | ||
1Hospital for Special Surgery, New York, NY, United States, 2GE Healthcare, San Diego, CA, United States |
||
Keywords: Pulse Sequence Design, Quantitative Imaging, Ultrashort echo time The relative accuracy of an ultrashort echo time (UTE) based T1ρ sequence was compared to a MAPSS based T1ρ sequence in the evaluation of articular cartilage. Scanning with similar parameters between the acquisitions was performed, producing similar T1ρ values in all cartilage sub-compartments except for the femoral trochlea (TrF). TrF MAPSS-T1ρ values were shorter than TrF UTE-T1ρ values (10.91% difference, p = 0.0028). Overall, there was reasonable agreement between the two sequences indicating that UTE-T1ρ may be a promising method to use in place of conventional MAPSS sequences to quantify T1ρ values of articular cartilage. |
| 2200 | Computer 26
|
Decoding of 3T and 7T bSSFP profile asymmetries for T1, T2, and fraction quantification in two-compartment systems |
| Nils Marc Joel Plähn1, Adèle Mackowiak2, Berk Açikgöz3, and Jessica Bastiaansen3 | ||
1Department of Diagnostic, Interventional and Pediatric Radiology (DIPR), Inselspital, Bern Universit, Bern, Switzerland, 2Department of Radiology, Lausanne University Hospital (CHUV) and University of Lausanne (UNIL), Lausanne, Switzerland, 3Department of Diagnostic, Interventional and Pediatric Radiology (DIPR), Inselspital, Bern University Hospital, University of Bern, Bern, Switzerland |
||
Keywords: Data Analysis, High-Field MRI, phase-cycled bSSFP, Aceton fraction quantification A novel Off-Resonant encoded Analytical parameter quantification using Complex Linearized Equations (ORACLE) method using phase-cycled bSSFP profiles was developed. The approach decodes complex asymmetry profiles in multi-compartment systems for simultaneous proton fraction, T1/T2 ratio, T1 and T2 quantification. The approach was validated in simulations and in an acetone-water phantom at 3T and 7T. Simulations and experiments validated the proposed method for multi-parameter quantification using phase-cycled bSSFP for two compartment singlet systems with high accuracy and precision. This provides the first step towards proton fraction quantification of more complex multiplet-systems, such as fat or myelin, exploiting complex asymmetry profiles. |
| 2201 | Computer 27
|
Split spin-echo acquisition for ultrashort T2* mapping: a simulation study |
| Mikhail Zubkov1 and Irina Melchakova1 | ||
1School of Physics and Engineering, ITMO University, Saint Petersburg, Russian Federation |
||
Keywords: Data Acquisition, Pulse Sequence Design MRI serves as a non-invasive way of assessing the degree of iron overload. This is done via T2*-mapping commonly performed with muti-echo gradient echo sequences. This approach fails in cases with extreme iron overload where the relaxation times drop down to submillisecond range. A method allowing ultra-short T2* mapping is simulated, employing a temporal acquisition offset in a spin-echo pulse sequence. Two acquisitions are suggested: one – before the spin-echo centre, and another – after, resulting in two different T2*-weighted images, acquired with half-line radial encoding. This method allows obtaining T2*-weighted images with weightings starting effectively at zero echo time. |
| 2202 | Computer 28
|
VUDU-SAGE: Efficient T2 and T2* Mapping using Joint Reconstruction for Motion-Robust, Distortion-Free, Multi-Shot, Multi-Echo EPI |
| Jaejin Cho1,2, Tae Hyung Kim3, Avery JL Berman4, Yohan Jun1,2, Xiaoqing Wang1,2, Borjan Gagoski2,5, and Berkin Bilgic1,2,6 | ||
1Athinoula A. Martinos Center for Biomedical Imaging, Charlestown, MA, United States, 2Harvard Medical School, Boston, MA, United States, 3Computer Engineering, Hongik University, Seoul, Korea, Republic of, 4Carleton University, Ottawa, ON, Canada, 5Fetal-Neonatal Neuroimaging & Developmental Science Center, Boston Children’s Hospital, Boston, MA, United States, 6Harvard/MIT Health Sciences and Technology, Cambridge, MA, United States |
||
Keywords: Data Acquisition, Quantitative Imaging, T2 and T2* mapping We demonstrate T2 and T2* mapping and joint reconstruction for multi-echo, multi-shot EPI data from a variable flip angle, blip-up and -down undersampling (VUDU) for spin and gradient echo (SAGE) acquisition. VUDU-SAGE employs a blip-up and -down acquisition strategy for correcting B0 distortion, and FLEET-ordering for motion-robust multi-shot EPI while maximizing signal using variable flip angles. VUDU-SAGE acquires five echoes consisting of two gradient-echo, two mixed, and one spin-echo contrasts. We jointly reconstruct all echoes and estimate T2 and T2* maps using Bloch dictionary matching. In-vivo experiment presents T2 and T2* map at 1x1x4mm3 resolution with a 9-second acquisition. |
| 2203 | Computer 29
|
Effects of Nucleus Pulposus Tissue Hydration on T1ρ and T2 Relaxation Times and Mechanical Properties |
| Megan Co1, Brian Raterman2, Arunark Kolipaka1,2, and Benjamin A Walter1,3 | ||
1Department of Biomedical Engineering, The Ohio State University, Columbus, OH, United States, 2Department of Radiology, The Ohio State University, Columbus, OH, United States, 3Spine Research Institute, The Ohio State University, Columbus, OH, United States |
||
Keywords: Data Acquisition, Quantitative Imaging, Spine, Intervertebral Disc T1ρ and T2 mapping have been developed to quantitively assess intervertebral disc (IVD) degeneration by quantifying proteoglycan and water content, respectively. In addition, magnetic resonance elastography (MRE) has been validated to quantify shear stiffness. This study determines how water content affects MRI relaxation times and MRE-derived mechanical properties. Our results showed that hydration has an influence on T1ρ and T2 relaxation times and MRE-derived shear stiffnesses and that there is a correlation between relaxation times and shear stiffnesses. This highlights MRI as a non-invasive technique to quantify tissue composition and mechanical properties to assess degenerative changes within the IVD. |
| 2204 | Computer 30
|
Unsupervised Learning-based Pulse Sequence Optimization framework for Magnetic Resonance Fingerprinting |
| Peng Li1, Yinghao Zhang1, Xin Lu2, and Yue Hu1 | ||
1Harbin Institute of Technology, Harbin, China, 2De Montfort University, Leicester, United Kingdom |
||
Keywords: Pulse Sequence Design, MR Fingerprinting The optimal design of the Magnetic resonance fingerprinting (MRF) sequence is still challenging due to the optimization of high-degrees-of-freedom acquisition parameters. In this paper, we propose a novel unsupervised learning-based pulse sequence design framework for efficient MRF sequence optimization. Specifically, we propose a novel pulse sequence generation network (PSG-Net) that fully exploits the sequence correlation to generate the optimal pulse sequence from a zero-initialized input. To achieve improved precision of parameter estimation, we use a predefined pulse sequence performance evaluation function that can directly represent tissue quantification separability as the loss function to update the parameters of the PSG-Net. |
| 2205 | Computer 31
|
Phase-based T2 mapping using dual phase-cycled balanced SSFP imaging |
| Merijn Berendsen1, Maša Bozic-Iven1,2, Joao Tourais1, Chiara Coletti1, Ingo Hermann1,2, and Sebastian Weingärtner1 | ||
1Delft University of Technology (TU Delft), Delft, Netherlands, 2Computer Assisted Clinical Medicine, Medical Faculty Mannheim, Heidelberg University, Mannheim, Germany |
||
Keywords: Pulse Sequence Design, Pulse Sequence Design Phased-based techniques have shown promise for scan time efficient quantification of T2 relaxation times. However, previously proposed methods use an unbalanced Gradient Echo approach, which shows residual sensitivity to B1+ and T1 changes. In this work, a dual phase-cycled balanced SSFP approach is proposed. Simulations and phantom measurements show its resilience against field inhomogeneities. Excellent agreement between phase values was obtained in phantom and simulations. Initial phantom acquisitions at 3T yield T2 maps with visually high map quality and close agreement to a spin-echo reference. Future evaluation of in vivo robustness and inter- and intra-subject repeatability is warranted. |
| 2206 | Computer 32
|
Fast, High-Resolution, and Optimal Contrast MRI |
| Yao Sui1,2, Onur Afacan1,2, Camilo Jaimes1,2, Ali Gholipour1,2, and Simon K Warfield1,2 | ||
1Harvard Medical School, Boston, MA, United States, 2Boston Children's Hospital, Boston, MA, United States |
||
Keywords: Image Reconstruction, Quantitative Imaging Any MRI practices prefer high resolution, high signal-to-noise ratio, short scan time, and high contrast. Unfortunately, fast scan leads to low resolution while high-resolution scan results in a reduced signal-to-noise ratio. In particular, it is challenging for radiologists and technologists, who perform MRI scans, to find optimal sequence parameters for each patient, leading to sub-optimal contrast. We developed a new methodology that enables fast and high-resolution brain MRI with improved signal-to-noise ratio and optimal contrast between white-matter and gray-matter for each individual patient, based on quantitative imaging and reconstruction techniques. Experiments on clinical data demonstrated the advantages of our approach. |
| 2207 | Computer 33
|
An improved postprocessing method for more accurate R2* measurements in the presence of macroscopic B0 field variations |
| Chu-Yu Lee1, Olivia Lullmann2, Emily J Steinbach2, Daniel R Thedens1, Lyndsay A Harshman2, and Vincent A Magnotta1 | ||
1Department of Radiology, The University of Iowa, Iowa City, IA, United States, 2Department of Pediatrics, The University of Iowa, Iowa City, IA, United States |
||
Keywords: Data Processing, Relaxometry, R2*; Macroscopic field; Correction Macroscopic field variations result in an overestimate of R2*. One common approach to correct the bias is to utilize a three-parameter fit; however, the fitting is subject to overfitting. This study presents a new two-stage fitting of the three-parameter model by assuming that the macroscopic field variation is slowly varying in space. Using a numerical simulation and in vivo mice data, we demonstrate that the proposed method allows for a more accurate R2* measurement, and is less sensitive to noise. |
| 2208 | Computer 34
|
Sub-second T2 mapping of the whole brain via multiband SENSE multiple overlapping-echo detachment imaging and deep learning |
| Simin Li1, Taishan Kang2, Jian Wu1, Weikun Chen1, Zhigang Wu3, Jiazheng Wang3, Congbo Cai1, and Shuhui Cai1 | ||
1Department of Electronic Science, Xiamen University, Xiamen, China, 2Department of Radiology, Zhongshan Hospital of Xiamen University, School of Medicine, Xiamen University, Xiamen, China, 3MSC Clinical & Technical Solutions, Philips Healthcare, Beijing, China |
||
Keywords: Image Reconstruction, Quantitative Imaging Most quantitative magnetic resonance imaging (qMRI) methods are time-consuming. Multiple overlapping-echo detachment (MOLED) imaging can achieve quantitative parametric mapping for a single slice within hundred milliseconds. To further accelerate MOLED, we combine MOLED with multiband SENSE (MB-SENSE) technology to achieve simultaneous multi-slice T2 mapping. To solve the problem of reconstructed image quality degraded caused by a high multiband factor MB, a plug-and-play (PnP) approach with prior denoisers was applied for image restoration to realize denoising at a high MB. The proposed multiband multiple overlapping-echo detachment (MB-MOLED) imaging can achieve sub-second T2 mapping of the whole brain with a high MB. |
| 2209 | Computer 35
|
Reduction of ADC bias with deep learning-based acceleration in diffusion-weighted MRI: A phantom validation study |
| Teresa Nolte1, Masami Yoneyama2, Chiara Morsch1, Alexandra Barabasch1, Maximilian Schulze-Hagen1, Johannes M. Peeters3, Christiane Kuhl4, and Shuo Zhang5 | ||
1Diagnostic and Interventional Radiology, Uniklinik RWTH Aachen University, Aachen, Germany, 2Philips Japan, Tokyo, Japan, 3Philips Healthcare, Best, Netherlands, 4Diagnostic and Interventional Radiology, University Hospital RWTH Aachen, Aachen, Germany, 5Philips GmbH Market DACH, Hamburg, Germany |
||
Keywords: Image Reconstruction, Diffusion/other diffusion imaging techniques, Noise Further acceleration of diffusion MRI in clinical examinations is desired but challenging mainly due to low signal and associated potential bias in the quantitative apparent diffusion coefficient (ADC) values. Artificial intelligence-based denoising and image reconstruction may provide a solution to address this challenge. We investigate and compare different image reconstruction methods, including conventional parallel imaging, compressed sensing, and a deep learning-based technique, in ADC accuracy and precision using a diffusion phantom with illustration of the principle in numeric simulation. |
| 2210 | Computer 36
|
Synthetic MRI derived quantitative mapping in predicting Lymphovascular interstitial infiltration status of cervical cancer |
| Zebo Huang1, Weiqiang Dou2, and Wenwei Tang1 | ||
1Nanjing Maternity and Child Health Care Hospital, Nanjing, China, 2GE Healthcare, MR Research China, Beijing, P.R. China, Beijing, China |
||
Keywords: MR Fingerprinting/Synthetic MR, Uterus, LVSI This study aimed to investigate whether synthetic-MRI derived quantitative maps can predict lymphovascular interstitial infiltration (LVSI) status in cervical cancer. 49 patients with cervical cancer were recruited with the status of LVSI confirmed by pathology. Synthetic MRI derived T1, T2 and PD mapping were obtained for each patient. Statistical differences were shown in T1 and PD values between LVSI-positive and LVSI-negative patients. An optimal diagnostic efficacy was further shown for T1+PD with high AUC of 0.777. With these findings, it can be concluded that relaxation maps derived from synthetic MRI may be helpful for predicting LVSI status in cervical cancer. |
| 2211 | Computer 37
|
Reliability and reproducibility of whole-brain quantitative MR relaxometry using different-channel-number coils |
| Ling Sang1, Weiyin Vivian Liu2, Hu Chen3, and Wen Chen1 | ||
1Department of Radiology, Taihe Hospital, Wuhan, China, 2GE Healthcare, Beijing, China, 3Department of Radiology,Taihe Hospital, Wuhan, China |
||
Keywords: Data Acquisition, Relaxometry Brain MAGiC imaging became widely applied in multi-center cooperation and in an individual longitudinal follow-up; however, acquisition with a backup coil at an emergency situation (i.e., a sudden coil breakdown) or a coil with different numbers of channels has not been explored on a 1.5T scanner yet. In contrast to small variation of PD values in the whole brain, T2 values varies relatively dramatically especially in cerebellum and showed statistically different between scans using different coils. It is worth noting that existence of inter-coil relaxometry difference should be careful for diagnosis. |
| 2212 | Computer 38
|
Analysis of the Discretization Error vs. Estimation Time Tradeoff of MRF Dictionary Matching and the Advantage of the Neural Net-based Approach |
| Chinmay Rao1, Jakob Meineke2, Nicola Pezzotti3, Marius Staring1, Matthias van Osch1, and Mariya Doneva2 | ||
1Leiden University Medical Center, Leiden, Netherlands, 2Philips Research, Hamburg, Germany, 3Philips Research, Eindhoven, Netherlands |
||
Keywords: MR Fingerprinting/Synthetic MR, MR Fingerprinting Traditional MR fingerprinting involves matching the acquired signal evolutions against a dictionary of expected tissue fingerprints to obtain the corresponding tissue parameters. Since this dictionary is essentially a discrete representation of a physical model and the matching process amounts to brute-force search in a discretized parameter space, there arises a tradeoff between discretization error and parameter estimation time. In this work, we investigate this tradeoff and show via numerical simulation how a neural net-based approach solves it. We additionally conduct a phantom study using 1.5T and 3T data to demonstrate the consistency of neural net-based estimation with dictionary matching. |
| 2213 | Computer 39
|
Reproducing the effect of Steady-State stabilization on MR signals and images and its relation to pulse sequences by MRI simulation |
| Noriyuki Tawara1 and Daiki Tamada2 | ||
1Department of Radiological Sciences, Faculty of Health Sciences, Japan Healthcare University, Sapporo, Japan, 2Department of Radiology, Yamanashi University, Yamanashi, Japan |
||
Keywords: Visualization, Visualization The objective of this study was to reproduce the generation process of Steady-State and FLASH Band in MRI phenomena that cannot be reproduced by the actual equipment because of the restrictions imposed by venders, in order to directly confirm their relationship with pulse sequences. MRI simulation can reproduce Bloch equation faithfully, and thus it is possible to reproduce the relation between pulse sequences and the phenomena related to MRI as numerical data. |
| 2214 | Computer 41
|
The effect of MultiBand acquisition on cerebral inversion recovery intravoxel incoherent motion imaging |
| Noa van der Knaap1,2,3, Paulien H.M. Voorter1,3, Marcel J.H. Ariës1,2, and Jacobus F.A. Jansen1,3,4 | ||
1School for Mental Health & Neuroscience, Maastricht University, Maastricht, Netherlands, 2Department of Intensive Care, Maastricht University Medical Center, Maastricht, Netherlands, 3Department of Radiology and Nuclear Medicine, Maastricht University Medical Center, Maastricht, Netherlands, 4Department of Electrical Engineering, Eindhoven University of Technology, Eindhoven, Netherlands |
||
Keywords: Parallel Imaging, Parallel Imaging The MultiBand (MB) imaging technique can reduce scan time considerably, which can be especially relevant for clinical application of techniques with extensive scan protocols, such as intravoxel incoherent motion (IVIM) imaging. However, quantitative IVIM parameter estimates may be affected by the use of MB. This study is a first step to assess the comparability between IVIM acquisitions with and without MB, which has considerable implications for interpretation of IVIM results across different datasets. |
| 2215 | Computer 42
|
Assessment of neurochemistry in human-derived cerebral organoids using high-resolution magic angle spinning NMR |
| Jamie Near1,2, Vorapin Chinchalongporn3, Rajshree Ghosh Biswas4, Maggie Wu1, Martin Wilson5, Andre Simpson4, and Carol Schuurmans3 | ||
1Physical Sciences, Sunnybrook Research Institute, Toronto, ON, Canada, 2Medical Biophysics, University of Toronto, Toronto, ON, Canada, 3Biological Sciences, Sunnybrook Research Institute, Toronto, ON, Canada, 4Chemistry, University of Toronto, Toronto, ON, Canada, 5University of Birmingham, Birmingham, United Kingdom |
||
Keywords: Data Acquisition, Tissue Characterization, NMR Cerebral organoids are self-organizing three-dimensional clusters of brain tissue, derived from human pluripotent stem cells. A new and rapidly advancing technology, cerebral organoids serve as an important model system for human brain research. In this study, we used high-resolution magic angle spinning (HR-MAS) NMR to characterize the neurochemical profile of ~100-day old cerebral organoids. High-quality spectra were obtained with excellent spectral resolution. More than 17 metabolites were detected, including many resonances commonly observed in the brain in vivo (e.g. choline, creatine, glutamate, GABA, etc.). Notably, NAA was absent. Future work will assess cerebral organoids at later stages of maturity. |
| 2216 | Computer 43
|
To trigger or not to trigger – removing cardiac triggering in diffusion MRI of the cervical spinal cord saves time without sacrificing quality |
| Kurt G Schilling1, Kristin P O'Grady1, Anna J.E. Combes1, Grace Sweeney1, Logan Prock1, Julien Cohen Adad2, Bennett A Landman3, and Seth A Smith1 | ||
1Vanderbilt University Medical Center, Nashville, TN, United States, 2Polytechnique Montreal, Montreal, QC, Canada, 3Vanderbilt University, Nashville, TN, United States |
||
Keywords: Data Acquisition, Spinal Cord Cardiac triggering is commonly used in diffusion MRI protocols of the cervical spinal cord to reduce cardiac-related motion. However, this dramatically increases scan time and can limit high angular resolution multi-shell experiments. We test whether advances in preprocessing motion correction and fitting procedures may overcome cardiac-related motion artifacts. We find that removing cardiac triggering regains significant scan time with no increase in prevalence of artifacts, while providing similar quantitative indices with comparable reproducibility. In summary, removing cardiac triggering for cervical spinal cord diffusion saves time without sacrificing image quality. |
| 2217 | Computer 44
|
Vendor neutral pulse programming: Running gammaSTAR sequences on Philips Hardware |
| Martijn Nagtegaal1,2, Simon Konstandin3, Daniel Hoinkiss3, Nicolas Gross-Weege4, Volkmar Schulz3,4, Matthias Günther3,5,6, and Matthias J.P. van Osch2 | ||
1Department of Imaging Physics, Delft University of Technology, Delft, Netherlands, 2C.J. Gorter MRI Center, Radiology Department, Leiden University Medical Center, Leiden, Netherlands, 3Imaging Physics, Fraunhofer Institute for Digital Medicine MEVIS, Bremen, Germany, 4RWTH Aachen, Aachen, Germany, 5mediri GmbH, Heidelberg, Germany, 6University of Bremen, Bremen, Germany |
||
Keywords: Pulse Sequence Design, Pulse Sequence Design, Vendor independent pulse programming Pulse sequence design and development in MR is currently hindered by the possibilities to freely share pulse sequences and test these on scanners of different vendors. Vendor independent pulse programming environments provide a shareable, open, and reproducible way of pulse sequence development. An important step in this is the support on scanners of all main vendors of these protocols. In this work we show for the first time the execution of gammaSTAR imaging protocols on a Philips scanner. The obtained FLASH and RARE based images are very similar to the images obtained using the vendor’s implementation. |
| 2218 | Computer 45
|
Multi-Echo EPI performed on NexGen 7T scanner increases spatial resolution and shortens TE |
| Alexander JS Beckett1,2, An T Vu3,4, Samantha J Ma5, Essa Yacoub6, and David A Feinberg1,2 | ||
1Helen Wills Neuroscience Institute, University of California, Berkeley, CA, United States, 2Advanced MRI Technologies, Sebastopol, CA, United States, 3Radiology, University of California, San Francisco, CA, United States, 4San Francisco Veteran Affairs Health Care System, San Francisco, CA, United States, 5Siemens Medical Solutions USA, Inc., Malvern, PA, United States, 6Center for Magnetic Resonance Research, University of Minnesota, Minneapolis, MN, United States |
||
Keywords: Data Acquisition, fMRI Multi-echo (ME) EPI can increase BOLD sensitivity and reduce signal drop-out compared with standard gradient echo (GE) EPI, however spatial resolution and number of echo images are limited by the echo train length of each image. The powerful gradients on the NexGen 7T scanner allow for shorter echo spacing, and hence a greater number of echoes collected at high-resolution (1.6mm isotropic) as compared to standard 7T systems. ME-EPI collected at these resolutions can be separated into TE-dependent and TE-independent components using ME-ICA, showing promise for ME functional connectivity studies at 7T. |
| 2219 | Computer 46
|
Characterizing Artifacts in Multi-dimensional MR Fingerprinting with High Efficiency for Sequence Optimization: Systematic Error Index |
| Siyuan Hu1, Debra McGivney1, Zhilang Qiu1, and Dan Ma1 | ||
1Case Western Reserve University, Cleveland, OH, United States |
||
Keywords: Pulse Sequence Design, MR Fingerprinting It is critical to characterize the dominating systematic errors caused by undersampling and field inhomogeneity to design robust MRF scans. However, characterizing such errors by direct simulations of aliasing artifacts is computationally expensive and impractical for sequence optimization for multi-dimensional MRF (mdMRF) scans with higher dimensions. We propose the Systematic Error Index, a model to characterize systematic errors with high computational efficiency. We demonstrate accurate and robust in vivo results from the optimized MRF and mdMRF scans obtained from the proposed SEI-based optimization framework. |
| 2220 | Computer 47
|
Ex vivo lymph node staging by a portable low-field MRI scanner |
| Anke Christenhusz1, Frank F.J. Simonis1, Bennie ten Haken1, Justin C.A. te Wildt1, Sadaf Sadamzadeh1, Anneriet E. Dassen2, and Lejla Alic1 | ||
1Magnetic Detection and Imaging, University of Twente, Enschede, Netherlands, 2Surgery, Medisch Spectrum Twente, Enschede, Netherlands |
||
Keywords: Data Acquisition, Cancer, Lymph node, low-field, portable scanner Sentinel lymph node (LN) biopsy facilitated by magnetic nanoparticles (MNP) is introduced for breast cancer patients eligible for breast conserving surgery. As m1etastatic depositions potentially introduce changes in heterogeneity of MNP-enhanced MRI, a portable low-field MRI scanner can be used for ex vivo perioperative LN staging. Therefore, we assessed the changes in T1w, T2w and STIR images due to iron deposition in LN. 2Clinically relevant LN segments, such as fat and iron depositions, identified in the pathology images, were also observable in MR scans. |
| 2221 | Computer 48
|
Imaging Biomarkers of Therapeutic Response to Melanoma to Kinase Inhibitors |
| Pradeep Kumar Gupta1, Stepan Orlovskiy1, Jyoti S Tomar1, Fernando Arias-Mendoza1, David S. Nelson1, Stephen Pickup1, Alexander A. Shestov1, Dennis B. Leeper2, Jerry D Glickson1, and Kavindra Nath1 | ||
1Departments of Radiology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, United States, 2Department of Radiation Oncology, Thomas Jefferson University, Philadelphia, PA, United States |
||
Keywords: Data Acquisition, Cancer In vivo 1H/31P MRS were used to monitor the effects of Kinase Inhibitor (KI) therapy in two metabolically different human melanoma xenografts models. Our goal is to determine the metabolic changes in human melanoma xenograft models when treated with two KIs, a BRAF and a MEK inhibitor. KI combination is more effective than single-drug therapy. Differences in relative levels of metabolites and bioenergetics between two human melanoma xenografts models may produce differential therapeutic responses to BRAF and MEK inhibitors. In melanoma, metabolic changes in response to targeted kinase inhibitor therapy occur rapidly and are connected to subsequent tumor response. |
| 2222 | Computer 49
|
Merging Cartesian and non-Cartesian sampling through GoLF-SPARKLING |
| Chaithya G R1,2, Guillaume Daval-Frérot1,2,3, Aurélien Massire3, Alexandre Vignaud1, and Philippe Ciuciu1,2 | ||
1Neurospin, CEA Paris Saclay, Gif-sur-Yvette, 91191, France, 2Inria, MIND, Palaiseau, 91120, France, 3Siemens Healthineers, Saint-Denis, 93210, France |
||
Keywords: New Trajectories & Spatial Encoding Methods, Brain Cartesian sampling can acquire a given k-space region with minimum redundancy, while non-Cartesian sampling can help achieve a larger k-space coverage. Through generalized affine constraints in SPARKLING and an adapted target sampling density, for the first time non-Cartesian and Cartesian sampling are merged within same trajectory, giving the best of both worlds. With Gridding of Low Frequencies (GoLF), we get SPARKLING k-space trajectories which carry out Cartesian sampling at the center of k-space. This approach paves the way for designing new kind of compound sampling patterns, which enforces Cartesian and non-Cartesian sampling within the same trajectory. |
| 2223 | Computer 50
|
A Frequency Modulation HArd Pulse ENcoding Sequence for Ultra-Short Echo Time High-Resolution Three-Dimensional MR Imaging |
| Jun Zhao1,2, Yupeng Cao1, Weinan Tang3, Quelu Chen4, Wentao Liu1, and Dong Han1 | ||
1Key Laboratory of Biological Effects and Safety, National Center for Nanoscience and Technology, beijing, China, 2School of Future Technology, University of Chinese Academy of Sciences, Beijing, China, 3Wandong Medical Inc, Beijing, China, Beijing, China, 4Department of Radiology, Wenzhou Central Hospital, Affiliated Dingli Clinical Institute of Wenzhou Medical University, Wenzhou, China |
||
Keywords: Pulse Sequence Design, Lung, hard pulse encoding, variable TE, frequency modulation, spiral Due to its short T2* value and low proton density, MR imaging of the lung is quite challenging. Recently, ultra-short echo time (UTE) techniques, such as stack-of-spiral, were used to conduct lung MRI with good image quality. Here, based on stack-of-spiral, a frequency modulation hard pulse encoding (HAPEN) 3D UTE was proposed to optimize the echo time of different phase encoding steps in the slice direction. As a result, HAPEN can achieve shorter TE and get better SNR in human lung MRI compared to the traditional stack-of-spiral UTE sequence. |
| 2224 | Computer 51
|
Optimized, volumetric, isotropic-resolution T2W 2D FLAIR with high temporal and SNR efficiencies |
| Dahan Kim1, Tzu Cheng Chao1, Dinghui Wang1, and James G Pipe1 | ||
1Department of Radiology, Mayo Clinic, Rochester, MN, United States |
||
Keywords: New Trajectories & Spatial Encoding Methods, New Trajectories & Spatial Encoding Methods, Spiral, FLAIR, LQ We describe our novel T2W 2D FLAIR sequence whose temporal and SNR efficiencies are maximized by (1) an efficient IR acquisition that minimizes sequence deadtime and (2) localized quadratic encoding which eliminates SNR-inefficiencies of multi-pass 2D acquisitions. These improvements allowed our FLAIR scan to shorten the scan time while achieving higher SNR than standard SE and 2D-TSE scans of identical/equivalent scan time. |
| 2225 | Computer 52
|
High-level and modular description of MRI sequences using domain-specific language |
| Jörn Huber1, Daniel Christopher Hoinkiss1, Christina Plump2,3, Christoph Lüth2,3, Rolf Drechsler2,3, and Matthias Günther1,4,5 | ||
1Fraunhofer Institute for Digital Medicine MEVIS, Bremen, Germany, 2German Research Center for Artificial Intelligence DFKI, Bremen, Germany, 3University of Bremen, Bremen, Germany, 4Faculty 1 (Physics/Electrical Engineering), University of Bremen, Bremen, Germany, 5mediri GmbH, Heidelberg, Germany |
||
Keywords: Pulse Sequence Design, Software Tools This work demonstrates the formulation of MRI sequences in a high-level Domain Specific Language (DSL). The DSL approach reduces the complexity of MR sequence programming and enables optimization of preconfigured DSL parameters using MR simulations. Finally, using the gammaSTAR framework, DSL sequences can be directly run on real MR scanners. |
| 2226 | Computer 53
|
Cardiac-Gated Rosette Pulse Sequence Development for Off-Resonance Frequency Imaging |
| Julian Bertini1, Mira Liu1, Chisondi Simba Warioba1, Yu Fen Chen2, and Timothy Carroll1 | ||
1University of Chicago, Chicago, IL, United States, 2Department of Radiology, Northwestern Feinberg School of Medicine, Chicago, IL, United States |
||
Keywords: Pulse Sequence Design, Phantoms Intracranial atherosclerotic disease (ICAD) is a leading cause of preventable ischemic stroke. Long-term management of chronic ICAD would benefit from directly quantifying key risk factors, such as the oxygen extraction fraction (OEF) and cerebrovascular reserve (CVR). The accurate measurement of the phase signal is vital to measuring OEF using MRI and seeing these effects through the cardiac cycle promises to elucidate CVR. A cardiac-gated, multi-shot, multi-echo rosette pulse sequence is developed and validated with simulation and a phantom study. The proposed pulse sequence and reconstruction pipeline produces strong off-resonance frequency measurements with a scan time under two minutes. |
| 2227 | Computer 54
|
Modified non-contrast enhanced spatially-selective time-resolved vessel imaging by using cylinder-shaped pre-saturation pulse train |
| Masahiro Takizawa1, Takashi Nishihara1, and Chikako Moriwake1 | ||
1FUJIFILM Healthcare Corporation, 2-1, Shintoyofuta, Japan |
||
Keywords: Pulse Sequence Design, Lung Cylinder-shaped pre-saturation pulse train is modified to achieve non subtract scheme for non-contrast enhanced spatially-selective and time-resolved vessel imaging. The target vessel is selected by cylinder-shaped pre-saturation, and the dynamics of blood flow in the target vessel is observed by changing the number of applied pre-saturation pulses. The developed pulse train was demonstrated to visualize dynamics of a target pulmonary vessel in the lung. |
| 2228 | Computer 55
|
Reduction of artifacts and background signals in ex vivo mouse embryo MRI by potato starch suspension |
| Tomokazu Tsurugizawa1,2,3, Takuma Kumamoto4, and Yoshichika Yoshioka3,5 | ||
1Human Informatics and Interaction Research Institute, National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba, Japan, 2Faculty of Engineering, University of Tsukuba, Tsukuba, Japan, 3Center for Information and Neural Networks (CiNet), Osaka University and National Institute of Information and Communications Technology (NICT), Osaka, Japan, 4Developmental Neuroscience Project, Department of Brain & Neurosciences, Tokyo Metropolitan Institute of Medical Science, Tokyo, Japan, 5Graduate School of Frontier Biosciences, Osaka University, Osaka, Japan |
||
Keywords: Data Acquisition, Ex-Vivo Applications The high-field MRI enables to investigate the microstructure in the mouse embryo. The proton-free fluid is used for the surrounding liquid around the specimen in MR-microimaging, but the potential issue of the image quality remains due to the air bubbles on the edge of the specimen and the motion artifact. Here, we demonstrated that the potato starch suspension with phosphate-buffered saline showed a low T1 and T2 signal intensity and strongly prevent the motion of the embryo during the scanning. These results indicate the utility of potato starch suspension for MR-microimaging of mouse embryos. |
| 2229 | Computer 56
|
Design of a Two-dimensional Ultrashort Echo Time Simultaneous Multi-slice Pulse Sequence |
| Jason Andrew Reich1, Erin MacMillan2,3,4, and Rebecca Feldman1,5 | ||
1Computer Science, Mathematics, Physics and Statistics, University of British Columbia, Kelowna, BC, Canada, 2UBC MRI Research Centre, Department of Radiology, Faculty of Medicine, University of British Columbia, Vancouver, BC, Canada, 3SFU ImageTech Lab, Simon Fraser University, Surrey, BC, Canada, 4Philips Canada, Mississauga, ON, Canada, 5Biomedical Engineering and Imaging Institute, Icahn School of Medicine at Mount Sinai, New York, NY, United States |
||
Keywords: Pulse Sequence Design, New Signal Preparation Schemes Simultaneous multi-slice (SMS) pulse sequences have allowed for reductions in scan time with minimal signal-to-noise ratio loss. However, when ultrashort echo times (UTEs) are desired, SMS pulse sequences have been challenging to implement. In this work, we explore a UTE SMS pulse sequence that makes use of a power independent of number of slices prepulse to shape the transverse magnetization profile and a whole volume hard excitation to excite the remaining longitudinal magnetization. The novel pulse sequence is estimated to reduce scan times by a factor of approximately 7.6 when compared to UTE pulse sequences with three-dimensional acquisitions. |
| 2230 | Computer 57
|
Application of 3D MR MENSA in preoperative evaluation of lumbar disc herniation: a prospective study |
| Xuelin Pan1, Yuting Wen1, Zhenlin Li1, Xin Rong2, and Miaoqi Zhang3 | ||
1Department of Radiology, West China Hospital of Sichuan University, Chengdu, Sichuan, China, 2Department of Orthopedics, West China Hospital of Sichuan University, Chengdu, Sichuan, China, 3MR Research, GE Healthcare, Beijing, China |
||
Keywords: Data Acquisition, Neuro In this work, we propose a preoperative magnetic resonance examination of patients with lumbar disc herniation using the 3D MENSA sequence. This sequence was superior to cube, cube stir sequence in subjective and objective evaluation.The preoperative 3D MRI MENSA sequence is able to clearly depict the nerve roots and offer desirable contrast between the nerve roots, ligamentum flavum, bone, and intervertebral discs. Patients with lumbar degeneration can effectively benefit from the MENSA sequence since it provides informative imaging information to help understand disc herniation and compression of adjacent tissues when developing preoperative surgical strategies. |
| 2231 | Computer 58
|
R2* Estimation by Multispectral Fat-Water Models for GRE and UTE Acquisitions Using Virtual Liver Iron Overload Model and Monte Carlo Simulations |
| Prasiddhi Neupane1, Utsav Shrestha1, and Aaryani Tipirneni-Sajja1,2 | ||
1Biomedical Engineering, The University of Memphis, Memphis, TN, United States, 2St. Jude Children's Research Hospital, Memphis, TN, United States |
||
Keywords: Data Analysis, Data Analysis Multispectral fat-water-R2* models are used for the confounder-free assessment of hepatic iron overload. In this study, Monte Carlo-based virtual liver iron overload models were created, MRI signals were synthesized for GRE and UTE acquisitions, and the R2* values estimated using the monoexponential and the multispectral fat-water models were analyzed. Our results demonstrate that both multispectral models exhibit high accuracy and precision for UTE acquisition at both 1.5T and 3T. |
| 2232 | Computer 59
|
Model-Based Dynamic B0 Compensation of a 0.35T MRI-Linac using Inclinometer Data |
| Austen Curcuru1, Taeho Kim2, Deshan Yang3, and H. Michael Gach1,2,4 | ||
1Biomedical Engineering, Washington University in Saint Louis, Saint Louis, MO, United States, 2Radiation Oncology, Washington University School of Medicine, Saint Louis, MO, United States, 3Duke University, Durham, NC, United States, 4Radiology, Washington University School of Medicine, Saint Louis, MO, United States |
||
Keywords: Artifacts, MR-Guided Interventions Balanced steady state free procession (bSSFP) sequences are commonly used for real-time imaging during MRI guided radiotherapy treatments (MR-IGRT). bSSFP sequences offer high temporal resolution and SNR but are sensitive to B0 fluctuations which lead to imaging artifacts during radiotherapy (RT) gantry rotatation.1 Previous work demonstrated that RT gantry rotation induced artifacts could be significantly reduced by compensating for B0 fluctuations using data from a free induction decay (FID) navigator prior to each bSSFP image.2 A model-based approach to B0 compensation using the gantry inclinometer was evaluated to reduce the SNR loss and temporal resolution associated with the navigator approach. |
| 2233 | Computer 60
|
Effectiveness of T2-mapping imaging in assessing disease severity in Spinal muscular atrophy: preliminary study |
| Yingyi Hu1,2, Yang Huang2,3, Taiya Chen1,2, Xinguo Lu4, Diangang Fang2, Kan Deng5, and Zhiyong Li2 | ||
1China Medical University, Shenyang, China, 2Department of Radiology, Shenzhen Children's Hospital, Shenzhen, China, 3Shantou University Medical College, Shantou, China, 4Internal Medicine-Neurology, Shenzhen Children's Hospital, Shenzhen, China, 5Philips Healthcare, Guangzhou, China |
||
Keywords: Data Analysis, Quantitative Imaging, T2-mapping To evaluate the potential of T2-mapping imaging as a qMRI marker for disease severity in Spinal muscular atrophy (SMA), we compared the T2 with muscle fat fraction (MFF), and clinical assessment of 13 muscles in the pelvis and thighs of 20 patients with SMA. A significant correlation was found between the mean T2 of all muscles and the patient’s clinical evaluation and MFF. Moreover, the highest mean T2 was found in the gluteus maximus, while the lowest in the adductor longus. Therefore, T2 mapping can be used as a quantitative and objective MRI technique to assess disease severity in SMA. |
| 2352 | Computer 1
|
Pre and Post contrast Simultaneous Parametric Mapping of Glioblastomas from routine T1 weighted images for Quantitative Enhancement Assessment |
| Elisa Moya-Sáez1,2, Rodrigo de Luis-García1, Juan A. Hernández-Tamames2, and Carlos Alberola-López1 | ||
1University of Valladolid, Valladolid, Spain, 2Erasmus MC, Rotterdam, Netherlands |
||
Keywords: MR Fingerprinting/Synthetic MR, Machine Learning/Artificial Intelligence Gadolinium based contrast agents (GBCAs) have the ability to uncover blood brain barrier damage, which appears in the images as contrast enhancement caused by the leakage into the perivascular tissues. However, in clinical practice, this assessment is performed by visual comparison between the weighted images obtained before and after the GBCA injection; enhancement quantification is still an unmet need. In this work we propose a deep learning approach for the computation of pre- and post-contrast parametric maps from conventional T1 weighted images. Results show how those maps can enable an automatic quantification of the tumor enhancement. |
| 2353 | Computer 2
|
Synthetic MRI and FSE-PROPELLER duo diffusion-weighted imaging to differentiate malignant from benign head and neck tumors |
| Baohong Wen1, Yong Zhang1, Jingliang Cheng1, Tianyong Xu2, Xiaoxia Hua2, and Lizhi Xie2 | ||
1the First Affiliated Hospital of Zhengzhou University, Zhengzhou, China, 2GE Healthcare, Shanghai, China |
||
Keywords: MR Fingerprinting/Synthetic MR, Head & Neck/ENT Accurate determination of the preoperative classification of histology of head and neck tumor remains a challenge. The aim of the study was to evaluate the feasibility and capability of synthetic MRI and stimulus and fast spin echo diffusion-weighted imaging (FSE-PROPELLER duo-DWI) for the differentiation of malignant from benign head and neck tumors. The results quantitively demonstrated the feasibility and also showed that T2 value is comparable to ADC value, and the combination of T2 and ADC values could provide improved diagnostic efficacy. |
| 2354 | Computer 3
|
Magnetic resonance fingerprinting in patients with Parkinson's disease |
| Yaping Wu1, Qian Xie2, Ge Zhang1, Yan Bai1, Xipeng Yue1, Wei Wei1, Fangfang Fu1, Nan Meng1, Xianchang Zhang3, Yusong Lin4,5,6, and Meiyun Wang1 | ||
1Department of Medical Imaging, Henan Provincial People’s Hospital, Zhengzhou, China, 2School of Computer and Artificial Intelligence, Zhengzhou University, Zhengzhou, China, 3MR Collaboration, Siemens Healthineers Ltd., Beijing, China, 4School of Cyber Science and Engineering, Zhengzhou University, Zhengzhou, China, 5Collaborative Innovation Center for Internet Healthcare, Zhengzhou University, Zhengzhou, China, 6Hanwei IoT Institute, Zhengzhou University, Zhengzhou, China |
||
Keywords: MR Fingerprinting/Synthetic MR, MR Fingerprinting, Parkinson's disease This study investigated the utility of 3D high resolution magnetic resonance fingerprinting (MRF) to detect potential brain changes in Parkinson’s disease (PD). MRF T1 and T2 maps of the brains of 22 PD patients and 22 volunteers were non-linearly normalized into the MNI space to explore the group differences using whole-brain analysis. PD patients had significantly higher T1 and T2 values in the right inferior temporal gyrus anterior/posterior division, right temporal fusiform cortex anterior division, right planum polare and red nucleus. Our findings suggest that quantitative parameters acquired from 3D MRF may provide additional information for precise diagnosis of PD. |
| 2355 | Computer 4
|
Role of MULTIPLEX MRI Study in Evaluating Brain Tumors |
| Min Gao1, Jun Liu1, Liyun Zheng2, and Yongming Dai3 | ||
1Department of Radiology, The Second Xiangya Hospital of Central South University, Changsha, China, 2Shenzhen United Imaging Research Institute of Innovative Medical Equipment, Shenzhen, China, 3MR Collaboration, Central Research Institute, United Imaging Healthcare, Shanghai, China |
||
Keywords: MR Fingerprinting/Synthetic MR, MR Fingerprinting Multi-parametric MR imaging methods have been intensively developed and investigated throughout the last decade. The purpose of this study was to utilize the single-scan 3D multi-parametric MRI technique, MULTIPLEX, to characterize and grade intracranial tumors. According to the results, MULTIPLEX MRI can effectively decrease the scan time while obtaining multi-parametric images and mappings. Besides, the simultaneous quantitative estimation of multiple MR parameters can reliably characterize and grade intracranial tumors. |
| 2356 | Computer 5
|
Reducing femoral flow artefacts in radial MR Fingerprinting: A comparison of two methods applied to prostate imaging |
| Kaia Ingerdatter Sørland1, Christopher George Trimble1, Tone Frost Bathen1,2, Mattijs Elschot1,2, and Martijn A. Cloos3 | ||
1Department of Circulation and Medical Imaging, Norwegian University of Science and Technology, Trondheim, Norway, 2Department of Radiology and Nuclear Medicine, St. Olavs hospital, Trondheim University Hospital, Trondheim, Norway, 3Centre for Advanced Imaging, The University of Queensland, Brisbane, Australia |
||
Keywords: MR Fingerprinting/Synthetic MR, Artifacts High acceleration factors in radial magnetic resonance fingerprinting (MRF) of the prostate lead to strong streaking artefacts from flow in the femoral blood vessels, distorting quantitative T1 and T2 measurements. We here compare two approaches to mitigate these artefacts, namely incorporating regional saturation bands in the MRF sequence (Sat-MRF) or applying region-optimized virtual coils to suppress signal from select spatial regions before image reconstruction (ROVir-MRF). Results from seven asymptomatic volunteers show that both methods efficiently reduce signal in the region of the femoral vessels, but there are differences in retained prostate signal to noise ratio and sensitivity to T1. |
| 2357 | Computer 6
|
Correlation of quantitative synthetic mapping imaging with cervical squamous carcinoma pathology |
| Qi Zhao1, Zhihua Pan1, Mingli Jin1, Kebin Yu1, Yin Jiang1, and Miaoqi Zhang2 | ||
1The 2nd Affiliated Hospital of Chengdu Medical College Nuclear Industry 416 Hospital, Chengdu, China, 2GE Healthcare, Beijing, China |
||
Keywords: MR Fingerprinting/Synthetic MR, MR Value Clinical and pathological staging are vital for management of Cervical squamous carcinoma (CSC). In this study, using synthetic MR technique, we investigate quantitative T1, T2 and PD measurements in different clinical and pathological stage groups. Performance of those quantitative measurements in diagnosing poorly-differentiated CSC is also investigated. We find T2 values are higher in late clinical stage CSC. Results also show a correlation between quantitative measurements and the degree of pathological differentiation. ROC analysis demonstrate that T1 and T2 values can differentiate poorly-differentiated CSC from intermediate and well differentiated CSC1. |
| 2358 | Computer 7
|
A nested iteration artificial neural network approach for efficient high dimensional parameter estimation in 31P-MRF |
| Mark Widmaier1,2,3, Zirun Wang1, Song-I Lim1,2,3, Daniel Wenz1,2, and Lijing Xin1,2 | ||
1CIBM Center for Biomedical Imaging, Lausanne, Switzerland, 2Animal Imaging and Technology, Ecole Polytechnique Federale de Lausanne (EPFL), Lausanne, Switzerland, 3Laboratory of functional and metabolic imaging, Ecole Polytechniqe Federale de Lausanne (EPFL), Lausanne, Switzerland |
||
Keywords: MR Fingerprinting/Synthetic MR, Machine Learning/Artificial Intelligence, MRS, Phosporus, Creatine Kinase, MRF A new magnetization transfer 31P Magnetic Resonance Fingerprinting (31P-MRF) technique is emerging to measure the creatine kinase (CK) chemical exchange rate kCK. The inherent obstacle of the exponential growth in the size of dictionaries with the number of free parameters, was overcome by introducing the nested iteration interpolation method (NIIM). To further reduce the processing time and cope with a nonlinear behaviour, we employed an artificial neuron network (ANN), instead of an interpolation method (IPM) as in the original approach. The nested iteration ANN method (NIAN) is compared with NIIM using simulation data and in vivo 31P-MRF data. |
| 2359 | Computer 8
|
MR Vascular Fingerprinting with 3D realistic blood vessel structures and machine learning to assess oxygenation changes in human volunteers |
| Aurélien Delphin1, Thomas Coudert1, Audrey Fan2, Michael E. Moseley3, Greg Zaharchuk3, and Thomas Christen1 | ||
1Univ. Grenoble Alpes, INSERM U1216, Grenoble Institut Neurosciences, GIN, Grenoble, France, 2Biomedical Engineering, University of California Davis, Davis, CA, United States, 3Department of Radiology, Stanford University, Stanford, CA, United States |
||
Keywords: MR Fingerprinting/Synthetic MR, Oxygenation The MR vascular fingerprinting (MRvF) approach extends the concept of MR fingerprinting to the study of microvascular properties and functions. Encouraging results have been obtained in healthy human volunteers as well as in stroke and tumor models in rats. However, it has been suggested that the method has a low sensitivity to blood oxygenation measurements. We improved the MRvF approach by using simulations with 3D realistic blood vessels from animal microscopy, new fingerprint-pattern organization and machine learning tools. The method was tested in retrospective data acquired in healthy-human volunteers while breathing different gas mixtures (Hyperoxia (100%O2), Normoxia (21%O2), hypoxia (14%O2)). |
| 2360 | Computer 9
|
Latent Diffusion Models Allow Generation of Synthetic Breast MRI DCE-MIPs |
| Lukas Folle1, Lorenz Kapsner2,3, Andreas Maier1, Michael Uder2, Sabine Ohlmeyer2, and Sebastian Bickelhaupt2 | ||
1Computer Science, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany, 2Institute of Radiology, Universitätklinikum Erlangen-Nürnberg, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany, 3Universitätklinikum Erlangen-Nürnberg, Friedrich-Alexander-Universität Erlangen-Nürnberg, Medizinisches Zentrum für Informations- und Kommunikationstechnik, Erlangen, Germany |
||
Keywords: MR Fingerprinting/Synthetic MR, Breast The training of neural networks for classification or segmentation of medical images requires large amounts of training data. Sharing of these datasets is commonly difficult due to legislation and privacy constraints of medical data. In this work, we demonstrate the utility of latent diffusion models that allow the generation of synthetic samples of dynamic contrast-enhanced breast MRI-derived maximum intensity projections of subtraction series. Whilst the image quality of the generated data is high as demonstrated by a radiologist evaluation, further steps are envisioned to derive specific compounds of data, e.g., BI-RADS, FGT, or BPE classes. |
| 2361 | Computer 10
|
3D High Resolution MR Fingerprinting for prostate cancer |
| Jesus Ernesto Fajardo1, Anna Lavrova1, Vikas Gulani1, and Yun Jiang1,2 | ||
1Department of Radiology, University of Michigan, Ann Arbor, MI, United States, 2Department of Biomedical Engineering, University of Michigan, Ann Arbor, MI, United States |
||
Keywords: MR Fingerprinting/Synthetic MR, MR Fingerprinting, prostate cancer We developed a 3D MR Fingerprinting method with B0 correction for imaging the prostate gland. T1 and T2 maps with the spatial resolution of 1 x 1 x 3 mm3 were obtained from phantom and in-vivo experiments, demonstrating the potential of performing accurate and high-resolution tissue quantification of prostate cancer. |
| 2362 | Computer 11
|
Towards Prospective Free-Breathing Abdominal Magnetic Resonance Fingerprinting |
| Madison Kretzler1, Xinzhou Li2, Leonardo Kayat Bittencourt3,4, Mark Griswold1, and Rasim Boyacioglu1 | ||
1Radiology, Case Western Reserve University, Cleveland, OH, United States, 2MR R&D Collaborations, Siemens Medical Solutions USA, Inc., St. Louis, MO, United States, 3Radiology, University Hospitals Cleveland Medical Center, Cleveland, OH, United States, 4School of Medicine, Case Western Reserve University, Cleveland, OH, United States |
||
Keywords: MR Fingerprinting/Synthetic MR, Liver, prospective, Free-breathing, Pilot tone A Pilot Tone (PT) based comparison study is presented for abdominal free-breathing MRF using a previously validated PT extraction and vendor developed extraction methods. The initial steps towards prospective processing of the data for immediate availability of the free-breathing MRF quantitative maps on the console are demonstrated. In vivo comparison results for various free-breathing patterns are shown with consistent agreement between respiratory motion in PT extraction methods. Extraction of the respiration signal from PT via the vendor reconstruction software is the first step towards prospective free-breathing abdominal MRF. |
| 2363 | Computer 12
|
Repeatability and reproducibility of MRF-based Myelin Water Fraction maps of healthy human brains |
| Marta Lancione1, Matteo Cencini1, Guido Buonincontri1, Jan Kurzawski2, Joshua D Kaggie3, Tomasz Matys3, Ferdia A Gallagher3, Graziella Donatelli4,5, Paolo Cecchi4,5, Mirco Cosottini6, Nicola Martini7, Francesca Frijia7, Domenico Montanaro1, Pedro A Gómez4, Rolf F Schulte8, Alessandra Retico9, Laura Biagi1, and Michela Tosetti1 | ||
1IRCCS Stella Maris, Pisa, Italy, 2New York University, New York, NY, United States, 3University of Cambridge, Cambridge, United Kingdom, 4IMAGO7 Foundation, Pisa, Italy, 5Azienda Ospedaliero-Universitaria Pisana, Pisa, Italy, 6University of Pisa, Pisa, Italy, 7Fondazione Toscana Gabriele Monasterio, Pisa, Italy, 8GE Healthcare, Munich, Germany, 9Istituto Nazionale di Fisica Nucleare, Pisa, Italy |
||
Keywords: MR Fingerprinting/Synthetic MR, MR Fingerprinting, Myelin, Reproducibility Myelin Water Fraction (MWF) can measure White Matter (WM) myelination and integrity and can be quantified using Magnetic Resonance Fingerprinting (MRF) which allows short scan time. In this work, we assessed the repeatability and the reproducibility of MWF using 3D SSFP MRF in a traveling head study performed on healthy volunteers scanned at five different scanners from the same vendor. We computed coefficients-of-variation to estimate voxelwise variability in WM and GLM analysis to measure biases. We reported a variability of ∼3% for repeated scans at the same site and ∼5% for different sites, with an average bias of 5%. |
| 2364 | Computer 13
|
Efficient MRF at 0.55T with long readouts and concomitant field effects correction |
| Zhibo Zhu1, Nam Gyun Lee2, Ye Tian1, and Krishna S. Nayak1 | ||
1Department Of Electrical and Computer Engineering, University of Southern California, Los Angeles, CA, United States, 2Department Of Biomedical Engineering, University of Southern California, Los Angeles, CA, United States |
||
Keywords: MR Fingerprinting/Synthetic MR, Brain We demonstrate FISP-MRF at 0.55T with improved efficiency using longer spiral readouts and concomitant field effect correction. 2D axial FISP-MRF is performed with 3 different readout durations and at 2 different off-isocenter locations, reconstructed without and with MaxGIRF concomitant field correction. Spatial blurring induced by concomitant field effect was successfully mitigated. In-vivo MRF white matter T2 achieved tighter distribution with longer readouts, e.g., the coefficient of variation decreased from 9.9% to 4.9%. |
| 2365 | Computer 14
|
Inline calculation of Extracellular Volume Maps using Cardiac Magnetic Resonance Fingerprinting |
| Alexander Fyrdahl1,2, Jenny Castaings2, Rebecka Steffen Johansson1, Kelvin Chow3, Jannike Nickander1,2, Nicole Seiberlich4, and Jesse I Hamilton4 | ||
1Department of Molecular Medicine and Surgery, Karolinska Institutet, Stockholm, Sweden, 2Department of Clinical Physiology, Karolinska University Hospital, Stockholm, Sweden, 3MR R&D Collaborations, Siemens Medical Solutions USA Inc., Chicago, IL, United States, 4Department of Radiology, University of Michigan, Ann Arbor, MI, United States |
||
Keywords: MR Fingerprinting/Synthetic MR, Cardiovascular This work describes a pipeline for fully inline generation of T1, T2 and extracellular volume (ECV) maps from cardiac Magnetic Resonance Fingerprinting (MRF) data. Reconstruction of multi-parametric tissue property maps using cardiac MRF have recently been made possible directly on the scanner by replacing the need to perform Bloch equation simulations with a neural network. MRF-derived T1, T2 and ECV values were acquired and compared to MOLLI and T2p-SSFP in 5 successive clinical patients. |
| 2366 | Computer 15
|
Highly efficient T1, T2 and Diffusion-prepared radial Magnetic Resonance Fingerprinting |
| Carlos Velasco1, Carlos Castillo-Passi1,2, Nadia Chaher1, Alkystis Phinikaridou1, René M. Botnar1,2, and Claudia Prieto1,2 | ||
1School of Biomedical Engineering & Imaging Sciences, King's College London, London, United Kingdom, 2Institute for Biological and Medical Engineering, Pontificia Universidad Católica de Chile, Santiago, Chile |
||
Keywords: MR Fingerprinting/Synthetic MR, Diffusion/other diffusion imaging techniques In this study we present a golden-angle-radial efficient magnetic resonance fingerprinting approach that enables simultaneous T1, T2 and ADC quantification in a single scan of ~18s and offers the possibility to be extended to a multi-echo acquisition for additional water-fat separation estimation. A quantitative comparison between reference maps and the proposed MRF maps has shown excellent agreement, and a proof-of-concept brain MRF multiparametric T1, T2 and ADC has shown comparable image quality compared to the longer and sequential clinical reference scans. |
| 2367 | Computer 16
|
Spatio-Temporal Reconstruction Neural Network for 3D MR fingerprinting of the Prostate |
| Jae-Yoon Kim1, Jae-Hun Lee1, Dongyeob Han2, Moon Hyung Choi3, and Dong-Hyun Kim1 | ||
1Department of Electrical and Electronic Engineering, Yonsei Univ., Seoul, Korea, Republic of, 2Siemens Healthineers Ltd, Siemens Korea, Seoul, Korea, Republic of, 3Department of Radiology, Eunpyeong St. Mary's Hospital, College of Medicine,The Catholic University of Korea, Seoul, Korea, Republic of |
||
Keywords: MR Fingerprinting/Synthetic MR, Image Reconstruction, Prostate Magnetic Resonance Fingerprinting (MRF) is a technology that computes T1, T2 parameters from time-evaluated signals. However, long scanning time in obtaining fully-sampled data is a challenging point while reducing the sampling rate results in poor reconstructed data quality. Here, we propose a spatio-temporal deep learning network for reconstruction from the under-sampled MRF data. According to the retrospective reconstructed results, the proposed method could produce the T1 and T2 maps of high fidelity similar to the fully-sampled ground-truth. |
| 2368 | Computer 17
|
Characterization of age- and gender-dependent differences in intervertebral disc strain using 3D-GRASP MRI under static mechanical loading |
| Rajiv G Menon1, Hector L de Moura1, Richard Kijowski1, and Ravinder R Regatte1 | ||
1Radiology, NYU Grossman School of Medicine, New York, NY, United States |
||
Keywords: MR Fingerprinting/Synthetic MR, Tissue Characterization, Intervertebral disc The goal of this study was to assess the effect of age and gender in IVD characterization using a multiparameter MR Fingerprinting technique that can quantify T1, T2 and T1rho in a clinically feasible time. Seventeen healthy subjects were recruited. The goal of this study was to assess the effect of age and gender in IVD characterization using a multiparameter MR Fingerprinting technique that can quantify T1, T2 and T1rho in a clinically feasible time. This study suggests that the use of MR fingerprinting is a useful tool to gain insights to pathology resulting in IVD degeneration. |
| 2369 | Computer 18
|
Synthetic MRI derived relaxation mapping in evaluating cartilaginous degeneration caused by prolonged excessive exercise |
| Likai Yu1, Xiaoqing Shi1, Lishi Jie1, Yibao Wei1, Weiqiang Dou2, Shaowei Liu3, and Nongshan Zhang3 | ||
1Nanjing University of Chinese Medicine, Nanjing, China, 2GE Healthcare, MR Research China, Beijing, China, 3Jiangsu Province Hospital of Chinese Medicine, Nanjing, China |
||
Keywords: Data Analysis, Osteoarthritis The purpose of this study was to investigate whether prolonged overexertion can lead to cartilage degeneration through Synthetic MRI derived relaxation maps. In this study, 30 participants were recruited and measured with T1 and T2 mapping derived by Synthetic MRI. There was a statistically significant trend in cartilage T1 and T2 values in the long-term exercise group compared to the normal exercise group. Based on altered T1 and T2 relaxation properties of cartilage tissues, we conclude that chronic overexercise may lead to cartilage degeneration. |
| 2370 | Computer 19
|
Anatomical priors informed tractography. Evaluation on the DiSCo synthetic dataset |
| Thomas Durantel1, Julie Coloigner1, Emmanuel Caruyer1, Gabriel Girard2,3, and Olivier Commowick1 | ||
1Univ Rennes, INRIA, CNRS, INSERM, IRISA UMR 6074, Empenn ERL U-1228, F-35000, Rennes, France, 2Signal Processing Laboratory (LTS5), Ecole Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland, 3Department of Computer Science, Université de Sherbrooke, Sherbrooke, QC, Canada |
||
Keywords: Data Processing, Tractography & Fibre Modelling White matter fiber tracking using diffusion magnetic resonance imaging (DWI) provides a powerful approach to map brain connections, but they are not completely reliable both in academic and clinical context. This is because most methods infer global connectivity from local information,leading to a high rate of false positive. We propose a method to include anatomical prior in the tractography algorithm. We show on synthetic DiSCo data that including this prior on two state-of-the-art tractography algorithms could improve the overall shape of the tractograms. |
| 2371 | Computer 20
|
Synthetic MRI in differentiating benign from metastatic retropharyngeal lymph nodes: combination with diffusion-weighted imaging |
| Peng Wang1, Shudong Hu2, Weiqiang Dou3, Jie Shi3, and Heng Zhang1 | ||
1Affiliated hospital of Jiangnan university, Wuxi, China, 2ffiliated hospital of Jiangnan university, Wuxi, China, 3GE Healthcare, MR Research China, Beijing, China |
||
Keywords: Head & Neck/ENT, Quantitative Imaging This study aimed to evaluate MAGiC imaging (one synthetic MRI technique, syMRI) and its combination with diffusion-weighted imaging (DWI) in discriminating benign from metastatic retropharyngeal lymph nodes (RLNs). With MAGiC derived relaxation parameters, 58 patients with 21 benign and 42 metastatic RLNs were measured. The resultant T1, T2, PD and T1SD values showed significant different values between benign and metastatic RLNs with an optimal diagnostic performance from T1SD. Moreover, the combination of MAGiC, DWI, and morphological features demonstrated a significantly improved performance on overall diagnosis. |
| 2372 | Computer 21
|
B1+ inhomogeneity mitigation in slice-selective excitation at 7T with a single-channel RF coil augmented by ΔB0 shim array fields |
| Molin Zhang1, Georgy Guryev1, Nick Arango1, Jason Stockmann2,3, Jacob White1, and Elfar Adalsteinsson1,4,5 | ||
1EECS, MIT, Cambridge, MA, United States, 2Athinoula A. Martinos Center for Biomedical Imaging, Charlestown, MA, United States, 3Department of Radiology, Harvard Medical School, Charlestown, MA, United States, 4Harvard-MIT Health Sciences and Technology, MIT, Cambridge, MA, United States, 55 Institute for Medical Engineering and Science, MIT, Cambridge, MA, United States |
||
Keywords: Pulse Sequence Design, RF Pulse Design & Fields We explored the utility of shim array fields in the mitigation of in-plane B1+ inhomogeneity for a slice-selective excitation at 7T and achieved better magnitude of transverse magnetization compared with birdcage transmit using conventional single-channel RF excitation. Approximately 2:1 range of B1+ was mitigated, at the cost of increased peak RF power, shim array current demands up to 20 A, and minor out-of-slice sidelobes. |
| 2373 | Computer 22
|
Echo-Volume Imaging with Restricted Field-of-View and (k, t)-Space Undersampling: A Fast fMRI Acquisition Technique |
| Qingfei Luo1, Kaibao Sun1, Guangyu Dan1,2, Muge Karaman1,2, and Xiaohong Joe Zhou1,2,3 | ||
1Center for Magnetic Resonance Research, University of Illinois at Chicago, Chicago, IL, United States, 2Department of Biomedical Engineering, University of Illinois at Chicago, Chicago, IL, United States, 3Departments of Radiology and Neurosurgery, University of Illinois at Chicago, Chicago, IL, United States |
||
Keywords: Pulse Sequence Design, fMRI Echo-volume imaging (EVI) can offer higher acquisition speed than echo-planar imaging (EPI) but is more sensitive to image distortion and blurring. In this study, we develop an EVI-based fast fMRI acquisition technique by employing three-dimension restricted field-of-view imaging (k-t rFOV-EVI) and (k, t)-space undersampling. Our human fMRI experiments covering the visual cortex demonstrate that k-t rFOV-EVI can provide higher image quality and fMRI detection sensitivity than the simultaneous multi-slice EPI at 2.5-mm-isotropic spatial resolution with a temporal resolution of 240 ms. |
| 2374 | Computer 23
|
Asymmetric echo variable flip angle RARE for n-echo Dixon based PDFF @ 9.4T |
| Wan-Ting Zhao1, Karl-Heinz Herrmann1, Renat Sibgatulin1, Janine Arlt2, Weiwei Wei2, Uta Dahmen2, and Jürgen R. Reichenbach1 | ||
1Medical Physics Group, Institute of Diagnostic Radiology, University hospital Jena, Jena, Germany, 2Experimental Transplantation Surgery, Department of General, Visceral and Vascular Surgery, University hospital Jena, Jena, Germany |
||
Keywords: Pulse Sequence Design, Preclinical This work demonstrates the use of echo-shifting in combination with an efficient variable flip angle 3D RARE to reconstruct fat fraction distribution in a liver sample, resolving the fine structure of the fat deposits in the tissue with a Dixon based analysis at 9.4T. |
| 2375 | Computer 24
|
Tailored multi-dimensional partial saturation pulses for inner/outer-volume spoiled steady-state imaging |
| Jon-Fredrik Nielsen1 and Kawin Setsompop2,3 | ||
1fMRI Laboratory and Biomedical Engineering, University Of Michigan, Ann Arbor, MI, United States, 2Radiology, Stanford University, Palo Alto, CA, United States, 3Electrical Engineering, Stanford University, Palo Alto, CA, United States |
||
Keywords: Pulse Sequence Design, RF Pulse Design & Fields We propose to use short (~1–1.5ms) multi-dimensional spatially selective RF pulses that saturate unwanted signal regions in spoiled gradient echo (SPGR/FLASH/T1-FFE) imaging. These pulses can be inserted at regular intervals in the sequence, e.g., immediately before each imaging excitation pulse. By reducing signal contamination from outside the region of interest (ROI), these pulses should enable more rapid and robust T1- and T2*-weighted imaging in many body and neuro applications. |
| 2376 | Computer 25
|
Tracking Abdominal Organ Motion in Real-time T2-weighted MRI with 2 Simultaneously Acquired Orthogonal Slices |
| Samantha Hickey1, Andreas Reichert1, Lars Bielak1,2, Wolfgang Ptacek3, Thomas Bortfeld4, and Michael Bock1 | ||
1Division of Medical Physics, Dept. of Diagnostic and Interventional Radiology, University Medical Center Freiburg, Faculty of Medicine, University of Freiburg, Freiburg, Germany, 2German Cancer Consortium (DKTK), Partner Site Freiburg, Freiburg, Germany, 3ACMIT Gmbh, Wiener Neustadt, Austria, 4Division of Radiation Biophysics, Department of Radiation Oncology, Massachusetts General Hospital and Harvard Medical School, Boston, MA, United States |
||
Keywords: Pulse Sequence Design, Radiotherapy, Interventional, T2-weighted Radiotherapy, cryoablation, and percutaneous needle biopsy procedures profit from MR-guidance with real-time lesion tracking. T2-weighted images often provide superior lesion contrast, but the temporal resolution of conventional T2-weighted sequences is insufficient for real-time MRI. Here, we present a fast, T2-weighted sequence Ortho-SSFP-Echo, which simultaneously acquires two orthogonal images, to track breathing induced kidney motion. |
| 2377 | Computer 26
|
Fast Magnetization-Prepared Gradient Echo Sequences for 3D-T1rho Mapping of the Knee Joint Using Optimized Variable Flip-Angles |
| Marcelo V. W. Zibetti1, Hector L. De Moura1, Mahesh B. Keerthivasan2, and Ravinder R. Regatte1 | ||
1Radiology, NYU Grossman School of Medicine, New York, NY, United States, 2Siemens Medical Solutions, Malvern, PA, United States |
||
Keywords: Pulse Sequence Design, Quantitative Imaging We proposed an efficient magnetization-prepared gradient echo (MP-GRE) sequence that uses optimized variable flip-angles (OVFA) to reduce acquisition time by 4x while increasing SNR when compared to magnetization-prepared angle-modulated partitioned k-space spoiled GRE snapshots (MAPSS), typically used for T1rho mapping. The proposed OVFA based sequence can improve the spatial resolution of T1rho mapping by 4x, with nearly same SNR and scan time as MAPSS. |
| 2378 | Computer 27
|
WHIRLED PEAS: Analytical Equations for Spiral Trajectories and Matching Gradient Waveforms |
| James G Pipe1 | ||
1Radiology, Mayo Clinic, Rochester, MN, United States |
||
Keywords: Pulse Sequence Design, New Trajectories & Spatial Encoding Methods, Spiral, WHIRL, analytical This work introduces exact time-dependent analytical solutions, for both gradient waveforms and k-space trajectories, for an optimized spiral-based imaging strategy based on maximum frequency, perpendicular slew rate, and gradient magnitude constraints. |
| 2379 | Computer 28
|
View ordering in PROPELLER for MSK applications |
| Henric Rydén1, Adam van Niekerk1, Matea Borbas2, and Mikael Skorpil3 | ||
1Clinical Neuroscience, Karolinska Institutet, Stockholm, Sweden, 2Karolinska University Hospital, Stockholm, Sweden, 3Karolinska Institutet, Stockholm, Sweden |
||
Keywords: Pulse Sequence Design, Data Acquisition, FSE, RARE, PROPELLER A novel view ordering scheme for RARE (FSE/PROPELLER) imaging is proposed and compared with the vendor solution, particularly suited for MSK applications where intermediate TE is desired. |
| 2380 | Computer 29
|
DeepRF for simulatneous multislice pulses |
| Jiye Kim1, Hongjun An1, Chungseok Oh1, Berkin Bilgic2,3, and Jongho Lee1 | ||
1Department of Electrical and Computer Engineering, Seoul National University, Seoul, Korea, Republic of, 2Department of Radiology, Havard Medical School, Boston, MA, United States, 3Athinoula A. Martinos Center for Biomedical Imaging, Charlestown, MA, United States |
||
Keywords: Pulse Sequence Design, Machine Learning/Artificial Intelligence DeepRF1 is a recently proposed RF pulse design method2-5 using deep reinforcement learning and optimization, generating an RF pulse defined by a reward (e.g., slice profile and energy constraint) from self-learning. Here, we proposed an improved algorithm for DeepRF that incorporates a modulation function to design an simultaneous multislice6 RF pulse. The new algorithm is tested and compared with the original multiband9 pulses, reporting reduced RF energy while preserving the characteristics of the original slice profile. Additionally, a multiPINS8 like inversion pulse is designed to demonstrate the usefulness of DeepRF for a non-constant slice selective gradient.
|
| 2381 | Computer 30
|
Variable-Density Velocity-Selective Preparation for Non-Contrast-Enhnaced MR Angiography |
| Minyoung Kim1,2, Jaeseok Park3,4, Seunghong Choi5, and Taehoon Shin1,2 | ||
1Department of Mechanical and Biomedical Engineering, Ewha Womans University, Seoul, Korea, Republic of, 2Graduate Program in Smart Factory, Ewha Womans University, Seoul, Korea, Republic of, 3Department of Biomedical Engineering, Sungkyunkwan University, Suwon, Korea, Republic of, 4Department of Intelligent Precision Healthcare Convergence, Sungkyunkwan University, Suwon, Korea, Republic of, 5Department of Radiology, Seoul National University Hospital, Seoul National University College of Medicine, Seoul, Korea, Republic of |
||
Keywords: Pulse Sequence Design, Pulse Sequence Design, Variable Density Sampling Excitation Velocity-selective (VS) magnetization preparation has shown great promise for non-contrast-enhanced MR angiography. Under the excitation k-space formalism, VS preparation pulse allocated RF weights to k-space at uniform intervals which leads to the aliased excitation at the inverse of the k-space interval (termed velocity FOV). In this study, we proposed a new version of VS preparation pulse with a diffused pattern of aliased saturation. The initial in-vivo test of the new version of VS-MRA shows a reduced effect of the aliased saturation compared with conventional VS-MRA. |
| 2382 | Computer 31
|
Optimized Phase Cycling for Coherence Pathway Selection in Pi Echo Planar Imaging |
| Mark Armstrong1 and Dan Xiao1 | ||
1University of Windsor, Windsor, ON, Canada |
||
Keywords: Pulse Sequence Design, New Trajectories & Spatial Encoding Methods Pi Echo Planar Imaging (PEPI) has a significantly reduced gradient duty cycle compared to FSE. It has a great potential for low field MRI where concomitant fields are significant. However, its application is limited due to the requirement of a near perfect refocusing pulse to avoid coherence pathway artifacts. In this work an optimized phase cycling scheme is proposed which minimizes coherence pathway contributions to the signal. This method has been validated in simulations and experiments showing a reduced sensitivity to flip angle compared to the conventional XY-16 method. Artifact-free 2D PEPI was achieved. |
| 2383 | Computer 32
|
Double-echo phase-incrementing SSel-MQC (pi-SSelMQC) in biomarker imaging with full signal recovery and excellent lipid and water suppression |
| Qiuhong He1, Hong Yuan2,3, and Yen-Yu Ian Shih2,4,5 | ||
1The School of Health Sciences, Purdue University, West Lafayette, IN, United States, 2Biomedical Research Imaging Center, University of North Carolina at Chapel Hill, Chapel Hill, NC, United States, 3Radiology, University of North Carolina at Chapel Hill, Chapel Hill, NC, United States, 4Neurology, University of North Carolina at Chapel Hill, Chapel Hill, NC, United States, 5Center for Animal MRI, University of North Carolina at Chapel Hill, Chapel Hill, NC, United States |
||
Keywords: Pulse Sequence Design, Spectroscopy A double-echo pi-SSelMQC method was developed to recover 100% biomarker signals, as compared to spin-echo pulse sequences, with excellent water and lipid suppression. Signals from both ZQ→DQ and DQ→ZQ coherence transfer pathways were detected. By synchronization of phase-encoding steps and RF phase increments of either MQ-preparation or MQ-transfer pulse, the biomarker images from different MQ-pathways were resolved and shifted away from residual lipid and water signals. Full lactate and polyunsaturated fatty acids (PUFA) signals were recovered in pi-SSelMQC imaging using yogurt phantoms, vegetable oil, and murine 344SQ lung tumors grown subcutaneously on the right thigh of syngeneic 129X1/SvJ male mice. |
| 2384 | Computer 33
|
Quadratic phase-modulated xSPEN (QxSPEN): A new route to high resolution single-shot imaging |
| Ke Dai1,2, Zhiyong Zhang2, and Lucio Frydman1,3 | ||
1Department of Chemical and Biological Physics, Weizmann Institute of Science, Rehovot, Israel, 2School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China, 3Azrieli National Center for Brain Imaging, Weizmann Institute of Science, Rehovot, Israel |
||
Keywords: Pulse Sequence Design, New Trajectories & Spatial Encoding Methods Single-shot cross-term SPatiotemporal ENcoding (xSPEN) is a single-shot approach to MRI with exceptional resilience to field inhomogeneities. xSPEN’s non-Fourier nature and sinc-like point-spread function demands SNR/resolution compromises which, so far, no post-processing has managed to solve. This study shows that introducing a quadratic phase modulation in conjunction with the hyperbolic phase modulation demanded by xSPEN can solve this, enabling the use of deconvolution principles providing resolution enhancement. The principles and examples of the ensuing quadratic xSPEN (QxSPEN) experiment, are presented on synthetic, phantom and human brain single-shot data. |
| 2385 | Computer 34
|
A new pulse sequence to selectively probe the signals of γ-aminobutyric acid |
| Xue Yang1, Ying Liu2, Caixia Fu3, He Wang2, Ying-Hua Chu4, Da-Xiu Wei1, and Ye-Feng Yao1 | ||
1East China Normal University, Shanghai, China, 2Fudan University, Shanghai, China, 3Siemens Shenzhen Magnetic Resonance Ltd., Shanghai, China, 4MR Collaboration, Siemens Healthineers Ltd., Shanghai, China |
||
Keywords: Pulse Sequence Design, Spectroscopy A new pulse sequence was developed to selectively probe the signals of γ-aminobutyric acid (GABA). Different with the previous pulse sequences, the signal selectivity of this pulse sequence is achieved by preparation and reconversion of the 1H spin singlet order (SSO) of GABA. The optimal control method was used in the design of the pulse sequence. By using the developed pulse sequence, the 1H signals of GABA in human brains were selectively probed. |
| 2386 | Computer 35
|
Off-resonance and B1+ Resilience of Relaxation Along a Fictitious Field (RAFF) pulses |
| Roeland Christiaan Naaktgeboren1, Chiara Coletti2, Christal van de Steeg-Henzen3, and Sebastian Weingärtner2 | ||
1Department of Imaging Physics, Delft University of Technology, Delft, Netherlands, 2Delft University of Technology, Delft, Netherlands, 3HollandPTC, Delft, Netherlands |
||
Keywords: Pulse Sequence Design, Cartilage, RAFF Spin-lock relaxation times can provide valuable markers for pathological remodelling, but their acquisition with constant amplitude spin-lock pulses is limited by SAR and sensitive to field inhomogeneities. Relaxation along a Fictitious Field (RAFF) can be used to measure spin-lock relaxation with reduced SAR burden. In this work, we evaluate the resilience of RAFF against field B0 and B+1 field inhomogeneities. Yet another RAFF (yaRAFF) pulse is introduced for fictitious field spin-locking with increased effective field strength. yaRAFF shows >5.9% reduced susceptibility for field inhomogeneities compared with RAFF in simulations and phantom measurements. In vivo images of yaRAFF in the calf show higher precision for small off-resonances, the mapping quality for large off-resonance is maintained for in phantom and in vivo TRAFF2 times at 3T. |
| 2387 | Computer 36
|
Flow-suppressed 2D Spin-Echo EPI with high B1-insensitivity using Hyperbolic Secant Pulses |
| Jae-Youn Keum1, Jeong Hee Yoon2, Michael Garwood3, and Jang-Yeon Park1,4 | ||
1Department of Biomedical Engineering, Sungkyunkwan University, Suwon, Korea, Republic of, 2Department of Radiology, Seoul National University Hospital and College of Medicine, Seoul, Korea, Republic of, 3Department of Radiology, Center for Magnetic Resonance Research, Minneapolis, MN, United States, 4Department of Intelligent Precision Healthcare Convergence, Sungkyunkwan University, Suwon, Korea, Republic of |
||
Keywords: Pulse Sequence Design, Diffusion/other diffusion imaging techniques, Hyperbolic-Secant Pulse Blood flow artifacts sometimes occur in 2D spin-echo sequence despite its intrinsic flow-suppression effect. Pre-saturation pulses have widely been used for flow-suppression, but it has several disadvantages such as requirement of additional RF pulses and being not effective at suppressing relatively slow blood flow signal. In this study, we propose a more effective flow-suppressed spin-echo sequence using hyperbolic-secant (HS) pulses for π/2-excitation and π-refocusing. The proposed method was applied to 2D spin-echo diffusion EPI for liver imaging. |
| 2388 | Computer 37
|
Diffusion-Weighted Half-Fourier Acquisition Single-Shot Turbo Spin-Echo (HASTE) Imaging with improved sensitivity |
| Aidin Arbabi1, Vitaliy Khlebnikov1, Jose P. Marques1, and David. G. Norris1 | ||
1Radboud University, Nijmegen, Netherlands |
||
Keywords: Pulse Sequence Design, Brain Half-Fourier acquisition single-shot turbo spin-echo diffusion-weighted is a clinically established magnetic resonance imaging tool for the detection of small lesions, particularly cholesteatoma. However, in the standard approach, half of the available signal is sacrificed through displacing one echo parity out of the acquisition window to fulfill the Carr-Purcell-Meiboom-Gill condition. We present a selective parity approach to tackle this problem. The proposed method features a near full sensitivity, a low specific absorption rate for long echo trains, and more than two-fold increase in signal-to-noise ratio, compared to the manufacturer's method under the same conditions. |
| 2389 | Computer 38
|
FGPA based accelerator for pMRI using coil compression |
| Tayaba Gul1, Omiar Inam2, Abdul Basit2, Nimra Naeem3, and Hammad Omer2 | ||
1Electrical and Computer Engineering Department, Comsats University Islamabd, Islamabad, Pakistan, 2Electrical and Computer Engineering Department, Comsats University Islamabad, Islamabad, Pakistan, 3Comsats University Islamabad, Islamabad, Pakistan |
||
Keywords: Parallel Imaging, Cardiovascular Parallel MRI (e.g., SENSE, GRAPPA) accelerates the data acquisition in modern clinical scanners using multiple receiver coils. However, it results in more computational demands on general purpose computers. Coil compression is a promising way to address the computational cost and memory requirements associated with a large number of receiver coils. In this work, a novel FPGA based hardware accelerator is designed to perform coil compression using QR decomposition. In-vivo reconstruction results from 30 coil cardiac data set, show that the proposed accelerator elevates the speed and memory constraints while preserving the image quality. |
| 2390 | Computer 39
|
Explainable machine learning for microstructural imaging of neonatal brain |
| Yihan Wu1,2, Hamza Kebiri1,3, Ali Gholipour1, and Davood Karimi1 | ||
1Harvard Medical School & Boston Children's Hospital, Boston, MA, United States, 2Zhejiang University, Hangzhou, China, 3Center for Biomedical Imaging & Lausanne University Hospital, Lausanne, Switzerland |
||
Keywords: Machine Learning/Artificial Intelligence, Diffusion/other diffusion imaging techniques, NODDI Deep learning has a great potential for estimating brain tissue micro-structure from diffusion MRI measurements. However, it is hard to understand and interpret how these models work. Therefore, until now these deep learning models have been designed using ad-hoc approaches. In this work, we propose a method for determining the contribution of different measurements to the prediction produced by these deep learning models. We apply this method for estimating the parameters of the NODDI model for the neonatal brain. Results show that this method is highly effective in determining the subsets of the measurements that result in lower estimation error. |
| 2391 | Computer 41
|
ScanHub: Open-Source Platform for MR Scanner Control, Acquisitions and Postprocessing |
| David Schote1,2, Johannes Behrens2, Lukas Winter1, Christoph Kolbitsch1, and Christoph Dinh2 | ||
1Physikalisch-Technische Bundesanstalt (PTB), Braunschweig and Berlin, Germany, 2Brain-Link UG (haftungsbeschränkt), Landau i.d. Pfalz, Germany |
||
Keywords: Software Tools, Software Tools ScanHub (https://github.com/brain-link/scanhub-ui) is an open, generic solution for cloud-based medical data acquisition and processing. Functionalities are subdivided into microservices supporting use cases in the clinical as well as in the research context. The platform capabilities are demonstrated with an exemplary MRI workflow. As a proof of principle MR data acquisition was simulated with an open-source Bloch solver. A reconstruction of the simulated raw data is provided by a microservice. The whole acquisition process is controlled via a web-based UI; from deploying a pulse sequence to the organization and visualization of reconstructed and DICOMized results. |
| 2392 | Computer 42
|
NiGSP: Graph Signal Processing on multimodal MRI data. |
| Stefano Moia1,2, Julia Brügger1,3, Philip Egger1,3, Giorgia Giulia Evangelista1,3, Friedhelm Cristoph Hummel1,3,4, Maria Giulia Preti1,2, and Dimitri Van De Ville1,2 | ||
1Neuro-X Institute, Ecole polytechnique fédérale de Lausanne, Geneva, Switzerland, 2Department of Radiology and Medical Informatics (DRIM), Faculty of Medicine, University of Geneva, Geneva, Switzerland, 3Neuro-X Institute, EPFL Valais, Clinique Romande de Réadaptation, Sion, Switzerland, 4Department of Clinical Neurosciences, Geneva University Hospital (HUG), Geneva, Switzerland |
||
Keywords: Software Tools, Brain Connectivity NiGSP is a python-based toolbox aimed at facilitating the adoption of graph signal processing with an emphasis on multimodal brain imaging data. We present its standard workflow, that allows basic filtering operations and metrics computations, and we introduce a novel application to cerebral stroke consisting in the creation of a subject-specific anatomical lesion-based filter to be applied on functional MRI timeseries. |
| 2393 | Computer 43
|
CMRsim - A Python package for MRI simulations incorporating complex organ motion and flow |
| Jonathan Weine1, Charles McGrath1, and Sebastian Kozerke1 | ||
1University and ETH Zurich, Zurich, Switzerland |
||
Keywords: Software Tools, Simulations, Cardiovascular We present CMRsim, a new Python package for MR simulations efficiently incorporating complex organ motion and flow. The MR signal can be calculated using both Bloch simulations as well as analytical signal models. The package leverages TensorFlow2 for GPU acceleration and thereby facilitates fast simulations while also providing a compilation-free framework for prototyping and community contributions. Significant effort has been put on maintainability and software engineering best practices. The API reference as well as information on installation and how to get started is publicly available at: https://people.ee.ethz.ch/~jweine/cmrsim/latest/index.html Keywords: Open-source, Simulation, Cardiovascular, Motion, Reproducible science, Digital twin |
| 2394 | Computer 44
|
Open-source tools for multi-center multi-platform 0.55T relaxometry studies |
| Bilal Tasdelen1, Rajiv Ramasawmy2, Kathryn E Keenan3, Adrienne Campbell-Washburn2, and Krishna S Nayak1 | ||
1Ming Hsieh Department of Electrical and Computer Engineering, University of Southern California, Los Angeles, CA, United States, 2Cardiovascular Branch, Division of Intramural Research, National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, MD, United States, 3NIST: National Institute of Standards and Technology, Boulder, CO, United States |
||
Keywords: Software Tools, Software Tools, open-source We demonstrate quantitative T1 and T2 mapping tools, built from open-source components Pulseq and qMRLab, suitable for multi-center multi-platform studies. The end-to-end solution includes acquisition, reconstruction, ROI analysis, and repeatability testing. The tools are demonstrated in the context of 0.55T relaxometry at two different sites and three scanners with two different hardware and software specifications. Relaxometry results are validated against results derived from vendor-provided sequences. |
| 2395 | Computer 45
|
Virtual Scanner Games |
| Gehua Tong1, Rishi Ananth2, Jason Stockmann3, Vlad Negnevitsky4, Benjamin Menküc5, Akbar Alipour6, John Thomas Vaughan, Jr.1,7, Sairam Geethanath7,8, Gaurav Verma9, and Gaurav Verma9 | ||
1Biomedical Engineering, Columbia University, New York, NY, United States, 2College of Arts and Sciences, University of Washington, Seattle, WA, United States, 3A.A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Charlestown, MA, United States, 4Oxford Ionics Ltd, Oxford, United Kingdom, 5University of Applied Sciences and Arts Dortmund, Dortmund, Germany, 6BioMedical Engineering and Imaging Institute, Dept. of Diagnostic, Molecular and Interventional Radiology, Icahn School of Medicine at Mt. Sinai, New York, NY, United States, 7Columbia Magnetic Resonance Research Center, Columbia University, New York, NY, United States, 8Accessible MR Laboratory, BioMedical Engineering and Imaging Institute, Dept. of Diagnostic, Molecular and Interventional Radiology, Icahn School of Medicine at Mt. Sinai, New York, NY, United States, 9Radiology, Icahn School of Medicine at Mount Sinai, New York, NY, United States |
||
Keywords: Software Tools, Simulations, Education Access to MRI in low-resource regions is often limited by the lack of local expertise required to sustain long-term MR operation. To help develop interest and understanding in a wider audience from a younger age, we developed Virtual Scanner Games. This web application can be distributed by USBs to run locally and provides eight educational interactive games to introduce MR fundamentals to students at the high school level or higher. |
| 2396 | Computer 46
|
SimMRI – A web-based MR Image Simulator for easy accessible MRI teaching |
| Christian Tönnes1, Christian Licht1, Lothar R. Schad1, and Frank G. Zöllner1 | ||
1Chair for Computer Assisted Clinical Medicine, MIiSM, Medical Faculty Mannheim, Heidelberg University, Mannheim, Germany |
||
Keywords: Software Tools, Simulations, Education SimMRI is a Web-Based MRI simulator developed for teaching students. It allows the simulation of multiple sequences for 1H and 23Na imaging on different brain datasets. The software enables users to customize parameters for each sequence, add noise, or change the voxel count for every dimension of the image. Additionally, the software provides compressed sensing reconstruction with three k-space subsampling schemes. The computation is solely performed on the client and therefore well suited for large student groups. The code, written in JavaScript, HTML, CSS, and c compiled to WebASM, is open source and supports easy inclusion of new sequences. |
| 2397 | Computer 47
|
Magnetic Resonance Image Processing and Analysis Platform (MRI-PAP): A Windows-based Automated Pipeline for Multiple Sclerosis |
| Alia Khaled1,2, Ahmed Bayoumi1, Joseph Thomas1, Refaat Gabr3, Khader Hasan3, Jerry Wolinsky1, and John Lincoln1 | ||
1Department of Neurology, University of Texas Health Sciences Center, Houston, TX, United States, 2Department of Cancer Systems Imaging, University of Texas MD Anderson Cancer Center, Houston, TX, United States, 3Department of Diagnostic and Interventional Imaging, University of Texas Health Sciences Center, Houston, TX, United States |
||
Keywords: Software Tools, Data Processing, Pipeline Neurologists and researchers rely on several software tools for neuroimaging data analysis. Limitations in these tools often necessitate the use of more than one program, spanning different coding languages and operating systems, which requires time, effort, and programming skills. MRI-PAP is a collection of pipelines with an easy-to-use graphical user interface (GUI) designed for MRI processing and analysis of multiple sclerosis patient images. These automated pipelines accept DICOM images as input and take them through a series of processing steps. MRI-PAP pipelines save time by accessing different packages from a single GUI and can be used without programming skills. |
| 2398 | Computer 48
|
CMRSeq - A Python package for intuitive sequence design |
| Jonathan Weine1, Charles McGrath1, and Sebastian Kozerke1 | ||
1University and ETH Zurich, Zurich, Switzerland |
||
Keywords: Software Tools, Pulse Sequence Design We present CMRseq, a new python package for MR sequence definitions. The package builds upon successful concepts of previous frameworks, while leveraging Python functionality to increase usability and maintainability. CMRseq features physical unit checks, adheres to coding style conventions and includes unit-test coverage and continuous integration infrastructure for documentation. Import and export functionality to facilitate compatibility with popular formats is also included. The API reference as well as information on installation and how to get started is publicly available at: https://people.ee.ethz.ch/~jweine/cmrseq/latest/index.html Keywords: Open source, reproducible science, sequence design, software tools, visualization |
| 2399 | Computer 49
|
MRI simulation of moving objects based on the Lagrange description of Bloch equations |
| Katsumi Kose1, Ryoichi Kose1, Daiki Tamada2, and Utaroh Motosugi3 | ||
1MRIsimulations Inc., Tokyo, Japan, 2University of Yamanashi, Chuo, Japan, 3Kofu Kyoritsu Hospital, Kofu, Japan |
||
Keywords: Software Tools, Motion Correction, motion simulation To investigate the possibility and limitation of MRI simulations based on the Lagrange description, MRI experiments and simulations of moving objects were performed. As a result, we were able to obtain simulated MR images that almost reproduced the experimental results within practical computation time. However, it was found that the T2 coherence effect should be reduced for moving objects. It was also found that more precise simulation is necessary to reproduce the detailed motion artifacts. |
| 2400 | Computer 50
|
A toolbox of chi-separation for magnetic susceptibility source separation |
| Jun-Hyeok Lee1, Hyeong-Geol Shin2,3, Minjun Kim2, Jongho Lee2, and Se-Hong Oh1 | ||
1Biomedical Engineering, Hankuk University of Foreign Studies, Yongin, Korea, Republic of, 2Electrical and Computer Engineering, Seoul National University, Seoul, Korea, Republic of, 3Johns Hopkins University School of Medicine & Kennedy Krieger Institute, Baltimore, MD, United States |
||
Keywords: Software Tools, Quantitative Susceptibility mapping, Chi-separation The susceptibility maps generated by QSM and $$$\chi$$$-separation show magnetic susceptibility distributions, which have been proposed as important biomarkers for brain disorders. However, susceptibility mapping requires a complicated multi-step procedure that is difficult for inexperienced researchers. There is a need to design a convenient application that can easily conduct susceptibility mapping. In this work, we developed a MATLAB based GUI software, called the ‘$$$\chi$$$-separation toolbox’ that generates high-quality susceptibility maps in just a few clicks. The GUI of our toolbox is intuitive and user-friendly that it helps researchers to conduct QSM and $$$\chi$$$-separation processing without difficulty. |
| 2401 | Computer 51
|
Automated High Order Shim for Neuroimaging Studies |
| Jia Xu1, Baolian Yang2, Douglas Kelley2, and Vincent A. Magnotta1,3,4 | ||
1Radiology, University of Iowa, Iowa City, IA, United States, 2GE Healthcare, Waukesha, WI, United States, 3Psychiatry, University of Iowa, Iowa City, IA, United States, 4Biomedical Engineering, University of Iowa, Iowa City, IA, United States |
||
Keywords: Software Tools, Shims In this work, we proposed an automated High Order Shim procedure for neuroimaging studies. The proposed shimming procedure is fully automated and hence eliminates variability between operators. The procedure performs automated real-time brain extraction to define the region of interest (ROI) of the field map to be used in the shimming algorithm. Automated High Order Shim has fewer image distortions and narrower spectral linewidths than linear shimming and manual high-order shimming, suggesting its superior performance in correcting B0 field homogeneity. The shimming performance was assessed by acquiring EPI-based images and MR spectroscopy at both 3T and 7T field strengths. |
| 2402 | Computer 52
|
Open-source dynamic MRI workflow for reproducible research |
| Prakash Kumar1, Bilal Tasdelen1, and Krishna S Nayak1 | ||
1Ming Hsieh Department of Electrical and Computer Engineering, University of Southern California, Los Angeles, CA, United States |
||
Keywords: Software Tools, Data Acquisition This work presents an open-source workflow for dynamic MR acquisition and reconstruction. The entire workflow consists of vendor-agnostic tools to allow for immediate sharing of data, pulse sequence, and reconstruction code across sites. This work is an adaptation of Open-Source MR Imaging and Reconstruction Workflow to dynamic imaging (Veldmann et. al, MRM 88(6):2395-2407). Using the FIRE interface, we demonstrate a Gadgetron loader of acquisition metadata including k-space trajectories and present a pipeline to convert pulse sequences developed through HeartVista’s SpinBench into Pulseq for easy sharing of sequences. |
| 2403 | Computer 53
|
A diffusion-weighted MRI pulse sequence development toolbox in the open source GinkgoSequence framework |
| Anaïs Artiges1, Éléa Granier1, Ivy Uszynski1, Franck Mauconduit1, Philippe Ciuciu2, and Cyril Poupon1 | ||
1BAOBAB, NeuroSpin, Paris-Saclay University, CNRS, CEA, Gif-sur-Yvette, France, 2Mind, NeuroSpin, Paris-Saclay University, INRIA, CEA, Gif-sur-Yvette, France |
||
Keywords: Software Tools, Pulse Sequence Design Developing MRI pulse sequences demands to access proprietary sequences, to open up parts of the code provided by the manufacturer, and to deal with intellectual property issues. To address these limitations, we have provided the community with GinkgoSequence, a modular and open-source MRI pulse sequence development framework for Siemens devices. In this work, we present the addition of a new diffusion-weighted sequences development toolbox in the existing GinkgoSequence framework. It includes trapezoidal and free-waveform diffusion gradients, fat saturation, echo-planar reading, partial Fourier and GRAPPA accelerations, and stimulated echo acquisition mode (STEAM). To assess its quality, it has been tested in-vivo. |
| 2404 | Computer 54
|
PyGRASP: A standalone python image reconstruction library for DCE-MRI acquired with radial sampling |
| Cemre Ariyurek1, Aziz Koçanaoğulları1, Can Taylan Sari1, Serge Vasylechko1, Onur Afacan1, and Sila Kurugol1 | ||
1Quantitative Intelligent Imaging Lab (QUIN), Department of Radiology, Boston Children's Hospital and Harvard Medical School, Boston, MA, United States |
||
Keywords: Software Tools, DSC & DCE Perfusion Dynamic contrast enhanced MRI (DCE-MRI) is capable of quantitative assessment of renal function and evaluation of the detailed anatomy of the urinary tract and renal arteries in patients. To reconstruct the dynamic volumes in DCE-MRI, Golden-angle RAdial Sparse Parallel (GRASP) reconstruction algorithm is commonly used. In this software, we have developed an efficient open-source, purely Python-based, standalone GRASP reconstruction library called pyGRASP that allows researchers to access the source code for development, facilitating flexible deployment with readable code and no compilation, and easy utilization without requirement of a steep learning curve. |
| 2405 | Computer 55
|
qMRpy: a cross-platform, extensible and easy-to-use open-source Python toolbox for efficient estimation of quantitative MR properties |
| Matteo Cencini1, Marta Lancione1, Luca Peretti1, and Michela Tosetti1 | ||
1IRCCS Stella Maris, Pisa, Italy |
||
Keywords: Software Tools, Software Tools Quantitative MRI methods improve tissue characterization and allow for better longitudinal assessment, but need efficient and reliable fitting routines. Currently available qMR frameworks, while covering a wide range of applications, suffer either from a relatively low efficiency or require non-trivial compilation steps. Here, we propose an open-source framework for qMR quantification which maintains a good computational efficiency without sacrificing the readability and customizability of the code. |
| 2406 | Computer 56
|
QMRI-neuropipe: A flexible software framework for the analysis of quantitative MRI data |
| Douglas C Dean III1,2,3, Nagesh Adluru3, and Jose Guerrero3 | ||
1Pediatrics, University of Wisconsin–Madison, Madison, WI, United States, 2Medical Physics, University of Wisconsin–Madison, Madison, WI, United States, 3Waisman Center, University of Wisconsin–Madison, Madison, WI, United States |
||
Keywords: Software Tools, Data Processing Quantitative MRI provides a unique opportunity to characterize the underlying tissue and establish measurements that canserve as biomarkers. However, many of these methods require specialized workflows and and tools which limit their broader adoption. Here, we present QMRI-neuropipe, an open-source, flexible framework that provides a wide selection of methods, algorithms, and tools for processing and analyzing multiple qMRI datatypes. QMRI-neuropipe supports the Brain Imaging Datset Standard (BIDS) and currently supports processing of diffusion and relaxometry datasets. Future developments will aim to incorporate alternative quantitative MRI methods (e.g. MT, QSM, etc.) into the QMRI-neuropipe framework. |
| 2407 | Computer 57
|
Simulation tool for non-Fourier MRSI reconstruction |
| Carina Graf1 and Christopher T Rodgers2 | ||
1Wolfson Brain Imaging Centre, Department of Clinical Neuroscience, University of Cambridge, Cambridge, United Kingdom, 2University of Cambridge, Cambridge, United Kingdom |
||
Keywords: Software Tools, Spectroscopy, non-cartesian MRSI With the wider availability of ultra-high field MR systems, metabolic imaging using MRSI is a fast-developing area of research. Frequently used sequences use non-uniformly sampled trajectories to achieve high-acceleration factors e.g. concentric-ring trajectories (CRT). In this work, we demonstrate the from-scratch implementation of a simulation and reconstruction tool for CRT-MRSI using (1) a regridding, (2) an iterative L2 linear solver as well as (3) applying the non-uniform FFT implementation from the BART-toolbox to simulated non-cartesian MRSI data with a known Fourier-transform pair. The exact sampled data permits the evaluation of the impact of different reconstruction implementations. |
| 2408 | Computer 58
|
Continuous automated MRS data analysis workflow |
| Aaron T. Gudmundson1,2, Helge J. Zöllner1,2, Christopher W. Davies-Jenkins1,2, Erik G. Lee3,4, Timothy J. Hendrickson3,4, Richard A. E. Edden1,2, and Georg Oeltzschner1,2 | ||
1The Russell H. Morgan Department of Radiology and Radiological Science, The Johns Hopkins University School of Medicine, Baltimore, MD, United States, 2Kennedy Krieger Institute, F. M. Kirby Research Center for Functional Brain Imaging, Baltimore, MD, United States, 3Masonic Institute for the Developing Brain, University of Minnesota, Minneapolis, MN, United States, 4Informatics Institute, University of Minnesota, Minneapolis, MN, United States |
||
Keywords: Software Tools, Spectroscopy, Automation Large-scale application of magnetic resonance spectroscopy (MRS) is limited by demanding data processing and reliance on expert knowledge. We have developed an open-source continuous automated analysis workflow for MRS data, substantially reducing the effort for manual data analysis and quality control. This workflow is of particular interest for application-oriented non-expert users and large-scale multi-center multi-vendor MRS studies, and should substantially protect against loss of data from acquisition protocol errors. |
| 2409 | Computer 59
|
Novel 3D Post-Processing for 1H Magnetic Resonance Spectroscopy: Fast and Accurate Metabolite Ratios |
| Dale H. Mugler1 | ||
1Neuroscience, Medical University of South Carolina, Charleston, SC, United States |
||
Keywords: Software Tools, Spectroscopy, Data Analysis, Metabolism Brain tumor metabolic maps are one application of non-invasive MR Spectroscopy (MRS), although there are many other areas of patient treatment and surgery planning that would benefit from improved MRS analysis. A fast, accurate new method for MRS is used here to estimate the ratio of Choline to NAA from simulated spectra modeled on those near a brain tumor, using simple formulas for determining the FID amplitudes that relate to metabolite intensities and concentrations. No approximate numerical integration is required. The time of computation of 0.052 seconds per voxel speeds the construction of a metabolic 3D brain map. |
| 2410 | Computer 60
|
GPU-accelerated JEMRIS for extensive MRI simulations. |
| Aizada Nurdinova1,2, Stefan Ruschke2, Jonathan Stelter2, Michael Gestrich3, and Dimitrios Karampinos2 | ||
1Biomedical Physics, Stanford University, Stanford, CA, United States, 2Department of Diagnostic and Interventional Radiology, School of Medicine, Technical University of Munich, Munich, Germany, 3Altair Engineering GmbH, Boeblingen, Germany |
||
Keywords: Simulations, Software Tools Bloch simulations is a powerful tool in MRI research and education as it allows predicting the signal evolutions in magnetic resonance experiments for various applications. However, Bloch simulations require solving ordinary differential equations for the whole spin ensemble and can thus become infeasible for realistic scenarios with large total number of spins. Therefore, the present work investigates the potential of a GPU-based implementation for accelerated Bloch simulations as an extension to the open source Juelich Extensible MRI Simulator (JEMRIS) in comparison to the existing MPI CPU implementation. |
| 2411 | Computer 61
|
MRI Radiomics Signature to Predict Lymph Node Metastasis after Neoadjuvant Chemoradiation Therapy in locally advanced Rectal Cancer |
| Zhu Fang1, Hong Pu1, Xiaoli Chen2, Yi Yuan1, Feng Zhang1, and Hang Li1 | ||
1Department of Radiology, Sichuan Academy of Medical Sciences and Sichuan Provincial People's Hospital, Chengdu, China, 2Department of Radiology, Affiliated Cancer Hospital of Medical School, University of Electronic Science and Technology of China, Sichuan Cancer Hospital, Chengdu, China |
||
Keywords: Radiomics, Radiomics To develop and validate clinical-radiomics models for predicting lymph node metastasis following neoadjuvant chemoradiation therapy in locally advanced rectal cancer .83 patients were retrospectively enrolled.pre-,post- and delta radiomics signatures of T2WI and ADC images were constructed by support vector machine model.These models were applied to predict LNM and 5-year disease-free survival.The clinical-deltaADC radiomics combined model presented good performance for predicting post-CRT LNM in the training cohort (AUC=0.895) and validation cohort (AUC=0.900). In ypT0-T2 stage, this model could predict 5-year RFS. Clinical-deltaADC radiomics combined model has good performance to predict LNM after nCRT and helped identify patients with poor prognosis. |
| 2412 | Computer 62
|
The value of whole volume radiomics machine learning model based on multi-parameter MRI in predicting triple negative breast cancer |
| Tingting Xu1, Xueli Zhang1, Ting Hua1, Guangyu Tang1, lin Zhang1, and Xiance Zhao2 | ||
1Radiology, Shanghai Tenth People’s Hospital, School of Medicine, Tongji University, Shanghai, China, 2Philips Healthcare, Shanghai, China |
||
Keywords: Radiomics, Machine Learning/Artificial Intelligence, Triple-negative breast cancer. DCE-MRI. ADC maps 43 TNBCs and 84 Non-TNBCs were allocated in this retrospective study.The lesions were manually segmented with ITK-SNAP software then whole volume radiomics features were extracted with Radcloud radiomics platform based on DCE-MRI and ADC maps, respectively. Three prediction models were constructed by using support vector machine (SVM) classifier, including Model A (based on the selected features of ADC maps), Model B (based on the selected features of DCE-MRI), and Model C (based on the selected features of both combined). The radiomics features model combining DCE-MRI and ADC maps can improve the diagnostic performance of predicting TNBC. |
| 2413 | Computer 63
|
MRI biomarkers of neurofluid production and egress provide support for aberrations in the neurofluid circuit in Huntington’s disease |
| Kilian Hett1, Colin D. McKnight2, Jarrod Eisma1, Adreanna B. Hernandez1, Melanie Leguizamon1, Jason Elenberger1, Ciaran M. Considine1, Daniel O. Claassen1, and Manus J. Donahue1 | ||
1Department of Neurology, Vanderbilt University Medical Center, Nashville, TN, United States, 2Department of Radiology, Vanderbilt University Medical Center, Nashville, TN, United States |
||
Keywords: Data Analysis, Neurodegeneration The overall goal of this work is to apply novel MRI methods, sensitive to CSF production and clearance, to test fundamental hypotheses regarding altered neurofluid circulation in patients with Huntington’s disease (HD). In participants with HD and healthy age- and sex-matched controls, we applied multi-modal MRI to investigate deviations of neuroimaging biomarkers of the neurofluid circuit at different locations: choroid plexus, cerebral aqueduct, and PSD. Analyses show a significant increase in ChP volume and PSD volume, and a decrease in ChP perfusion and CSF net flow, in HD participants relative to healthy controls. |
| 2414 | Computer 64
|
Associated high-order resting-state functional connectivity network for diagnosis of schizophrenia |
| Dafa Shi1, Haoran Zhang1, Guangsong Wang1, and Ke Ren1 | ||
1Department of Radiology, Xiang’an Hospital of Xiamen Uneversity,School of Medicine, Xiamen University, Xiamen, China |
||
Keywords: Data Analysis, Brain Connectivity, high-order functional connectivity Schizophrenia (SZ) is one of the most prevalent mental disorders; however, its accurate diagnosis is difficult in clinical practice. Currently, the underlying mechanism of SZ remains poorly understood. The associated higher-order functional connectivity (HOFC) which constructed based on the conventional FC is promising for understanding pathological changes of brain connectome. In our study,we found the model constructed with associated HOFC outperformed the model constructed with conventional FC. SZ-related brain regions were widely distributed in frontal, parietal, insula, occipital, subcortical, and limbic lobes, which are the core brain areas of the subcortical, fronto-parietal, sensorimotor, limbic, and default mode networks. |
| 2415 | Computer 65
|
Volumetric analysis of Brain Regions: A cross-sectional study on a large healthy dataset |
| Nasrin Akbari1, Saurabh Garg1, Thanh-Duc Nguyen1, Saqib Basar1, Mostafa Fatehi2, Yosef Chodakiewitz3, Rajpaul Attariwala1,3, Sean London3, and Sam Hashemi1,3 | ||
1Voxelwise Imaging Technology Inc, Vancouver, BC, Canada, 2Vancouver General Hospital, Vancouver, BC, Canada, 3Prenuvo, Vancouver, BC, Canada |
||
Keywords: Data Analysis, Brain, A cross-sectional study, Brain quantification, Brain volume analysis Cross-sectional neuroimaging studies have indicated hemispheric asymmetry, gender differences and a decline in brain volume with advancing age. However, these studies often have a small sample size (≤ 500) and unbalanced groupings, resulting in contradictory findings that might not be reproducible. Precise estimates on a large cohort will facilitate the assessment and quantification of brain volume, resulting in improved diagnostics and establishing statistical norms. In this study we analyze 6739 individuals to draw normative conclusions about typical patterns of age-related brain changes. We provide whole-brain and regional volume aging-curves for males and females for raw and head-size adjusted data. |
| 2416 | Computer 66
|
Classification of low- and high-grade gliomas through multimodal temporal MRI and PET data |
| Marianna Inglese1,2, Matteo Ferrante1, Tommaso Boccato1, Shah Islam3, Matthew Williams4,5, Adam D Waldman6, Kevin O'Neill7, Eric O Aboagye3, and Nicola Toschi1,8 | ||
1Biomedicine and Prevention, University of Rome Tor Vergata, Rome, Italy, 2Surgery and Cancer, Imperial College London, London, United Kingdom, 3Department of Surgery and Cancer, Imperial College London, London, United Kingdom, 4Computational Oncology Group - Department of Surgery and Cancer, Imperial College London, London, United Kingdom, 5Institute for Global Health Innovation, Imperial College London, London, United Kingdom, 6Centre for Clinical Brain Sciences, University of Edinburgh, Edimburgh, United Kingdom, 7Imperial College Healthcare NHS Trust, London, United Kingdom, 8Department of Radiology, Athinoula A. Martinos Center for Biomedical Imaging, Boston, NY, United States |
||
Keywords: Data Processing, Modelling, Deep Learning, convolutional filters Stratifying human brain gliomas using imaging techniques is extremely challenging. Valuable insight into the characterization and classification of gliomas can be provided by integrating two imaging modalities, i.e. 18F-FPIA PET and MRI. This study introduces a new approach for glioma stratification based on the extraction of temporal features from tissue time activity curves (TACs) extracted from dynamic PET/MRI data. We exploit tissue-specific biochemical properties embedded in the TACs through deep learning and achieve good discrimination results while foregoing pharmacokinetic fitting and hence invasive measurement of the AIF. |
| 2417 | Computer 67
|
Implementation of a convolutional neural network for brainstem landmark detection and co-registration |
| Owen Bleddyn Woodward1, Ian Driver1, Michael Germuska1, and Richard Wise2 | ||
1CUBRIC, Cardiff University, Cardiff, United Kingdom, 2Department of Neuroscience, Imaging and Clinical Science, 'G. D'Annunzio University' of Chieti-Pescara, Chieti, Italy |
||
Keywords: Data Processing, Machine Learning/Artificial Intelligence, Co-registration Accurate brainstem co-registration is important when analysing brainstem functional MRI data. We trained a convolutional neural network (CNN) to detect a set of brainstem landmarks and to define a brainstem region-of-interest and used these to co-register the brainstem between functional and anatomical space using previously developed landmark-based and automated brainstem co-registration (LBC and ABC) methods. The use of CNNs to supply these features to LBC and ABC produced results that compared well to conventional methods. Similar CNNs could be applied to other brain regions and such methods may be useful to automate the analysis of large functional datasets. |
| 2418 | Computer 68
|
DECOMPOSE-QSM improves the tracking of iron-related neurodegenerative pathology |
| Maruf Ahmed1, Jingjia Chen1, Yann Cobigo2, Suzanne Baker3, and Chunlei Liu1,4 | ||
1EECS, University of California, Berkeley, Berkeley, CA, United States, 2University of California, San Francisco, San Francisco, CA, United States, 3Lawrence Berkeley National Laboratory, Berkeley, CA, United States, 4Helen Wills Neuroscience Institute, University of California, Berkeley, Berkeley, CA, United States |
||
Keywords: Data Processing, Quantitative Susceptibility mapping DECOMPOSE-QSM method is used to extract a sub-voxel level paramagnetic component susceptibility (PCS) in Alzheimer’s disease (AD) related neurodegeneration. Correlation between both PCS and thresholded QSM values with tau PET SUVr are studied. PCS shows stronger association strength with tau PET measures compared to susceptibility measured using only the thresholded positive voxels (positive QSM) in the QSM map in known regions related to tau pathological progression. By separating sub-voxel paramagnetic and diamagnetic sources, DECOMPOSE-QSM provides a more specific and sensitive measure of iron-related pathology. |
| 2419 | Computer 69
|
Exposure to Welding Fumes: Evaluating the metabolite-metal relationship |
| Gianna Nossa1, Humberto Monsivais2, Chang Guen Lee2, Jae Hong Park2, and Ulrike Dydak2,3 | ||
1Purdue University, West Lafayette, IN, United States, 2School of Health Science, Purdue University, West Lafayette, IN, United States, 3Department of Radiology and Imaging Sciences, Indiana University School of Medicine, Indianapolis, IN, United States |
||
Keywords: Data Analysis, Toxicity Chronic exposure to Mn and Fe, through inhalation of welding fumes, is known to cause neurotoxicity. In this MRS study, we aim to determine if toenail Mn and Fe levels can be predictive of neurotoxicity identified through changes in GABA, GSH, and/or Glu concentrations. Significant GABA and GSH correlations with Fe indicate that toenail Fe concentrations can reflect decreased GABA and GSH levels in the welders’ brain. This emphasizes that Fe should not be ignored and might contribute to oxidative stress in Mn neurotoxicity. |
| 2420 | Computer 70
|
Longitudinal Modeling of Stroke using Stochastic Distances and NODDI Diffusion Model |
| Anuja Sharma1 and Edward DiBella1 | ||
1Radiology and Imaging Sciences, University of Utah, Salt Lake City, UT, United States |
||
Keywords: Data Analysis, Machine Learning/Artificial Intelligence, Stroke diffusion modeling Limited prior work has explored longitudinal modeling of human brain stroke using advanced diffusion techniques. We aim to address this gap by analyzing longitudinal stroke data from diffusion spectrum imaging for modeling and predicting clinical markers of stroke recovery. Our proposed data analysis method uses a mixed-effect model which exploits stochastic distances from these images for improved regression model statistics and handling of imbalanced, inconsistent longitudinal data. We also demonstrate that this aids in differentiating between population-level and patient-level effects and the corresponding key contributing predictors at each of these levels. |
| 2421 | Computer 71
|
Comparison of Single-Slice Liver R2* Estimation with FerriSmart for Assessment of Hepatic Iron overload |
| Juan Pablo Esparza1, Cara Morin2,3, Chris Goode3, Zachary Abramson3, and Aaryani Tipirneni-Sajja1,3 | ||
1Biomedical Engineering, The University of Memphis, Memphis, TN, United States, 2Radiology, Cincinnati Children's Hospital Medical Center, Cincinnati, OH, United States, 3Diagnostic Imaging, St. Jude Children's Research Hospital, Memphis, TN, United States |
||
Keywords: Data Analysis, Liver FerriScan is considered the gold standard to quantify liver iron content (LIC) by MRI. Recently, FerriSmart, an AI—based algorithm was developed to permit faster postprocessing of R2-based LIC. To our knowledge, FerriSmart results have not been compared to R2*-LIC methods. In this work, R2*, R2*-LIC, and FerriSmart LIC were compared to evaluate their agreement. |
| 2422 | Computer 72
|
Magnetic resonance imaging radiomics to differentiate ovarian sex cord-stromal tumors and primary epithelial ovarian cancers |
| Cheng Meiying1, Tan Shifang1, Ren Tian2, Zhu Zitao3, Wang Kaiyu4, Zhang Lingjie1, Meng Lingsong1, Yang Xuhong5, Yang Zhexuan1, and Zhao Xin1 | ||
1Department of Radiology, The Third Affiliated Hospital of Zhengzhou University, Zhengzhou, China, 2Department of Information, The Third Affiliated Hospital of Zhengzhou University, Zhengzhou, China, 3Wuhan University, Wuhan, China, 4MR Research China, GE Healthcare, Beijing, China, 5Huiying Medical Technology, Beijing, China |
||
Keywords: Machine Learning/Artificial Intelligence, Urogenital Ovarian sex cord-stromal tumors (SCSTs) are rare nonepithelial neoplasms that usually are benign or at early stages, but sometimes they are confused with malignant tumors such as epithelial ovarian cancers (EOCs). We constructed five models including clinical model, conventional MR model, traditional model, radiomics model and mixed model based on logistic regression classifier to distinguish SCSTs and EOCs. The performance of each model was evaluated. The radiomics approach showed excellent prediction results, and the mixed model stood out among all the models. |
| 2423 | Computer 73
|
Investigating Racial Disparities for the Performance of mpMRI in Diagnosis of Prostate Cancer |
| Fatemeh Zabihollahy1, Qi Miao2, Ida Sonni1, Harrison Kim3, Robert E Reiter4, Steven S Raman1, and Kyunghyun Sung1 | ||
1Department of Radiological Sciences, David Geffen School of Medicine at University of California, Los Angeles, Los Angeles, CA, United States, 2The First Affiliated Hospital of China Medical University, Shenyang City, China, 3Department of Radiology, University of Alabama at Birmingham, Birmingham, AL, United States, 4Department of Urology, David Geffen School of Medicine at University of California, Los Angeles, Los Angeles, CA, United States |
||
Keywords: Data Analysis, Multimodal Race and ethnicity strongly impact the risk of prostate cancer (PCa). In this study, the differences in the performance of mpMRI for PCa diagnosis are investigated between African American (AA) and Caucasian American (CA) men. After matching commonly known clinical risk factors, the diagnostic performances were compared between two racial groups regarding cancer prevalence, detection rates, and positive predictive values. Our result demonstrated that there are significant differences between AA and CA men in detecting clinically significant PCa when stratified by different prostate zones. |
| 2424 | Computer 74
|
Evaluation of renal function in chronic kidney disease using histogram analysis based on diffusion multi-models |
| Guimian Zhong1, Mengzhu Wang2, Qijia Han1, and Zhiming Xiang1 | ||
1Department of Radiology, Guangzhou Panyu Center Hospital, Guangzhou, China, 2Siemens Healthineers Ltd, Guangzhou, China |
||
Keywords: Radiomics, Diffusion/other diffusion imaging techniques, Histogram analysis；magnetic resonance imaging；Renal function This study established and validated a predictive model based on histogram features of four diffusion models to identify and evaluate early renal impairment in CDK. The results suggested that the model based on the ADC and MK could distinguish the normal and mild CDK well, and could accurately and noninvasively evaluate and predict early CDK renal dysfunction. |
| 2425 | Computer 75
|
Novel subject-specific method of visualising group differences from multiple DTI metrics without averaging |
| Eryn E Kwon1,2,3, Maryam Tayebi1,3, Joshua McGeown1,2, Matthew McDonald1,2, Patrick McHugh1,4, Paul Condron1,2, Leigh Potter1,5, Davidson Taylor1,6, Jerome Maller7, Miao Qiao8, Alan Wang1,2,3, Poul Nielsen3,8, Justin Fernandez1,3,8, Miriam Scadeng1,2, Samantha Holdsworth1,2, and Vickie Shim1,3 | ||
1Mātai Medical Research Institute, Tairāwhiti/Gisborne, New Zealand, 2Faculty of Medical and Health Sciences & Centre for Brain Research, University of Auckland, Auckland, New Zealand, 3Auckland Bioengineering Institute, University of Auckland, Auckland, New Zealand, 4Tūranga Health, Tūranganui-a-kiwa, Tairāwhiti/Gisborne, New Zealand, 5Ngāti Porou, Ngāti Kahungunu, Rongomaiwahine, Rongowhakaata, Tairāwhiti/Gisborne, New Zealand, 6Ngai Tāmanuhiri, Rongowhakaata, Ngāti Porou, Tairāwhiti/Gisborne, New Zealand, 7General Electric Healthcare, Victoria, Australia, 8Department of Engineering Science, Faculty of Engineering, University of Auckland, Auckland, New Zealand |
||
Keywords: Data Analysis, Data Analysis, DTI, diffusion MRI, PCA, Mesh-fitting, subject-specific model, data aggregation Averaging is commonly used for data reduction/aggregation to summarise such high-dimensional data, resulting in information loss. However, individual variability makes group-wise comparisons difficult without data reduction/aggregation. To address these issues, we developed a novel technique that integrates diffusion tensor (DTI) metrics along the whole volumes of the fibre bundle using a mesh-fitting technique. Using the right Corticospinal Tract (rCST) as an example, we demonstrate the utility of the method in detecting differences in DTI metrics of contact-sports and non-contact-sports athletes. |
| 2426 | Computer 76
|
Radiomics Applied to Phase Contrast MRI Images Successfully Distinguishes Healthy Subjects and Multiple Sclerosis Patients |
| Eros Montin1,2, Marco Muccio1, Yulin Ge1, and Riccardo Lattanzi1,2,3 | ||
1Center for Advanced Imaging Innovation and Research (CAI2R) Department of Radiology, Radiology Department, New York University Grossman School of Medicine, New York, New York, USA, New York, NY, United States, 2Bernard and Irene Schwartz Center for Biomedical Imaging, Department of Radiology, New York University Grossman School of Medicine, New York, New York, USA, New York, NY, United States, 3Vilcek Institute of Graduate Biomedical Sciences, New York University Grossman School of Medicine, New York, New York, USA, New York, NY, United States |
||
Keywords: Radiomics, Multiple Sclerosis In this study, we applied imaged-based radiomic techniques to phase contrast (PC) MRI images to distinguish the blood flow through the neck-feeding arteries of healthy controls (HC) from MS patients. By applying a simple machine learning model, k nearest neighbor, we found that first order features of the arteries’ regions of interest (ROI), drawn on phase images, reported the best accuracy (0.80) in labeling MS patients and HCs. PC-MRI is a fast and reliable imaging technique that, in conjunction with radiomics, offers great clinical potential to further quantify the diagnosis of MS, currently relying on qualitative approaches. |
| 2427 | Computer 77
|
Artificial Neural Network for Intravoxel Incoherent Motion Fitting: an Application in Glioma |
| Lucas Murilo da Costa1, André Monteiro Paschoal2, and Renata Leoni1 | ||
1University of São Paulo, Ribeirão Preto, Brazil, 2Instituto de Radiologia e Departamento de Radiologia e Oncologia, Faculdade de Medicina, Universidade de Sao Paulo, São Paulo, Brazil |
||
Keywords: Data Processing, Diffusion/other diffusion imaging techniques We proposed an Artificial Neural Network (ANN) to fit the IVIM images and estimate diffusion, pseudo-diffusion, and perfusion fraction values. The ANN was first trained with simulated data and tested with images of 40 healthy controls and one patient with glioma. ANN provided parameter values similar to those obtained with a standard fitting method, but D* maps with less noise. The ANN fitting process was less time-consuming and was shown to be a possible tool for replacing a standard fitting algorithm. |
| 2428 | Computer 78
|
Post mortem magnetic resonance imaging simulation system |
| Hiroyuki Kabasawa1, Masatoshi Kojima2, Daisuke Yajima3, and Yohsuke Makino4 | ||
1Department of Radiological Sciences, Internationa University of Health and Walfare, Narita, Japan, 2Department of Legal Medicine, Chiba University, Chiba, Japan, 3Department of Forensic Medicine, International University of Health and Welfare, Narita, Japan, 4Department of Forensic Medicine, The University of Tokyo, Tokyo, Japan |
||
Keywords: Software Tools, Software Tools, post mortem imaging A MR Image simulation system for post mortem subject scanning is developed and validated. The system has MR console like user interface. It has body temperature input as well as tissue fixation condition input, so that post mortem MR imaging can be simulated for various scanning condition. The system was validated for T1 weighted image and FLAIR for low body temperature condition. The result showed the system could simulate the actual acquired image contrast change associated with body temperature change. The proposed system can be used to predict MR contrast change associated with body temperature as well as fixation condition. |
| 2429 | Computer 79
|
Automatic Quantification of Total Cerebral Blood Flow from Phase-Contrast MRI Using Deep Learning |
| Jinwon Kim1, Hyebin Lee2, Jinhee Jang2, and Hyunyeol Lee1 | ||
1School of Electronic and Electrical Engineering, Kyungpook National University, Daegu, Korea, Republic of, 2Department of Radiology, Seoul St. Mary’s Hospital, College of Medicine, The Catholic University of Korea, Seoul, Korea, Republic of |
||
Keywords: Machine Learning/Artificial Intelligence, Quantitative Imaging In this work, we aim to develop a deep learning (DL)-based processing pipeline that enables rapid and correct segmentation of brain-feeding arteries in neck phase-contrast (PC) MR images, thereby achieving accurate quantification of total cerebral blood flow (tCBF) in an automated manner. To this end, we implemented a U-Net architecture where magnitude/phase-combined PC images are provided for network training. The results suggest that the present, automated method yields accurate measurements of tCBF in comparison to ground truth values obtained from manual vessel segmentation. |
| 2430 | Computer 80
|
Adverse effect of non-linearity in gradient on MRI and their impact on localization and morphology of testing sample |
| Rajesh Harsh1, URVASHI SINGH1, and Dharmesh Verma1 | ||
1Medical Systems Division, S.A.M.E.E.R. Mumbai, Mumbai, India |
||
Keywords: Data Analysis, Artifacts Study demonstrate the analysis of surface morphology and sample localization with variation from iso-center in the presence of non-linearity in gradient. In these experiments, we have analyzed that gradient non-linearity plays a crucial role in surface morphology and localization of samples. For the end position 200 mm and -200 mm, size of the object appears small. Additionally, the results suggest that when a large sample size is taken into account, gradient non-linearity exhibits greater negative impacts. |
| 2529 | Computer 1
|
Cerebrospinal fluid as a zero reference regularization for functional quantitative susceptibility mapping |
| Chang-Jin Huang1, Hong-Yi Wu1, Changwei Wu2, Shen-Mou Hsu3, and Jyh-Hong Chen4 | ||
1Graduate Institute of Biomedical Electronics and Bioinformatics, National Taiwan University, Taipei, Taiwan, 2Research Center for Brain and Consciousness, Taipei Medical University, Taipei, Taiwan, 3Imaging Center for Integrated Body, Mind and Culture Research, National Taiwan University, Taipei, Taiwan, 4Department of Electrical Engineering, National Taiwan University, Taipei, Taiwan |
||
Keywords: Data Analysis, Quantitative Susceptibility mapping Functional quantitative susceptibility mapping (fQSM) providing complementary quantitative information for fMRI, has been applied to study brain functions. However, the sensitivity of fQSM suffers from the QSM reconstruction, especially solving an ill-posed deconvolution. To improve the sensitivity of fQSM, we first applied MEDI+0 in fQSM study. This method using cerebrospinal fluid as a zero reference regularization has been proven to reduce the variability of the susceptibility maps from rescan. The higher common voxel ratio and cosine similarity scores were obtained by MEDI+0 than by MEDI. Activated voxels were successfully detected by MEDI+0 during high cognition task from standard fMRI acquisition. |
| 2530 | Computer 2
|
Tumor Tissue Segmentation for Seed-based Mapping of Peritumoral Resting-State Connectivity in Patients with Glioblastomas |
| Kevin Marcus Rosenberg1,2,3 and Stefan Posse3,4,5 | ||
1Lovelace Medical System, Albuquerque, NM, United States, 2University of New Mexico, Albuquerque, NM, United States, 3Neurinsight LLC, Albuquerque, NM, United States, 4Neurology, University of New Mexico, Albuquerque, NM, United States, 5Physics and Astronomy, University of New Mexico, Albuquerque, NM, United States |
||
Keywords: Machine Learning/Artificial Intelligence, fMRI (resting state) In this study, we developed a model-free seed selection approach using deep learning-based tumor tissue segmentation in combination with iterative subject-specific seed-optimization which improves the specificity of peri-tumoral seed selection. The methodology automates seed placement in the vicinity of the tumor in the zone that is at risk during surgical resection without relying on neurofunctional brain atlases. Evaluation of cortical eloquence in different tumor subregions, such as edematous and infiltrative regions was feasible using a single MRI contrast. This computationally efficient approach was integrated into a real-time resting-state fMRI analysis pipeline to characterize peri-tumoral connectivity in patients with glioblastomas. |
| 2531 | Computer 3
|
MRGazerII: Camera-free Decoding Eye Movements from Functional Magnetic Resonance Imaging |
| Rongjie Hu1, Jie Liang1, Yiwen Ding2, Shuang Jian2, Xiuwen Wu1, Yanming Wang1, Yong Zhang3, Zhen Liang2, Bensheng Qiu1, and Xiaoxiao Wang1 | ||
1Center for Biomedical Imaging, University of Science and Technology of China, Hefei, China, 2Anhui Medical University, Hefei, China, 3GE Healthcare, Shanghai, China |
||
Keywords: Machine Learning/Artificial Intelligence, fMRI, eye movement Eye movements reflect changes in human behavior and thought to some extent, but many functional magnetic resonance imaging (fMRI) studies are limited by equipment and do not perform eye movement tracking. Recently, a deep learning method has been proposed for the regression problem of a single volume's gaze point. In this paper, we propose an end-to-end pipeline called MRGazerⅡ, which includes eye signal extraction, eye-movement behavior recognition and gaze point regression from fMRI scanning slices. The method was tested on the human connectome project (HCP) fMRI dataset and achieved the desired performance. |
| 2532 | Computer 4
|
Harmonization of Multi-Site rs-fMRI data using CovBat |
| Katherine Breedlove1,2, Hector Arciniega1,2, Sylvain Bouix3, Martha Shenton1,2, and Alexander Lin1,2 | ||
1Brigham and Women's Hospital, Boston, MA, United States, 2Harvard Medical School, Boston, MA, United States, 3University of Quebec, Montreal, QC, Canada |
||
Keywords: Software Tools, fMRI (resting state) Multisite studies are an important feature of modern neuroimaging research as they allow the study of larger and more diverse populations, which is essential to obtain clinically relevant insight and sensitivity to subtle effects. However, due to the variability among scanner platforms, even when utilizing the same acquisition protocols, harmonization is needed to remove site-to-site variability in order to improve signal-to-noise ratio and improve statistical power. By implementing CovBat as a BIDS application, we have created a turnkey application that quickly minimizes confounding signals in multisite fMRI studies. |
| 2533 | Computer 5
|
Preliminary Functional Quantitative Susceptibility Mapping with Multi-Echo EPI |
| Jannette Nassar1, Oliver C. Kiersnowski1, Patrick Fuchs1, and Karin Shmueli1 | ||
1Medical Physics and Biomedical Engineering, University College London, London, United Kingdom |
||
Keywords: Data Analysis, Brain, Quantitative Susceptiblity Mapping, fQSM, fMRI Functional QSM (fQSM) detects changes in blood oxygenation in response to neuronal activation, providing complementary information to conventional magnitude-based fMRI. For standard structural gradient-echo QSM, multi-echo (ME) acquisitions are more accurate than single-echoes. Preliminary work suggests this holds for ME-EPI. Previous fQSM studies used single-echo EPI with physiological noise correction but with ME-EPI, we observed fQSM activations in the visual cortex with a visual stimulus without physiological noise correction. ME and single-echo EPI fQSM were compared, showing that ME-EPI might be preferable. fQSM activations were weaker (maximum T-score=4 compared to 10 in fMRI) and more localised than fMRI, as expected. |
| 2534 | Computer 6
|
Spatial-temporal reconstruction using UNFOLD in looping-star silent fMRI |
| Haowei Xiang1, Jeffrey A. Fessler1, and Douglas C. Noll2 | ||
1EECS, University of Michigan, Ann Arbor, MI, United States, 2Biomedical Engineering, University of Michigan, Ann Arbor, MI, United States |
||
Keywords: Image Reconstruction, Brain Looping star is a silent MRI pulse sequence that can be used for quantitative susceptibility mapping (QSM), T2*-weighted imaging and fMRI. However, the sparse radial sampling of looping-star limits its spatial and temporal resolution in fMRI studies. This work proposes a customized looping-star fMRI protocol and a spatial temporal reconstruction method that removes the undersampling artifact from the repeating sampling pattern and improves the temporal resolution and quality of the time course and activation map. |
| 2535 | Computer 7
|
High-Frequency Resting-State Connectivity using Spectrally Segmented Regression of Movement, Physiological Noise and Spectral Residuals |
| Khaled Talaat1, Bruno Sa De La Rocque Guimaraes2, and Stefan Posse3,4 | ||
1Nuclear Engineering, University of New Mexico, Albuquerque, NM, United States, 2Nucelar Engineering, University of New Mexico, Albuquerque, NM, United States, 3Neurology, University of New Mexico, Albuquerque, NM, United States, 4Physics and Astronomy, University of New Mexico, Albuquerque, NM, United States |
||
Keywords: Data Processing, fMRI (resting state) Regression of filtering residuals is introduced to spectrally segmented regression of nuisance parameters in high-speed fMRI to enable the application of finite impulse response filters for spectral segmentation of regressors. This extension of our previously introduced method of spectrally and temporally segmented regression improves the removal of noise and mitigates the introduction of artefactual correlations in high frequency resting-state fMRI. Simulations and in-vivo data demonstrate significant advantages of spectrally segmented regression compared to whole-band regression when frequency-dependent errors are present in the regression model. High-frequency resting-state connectivity is detected with high sensitivity during normo-, hypo- and hypercapnic state. |
| 2536 | Computer 8
|
Improved BOLD activation using multiecho acquisition in simultaneous brain-spinal cord fMRI |
| Christine Sze Wan Law1, Ken Weber1, Merve Kaptan1, Dario Pfyffer1, Sean Mackey1, and Gary Glover1 | ||
1Stanford University, Stanford, CA, United States |
||
Keywords: Data Processing, Spinal Cord, fmri, multiecho, brain Functional activation within brain has been studied extensively using fMRI. But limiting investigation to only the brain provides a truncated view of the human central nervous system as it does not capture information exchange between brain and body periphery through spinal cord. Simultaneous brain-spinal cord fMRI provides a means to measure motor and pain activity across the central nervous system. BOLD signal, especially in spinal cord, usually suffers from poor signal-to-noise ratio (SNR) which can cause difficulty in detecting activation. Here, we demonstrate that multiecho acquisition improves spinal cord BOLD detection. |
| 2537 | Computer 9
|
Multi-echo EPI more effectively boosts BOLD sensitivity than sequence optimization in orbitofrontal cortex |
| Vahid Malekian1, Nadège Corbin1,2, Michael Moutoussis1,3, and Martina F. Callaghan1 | ||
1Wellcome Centre for Human Neuroimaging, UCL Queen Square Institute of Neurology, University College London, London, United Kingdom, 2Centre de Résonance Magnétique des Systèmes Biologiques, CNRS‐University Bordeaux, Bordeaux, France, 3Max Planck University College London Centre for Computational Psychiatry and Ageing Research, London, United Kingdom |
||
Keywords: Data Acquisition, fMRI Multi-echo fMRI can boost BOLD sensitivity relative to conventional single-echo fMRI, especially in high-susceptibility brain regions like orbito-frontal cortex (OFC). Another option is to optimise slice-tilts, z-shimming and k-space traversal to minimise susceptibility effects. In this study, we sought to determine if multi-echo EPI, which requires the use of parallel imaging to achieve reasonable echo times, would remain optimal in OFC when compared to an OFC-optimised single echo alternative. The relative performance is quantified via BOLD contrast-to-noise ratio and an additional comparison is made by incorporating the TE Dependent ANAlysis (TEDANA) denoising approach. Multi-echo increased BOLD CNR, particularly following denoising. |
| 2538 | Computer 10
|
Synthesis of Respiratory Phase from Image Phase in Resting-State Functional Magnetic Resonance Imaging |
| Alexander Jaffray1, Michelle Medina1, Christian Kames1, and Alexander Rauscher1,2 | ||
1Physics, UBC MRI Research Centre, Vancouver, BC, Canada, 2Department of Pediatrics, UBC, Vancouver, BC, Canada |
||
Keywords: Data Processing, fMRI (resting state), Breathing Belt, Respiratory Phase Functional Magnetic Resonance Imaging (fMRI) is routinely acquired using gradient-echo sequences with long echo times and short repetition time. Such sequences encode information about the magnetic field in the often discarded image phase. We demonstrate a method for processing the phase of reconstructed fMRI data to isolate temporal fluctuations in the harmonic fields associated with respiration by solving a blind source separation problem. Computed respiratory phase from the fMRI-derived field fluctuations was shown to be in strong agreement with breathing belt data acquired during the same scan. This work thus presents a hardware free measurement of respiratory phase. |
| 2539 | Computer 11
|
Bayesian Longitudinal Tensor Response Regression to model task-fMRI based neuroplasticity in post-stroke aphasia |
| Venkatagiri Krishnamurthy1,2, Alec Reinhardt3,4, Serena Song2, Joo Han2, M. Lawson Meadows2, Bruce Crosson2, and Suprateek Kundu3,4 | ||
1Dept. of Medicine, Emory University, Atlanta, GA, United States, 2Atlanta VA Medical Center, Decatur, GA, United States, 3Dept. of Biostatistics, Emory University, Atlanta, GA, United States, 4Dept. of Biostatistics, UT MD Anderson Cancer Center, Houston, TX, United States |
||
Keywords: Data Analysis, fMRI (task based), Neuroplasticity Stroke is inherently complex due to the heterogeneity in lesion location, size in addition to other clinical comorbidities. Further, longitudinal three-dimensional fMRI datasets add more complexity towards estimating sensitive and robust biomarkers of neuroplasticity. In this study we propose an innovative extension of Bayesian Tensor Response Regression (BTRR) approach to estimate neuroplasticity that is more sensitive, accurate and reliable compared to traditional voxel-wise approach. Results from our longitudinal aphasia-treatment study not only show that the BTRR approach is more superior but is also able to derive plasticity estimates that are sensitive to treatment differences that are subject and time-point specific. |
| 2540 | Computer 12
|
Comparison of test–retest reliability of structural, resting state functional, and diffusion tensor magnetic resonance imaging |
| Di Hu1, Yanqiu Lv1, Dandan Zheng2, Geli Hu2, Peng Sun2, and Yun Peng1 | ||
1The Department of Radiology,, Beijing Children's Hospital, Capital Medical University, National Center for Children's Health, Beijing, China, 2Clinical & Technical Support, Philips Healthcare, Beijing, China |
||
Keywords: Data Analysis, Reproductive We measured imaging reproducibility in structural, resting state fMRI and diffusion tensor scans across different time points and scanners in healthy volunteers. First, we assessed structural imaging variability by calculating volume for seven subcortical structures. Second, we evaluated across-scanner and across-time reliability of rsfMRI by assessing temporal signal-to-noise ratio of five networks. Finally, we assessed variability in diffusion metrics across scanners and time points. Our results provide statistical validations for longitudinal work on multiple systems, especially for structural study. The influence of different equipment in rsfMRI and DTI related research may be considered, especially DMN and GCC analysis involved. |
| 2541 | Computer 13
|
Impact of B0 field imperfections correction on BOLD sensitivity in 3D-SPARKLING fMRI data |
| Zaineb Amor1, Caroline Le Ster1, Chaithya G.R.1,2, Guillaume Daval-Frérot1,2,3, Bertrand Thirion1,2, Nicolas Boulant1, Franck Mauconduit1, Christian Mirkes4, Philippe Ciuciu1,2, and Alexandre Vignaud1 | ||
1Université Paris-Saclay, CEA, NeuroSpin, CNRS, Gif-sur-Yvette, France, 2Université Paris-Saclay, Inria, MIND, Palaiseau, France, 3Siemens Healthineers, Saint-Denis, France, 4Skope Magnetic Resonance Technologies AG, Zurich, Switzerland |
||
Keywords: Data Acquisition, fMRI Static and dynamic ∆B0 field imperfections are detrimental for fMRI applications as they degrade the temporal SNR (tSNR) and the sensitivity to the BOLD contrast. In this work we propose an experimental protocol for field imperfection monitoring and correction on 3D-SPARKLING fMRI data using the Skope Clip-on field camera in an alternative setting challenging its long TR constraint. We demonstrate the viability of our protocol and the reproducible gain in image quality, tSNR and retinotopic maps when correcting static and dynamic field imperfections on resting-state and task-based fMRI data for 3 healthy volunteers |
| 2542 | Computer 14
|
Noise Reduction for Task-Based Functional Magnetic Resonance Imaging using NORDIC and KWIA |
| Ru Zhang1, Zhifeng Chen2, Chengyang Zhao1, Danny J.J. Wang1, and Kay Jann1 | ||
1Mark and Mary Stevens Neuroimaging and Informatics Institute, University of Southern California, Los Angeles, CA, United States, 2Monash Biomedical Imaging, Department of Data Science & AI, Faculty of IT, Monash University, Clayton, Australia |
||
Keywords: Image Reconstruction, Image Reconstruction We quantitatively evaluate the performance of NORDIC and KWIA methods in denoising of high-resolution task-fMRI data from 7T in comparison to the standard manufacturer reconstruction. We found that both techniques improve single run statistics as well as improved consistency across runs. Analysis of tSNR confirmed improved data quality in the time domain that facilitates time-series analysis for task-fMRI. |
| 2543 | Computer 15
|
Increased BOLD resting-state fluctuation amplitude following upregulation of inhibitory activity with the GABA agonist alprazolam |
| Fanny Munsch1, Manuel Taso1, Daniel H. Wolf2, Daniel Press1, Stephanie Buss1, John A. Detre2, and David C. Alsop1 | ||
1Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, United States, 2University of Pennsylvania, Philadelphia, PA, United States |
||
Keywords: Data Analysis, fMRI (resting state), Pharmacological MRI | Drugs | GABA | Brain connectivity Neural activity reflects a complex interplay between excitatory and inhibitory signaling. Yet, the BOLD signal is typically interpreted as related to excitatory activity. However, GABAergic neurotransmission also requires metabolism and BOLD may be sensitive to this inhibitory activity that can promote coordinated activity across networks. Here, we evaluated the changes in BOLD rsfMRI and rsCBF metrics (amplitude of low frequency fluctuations (ALFF), intrinsic connectivity (IC) and local correlations (LC)) after oral administration of alprazolam, a GABA agonist, in healthy volunteers. We observed a highly significant increase of BOLD ALFF across the cortex that suggests increasing inhibition can increase BOLD signals. |
| 2544 | Computer 16
|
3D Multi-echo spiral acquisition with model-based reconstruction for fMRI |
| Zidan Yu1, Christoph Rettenmeier1, and V. Andrew Stenger1 | ||
1Department of Medicine, University of Hawaii, Honolulu, HI, United States |
||
Keywords: Data Acquisition, fMRI Multi-echo fMRI is of current interest due to its potential for more accurate BOLD contrast with reduced artifacts. Furthermore, spiral trajectories are advantageous for fMRI because of high sampling efficiency. However, images are susceptible to blurring due to B0 inhomogeneity. In this study, we present a 3D multi-echo spiral sequence with a model-based reconstruction method that combines under sampled spiral data from different echoes for increased sampling efficiency and reduced B0 artifacts. Multi-echo and single-echo spiral fMRI data including T2* maps were acquired and compared at 3T demonstrating the method. |
| 2545 | Computer 17
|
A consensus Protocol for task-free Anesthetized Rat functional Magnetic Resonance Imaging and Data Analysis |
| Roël Matthijs Vrooman1, Gabriel Desrosiers-Gregoire2,3, Gabriel Devenyi2,4, Yen-Yu Ian Shih5,6,7, Sung-Ho Lee5,6, Monica van den Berg8,9, Georgios Keliris8,9, and Joanes Grandjean1,10 | ||
1Donders Institute for Brain, Cognition and Behaviour, RadboudUMC, Utrecht, Netherlands, 2Cerebral Imaging Centre, Douglas Mental Health University Institute, Verdun, QC, Canada, 3Integrated Program in Neuroscience, McGill University, Montreal, QC, Canada, 4Department of Psychiatry, McGill University, Montreal, QC, Canada, 5Center for Animal MRI, The University of North Carolina at Chapel Hill, Chapel Hill, NC, United States, 6Neurology, The University of North Carolina at Chapel Hill, Chapel Hill, NC, United States, 7Biomedical Engineering, The University of North Carolina at Chapel Hill, Chapel Hill, NC, United States, 8Bio-imaging lab, University of Antwerp, Antwerpen, Belgium, 9µNEURO Research Centre of Excellence, University of Antwerp, Antwerpen, Belgium, 10Department for Medical Imaging, Radboud University Medical Center, Nijmegen, Netherlands |
||
Keywords: Data Acquisition, fMRI (resting state), Rat, Rodent fMRI in rats is performed under diverse protocols that span different animal handling practices and preprocessing steps. The disparity in image acquisition and preprocessing hampers the interoperability and comparison of the results between studies. Importantly, it is unclear what factors lead to superior acquisition and detection of functional networks. We detail a consensus-based task-free fMRI acquisition protocol in the rat that has been validated in 21 datasets. Along with the animal handling protocol and MRI sequence, we provide a detailed path for data conversion, preprocessing, and analysis. |
| 2546 | Computer 18
|
A Novel Approach to Estimate Whole-Brain Dynamic Cerebral Autoregulation using Task-fMRI |
| Joo Han1, Justin D. Sprick2,3, Lisa C. Krishnamurthy1,4, Serena Song1, and Venkatagiri Krishnamurthy1,5 | ||
1Center for Visual and Neurocognitive Rehabilitation, Atlanta VAMC, Decatur, GA, United States, 2Division of Renal Medicine, Department of Medicine, Emory University, Atlanta, GA, United States, 3University of North Texas Health Science Center, Denton, TX, United States, 4Department of Physics & Astronomy, Georgia State University, Atlanta, GA, United States, 5Division of Geriatrics and Gerontology, Department of Medicine, Emory University, Atlanta, GA, United States |
||
Keywords: Signal Modeling, fMRI (task based) The limitation of estimating Dynamic Cerebral Autoregulation (dCA) using Transcranial Doppler ultrasound is the lack of whole-brain estimation capabilities. In this study, the data acquisition of whole-brain task-fMRI scan during a novel Passive Cyclical Leg Raise (PCLR) task allows us to induce blood pressure (BP) changes. A surrogate BP signal was acquired from the brain stem and depicts expected fluctuations based on the PCLR-task design. The whole-brain dCA was estimated using voxel-wise TFA to obtain the transfer gain, coherence, and phase-offset. These results show similar significant brain regions but with distinct values based on the participant’s varying level of constitution. |
| 2547 | Computer 19
|
Network representation of fMRI timeseries using visibility graphs |
| Govinda Poudel1 | ||
1Australian Catholic University, Melbourne, Australia |
||
Keywords: Data Analysis, fMRI (resting state) Timeseries can be mapped into graphs by linking visibility of the signal at each time-point with respect to others – an approach known as visibility graph. We used visibility graph analysis to convert fMRI data into complex networks. We then evaluated the graph theoretical properties of this network and characterised their robustness to motion and test-retest reliability between sessions. We show that time-series network features such as average degree, average path length, and clustering coefficient are highly sensitive to motion. However, on a low-motion dataset they have a good reliability between sessions. |
| 2548 | Computer 20
|
Physiological contributions of ECG-derived respiration to BOLD fluctuations during resting-state and respiratory modulations |
| Inês Esteves1, Ana R. Fouto1, Amparo Ruiz-Tagle1, Gina Caetano1, and Patrícia Figueiredo1 | ||
1ISR-Lisboa and Department of Bioengineering, Instituto Superior Técnico – Universidade de Lisboa, Lisbon, Portugal |
||
Keywords: fMRI, Multimodal, ECG, respiration Physiological signal acquisition during fMRI may be used for multiple purposes, though it usually requires additional setup which may increase complexity and cause subject discomfort. Since ECG is modulated by respiration, an ECG-derived respiration (EDR) may be obtained without needing extra equipment for EEG-fMRI studies, which inherently use the ECG. In this work, EDR signals were computed for resting-state and two respiratory challenges modulating respiration patterns, to validate their use in the MR environment. Their performance for estimating physiological regressors of BOLD-fMRI signals was similar to the one obtained by using concurrently acquired respiratory signals. |
| 2705 | Computer 1
|
The dependence of the macrovascular transverse R2’ relaxation and resultant BOLD fMRI signal on vascular position: A simulation study |
| Xiaole Zhong1,2 and J. Jean Chen1,2,3 | ||
1Department of Medical Biophysics, University of Toronto, Toronto, ON, Canada, 2Rotman Research Institute, Baycrest Hospital, Toronto, ON, Canada, 3Institute of Biomedical Engineering, University of Toronto, Toronto, ON, Canada |
||
Keywords: Signal Modeling, Simulations The R2’ effect is the foundation of BOLD fMRI contrast. R2’ is disproportionately sensitive to the presence of macrovasculature, which disrupts the homogeneity of the main magnetic field (B0), resulting in BOLD signal intensities that may heavily depend on macrovascular orientation and volume but also on vascular position. We simulate the BOLD signal strength in a voxel containing macrovasculature at various vascular positions within the voxel. This simulation highlights an additional source of variability in the macrovascular contributions to R2’-weighted MRI. |
| 2706 | Computer 2
|
Brain Decoding and Reconstruction of concepts of visual stimuli from fMRI through deep diffusion models |
| Matteo Ferrante1, Tommaso Boccato2, and Nicola Toschi3,4 | ||
1Biomedicine and prevention, University of Rome Tor Vergata, Roma, Italy, 2Biomedicine and prevention, University of Rome Tor Vergata, Rome, Italy, 3BioMedicine and prevention, University of Rome Tor Vergata, Rome, Italy, 4Department of Radiology,, Athinoula A. Martinos Center for Biomedical Imaging and Harvard Medical school, Boston, MA, USA, Boston, MA, United States |
||
Keywords: Machine Learning/Artificial Intelligence, Neuroscience, brain decoding, fMRI In vision, the brain is a feature extractor which works from images. We hypothesize that fMRI can mimic the latent space of a classifier, and employ deep diffusion models with BOLD data from the occipital cortex to generate images which are plausible and semantically close to the visual stimuli administered during fMRI. To this end, we mapped BOLD signals onto the latent space of a pretrained classifier and used its gradients to condition a generative model to reconstruct images. The semantic fidelity of our BOLD response to visual stimulus reconstruction model is superior to the state of the art. |
| 2707 | Computer 3
|
A universal B1 shim for the human cerebellum |
| Emma Brouwer1,2, Wietske van der Zwaag1,2, and Nikos Priovoulos1,2 | ||
1Spinoza Centre for Neuroimaging, Amsterdam, Netherlands, 2Computational Cognitive Neuroscience and Neuroimaging, Netherlands Institute for Neuroscience, Amsterdam, Netherlands |
||
Keywords: Data Acquisition, Shims, B1 shim The human cerebellum is underexplored in-vivo due to a lack of acquisition methods that effectively portray it. Its dense architecture necessitates high resolution, which benefits from 7T imaging. Unfortunately, at higher field strengths, images suffer from severely destructive B1-interference with typical coil geometries. Subject-optimized parallel-transmit approaches can mitigate B1 artifacts but increase scan-time. We designed a universal B1-shim specific to the cerebellum and immediately applicable across individuals for high-resolution 7T fMRI. We demonstrate improvements in tSNR and functional responses in cerebellar ROIs during a hand-motor task compared to a quadrature mode shim setting. |
| 2708 | Computer 4
|
Statistical evaluation of complexity tests for fMRI timeseries data |
| Dilmini Wijesinghe1, Danny JJ Wang1, and Kay Jann1 | ||
1USC Mark and Mary Stevens Neuroimaging and Informatics Institute, Keck School of Medicine at USC, Los Angeles, CA, United States |
||
Keywords: Software Tools, fMRI (resting state), Complexity, Higuchi Fractal Dimension, Hurst Exponent, Lempel-Ziv Complexity, Approximate Entropy, Multiscale Sample Entropy, Fuzzy Entropy, Permutation Entropy Complexity measures of rs-fMRI signals based on non-linear timeseries analyses have been proposed for quantifying the predictability of fMRI signals. There are multiple mathematical methods for evaluating complexity in timeseries signals. This study evaluates the optimal parameter settings in seven complexity tests by analyzing mean complexity measures of grey matter (GM) and CSF across multiple complexity tests with different parameter settings and different fMRI acquisition protocols. Furthermore, complexity evaluation was performed on surrogate timeseries data. Overall, Multiscale Sample Entropy, Fuzzy Entropy, and Hurst Exponent consistently showed an increase in mean complexity of GM compared to CSF in original data. |
| 2709 | Computer 5
|
Using BOLD-fMRI to Compute Respiration Volume per Time (RVT) and Respiration Variation (RV) with Convolutional Neural Network (CNN) in Children |
| Abdoljalil Addeh1,2, Fernando Vega1,2, Rebecca J. Williams2,3,4, Ali Golestani5, G. Bruce Pike2,3,4, and M. Ethan MacDonald 1,2,3,6 | ||
1Department of Biomedical Engineering, Schulich School of Engineering, University of Calgary, Calgary, AB, Canada, 2Hotchkiss Brain Institute, Cumming School of Medicine, University of Calgary, Calgary, AB, Canada, 3Department of Radiology, Cumming School of Medicine, University of Calgary, Calgary, AB, Canada, 4Department of Clinical Neurosciences, Cumming School of Medicine, University of Calgary, Calgary, AB, Canada, 5Department of Medical Physics, Alberta Heath Services, Calgary, AB, Canada, 6Department of Electrical & Software Engineering, Schulich School of Engineering, University of Calgary, Calgary, AB, Canada |
||
Keywords: Data Processing, fMRI, Physiological Noise In many fMRI studies, respiratory signals are unavailable or do not have acceptable quality. Consequently, the direct removal of low-frequency respiratory variations from BOLD signals is not possible. This study proposes a one-dimensional CNN model for reconstruction of two respiratory measures including RV and RVT. Results show that a CNN can capture informative features from the BOLD signals and reconstruct accurate RV and RVT timeseries. It is expected that application of the proposed method will lower the cost of fMRI studies, reduce complexity, and decrease the burden on participants because they will not be required to wear a respiratory bellows |
| 2710 | Computer 6
|
Benchmarking local low rank denoising methods for task-based fMRI data analysis |
| Pierre-Antoine Comby1, Zaineb Amor1, Alexandre Vignaud1, and Philippe Ciuciu1 | ||
1Neurospin (CEA), Gif-sur-Yvette, France |
||
Keywords: Data Processing, fMRI (task based), Denoising, Benchmak Local-low-rank denoising in task-based fMRI increases sensitivity to statistical detection of neural activity, without harming specificity. We compared 5 methods (NORDIC, MP-PCA, Hybrid-PCA, Optimal-Threshold, Hybrid-OT) in 4 preprocessing configuration (denoising on magnitude/complex data, before/after motion correction) and their effect on downstream analysis. For best performance, the denoising shoud be done prior to motion correction, and using complex-valued data is only valuable in some settings. In average (n=6), up to 8 times more activations can be detected (p < 0.05, controlling for FDR). We also provide open-source implementations for broader use of Local-Low-Rank denoising methods in fMRI. |
| 2711 | Computer 7
|
Representation learning of resting state functional MRI using a volumetric variational autoencoder model (3D VAE) |
| Scott Peltier1, Michelle Karker2, Kuan Han2, Doug Noll2, and Zhongming M Liu2 | ||
1Functional MRI Laboratory, University of Michigan, Ann Arbor, MI, United States, 2Biomedical Engineering, University of Michigan, Ann Arbor, MI, United States |
||
Keywords: Machine Learning/Artificial Intelligence, Machine Learning/Artificial Intelligence In this work, we consider whether a VAE model trained with volumetric fMRI data (rather than a cortical subset of the data) is capable of encoding fMRI into low-dimensional representations, decoding these representations back into volumetric fMRI space, and also generating new fMRI patterns from the latent space. For 3D VAE model training, validation, and testing, volumetric resting-state fMRI data was used from the Human Connectome Project minimally preprocessed pipeline. We find the 3D VAE is able to accurately represent the spatial and temporal information in the data. In addition, it is able to synthesize realistic resting-state networks. |
| 2712 | Computer 8
|
Using interpretable deep learning on task fMRI data to understand brain regions related to working memory - a repeatability study |
| Tianyun Zhao1, Philip Tubiolo1,2, Thomas Hagan1, John C. Williams1,2, Jared Van Snellenberg2,3,4, and Chuan Huang1,4 | ||
1Biomedical Engineering, Stony Brook University, Stony Brook, NY, United States, 2Psychiatry and Behavioral Health, Renaissance School of Medicine at Stony Brook University, Stony Brook, NY, United States, 3Psychology, Stony Brook Univeristy, Stony Brook, NY, United States, 4Radiology, Renaissance School of Medicine at Stony Brook University, Stony Brook, NY, United States |
||
Keywords: Machine Learning/Artificial Intelligence, fMRI (task based) Deep learning, especially convolutional neural networks (CNN), has been shown to be able to identify the non-linear relation between functional magnetic resonance imaging (fMRI) and task performance. CNN can generate an interpretable result called saliency map highlighting regions that are important for task performance. It can uncover other neural processes that linear modeling cannot due to the high dimensionality nature of the fMRI. The CNN result can be presented as a saliency Previously, we developed a pipeline to produce the saliency map for working memory tasks. In this work, we further evaluated the repeatability of our pipeline. |
| 2713 | Computer 9
|
fMRI signal characteristics underlying human white-matter functional connectivity |
| Nayana Menon1, Jonathan Polimeni2, and J. Jean Chen1,3,4 | ||
1Rotman Research Institute, Baycrest Health Sciences, Toronto, ON, Canada, 2Athinoula A. Martinos Center for Biomedical Imaging, Department of Radiology, Harvard Medical School, Massachusetts General Hospital, Boston, MA, United States, 3Department of Medical Biophysics, University of Toronto, Toronto, ON, Canada, 4Department of Biomedical Engineering, University of Toronto, Toronto, ON, Canada |
||
Keywords: Data Analysis, White Matter, fMRI, functional connectivity, Fourier analysis, Human Connectome Project In this study, we illustrate the BOLD fMRI signal characteristics underlying functional correlations in the white matter of healthy adults. We use a group-level white-matter only Pearson’s correlation matrix to identify several high-connectivity regions of interest. Our results show that the BOLD signal spectra of connected white-matter regions contain a low-frequency peak unseen in unconnected regions. |
| 2714 | Computer 10
|
Joint Slice-Selective Pulse Design for Segmented FLEET-EPI using a Differential Bloch Simulator and Conjugate Gradient-Optimal Control |
| Imam Ahmed Shaik1, Mukund Balasubramanian2,3, Avery J.L. Berman4,5, Jonathan R. Polimeni2,6,7, and William Grissom1 | ||
1Institute of Imaging Science, Vanderbilt University, Nashville, TN, United States, 2Harvard Medical School, Boston, MA, United States, 3Boston Children's Hospital,, Boston, MA, United States, 4Carleton University, Ottawa, ON, Canada, 5University of Ottawa Institute of Mental Health Research, Ottawa, ON, Canada, 65Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital,, Charlestown, MA, United States, 7Harvard-MIT Division of Health Sciences and Technology,Massachusetts Institute of Technology, Cambridge, MA, United States |
||
Keywords: Data Acquisition, fMRI Multishot EPI-based fMRI can achieve high spatial resolutions to resolve functional activity at the level of layers and columns, but suffers sensitivity to motion and phase changes between shots. Reordering the shots in a multislice stack so that each slice’s shots are acquired sequentially (VFA-FLEET-EPI) reduces this sensitivity but requires specialized RF pulses to maintain consistent signal between shots as longitudinal magnetization evolves. In this work we show that designing these pulses jointly using an autodiff optimal control algorithm yields more consistent signal across shots which reduces ghosting compared to VFA sinc and recursively designed SLR pulses. |
| 2715 | Computer 11
|
Manual registration and customized template for rodent fMRI data spatial normalization |
| Wen-Ju Pan1, Nmachi Anumba1, Nan Xu1, Lisa Meyer-Baese1, and Shella Keilholz1 | ||
1Emory University/Georgia Institute of Technology, Atlanta, GA, United States |
||
Keywords: Data Analysis, Brain Rodent EPI image qualities may vary across coil types, coil positioning and different animals that challenge atlas registration. We proposed an accurate registration with study-group customized EPI template and initially manual registration with an assistance of the newly-introduced tissue-boundary atlas. Our studies demonstrated some visible mismatching in local anatomic structures by the standard registration methods for rodent data which were effectively corrected by the presented method. |
| 2716 | Computer 12
|
Functional MRI denoising Using Data-Driven Multi-Step Deep Neural Network |
| Sina Ghaffarzadeh1, Vahid Malekian2, Faeze Makhsousi3, and Seyyed Ali Seyyedsalehi3 | ||
1Biomedical Engineering, Amirkabir University of Technology (Tehran Polytechnic), Tehran, Iran (Islamic Republic of), 2University College London, London, United Kingdom, 3Amirkabir University of Technology (Tehran Polytechnic), Tehran, Iran (Islamic Republic of) |
||
Keywords: Data Processing, Brain In this study a novel method for sampling the active and noisy areas is proposed by using the purification of gray and non-gray matter areas of fMRI data. Also, a data-driven network is proposed in a parallel, multi-step and integrated manner for optimal noise reduction of t-fMRI data. Besides, the proposed method reduces substantially physiological noise without considering the specific noise source and only by using the ROI of noise and activity. Based on the results, the proposed method provides a more accurate and improved activity map than previous methods, which increases the power of activity analysis in fMRI data. |
| 2717 | Computer 13
|
BISEPI high-resolution fMRI at 7T with NORDIC-PCA |
| Guoxiang Liu1,2, Takashi Ueguchi1,2, and Seiji Ogawa1,3 | ||
1NICT, Osaka, Japan, 2Graduate School of Frontier Biosciences, Osaka University, Osaka, Japan, 3Tohoku Fukushi University, Sendai, Japan |
||
Keywords: Image Reconstruction, fMRI (task based), Highe resolution Recently, a new noise reduction technique NOise Reduction with Distribution Corrected (NORDIC) PCA has been proposed for High resolution fMRI, and the improvements in keys including temporal-SNR, BOLD detectability have been demonstrated using this technique. In this work, we applied this technique to 0.4-mm isotropic human fMRI data acquired with our Block-interleaved segmented EPI, and compare the BOLD detectability of the data with and without NORDIC-performed. |
| 2718 | Computer 14
|
Depth-dependent effects of thermal and physiological noise reduction in BOLD fMRI |
| Maria Guidi1,2,3, Giovanni Giulietti4,5, Harald E. Moeller2, David G. Norris3, and Federico Giove1,4 | ||
1MARBILab, Enrico Fermi Research Center, Rome, Italy, 2Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig, Germany, 3Donders Centre for Cognitive Neuroimaging, Radboud University, Nijmegen, Netherlands, 4Fondazione Santa Lucia IRCCS, Rome, Italy, 5SAIMLAL Department, Sapienza University, Rome, Italy |
||
Keywords: Data Processing, Data Analysis, Denoising, Layers In this study, we evaluated the effect of common denoising steps (NORDIC, regression for motion parameters, RETROICOR and aCompCor) on a high-resolution resting-state BOLD fMRI dataset. We extracted the temporal standard deviation and the spectral power density at different cortical depths in the primary motor cortex and found that each denoising algorithm had a distinct signature on the profile shape. We further estimated the effect of denoising by calculating the temporal signal-to-noise ratio and delta variation signal (DVARS) for different tissue types and found that NORDIC and aCompCor had the largest impacts on the metrics considered. |
| 2719 | Computer 15
|
Validating NORDIC denoising on high-resolution fMRI data at 7 T |
| Viktor Pfaffenrot1 and David Norris1,2 | ||
1Erwin L. Hahn Institute for Magnetic Resonance Imaging, University of Duisburg-Essen, Essen, Germany, 2Donders Institute for Brain, Cognition and Behavior, Radboud Iniversity Nijmegen, Nijmegen, Netherlands |
||
Keywords: Data Processing, fMRI, denoising, laminar fMRI, NORDIC PCA High-resolution fMRI is dominated by thermal noise. The NORDIC PCA filter is a promising method to remove thermal noise but its effect on the spatial fidelity of layer-dependent profiles is sparsely investigated. We evaluated the performance of NORDIC with several parameter setting for laminar fMRI at 7 T and sought to investigate the effect NORDIC has on the shape of the profiles. Our results suggest that NORDIC indeed affects high-resolution profiles but the overall effect can be seen to be small and can be further reduced with careful parameter settings. |
| 2720 | Computer 16
|
Where is each finger area in brain? Enhanced characterization of the somatosensory area using high-resolution EPIK at 3T |
| Sung Suk Oh1, Min Cheol Chang2, Soyoung Kwak2, Jong-ryul Choi3, N. Jon Shah4,5,6,7, and Seong Dae Yun4 | ||
1Medical Device Development Center, K-MEDI hub, Daegu, Korea, Republic of, 2Department of Rehabilitation Medicine, Yeungnam University, Daegu, Korea, Republic of, 3K-MEDI hub, Daegu, Korea, Republic of, 4Institute of Neuroscience and Medicine 4, INM-4, Forschungszentrum Juelich, Juelich, Germany, 5Institute of Neuroscience and Medicine 11, INM-11, Forschungszentrum Juelich, Juelich, Germany, 6JARA - BRAIN - Translational Medicine, Aachen, Germany, 7Department of Neurology, RWTH Aachen University, Aachen, Germany |
||
Keywords: Data Acquisition, fMRI (task based) The tactile stimulus of fingers in fMRI can effectively reveal the somatosensory functional region. However, this research requires a high-resolution fMRI technique, enabling a distinct delineation of functional areas for each finger. This work demonstrates the use of high-resolution EPIK (1.25 x 1.25 mm2) for enhanced characterization of somatosensory areas at 3T. EPIK provides improved spatial resolution than EPI (1.72 x 1.72 mm2) with increased brain coverage while keeping the same temporal resolution as EPI. The activation region from EPIK was shown to be more locally specific, thereby enabling a clearer recognition of functional areas for each finger. |
| 2721 | Computer 17
|
Spin-echo-based generalized Slice Dithered Enhanced Resolution (gSLIDER) for mesoscale fMRI at 3 Tesla |
| Salvatore John Torrisi1,2,3, Congyu Liao4, Jennifer Townsend1,3, and An (Joseph) Vu1,3 | ||
1Radiology, SF VA Medical Center, San Francisco, CA, United States, 2Northern California Institute of Research and Education, San Francisco, CA, United States, 3Radiology, University of California, San Francisco, San Francisco, CA, United States, 4Division of Radiological Sciences Laboratory, Stanford University, Palo Alto, CA, United States |
||
Keywords: Pulse Sequence Design, Contrast Mechanisms, Neuro We demonstrate sub-millimeter generalized Slice Dithered Enhanced Resolution (gSLIDER) for high-resolution spin-echo (SE) fMRI at 3T. Activations were significantly greater than standard SE fMRI, demonstrating the suitability of this method for high-resolution, mesoscale fMRI. |
| 2722 | Computer 18
|
The ALFF alterations in maintenance hemodialysis patients with sleep disorder:a rs-fMRI study combined with machine learning analysis |
| Menghan Feng1, Yue Zhang2, Zeying Wen3,4, Yuchi Wu2, Chengwei Fu3, Kan Deng5, Qizhan Lin2, and Bo Liu2 | ||
1Xin-Huangpu Joint Innovation Institute of Chinese Medicine in Guangdong Province, Guangzhou, China, 2The Second Affiliated Hospital of Guangzhou University of Chinese Medicine, Guangzhou, China, 3Guangzhou University of Chinese Medicine, Guangzhou, China, 4The First Affiliated Hospital of Henan University of Chinese Medicine, Zhengzhou, China, 5Philips Healthcare, Guangzhou, China |
||
Keywords: fMRI (resting state), Kidney The prevalence of SD is high in chronic kidney disorder patients who suffering maintenance hemodialysis whereas little is known about the neuropathologic mechanism. We investigated the ALFF alterations between MHDSD patients and HCs and those meaningful features were used for constructing discriminating model based the SVM algorithm to classify the MHDSD patients. We found the aberrant spontaneous activities in the DMN, AN, CEN, and VIN in MHDSD patients. Additionally, the classifier indicated the discriminative ALFF features in the above regions demonstrated good performance. This will contribute to well understanding the neuropathological mechanism and seeking biomarkers for discrimination in MHDSD patients. |
| 2723 | Computer 19
|
Reliability and sensitivity of fMRI at 3T and 1.5 T: metrics toward fMRI guided individualized precise TMS treatment |
| Qiu Ge1,2,3, MinLiang Yao4, Juan Yue1,2,3,4, Xue Yang1,2,3, YueJiao Ding1,2,3, Yong Zhang5, and Yufeng Zang1,2,3,4 | ||
1Center for Cognition and Brain Disorders, The Affiliated Hospital of Hangzhou Normal University, Hangzhou, China, 2Zhejiang Key Laboratory for Research in Assessment of Cognitive Impairments, HangZhou, China, 3Institute of Psychological Sciences, Hangzhou Normal University, HangZhou, China, 4Hangzhou Normal University Affiliated Deqing Hospital, HangZhou, China, 5Yong Zhang, GE Healthcare, Shanghai, China |
||
Keywords: Data Acquisition, fMRI FDA cleared a TMS treatment protocol with fMRI-guided individualized targeting for major depressive disorder. This protocol was based on 3T scanner. However, 1.5 T MRI is more widely available in hospitals. We comprehensively tested the reliability and sensitivity of fMRI metrics related to TMS treatment on 1.5T and 3T scanners. In general, the sensitivity of 3T is larger than 1.5T. However, the reproducibility of peak FC location and the activation location, the reliability of RS-fMRI local metrics, are similar for 1.5T and 3T. In conclusion, 1.5 fMRI meets the needs for guiding individualized precise rTMS treatment. |
| 2724 | Computer 20
|
When BOLD Optimized ME-EPI Echo Combination Isn't BOLD Optimized |
| R. Allen Waggoner1, Chisato Suzuki1, Ken-ichi Ueno1, and Keiji Tanaka1 | ||
1RIKEN Center for Brain Science, Wako-shi, Saitama, Japan |
||
Keywords: Artifacts, fMRI, Multi-Echo, Multi-Shot In regions suffering from magnetic susceptibility effects, the use of BOLD optimized echo combination for averaging Single-Shot/Multi-Echo EPI data across echos, leads to voxels with improves signal but reduced BOLD sensitivity. The use of Multi-shot/Multi-Echo EPI can suppress the susceptibility effects allowing BOLD optimized echo averaging that retains BOLD sensitivity. |
| 2880 | Computer 1
|
Three Delineating Radiomics Models Based on MR-determined Metastatic Lymph Nodes for Prognostic Prediction in Nasopharyngeal Carcinoma |
| Manqian Huang1, Kan Deng2, Hui Xie1, Wenjie Huang1, Xiaoyi Wang3, Chao Luo1, Shuqi Li1, Chunyan Cui1, Huali Ma1, Lizhi Liu1, and Haojiang Li1 | ||
1Sun Yat-sen University Cancer Center, Guangzhou, China, 2Philips Healthcare, Guangzhou, China, 3Hainan General Hospital (Hainan Affiliated Hospital of Hainan Medical University), Haikou, China |
||
Keywords: Radiomics, Cancer, Nasopharyngeal Carcinoma The purpose of this study was to assess the performance of radiomic models based on MR-determined metastatic lymph nodes phenotype in predicting the prognosis of nasopharyngeal carcinoma (NPC) with three different feature selection methods: all lymph nodes (ALN), the largest lymph node (LLN), and the largest slice of the largest lymph node (LSLN). The results showed that LSLN radiomic features showed better accuracy in predicting the overall survival (OS). |
| 2881 | Computer 2
|
MRS combined with enhanced silhouette in the prediction of malignant glioma radiomics classification |
| jing song1, hui qian zong1, ya zhang1, jing wang1, hong yang wei1, and li zhi xie2 | ||
1The Second Hospital of Hebei Medical University, Shijiazhuang, China, 2GE Healthcare, Beijing, China |
||
Keywords: Radiomics, Brain, Magnetic resonance spectroscopy Question: Among the studies using radiomics methods to predict glioma grading, most of them are based on conventional magnetic resonance imaging sequences, and functional magnetic resonance imaging is less studied. Methods: This study predicted malignant glioma grading based on magnetic resonance structural images and magnetic resonance spectroscopy using an radiomics approach. Results: The test set AUC of the model constructed based on T1-enhanced images and the ratio of three metabolites of MRS was 0.95. Conclusion: Radiomics based on T1-CE and MRS has a good performance in identifying both grade III and grade IV gliomas. |
| 2882 | Computer 3
|
Radiomic feature reliability of protein-based amide proton transfer-weighted images of brain tumors acquired with compressed sensing |
| Jingpu Wu1,2, Yiqing Shen1,3, Qianqi Huang3, Pengfei Guo1,3, Jinyuan Zhou1, and Shanshan Jiang1 | ||
1Department of Radiology, School of Medicine, Johns Hopkins University, Baltimore, MD, United States, 2Department of Applied Mathematics and Statistics, Whiting School of Engineering, Johns Hopkins University, Baltimore, MD, United States, 3Department of Computer Science, Whiting School of Engineering, Johns Hopkins University, Baltimore, MD, United States |
||
Keywords: Radiomics, Radiomics Sensitivity encoding (SENSE) is a conventional practice for accelerating APTw image acquisition. To achieve an even higher acceleration, SENSE with compressed sensing (CS-SENSE) was introduced. However, its effect on the radiomic features extracted from the APTw images was yet studied. Here we extracted radiomic features from both SENSE- and CS-SENSE-APTw images and evaluated their reliability in different regions of interest (ROIs). Moreover, filters play an important part in emphasizing specific image characteristics in radiomics. The impact of filters on the radiomic features were also discussed. Our results provided a comprehensive reference for radiomic analyses where CS is implemented for acceleration. |
| 2883 | Computer 4
|
Reproducibility of radiomics features between MRI-derived synthetic-CT and true CT in prostate MR-guided radiotherapy |
| Yihang Zhou1, Jing Yuan1, Cindy Xue1, Bin Yang2, Kin Yin Cheung2, and Siu Ki Yu2 | ||
1Research Department, Hong Kong Sanatorium and Hospital, Hong Kong, China, 2Medical Physics Department, Hong Kong Sanatorium and Hospital, Hong Kong, China |
||
Keywords: Radiomics, Prostate, MRCAT, synthetic-CT, reproducibility MR-guided radiotherapy and radiomics have gained considerable attention. Synthetic-CT (sCT) derived from MRI has been adopted to facilitate MRI-only radiotherapy planning. Researches have evaluated the sCT both qualitatively and quantitatively. However, few studies have looked into the potential that sCT offers at radiomics level. We hypothesize that sCT generated from MR can faithfully reproduce radiomics features compared to those extracted from true planning-CT. We aim to investigate the reproducibility of radiomics features derived from a commercially available sCT generation (MRCAT) acquired on a 1.5T MR-simulator to those obtained from a CT simulation scan in a cohort of prostate cancer patients. |
| 2884 | Computer 5
|
Prediction of overall survival for astrocytoma with whole-tumor radiomics analysis based on diffusion kurtosis imaging |
| Yan Tan1, Dawei Tian1, Wenqiao Zheng1, Xiaochun Wang1, and Hui Zhang1 | ||
1Department of Radiology, First Hospital of Shanxi Medical University, Taiyuan, China |
||
Keywords: Radiomics, Diffusion/other diffusion imaging techniques For investigating the usefulness of radiomics signature based on diffusion kurtosis imaging in overall survival prediction for astrocytoma, radiomics features extracted from whole-tumor on MK and MD images were selected to construct radiomics signature. Then the radiomics signature was integrated with clinical-genetic risk factors to develop a combined model, which was represented as nomogram. The results showed that the radiomics signature have offered clinically relevant prognostic information for astrocytoma and further stratified patients into different risk groups. The combined model achieved the highest predictive performance and facilitated clinical decision-making by nomogram, demonstrating the incremental value of radiomics signature. |
| 2885 | Computer 6
|
Potential anti-HER2 target therapy beneficiaries: can MRI radiomics identify the status of HER2-low in breast cancer? |
| Xiaoqian Bian1, Zhibin Yue2, Siyao Du1, Yan Xu3, Yang Song3, Min Zhao4, and Lina Zhang1 | ||
1The First Hospital of China Medical University, Shenyang, China, 2Department of Biomedical Engineering, School of Intelligent Medicine, China Medical University, Shenyang, China, 3MR Scientific Marketing, Siemens Healthineers, Shanghai, China, 4Pharmaceutical Diagnostics, GE Healthcare, Beijing, China |
||
Keywords: Radiomics, Breast Recently, HER2-low expression tumor was proposed as a new entity in the field of breast cancer and radiomics studies related to HER2-low are rare. This study investigated whether multiparametric MRI-based radiomics can distinguish HER2- low from HER2-positive and HER2-zero in breast cancer. Results showed that clinico-radiomics nomograms including hormone receptor status and radiomics signatures combined contrast-enhanced T1-weighted and apparent diffusion coefficient map performed best in both prediction tasks: HER2-positive vs. HER2-negative and HER2-low vs. HER2-zero. This suggests that multiparametric MRI radiomics achieved effective prediction of HER2-low, which might be further guidance of the anti-HER2 targeted therapy in breast cancer. |
| 2886 | Computer 7
|
Effects of SMS accelerating factor on stability of radiomics features from quantitative parametric maps of IVIM and DKI in Cervical Cancer |
| Ai Shuangquan1,2, Peng Wei1, He Yaoyao1, Zhang Huiting3, Grimm Robert4, Zhang Zhaoxi1, Peng Lin5, Liu Yulin1, and Yuan Zilong1 | ||
1Department of Radiology, Hubei Cancer Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China, 2School of Biomedical Engineering, South-Central Minzu University, Wuhan, China, 3MR Scientific Marketing, Siemens Healthcare, Wuhan, China, Wuhan, China, 4MR Application Predevelopment, Siemens Healthcare GmbH, Erlangen, Germany, Erlangen, Germany, 5Department of Gynecology, Hubei Cancer Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China |
||
Keywords: Radiomics, Cancer, Cervical Cancer The aim of this study was to explore the effects of SMS on radiomics features in quantitative parametric maps based on IVIM and DKI models. Meanwhile, the effects of whole-tumor and maximal-tumor layer delineation on radiomics features were compared, whether stable features are associated with clinical staging of cervical cancer was analyzed. The results showed that SMS had the greatest effect on features extracted from D* and f map from IVIM model and the least effect on features extracted from ADC map; SMS had a greater effect on the radiomics features extracted by maximum-tumor layer delineation than by whole-tumor delineation. |
| 2887 | Computer 8
|
The value of regional radiomics score based on postoperative conventional MRI in evaluation of glioma recurrence |
| Jinfa Ren1, Xiaoyang Zhai1, Dongming Han1, Huijia Yin1, Ruifang Yan1, and Kaiyu Wang2 | ||
1Department of MR, The First Affiliated Hospital of Xinxiang Medical University, Weihui, China, 2GE Healthcare, MR Research China, Beijing, China |
||
Keywords: Radiomics, Radiomics Early diagnosis of postoperative glioma recurrence is difficult. But emerging measurements of radiomics could provide a powerful tool for this dilemma. We used the least absolute shrinkage and selection operator to select features and generate radiomics scores based on multiple modalities of conventional MRI to discriminate recurrence from treatment-related effects. We found that tumor recurrence could be independently identified by features from both the postoperative enhanced region and edematous region with a best performance of the combined one. |
| 2888 | Computer 9
|
Comparison of radiomics-based machine learning survival models in predicting prognosis of glioblastoma |
| Jixin Luan1, Chuanchen Zhang2, Bin Liu1, Aocai Yang1, Kuan Lv3, Pianpian Hu3, and Guolin Ma1 | ||
1China-Japan Friendship Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing, China, 2Department of Radiology, Liaocheng People's Hospital, Shandong First Medical University & Shandong Academy of Medical Sciences, Liaocheng, China, 3China-Japan Friendship Hospital, Beijing, China |
||
Keywords: Radiomics, Cancer In this study, we aimed to compare the performance of radiomics-based machine learning survival models in predicting the prognosis of glioblastoma multiforme (GBM) patients. The Cox proportional-hazards model (Cox-PH) and SurvivalTree, Random survival forest (RSF), DeepHit, DeepSurv four machine learning models were constructed, and the performance of the models was evaluated using C-index. We found that deep learning algorithms based on radiomics in predicting the overall survival of GBM patients, and the DeepSurv model showed the best predictive ability. |
| 2889 | Computer 10
|
Rado – A Cloud-Based Toolbox for Radiomics Analysis |
| Eros Montin1,2 and Riccardo Lattanzi1,2,3 | ||
1Center for Advanced Imaging Innovation and Research (CAI2R) Department of Radiology, Radiology Department, New York University Grossman School of Medicine, New York, New York, USA, New York, NY, United States, 2Bernard and Irene Schwartz Center for Biomedical Imaging, Department of Radiology, New York University Grossman School of Medicine, New York, New York, USA, New York, NY, United States, 3Vilcek Institute of Graduate Biomedical Sciences, New York University Grossman School of Medicine, New York, New York, USA, New York, NY, United States |
||
Keywords: Radiomics, Software Tools Rado is a web-based application designed to fully automatically perform a radiomic analysis. Rado allows users to extract features from medical images, apply feature selections, and mine the dataset using the most common classifiers and regressors. It also offers the option to augment the dataset by means of rigid transformations.The current version of Rado allows users to interact with the dataset by means of restful APIs or using a standardized web GUI. The application will be distributed via the Cloud MR portal, which allows running the feature extraction on sparse servers as well as on local computers. |
| 2890 | Computer 11
|
A radiomic approach to the diagnosis of femoroacetabular impingement |
| Eros Montin1,2, Richard Kijowski3, and Riccardo Lattanzi1,2,4 | ||
1Center for Advanced Imaging Innovation and Research (CAI2R) Department of Radiology, Radiology Department, New York University Grossman School of Medicine, New York, New York, USA, New York, NY, United States, 2Bernard and Irene Schwartz Center for Biomedical Imaging, Department of Radiology, New York University Grossman School of Medicine, New York, New York, USA, New York, NY, United States, 3Department of radiology, New York University Grossman School of Medicine, New York, NY, United States, New York, NY, United States, 4Vilcek Institute of Graduate Biomedical Sciences, New York University Grossman School of Medicine, New York, New York, USA, New York, NY, United States |
||
Keywords: Radiomics, MSK The results of this study showed that radiomic can automatically distinguish a healthy joint from one with impingement using water-only Dixon MRI. To our knowledge, this is the first application of radiomic for FAI diagnosis. Our radiomic analysis achieved an accuracy greater than 97%, which is higher than the 90% accuracy for detecting FAI reported for standard diagnostic tests (90%). Combining our proposed radiomic analysis with methods for automated joint segmentation could be used to rapidly identify patients with FAI, avoiding time-consuming radiological measurements of bone morphology. |
| 2891 | Computer 12
|
Radiomics-based Machine Learning for Predicting Clinically Significant Cancer in Multicenter Cohort: Comparison to PI-RADS Reading |
| Gabriel Addio Nketiah1,2, Mohammed RS Sunoqrot 1,3, Elise Sandsmark3, Sverre Langørgen 3, Kirsten M Selnæs 1,3, Helena Bertilsson 1,4, Mattijs Elschot 1,3, and Tone F Bathen1,3 | ||
1Department of Circulation and Medical Imaging, Norwegian University of Science and Technology, Trondheim, Norway, 2Department of Radiology and Nuclear Medicine, St. Olavs Hospital, Trondheim University Hospital,, Trondheim, Norway, 3Department of Radiology and Nuclear Medicine, St. Olavs Hospital, Trondheim University Hospital, Trondheim, Norway, 4Department of Urology, St. Olavs Hospital, Trondheim University Hospital, Trondheim, Norway |
||
Keywords: Machine Learning/Artificial Intelligence, Prostate Synopsis: Recently, predictive machine learning models have shown promise for prostate cancer diagnosis. The utility of MRI radiomic features for prostate cancer detection and classification has been shown several studies, but mostly using relatively small and single centre cohort. In this study, we showed that radiomics-based machine learning can perform relatively well compared to clinical practice, especially in large multicentre settings. On the patient-level analysis, the areas under the receiver-operating curves for PI-RADS reading by a radiologist and machine learning model were 90% and 89%, respectively. |
| 2892 | Computer 13
|
The Impact of Compressed Sensing on Radiomics of Fast Anatomical MRI of Rat Brain at 7T |
| Arda Arpak1, Esra Sümer1, Asım Samlı1, Pınar S. Özbay1, Uluç Pamuk2,3, Kübra Çağlar2, Elif M. Demir2, Mehmet Yumak2, Can A. Yucesoy1, and Esin Ozturk-Isik1 | ||
1Institute of Biomedical Engineering, Boğaziçi University, İstanbul, Turkey, 2Targeted Therapy Technologies Experimental Animal Imaging Laboratory, Boğaziçi University, İstanbul, Turkey, 3Department of Biomedical Engineering, İstinye University, İstanbul, Turkey |
||
Keywords: Image Reconstruction, Preclinical, Radiomics This study aims to investigate the effect of compressed sensing reconstruction on radiomics parameters of rat brain MRI at 7T. T2 weighted MRI of rat brain were randomly under-sampled with R=2.2 and reconstructed using compressed sensing. Brains were segmented after registration to a Wistar rat brain atlas. The radiomics of original and accelerated MRI data were compared in seven scans of rat brains using a Wilcoxon signed-rank test. Stable radiomics features were identified using intraclass correlation coefficients. The findings of this study indicated that not all radiomics features of rat brain were robust to compressed sensing acceleration. |
| 2893 | Computer 14
|
Preoperative MRI-based radiomic-clinical nomogram to predict residual tumor for advanced high-grade serous ovarian carcinoma |
| Jingjing Lu1, Songqi Cai1, Fang Wang2, Pu-Yeh Wu3, Xianpan Pan2, Jinwei Qiang4, Haiming Li5, and Mengsu Zeng1 | ||
1Department of Radiology, Zhongshan Hospital, Fudan University, Shanghai, China, 2Department of Research and Development, Shanghai United Imaging Intelligence Co., Ltd, Shanghai, China, 3GE Healthcare, Beijing, China, 4Department of Radiology, Jinshan Hospital, Fudan University, Shanghai, China, 5Department of Radiology, Fudan University Shanghai Cancer Center, Shanghai, China |
||
Keywords: Machine Learning/Artificial Intelligence, Radiomics, Ovarian carcinoma, Residual tumor prediction Residual tumor (RT) status is associated with the prognosis and survival rate of patients with high-grade serous ovarian carcinoma (HGSOC). However, current RT status prediction approach through laparoscopy has disadvantages of invasiveness, high cost and incidence of tumor metastases. In this study, we proposed a radiomic-clinical nomogram, based on multiple-sequence MRI combined with score of abdominal metastases and clinical markers, for preoperative prediction of RT status. We demonstrated that the radiomic-clinical nomogram had satisfactory prediction performance in all cohorts (AUC = 0.900-0.936). The clinical application value of the nomogram was further confirmed by decision curves. |
| 2894 | Computer 15
|
Different multiparametric MRI-based radiomics models for differentiating stage IA endometrial cancer from benign endometrial lesions: A multicenter study |
| Qiu Bi1, Kunhua Wu1, and Yunzhu Wu2 | ||
1the First People’s Hospital of Yunnan Province, Kunming, China, 2Siemens Healthcare, Shanghai, China |
||
Keywords: Artifacts, Uterus In the study, age and irregular vaginal bleeding were the valid predictive parameters in clinical model. On the basis of several common machine learning algorithms, the diverse multiparametric MRI-based radiomics models were developed to differentiate stage IA EC from benign endometrial lesions, and LR algorithm model were selected as the optimal radiomics model with the highest AUC and accuracy. Compared with clinical model and radiologist, the optimal radiomics model and the compositive models combining clinical parameters with radiomics features, like the nomogram, stacking model, and ensemble model showed better diagnostic performance and achieved good clinical net benefits. The nomogram had a higher AUC than that of the optimal radiomics model, and revealed more stable discrimination efficiency and better generalization ability than stacking and ensemble modals. |
| 2895 | Computer 16
|
nnFAE: An Extended Module for FeAture Explorer (FAE) for Radiomic Feature Processing |
| Yang Song1, Chengxiu Zhang2, Jing Zhang2, Shaoyu Wang1, Xu Yan1, Yefeng Yao2, and Guang Yang2 | ||
12. MR Scientific Marketing, Siemens Healthcare, Shanghai, China, 21. Shanghai Key Laboratory of Magnetic Resonance, East China Normal University, Shanghai, China |
||
Keywords: Machine Learning/Artificial Intelligence, Software Tools The biological meaning, model robustness and the harmonization of the features are the focuses in the current radiomics development. We designed a software named nnFAE which extends the open-source FeatureExplorer (FAE) to extract habitats features using multi-parameter MR images, to extract robust features recommended by IBSI, and to harmonize features from multi-vendors, etc. nnFAE has a graphic user interface to process the images and feature matrix in batch and can be used readily in radiomics studies. |
| 2896 | Computer 17
|
Application of radiomics approach on predicting freezing of gait in Parkinson’s disease based on rs-fMRI indices |
| Miaoran Guo1, Hu Liu1, and Guoguang Fan1 | ||
1The First Hospital of China Medical University, Shenyang, China |
||
Keywords: Machine Learning/Artificial Intelligence, fMRI (resting state), Parkinson’s disease, freezing of gait, feedforward neural network, receiver operating characteristic.
|
| 2897 | Computer 18
|
Feasibility of texture analysis of cocaine use disorder patients’ MRI data to predict early brain changes |
| Pavan a Poojar1, Priti Balchandani2, Yasmin Hurd3, Shilpa Taufique3, and Sairam Geethanath1 | ||
1Accessible Magnetic Resonance Laboratory, Biomedical Imaging and Engineering Institute, Department of Diagnostic, Molecular and Interventional Radiology, Icahn School of Medicine at Mount Sinai, New York, NY, United States, 2Biomedical Imaging and Engineering Institute, Department of Diagnostic, Molecular and Interventional Radiology, Icahn School of Medicine at Mount Sinai, New York, NY, United States, 3Department of Psychiatry, Addiction Institute at Mount Sinai, New York, NY, United States |
||
Keywords: Data Analysis, Brain Substance use disorder (SUD) affects the structure, function, and metabolism of the brain, and MR imaging helps to track and manage the therapeutic efficacy. We analyzed cortical white matter and Amygdala regions for 5 patients with cannabis use disorder using changes in texture and volume. Each patient was scanned 4 times(at 0,2,12 and 24 weeks) with 3D fast spin echo on a 3T scanner. We found increased textural changes across intervals for cortical white matter. However, for Amygdala, volumetric changes were greater in weeks 2 and 12.Both these changes help in the early monitoring of the effectiveness of the therapy. |
| 2898 | Computer 19
|
Prediction of Genomic Signature of Prostate Lesion Radiosensitivity by mpMRI Radiomics and Machine Learning |
| Evangelia I Zacharaki1, Mohammad Alhusseini 1, Adrian L Breto1, Isaac L Xu1, Ahmad Algohary1, Wendi Ma 1, Sandra M Gaston 1, Matthew C Abramowitz 1, Alan Dal Pra 1, Sanoj Punnen1, Alan Pollack 1, and Radka Stoyanova 1 | ||
1University of Miami, Miami, FL, United States |
||
Keywords: Radiomics, Prostate, multi-parametric MRI, prostate cancer radiosensitivity, genomic siganture, PORTOS Genomic classifiers, such as PORTOS, have shown great promise in the prediction of prostate cancer radiosensitivity. However, the spatial heterogeneity of prostate cancer may confound genomic assessment due to tumor sampling error. We aimed to develop a model predictive of PORTOS genomic signature using multiparametric MRI (mpMRI) radiomics features and machine learning. Lesions were localized based on Habitat Risk Score maps. Eight radiomic features were selected (out of 167) including T2, ADC, high B-value intensity and texture variables and used to build logistic regression models through cross-validation. Our analysis shows association between the radiomics profile and prostate lesion radiosensitivity phenotype. |
| 2899 | Computer 21
|
Six-fold enhancement in spatial-resolution of Pseudo-Continuous Arterial Spin Labeling Perfusion Mapping using a Cryogenic Coil at 9.4T |
| Sara Pires Monteiro1,2, Lydiane Hirschler3,4, Emmanuel L. Barbier4, Patrícia Figueiredo2, and Noam Shemesh1 | ||
1Champalimaud Research, Champalimaud Foundation, Lisbon, Portugal, 2Institute for Systems and Robotics - Lisboa and Department of Bioengineering, Instituto Superior Técnico, Universidade de Lisboa, Lisbon, Portugal, 3C.J. Gorter MRI Center for High Field MRI, Department of Radiology, Leiden University Medical Center, Leiden, Netherlands, 4Inserm, Grenoble Institut des Neurosciences, Université Grenoble Alpes, Grenoble, France |
||
Keywords: Data Acquisition, Arterial spin labelling, Preclinical High resolution CBF mapping using ASL can benefit multitude applications, yet its spatiotemporal resolution is limited in pre-clinical settings and at higher fields. To address this issue, we explored the possibility of improving the resolution of pCASL by using a cryogenic coil at 9.4T in the rat brain. Compared to the current state-of-the-art measurements, we managed to enhance the resolution of pCASL images by a factor of at least 6. The interpulse phase optimizations applied at the labelling plane are crucial for higher and stable inversion efficiency. |
| 2900 | Computer 22
|
Denoising single and multi-delay 3D pCASL using SWIN Transformer |
| Qinyang Shou1, Chenyang Zhao1, Xingfeng Shao1, and Danny JJ Wang1 | ||
1Laboratory of Functional MRI Technology (LOFT), Stevens Neuroimaging and Informatics Institute, University of Southern California, Los Angeles, CA, United States |
||
Keywords: Machine Learning/Artificial Intelligence, Machine Learning/Artificial Intelligence We developed a Transformer-based deep learning denoising model to improve the SNR for both single and multi-delay perfusion images acquired using 3D pseudo-continuous arterial spin labeling (pCASL). This method can significantly improve SNR (~2-fold) of the perfusion images without introducing bias for CBF and ATT quantification for both single-delay and multi-delay 3D pCASL. Further training and testing of this model on clinical datasets acquired on different vendor platforms is warranted. |
| 2901 | Computer 23
|
Improved perfusion-weighted images at 7T combining pTx and low B1+ Adiabatic pulses |
| Ícaro Agenor Ferreira de Oliveira1,2, Robin Schnabel1, Matthias JP van Osch3, Lydiane Hirschler3, and Wietske van der Zwaag1,2 | ||
1Spinoza Centre for Neuroimaging, Amsterdam, Netherlands, 2Computational Cognitive Neuroscience and Neuroimaging, Netherlands Institute for Neuroscience, Amsterdam, Netherlands, 3Radiology, Leiden University Medical Center, Leiden, Netherlands |
||
Keywords: Data Acquisition, High-Field MRI A FAIR acquisition at 7T, using a TR-FOCI pulse for background suppression, was set up on a pTx (32Rx8Tx) system and on the standard (32Rx2Tx) system. CBF-imaging results were compared in terms of CBF-weighted signal intensities and drop-off with increasing thickness of the slab-selective inversion slab. The higher and more homogeneous B1 of the pTx system translated to higher CBF-weighted signal intensities. Both setups allowed functional mapping of the hand region in the motor cortex in approximately 5 minutes. |
| 2902 | Computer 24
|
Subject-Specific Circular Artifacts in ASL Perfusion Imaging with 3D Stack-of-Spirals FSE Readout |
| Yichen Hu1, Junpu Hu2, Zhongqi Zhang1, Hui Liu1, Qi Liu1, Yongquan Ye1, and Jian Xu1 | ||
1United Imaging, Houston, TX, United States, 2United Imaging, Shanghai, China |
||
Keywords: Artifacts, Arterial spin labelling In perfusion images of ASL scans of brain, artifacts originated from unlabeled blood inflow can be effectively suppressed by inferior saturation. In our study with spiral-based readout, artifacts in the patterns of bright spots and concentric rings were determined to arise from arterial blood inflow. We discovered that the emergence of the artifacts is directly related to the carotid artery anatomies of human subjects. The artifacts appeared persistently for just one out of four participating volunteers. The application of inferior saturation should be favored to avoid such artifacts, which had been a fact not fully recognized from recent ASL studies. |
| 2903 | Computer 25
|
Deep Learning-denoised Isotropic 2mm Whole Brain pseudo-Continuous Arterial Spin Labeling at 7T |
| Chenyang Zhao1, Qinyang Shou1, Xingfeng Shao1, and Danny JJ Wang1,2 | ||
1Mark & Mary Stevens Neuroimaging and Informatics Institute, Keck School of Medicine, University of Southern California, Los Angeles, CA, United States, 2Department of Neurology, Keck School of Medicine, University of Southern California, Los Angeles, CA, United States |
||
Keywords: Machine Learning/Artificial Intelligence, Arterial spin labelling, Noise reduction Optimized pseudo-Continuous Arterial Spin Labeling has been implemented at 7T. To achieve whole brain high-resolution (2mm isotropic) perfusion imaging at 7T, however, requires prolonged scan time with an increased number of segments. A deep learning (DL) model was trained to boost the signal-to-noise ratio (SNR) for a scan with fewer repetitions and thus a shorter scan time. The analysis of SNR and temporal SNR suggests that at least 3 repetitions are needed to make a high-SNR prediction comparable to the full scan without compromising quantification accuracy. With DL denoising, the original 12 mins scan can be finished in 4 min. |
| 2904 | Computer 26
|
Increasing temporal SNR, sharpness, specificity, and sensitivity of ASLfMRI using the partial separability model |
| Charles John Marchini1 and Brad Sutton 1 | ||
1Bioengineering, University of Illinois Urbana-Champaign, Urbana, IL, United States |
||
Keywords: Sparse & Low-Rank Models, Arterial spin labelling ASLfMRI temporal SNR (tSNR) and sharpness in the inferior-superior direction was improved by using partial separability, a low rank model. The method requires an additional novelty to the partial separable model which allows time points with no corresponding imaging data, only temporal navigator data, to be reconstructed. A finger tapping task was used to demonstrate the detection of cerebral blood flow to the motor cortex. Mean squared error, structural similarity index, and the area under the curve of a receiver operating characteristic curve was also improved as shown by using a simulation of ASLfMRI data. |
| 2905 | Computer 27
|
Accelerated Brain ASL using 3D Gradient and Spin-Echo pseudo-Continuous ASL (GraSE-pCASL) with Compressed SENSE. |
| Yutaka Hamatani1, Kayoko Abe2, Masami Yoneyama3, Johannes M Peeters4, Kim van de Ven4, Michinobu Nagao2, Yasuhiro Goto1, Isao Shiina1, Kazuo Kodaira1, Takumi Ogawa1, Mana Kato1, and Shuji Sakai2 | ||
1Department of Radiological Services, Tokyo Women's Medical University Hospital, Tokyo, Japan, 2Department of Diagnostic imaging & Nuclear Medicine, Tokyo Women's Medical University Hospital, Tokyo, Japan, 3Philips Japan, Tokyo, Japan, 4Philips Healthcare, Best, Netherlands |
||
Keywords: Data Acquisition, Arterial spin labelling In this study, 3D GraSE-pCASL was combined with Compressed SENSE (CS) to accelerate the acquisition time of perfusion images. The results showed that accelerated 3D CS-GraSE-pCASL could provide sufficient perfusion information in half the acquisition time compared to the conventional method. This technique may be useful in the diagnosis of cerebral blood flow disorders especially for pediatric patients and/or patients with cognitive impairments. |
| 2906 | Computer 28
|
Comparison of velocity-selective-inversion arterial spin labeling schemes |
| Ke Zhang1, Simon M.F. Triphan1, Oliver Sedlaczek1,2, Christian Ziener2, Hans-Ulrich Kauczor1, Heinz-Peter Schlemmer2, and Felix T. Kurz2 | ||
1Department of Diagnostic and Interventional Radiology, Heidelberg University Hospital, Heidelberg, Germany, 2Department of Radiology, German Cancer Research Center, Heidelberg, Germany |
||
Keywords: Pulse Sequence Design, Arterial spin labelling Velocity-selective pulses include VS saturation pulses (VSS) and VS Inversion (VSI) pulses. Previous study (1) concluded that both dual-sBIR8-VSS and sinc-VSI achieved the highest SNR efficiency among the VS labeling schemes. Overall, the dual-sBIR8-VSS pulse was the most robust against field imperfections, whereas sinc-modulated VSI pulse showed greater tSNR and was the best among the VSI methods. In this study, VSI sequence with rectangular small flip-angle RF pulses (rect-VSI), sinc-VSI with and without VS gradients during the control condition are compared. Bloch simulation and in vivo experiments for their robustness against B1, B0 variation and eddy current (EC) are investigated. |
| 2907 | Computer 29
|
Rapid Bayesian inference for perfusion quantification using ASL-MRI with a VAE-based neural network structure |
| Yechuan Zhang1, Michael Chappell2,3,4,5,6, and Jian-Qing Zheng1 | ||
1University of Oxford, Oxford, United Kingdom, 2Institute of Biomedical Engineering, Department of Engineering Science, University of Oxford, Oxford, United Kingdom, 3Wellcome Centre for Integrative Neuroimaging, FMRIB, Nuffield Department of Clinical Neuroscience, University of Oxford, Oxford, United Kingdom, 4Mental Health and Clinical Neuroscience, School of Medicine, University of Nottingham, Nottingham, United Kingdom, 5Sir Peter Mansfield Imaging Centre, School of Medicine, University of Nottingham, Nottingham, United Kingdom, 6Nottingham Biomedical Research Centre, Queen’s Medical Centre, University of Nottingham, Nottingham, United Kingdom |
||
Keywords: Data Analysis, Arterial spin labelling A Variational Autoencoder (VAE) based framework was created to solve perfusion parameter estimation problem for ASL non-linear forward models. The ultimate goal was to build up an efficient and uncertainty-aware framework for parameter estimation problem in medical imaging, using the concept from Variaitonal Bayes (VB) which was already applied to ASL. Evaluation was performed using simulation and real data experiments with a bi-exponential model and two ASL-MRI forward models with different complexity. Compared with Markov Chain Monte Carlo (MCMC) and analytical VB (aVB), our VAE-based model achieved comparable accuracy, and hundredfold improvement in computational time. |
| 2908 | Computer 30
|
Navigator-based slice tracking for liver pCASL using spin-echo EPI acquisition |
| Ke Zhang1, Simon M.F. Triphan1, Oliver Sedlaczek1,2, Christian Ziener2, Hans-Ulrich Kauczor1, Heinz-Peter Schlemmer2, and Felix T. Kurz2 | ||
1Department of Diagnostic and Interventional Radiology, Heidelberg University Hospital, Heidelberg, Germany, 2Department of Radiology, German Cancer Research Center, Heidelberg, Germany |
||
Keywords: Pulse Sequence Design, Arterial spin labelling Liver perfusion is an important physiological parameter in health and disease (1,2). In the measurement of liver perfusion using arterial spin labelling (ASL), respiratory motion is a major challenge. In this study, respiratory motion information is acquired from a projection signal and used to adjust the position of the excited slice in real time. The feasibility of free-breathing multi-slice liver perfusion imaging using spin-echo EPI based pseudo-continuous ASL (pCASL) with navigator-based slice tracking method is investigated. |
| 2909 | Computer 31
|
A novel Fourier-based perspective to analyze and compensate for corruption in arterial spin labeling MRI |
| Seon-Ha Hwang1 and Sung-Hong Park1 | ||
1Department of Bio and Brain Engineering, Korea Advanced Institute of Science and Technology, Daejeon, Korea, Republic of |
||
Keywords: Data Processing, Arterial spin labelling A new perspective in Fourier domain to understand the ASL signal and compensate for corruption artifacts is proposed. The averaged perfusion-subtraction image corresponds to a single frequency component and the other frequency components could be interpreted as non-perfusion signals. In this viewpoint, the multiple corruptions could be compensated by finding the optimal perfusion frequency component to minimize the zigzag amplitude in the non-perfusion signals. The proposed compensation reduced the corruption without eliminating any measured data, preserving SNR. This new perspective would open up the possibility of processing the ASL signal in the Fourier domain rather than the time domain. |
| 2910 | Computer 32
|
Combined Parkinson’s Disease Related Patterns using ASL MRI and FDG PET |
| Yu Zeng1, Zizhao Ju2, Weiying Dai3, Yong Zhang4, David Alsop5, Chuantao Zuo2, and Li Zhao1 | ||
1College of Biomedical Engineering & Instrument Science, Zhejiang University, Zhejiang, China, 2PET Center & National Center for Neurological Disorders & National Clinical Research Center for Aging and Medicine, Huashan Hospital, Fudan University, Shanghai, China, 3Department of Computer Science, State University of New York at Binghamton, Binghamton, NY, United States, 4GE Healthcare, Shanghai, China, 5Radiology, Beth Israel Deaconess Medical Center and Harvard Medical school,, Bostom, MA, United States |
||
Keywords: Data Analysis, Parkinson's Disease, Parkinson's disease-related pattern (PDRP), multimodality Parkinson's disease-related patterns (PDRPs) of metabolism and perfusion have been reported to reflect brain abnormalities in Parkinson’s disease (PD). However, differences between the glucose metabolism PDRP derived using FDG-PET and the perfusion PDRP derived using ASL have not been compared directly in the same patient cohort. In this work, PDRPs were compared using PET and ASL images of 43 PD patients and 28 health controls using Scaled Subprofile Model analysis. In addition, a new method was proposed to build a multi-modality pattern. Our primary results showed that combined metabolism-perfusion PDRP resulted in superior accuracy than ASL- and PET-derived PDRP alone. |
| 2911 | Computer 33
|
Identifying Brain Resting-state Networks from Arterial Spin Labeling with Spectral Clustering |
| Jason Barrett1, Haomiao Meng2, Zongpai Zhang1, Song Chen1, Li Zhao3, David Alsop3, Xingye Qiao2, and Weiying Dai1 | ||
1Computer Science, Binghamton University, Vestal, NY, United States, 2Mathematical Sciences, Binghamton University, Vestal, NY, United States, 3Department of Radiology, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, MA, United States |
||
Keywords: Data Analysis, fMRI (resting state), network detection We propose a spectral clustering algorithm (SCA) based on the Pearson correlation metric (SCA-PC) to identify large-scale brain networks in arterial spin labeling (ASL) images. It was shown to be more robust to Gaussian distributed noise sources based on simulations. We studied the robustness of SCA-PC vs. the traditional SCA method based on a Euclidean distance metric (SCA-ED) for deriving resting-state networks from real human fMRI data. Our results indicate that SCA-PC can derive better brain networks from ASL data than traditional SCA-ED. |
| 2912 | Computer 34
|
MT-DICE for Simultaneous Quantification of Perfusion, Permeability, and Susceptibility |
| Yang Chen1,2, Zhehao Hu1, Jiayu Xiao1, Mark Shiroishi1, Wensha Yang1, Debiao Li3, Anthony G. Christodoulou3, and Zhaoyang Fan1 | ||
1Department of Radiology, Keck School of Medicine, University of Southern California, Los Angeles, CA, United States, 2Department of Biomedical Engineering, University of Southern California, Los Angeles, CA, United States, 3Cedars-Sinai Medical Center, Los Angeles, CA, United States |
||
Keywords: Quantitative Imaging, Permeability, perfusion, susceptibility We have recently developed an MR MultiTasking based Dynamic Imaging for Cerebrovascular Evaluation (MT-DICE) technique that provides DCE- and leakage-corrected DSC-MRI parameters simultaneously with one scan and a single-dose contrast injection. This work further expanded the technique by incorporating quantitative susceptibility mapping. The refined technique was validated in healthy volunteers and patients, demonstrating the good agreement with reference methods. |
| 2913 | Computer 35
|
Intensity-based Deep Learning for SPION concentration estimation in MR imaging |
| Alberto Di Biase1,2, Shuang Liu3, Masaki Sekino3, and Pablo Irarrázabal1,2 | ||
1Department of Electrical Engineering, Pontificia Universidad Católica de Chile, Santiago, Chile, 2Biomedical Imaging Center, Pontificia Universidad Católica de Chile, Santiago, Chile, 3Department of Bioengineering, School of Engineering, University of Tokyo, Tokyo, Japan |
||
Keywords: Machine Learning/Artificial Intelligence, Quantitative Imaging, SPION SPION is a contrast agent with a wide range of biomedical applications. A new Deep Learning based method is presented for the quantification of SPION from intensity images. This contrast agent cause off-resonance artifacts, distorting the image. The field map is encoded in the difference of two images taken alternating the direction of the slice selection gradient. The network was trained on simulated data. The network is based on U-net and uses only 2D convolution to process the whole 3D volume, interpreting the last dimension as filters. Results are shown in simulations and on phantoms acquired on a 7T scanner. |
| 2914 | Computer 36
|
Resolution Dependency Study of Cerebral Blood Volume and Vessel Size Index |
| Yelim Gong1, DongKyu Lee1, and HyungJoon Cho1 | ||
1BME, UNIST, Ulsan, Korea, Republic of |
||
Keywords: Data Analysis, Blood vessels To obtain cerebral vasculature related information, various kinds of cerebrovascular magnetic resonance imaging (MRI) technique are being applied. However, it turned out that cerebral blood volume (CBV) and vessel size index (VSI) values differ depending on their spatial resolutions. To find an adjustment solution for this phenomenon, the resolution dependency must be confirmed in advance. Therefore, this study aims to investigate the resolution dependency of CBV and VSI through whole brain and region of interest based analysis. The results show that micro CBV and VSI values are resolution-dependent, while total CBV did not show any distinct patterns. |
| 2915 | Computer 37
|
3D Artificial Cerebral Blood Volume Generation from T1W Structural MRI |
| Vishwanatha Mitnala Rao1, Scott A Small2, and Jia Guo3 | ||
1Biomedical Engineering, Columbia University, Acton, MA, United States, MA, United States, 2Department of Neurology, Columbia University Medical Center, New York, NY, United States, 3Department of Psychiatry, Columbia University, New York, NY, United States |
||
Keywords: Machine Learning/Artificial Intelligence, Brain While gadolinium-based contrast agents are necessary to generate a quantitative mapping of brain metabolism, they are invasive with unclear long-term side-effects. As such, convolutional neural networks (CNNs) have been explored as a method to generate artificial cerebral blood volume (aCBV) maps from T1W structural MRI scans. However, prior implementations process MRI in 2D slices, severely limiting output resolution, production time, and utility. In this study, we propose a 3D CNN-Transformer hybrid aCBV generation tool that outperforms both 2D and 3D implementations of the prior state-of-the-art model (PSNR: 29.46, P.R.: 0.836, SSIM: 0.875, S.R.: 0.681). |
| 2916 | Computer 38
|
Comparison of Blood Flow Measurement of Posterior Cerebral Circulation at Vertebral and Basilar Arteries Using Phase Contrast MRI |
| Feng Xu1,2 and Qin Qin1,2 | ||
1The Russell H. Morgan Department of Radiology and Radiological Science, Johns Hopkins University, Baltimore, MD, United States, 2F.M. Kirby Research Center for Functional Brain Imaging, Kennedy Krieger Institute, Baltimore, MD, United States |
||
Keywords: Data Acquisition, Velocity & Flow Phase contrast (PC) MRI provides a quantitative measurement of the total flux of blood flow. As it is not easy to find a perpendicular plane to the vertebral arteries (VA) at the base of the brain, especially in the elderly and people who developed tortuous arteries, we explored the alternative imaging location at the basilar artery (BA) where two VA merged in front of the pon. The flux of BA shows an excellent correlation (ρ=0.9, P<0.01) with that of VA, suggesting that BA is a practical choice for measuring the global blood flow of posterior cerebral circulation using PC MRI. |
| 2917 | Computer 39
|
Visualization of Collateral Vessels in 4D TRANCE MR Angiography Compaired with 3D TOF in Moyamoya Disease |
| Zhihua Chen1, Kaiyin Liang1, Yongkun Lan1, and Wen Zhou1 | ||
1PEKING UNIVESITY SHENZHENG HOSPITAL, Shen Zheng, China |
||
Keywords: Visualization, Blood vessels Compared with 3D-CE-MRA, TOF technology isn't conducive to display slow blood flow. Moreover, TOF angiography is a static approach that lacks flow dynamic information. In our study, the 4D TRANCE MRA depicted the proximal arteries better than 3D TOF and may replace TOF angiography. Our study also indicated that LMA collaterals were better visualized in 4D TRANCE angiography and visualization of distal cerebral arteries and collateral vessels with Moyamoya disease with 4D TRANCE MR angiography was good to excellent. Besides conventional MRA to show the shape and distribution of blood vessels, it can also provide hemodynamic information of vascular diseases. |
| 2918 | Computer 40
|
MRI evidence for reduced human brain oxygen metabolism during midazolam sedation |
| Hannah L Chandler1, Ian Driver2, Sharmila Khot1, Zoltán Auer3, Murthy Varanasi3, Neeraj Saxena1,3, Michael Germuska2, and Richard G Wise1,4,5 | ||
1Cardiff University Brain Research Imaging Centre (CUBRIC), Department of Psychology, Cardiff University, Cardiff, United Kingdom, 2Cardiff University Brain Research Imaging Centre (CUBRIC), School of Physics and Astronomy, Cardiff University, Cardiff, United Kingdom, 3Department of Anaesthetics, ICU and Pain medicine, Cwm Taf Morgannwg University Health Board, Merthyr Tydfil, United Kingdom, 4Neurosciences, Imaging and Clinical Sciences, University “G. d'Annunzio” of Chieti-Pescara, Chieti, Italy, 5Institute for Advanced Biomedical Technologies (ITAB), G. d’Annunzio University’ of Chieti-Pescara, Chieti, Italy |
||
Keywords: Data Analysis, Arterial spin labelling We investigated the effects of mild sedation with the type A GABA receptor positive allosteric modulator (GABAA-R PAM) midazolam, on cerebral blood flow (CBF), oxygen extraction fraction (OEF) and the rate of cerebral metabolic oxygen consumption (CMRO2) in the healthy human brain using ASL and TRUST MRI. CMRO2 was significantly reduced during sedation compared to wakefulness but no statistically significant change in CBF or OEF was detected. Our data are consistent with prior imaging evidence of reduced brain energy consumption during sedation with midazolam. |
| 2919 | Computer 41
|
DC-Swin: Deep Cascade of Swin Transformer with Sensitivity Map for Parallel MRI Reconstruction |
| Naoto Fujita1 and Yasuhiko Terada1 | ||
1Graduate School of Science and Technology, University of Tsukuba, Tsukuba, Japan |
||
Keywords: Image Reconstruction, Machine Learning/Artificial Intelligence Deep learning (DL) reconstruction networks are predominantly architectures that unroll traditional iterative algorithms and tend to perform better than non-unrolled models. Both types of models use convolutional neural networks (CNNs) as building blocks, but CNNs have the disadvantage of focusing on local relationships in the image. To overcome this, hybrid models have been proposed that combine CNNs with Transformers that focus on long-range dependencies. However, these hybrid transformers have been limited to non-unrolled reconstruction networks. Here, we propose an unrolled reconstruction network using a hybrid Transformer, Deep Cascade of Swin Transformer (DC-Swin), and verify that DC-Swin has high performance. |
| 2920 | Computer 42
|
2.5D Networks for Physics-Guided Deep Learning Reconstruction of 3D Non-Cartesian MRI from Limited Training Data |
| Chi Zhang1,2, Davide Piccini3,4, Omer Burak Demirel1,2, Gabriele Bonanno4, Steen Moeller2, Burhaneddin Yaman1,2, Matthias Stuber3,5, and Mehmet Akçakaya1,2 | ||
1Department of Electrical and Computer Engineering, University of Minnesota, Minneapolis, MN, United States, 2Center for Magnetic Resonance Research, University of Minnesota, Minneapolis, MN, United States, 3Department of Diagnostic and Interventional Radiology, Lausanne University Hospital and University of Lausanne, Lausanne, Switzerland, 4Advanced Clinical Imaging Technology, Siemens Healthineers International, Lausanne, Switzerland, 5Center for Biomedical Imaging, Lausanne, Switzerland |
||
Keywords: Image Reconstruction, Machine Learning/Artificial Intelligence Although recent studies enabled physics-guided deep learning (PG-DL) reconstruction of 3D non-Cartesian MRI, it suffers from blurring, partially due to limited training data. In this study we propose 2.5D PG-DL using three 2D CNNs on orthogonal views for 3D reconstruction to efficiently exploit the limited training data. Results on 3D kooshball coronary MRI show the proposed strategy noticeably improves image sharpness. |
| 2921 | Computer 43
|
Greedy Learning for Memory-Efficient Self-Supervised MRI Reconstruction |
| Arushi Gupta1, Batu M Ozturkler2, Arda Sahiner2, Tolga Ergen2, Arjun D Desai2,3, Shreyas Vasanawala3, John M Pauly2, Morteza Mardani2, and Mert Pilanci2 | ||
1California Institute of Technology, Pasadena, CA, United States, 2Electrical Engineering, Stanford University, Stanford, CA, United States, 3Radiology, Stanford University, Stanford, CA, United States |
||
Keywords: Image Reconstruction, Image Reconstruction Deep learning (DL) has recently shown state-of-the-art performance for accelerated MRI reconstruction. However, supervised learning requires fully-sampled training data, and training these networks with end-to-end backpropagation requires significant memory for high-dimensional imaging. These challenges limit the use of DL in high-dimensional settings where access to fully-sampled data is unavailable. Here, we propose self-supervised greedy learning for memory-efficient MRI reconstruction without fully-sampled data. The method divides the end-to-end network into smaller network modules and independently calculates a self-supervised loss for each subnetwork. The proposed method generalizes as well as end-to-end learning without fully-sampled data with at least 7x less memory usage. |
| 2922 | Computer 44
|
High Resolution MR Reconstruction with Functionally Separate Neural Networks |
| Hideaki Kutsuna1, Shun Uematsu2, and Kensuke Shinoda2 | ||
1MRI Systems Development Department, Canon Medical Systems Corporation, Kanagawa, Japan, 2MRI Systems Development Department, Canon Medical Systems Corporation, Tochigi, Japan |
||
Keywords: Image Reconstruction, Machine Learning/Artificial Intelligence, Super resolution The authors propose a new reconstruction method to obtain higher resolution images from an MR acquisition. The method incorporates MR physics and two neural networks, which are functionally separate, for denoising and upsampling. The proposed method was evaluated by applying it to both retrospectively and prospectively undersampled data. The result showed that the proposed technique is capable of reconstructing higher resolution images over a conventional method, by multiplying the matrix size while keeping more detail structure in the originally sampled data. |
| 2923 | Computer 45
|
Improving variational network based 2D MRI reconstruction via feature-space data consistency |
| Ilias Giannakopoulos1, Patricia Johnson1, Riccardo Lattanzi1, and Matthew Muckley2 | ||
1The Bernard and Irene Schwartz Center for Biomedical Imaging (CBI), Department of Radiology, New York University Grossman School of Medicine, New York, NY, United States, 2Facebook AI Research, Meta, New York, NY, United States |
||
Keywords: Image Reconstruction, Parallel Imaging, Compressed Sensing, Deep Learning Deep learning (DL) methods have enabled state-of-the-art reconstructions of magnetic resonance images of highly undersampled acquisitions. The end-to-end variational network (E2E VarNet) is a DL method that can output high quality reconstructions through an unrolled gradient descent algorithm. Nevertheless, the network discards a lot of high-level feature representations of the image to perform data consistency in the image space. Here, we adapted the E2E VarNet architecture to perform the data consistency in a feature space. We trained the proposed network using the fastMRI brain dataset and observed 0.0013 SSIM improvement for eight-fold accelerations.
|
| 2924 | Computer 46
|
Accelerate Single-Channel 3D MRI through Undersampling and Deep Neural Network Reconstruction |
| Christopher Man1,2, Vick Lau1,2, Jiahao Hu1,2, Junhao Zhang1,2, Linfang Xiao1,2, and Ed X. Wu1,2 | ||
1Laboratory of Biomedical Imaging and Signal Processing, The University of Hong Kong, Hong Kong SAR, China, 2Department of Electrical and Electronic Engineering, The University of Hong Kong, Hong Kong SAR, China |
||
Keywords: Image Reconstruction, Image Reconstruction 3D MRI data contains more redundant information than 2D MRI data, which is favourable for reconstruction. However, deep learning reconstruction of 3D MRI data remains to be explored due to the computational burden that scales exponentially with spatial dimensions. This study presents a deep learning method to reconstruct single-channel 3D MRI data with uniform undersampling along two phase-encoding directions, in which conventional multi-channel parallel imaging methods are generally not applicable. The results demonstrate the robust reconstruction for single-channel 3D MRI data at high acceleration and in the presence of anomaly. |
| 2925 | Computer 47
|
Deep Unfolding MR reconstruction - weighting the k-space sampling Density in training Loss (wkDeLo) |
| Parna Eshraghi Boroojeni1, Wejie Gan2, Jiaming Liu2, Yuyang Hu2, Yasheng Chen3, Paul Commean4, Cihat Eldeniz5, Tongyao Wang6, Gary Skolnick7, Corinne Merrill7, Kamlesh Patel7, Hongyu An6, and Ulugbek Kamilov8 | ||
1Washington University in Saint Louis, Saint Louis, MO, United States, 2Engineering - Computer Science and Engineering, Washington University in Saint Louis, Saint Louis, MO, United States, 3Neurology, Washington University in Saint Louis, Saint Louis, MO, United States, 4Radiology - Main - Research - Radiological Sciences, Washington University in Saint Louis, Saint Louis, MO, United States, 5Radiology - Research Imaging Facilities - MR Facility, Washington University in Saint Louis, Saint Louis, MO, United States, 6Radiology - Main - Research - Radiological Sciences - Biomedica, Washington University in Saint Louis, Saint Louis, MO, United States, 7Surgery - Plastics, Washington University in Saint Louis, Saint Louis, MO, United States, 8Engineering - Electrical & Systems Engineering, Washington University in Saint Louis, Saint Louis, MO, United States |
||
Keywords: Image Reconstruction, Image Reconstruction We develop a self-supervised and physics-guided deep unfolding (DU) network for MR image reconstruction by weighting the k-space sampling Density in network training Loss (wDeLo). We have demonstrated that high-quality MR images at a spatial resolution of 0.6x0.6x0.8 mm3 could be achieved using an acquisition time of 1 minute (x6.25 acceleration) or 45 seconds (8x acceleration). Compared to SSDU and its uniform weighted counterpart (un-wDeLo), the wDeLo method significantly improves PSNR and SSIM. It had fewer artifacts, lower noise, and preserved image sharpness. |
| 2926 | Computer 48
|
A Regularized Conditional GAN for Posterior Sampling in MR Image Reconstruction |
| Matthew Charles Bendel1, Rizwan Ahmad2, and Philip Schniter1 | ||
1Dept. ECE, The Ohio State University, Columbus, OH, United States, 2Dept. BME, The Ohio State University, Columbus, OH, United States |
||
Keywords: Image Reconstruction, Machine Learning/Artificial Intelligence For magnetic resonance (MR) image reconstruction, Fourier-domain measurements are collected at rates far below Nyquist to reduce clinical exam time. Because many plausible reconstructions exist that are consistent with a given measurement, we use machine learning to sample from the posterior distribution rather than generate a single image reconstruction. Many such works leverage score-based generative models (SGMs), which seek to iteratively denoise a random input but require many minutes to generate each sample. We propose a conditional generative adversarial network (GAN) that generates hundreds of posterior samples per minute and outperforms the current state-of-the-art SGM for multi-coil MR posterior sampling. |
| 2927 | Computer 49
|
Acquisition Adaptive Unrolled Deep Learning Framework for Parallel MRI |
| Aniket Pramanik1, Sampada Bhave2, Saurav Sajib2, Samir Sharma2, and Mathews Jacob1 | ||
1University of Iowa, Iowa City, IA, United States, 2Canon Medical Research USA, Mayfield Villlage, OH, United States |
||
Keywords: Image Reconstruction, Machine Learning/Artificial Intelligence Model-based Deep Unrolled Networks offer high quality reconstructions but the performance degrades if any mismatch occurs in acquisition settings. Different networks for different acquisition settings require extensive training data, in addition to making the clinical deployment difficult. We propose a single unrolled deep-learning algorithm called as Ada-MoDL, whose parameters are conditioned on the acquisition information (metadata) of a dataset, using a Multi-Layer Perceptron (MLP) that maps the metadata to network parameters. Ada-MoDL outperforms models trained for a specific acquisition setting or a single model trained with all the available contrasts, when the training data in each acquisition setting is limited. |
| 2928 | Computer 50
|
Posterior Sampling for Accelerated Multicoil MRI Reconstruction using a Conditional Normalizing Flow |
| Jeffrey Wen1, Rizwan Ahmad2, and Philip Schniter1 | ||
1Electrical Engineering, The Ohio State University, Columbus, OH, United States, 2Biomedical Engineering, The Ohio State University, Columbus, OH, United States |
||
Keywords: Image Reconstruction, Machine Learning/Artificial Intelligence For accelerated MR image reconstruction, machine learning (ML)-based methods outperform traditional sparsity-based methods by exploiting large datasets to learn effective priors. However, most ML methods output only a single image reconstruction when in fact there may be many plausible reconstructions given the measurement and prior. To extract this diagnostically relevant information, we propose to explore the space of plausible images, i.e., to sample the posterior, using ML. Among ML methods, conditional normalizing flows (CNFs) stand out for rapid sample generation and simple likelihood-based training. In this work, we present the first CNF for posterior sample generation in accelerated multicoil MRI. |
| 2929 | Computer 51
|
MixRecon: A neural network mixing CNN and Transformer utilizes hybrid representations of image features for Accelerated MRI Reconstruction |
| Hongjian Kang1, Liping Zhang1, and Weitian Chen1 | ||
1Department of Imaging and Interventional Radiology, The Chinese University of Hong Kong, Hong Kong, Hong Kong |
||
Keywords: Image Reconstruction, Machine Learning/Artificial Intelligence Accelerated reconstruction for magnetic resonance imaging (MRI) is a challenging ill-posed problem because of the excessive under-sampling operation in k-space. Existing CNN-based and Transformer-based solutions face difficulties obtaining powerful representation due to relatively unitary local or global feature modeling capability. In this study, we develop a dual-branch network that can simultaneously exploit the complementarity of the two-style features by leveraging the merits of CNN and Transformer, to generate high-quality reconstruction from zero-filled images in the spatial domain. Qualitative and quantitative results from the fastMRI dataset demonstrate that the proposed method can achieve improved performance compared with other benchmark methods. |
| 2930 | Computer 52
|
Model-based neural ODE network for parallel MRI reconstruction |
| Jun-Hyeok Lee1, Gawon Lee1, and Se-Hong Oh1 | ||
1Department of Biomedical Engineering, Hankuk University of Foreign Studies, Yongin, Korea, Republic of |
||
Keywords: Image Reconstruction, Parallel Imaging, Neural ODE In parallel MRI, DNN-based models have recently outperformed conventional reconstruction techniques and can reconstruct high-quality MRI images, especially at high acceleration factors. We propose a model-based neural ODE network to reconstruct artifact-free MR images from under-sampled k-space data. We replaced the existing U-Net with a modified U-Net framework using neural ODEs with E2E-VarNet as the backbone. Our network solves unrolled iterations of reconstruction optimization with neural ODEs, and each neural ODE uses a gradient update step as a dynamics step. Our approach showed the improved reconstruction performance comparable to the SOTA method with few parameters. |
| 2931 | Computer 53
|
Multicoil Deep Equilibrium MRI Reconstruction |
| Muntasir Shamim1, Chenwei Tang2, Leonardo Rivera2, and Kevin M Johnson2 | ||
1Electrical and Computer Engineering, University of Wisconsin Madison, Madison, WI, United States, 2Medical Physics, University of Wisconsin Madison, Madison, WI, United States |
||
Keywords: Image Reconstruction, Brain Deep equilibrium (DEQ) image reconstruction models have demonstrated improved performance while remaining memory efficient for single coil MRI reconstructions; however, DEQ models are known to be susceptible to model drift due to variabilities in the forward model. In the case of multicoil MRI, the forward model is changed due to the sampling patterns and coil sensitivity encoding. In this work, we implemented a multicoil DEQ model and analyzed model drift with respect to the coil sensitivities. |
| 2932 | Computer 54
|
Recovery with self-calibrated denoisers from multiple undersampled images (ReSiDe-M) |
| Sizhuo Liu1, Philip Schniter1, and Rizwan Ahmad1 | ||
1The Ohio State University, Columbus, OH, United States |
||
Keywords: Image Reconstruction, Machine Learning/Artificial Intelligence Recovery with a self-calibrated denoiser (ReSiDe) is an unsupervised learning method based on the plug-and-play (PnP) framework. In ReSiDe, denoiser training and a call to the denoising subroutine are performed in each iteration of PnP. However, ReSiDe is computationally slow, and its performance is sensitive to the noise level selected to train the denoiser. Here, we extend ReSiDe from single-image to multi-image recovery (ReSiDe-M), improving both performance and computation speed. We also propose an auto-tuning method to select the noise level for denoiser training. Using data from fastMRI and MRXCAT perfusion phantom, we compare ReSiDe-M with other unsupervised methods. |
| 2933 | Computer 55
|
Accelerating MRI Using Vision Transformer with Unpaired Unsupervised Training |
| Peizhou Huang1, Hongyu Li2, Ruiying Liu2, Xiaoliang Zhang1, Xiaojuan Li3, Dong Liang4, and Leslie Ying1,2 | ||
1Biomedical Engineering, State University of New York at Buffalo, Buffalo, NY, United States, 2Electrical Engineering, State University of New York at Buffalo, Buffalo, NY, United States, 3Program of Advanced Musculoskeletal Imaging (PAMI), Cleveland Clinic, Cleveland, OH, United States, 4Paul C. Lauterbur Research Center for Biomedical Imaging, SIAT CAS, Shenzhen, China |
||
Keywords: Image Reconstruction, Machine Learning/Artificial Intelligence, Unpaired unsupervised training, Transformer In this abstract, we propose a novel deep-learning reconstruction method that enables training with only unpaired undersampled k-space data without the ground truth. The network utilizes a statistical model for the undersampling artifacts to enable unsupervised learning, and the generative adversarial network to enable unpaired training. In addition, the physics model is incorporated into the transformer network by unrolling the underlying optimization problem. Experiment results based on the fastMRI knee dataset exhibit marked improvements over the existing state-of-the-art reconstructions. |
| 2934 | Computer 56
|
On Training Model Bias of Deep Learning based Super-resolution Frameworks for Magnetic Resonance Imaging |
| Mamata Shrestha1, Nian Wang2,3, and Ukash Nakarmi1 | ||
1Computer Science and Computer Engineering, University of Arkansas, Fayetteville, AR, United States, 2Radiology and Imaging Sciences, Indiana University School of Medicine, Indianapolis, IN, United States, 3Stark Neurosciences Research Institute, Indianapolis, IN, United States |
||
Keywords: Image Reconstruction, Data Processing, Image super-resolution, model bias, deep learning Deep Learning based image super-resolution methods are biased towards training data modeling. Generalizability of DL based super-resolution frameworks can be improved by introducing variability and diversity in the training data. |
| 2935 | Computer 57
|
Deep Learning-based MRI Reconstruction with Artificial Fourier Transform(AFT)-Net |
| Yanting Yang1, Andrew F. Laine1, and Jia Guo2,3 | ||
1Department of Biomedical Engineering, Columbia University, New York, NY, United States, 2Department of Psychiatry, Columbia University, New York, NY, United States, 3Mortimer B. Zuckerman Mind Brain Behavior Institute, Columbia University, New York, NY, United States |
||
Keywords: Image Reconstruction, Machine Learning/Artificial Intelligence Conventional medical image reconstruction methods are less parametric and lack generality due to random error and noise. A novel artificial Fourier transform (AFT) framework is developed which determines the mapping between k-space and i-space like DFT while can be fine-tuned with further training. The flexibility of AFT allows it to be simply incorporated into any existing deep learning network as learnable or static blocks. Reconstruction and denoising tasks are combined into a unified network that simultaneously enhances the image quality. AFT-Net achieves competitive results compared with other methods and proofs to be more robust to additional noise and contrast differences. |
| 2936 | Computer 58
|
When System Model meets Image Prior: An Unsupervised Deep Learning Architecture for Accelerated Magnetic Resonance Imaging |
| Ibsa Kumara Jalata1 and Ukash Nakarmi1 | ||
1Computer Science and Computer Engineering, University of Arkansas, Fayetteville, AR, United States |
||
Keywords: Image Reconstruction, Data Processing Unsupervised deep learning framework that integrates system priors using unrolled optimization and general image priors can reconstruct high quality Magnetic Resonance images comparable to supervised methods from highly undersampled k-space data. We develop an unsupervised deep learning framework that integrates system priors in MR acquisition and image priors to reconstruct high quality MR images from highly undersampled k-space data without using ground truth images. |
| 2937 | Computer 59
|
Self-Supervised MR Image Reconstruction with Stage-by-Stage Data Refinement |
| Xue Liu1, Cheng Li1, Yu Zhang1, Haoran Li1, Yeqi Wang1, Hairong Zheng1, and Shanshan Wang1,2,3 | ||
1Paul C. Lauterbur Research Center for Biomedical Imaging, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China, 2Peng Cheng Laboratory, Shenzhen, China, 3Guangdong Provincial Key Laboratory of Artificial Intelligence in Medical Image Analysis and Application, Guangdong Provincial People’s Hospital, Guangdong Academy of Medical Sciences, Guangzhou, China |
||
Keywords: Image Reconstruction, Machine Learning/Artificial Intelligence Deep learning-based, especially fully supervised learning-based, methods have shown unprecedented performance in MR image reconstruction. Fully sampled data are utilized as references to supervise the learning process. However, it is challenging to acquire fully sampled data in many real-world application scenarios. Unsupervised approaches are required. Here, we propose an iterative data refinement method for enhanced self-supervised MR image reconstruction. Different from Yaman's self-supervised learning method (SSDU), training data in our method are refined iteratively during model optimization to progressively eliminate the data bias between the undersampled reference data and fully sampled data. Better reconstruction results are obtained. |
| 2938 | Computer 60
|
Accelerated MRI Reconstruction Using a Lightweight Recurrent Transformer: ReconFormer |
| Pengfei Guo1, Yiqun Mei1, Jinyuan Zhou2, Shanshan Jiang2, and Vishal M. Patel1 | ||
1Department of Computer Science, Johns Hopkins University, Baltimore, MD, United States, 2Department of Radiology, Johns Hopkins University, Baltimore, MD, United States |
||
Keywords: Image Reconstruction, Image Reconstruction Accelerating MRI reconstruction process is a challenging ill-posed inverse problem due to the excessive under-sampling operation in k-space. While state-of-the-art algorithms have shown a great progress based on convolutional neural networks (CNN), transformers for MRI reconstruction has not been fully explored in the literature. We propose a recurrent transformer model, namely, ReconFormer, for MRI reconstruction which can iteratively reconstruct high fertility magnetic resonance images from highly under-sampled k-space data. We validate the effectiveness of ReconFormer on multiple datasets with different magnetic resonance sequences and show that it achieves significant improvements over the state-of-the-art methods with better parameter efficiency. |
| 2939 | Computer 61
|
Through-plane Super-resolution of Prostate MRI Using Diffusion Models |
| Kai Zhao1, Jiahao Lin1, and Kyung Hyun Hyun Sung1 | ||
1Department of Radiological Sciences, University of California, Los Angeles, Los Angeles, CA, United States |
||
Keywords: Data Analysis, Prostate, super resolution We proposed a generative models-based method for through-plane super-resolution of multi-slice MRI scans. We proposed a novel slice-profile transformation that synthesizes low/high through-plane resolution slices for model training without accessing to isotropic datasets. Expert ready study, visual and quantitative comparisons reveal our method produce superior super-resolution results in terms of sharpness, artifacts, noise level and overall image quality. |
| 2940 | Computer 62
|
Deep learning of 3D T2-weighted MRI provides support for arachnoid granulation hypertrophy in patients with Parkinson’s disease |
| Melanie Leguizamon1, Colin McKnight2, Jarrod Eisma1, Alex Song1, Jason Elenberger1, Daniel Claassen1, Manus Donahue1, and Kilian Hett1 | ||
1Department of Neurology, Vanderbilt University Medical Center, Nashville, TN, United States, 2Department of Radiology, Vanderbilt University Medical Center, Nashville, TN, United States |
||
Keywords: Data Analysis, Neurodegeneration We apply novel deep learning algorithms using non-contrasted T2-weighted MRI to test hypotheses regarding arachnoid granulation (AG) hypertrophy in patients with Parkinson’s disease (PD). Using this method, we identify AGs protruding into the superior sinus lumen, which may serve as a site of CSF egress implicated in the neurofluid clearance system. Results suggest a significant increase in AG volume in the parietal and frontal lobes of PD participants compared to age-matched healthy controls, potentially indicative of reduced neurofluid clearance efficiency secondary to macromolecular aggregation along the CSF circuit. |
| 2941 | Computer 63
|
Deep learning improves the estimation of fiber orientation distribution for tractography in the human |
| Zifei Liang1, Patryk Filipiak1, Steven H Baete1, Yulin Ge1, Leslie Ying2, and Jiangyang Zhang1 | ||
1Radiology, NYU Langone health, new york, NY, United States, 2the State University of New York, Buffalo, NY, United States |
||
Keywords: Visualization, Brain Connectivity, Diffusion MRI tractography Although diffusion MRI (dMRI) tractography can map brain connectivity non-invasively, accurate tractography in the human brain remains challenging due to inherent and technical limitations. In this study, we demonstrate a deep learning (DL) based approach for improving the estimation of fiber orientation distribution (FOD) from dMRI data. Trained with augmented whole brain tractography results from high-resolution dMRI data, the DL approach outperformed conventional FOD estimation methods in crossing fiber regions with dMRI data at spatial and angular resolutions comparable to routine clinical scans. The approach can potentially shorten the dMRI acquisition necessary for accurate tractography and connectome analysis. |
| 2942 | Computer 64
|
Automated hybrid approach to diagnose Parkinson's disease via deep learning and radiomics |
| Hongyi Chen1, Xueling Liu2, Yuxin Li1,2, Puyeh Wu3, and Daoying Geng1,2 | ||
1Academy for Engineering and Technology, Fudan University, Shanghai, China, 2Huashan Hospital, State Key Laboratory of Medical Neurobiology, Fudan University, Shanghai, China, 3GE Healthcare, Beijing, China |
||
Keywords: Data Analysis, Parkinson's Disease In this study, we constructed a hybrid machine learning model utilizing CNN and radiomics features based on NM-sensitive setMag images. The hybrid features improved the diagnostic performance in distinguishing PD patients from HC, as demonstrated in the SVM classifier, which demonstrated 95.7% accuracy, 92.9% sensitivity, and 100% specificity. The interpretability of the radiomics approach is better because radiomics features provide more interpretable biomarkers, while the CNN approach extracts deeper features from images. Furthermore, visualizing regions that influence classification decisions via saliency map can also enhance the interpretability of the CNN approach. |
| 2943 | Computer 65
|
Biophysical model estimation of white matter using machine learning combined dictionary-based non-negative linear least square optimization |
| Hesam Abdolmotalleby1 and Merry Mani1 | ||
1University of Iowa, Iowa City, IA, United States |
||
Keywords: Data Processing, Data Processing We estimated biophysical model parameters using a dictionary of diffusion signals. The dictionary is generated based on the set of biophysical model parameters. We then rotated this dictionary along fascicle direction derived from spherical deconvolution and tried to fit it to the diffusion signal using non-negative linear least square optimization to find the associated dictionary weights. These weights went through a deep learning network to find model parameters. Our approach shows reasonable accuracy and reliability for biophysical model estimation. It is also possible to extend our approach to more complex and general biophysical models to achieve higher specificity. |
| 2944 | Computer 66
|
Association of neuroanatomical changes with neuropsychological changes in Treated Adult HIV-Positive Patients |
| Ajin Joy1, Rajakumar Nagarajan1, Eric Daar2,3, Jhelum Paul1, Santosh K Yadav4, Mario Guerrera2, Paul M. Macey5, and M. Albert Thomas1 | ||
1Radiological Sciences, University of California, Los Angeles, Los Angeles, CA, United States, 2Division of HIV Medicine, Lundquist Institute at Harbor-UCLA Medical Center, Los Angeles, CA, United States, 3Medicine, University of California, Los Angeles, Los Angeles, CA, United States, 4Department of Radiology and Radiological Science, School of Medicine,, Johns Hopkins University, Baltimore, MD, United States, 5School of Nursing and Brain Research Institute, University of California, Los Angeles, Los Angeles, CA, United States |
||
Keywords: Data Processing, Brain This study was designed to use the structural MRI data to characterize volumetric changes in gray matter and white matter, and to determine variations in cortical thickness, in a group of HIV-infected individuals to understand how their brain structural changes are associated with their neuropsychological state compared to a group of healthy individuals with similar social, behavioral backgrounds. Despite the demographic similarities and cART, we observed reduced gray and white matter volumes, as well as altered cortical thickness in HIV-infected participants compared with healthy controls. In addition, the neuroanatomic changes in HIV-infected patients showed statistically significant correlations with memory scores. |
| 2945 | Computer 67
|
The effects of brain diffusion MRI preprocessing pipelines in a clinical study of neurodegenerative disease |
| Kouhei Kamiya1,2, Sayori Hanashiro3, Osamu Kano3, Koji Kamagata2, Masaaki Hori1,2, and Shigeki Aoki2 | ||
1Department of Radiology, Toho University, Faculty of Medicine, Tokyo, Japan, 2Department of Radiology, Juntendo University, Faculty of Medicine, Tokyo, Japan, 3Department of Neurology, Toho University, Faculty of Medicine, Tokyo, Japan |
||
Keywords: Data Processing, Diffusion/other diffusion imaging techniques, Image preprocessing The impacts of different diffusion MRI preprocessing pipelines on the statistical results of clinical studies were investigated in amyotrophic lateral sclerosis. The results illustrate the influences of popular preprocessing tools on the effect size of the disease in DKI metrics. |
| 2946 | Computer 68
|
Generative diffusion model for DSC MRI artifact correction |
| Muhammad Asaduddin1, Eung Yeop Kim2, and Sung-Hong Park1 | ||
1Bio and Brain Engineering, Korea Advanced Institute of Science and Technology, Daejeon, Korea, Republic of, 2Department of Radiology, Samsung Medical Center, Sungkyunkwan University College of Medicine, Seoul, Korea, Republic of |
||
Keywords: Data Processing, DSC & DCE Perfusion Dynamic susceptibility contrast (DSC) MRI may suffer from artifacts due to long acquisition time. Past methods are limited in their performance and may change the contrast passage timing. In this work, we present a generative diffusion model that can restore signal loss and movement artifacts. We showed the generated DSC MRI images to have proper post-contrast vessel and grey matter structure with accurate contrast agent arrival/washout timing. The brain shape was also accurately generated as shown by the DICE score. This approach could provide a solution to restore a corrupted DSC MRI data while maintaining accurate contrast passage timing. |
| 2947 | Computer 69
|
Image distortion correction for diffusion MRI using U-Net and Transformer |
| Tsuyoshi Ueyama1,2,3, Erika Takahashi1, Naoto Fujita1, Yuichi Suzuki2, Hideyuki Iwanaga2, Osamu Abe4, and Yasuhiko Terada1 | ||
1Graduate School of Science and Technology, University of Tsukuba, Tsukuba, Ibaraki, Japan, 2Radiology center, The University of Tokyo Hospital, Tokyo, Japan, 3School of Medicine, Stanford University, Palo Alto, CA, United States, 4Department of Radiology, The University of Tokyo Hospital, Tokyo, Japan |
||
Keywords: Data Processing, Machine Learning/Artificial Intelligence, Diffusion weighted image/Diffusion tensor image Although several end-to-end deep neural networks have been proposed to correct image distortion directly from distorted images, no study has verified the distortion correction performance for high b-values diffusion-weighted image (DWI) and diffusion tensor image (DTI) parameters. For example, the U-Net-based Synb0-DisCo was only validated for distortion correction of b0 images. Here, we used two networks, U-Net and Trans-DisCo, to verify distortion correction performance for DWIs and DTI parameter images. Trans-DisCo is our proposed model that replaces the convolutional neural network in U-Net with Swin Transformer, and we have shown that it outperforms U-Net. |
| 2948 | Computer 70
|
MP-PCA image denoising technique for high resolution quantitative susceptibility mapping (QSM) of the human brain in vivo |
| Liad Doniza1, Neta Stern2, Dvir Radunsky2, Coral Helft3, Patrick Fuchs4, Anita Karsa4, Karin Shmueli4, and Noam Ben-Eliezer2,3,5 | ||
1Department of Electrical Engineering, Tel Aviv University, Tel Aviv, Israel, 2The Department of Biomedical Engineering, Tel-Aviv University, Tel-Aviv, Israel, 3Sagol School of Neuroscience, Tel-Aviv University, Tel-Aviv, Israel, 4Department of Medical Physics and Biomedical Engineering, University College London, London, United Kingdom, 5Center for Advanced Imaging Innovation and Research (CAI2R), New-York University Langone Medical Center, New York, NY, United States |
||
Keywords: Data Processing, Susceptibility, Denoising Quantitative susceptibility mapping (QSM) has many clinical applications such as distinguishing between acute and chronic multiple sclerosis (MS) lesions, and probing microbleeds in traumatic brain injury. High scan resolutions improve diagnostic quality and reduce partial volume artifacts albeit at a price of a lower signal-to-noise ratio (SNR). In this study, we introduce a principal component analysis (PCA) denoising algorithm for QSM data, showing the ability to generate QSM maps of the human brain at 0.6x0.6x0.6 mm3 resolution at 3T in vivo. |
| 2949 | Computer 71
|
Manifold Learning and Dimensionality Estimation for the Human Functional Connectome |
| Javier Gonzalez-Castillo1, Isabel Fernandez1, Daniel A Handwerker1, Ka-Chun Lam2, Francisco Pereira2, and Peter A Bandettini1,3 | ||
1Section on Functional Imaging Methods, NIMH, Bethesda, MD, United States, 2Machine Learning Team, NIMH, Bethesda, MD, United States, 3FMRI Core Facility, NIMH, Bethesda, MD, United States |
||
Keywords: Machine Learning/Artificial Intelligence, fMRI, Manifold Learning, TSNE, UMAP, Laplacian Eigenmaps, Intrinsic Dimension The exploration of time-varying aspects of the human functional connectome (FC) is challenging because the high dimensionality of connectivity matrices precludes direct visual inspection in a meaningful manner. Dimensionality reduction helps circumvent this problem, yet effective application requires a-priori knowledge of the intrinsic dimension of the data and careful selection of algorithmic hyperparameters. Here, we first estimate the intrinsic dimension of FC data. Next, we use data with known cognitive state changes to evaluate the effectiveness of Laplacian Eigenmaps, T-SNE and UMAP to generate informative low dimensional representations of time-varying FC data for explorative and predictive purposes. |
| 2950 | Computer 72
|
Realizing a robust multi-task training strategy with deep learning: application in Spine MR image quality assessment |
| Deepa Anand1, Dattesh Shanbhag1, Chitresh Bhushan2, and Uday Patil3 | ||
1GE Healthcare, Bangalore, India, 2GE Healthcare, Niskayuna, NY, United States, 3GE Healthcaer, Bangalore, India |
||
Keywords: Machine Learning/Artificial Intelligence, Machine Learning/Artificial Intelligence Multi-task training is attractive in AI applications (from memory, processing time and potential data reduction), where tasks have commonality in terms of features, but still requires differentiation for individual outputs derived. In this work we present a methodology to implement robust multi-task learning framework considering various strategies (parallel, iterative and sequential). We tested the approach for image quality assessment of spine MRI localizer images. We demonstrate that the sequential training is the most effective, in preserving an accuracy above the acceptable level while allowing for a save in number of model parameters (50%). |
| 2951 | Computer 73
|
Respiratory Motion Artifact Simulation for DL application in Cardiac MR Image Quality Assessment |
| Sina Amirrajab1, Yasmina Al Khalil1, Josien Pluim1, Marcel Breeuwer1,2, and Cian M. Scannell1 | ||
1Biomedical Engineering, Eindhoven University of Technology, Eindhoven, Netherlands, 2MR R&D - Clinical Science, Philips Healthcare, Eindhoven, Netherlands |
||
Keywords: Machine Learning/Artificial Intelligence, Artifacts, Respirator Artifact Simulation To tackle data scarcity for training a deep-learning algorithm for cardiac MR image quality assessment, we develop a k-space method for simulating respiratory motion artifacts with different levels of severity on artifact-free publicly available cardiac MRI data. The benefit of such simulated data is investigated, demonstrating the usefulness of training a feature extractor with the simulated artifacts for image quality classification. Our proposed method achieved the test accuracy of 0.625 and Cohen's Kappa of 0.473 (n=120 images), ranking third in task one for the CMRxMotion challenge of MICCAI 2022. |
| 2952 | Computer 74
|
Deep Learning based prediction of the planes for automated planning of MRI imaging of cervical neural foramina and lumbar pars interarticularis |
| Chitresh Bhushan1, Dattesh D Shanbhag2, Uday Patil2, Trevor Kolupar3, and Maggie Fung3 | ||
1AI and Medical Imaging, GE Research, Niskayuna, NY, United States, 2GE Healthcare, Bangalore, India, 3GE Healthcare, Waukesha, WI, United States |
||
Keywords: Machine Learning/Artificial Intelligence, Visualization, Scan Planning, Spine, Reformatting We present a generalized DL-based intelligent slice placement framework for planes of cervical neural foramina (CF) and lumbar pars interarticularis (PI) for spine MRI. CF and PI scan improves assessment of foraminal stenosis and lumbar spondylolysis respectively, but requires highly skilled operator for accurate prescription. Our approach enables automatic patient-specific scan plane prescription from routine axial T2W images. In our test, it achieves mean error of <0.7 mm and <0.2 degrees and demonstrates similar/better contiguous anatomical visualization to manual scans on retrospective reformatting of 3D data. These results indicate that our approach is accurate and suitable for clinical usage. |
| 2953 | Computer 75
|
Deep learning-based saturation band placement in spine localizer images |
| Ashish Saxena1, Chitresh Bhushan2, Soumya Ghose2, Patil Uday1, Kameswari Padmanabhan3, Sanjay NT3, and Dattesh Shanbhag3 | ||
1GE Research, Bengaluru, India, 2GE Research, Niskayuna, NY, United States, 3GE Healthcare, Bengaluru, India |
||
Keywords: Machine Learning/Artificial Intelligence, Visualization, workflow Saturation band placement is important for obtaining good quality MRI images in presence of structures which can generate artifacts such as cardiac regions or large pulsating blood vessels. In anatomy such as spine, saturation-band placement can be time consuming since technologist has to ensure that it doesn’t overlap with vertebrae regions. In this study, we demonstrated an automated deep-learning method to accurately place the saturation band on 2D three-plane localizer images with no intervention from the technologist. |
| 2954 | Computer 76
|
Within Patient Contrast Adjustment Through a Self-Consistent Deep Learning Model when Imaging Near Metal: An Example in Angiography |
| Kevin M. Koch1,2 and Andrew S. Nencka1,2 | ||
1Radiology, Medical College of Wisconsin, Milwaukee, WI, United States, 2Center for Imaging Research, Medical College of Wisconsin, Milwaukee, WI, United States |
||
Keywords: Machine Learning/Artificial Intelligence, Vessels, Metal Artifact Multi-spectral imaging around metallic implants allows for a unique application of subject-specific deep learning applications because both high-resolution standard acquisitions and multi-spectral acquisitions are performed with significant regions of artifact-free overlap. A deep neural network can be trained in that region of artifact-free overlap within a single patient to yield a contrast transform on the multi-spectral images and achieve similar contrast to traditional acquisitions in the region of those acquisitions obscured by metal artifacts. In this proof of concept, a preliminary example of synthesizing time of flight-like contrast from a multi-spectral acquisition with vascular flow voids was shown. |
| 2955 | Computer 77
|
Real-time style transfer for quantitative susceptibility mapping using unsupervised learning |
| Zhuang Xiong1, Yang Gao2, and Hongfu Sun1 | ||
1the University of Queensland, Brisbane, Australia, 2Central South University, Changsha, China |
||
Keywords: Machine Learning/Artificial Intelligence, Machine Learning/Artificial Intelligence, unsuperivised learning Existing supervised deep learning methods for quantitative susceptibility mapping (QSM) could lead to degraded results when applied to phase images acquired with different scan parameters, such as image resolution and acquisition orientation. This work proposes a novel unsupervised learning method incorporating style transfer and deep image prior to enabling the reconstruction of susceptibility maps from local field maps acquired with any scan parameters. To speed up the inference, a pre-training strategy is also proposed, reducing reconstruction time from minutes to seconds. |
| 2956 | Computer 78
|
CloudBrain-LabelAI: An Online Intelligent Medical Imaging Annotation and Training Platform |
| Bangjun Chen1, Jian Wang1, Yirong Zhou1, Yu Hu1, Shuxian Niu1, Biao Qu2, Di Guo3, Jingjing Xu1, Jiyang Dong1, and Xiaobo Qu1 | ||
1Biomedical Intelligent Cloud R&D Center, Department of Electronic Science, National Institute for Data Science in Health and Medicine, Xiamen University, Xiamen, China, 2Department of Instrumental and Electrical Engineering, Xiamen University, Xiamen, China, 3School of Computer and Information Engineering, Xiamen University of Technology, Xiamen, China |
||
Keywords: Machine Learning/Artificial Intelligence, Segmentation Deep learning and cloud computing technologies have shown remarkable performance in medical imaging research. Using deep learning for clinical research requires a programming foundation, while the annotation of image data is a highly specialized and time-consuming process. In this work, we develop a high-performance online medical image annotation and training platform (CloudBrain-LabelAI). It provides medical image researchers with an efficient image annotation platform for the rapid construction of datasets for deep learning. Meanwhile, it provides codeless image segmentation training and prediction based on cloud computing, which greatly reduces the threshold of medical image segmentation efforts and simplifies the overall workflow. |
| 2957 | Computer 79
|
Self-contained comprehensive quantification of dynamic contrast-enhanced MRI using physics+kinetics-based network learning (PKNet) |
| Soudabeh Kargar1, Ouri Cohen1, Sungmin Woo2, Hebert Alberto Vargas2, and Ricardo Otazo1 | ||
1Medical Physics, MSKCC, New York, NY, United States, 2Radiology, MSKCC, New York, NY, United States |
||
Keywords: Machine Learning/Artificial Intelligence, DSC & DCE Perfusion, Deep Learning This work develops a deep learning technique that is trained according to physics and kinetics for self-contained comprehensive quantification of dynamic contrast-enhanced MRI. In addition to perfusion parameters, patient-specific parameters that affect the quantification are estimated, including bolus arrival time, T1, steady state magnetization, and AIF. The DCE-MRI quantification network was tested on a patient with cervical cancer and demonstrated high concordance between two scans separated by 24 hours. Physics plus kinetics informed network learning (PKNet) enabled the quantification of multiple parameters which has the potential to increase reproducibility of quantitative DCE-MRI, a long-desired goal. |
| 2958 | Computer 80
|
Predicting the Risk of Non-Saccular Aneurysms Based on the Dynamic Contrast-Enhanced Magnetic Resonance Imaging |
| Yan Li1, Ziming Xu1, Linggen Dong2, Yajie Wang1, Peng Liu2, Ming Lv2, and Huijun Chen1 | ||
1Center for Biomedical Imaging Research, Tsinghua University, Beijing, China, 2Department of Neurosurgery, Beijing Tiantan Hospital, Beijing Neurosurgical Institute, Capital Medical University, Beijing, China |
||
Keywords: Data Analysis, DSC & DCE Perfusion Assessment of intracranial aneurysm rupture risk is clinically critical. Recent studies have shown that the aneurysm wall permeability (Ktrans) and wall enhancement index are independent risk factors for predicting saccular aneurysm rupture. However, this conclusion has not been confirmed in non-saccular aneurysms. Our study suggested that Ktrans was significantly associated with the aneurysm size and PHASES score in non-saccular aneurysms. But Ktrans did not associate with the wall enhancement index in non-saccular aneurysms. Therefore, Ktrans can provide an independent quantitative indicator for assessing the risk of rupture of in non-saccular intracranialaneurysms. |
| 3057 | Computer 1
|
Cortical Blood Perfusion during Preclinical Migraine at 21.1 T using FAIR-PASL |
| Dayna Leigh Richter1,2 and Samuel Colles Grant1,2 | ||
1National High Magnetic Field Laboratory, Florida State University, Tallahassee, FL, United States, 2Chemical & Biomedical Engineering, FAMU-FSU College of Engineering, Tallahassee, FL, United States |
||
Keywords: Neurodegeneration, Arterial spin labelling, Migraine, FAIR-EPI Migraine is a neurological disorder with neurovascular implications. Blood flow changes based on brain's activity and needs, so excessive neural activation during migraine is likely to make significant alterations to cortical blood perfusion. Therefore, cortical blood perfusion was measured using FAIR-PASL to monitor the progression in preclinical migraine in female Sprague-Dawley rats. Parameters for inversion RF pulse shape and slice thickness were optimized to achieve more robust fits for monitoring cortical blood perfusion during preclinical migraine. |
| 3058 | Computer 2
|
A python-based post-processing toolkit for rodent perfusion MRI |
| Jinyuan Zhang1,2, Yishuang Yang1,2, Rong Xue1,2, Yan Zhuo1,2, and Zihao Zhang1,3 | ||
1State Key Laboratory of Brain and Cognitive Science, Institute of Biophysics, Chinese Academy of Sciences, Beijing, China, 2University of Chinese Academy of Sciences, Beijing, China, 3Institute of Artificial Intelligence, Hefei Comprehensive National Science Center, Hefei, China |
||
Keywords: Software Tools, Software Tools We developed a python-based post-processing toolkit for rodent perfusion MRI with an easy-to-use graphical user interface (GUI). Until now, this toolkit provides interfaces for the post-processing of dynamic susceptibility contrast (DSC) MRI and flow-sensitive alternating inversion recovery (FAIR) pulsed arterial spin labeling (pASL). For each modality, the toolkit has function modules including image viewer, display of time series data, ROI tools, and visualization of quantitative parameter maps. This toolkit is open source and welcomes added features. It will benefit preclinical studies using perfusion MRI. |
| 3059 | Computer 3
|
Relative Cerebral Blood Volume Differences Between Adult-Type Diffuse Glioma Subgroups According to WHO 2021: A Study of 146 Gliomas |
| Buse Buz-Yalug1, Ayca Ersen Danyeli2,3, M. Cengiz Yakicier3,4, M. Necmettin Pamir3,5, Koray Ozduman3,5, Alp Dincer3,6, and Esin Ozturk-Isik1,3 | ||
1Institute of Biomedical Engineering, Bogazici University, Istanbul, Turkey, 2Department of Medical Pathology, Acibadem University, Istanbul, Turkey, 3Brain Tumor Research Group, Acibadem University, Istanbul, Turkey, 4Department of Molecular Biology and Genetics, Acibadem University, Istanbul, Turkey, 5Department of Neurosurgery, Acibadem University, Istanbul, Turkey, 6Department of Radiology, Acibadem University, Istanbul, Turkey |
||
Keywords: Machine Learning/Artificial Intelligence, Machine Learning/Artificial Intelligence The main purpose of this study is to analyze the relative cerebral blood volume (rCBV) differences of adult diffuse glioma subgroups defined in the updated WHO 2021 brain tumor classification. The IDH wildtype group had statistically significantly higher rCBV values, and IDH mutational subgroups were classified with 82.5% accuracy (precision = 82.1%, recall = 82.9%), while the accuracies of glioblastoma and oligodendroglioma classification was 81.9%, glioblastoma and astrocytoma classification was 83.3%, and astrocytoma and oligodendroglioma classification was 77.5%. |
| 3060 | Computer 4
|
The impact of naproxen on Cerebral Blood Flow in preclinical Alzheimer’s disease |
| Safa Sanami1,2, Brittany N Intzandt3, Julia Huck1, PREVENT-AD Research Group4, and Claudine J Gauthier2,5,6 | ||
1Physics, Concordia University, Montreal, QC, Canada, 2Centre de Recherche de l’Institut de Cardiologie de Montr´eal, Montreal, QC, Canada, 3Sunnybrook Research Institute, Toronto, ON, Canada, 4Douglas Mental Health Institute, Montreal, QC, Canada, 5Concordia University, Montreal, QC, Canada, 6PERFORM Centre, Concordia University, Montreal, QC, Canada |
||
Keywords: Data Processing, Alzheimer's Disease Alzheimer’s Disease (AD) is the most common form of dementia and displays a long preclinical phase. The use of non-steroidal anti-inflammatory drugs (NSAIDs) in this phase has a protective effect and decreases the risk of developing AD cross-sectionally. However, the effects of NSAIDs therapy long-term on cerebral hemodynamics in this early phase is unclear. This is the first study to assess whether longitudinal use of NSAIDs (i.e., naproxen) has an impact on cerebral blood flow, and whether this is associated with a change in cognition and/or cerebrospinal fluid markers. |
| 3061 | Computer 5
|
Assessing cerebral perfusion: transfer function analysis of the BOLD response to hypoxia-induced changes in deoxyhemoglobin |
| James Duffin1,2, Ece Su Sayin1,3, Olivia Sobczyk2,3, Julien Poublanc4, Harrison T. Levine1,3, David J. Mikulis4, and Joseph A. Fisher1,5 | ||
1Physiology, University of Toronto, Toronto, ON, Canada, 2Department of Anaesthesiology and Pain Management, University of Toronto, Toronto, ON, Canada, 3Joint Department of Medical Imaging and the Functional Neuroimaging Lab,, University Health Network, Toronto, ON, Canada, 4Joint Department of Medical Imaging and the Functional Neuroimaging Lab, University Health Network, Toronto, ON, Canada, 5Department of Anaesthesiology and Pain Management, University Health Network, Toronto, ON, Canada |
||
Keywords: Data Analysis, Brain We used hypoxia-induced changes in deoxyhemoglobin concentration as a susceptibility contrast agent. Transfer function analysis of the resulting changes in the BOLD signal for each voxel provided cerebral perfusion measures that indicate the distribution of the strength of the signal response to the contrast agent (gain) and the lag of the response (phase or time). We relate the gain and lag to conventional resting hemodynamic measures cerebral blood volume and mean transit time, respectively. |
| 3062 | Computer 6
|
Deep Learning-based Automatic Perfusion Phase Identification for Dynamic T1-weighted Liver MRI |
| Robert Grimm1, Malte Müller2, Cornelius Jacob1, Sabine Mollus1, Christian Tietjen3, Moon Hyung Choi4, Kazuki Oyama5, Thomas Weikert6, Andrew D Hardie7, Jeong Hee Yoon8, Heinrich von Busch3, Gregor Thoermer1, and Volker Daum2 | ||
1MR Application Predevelopment, Siemens Healthcare GmbH, Erlangen, Germany, 2Chimaera GmbH, Erlangen, Germany, 3Digital and Automation, Siemens Healthcare GmbH, Erlangen, Germany, 4Eunpyeong St. Mary’s Hospital, Catholic University of Korea, Seoul, Korea, Republic of, 5Department of Radiology, Shinshu University Hospital, Nagano, Japan, 6Department of Radiology, Universitätsspital Basel, Basel, Switzerland, 7Radiology and Radiological Science, Medical University of South Carolina, Charleston, SC, United States, 8Seoul National University Hospital and College of Medicine, Seoul, Korea, Republic of |
||
Keywords: Data Analysis, Machine Learning/Artificial Intelligence A deep learning-based approach for automatic identification of the perfusion phases in dynamic T1-weighted liver MRI is presented. First, an encoder model combined with two dense layers was trained to classify each image into pre-contrast, arterial, portal-venous, late, or hepatobiliary phase. In a second pass, classification errors are detected and adjusted, based on the expected occurrence order and relative timing to the arterial phase. The AI model reached sensitivities of 67% to 99%. Most common mis-classifications were confusions of the portal-venous or late phase with the adjacent phases. By the rule-based adjustments, the classification performance was raised to >95% accuracy. |
| 3063 | Computer 7
|
The ERIC phantom - Extra-dimensional respiration and inflow of contrast: an open-source DCE digital simulation tool |
| Eric Schrauben1 | ||
1Radiology and Nuclear Medicine, Academic Medical Centre, Amsterdam, Netherlands |
||
Keywords: Software Tools, DSC & DCE Perfusion An open-source MATLAB-based software tool was developed for digital simulation (using the MRXCAT framework) of DCE MRI acquisition and reconstruction strategies, termed the ERIC phantom: extra-dimensional respiration and inflow of contrast. The ERIC phantom is a user-friendly graphical user interface, allowing for variable respiratory motion, acquisition parameters and trajectories, motion correction strategies, and compressed sensing reconstructions. This realistic abdominal DCE phantom can be used to investigate varying strategies for producing high quality DCE reconstructions in the presence of respiratory motion and contrast inflow. |
| 3064 | Computer 8
|
Normal left ventricular flow dynamics in a paediatric population assessed by 4D Flow MRI |
| Fraser Maurice Callaghan1,2, Barbara Burkhardt2,3, Julia Geiger2,4, Emanuela Valsangiacomo Buechel2,3, and Christian Kellenberger2,4 | ||
1University Children's Hospital Zurich, Zurich, Switzerland, 2Children’s Research Center, University Children's Hospital Zurich, Zurich, Switzerland, 3Division of Pediatric Cardiology, University Children's Hospital Zurich, Zurich, Switzerland, 4Department of Diagnostic Imaging, University Children's Hospital Zurich, Zurich, Switzerland |
||
Keywords: Data Analysis, Cardiovascular, 4D flow 4D flow MRI provides a rich dataset that can make analysis challenging. A new technique of left ventricle (LV) registration is presented and compared with established techniques for analysis of flow dynamics in a normal paediatric population. Averaging of group flow dynamics in a common space permits additional analysis, complementary to established techniques. Our technique demonstrated flow dynamics on a common LV space and identified subtle differences in LV velocities between normal male and female subjects of a paediatric population. This represents a new tool in 4D flow data analysis for potential biomarker development. |
| 3065 | Computer 9
|
Intracranial Wall Permeability for Prediction of Aneurysm Growth Using Dynamic Contrast-Enhanced MRI |
| Ziming Xu1, Linggen Dong2, Longhui Zhang2, Yajie Wang1, Jiaqi Dou1, Peng Liu2, Ming Lv2, and Huijun Chen1 | ||
1Center for Biomedical Imaging Research, School of Medicine, Tsinghua University, Beijing, China, 2Department of Neurosurgery, Beijing Tiantan Hospital, Beijing Neurosurgical Institute, Capital Medical University, Beijing, China |
||
Keywords: Data Analysis, DSC & DCE Perfusion Size of intracranial aneurysm (IA) is considered the most important clinical factor to determine the risk of IA rupture. Accurate prediction of aneurysm growth is crucial for preventive management but remains challenging. Herein, we studied the potential predictive roles of demography and imaging characteristics in a radiological follow-up study. And wall permeability calculated by DCE-MRI was firstly observed to accurately identify the IAs with a relatively high risk of aneurysm growth (AUC = 0.875). Our study demonstrated that higher wall permeability was associated with aneurysm growth and the regions with higher wall permeability co-localized with the direction of aneurysm growth. |
| 3066 | Computer 10
|
Association between 3D Geometry of Vertebrobasilar Arteries and Basilar Artery Atherosclerosis: An MRI Study |
| Jiachen Liu1, Rui Shen1, Dandan Yang2, Miaoxin Yu3, and Xihai Zhao1 | ||
1Center for Biomedical Imaging Research, Department of Biomedical Engineering, Tsinghua University, Beijing, China, 2Department of Radiology, Beijing Geriatric Hospital, Beijing, China, 3Department of Neurology, Beijing Tiantan Hospital, Capital Medical University, Beijing, China |
||
Keywords: Data Analysis, Atherosclerosis Geometry of VBA has been proved to be associated with the presence of BA plaques, yet relative studies are lack of 3D information due to methodological difficulties. We proposed a semi-automated algorithm for quantitative measurement, which can evaluate local diameter, local curvature and angles between vessels directly in 3D space. We used the proposed algorithm to measure the 3D geometric metrics and found significant association between VBA geometry and the presence of BA plaques. Normalized diameter variation of BA and BA & PCA-R angle projected to lateral view were also associated with the presence of BA plaques independently. |
| 3067 | Computer 11
|
Automated Intracranial Artery Labeling in Patients with Cerebrovascular Steno-occlusive Diseases |
| Lixin Liu1, Yi Lv2, Peirong Jiang3, He Wang1,4,5, and Zhensen Chen1,5 | ||
1Institute of Science and Technology for Brain-Inspired Intelligence,Fudan University, Shanghai, China, 2School of Compute Science and Technology, Beijing Institute of Technology, Beijing, China, 3Department of Radiology, Fujian Medical University Union Hospital, Fuzhou, China, 4Human Phenome Institute, Fudan University, Shanghai, China, 5Key Laboratory of Computational Neuroscience and Brain-Inspired Intelligence (Fudan University), Ministry of Education, Shanghai, China |
||
Keywords: Data Processing, Blood vessels Labeling of intracranial arteries is important for computer-aided diagnosis of cerebrovascular diseases and quantitative analysis of intracranial vasculature. Performance of the previously proposed automated intracranial artery labeling method based on Graph Neural Network (GNN) is limited in datasets with overt cerebrovascular steno-occlusive diseases. In this study, we improved the generalizability of the GNN-based method by using dedicated data augmentation and spatial normalization strategy. The results show that our method is more robust than the previous method in ischemic stroke patients with overt intracranial stenosis or occlusion. |
| 3068 | Computer 12
|
Automated Localization of the Extracranial Carotid Artery in Black Blood Contrast MR Images Using a Deep Learning Approach |
| SeyyedKazem HashemizadehKolowri1, Nadin Zanaty2,3, Gador Canton2, Niranjan Balu2, Thomas S. Hatsukami2, and Chun Yuan1,2 | ||
1Radiology and Imaging Sciences, University of Utah, SALT LAKE CITY, UT, United States, 2Department of Radiology, University of Washington, Seattle, WA, United States, 3Radiology, Zagazig University, Zagazig, Egypt |
||
Keywords: Machine Learning/Artificial Intelligence, Cardiovascular, Vessel Wall Imaging In this work, a deep learning approach for automated localization of carotid arteries in black blood contrast MR data is proposed. This is the first step in automated analysis of vessel wall imaging data. Carotid arteries supply oxygenated blood to the brain and are susceptible to atherosclerosis, so their vessel wall imaging is of significant importance in clinical evaluations. However, currently only qualitative assessment of VW imaging data relying on visual inspection is implemented in clinics that are not scaleable. Therefore, developing automated image processing tools to quantitatively analyze vessel wall imaging data can have a major clinical impact. |
| 3069 | Computer 13
|
Automated modeling and morphologic analysis of deep medullary veins at 7T MRI in patients with cognitive impairment |
| Zhixin Li1,2,3, Jingyuan Zhang1,2,3, Li Liang 4, Jing An5, Hairong Qian4, Rong Xue1,2,3, Yan Zhuo1,2,3, and Zihao Zhang1,2,6 | ||
1State Key Laboratory of Brain and Cognitive Science, Institute of Biophysics, Chinese Academy of Sciences, Beijing, China, Beijing, China, 2University of Chinese Academy of Sciences, Beijing, China, Beijing, China, 3The Innovation Center of Excellence on Brain Science, Chinese Academy of Sciences, Beijing, China, Beijing, China, 4Department of Neurology, the Sixth Medical Center, Chinese PLA General Hospital, Beijing, China, 5Siemens Shenzhen Magnetic Resonance Ltd., Shenzhen, China, Shenzhen, China, 6Institute of Artificial Intelligence, Hefei Comprehensive National Science Center, Hefei, China, Hefei, China |
||
Keywords: Machine Learning/Artificial Intelligence, Modelling Deep medullary veins (DMVs) support cerebral venous drainage. They may display abnormal changes in patients with cognitive impairment. They can be visualized by multi-echo gradient echo imaging at 7T. This study proposed a segmentation and tracking method based on deep learning and shortest-path optimization. It automatically quantified the morphologic parameters of DMVs from the vascular model. These characteristics of DMVs correlated with the patients’ cognitive scores, and might reflect the pathology of vascular lesions in cognitive impairment. |
| 3070 | Computer 14
|
Carotid arterial stiffening is associated with reduced downstream territorial perfusion of internal carotid artery in elderly adults |
| Yining He1, Jianing Tang1,2, Tianrui Zhao1,2, and Lirong Yan1,3 | ||
1Radiology, Northwestern University, Chicago, IL, United States, 2Biomedical Engineering, Northwestern University, Evanston, IL, United States, 3Neurology, University of Southern California, Los Angeles, CA, United States |
||
Keywords: Blood vessels, Aging Arterial stiffness is an important risk marker for poor brain aging, vascular disease, and dementia. Greater arterial stiffness leads to the transmission of excessive pulsations from the greater vessels into the downstream capillary and tissue causing microvascular dysfunction. However, previous studies have mainly focused on central or peripheral pulse wave velocity assessment. The present study has demonstrated a significant association of the PWV of the feeding arteries to the brain and its downstream territorial perfusion assessed by using two new MRI techniques, which could provide valuable insight into the neurovascular pathology of aging and brain dysfunction. |
| 3071 | Computer 15
|
Retrospective quantification pharmacokinetics of clinical breast DCE-MRI using deep learning |
| Chaowei Wu1,2, Lixia Wang1, Nan Wang1,3, Stephen Pandol4, Anthony G Christodoulou1,2, Yibin Xie1, and Debiao Li1,2 | ||
1Biomedical Imaging Research Institute, Cedars-Sinai Medical Center, Los Angeles, CA, United States, 2Department of Bioengineering, University of California, Los Angeles, Los Angeles, CA, United States, 3Radiology Department, Stanford University, Stanford, CA, United States, 4Division of Digestive and Liver Diseases, Cedars-Sinai Medical Center, Los Angeles, CA, United States |
||
Keywords: Machine Learning/Artificial Intelligence, DSC & DCE Perfusion Standard-of-care DCE-MRI suffers from a limited number of contrast phases and low temporal resolution, preventing the quantification of pharmacokinetic parameters. Quantitative DCE-MRI techniques have not yet been widely applied in the clinic due to the limited availability of specialized sequences and image reconstruction. To tackle this problem, we proposed to improve the temporal resolution of multi-phasic DCE-MRI by deep learning post-processing and demonstrated promising results in tumor delineation in the Duke-Breast-Cancer-MRI dataset. |
| 3072 | Computer 16
|
The application of Quantitative Perfusion Analysis of GRASP for assessing pathologic prognostic factors in rectal cancer |
| Mi Zhou1, Meining Chen2, and Yuting Wang1 | ||
1Department of Radiology, Sichuan Provincial People's Hospital, chengdu, China, 2Department of MR Scientific Marketing, Siemens Healthcare, Shanghai, China |
||
Keywords: Data Analysis, Perfusion, rectum Colorectal cancer’s pathological prognostic factors directly affect the patient's prognosis. Golden-angle RAdial Sparse Parallel (GRASP) imaging was invented to calculate the accurate perfusion parameters including influx forward volume transfer constant (Ktrans), rate constant (Kep), and plasma volume fraction (Ve). We found GRASP parameters showed significant differences in many prognostic factors, including histology type, extramural venous invasion (EMVI), lymphovascular invasion (LVI), tumor deposit (TD) and lymph node metastasis (LNM), and were independent factors for histology type, LNM, LVI and EMVI. GRASP parameters were strongly associated with preoperative prognostic factors in rectal cancer, providing useful information for treatment and follow-up protocol. |
| 3073 | Computer 17
|
Evaluation of multi-echo-based Hybrid-EPI (HEPI) technique for measuring brain oxygenation |
| Krishnapriya Venugopal1, Esther A.H Warnert1, Daniëlle van Dorth2, Marion Smits1, Matthias J.P van Osch2, Dirk H.J Poot1, and Juan Antonio Hernandez-Tamames1,3 | ||
1Radiology and Nuclear Medicine, Erasmus MC, Rotterdam, Netherlands, 2Radiology, Leiden University Medical Center, Leiden, Netherlands, 3Medical Imaging, TU Delft, Delft, Netherlands |
||
Keywords: Data Acquisition, Brain, Oxygenation The reversible component of transverse relaxation time, R2’, enables the measurement of blood oxygenation, an important biomarker in several diseases. We propose a multi-echo HEPI technique to estimate R2’. HEPI combines GRE and SE and hence provides R2* and R2, when acquired at different echo-times. The accuracy of ME-HEPI in measuring R2’ is evaluated in a phantom in this work. The sensitivity of HEPI to different oxygenation levels, attained using respiratory challenge MRI, is studied in a healthy subject. This is also investigated using a simulation tool that simulates microvasculature and MR signal for varying oxygenation in the vessels.
|
| 3074 | Computer 18
|
Reliable blood-brain barrier water exchange estimates in the rat brain using a crusher-compensated exchange rate (CCXR) model |
| Yolanda Ohene1,2, Elizabeth Powell3, Samo Lasič4,5, Geoff J. M. Parker3,6, Laura M. Parkes1,2, and Ben R. Dickie2,7 | ||
1Division of Psychology, Communication and Human Neuroscience, University of Manchester, Manchester, United Kingdom, 2Geoffrey Jefferson Brain Research Centre, University of Manchester, Manchester, United Kingdom, 3UCL, London, United Kingdom, 4Danish Research Centre for Magnetic Resonance, Copenhagen, Denmark, 5Random Walk Imaging, Åkarp, Sweden, 6Bioxydyn Limited, Manchester, United Kingdom, 7Division of Informatics, University of Manchester, Manchester, United Kingdom |
||
Keywords: Data Processing, Permeability Filter exchange imaging (FEXI) is a promising technique for measuring water exchange across the blood-brain barrier (BBB). However, the application of FEXI for the rodent brain requires thinner slices and therefore higher crusher gradients which lead to a progressive underestimation of the apparent exchange rate (AXR). Here, we implement a crusher-compensated exchange rate (CCXR) model which reduces the bias induced by the crusher gradients and allows more accurate estimates of BBB water exchange in the rat brain. |
| 3075 | Computer 19
|
Early Prediction of Response to Neoadjuvant Systemic Therapy of Triple Negative Breast Cancer using Radiomics on DCE-MRI |
| Bikash Panthi1, Sanaz Pashapoor1, Beatriz E. Adrada1, Rosalind P. Candelaria1, Mary S. Guirguis1, Miral M. Patel1, Rania M. Mohamed1, Medine Boge1, Zijian Zhou1, Jong Bum Son1, Ken-Pin Hwang1, Huong T.C. Le-Petross1, Jessica W.T. Leung1, Marion E. Scoggins1, Gary J. Whitman1, Zhan Xu1, Deanna L. Lane1, Tanya Moseley1, Frances Perez1, Jason White1, Elizabeth Ravenberg1, Alyson Clayborn1, Huiqin Chen1, Jia Sun1, Peng Wei1, Alastair Thompson2, Stacy Moulder1, Anil Korkut1, Lei Huo1, Kelly K. Hunt1, Jennifer K. Litton1, Vicente Valero1, Debu Tripathy1, Wei Yang1, Clinton Yam1, Gaiane M Rauch1, and Jingfei Ma1 | ||
1MD Anderson Cancer Center, Houston, TX, United States, 2Baylor College of Medicine, Houston, TX, United States |
||
Keywords: Radiomics, fMRI, Treatment response We developed models based on radiomic features from dynamic contrast enhanced (DCE) MR images and demonstrated that these models have potential to serve as non-invasive biomarkers for early prediction of pathologic complete response (pCR) in triple negative breast cancer (TNBC) patients undergoing neoadjuvant systemic therapy (NAST). |
| 3076 | Computer 21
|
CNN-based Estimation of Spatio-Temporal Regularization Parameter-Maps for TV-Reconstruction in Dynamic Cardiac MRI |
| Andreas Kofler1, Clemens Sirotenko2, Felix Frederik Zimmermann1, David Schote1, Christoph Kolbitsch1,3, Fatima Antarou Ba4, Fabian Altekrüger4,5, Evangelos Papoutsellis6,7, and Kostas Papafitsoros8 | ||
1Physikalisch-Technische Bundesanstalt (PTB), Braunschweig and Berlin, Germany, 2Weierstraß-Institut für Angewandte Analysis und Stochastik, Berlin, Germany, 3King’s College London, London, United Kingdom, 4Technische Universität Berlin, Berlin, Germany, 5Humboldt Universität zu Berlin, Berlin, Germany, 6Science and Technology Facilities Council (STFC), Oxford, United Kingdom, 7Finden Ltd, Oxford, United Kingdom, 8Queen Mary University of London, London, United Kingdom |
||
Keywords: Image Reconstruction, Machine Learning/Artificial Intelligence, Interpretable Machine Learning, Algorithmic Unrolling, Iterative Neural Networks We propose a method for estimating spatio-temporal regularization parameter-maps to be used for dynamic cardiac MR image reconstruction using total variation (TV)-minimization. Based on recent developments in algorithmic unrolling using Neural Networks (NNs), our approach uses two sub-networks. The first one predicts a spatio-temporal regularization parameter-map from an input image. Then, a second sub-network approximately solves a TV-reconstruction problem which is formulated with the estimated regularization parameter-map. We show that the proposed method can be used to further improve the TV-reconstructions compared to using only one single scalar regularization parameter or two regularization parameters for space and time. |
| 3077 | Computer 22
|
Combining HOSVD-based joint reconstruction with shifted acquisition sampling pattern enables liver DWI in a single breath-hold |
| Guohao Jiang1, Linzheng Hong1, Zhaopeng Li2, and Jun Xie2 | ||
1ShanghaiTech University, Shanghai, China, 2United Imaging Healthcare, Shanghai, China |
||
Keywords: Image Reconstruction, Diffusion/other diffusion imaging techniques Fast liver DWI scan has significant clinical relevance, however, motion artifacts, long scan time and low SNR remain as major technical challenges for liver DWI. In this study, we proposed a joint recon method based on global and local high order tensor SVD (HOSVD), combining with shifted acquisition sampling pattern across diffusion directions. The proposed method improves DWI image quality and enables fast liver DWI scan within a single breath-hold. |
| 3078 | Computer 23
|
Accelerated 3D dynamic upper-airway MRI in naturally sleeping obstructive sleep apnea patients. |
| Wahidul Alam1, Junjie Liu2, and Sajan Goud Lingala1,3 | ||
1Roy J. Carver Department of Biomedical Engineering, University of Iowa, iowa city, IA, United States, 2Department of Neurology, University of Iowa, iowa city, IA, United States, 3Department of Radiology, University of Iowa, Iowa city, IA, United States |
||
Keywords: Image Reconstruction, Head & Neck/ENT Obstructive sleep apnea (OSA) is characterized by breathing-related obstructions of the upper airway during sleep. In this work, we develop a motion resolved extra-dimension sparsity-based approach to resolve the kinematics of upper-airway in OSA during natural sleep. This approach is demonstrated on two adult OSA patients: one undergoing a Bi-directional positive airway pressure therapy during MRI scanning, and one without therapy |
| 3079 | Computer 24
|
Water/Fat separation with spatio-temporal EPI-based acquisition and reconstruction in body imaging |
| Xuetong Zhou1,2, Philip K. Lee2, and Brian A. Hargreaves1,2,3 | ||
1Department of Bioengineering, Stanford University, Stanford, CA, United States, 2Department of Radiology, Stanford University, Stanford, CA, United States, 3Department of Electrical Engineering, Stanford University, Stanford, CA, United States |
||
Keywords: Image Reconstruction, Image Reconstruction Water/fat separation is a reliable fat suppression technique. However, it is challenging in EPI-based imaging due to the large displacement along phase-encoding direction. We demonstrate the artifacts resulting from the poor-conditioning reconstruction of water/fat separation with EPI-based acquisition in body imaging. As a solution, unlike conventional multi-echo acquisition methods, our approach encodes the spectral components (water/fat) by acquiring and reconstructing the spatial and echo-time dimensions jointly. EPTI sampling trajectories and water/fat masking were used to improve the conditioning of the reconstruction. Phantom and in vivo (brain and breast) experiments were performed to show both results and limitations of this approach. |
| 3080 | Computer 25
|
High resolution 4D-cine MRI of pulsatile aneurysmal motion with radial acquisition at 7T |
| Thai Akasaka1, Koji Fujimoto2, Martijn Cloos3, Tomohisa Okada1, Shinichi Urayama1, and Isa Tadashi1 | ||
1Human Brain Research Center, Kyoto University, Kyoto, Japan, 2Real World Data Research and Development, Kyoto University, Kyoto, Japan, 3University of Queensland, Brisbane, Australia |
||
Keywords: Image Reconstruction, Image Reconstruction The pulsation of intracranial aneurysms (IAs) is a novel risk factor of rupture. We aim to visualize the pulsation of IAs with 7T MRI using GRASP, a relatively novel MRI technique that provides high spatial and temporal resolution. Using thsis method, we were able to depict the pulsation of an aneurysm phantom as a 4D-cine movie. This technique may be useful in predicting the probability of IA rupture more accurately. |
| 3081 | Computer 26
|
Unsupervised denoising of prostate DWI |
| Laura Pfaff1,2, Fabian Wagner1, Julian Hossbach1,2, Elisabeth Preuhs1, Fasil Gadjimuradov1,2, Thomas Benkert2, Dominik Nickel2, Tobias Wuerfl2, and Andreas Maier1 | ||
1Pattern Recognition Lab, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany, 2Magnetic Resonance, Siemens Healthcare GmbH, Erlangen, Germany |
||
Keywords: Image Reconstruction, Diffusion/other diffusion imaging techniques The diagnostic value of diffusion-weighted MR images is often degraded by their inherently low signal-to-noise ratio (SNR), especially for high b-values. In this context, the application of learning-based denoising methods is difficult since most methods require noise-free target images for training. We show how to denoise and evaluate diffusion-weighted MR images in a self-supervised manner by exploiting an adapted version of Stein’s unbiased risk estimator and specific properties of the data. Both quantitative and qualitative evaluations indicate increased performance over state-of-the-art unsupervised denoising methods. |
| 3082 | Computer 27
|
Non-contrast high-resolution 4D-peripheral MR angiography using retrospective echo planar imaging with Compressed SENSE |
| Yasuhiro Goto1, Michinobu Nagao2, Masami Yoneyama3, Johannes M Peeters4, Yasutomo Katsumata3, Isao Shiina1, Kazuo Kodaira1, Yutaka Hamatani1, Takumi Ogawa1, Mana Kato1, and Shuji Sakai2 | ||
1Department of Radiological Services, Tokyo Women's Medical University, Tokyo, Japan, 2Department of Diagnostic imaging & Nuclear Medicine, Tokyo Women's Medical University, Tokyo, Japan, 3Philips Japan, Tokyo, Japan, 4Philips Healthcare, Best, Netherlands |
||
Keywords: Image Reconstruction, Blood vessels In this study demonstrated that the higher reduction factor (R=10.0) with a one-minute scan still provided sufficient image quality with significantly faster scan time compared with conventional REPI-SENSE. REPIX would be useful for further assessment of PAD pathology even with multiple VENC acquisitions still in a clinically feasible scan time.REPIX 4D-MRA well depicted peripheral arteries clearly in a significantly short time thanks to Compressed SENSE, compared with conventional SENSE. |
| 3083 | Computer 28
|
Accelerated whole-heart MRI for congenital heart disease patients using a motion-corrected deep learning reconstruction network |
| Andrew Phair1, Anastasia Fotaki1, Lina Felsner1, Haikun Qi2, René M. Botnar1, and Claudia Prieto1 | ||
1School of Biomedical Engineering and Imaging Sciences, King's College London, London, United Kingdom, 2School of Biomedical Engineering, ShanghaiTech University, Shanghai, China |
||
Keywords: Image Reconstruction, Cardiovascular A deep learning reconstruction framework, trained in an end-to-end fashion and incorporating both a non-rigid respiratory motion estimation network and a motion-informed model-based reconstruction network, has been previously demonstrated to enable good quality images from seven-fold undersampled acquisitions for coronary magnetic resonance angiography applications. Herein, we apply the framework to whole-heart MRI scans of patients with congenital heart disease, enabling fast reconstruction of 7×-accelerated acquisitions and achieving image quality comparable to that of state-of-the-art patch-based low-rank iterative techniques. |
| 3084 | Computer 29
|
Dual Domain Deep Learning Framework for Cardiac MR Image Reconstruction |
| Faisal Najeeb1, Madiha Arshad 1, Muhammad Shafique1, and Hammad Omer1 | ||
1Electrical and Computer Engineering, COMSATS University Islamabad, Islamabad, Pakistan |
||
Keywords: Image Reconstruction, Artifacts, Deep Learning, Compressed Sensing , Parallel MRI Most deep learning methods apply U-Net either in image or k-space domain. Nevertheless, these methods have limitations: (1) Directly applying U-Net in k-space domain is not optimal for extracting features; (2) conventional image-domain oriented U-Net does not fully utilize the information of encoder part of the network for extracting features in the decoder part. In this paper, a dual-domain deep learning-based approach is presented, incorporating multi-coil data consistency layers for the reconstruction of cardiac MR images from 1-D Variable Density (VD) under-sampled data. Experiments show superior reconstruction results of the proposed method than conventional Compressed Sensing (CS) method. |
| 3085 | Computer 30
|
Towards motion-resolved interleaved radial 23Na/1H magnetic resonance imaging of the human heart at 7 Tesla |
| Jörn Huber1, Laurent Ruck2,3, Matthias Günther1,4,5, Armin Nagel2,6, and Simon Konstandin1,4 | ||
1Fraunhofer Institute for Digital Medicine MEVIS, Bremen, Germany, 2Institute of Radiology, University Hospital Erlangen, Erlangen, Germany, 3Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU), Erlangen, Germany, Erlangen, Germany, 4mediri GmbH, Heidelberg, Germany, 5Faculty 1 (Physics/Electrical Engineering), University of Bremen, Bremen, Germany, 6Division of Medical Physics in Radiology, German Cancer Research Center (DKFZ), Heidelberg, Germany, Erlangen, Germany |
||
Keywords: Image Reconstruction, Myocardium Motion correction in interleaved 23Na/1H MRI of the human heart is important to improve the diagnostic reliability of reconstructed images and derived quantitative parameters. Therefore, this work demonstrates and compares the application of different reconstruction techniques to undersampled and motion-gated 23Na/1H MRI data at 7 T. |
| 3086 | Computer 31
|
FIRE and MATLAB for Free-Breathing Segmented LGE – A Novel Sequence and Reconstruction Approach for Selecting and Correcting Data by Respiration |
| Wolfgang G Rehwald1,2, Kelvin Chow3, Rafael Rojas4, Nestor Mena4, Diana Alexandrov4, George Gamoneda4, David Wendell4, Ryan Seward4, Jeana Dement4, Vera Kimbrell4, Han Kim4, Indraneel Borgohain5, Igor Klem4, and Raymond Kim4 | ||
1Siemens Healthineers, Durham, NC, United States, 2Duke Cardiovascular MR Center, Duke University, Durham, NC, United States, 3Siemens Healthineers, Chicago, IL, United States, 4Duke Cardiovascular MR Center, Duke University Hospital, Durham, NC, United States, 5Siemens Healthineers, Princeton, NJ, United States |
||
Keywords: Image Reconstruction, Motion Correction, Reordering We introduce a free breathing segmented LGE technique combining a new acquisition, reordering and reconstruction scheme. The on-scanner reconstruction uses MATLAB and FIRE. The method produces image quality (IQ) and SNR otherwise only obtainable with breath held segmented interleaved LGE. In 27 patients, we show that IQ is superior to free breathing segmented interleaved LGE with multiple averages. SNR is higher compared to averaged motion corrected single shots when matching spatial and temporal resolution, the number of used measurements per image, and the readout type. Being a segmented technique, temporal and spatial resolution limitations of single shots do not apply.
|
| 3087 | Computer 32
|
Super-resolution reconstruction of time-resolved four-dimensional computed tomography (TR-4DCT) with multiple breaths based on TR-4DMRI |
| Yilin Liu1, Asala Ahamd1, Xingyu Nie2, and Guang (George) Li1 | ||
1Medical Physics, Memorial Sloan Kettering Cancer Center, New York, NY, United States, 2Radiology, University of Kentucky, Lexington, KY, United States |
||
Keywords: Image Reconstruction, Radiotherapy, time-resolved 4DMRI, time-resolved 4DCT, multi-breathing cycles This study has demonstrated that the feasibility to reconstruct multiple-breath TR-4DCT via the super-resolution reconstruction framework through either CT4D→(MRBH→MRFB) or (CT4D←MRBH)→ MRFB deformable image registration. Using TR-4DCT, potential dosimetry consequences in radiotherapy of lung, liver, and pancreatic patients due to patient breathing irregularities can be readily assessed. |
| 3088 | Computer 33
|
Convergent and Interpretable Dynamic Cardiac MR Image Reconstruction with Neural Networks-based Convolutional Dictionary Learning |
| Andreas Kofler1, Christian Wald2, Tobias Schaeffter1,3,4, Markus Haltmeier5, and Christoph Kolbitsch1,3 | ||
1Physikalisch-Technische Bundesanstalt (PTB), Braunschweig and Berlin, Germany, 2Charité - Universitätsmedizin Berlin, Berlin, Germany, 3King’s College London, London, United Kingdom, 4Technical University of Berlin, Berlin, Germany, 5University of Innsbruck, Innsbruck, Austria |
||
Keywords: Image Reconstruction, Machine Learning/Artificial Intelligence, Signal Processing, Sparsity-based Methods, Convolutional Dictionary Learning In this work we consider three different variants physics-informed Neural Networks (PINNs) which use the convolutional dictionary learning framework for image reconsruction in dynamic cardiac MRI. Although all three NNs share the same mechanism for regularization, the iterative schemes differ because they are derived from different problem formulations. We compare the methdos in terms of reconstruction performance as well as stability with respect to the number of iterations. All three methods yield similarly accurate reconstructions. However, by construction, only one of the three methods defines a convergent reconstruction algorithm and is therefore stable w.r.t. to the number of iterations. |
| 3089 | Computer 34
|
Two-Stage Kalman Filtering as a Framework for Accelerated Cardiac MRI |
| Aaron Curtis1,2 and Hai-Ling Margaret Cheng1,2,3 | ||
1Electrical and Computer Engineering, University of Toronto, Toronto, ON, Canada, 2Translational Biology & Engineering Program, Ted Rogers Centre for Heart Research, Toronto, ON, Canada, 3Institute of Biomedical Engineering, University of Toronto, Toronto, ON, Canada |
||
Keywords: Image Reconstruction, Heart Robust, real-time dynamic cardiac MRI (CMR) would provide information on the temporal signatures of disease that we currently cannot assess. We present a novel Kalman filtering framework that uses a priori statistics derived from a single cardiac cycle to adaptively predict temporal cardiac dynamics. Kalman filtering is ideal, as it ameliorates noise introduced from our maximum acceleration factor of 60, guarantees reconstruction fidelity, and enables flexible undersampling. Furthermore, reconstruction may be performed at an even higher temporal resolution than the training data. As such, our algorithm can be a foundation for true real-time dynamic CMR. |
| 3090 | Computer 35
|
4D Cardiac MR Image Reconstruction by Deep Learning with Wavelet Transform |
| Junhao Zhang1,2, Yujiao Zhao1,2, Jiahao Hu1,2,3, Ye Ding1,2, Christopher Men1,2, Vick Lau1,2, Alex T.L.Leong1,2, and Ed X. Wu1,2 | ||
1Laboratory of Biomedical Imaging and Signal Processing, the University of Hong Kong, HongKong, China, 2Department of Electrical and Electronic Engineering, the University of Hong Kong, HongKong, China, 3Department of Electrical and Electronic Engineering, Southern University of Science and Technology, Shenzhen, China |
||
Keywords: Data Analysis, Cardiovascular, cardiac reconstruction,deep learning We present a CNN-based deep learning model to reconstruct the cardiac cine images from undersampled single-channel 4D MR data. The wavelet transform and spatial-temporal attention mechanisms are introduced in the model. The proposed model could reconstruct the cardiac images and recover the cardiac wall motion more robustly than the low-rank plus sparsity (i.e., L+S) reconstruction method. This approach presents one promising solution for accelerated cardiac dynamic imaging with one single channel through deep learning from the sparsity in spatial and temporal domains. |
| 3091 | Computer 36
|
Tracking the moving stomach using MRI and neural ordinary differential equations |
| Xiaokai Wang1, Jiayue Cao2, Kuan Han2, Minkyu Choi2, Yushi She2, Ulrich Scheven2, and Zhongming Liu2 | ||
1Biomedical Engineering, University of Michgan, Ann Arbor, MI, United States, 2University of Michigan, Ann Arbor, MI, United States |
||
Keywords: Data Analysis, Data Analysis, Neural Network We describe a method, namely neural ordinary differential equations, to track the movement of the stomach based on dynamic and contrast-enhanced gastrointestinal MRI. This model uses a neural network to learn the continuous biomechanical process that drives the shape change of the stomach wall over the course of digestion. This method allows us to represent gastric motor events on a generic surface template of the stomach and to further reveal the pattern of gastric motility with higher specificity and resolution than are previously attainable in vivo. This method should be also applicable to other organs, such as the heart. |
| 3092 | Computer 37
|
Improvement of the time-resolved 4DMRI image quality through super-resolution reconstruction using compressed sensing |
| Can Wu1, Guang (George) Li1, and Yilin Liu1 | ||
1Medical Physics, Memorial Sloan Kettering Cancer Center, New York, NY, United States |
||
Keywords: Image Reconstruction, Radiotherapy, TR-4DMRI, multi-breathing cycles, breathing irregularities, image quality, compressed sensing Since the development of respiratory-correlated four-dimensional computed tomography (RC-4DCT) [1] and magnetic resonance imaging (RC-4DMRI) [2,3], patient-specific respiratory-induced tumor motion has been incorporated in radiotherapy for treating mobile tumors, such as lung, liver and pancreatic cancer. However, the RC-based snapshot 4D imaging only provides one breathing cycle, often contains severe binning motion artifacts, and may not represent tumor motion over 20-minute treatment, affecting treatment outcomes. Therefore, respiratory motion irregularities remain a challenge in radiotherapy. In this study, we report an improved time-resolved 4DMRI technique that captures multi-breath and can be used clinically and quantifies tumor motion irregularities. |
| 3093 | Computer 38
|
High temporal resolution DCE-MRI improves the performance of model fitting parameters |
| Xin Li1, Travis Rice-Stitt2, Lina Gao3, Marina Aguiñaga4, Kevin R Turner5, Bryan Foster6, Fergus Coakley6, Mark Garzotto4,7, and Ryan Kopp4,7 | ||
1Advanced Imaging Research Center, Oregon Health & Science University, Portland, OR, United States, 2Pathology, Oregon Health & Science University, Portland, OR, United States, 3Knight Biostatistics Shared Resources, Oregon Health & Science University, Portland, OR, United States, 4Urology, Oregon Health & Science University, Portland, OR, United States, 5Providence Health and Service, Portland, OR, United States, 6Diagnostic Radiology, Oregon Health & Science University, Portland, OR, United States, 7Portland VA Medical Center, Portland, OR, United States |
||
Keywords: Data Analysis, DSC & DCE Perfusion Recent advance in data acquisition makes fast DCE-MRI data acquisition feasible. This work investigated the impact of DCE-MRI temporal resolution on the data’s pharmacokinetic modeling and the water exchange effect imparted into the model parameters. Using both the Tofts model and the water-exchange sensitized Shutter-Speed model, our results showed that DCE data with higher temporal resolution (shorter intersample interval) is beneficial in model parameter precision and in quantifying transcytolemmal water exchange effect. The later may offer additional lesion-detection specificity in clinical prostate MRI. |
| 3094 | Computer 39
|
Implementation of Hyperpolarized 129Xe Accelerated MRI in Phantom and Human Lungs: Preliminary Study and Troubleshooting |
| Samuel Perron1, Matthew S. Fox1,2, and Alexei V. Ouriadov1,2,3 | ||
1Physics and Astronomy, University of Western Ontario, London, ON, Canada, 2Lawson Health Research Institute, London, ON, Canada, 3School of Biomedical Engineering, University of Western Ontario, London, ON, Canada |
||
Keywords: Data Acquisition, Low-Field MRI Accelerated MRI could significantly improve image quality of lung imaging without increasing costs, especially for low field strengths. The signal decay of a series of undersampled xenon-129 images is fitted to the Stretched-Exponential-Model to yield higher SNR. The proposed method was implemented in undersampled phantom images at low field (0.074T) and human lung images at high field (3T) using the FGRE pulse sequence with hyperpolarized 129Xe; SNR was significantly improved within the same scan duration compared to a fully-sampled image. Potential issues with the compressed-sensing reconstruction are identified and possible solutions presented to promote robustness of method. |
| 3095 | Computer 40
|
Slab-selection in free-running cardiac and respiratory motion-resolved bSSFP 5D whole-heart MRI |
| Robin Ferincz1, Ludovica Romanin1, Jérôme Yerly2, Davide Piccini1,3,4, Matthias Stuber2, and Christopher William Roy1 | ||
1Department of Radiology, Lausanne University Hospital (CHUV) and University of Lausanne (UNIL), Lausanne, Switzerland, 2Center for Biomedical Imaging (CIBM), Lausanne, Switzerland, 3Advanced Clinical Imaging Technology (ACIT), Siemens Healthineers International AG, Lausanne, Switzerland, 4LTS5, École Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland |
||
Keywords: Data Acquisition, Artifacts Recent advances have enabled high resolution cardiac and respiratory motion-resolved 5D whole-heart MRI using non-contrast enhanced non-selective 3D radial bSSFP. However, the nature of bSSFP, the subject dependent anatomy, as well as the underlying sparse reconstruction can lead to banding and streaking artifacts, which degrade image quality and reduce diagnostic utility. In this work, the impact of slab-selective RF pulses in a programmable orientation that is independent of the k-space trajectory is assessed in a cohort of healthy volunteers. Preliminary results suggest that a subject-specific slab orientation can reduce artifacts and improve image quality. |
| 3096 | Computer 41
|
Geometric Constrained Deep Learning for Motion Correction of Fetal Brain MR Images |
| Laifa Ma1,2, Liangjun Chen1, Fenqiang Zhao1, Zhengwang Wu1, Li Wang1, Weili Lin1, He Zhang3, Kenli Lin2, and Gang Li1 | ||
1University of North Carolina at Chapel Hill, Chapel Hill, NC, United States, 2Hunan University, Changsha, China, 3Fudan University, Shanghai, China |
||
Keywords: Image Reconstruction, Brain Robust motion correction of fetal brain MRI slices is crucial for fetal brain volume reconstruction. However, conventional methods can only handle a limited range of motion. Hence, a deep learning model based on prior geometric constraints is proposed to predict the motion of 2D slices. It consists of a global and a relative motion estimation network. Sharing features between two networks make the model to learn more unique feature representations for global motion correction. Moreover, we present a control point-based approach to simulate complex fetal motion trajectories. The experimental results demonstrate that the proposed method is effective and efficient. |
| 3097 | Computer 42
|
Effect of pathology on quantitative metrics of image reconstruction using a deep learning-based brain MRI reconstruction model |
| Shengjia Chen1, Patricia Johnson1, and Yvonne W. Lui1 | ||
1Department of Radiology, New York University Langone Health, New York, NY, United States |
||
Keywords: Image Reconstruction, Brain We evaluate image quality in brain MR images with pathology, reconstructed by a deep learning-based image reconstruction algorithm. We have two main contributions: 1) a procedure for evaluating the image reconstruction quality of images, both globally and in patches with labelled pathology, and 2) report quantitative differences between two groups of reconstructed images (abnormal vs. normal). The pathology evaluation results find pathology regions have more losses and lower structural similarity when compared to normal patches and entire normal brains. |
| 3098 | Computer 43
|
Accelerated acquisition and DL reconstruction to enable single shot fast spin echo (SSFSE) imaging of diagnostic quality using single coil |
| Sudhanya Chatterjee1, Florintina C1, Rohan Patil1, Sajith Rajamani1, Rajagopalan Sundaresan1, Uday Patil1, Preetham Shankapal1, Suresh Emmanuel Joel1, Ramesh Venkatesan1, and Harsh Agarwal1 | ||
1GE Healthcare, Bangalore, India |
||
Keywords: Image Reconstruction, Image Reconstruction, artificial intelligence, abdomen, SSFSE Single shot fast spin echo (SSFSE) is a popular imaging approach for acquisition of high-resolution MR images in motion sensitive areas such as abdomen. In certain clinical settings, use of multiple coils setup for acquisition is not feasible (such as abdominal scans for obese subjects in non-wide bore MRI scanners). SSFSE imaging with single coil using the popular partial Fourier approach only presents risk of excessive blurring in the images. In this study we present a method to enable SSFSE T2 imaging using single coil. Proposed method is evaluated on prospectively accelerated data. |
| 3099 | Computer 44
|
SSMo-QSM: A self-supervised learning method for model-based quantitative susceptibility mapping reconstruction |
| Jie Feng1, Ming Zhang1, Ruimin Feng1, and Hongjiang Wei1 | ||
1School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China |
||
Keywords: Image Reconstruction, Quantitative Susceptibility mapping This study, SSMo-QSM, introduced a self-supervised strategy to achieve a model-based deep learning QSM reconstruction for dealing with the imperfect ground truth in supervised learning methods. In SSMo-QSM, the direct frequency domain division results between phase and dipole kernel at untruncated areas of TKD are randomly separated into two subsets. One is used as the data consistency of the unrolled network and the other is used to define the loss function for training, respectively. The preliminary results of synthetic data suggested that SSMo-QSM performed comparably against supervised methods on accurate susceptibility mapping with suppressed streaking artifact. |
| 3100 | Computer 45
|
Deep Magnetic Resonance Fingerprinting Based on local and global vision transformer |
| Peng Li1, Xiaodi Li1, Xin Lu2, and Yue Hu1 | ||
1Harbin Institute of Technology, Harbin, China, 2De Montfort University, Leicester, United Kingdom |
||
Keywords: Image Reconstruction, MR Fingerprinting Magnetic resonance fingerprinting (MRF) can achieve simultaneous imaging of multiple tissue parameters. However, the size of the tissue fingerprint dictionary used in MRF grows exponentially as the number of tissue parameters increases, which may result in prohibitively large dictionaries that require extensive computational resources. Existing CNN-based methods obtain parameter reconstruction patch-wisely, using only local information and resulting in limited reconstruction speed. In this paper, we propose a novel end-to-end local and global vision transformer (LG-VIT) for MRF parameter reconstruction. The proposed method enables significantly fast and accurate end-to-end parameter reconstruction while avoiding the high computational cost of high-dimensional data. |
| 3101 | Computer 46
|
B0 Inhomogeneity Distortion Corrected Image Reconstruction with Deep Learning on An Open Bore MRI-Linac |
| Shanshan Shan1,2, Yang Gao3,4, Meng Ma5, Hongping Gan5, David Waddington2, Brendan Whelan2, Paul Liu2, Chunyi Liu1, Mingyuan Gao1, and Feng Liu4 | ||
1Center for Molecular Imaging and Nuclear Medicine, State Key Laboratory of Radiation Medicine and Protection,School for Radiological and Interdisciplinary Sciences (RAD-X), Soochow University, Suzhou, China, 2ACRF Image X Institute, Faculty of Medicine and Health, The University of Sydney, Sydney, Australia, 3School of Computer Science and Engineering, Central South University, Changsha, China, 4School of Information Technology and Electrical Engineering, University of Queensland, Brisbane, Australia, 5School of Software, Northwestern Polytechnical University, Suzhou, China |
||
Keywords: Image Reconstruction, Brain MRI-Linac systems require real-time anatomical images with high geometric fidelity to localize and track tumours during radiotherapy treatments. Image distortions caused by B0 field inhomogeneity and slow MR acquisition hinder the application of real-time MRI-guided radiotherapy. Here, we develop and investigate a deep learning-based reconstruction pipeline to reconstruct B0 inhomogeneity distortion-corrected images (B0ReconNet) directly from k-space. MR acceleration techniques such as compressed sensing (CS) were integrated into B0ReconNet to further reduce acquisition time. Simulated and experimental data with fully sampled and retrospectively subsampled acquisitions on a 1T open bore MRI-Linac were used to validate the proposed method. |
| 3102 | Computer 47
|
Accelerated Propeller FSE-DWI with Unrolled Deep Learning Reconstruction at 1.5T Clinical MRI |
| Uten Yarach1, Atita Suwannasak1, and Prapatsorn Sangpin2 | ||
1Radiologic Technology, Chiang Mai University, Chiang Mai, Thailand, 2Philips Healthcare (Thailand), Bangkok, Thailand |
||
Keywords: Image Reconstruction, Brain Fast spin echo diffusion magnetic resonance imaging (FSE-DWI) is often referred to as a standard for MRI diagnosis of Cholesteatoma. However, the acquired data require multiple steps during image reconstruction which turn out high residual artifacts. In this work, we develop rapid reconstruction for propeller FSE-DWI to improve its signal-to-noise ratio (SNR) through unrolled deep learning (DL) framework. Results show that the proposed unrolled DL reconstruction enables increasing bout 2x SNR compared to SNR obtained by online reconstructed images. Moreover, its speed is about 200x faster than conventional locally low rank constraint reconstruction. |
| 3103 | Computer 48
|
Unsupervised model for removing Nyquist/motion artifacts in SPatiotemporal Encoding MRI |
| Qingjia Bao1, Liyang Xia2, Kewen Liu2, Xinjie Liu1, Peng Sun3, Lucio Frydman4, and Chaoyang Liu1 | ||
1Innovation Academy for Precision Measurement Science and Technology, Wuhan, China, 2Wuhan University of Technology, Wuhan, China, 3Philips Healthcare, Beijing, China, 4Weizmann Institute of Science, Rehovot, Israel |
||
Keywords: Image Reconstruction, Susceptibility Although a major advantage of SPEN vs EPI is a higher immunity to artifacts, it suffers from Nyquist or motion artifacts. We proposed a new unsupervised CNN model that takes advantage of both physical model and Deep learning. The model consists of three parts: phase feature extraction module, which can extract the phase features of even/odd phase differences or motion-caused phase differences in multi-shot echo data. Then, the phase maps are generated with these phase difference features. Lastly, the phase correction modules to remove artifacts. The results show that the proposed model can effectively correct Nyquist/motion artifacts in single-shot/multi-shot SPEN. |
| 3104 | Computer 49
|
Task-based evaluation of deep learning-based reconstruction for highly-accelerated 3D T1-weighted brain MRI scans |
| Sangtae Ahn1, Chitresh Bhushan1, John Huston2, J. Kevin DeMarco3, Robert Y. Shih3,4, Joshua D. Trzasko2, Rafi Brada5, Graeme Mckinnon6, Isabelle Heukensfeldt Jansen1, Dan Rettmann7, Brian Burns8, Ty A. Cashen6, Nir Mazor5, Xucheng Zhu8, and Thomas K. Foo1 | ||
1GE Research, Niskayuna, NY, United States, 2Mayo Clinic College of Medicine, Rochester, MN, United States, 3Walter Reed National Military Medical Center, Bethesda, MD, United States, 4Uniformed Services University of the Health Sciences, Bethesda, MD, United States, 5GE Research, Herzliya, Israel, 6GE HealthCare, Waukesha, WI, United States, 7GE HealthCare, Rochester, MN, United States, 8GE HealthCare, Menlo Park, CA, United States |
||
Keywords: Image Reconstruction, Machine Learning/Artificial Intelligence, Brain, Neuro 3D MRI enables thin slices at the cost of long scan times, causing practical challenges. Recently, deep-learning (DL) techniques have successfully accelerated MR scans. However, it is challenging to characterize the image quality (IQ) performance of DL methods by conventional metrics because IQ depends on applications, i.e., how images are used. We evaluate the IQ performance of DL-Speed, our DL-based acceleration method, for 3D T1-weighted MPRAGE brain scans, in 1) post-reconstruction subcortical structure segmentation, and 2) a reader study. The results imply DL-Speed can accelerate scans with reduction factor R=10 while maintaining IQ comparable to standard parallel imaging with R=2.1. |
| 3105 | Computer 50
|
Improved diagnostic efficacy on structural abnormalities of knee using high-resolution deep learning-based 2D FSE images |
| Xiaxia Wu 1, Weiyin Vivian Liu2, and Yunfei Zha1 | ||
1Renmin Hospital of Wuhan University, Wuhan, China, 2GE Healthcare,MR Reaearch China,Beijing, Wuhan, China |
||
Keywords: Image Reconstruction, Cartilage Diagnostic performance was limited to image resolution and contrast between target tissues and surrounding tissues. A rapid knee imaging has been perused but no loss of image quality is critical. This study proposed a rapid knee imaging based on two-dimensional fast spin echo sequence and examined the reliability and diagnostic performance of deep learning-based reconstruction T1-, T2- and PD- weighted images on knee joint pathology via comparison of images with and without deep-learning reconstruction algorithm (DLR). Diagnostic efficacy on knee structural abnormalities of 2D DLR FSE sequence elevated using knee arthroscopy results as the gold standard. |
| 3106 | Computer 51
|
Motion-mitigated reconstruction of accelerated MRI by using an unfolded variational network |
| Zijian Zhou1,2, Haikun Qi1,2, and Peng Hu1,2 | ||
1School of Biomedical Engineering, ShanghaiTech University, Shanghai, China, 2Shanghai Clinical Research and Trial Center, Shanghai, China |
||
Keywords: Image Reconstruction, Motion Correction Motion-mitigated reconstruction of highly undersampled MRI was achieved by adding a motion estimation module to the data consistency part of the model-based unfolded variation network. The motion estimation module consisted of a pair of convolutional blocks with residual inputs and added only limited number of trainable parameters to the network. The network was trained and tested on synthesized motion-corrupted images from a publicly available knee dataset. The reconstructed images with the proposed motion estimation module were sharper, and details were better recovered, with the structural similarity and peak signal-to-noise ratio significantly improved. |
| 3107 | Computer 52
|
Does CS-SENSE acceleration influence the performance of an AI based synthetic CT algorithm? A volunteer study in the lumbar spine. |
| Yulia Shcherbakova1, Tijl A van der Velden1,2, and Peter R Seevinck1,2 | ||
1Imaging Division, UMC Utrecht, Utrecht, Netherlands, 2MRIguidance B.V., Utrecht, Netherlands |
||
Keywords: Image Reconstruction, Skeletal, AI, acceleration, CS-SENSE, spine, sCT, bone In this work, we investigated the performance of an AI based algorithm - synthetic CT generation - when subjected to compressed sensing-sensitivity encoding (CS-SENSE) accelerated gradient echo images. We performed MR experiments in five volunteers, using different CS-SENSE acceleration factors for the MR acquisitions. Our results showed that using CS-SENSE factors of 1.45 and 2 increased noise in the MR source images but did not compromise the sCT reconstruction on visual inspection, which was confirmed by quantitative metrics. However, CS-SENSE with a factor of 3 caused artifacts in the sCT images which may affect the safety and diagnostic performance of the product. |
| 3108 | Computer 53
|
Learned Tensor Low-CP-Rank and Bloch response manifold priors for Non-Cartesian MRF Reconstruction |
| Peng Li1, Xiaodi Li1, Xin Lu2, and Yue Hu1 | ||
1Harbin Institute of Technology, Harbin, China, 2De Montfort University, Leicester, United Kingdom |
||
Keywords: Image Reconstruction, MR Fingerprinting, Tensor Low-rank, CP Decomposition, Bloch Response Manifold, non-Cartesian We propose a deep unrolled network for non-Cartesian MRF reconstruction by unrolling the MRF reconstruction model regularized by the tensor low-rank and the Bloch resonance manifold priors. To avoid computationally burdensome singular value decomposition, we propose a learned CP decomposition module to exploit the tensor low-rank priors of MRF data. Inspired by the MRF imaging mechanism, we also propose a Bloch response manifold module to learn the mapping between reconstructed MRF data and the multiple parameter maps. Numerical experiments show that the proposed network can improve the reconstruction quality of MRF data and multi-parameter maps within significantly reduced computational time. |
| 3109 | Computer 54
|
DTI-Net: Unsupervised Diffusion Tensor Reconstruction Using Implicit Neural Representation |
| Yuting Shi1, Yuyao Zhang2, and Hongjiang Wei1 | ||
1Shanghai Jiao Tong University, Shanghai, China, 2ShanghaiTech University, Shanghai, China |
||
Keywords: Image Reconstruction, Machine Learning/Artificial Intelligence Diffusion tensor imaging (DTI) needs a large number of diffusion-weighted images (DWIs) to reliably reconstruct the diffusion measurements of the brain white matter, making the data acquisition time-consuming. Deep learning has emerged as a powerful technique to reduce the number of acquired DWIs. While most existing deep learning methods are supervised and need high-quality ground truth data as the training labels. Here, we proposed an unsupervised and subject-specific DTI reconstruction method called DTI-Net to significantly reduce the required number of DWIs, while also can simultaneously conduct the super-resolution reconstruction of the tensors. |
| 3110 | Computer 55
|
High efficient DTI reconstruction network with flexible diffusion directions |
| Zejun Wu1, Jiechao Wang1, Zunquan Chen1, Zhigang Wu2, Jianfeng Bao3, Shuhui Cai1, and Congbo Cai1 | ||
1Department of Electronic Science, Xiamen University, Xiamen, China, 2Philips Healthcare, Shenzhen, China, 3the First Affiliated Hospital of Zhengzhou University, Zhengzhou, China |
||
Keywords: Image Reconstruction, Diffusion Tensor Imaging Deep learning has been used in diffusion tensor imaging (DTI) to fast reconstruct diffusion parameters. However, diffusion-weighted images (DWIs) as network input must maintain diffusion gradient direction consistency during training and testing for deep-learning-based DTI parameter mapping. A dynamic-convolution-based network was developed to achieve generalized DTI parameter mapping for flexible diffusion gradient directions. This proposed method uses dynamic convolution kernels to embed diffusion gradient direction information into feature maps of the corresponding diffusion signal. The results indicate that the proposed method can reconstruct high-quality DTI-derived maps from six diffusion gradient directions. |
| 3111 | Computer 56
|
GROG Gridding using modified VGG-16 CNN model for Non-Cartesian MR Image Reconstruction |
| Muhammad Atif1, Madiha Arshad1, Yumna Bilal1, Omair Inam1, Hassan Shahzad2, and Hammad Omer1 | ||
1Medical Image Processing Research Group (MIPRG), Department of Electrical and Computer Engineering, COMSATS University, Islamabad, Pakistan, 2National Centre of Physics (NCP), Islamabad, Pakistan |
||
Keywords: Image Reconstruction, Image Reconstruction Self-Calibrating GROG (SC-GROG) is a gridding algorithm that maps the k-space MRI data from non-Cartesian to Cartesian domain. The main limitation of SC-GROG is its computational cost to calculate the GROG weights. This paper proposes a customized deep learning framework (based on VGG-16 CNN model) to calculate the 2D-Gridding weight sets for SC-GROG. Initially, the proposed model is trained on human head images, and later fine-tuning is performed using Golden-angle radial Liver Perfusion datasets. The results show that the proposed method significantly reduces the computation time for the estimation of GROG weights while maintaining the image quality. |
| 3112 | Computer 57
|
Accelerating 7T susceptibility-weighted imaging with complex-valued convolutional neural network |
| Caohui Duan1, Xiangbing Bian1, Kun Cheng1, Xiaoyu Wang1, Jinhao Lyu1, Xueyang Wang1, Jianxun Qu2, Xin Zhou3, and Xin Lou1 | ||
1Department of Radiology, Chinese PLA General Hospital, Beijing, China, 2MR Collaboration, Siemens Healthineers Ltd., Beijing, China, 3Key Laboratory of Magnetic Resonance in Biological Systems, State Key Laboratory of Magnetic Resonance and Atomic and Molecular Physics, National Center for Magnetic Resonance in Wuhan, Wuhan Institute of Physics and Mathematics, Innovation Academy for Precision Measurement Science and Technology, Chinese Academy of Sciences‒Wuhan National Laboratory for Optoelectronics, Wuhan, China |
||
Keywords: Image Reconstruction, Machine Learning/Artificial Intelligence Ultra-high field 7T susceptibility-weighted imaging (SWI) has shown great potential in visualizing and evaluating a broad range of pathology, but suffers from long acquisition times. In this study, a complex-valued convolutional neural network (ComplexNet) model was proposed to reconstruct highly accelerated 7T SWI data. The average reconstruction time of ComplexNet was 0.56 seconds per slice (45.16 seconds per participant). Meanwhile, ComplexNet can provide high-quality 7T SWI for visualizing subtle pathology, including cerebral microbleeds, asymmetric deep medullary veins, and swallow tail sign. |
| 3113 | Computer 58
|
Parallel non-Cartesian Spatial-Temporal Dictionary Learning Neural Networks (stDLNN) for Accelerating Dynamic MRI |
| Zhijun Wang1, Huajun She1, and Yiping P. Du1 | ||
1School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China |
||
Keywords: Image Reconstruction, Machine Learning/Artificial Intelligence Dynamic MRI shows promising clinical values and several applications have been investigated such as cardiac, pulmonary, and hepatic imaging. However, a successful application of dynamic MRI is hampered by its time-consuming acquisition. To improve the performance and interpretability for the accelerating reconstruction methods, we proposed the parallel non-Cartesian Spatial-Temporal Dictionary Learning Neural Networks (stDLNN), which combines the traditional spatial-temporal dictionary learning methods with the deep neural networks for accelerating dynamic MRI. It has favorable interpretability and provides better image quality than the state-of-the-art CS methods (L+S, BCS) and deep learning methods (DCCNN, PNCRNN), especially at high acceleration rate at R=25. |
| 3114 | Computer 59
|
Dynamic MRI using Learned Transform-based Tensor Low-Rank Network (LT$$$^2$$$LR-Net) |
| Yinghao Zhang1, Xiaodi Li1, Peng Li1, and Yue Hu1 | ||
1School of Electronics and Information Engineering, Harbin Institute of Technology, Harbin, China, China |
||
Keywords: Image Reconstruction, Machine Learning/Artificial Intelligence Tensor low-rank models have recently emerged as powerful alternative representations for three-dimensional dynamic MR datasets. In this paper, we introduce a novel learned transform-based tensor low-rank network for dynamic MRI based on the tensor singular value decomposition (t-SVD). Instead of manually designing the t-SVD-based transform, we propose to utilize CNN to adaptively learn the relatively optimal transformation from the dynamic MR dataset for more robust and accurate tensor low-rank representations. Experimental results on cardiac cine MRI reconstruction demonstrate the superior performance of the proposed framework compared with the state-of-the-art methods. |
| 3115 | Computer 60
|
Self-Supervised Image Reconstruction of 7T MP2RAGE for Multiple Sclerosis: 0.5mm Isotropic Resolution in 10 Minutes |
| Thomas Yu1,2,3, Francesco La Rosa4, Gian Franco Piredda1,5,6, Jonadab Dos Santos Silva4, Faye Bourie4, Henry Dieckhaus7, Govind Nair7, Patrick Liebig8, Jean-Philippe Thiran2,3,5, Tobias Kober1,2,3, Erin Beck4,7, and Tom Hilbert1,2,3 | ||
1Advanced Clinical Imaging Technology, Siemens Healthineers International AG, Lausanne, Switzerland, 2Signal Processing Laboratory 5 (LTS5), Ecole Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland, 3Department of Radiology, Lausanne University Hospital and University of Lausanne, Lausanne, Switzerland, 4Department of Neurology, Icahn School of Medicine at Mount Sinai, New York, NY, United States, 5Centre d’Imagerie Biomédicale (CIBM), EPFL, Lausanne, Switzerland, 6Human Neuroscience Platform, Fondation Campus Biotech Geneva, Geneva, Switzerland, 7National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, MD, United States, 8Siemens Healthineers International AG, Erlangen, Germany |
||
Keywords: Image Reconstruction, Machine Learning/Artificial Intelligence, Ultra-High Field MRI High-resolution 3D MR imaging is necessary for the detailed assessment of focal pathologies, such as cortical lesions. However, high-resolution demands tradeoffs with acceleration and SNR, which is difficult to address with standard machine learning reconstructions due to the infeasibility of collecting large datasets of fully sampled data. Using a dataset of high-resolution (0.5mm isotropic), 3D, 7T MP2RAGE scans of multiple sclerosis patients, we show that a self-supervised reconstruction from one scan, requiring no fully sampled data, has higher apparent SNR than a median of three scans, currently used for assessment, with comparable tissue contrast and lesion conspicuity. |
| 3116 | Computer 61
|
Improving the accuracy of Multipool-Lorentzian fitting in CEST MRI by use of a Pseudo-Voigt lineshape for direct water saturation |
| Markus Huemer1, Clemens Stilianu1, and Rudolf Stollberger1,2 | ||
1Institute of Biomedical Imaging, Graz University of Technology, Graz, Austria, 2BioTechMed Graz, Graz, Austria |
||
Keywords: Data Analysis, CEST & MT In Chemical Exchange Saturation Transfer (CEST) MRI one common analysis technique is Multipool-Lorentzian fitting of the resulting Z-Spectrum. Multiple Lorentzian lineshapes are fitted to the measured data. The lineshape of the direct water saturation (DS) can deviate substantially from a Lorentzian depending on the saturation scheme and $$$B_1^+$$$. To prevent a cross-influence of the DS on the CEST and increase the fit accuracy we propose the use of a Pseudo-Voigt lineshape for the water line. |
| 3117 | Computer 62
|
Deformable Groupwise Registration Using a Locally Low-Rank Dissimilarity for Myocardial Strain Estimation from Cine MRI Images |
| Haiyang Chen1 and Chenxi Hu1 | ||
1The Institute of Medical Imaging Technology, School of Biomedical Engineering, Shanghai Jiao Tong University (SJTU), Shanghai, China |
||
Keywords: Data Analysis, Data Processing, registration We propose a deformable groupwise registration method using a locally low-rank (LLR) dissimilarity to estimate myocardial strain from cine MRI images. The proposed method eliminates the drift effect commonly observed in the optical flow and sequentially pairwise registration, facilitating more accurate strain estimation in the diastolic phase. Compared to the globally low-rank dissimilarity, LLR dissimilarity shows slightly better tracking accuracy by imposing the low-rank property in local image regions rather than the whole image. Experiments on a large public cine MRI dataset demonstrates the accuracy of the proposed method on tracking and strain estimation. |
| 3118 | Computer 63
|
Digital Reference Objects with BART |
| Nick Scholand1,2, Martin Schilling3, Martin Heide3, and Martin Uecker1,2,3 | ||
1Institute of Biomedical Imaging, Graz University of Technology, Graz, Austria, 2German Centre for Cardiovascular Research (DZHK), Partner site Göttingen, Göttingen, Germany, 3Department of Interventional and Diagnostic Radiology, University Medical Center Göttingen, Göttingen, Germany |
||
Keywords: Software Tools, Software Tools Digital Reference Objects (DRO) are important for developing imaging and reconstruction techniques especially in medical imaging applications. They allow a stepwise increase in complexity during testing of methods by simulating reproducible datasets. In this work we present a variety of tools the BART toolbox provides for the creation and simulation of DROs. We present a workflow for designing DROs with Bézier curves using vector graphics applications which are then imported into BART. |
| 3119 | Computer 64
|
Tract-specific myelin mapping using magnetization transfer-prepared diffusion imaging: comparison with conventional MTR tractometry |
| Wen Da Lu1,2, Mark Cameron Nelson2,3, Ilana Ruth Leppert2, Pietro Bontempi4, Simona Schiavi4,5, Christopher Dennis Rowley1,2, Alessandro Daducci4, and Christine Lucas Tardif1,2,3 | ||
1Department of Biological and Biomedical Engineering, McGill University, Montreal, QC, Canada, 2McConnell Brain Imaging Centre, Montreal Neurological Institute and Hospital, Montreal, QC, Canada, 3Department of Neurology and Neurosurgery, McGill University, Montreal, QC, Canada, 4Department of Computer Science, University of Verona, Verona, Italy, 5Department of Neuroscience, Rehabilitation, Ophthalmology, Genetics, Maternal and Child Health (DINOGMI), University of Genoa, Genova, Italy |
||
Keywords: Data Analysis, Tractography & Fibre Modelling, Magnetization Transfer, Diffusion-Weighted Imaging, Tractometry Tractometry is a technique used to investigate the microstructure variations along white matter tracts or to reconstruct microstructure-weighted connectivity matrices. However, partial volume effects from crossing fibers bias the individual fiber measurements and conceal subtle differences. Here, we compare MTR tractometry to a novel approach using dual-encoded MT-weighted DWI analyzed using COMMIT to disentangle the MTR signal of individual white matter fibers. The results show a broader distribution in edge MTR values for the dual-encoding approach in comparison to tractometry. Both techniques show similar spatial MTR patterns, however the dual-encoded MT-weighted DWI approach shows a higher pattern of inter-individual variability. |
| 3120 | Computer 65
|
Automatic Localization of EEG Electrodes from MR Structural Images |
| Mohammadreza Rezaei-Dastjerdehei1 and Pierre LeVan2 | ||
1Department of Biomedical Engineering, University of Calgary, Calgary, AB, Canada, 2Department of Radiology, University of Calgary, Calgary, AB, Canada |
||
Keywords: Data Analysis, Segmentation, EEG Electrodes Localization, T1-weighted MRI The localization of EEG sources is one of the fundamental approaches to facilitate the interpretation of EEG data. However, the accuracy of source localization depends on the exact knowledge of the position of the electrodes on the scalp, which currently requires time-consuming and/or expensive approaches. Here, an automatic method is proposed that retrieves the electrode positions by localizing the curvature changes in T1-weighted MRI images caused by electrodes. The results show an average detection sensitivity of ~96.4%, with an average position error of 4.23 mm for all subjects. |
| 3121 | Computer 66
|
Habitats Analysis in Hepatocellular Carcinoma to Predict Microvascular Invasion by Intravoxel Incoherent Motion: A Pilot Study |
| Chenhui Li1, Jinhuan Xie1, Liling Long1, Huiting Zhang2, and Yang Song2 | ||
1The First Affiliated Hospital of Guangxi Medical University, Nanning, China, 2MR Scientific Marketing, Siemens Healthineers, Shanghai, China |
||
Keywords: Data Analysis, Segmentation, Habitats The method of delineating the ROI of whole lesions on quantitative parameter images and then averaging them for comparison did not accurately quantify heterogeneity. In this study, we adopted a habitats analysis method combined with tissue cellularity and blood flow information from IVIM model to segment whole tumor to four subregions to predict microvascular invasion (MVI) in hepatocellular carcinoma. The results show that habitats analysis predicts MVI positivity with an accuracy of 70.19%, and the averaged value of each parameter in whole tumor was not predictive for MVI. This provides a good starting point for further application of this method. |
| 3122 | Computer 67
|
A robust image analysis pipeline for high fidelity kurtosis and tensor fitting of 1.2mm isotropic infant brain diffusion MRI |
| Tianjia Zhu1,2, Minhui Ouyang1,3, Kay Sindabizera1, Juri Kim2, and Hao Huang1,3 | ||
1Department of Radiology, Children's Hospital of Philadelphia, Philadelphia, PA, United States, 2Department of Bioengineering, University of Pennsylvania, Philadelphia, PA, United States, 3Department of Radiology, University of Pennsylvania, Philadelphia, PA, United States |
||
Keywords: Data Processing, Pediatric, Neuro High-resolution infant diffusion MRI (dMRI) data poses specific challenges such as severe motion artifacts, long acquisition time for multi-shell data, and low signal-to-noise (SNR). We present a robust image analysis pipeline for high fidelity kurtosis and tensor fitting of 1.2mm isotropic infant brain dMRI that allows us to efficiently analyze in-vivo dMRI data despite the considerable technical challenges specific to infant imaging. State-of-the-art preprocessing including slice-to-volume motion correction and susceptibility-by-movement correction is combined with advanced self-supervised learning-based denoising to produce high fidelity diffusion tensor and diffusion kurtosis fitting. |
| 3123 | Computer 68
|
Arterial phase detection when utilizing stack-of-stars dynamic Data Set with various pseudo contrast enhancement |
| Yutaka Hoshiyama1, Chunqi Wang2, Hong Yang2, Hideki Ota3,4, Atsuro Masuda4, Masahiro Kawabata4, Hideaki Kutsuna1, Kensuke Shinoda1, and Yoshimori Kassai1,3 | ||
1MRI Systems Division, Canon Medical Systems Corporation, Ohtawara, Japan, 2Research & Development Center, Canon Medical Systems (China) Co., Ltd, Beijing, China, 3Department of Advanced MRI Collaboration Research, Tohoku University Graduate School of Medicine, Sendai, Japan, 4Department of Diagnostic Radiology, Tohoku University Hospital, Sendai, Japan |
||
Keywords: Data Analysis, Contrast Agent, Dynamic We propose an automated method to present candidates of the arterial phase as the optimal time points from the liver contrast-enhanced MRI images based on the time intensity curve (TIC) analysis of the several ROIs. In clinical situation, there may be some differences of TIC due to a speed of uptake and washout and a quality of the image. In this study, the robustness of our proposed automated method was investigated by utilizing stack-of-stars dynamic images with various pseudo contrast enhancement effects. The results showed that our proposed automated method was applicable for such effects. |
| 3124 | Computer 69
|
Subject-specific analysis reveals spatially heterogeneous white matter abnormalities in sports-related concussion |
| Ho-Ching Yang1, Mario Dzemidzic1,2, Qiuting Wen1, Larry D Riggen Jr3, Steven P Broglio4, Michael A McCrea5, Thomas W McAllister6, Jaroslaw Harezlak7, and Yu-Chien Wu1,8 | ||
1Department of Radiology and Imaging Sciences, Indiana University School of Medicine, Indianapolis, IN, United States, 2Department of Neurology, Indiana University School of Medicine, Indiana University School of Medicine, Indianapolis, IN, United States, 3Department of Biostatistics and Health Data Science, Indiana University School of Medicine, Indianapolis, India, 4Michigan Concussion Center, University of Michigan, Ann Arbor, MI, United States, 5Department of Neurosurgery, Medical College of Wisconsin, Milwaukee, WI, United States, 6Department of Psychiatry, Indiana University School of Medicine, Indianapolis, IN, United States, 7Department of Epidemiology and Biostatistics, Indiana University School of Public Health, Bloomington, IN, United States, 8Stark Neurosciences Research Initiative, Indiana University School of Medicine, Indianapolis, IN, United States |
||
Keywords: Data Analysis, Traumatic brain injury, Subject-specific analysis, Heterogeneous, Concussion Sport-related concussion (SRC) injury inclines to cause subject-specific brain region abnormalities. This study applied subject-specific analysis that accounts for inter-subject variation to investigate heterogeneity of white matter alterations in a sample from a large national multicenter study, the Concussion Assessment, Research and Education Consortium. Our results demonstrated that subject-specific white matter abnormalities in SRC can be uncovered by calculating the extreme Z-score maps based on a template generated from normally distributed diffusion tensor imaging metric values in non-contact sports controls. Our finding indicates that the SRC-induced white matter abnormalities show heterogeneous spatial distribution across participants. |
| 3125 | Computer 70
|
Multidimensional optimal denoising with block matching is robust to misalignment and contrast changes |
| Khoi Minh Huynh1, Sang Hun Chung1, Yueh Lee1, and Pew-Thian Yap1 | ||
1Department of Radiology and Biomedical Research Imaging Center, UNC Chapel Hill, Chapel Hill, NC, United States |
||
Keywords: Data Processing, Lung Denoising methods can leverage common information from multiple image volumes for better noise reduction. However, performance can degrade with inter-volume misalignment or contrast changes. We proposed a block-matching denoising method for effective noise removal and robustness to inter-volume differences. |
| 3126 | Computer 71
|
Improving interpretation accuracy for curved planar reformation of intracranial vessel wall MRI by reducing intra-plane rotation |
| Xin Wang1, Yin Guo2, Kaiyu Zhang2, Gador Canton3, Niranjan Balu3, Thomas S. Hatsukami4, Mahmud Mossa-Basha5, and Chun Yuan3 | ||
1Department of Electrical and Computer Engineering, University of Washington, Seattle, WA, United States, 2Department of Bioengineering, University of Washington, Seattle, WA, United States, 3Department of Radiology, University of Washington, Seattle, WA, United States, 4Department of Surgery, University of Washington, Seattle, WA, United States, 5Department of Radiology, University of North Carolina-Chapel Hill, Chapel Hill, NC, United States |
||
Keywords: Data Processing, Visualization This work aims to develop an algorithm for straightened curved planar reformation (CPR), i.e., extracting the 2D cross-sectional images along a vessel centerline in a vessel wall MRI scan. Previous methods determine the orientation of CPR images in a fixed way, leading to undesired rotation along the centerline. Our method instead calculates the local coordinate system with respect to previous slices, thus keeping the continuity of cross-sections. Our CPR results show a better readability of the straightened vessel image, especially when one wants to utilize the spatial information among adjacent CPR slices for a more comprehensive and accurate medical analysis. |
| 3127 | Computer 72
|
Improved pulmonary oxygen-enhanced MRI parameter precision enabled by hierarchical Bayesian inference |
| Josephine H Naish1,2, Marta Tibiletti1, Christopher Short3,4, Tom Semple3,4, Simon Padley3,4, Jane C Davies3,4, and Geoff JM Parker1,5 | ||
1Bioxydyn Ltd, Manchester, United Kingdom, 2MCMR, Manchester University NHS Foundation Trust, Manchester, United Kingdom, 3National Heart & Lung Institute, Imperial College London, London, United Kingdom, 4Royal Brompton Hospital, Guy's & St Thomas’ Trust, London, United Kingdom, 5Centre for Medical Image Computing, Department of Medical Physics and Biomedical Engineering, University College London, London, United Kingdom |
||
Keywords: Data Analysis, Lung, Bayes Lung oxygen-enhanced MRI can provide regional information relating to lung function, but voxel-wise parameter estimation is hampered by low SNR. Here we present a hierarchical Bayesian approach to voxel-wise parameter estimation implemented in R and the probabilistic programming language Stan. In both simulations and in OE-MRI data acquired in patients with cystic fibrosis, the Bayesian approach results in substantially less noisy parameter maps compared to conventional non-linear least squares estimation. |
| 3128 | Computer 73
|
Assessment of tumor subregion complexity using fractal analysis in breast cancer |
| Run Xu1, Guanwu Li1, Dan YU2, Yongming Dai2, and Xinyue Liang2 | ||
1Department of Radiology, Yueyang Hospital of Integrated Traditional Chinese and Western Medicine, Shanghai University of Traditional Chinese Medicine, Shanghai, China, 2Central Research Institute, United Imaging Healthcare, Shanghai, China |
||
Keywords: Data Analysis, Cancer This study aimed to develop multiparametric physiologic MRI-based spatial habitat analysis and to validate the association between the habitats and the molecular subtypes. The distinct tumor habitats were identified using multiple MRI-parameter maps, then applied to calculate the volume fraction and fractal dimension (FD) and to investigate their diagnosis value. The combination of FD and volume fraction improves the discriminatory ability for TNBC from non-TNBC with an AUC to 0.951 from 0.853. |
| 3129 | Computer 74
|
Identification of Subject-Specific Motion Variations via Atlas-Based Deformation Refinement from Real-Time MRI |
| Fangxu Xing1, Riwei Jin2, Imani Gilbert3, Georges El Fakhri1, Jamie L. Perry3, Bradley P. Sutton2, and Jonghye Woo1 | ||
1Radiology, Harvard Medical School, Boston, MA, United States, 2University of Illinois at Urbana-Champaign, Champaign, IL, United States, 3East Carolina University, Greenville, NC, United States |
||
Keywords: Data Analysis, Data Analysis, Speech, motion analysis, atlas Dynamic magnetic resonance imaging has become increasingly efficient at capturing speech deformations of the velopharyngeal region in real time. With previously developed dynamic vocal tract atlases, quantification of group deformation statistics of a population in speech has become possible in a common atlas space. However, subject-specific deformation characteristics as an underlying property tend to be hidden in the spatial and temporal alignment process of atlas construction. We present a registration-based deformation characterization method that extracts subject-specific motion variations in two layers of registration steps. A dataset of fifteen human subjects is processed to reveal unique deformation patterns of each subject. |
| 3130 | Computer 75
|
MRI for tissue classification of subcutaneous colon cancer using oxygen enhanced (OE), dynamic contrast enhanced (DCE) MRI and [18F]FAZA PET |
| Marta Vuozzo1, Max Zimmermann2, Mystele Tendonge3, Petros Martirosian2, Manfred Kneilling4, Fritz Schick2, Bernd Pichler3, and Andreas Schmid2 | ||
1Werner Siemens Imaging Center, University Hospital Tübingen, Tübingen, Germany, 2University Hospital Tübingen, Tübingen, Germany, 3Univeristy Hospital Tübingen, Tübingen, Germany, 4Univerity Hospital Tübingen, Tübingen, Germany |
||
Keywords: Data Analysis, Cancer, Multimodal Solid tumors exhibit intratumoral heterogeneity, which is related to therapy efficacy. Since biopsies represent a small part of the tumor, multimodal imaging techniques provide panoptic cancer characterization. We developed an acquisition and analysis protocol based on multiparametric MRI to classify and provide a holistic characterization of intratumoral heterogeneity and correlate it with positron emission tomography and histology. We applied MRI to characterize phenotypic changes induced by antiPD-L1 therapy response. Results reveal that models trained exclusively with MRI data provide biologically relevant maps of phenotypes showing intratumoral heterogeneity, but also allow non-invasive identification of tumors that respond or resist to therapy. |
| 3131 | Computer 76
|
Diaphragm Excursion Measured by Free-breathing 1H MRI in Patients with Asthma: Repeatability and Bronchodilator Response |
| Yonni Friedlander1,2, Dante P.I. Capaldi3, Norm Konyer1, Melanie Kjarsgaard2, Carmen Venegas2,4, Parameswaran Nair2,4, and Sarah Svenningsen1,2,4 | ||
1Imaging Research Centre, St. Joseph's Healthcare Hamilton, Hamilton, ON, Canada, 2Firestone Institute of Respiratory Health, St. Joseph's Healthcare Hamilton, Hamilton, ON, Canada, 3Deparment of Radiation Oncology, University of California San Francisco, San Francisco, CA, United States, 4Division of Respirology, McMaster University, Hamilton, ON, Canada |
||
Keywords: Data Analysis, Lung Using free-breathing 1H MRI, diaphragm excursion was measured in 9 patients with severe asthma at two visits 4-months apart. At each visit, patients with asthma were imaged before and after administration of a bronchodilator. Diaphragm excursion measured at the two visits were not significantly different from one another (mean bias=0.34cm, p=0.07) and were well correlated (r=0.53, p=0.004), demonstrating repeatability of the measurement. In addition, diaphragm excursion was significantly increased following bronchodilator administration (p=0.01). |
| 3132 | Computer 77
|
Improving Low Resolution MRI Contrast for Brain-Environment Neuroepidemiology |
| Jasmine D Cakmak1, Reza Azarpazhooh2, Alexander Khaw2, and Udunna Anazodo1,2 | ||
1McGill University, Montreal, QC, Canada, 2Western University, London, ON, Canada |
||
Keywords: Data Processing, Segmentation, Synthetic MRI Recycling large-scale clinical data to retrospectively study associations of brain imaging variables with emerging environmental risk factors is a sustainable approach in the growing field of neuroepidemiology. Here, we evaluated the utility of a deep learning tool to increase the resolution of clinical (1.5T) T1-weighted MRI by comparing global assessment of gray matter volume (GMV) and cortical thickness (CT) to 3T research scans. Overall, the resolved 1.5T images had higher (18%) GMV and CT compared to 3T and were more biased than unresolved images. Further analysis in larger cohorts using improved segmentation approaches could validate recycling of enhanced clinical scans. |
| 3133 | Computer 78
|
Head-to-head Comparison of Free Water Elimination Methods in Fornix |
| Ken Sakaie1, Katherine Koenig1, and Mark J. Lowe2 | ||
1The Cleveland Clinic, Cleveland, OH, United States, 2Imaging Institute, The Cleveland Clinic, Cleveland, OH, United States |
||
Keywords: Data Analysis, Alzheimer's Disease Diffusion Tensor Imaging measures in fornix may serve as a biomarker in Alzheimer’s disease. However, partial volume averaging between the fornix and surrounding cerebrospinal fluid imposes systematic bias that can conflate tissue microstructure with atrophy. Free Water Elimination methods have been proposed as a solution, but comparison between these methods is limited. We perform direct comparison between the approaches in terms of repeatability and bias. |
| 3134 | Computer 79
|
Microscopic fractional anisotropy with free water elimination |
| Nico J J Arezza1,2, Mohammad Omer1, and Corey A Baron1,2 | ||
1Medical Biophysics, Western University, London, ON, Canada, 2Centre for Functional and Metabolic Mapping, Robarts Research Institute, London, ON, Canada |
||
Keywords: Data Processing, Neuro Water diffusion anisotropy is a diffusion MRI metric that is sensitive to brain injury and neurodegeneration. Microscopic fractional anisotropy (μFA) is a new metric of diffusion anisotropy that is immune to crossing fiber effects, unlike traditional FA, which enables probing of axon integrity in gray matter and crossing white matter tracts. However, μFA is underestimated in voxels adjacent to cerebrospinal fluid (CSF) due to partial volume contamination, which is particularly problematic in cortical gray matter and tissue adjacent to ventricles. Here, we demonstrate a free water elimination method to remove the CSF signal from μFA measurements in four healthy volunteers. |
| 3135 | Computer 80
|
Feature selection to facilitate surgical planning from MRI of Placenta Accreta Spectrum Disorder |
| Joanna Chappell1, Nada Mufti1,2, Hassna Irzan1, Pat O’Brien3, George Attilakos3, Magda Sokolska4, Priya Narayanan3, Trevor Gaunt5, Paul Humpheries5, Premal Patel5, Elspeth Whitby6, Eric Jauniaux2,3, J. Ciaran Hutchinson5, Neil Sebire5, David Atkinson7, Giles Kendall2,4, Sebastien Ourselin 1, Tom Vercauteren1, Anna L David2,3, and Andrew Melbourne1 | ||
1School of Biomedical Engineering and Imaging Sciences (BMEIS), Kings College London, London, United Kingdom, 2Elizabeth Garret Anderson Institute for Women’s Health, University College London, London, United Kingdom, 3University College London Hospital, London, United Kingdom, 4Department of Medical Physics and Biomedical Engineering, University College London Hospitals, London, United Kingdom, 5Great Ormond Street Hospital for Children, London, United Kingdom, 6Sheffield Teaching Hospitals NHS Foundation Trust, Sheffield, United Kingdom, 7Centre for Medical Imaging, University College London, London, United Kingdom |
||
Keywords: Data Processing, Placenta Feature Selection Models provide a ranking of pathological MRI markers able to predict the outcome of Placenta Accreta Spectrum Disorder, which could be used to aid in clinical decision-making and improve maternal outcome. The potential being to reduce the workload of radiologists by establishing the most clinically relevant pathological MRI markers that predict outcome. Our results found three pathological markers to have the highest ranking to the outcomes with an average accuracy of 75% using a Random Forest Selection Model and Boruta algorithm. |
| 3233 | Computer 1
|
Conversion of legacy data of nonclinical trials using MRI into CDISC SEND: Toward Standardization of an Animal MR Imaging Data |
| Do-Wan Lee1, Young Chul Cho2, Mi-hyun Kim3,4, Yeon Ji Chae5, Su Jung Ham3, Yousun Ko1, Seongwon Na2, Youngbin Shin2, Nari Kim6, Dong‐Cheol Woo5,7, and Kyung Won Kim2 | ||
1Department of Radiology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea, Republic of, 2Biomedical Research Center, Asan Institute for Life Sciences, Asan Medical Center, Seoul, Korea, Republic of, 3Trialinformatics Inc., Seoul, Korea, Republic of, 4Department of Radiation Science & Technology, Jeonbuk National University, Jeonju, Korea, Republic of, 5Department of Convergence Medicine, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea, Republic of, 6Department of Medical Science, Asan Medical Institute of Convergence Science and Technology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea, Republic of, 7Convergence Medicine Research Center, Asan Institute for Life Sciences, Asan Medical Center, Seoul, Korea, Republic of |
||
Keywords: Visualization, Visualization In nonclinical animal research for new drug development, the US FDA mandates to manage data in accordance with clinical data interchange standards consortium (CDISC) standard for exchange of nonclinical data (SEND). Older nonclinical trial data, i.e., legacy nonclinical trial data, contain highly useful information but those data were created in various data formats by different researchers. The present study specifically describes how to implement the CDISC SEND data standards for the legacy nonclinical trial data with MRI. Standardization of animal MRI data with CDISC SEND enables us to utilize legacy data for next-generation drug development and new knowledge creation. |
| 3234 | Computer 2
|
IMPROVED EX-VIVO IMAGING USING BRAIN & CONTAINER-SPECIFIC 3D-PRINTED HOLDERS |
| Sethu K. Boopathy Jegathambal1, Ilana Ruth Leppert2, David A. Rudko2, and Amir Shmuel2 | ||
1Department of Biomedical Engineering, McGill University, Montreal, QC, Canada, 2McConnell Brain Imaging Center, Montreal Neurological Institute and Hospital, Montreal, QC, Canada |
||
Keywords: Data Acquisition, Ex-Vivo Applications, Brain Ex-Vivo MRI Depending on the medium, it might be difficult to maintain the brain stable during ex-vivo imaging. In addition, the SNR might not be homogeneous. We propose using brain and container-specific 3D-printed models for holding the brain. The method requires preliminary imaging of the container and the brain. Then, a 3D-printed model is created, with profiles that match the surface of the brain on one side and the inner surface of the container on the other. The holder is designed to maintain the brain stable in space, where the SNR is high and homogeneous, while allowing contact with the immersion fluid. |
| 3235 | Computer 3
|
Association of MR image quality measures with diagnostic accuracy and inter-reader agreement of PI-RADS for detection of prostate cancer |
| Pranav Sompalle1, Ansh Roge1, Michael Sobota1, Amogh Hiremath2, Sree Harsha Tirumani3, Leonardo Kayat Bittencourt3, Ryan Ward4, Justin Ream4, Andrei Purysko4, Anant Madabhushi5, Satish Viswanath1, and Rakesh Shiradkar5 | ||
1Case Western Reserve University, Cleveland, OH, United States, 2Biomedical Engineering, Case Western Reserve University, Cleveland, OH, United States, 3University Hospitals Cleveland Medical Center, Cleveland, OH, United States, 4Cleveland Clinic Foundation, Cleveland, OH, United States, 5Emory University, Atlanta, GA, United States |
||
Keywords: Data Processing, Data Processing, Image processing Radiologists’ ability to detect clinically-significant prostate cancer on MRI is affected by image quality. While subjective guidelines are provided by the prostate imaging reporting and data system (PI-RADS) and other recent methods, there is a need to develop objective and quantitative metrics of MRI quality. In this multi-reader study, we derive MRI quality metrics using an image processing-based open-source software tool and evaluate their association with radiologist-assigned PI-RADS scores to detect clinically-significant prostate cancer (csPCa). We observe that higher quality MRI is associated with improved detection of csPCa on MRI. |
| 3236 | Computer 4
|
Dual-site and dual-vendor comparison of cerebral oxygen extraction fraction by ASE and TRUST MRI in identical participants |
| Chunwei Ying1, Spencer Waddle2, Nkemdilim Igwe3, Niral J. Patel4, Alexander K. Song2, Charu Balamurugan4, Lori C. Jordan4, Dengrong Jiang5, Hanzhang Lu5, Andria L. Ford3, Manus J. Donahue2, and Hongyu An1,3 | ||
1Mallinckrodt Institute of Radiology, Washington University School of Medicine, St Louis, MO, United States, 2Department of Neurology, Vanderbilt University Medical Center, Nashville, TN, United States, 3Department of Neurology, Washington University School of Medicine, St Louis, MO, United States, 4Department of Pediatrics, Vanderbilt University Medical Center, Nashville, TN, United States, 5Department of Radiology, Johns Hopkins University School of Medicine, Baltimore, MD, United States |
||
Keywords: Data Analysis, Brain, repeatability; reproducibility We evaluated dual-site and dual-vendor test-retest repeatability and reproducibility of cerebral oxygen extraction fraction (OEF) measured with asymmetric spin echo (ASE) and T2-Relaxation-Under-Spin-Tagging (TRUST) MRI. The intra-site intra-class correlation coefficient (ICC) of global ASE-OEF were 0.952 and 0.919 for each site. The intra-site ICC of TRUST-OEF based on the bovine and the HbA model were 0.810 and 0.792 at site 1, and 0.928 and 0.924 at site 2. The inter-site inter-vendor correlation was 0.849 for ASE-OEF and 0.814 for TRUST-OEF based on the bovine model. ASE-OEF was significantly associated with TRUST-OEF (p<0.001). |
| 3237 | Computer 5
|
A Novel Deep Learning Tissue Normalization Method for Longitudinal Analysis of T2-Weighted MRI following Prostate Cancer Radiation Treatment |
| Ahmad Algohary1, Evangelia I. Zacharaki1, Mohammad Alhusseini1, Adrian Breto1, Veronica Wallaengen1, Isaac Xu1, Sandra Gaston1, Patricia Castillo1, Sanoj Punnen1, Benjamin Spieler1, Matthew Abramowitz1, Alan Dal Pra1, Oleksandr Kryvenko1, Alan Pollack1, and Radka Stoyanova1 | ||
1The University of Miami, Miami, FL, United States |
||
Keywords: Machine Learning/Artificial Intelligence, Prostate In this work, we introduce a novel automated triple-reference intensity normalization method for T2W images with the aim of obtaining consistent longitudinal measurements leading to improved quantitative assessment of radiation treatment outcome for prostate cancer patients. |
| 3238 | Computer 6
|
The Use of Quantitative Metrics and Machine Learning to Predict Radiologist Interpretations of Image Quality and Artifacts |
| Lucas McCullum1, John Wood2, Maria Gule-Monroe3, Ho-Ling Anthony Liu4, Melissa Chen3, Komal Shah3, Noah Nathan Chasen3, Vinodh Kumar3, Ping Hou4, Jason Stafford4, Caroline Chung5, Moiz Ahmad4, Christopher Walker4, and Joshua Yung4 | ||
1Medical Physics, MD Anderson Cancer Center, Houston, TX, United States, 2Enterprise Development and Integration, MD Anderson Cancer Center, Houston, TX, United States, 3Neuroradiology, MD Anderson Cancer Center, Houston, TX, United States, 4Imaging Physics, MD Anderson Cancer Center, Houston, TX, United States, 5Data Governance & Provenance, MD Anderson Cancer Center, Houston, TX, United States |
||
Keywords: Machine Learning/Artificial Intelligence, Data Analysis, Image Quality A dataset of 3D-GRE and 3D-TSE brain 3T post contrast T1-weighted images as part of a quality improvement project were collected and shown to five neuro-radiologists who evaluated each sequence for image quality and artifacts. The same scans were processed using the MRQy tool for objective, quantitative image quality metrics. Using the combined radiologist and quantitative metrics dataset, a decision tree classifier with a bagging ensemble approach was trained to predict radiologist assessment using the quantitative metrics. The resulting AUCs for each classification task were above 0.7 for all combinations of sufficiently represented classes and qualitative image metrics. |
| 3239 | Computer 7
|
Guided field Map Estimation and Image Correction for MRI with LargeMagnetic Field Inhomogeneity |
| Navid Reyhanian1, Michael Mullen2, Taylor Froelich2, Michael Garwood2, and Jarvis Haupt1 | ||
1Department of Electrical and Computer Engineering, University of Minnesota, Minneapolis, MN, United States, 2Center for Magnetic Resonance Research, University of Minnesota, Minneapolis, MN, United States |
||
Keywords: Data Analysis, Magnets (B0), optimization; static field inhomogeneity; artifact correction Many methods deployed for Magnetic Resonance Imaging (MRI) suffer from degraded image quality when the static magnetic field ($$$B_0$$$) is nonuniform, as $$$B_0$$$ inhomogeneity causes different types of image artifacts and distortions. Image correction methods require a prior estimate of the field map. In this work, we study the problem of static $$$B_0$$$ inhomogeneity estimation from distorted and undistorted image pairs, despite undistorted images being noisy. The estimated field map is later utilized to correct distorted images. The efficacy of the proposed approach is demonstrated for different encoding methods in the presence of large magnetic field inhomogeneity via extensive simulations. |
| 3240 | Computer 8
|
Semi-automated analysis to reduce operator variability in oxygen extraction fraction using susceptometry-based oximetry |
| Bo-Jhen Su1, Pei-Hsin Wu1, Rajiv S Deshpande2, and Felix W Wehrli2 | ||
1Department of Electrical Engineering, National Sun Yat-sen University, Kaohsiung, Taiwan, 2Department of Radiology, University of Pennsylvania, Philadelphia, PA, United States |
||
Keywords: Data Analysis, Oxygenation SvO2 and OEF were estimated via susceptometry-based oximetry (SBO). The impact of user-defined ROI was investigated in this study. The results show that (1) the consistency of SvO2 estimation depends on operators’ comprehension of MR-oximetry, (2) SvO2 might be under-estimated with manual ROI selection, and (3) the operator-induced bias can be eliminated by the semi-automatic analysis. The preliminary data indicate the feasibility of the constructed semi-automatic analysis for consistent estimation of SvO2 among operators and its potential for bulk analysis and clinical use. |
| 3241 | Computer 9
|
A Comprehensive Quality Assessment on Fast MRI Acquisition with united Compressed Sensing Techniques |
| Zhuoyang Gu1, Lianghu Guo1, Qing Yang1, Xinyi Cai1, Tianli Tao1, Sifan He1, Qian Wang1, He Qiang2, Dinggang Shen1, and Han Zhang1 | ||
1School of Biomedical Engineering, ShanghaiTech University, Shanghai, China, 2United Imaging Healthcare Co., Ltd., Shanghai, China |
||
Keywords: Parallel Imaging, Data Acquisition, Quality Assessment MRI acquisition from certain populations is difficult. For instance, infant MRI without sedation during wakefulness has a high failure rate due to head motion. Thus, a shorter scan time is required. Increasing studies focused on fast MRI acquisition but few studies examined their qualities. We argue that only calculating common image quality metrics (e.g., SNR) may not meet the satisfaction of brain MRI analysis with fast acquisition techniques. A whole-process quality assessment method is thus proposed comparing different fast acquisition techniques with practical guidelines. In future, comprehensive quality assessment on fast MRI acquisitions can become a study routine. |
| 3242 | Computer 10
|
Investigating the computational reproducibility of Neurodesk |
| Thuy Thanh Dao1, Angela Renton1,2, Aswin Narayanan3, Markus Barth1, and Steffen Bollmann1 | ||
1School of Information Technology and Electrical Engineering, The University of Queensland, Brisbane, Australia, 2Queensland Brain Institute, The University of Queensland, Brisbane, Australia, 3Australian National Imaging Facility, The University of Queensland, Brisbane, Australia |
||
Keywords: Software Tools, Data Processing Glatard et al [1] found that the operating system version affected the computational results of neuroimaging analysis pipelines. Their work demonstrated that this reproducibility issue is caused by the accumulated floating-point differences across operating systems. Neurodesk (www.neurodesk.org) approaches this problem by packaging every pipeline into a software container and thereby can control the underlying dependencies for each application regardless of the operating system. In this work, a brain tissue segmentation pipeline from the FMRIB Software Library (FSL) was compared with a containerised version from Neurodesk. The comparison revealed that Neurodesk provides computational reproducibility compared with local installations. |
| 3243 | Computer 11
|
ComBat scanner harmonization for fixel-based analysis |
| Remika Mito1,2, Heath Pardoe1, Robert Smith1, Jacques-Donald Tournier3,4, David Vaughan1,2,5, Mangor Pedersen6, and Graeme Jackson1,2,5 | ||
1Florey Institute of Neuroscience and Mental Health, Melbourne, Australia, 2Florey Department of Neuroscience and Mental Health, University of Melbourne, Melbourne, Australia, 3Biomedical Engineering Department, School of Biomedical Engineering and Imaging Sciences, King's College London, London, United Kingdom, 4Centre for the Developing Brain, School of Biomedical Engineering and Imaging Sciences, King's College London, London, United Kingdom, 5Department of Neurology, Austin Health, Melbourne, Australia, 6Auckland University of Technology, Auckland, New Zealand |
||
Keywords: Data Analysis, Diffusion/other diffusion imaging techniques, Data harmonization Diffusion MRI data are known to be sensitive to scanner and site effects, for which harmonization methods have been developed. However, harmonization methods have not yet been applied to many advanced diffusion imaging metrics. In this work, we apply the ComBat harmonization approach to fixel-based analysis. ComBat successfully minimizes scanner-related differences in fibre density and cross-section, and performs similarly to the inclusion of scanner as a nuisance regressor during whole-brain fixel-based analysis. Importantly, ComBat can now readily be used within the fixel-based framework, which will enable large multi-centre studies to implement this approach in the future. |
| 3244 | Computer 12
|
Reliability of brain volume measures of accelerated 3D T1-weighted images with deep learning-based reconstruction |
| Woojin Jung1, Seongjae Mun1, Jingyu Ko1, and Koung Mi Kang2 | ||
1AIRS Medical, Seoul, Korea, Republic of, 2Department of Radiology, Seoul National University Hospital, Seoul, Korea, Republic of |
||
Keywords: Machine Learning/Artificial Intelligence, Brain We validated the brain volumetric results with 3D T1 weighted images with accelerated scans reconstructed by FDA-cleared deep learning-based software (SwiftMR, AIRS Medical). Acceleration scans with three different acceleration levels were simulated using k-space undersampling, and the image quality and brain volume measures were evaluated. In addition, we acquired conventional and accelerated scans from each participant to evaluate the reliability between conventional and accelerated scans reconstructed by SwiftMR and inter-method reliability between different brain segmentation software. As a result, brain volume measures with accelerated scans with deep learning-based reconstruction were in good agreement with those of the corresponding conventional scan. |
| 3245 | Computer 13
|
Harmonization of the multisite ALPS-index based on DTI-ALPS using the combined association test |
| Yuya Saito1, Koji Kamagata1, Toshiaki Taoka2, Christina Andica1, Wataru Uchida1, Kaito Takabayashi1, Mana Owaki1,3, Seina Yoshida1,3, Keigo Yamazaki1,3, Shohei Fujita1,4, Akifumi Hagiwara1, Junko Kikuta1, Toshiaki Akashi1, Akihiko Wada1, Keigo Shimoji1, Masaaki Hori5, Shinji Naganawa6, and Shigeki Aoki1 | ||
1Department of Radiology, Juntendo University Graduate School of Medicine, Tokyo, Japan, 2Department of Innovative Biomedical Visualization (iBMV), Nagoya University Graduate School of Medicine, Aichi, Japan, 3Department of Radiological Sciences, Graduate School of Human Health Sciences, Tokyo Metropolitan University, Tokyo, Japan, 4Department of Radiology, University of Tokyo, Tokyo, Japan, 5Department of Radiology, Toho University Omori Medical Center, Tokyo, Japan, 6Department of Radiology, Nagoya University Graduate School of Medicine, Aichi, Japan |
||
Keywords: Data Processing, Data Processing, Diffusion MRI, ALPS-index, DTI-ALPS, Harmonization, multisite study Diffusion tensor image analysis along the perivascular space (DTI-ALPS) is a promising noninvasive method for indirectly evaluating the glymphatic system. However, the ALPS-index calculated from diffusion MRI data collected at multiple sites should be harmonized to avoid site-related effects. We applied the combined association test (ComBat), which uses regression of covariates with empirical Bayes, for harmonizing the ALPS-index. ComBat mitigated site-related effects, increased statistical power to differentiate Alzheimer’s disease and cognitive normal, and improved the correlation between the ALPS-index and cognitive function. Thus, ComBat harmonization can be applied to evaluate the glymphatic system using the ALPS-index in large multisite studies. |
| 3246 | Computer 14
|
Reliable and repeatable fully automated bolus arrival times and arterial inputs using 2 Hz whole brain multiband DSC-MRI at 7T |
| Daniel L Schwartz1, Xin Li2, Fred Lesage3, Andreas Linninger4, Mohammad Jamshidi4, Becky Stone5, and William Rooney5 | ||
1Advanced Imaging Research Center/Neurology, OHSU, PORTLAND, OR, United States, 2Advanced Imaging Research Center, OHSU, PORTLAND, OR, United States, 3Electrical Engineering, Polytechnique Montreal, Montreal, QC, Canada, 4Bioengineering, UIC, Chicago, IL, United States, 5Advanced Imaging Research Center, OHSU, Portland, OR, United States |
||
Keywords: Data Processing, DSC & DCE Perfusion, Whole-brain bolus arrival times Dynamic susceptibility contrast imaging has used bolus arrival time (BAT) in tissue, relative to an arterial input function (AIF), as a nuisance variable when deriving perfusion metrics (e.g. blood flow). However, BAT reflects physiological changes in both fluid dynamics of feeding vasculature and capillary inflow. This work presents a reproducible approach for reliable and repeatable quantification of an AIF and whole brain BAT, which itself may be germane to chronic and acute cerebrovascular disease (CVD), such as white matter hyperintensities, age-associated CVD, and stroke. |
| 3247 | Computer 15
|
Automated brain QSM computation pipeline deployed in the European Open Science Cloud |
| Miguel Guevara1, Davy Cam2, Jacques Badagbon2, Stephane Roche2, Michel Bottlaender1, Yann Cointepas1, Jean-François Mangin1, Ludovic de Rochefort2, and Alexandre Vignaud1 | ||
1Neurospin, Paris, France, 2Ventio, Marseille, France |
||
Keywords: Data Processing, Quantitative Susceptibility mapping, Cloud-computing Introduction of a pipeline for computing the Quantitative Susceptibility Mapping (QSM) from non-optimal MRI data in a secure cloud-infrastructure. The pipeline allows the computation of clean and exploitable QSM, thanks to a phase pre-processing to reduce the presence of artifacts in the 7T MRI data. |
| 3248 | Computer 16
|
dbdicom: an open-source python interface for reading and writing DICOM databases |
| Steven Sourbron1, Joao Almeida e Sousa1, Alexander Daniel2, Charlotte Buchanan2, Ebony Gunwhy1, Eve Lennie1, Kevin Teh1, Steve Shillitoe1, David Morris3, Andrew Priest4, David Thomas5, and Susan Francis2 | ||
1University of Sheffield, Sheffield, United Kingdom, 2University of Nottingham, Nottingham, United Kingdom, 3University of Edinburgh, Edinburgh, Scotland, 4University of Cambridge, Cambridge, United Kingdom, 5University College London, London, United Kingdom |
||
Keywords: Software Tools, Data Processing DICOM is the universally recognized standard for medical imaging, but reading and writing from DICOM databases remains a challenging task for most data scientists. The dbdicom package was developed to provide an intuitive programming interface for reading and writing data from entire DICOM databases, and replaces confusing DICOM-native concepts by language and notations that will be more familiar to data scientists working in python. The package is available under an open license but is in the early stages of development and currently rolled out in 3 multi-vendor, multi-center studies. |
| 3249 | Computer 17
|
Improved Image Quality with Deep Learning-Based Image Reconstruction for Multi-shot Diffusion-Weighted Imaging of the Prostate |
| Patricia S. Lan1, Xinzeng Wang2, Alessandro Scotti3, Praveen Jayapal4, Pingni Wang1, Arnaud Guidon5, and Andreas M. Loening4 | ||
1GE Healthcare, Menlo Park, CA, United States, 2GE Healthcare, Houston, TX, United States, 3GE Healthcare, Columbus, OH, United States, 4Department of Radiology, Stanford University, Stanford, CA, United States, 5GE Healthcare, Boston, MA, United States |
||
Keywords: Diffusion/other diffusion imaging techniques, Diffusion/other diffusion imaging techniques Diffusion-weighted imaging (DWI) is a key component of identifying prostate tumors on MRI. Multi-shot DWI techniques (e.g. MUSE) have been shown to enable high resolution prostate DWI and, compared to single-shot DWI, reduce distortion artifacts due to rectal gas and hip implants at the expense of increased scan time. In this study, we evaluated a CNN-based deep learning (DL) image reconstruction method for MUSE. Our results indicate that for high b-value images the DL-based reconstruction improved perceived image quality even with half the original NEX, suggesting potential scan time reduction using DL. |
| 3250 | Computer 18
|
Repeatability of very low field MRI |
| Pavan Poojar1 and Sairam Geethanath1 | ||
1Accessible Magnetic Resonance Laboratory, Biomedical Imaging and Engineering Institute, Department of Diagnostic, Molecular and Interventional Radiology, Icahn School of Medicine at Mount Sinai, New York, NY, United States, New York, NY, United States |
||
Keywords: Data Acquisition, Low-Field MRI Low-field MR scanners are low-cost, portable, and requires lesser space. In this study, we performed in vitro phantom repeatability study on a 48 mT scanner (Multiwave technologies) by scanning the phantom 3 sessions in a day (11 AM, 2 PM, and 5 PM) at 3 locations. During each scan, we measured the temperature, humidity, and F0 to see their effect on the image. We scanned 9 sequences in the study to monitor noise, F0 and B0. The difference in temperature and humidity was determined for each session.The line intensity profilesrevealed that VLF scanner produces consistent B0 maps and qualitative images. |
| 3251 | Computer 19
|
MAVRIC SL can reduce inter-rater variability of the measurements for the cervical spine segment height in ACDF patients on 3T MRI |
| Renjie Yang1, Changsheng Liu1, Weiyin Vivian Liu2, Liang Li1, and Yunfei Zha1 | ||
1Renmin Hospital of Wuhan University, Wuhan, China, 2GE Healthcare, Beijing, China |
||
Keywords: Artifacts, Surgery Subsidence assessment is an important indicator for postoperative follow-ups for ACDF patients and normally performed on radiographs. It is still challenging to image ACDF patients using conventional MRI for the measure of segment height is influenced by the metal artifacts. In this study, MAVRIC SL showed better inter-rater agreement on measurements of segment height compared to FSE T1WI. 20 cases on MAVRIC SL and 6 cases on FSE T1W1 among 22 patients showed the measurements of segment height was less than the maximum allowed difference (2mm). This study demonstrated that MAVRIC SL has a strong potential in the subsidence assessment. |
| 3252 | Computer 20
|
Automated Quality Assessment of Liver Magnetic Resonance Images with Fully Automatic Segmentation and Radiomics Approach |
| Hai Nan Ren1, Li Jun Qian1, Xu Hua Gong1, Yan Zhou1, and Yang Song2 | ||
1Radiology, Renji Hospital, Shanghai Jiao Tong University, School of Medicine, Shanghai, China, 2MR Scientific Marketing, Siemens Healthineers Ltd., Shanghai, China |
||
Keywords: Machine Learning/Artificial Intelligence, Machine Learning/Artificial Intelligence, Quality control This study investigated the feasibility and performance of quality assessment of hepatic magnetic resonance (MR) images using a deep-learning-based segmentation and radiomics approach. We used a pre-trained deep learning model to segment the liver on different contrast-enhanced MR phases and then extracted quantitative features to assess the image quality by a machine learning method. The results showed that the radiomics model had a high performance for image quality identification in both training and test sets. This suggests that it was feasible to automate the identification of image quality by using radiomics approaches. |
| 3409 | Computer 1
|
Slice-by-slice Dynamic Shimming Based on a Chemical Shift-Encoded Acquisition to Improve Fat Suppression in DWI |
| Aidan Tollefson1,2, Gaohong Wu3, Patricia Lan4, Arnaud Guidon5, Rianne A Van Der Heijden2, Daiki Tamada2, Ali Pirasteh1,2, and Diego Hernando1,2 | ||
1Medical Physics, University of Wisconsin-Madison, Madison, WI, United States, 2Radiology, University of Wisconsin-Madison, Madison, WI, United States, 3GE Healthcare, Waukesha, WI, United States, 4GE Healthcare, Menlo Park, CA, United States, 5GE Healthcare, Boston, MA, United States |
||
Keywords: Artifacts, Fat B0 inhomogeneities can lead to failures in chemical shift-based fat suppression, particularly in complex susceptibility environments. Conventional volumetric shimming methods are frequently unable to compensate for these B0 inhomogeneities, which leads to residual fat signal and artifacts in various applications. We propose a slice-by-slice shimming method that relies on information (water-only image, fat-only image, B0 fieldmap) derived from a rapid chemical shift-encoded acquisition. This method demonstrated improved fat suppression in diffusion weighted imaging (DWI) of the upper leg and may lead to improved reliability in other applications. |
| 3410 | Computer 2
|
Investigating the Impact of Motion and Associated B0 Changes on Oxygenation Sensitive MRI through Realistic Simulations |
| Hannah Eichhorn1,2, Kerstin Hammernik3,4, Samira M. Epp5,6, Dimitrios C. Karampinos7, Julia A. Schnabel1,2,8, and Christine Preibisch9 | ||
1Institute of Machine Learning in Biomedical Imaging, Helmholtz Center Munich, Munich, Germany, 2TUM School of Computation, Information and Technology; and TUM Institute for Advanced Study, Technical University of Munich, Munich, Germany, 3Lab for Artificial Intelligence in Healthcare and Medicine, Technical University of Munich, Munich, Germany, 4Department of Computing, Imperial College London, London, United Kingdom, 5Department of Neuroradiology, Neuroimaging Center, Technical University of Munich, Munich, Germany, 6Graduate School of Systemic Neurosciences, Ludwig-Maximilians-University, Munich, Germany, 7Department of Diagnostic and Interventional Radiology, School of Medicine, Technical University of Munich, Munich, Germany, 8School of Biomedical Imaging and Imaging Sciences, King’s College London, London, United Kingdom, 9Department of Neuroradiology, School of Medicine, Technical University of Munich, Munich, Germany |
||
Keywords: Artifacts, Oxygenation, Motion Simulation T2*-weighted gradient echo imaging is strongly impacted by subject head motion, amongst others due to motion-related changes in B0 inhomogeneities. Within the oxygenation-sensitive mqBOLD protocol, motion artifacts in T2*-weighted images lead to errors in derived parameter maps. To quantify these errors, we performed realistic motion simulations incorporating rigid body transformations and motion-related field inhomogeneity changes. Our results demonstrate the importance of including B0 inhomogeneities for realistic motion artifact patterns. Even small amounts of simulated motion resulted in substantial errors in derived T2* and R2’ parameter maps, which highlights the relevance of T2* motion correction within the mqBOLD technique. |
| 3411 | Computer 3
|
Independent Component Analysis for Noise Removal in MR Fingerprinting |
| Emma L Thomson1,2, Claudia A M Gandini Wheeler-Kingshott3,4,5, and Geoff J M Parker1,6,7 | ||
1Centre for Medical Image Computing, Department of Medical Physics and Biomedical Engineering, University College London, London, United Kingdom, 2UCL Queen Square Institute of Neurology, Faculty of Brain Sciences, London, United Kingdom, 3NMR Research Unit, Queen Square MS Centre, Department of Neuroinflammation, London, United Kingdom, 4Department of Brain & Behavioural Sciences, University of Pavia, Pavia, Italy, 5Brain Connectivity Centre Research Department, IRCCS Mondino Foundation, Pavia, Italy, 6NMR Research Unit, Queen Square MS Centre, Department of Neuroinflammation, UCL Queen Square Institute of Neurology, Faculty of Brain Sciences, London, United Kingdom, 7Bioxydyn Limited, Manchester, United Kingdom |
||
Keywords: Artifacts, MR Fingerprinting We propose the use of independent component analysis for the removal of coherent noise sources prior to matching for magnetic resonance fingerprinting (MRF). We tested this technique for the removal of reconstruction artefacts on images acquired with a spiral k-space acquisition to quantify intravascular T1, extravascular T1, B1+, cerebral blood volume (νb) and inter-vascular water exchange (1/τb). We demonstrate that removal of coherent noise sources in this way can improve the precision of measurements of parameters. |
| 3412 | Computer 4
|
Signal gain by reduction of the concomitant phase in double diffusion encoding by means of added oscillating gradients |
| Julian Rauch1,2, Frederik B. Laun3, Dominik Ludwig1, Maxim Zaitsev4, Mark E. Ladd1,2,5, Peter Bachert1,2, and Tristan A. Kuder1,2 | ||
1Division of Medical Physics in Radiology, German Cancer Research Center (DKFZ), Heidelberg, Germany, 2Faculty of Physics and Astronomy, Heidelberg University, Heidelberg, Germany, 3University Hospital Erlangen, Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU), Erlangen, Germany, 4Medical Physics, Department of Radiology, Faculty of Medicine, Medical Center University of Freiburg, Freiburg, Germany, 5Faculty of Medicine, Heidelberg University, Heidelberg, Germany |
||
Keywords: Artifacts, Artifacts Intravoxel dephasing generated by Maxwell or concomitant fields can cause image artifacts like signal voids or falsify the quantification. In this study, an optimization scheme to reduce concomitant field effects in diffusion sequences with single pairs of bipolar gradients on each axis is presented. Oscillating gradients are added onto the original gradient pulses with the aim of reducing the concomitant phase without significant changes in the sequence properties. The proposed method is evaluated in both measurements and simulations, and gives rise to a positive effect on the signal for arbitrary diffusion wave vector pairs. |
| 3413 | Computer 5
|
Optimised navigator correction of physiological field fluctuations in multi-echo GRE of the lumbar spinal cord at 3T |
| Laura Beghini1, Gergely David2,3, Martina D. Liechti3, Silvan Büeler3, and S. Johanna Vannesjo1 | ||
1Department of Physics, Norwegian University of Science and Technology, Trondheim, Norway, 2Spinal Cord Injury Center, Balgrist University Hospital, University of Zurich, Zurich, Switzerland, 3Dept. of Neuro-Urology, Balgrist University Hospital, University of Zurich, Zurich, Switzerland |
||
Keywords: Artifacts, Spinal Cord Multi-echo GRE sequences acquired in the spinal cord are sensitive to breathing-related field variations, leading to ghosting artifacts. While navigator-based corrections are often used in the brain to account for field variations, standard implementations often fail in the spinal cord due to the close proximity to the lungs. Here, we investigated optimised navigator corrections including a fast Fourier transform combined with region selection encompassing the spinal canal. Besides a visual improvement of image quality, the optimised navigator correction yielded statistically significant improvements in the CNR between WM and CSF and the SNR in the WM. |
| 3414 | Computer 6
|
RF Interference Manifested Zipper Detection And Removal Using Channel Compression |
| Megha Goel1, Preetham Shankpal1, Suresh Emmanuel Joel1, Rajagopalan Sundareshan1, and Harsh Agarwal1 | ||
1GE Healthcare, Bengaluru, India |
||
Keywords: Artifacts, Image Reconstruction, Zipper-removal Zipper artifacts are commonly seen in MR images due to spurious radio-frequency signals or improper RF-shielding. Zipper-riddled images lose diagnostic value and are usually sent for rescan. Here, we attempt to mitigate zipper artifacts in the post-processing pipeline after scan has been acquired. We do this after channel combination technique has limited zipper appearance to 1-2 pseudo-channels, which we detect and remove from channel-combination process. We evaluated this on various brain contrasts and confirm reduction of zipper presence visually in the images. Given that zippers manifest as bright/dark discontinuous lines irrespective of the anatomy/contrast, the method should be generalizable. |
| 3415 | Computer 7
|
Denoising of 7T MP2RAGE MRI with Preserving the Brain Signal |
| Kwan-Jin Jung1 | ||
1Biomedical Imaging Center, Beckman Institute, University of Illinois at Urbana-Champaign, Urbana, IL, United States |
||
Keywords: Artifacts, High-Field MRI, MP2RAGE, 7T MP2RAGE requires denoising of the synthesized T1-weighted image. The conventional denoising method is to add a constant in the synthesis that alters the brain signal. This affects brain segmentation. A new denoising method has been developed to denoise it with preserving the brain signal. The new method estimates the bias field from two inversion images and sets the noise floors at the low and high noise levels in proportion to the bias field. The denoise weight is calculated at each voxel using the two noise floors. The proposed denoising method worked reliably with preserving the brain signal. |
| 3416 | Computer 8
|
Effect of chemical shift displacement on Lipid Composition determination in localized 1H NMR Spectroscopy |
| Julian Mevenkamp1,2, Mijke Buitinga1,2, Pandichelvam Veeraiah3,4, and Vera B. Schrauwen-Hinderling1,2,5 | ||
1Nutrition & Movement Sciences, Maastricht University, Maastricht, Netherlands, 2Radiology & Nuclear Medicine, Maastricht University Medical Center, Maastricht, Netherlands, 3Scannexus, Maastricht, Netherlands, 4Faculty of Health Medicine & Life Sciences, Maastricht University, Maastricht, Netherlands, 5German Diabetes Center, Düsseldorf, Germany |
||
Keywords: Artifacts, Spectroscopy, Chemical Shift Displacement We showed that implementing prior knowledge about Chemical Shift Displacement Artifacts (CSDA) in localized MR-spectroscopy improves fit quality in the context of hepatic lipid composition measurements (LICO) performed with PRESS. Residuals of fitting PRESS spectra of peanut oil with CSDA prior knowledge show smaller residuals compared to assuming a purely Spin Echo based J-coupling evolution during the echo time. Furthermore, peanut oil LICO measured by PRESS and fitted with CSDA prior knowledge almost matched reference values determined by HR-NMR (PUFAHR-NMR = 23.5%, MUFAHR-NMR = 52.9%, SFAHR-NMR = 23.6% vs. PUFACSDA = 21.57%, MUFACSDA = 54.4%, SFACSDA = 24.0%). |
| 3417 | Computer 9
|
Isotropic sub-mm bSSFP imaging near metal at 0.55T |
| Kübra Keskin1, Nam G. Lee2, Jay Acharya3, and Krishna S. Nayak1,2 | ||
1Electrical and Computer Engineering, University of Southern California, Los Angeles, CA, United States, 2Biomedical Engineering, University of Southern California, Los Angeles, CA, United States, 3Radiology, University of Southern California, Los Angeles, CA, United States |
||
Keywords: Artifacts, Artifacts, Metal At conventional field strengths, the large magnetic susceptibility difference between metallic implants and surrounding tissues causes severe artifacts often requiring multi-spectral imaging. Recent low-field MRI systems offers new opportunities for imaging adjacent to metallic implants, including balanced SSFP. Here, we utilize phase-cycled bSTAR, a short-TR 3D dual-echo radial bSSFP sequence, for isotropic 3D banding artifact free imaging near metallic implants at 0.55T. We demonstrate diagnostic quality images in volunteers with wrist and spinal hardware with high SNR efficiency compared to TSE-based sequences. |
| 3418 | Computer 10
|
Demonstration of motion-dependent magnetic field inhomogeneity in the kidneys and its retrospective correction for renal DWI |
| Nima Gilani1, Artem Mikheev1, Inge Manuela Brinkmann2, Dibash Basukala1, Thomas Benkert3, Malika Kumbella1, James S. Babb1, Hersh Chandarana1, and Eric E. Sigmund1 | ||
1Department of Radiology, NYU Langone Health, New York, NY, United States, 2Siemens Medical Solutions USA Inc., New York, NY, United States, 3Siemens Healthcare GmbH, Erlangen, Germany |
||
Keywords: Artifacts, Kidney, TOPUP, Field Inhomogeneity Echo planar imaging is highly affected by field map inhomogeneity distortion artifact. Field map inhomogeneity has shown to be motion dependent in the kidneys. In the present work, we propose an alternative method for correction of magnetic field inhomogeneity for renal DWI in respiratory-resolved fashion. Specifically, we collect a series of forward and reverse phase encoded b=0 images to sample kidney motion caused by breathing, map the spatial and respiratory phase dependence of the magnetic field inhomogeneity, and correct each image of free-breathing DWI series according to their respiratory phase. |
| 3419 | Computer 11
|
Investigation of Susceptibility Gradient Artifacts in EPI at 7T |
| Chan Hong Moon1, Hoby Hetherington2,3, and Jullie W. Pan2,4 | ||
1MRRC, Department of Radiology, University of Pittsburgh, Pittsburgh, PA, United States, 2NEXTGEN, Department of Radiology, University of Missouri, Columbia, MO, United States, 3Resonance Research, Inc., Billerica, MA, United States, 4University of Missouri, Columbia, MO, United States |
||
Keywords: Artifacts, High-Field MRI, EPI, 7T, Susceptibility, GRAPPA, Aliasing With the development of high-field 7T scanners, interest in high resolution neuroimaging for EPI-based functional and diffusion MRI has increased. However, EPI artifacts at 7T (spatial distortion and signal loss) increase compared to that seen at 3T due to increased B0 inhomogeneity. Furthermore, EPI artifacts become more significant and complicated when parallel imaging is used. As a result, 7T fMRI and DWI suffers from B0 inhomogeneity-induced aliasing as well as the more common EPI artifacts. In this study, the source of B0 inhomogeneity artifact in parallel imaging was investigated and methods to reduce these artifacts are proposed/tested at 7T. |
| 3420 | Computer 12
|
Spatiotemporal B0 correction in Oscillating Steady State Imaging (OSSI) fMRI using Free Induction Decay Navigators (FIDnavs) |
| Mariama Salifu1, Douglas Noll1, and Melissa Haskell2,3 | ||
1Biomedical Engineering, University of Michigan, Ann Arbor, MI, United States, 2Electrical Engineering, University of Michigan, Ann Arbor, MI, United States, 3Hyperfine Research, Guilford, CT, United States |
||
Keywords: Artifacts, Shims, B0 correction Temporal field variations (e.g., physiological noise) in oscillating steady state imaging (OSSI) not only increases signal fluctuations as seen in GRE, but also alters signal characteristics that lead to inconsistent k-space acquisitions from TR to TR. This effectively reduces the overall functional contrast and temporal signal to noise ratio (tSNR). Here, we utilized free induction decay (FID) signal phase to estimate up to the first-order field inhomogeneity coefficients for prospective spatiotemporal B0 field correction in OSSI. |
| 3421 | Computer 13
|
EPI Nyquist Ghost Correction by Iteratively Enforcing Structural Low Rankness |
| Yilong Liu1, Mengye Lyu2, Yunlin Zhao1, Linfang Xiao3, and Ed X. Wu4,5 | ||
1Guangdong-Hongkong-Macau Institute of CNS Regeneration, Key Laboratory of CNS Regeneration (Ministry of Education), Jinan University, Guangzhou, China, 2College of Health Science and Environmental Engineering, Shenzhen Technology University, Shenzhen, China, 3Hangzhou Weiying Medical Technology Co., Ltd, Hangzhou, China, 4Laboratory of Biomedical Imaging and Signal Processing, The University of Hong Kong, Hong Kong, China, 5Department of Electrical and Electronic Engineering, The University of Hong Kong, Hong Kong, China |
||
Keywords: Artifacts, Artifacts, EPI, Nyquist ghost correction This study presents a Nyquist ghost correction method by iteratively enforcing low rankness of the block-wise Hankel matrix and updating the 1D linear model for phase correction. The proposed method was evaluated with multi-shot EPI data at both ultra-high field (7T) and low field (0.3T). Compared with existing SVD-based method, the proposed method requires much fewer iterations to achieve similar performance, provides a computationally efficient and robust solution to EPI Nyquist ghost correction. |
| 3422 | Computer 14
|
Preclinical spiral imaging with a high-performance gradient insert: a comparison of trajectory correction methods |
| Hannah Scholten1, Tobias Wech1, Sascha Köhler2, and Jürgen E. Schneider3 | ||
1Department of Diagnostic and Interventional Radiology, University Hospital Würzburg, Würzburg, Germany, 2Bruker BioSpin MRI GmbH, Ettlingen, Germany, 3Leeds Institute of Cardiovascular and Metabolic Medicine, University of Leeds, Leeds, United Kingdom |
||
Keywords: Artifacts, System Imperfections: Measurement & Correction, GIRF, GSTF, LTI We compared three different trajectory correction methods in spiral phantom images on a preclinical 7T scanner with a high-performance gradient insert. The first was an isotropic delay-correction, the second a gradient correction based on the gradient system transfer function (GSTF), and the third employed measured trajectories. In both an axial and a double-oblique slice, the measured trajectory resulted in the best image quality, while the delay-corrected images suffered from halo effects and signal bleeding. The GSTF-corrected images only had minor remaining artifacts, stemming from trajectory deviations due to non-linearities in the gradient chain. |
| 3423 | Computer 15
|
Cardiac T1 and T2 Mapping: The Importance of Rest Periods in Quantitative Tissue Property Mapping |
| Seth Garrett, B.S.1, Jesse Hamilton, Ph.D.1, and Nicole Seiberlich, Ph.D.1 | ||
1Michigan Medicine- Department of Radiology, University of Michigan, Ann Arbor, MI, United States |
||
Keywords: Artifacts, Cardiovascular The aim of this study is to assess whether the presence of rest periods in the form of breathhold commands affects T1 and T2 values measured in the myocardium using both clinically standard pulse sequences and cardiac Magnetic Resonance Fingerprinting (MRF). When mapping data are collected immediately following cine acquisitions, T1 maps from both MOLLI and MRF show significant errors, although T2 maps from both T2-prepared bSSFP and MRF are unaffected by the absence of a pause between sequences. |
| 3424 | Computer 16
|
Predicting MR Image Quality from Breathing Patterns |
| Soham S Vasanawala1 and John M Pauly1 | ||
1Electrical Engineering, Stanford University, Stanford, CA, United States |
||
Keywords: Artifacts, Motion Correction, abdomen, respiration Motion, particularly from breathing, compromises the quality of magnetic resonance images. In this work, we hypothesize that detected breathing patterns can be utilized to predict whether adequate MR image quality will be obtained. With a K-means clustering algorithm, 9 in 10 forty-second breathing waveforms were correctly predicted as either resulting in a high or low image quality image; this finding can save time from unnecessary scans. Other models achieved similar results as K-means clusterings. |
| 3425 | Computer 17
|
Clinical evaluation of k-space correlation informed motion artifact detection in segmented multi-slice MRI |
| Ikbeom Jang1,2, Malte Hoffmann1,2, Nalini Singh3,4, Yael Balbastre1, Lina Chen5, Marcio Aloisio Bezerra Cavalcanti Rockenbach5, Adrian Dalca1,2,3, Iman Aganj1,2, Jayashree Kalpathy-Cramer1,2, Bruce Fischl1,2,3,4, and Robert Frost1,2 | ||
1Department of Radiology, Massachusetts General Hospital, Boston, MA, United States, 2Department of Radiology, Harvard Medical School, Boston, MA, United States, 3Computer Science and Artificial Intelligence Laboratory, Massachusetts Institute of Technology, Cambridge, MA, United States, 4Harvard-MIT Health Sciences and Technology, Massachusetts Institute of Technology, Cambridge, MA, United States, 5Data Science Office, Mass General Brigham, Boston, MA, United States |
||
Keywords: Artifacts, Artifacts, Motion artifact, Image Quality, Neuroimaging Motion artifacts can negatively impact diagnosis, patient experience, and radiology workflow especially when a patient recall is required. Detecting motion artifacts while the patient is still in the scanner could potentially improve workflow and reduce costs by enabling immediate corrective action. We demonstrate in a clinical k-space dataset that using cross-correlation between adjacent phase-encoding lines can detect motion artifacts directly from raw k-space in multi-shot multi-slice scans. We train a split-attention residual network to examine the performance in predicting motion artifact severity. The network is trained on simulated data and tested on real clinical data. |
| 3426 | Computer 18
|
Large-scale simulation of rigid head motion artifacts in whole brain MRI data and the impact on automatic cortical segmentation |
| Hampus Olsson1, Jason Michael Millward1,2, Ludger Starke1, Tobias Klein1, Wenli Xu1, Sabrina Klix1, Christoph Lippert3,4, Thoralf Niendorf1,2, and Sonia Waiczies1,2 | ||
1Berlin Ultrahigh Field Facility (B.U.F.F.), Max Delbrück Center for Molecular Medicine in the Helmholtz Association, Berlin, Germany, 2Experimental and Clinical Research Center, a joint cooperation between the Charitë Medical Faculty and the Max Delbrück Center for Molecular Medicine in the Helmholtz Association, Berlin, Germany, 3Digital Health & Machine Learning Research Group, Hasso Plattnet Institut for Digital Engineering, Potsdam, Germany, 4Hasso Plattner Institute for Digital Health at Mount Sinai, Icahn School of Medicine at Mount Sinai, New York, NY, United States |
||
Keywords: Artifacts, Artifacts MRI datasets from epidemiological studies such as the German National Cohort (GNC) are increasingly used in machine learning. These datasets show a lower prevalence of motion artifacts than encountered in clinical practice. We simulated realistic motion artifacts caused by rigid head motion on the GNC MR data (50 subjects) by modifying the TorchIO data augmentation framework. Five levels of artifact severity were simulated. We benchmarked our results to empirical measurements using the standard GNC MPRAGE imaging protocol. The robustness to motion of FreeSurfer and SynthSeg for cortical segmentation was investigated, indicating an improved performance using SynthSeg. |
| 3427 | Computer 19
|
Streak Artifact Reduction using Convolutional Neural Network: Clinical Feasibility Study in Liver MR Imaging |
| Satoshi Funayama1, Shintaro Ichikawa1, Yukichi Tanahashi1, Takanobu Ikeda1, Koh Kubota1, Masaya Kutsuna1, and Satoshi Goshima1 | ||
1Hamamatsu University School of Medicine, Hamamatsu, Japan |
||
Keywords: Artifacts, Machine Learning/Artificial Intelligence, Liver, Motion Correction Radial sampling enables free free-breathing abdominal MR imaging. Meanwhile, while it suffers from streak artifacts. We propose streak artifact reduction using convolutional neural network (SARC) which utilize Hough domain. In Hough domain, a streak becomes like a dot which is more localized compared with image domain. The network was trained in end-to-end manner. The SARC was show better image quality in objective image quality metrics and visual evaluation by a radiologist. SARC showed feasibility for clinical MR imaging. |
| 3428 | Computer 20
|
GDCNet: deep learning model for self-consistent geometric distortion correction of EPI images without a field map |
| Marina Manso Jimeno1,2, John Thomas Vaughan1,2, and Sairam Geethanath2,3 | ||
1Department of Biomedical Engineering, Columbia University, New York, NY, United States, 2Columbia Magnetic Resonance Research Center, Columbia University, New York, NY, United States, 3The Biomedical Engineering Institute (Department of Diagnostic, Molecular and Interventional Radiology), Icahn School of Medicine at Mount Sinai, New York, NY, United States |
||
Keywords: Artifacts, fMRI, Geometrid Distortion, B0 inhomogeneity GDCNet is a “self-consistent” deep learning (DL) model for distortion correction of EPI fMRI images and field map estimation. It only requires the EPI images for correction, saving acquisition time and avoiding motion-related correction errors. The two supervised U-Nets for forward modelling and distortion correction have been tested in silico and in vivo on a publicly-available and a prospectively-acquired dataset. The in silico models demonstrated generalization capabilities and achieved a mean RMSE of 2.56 x10-2 as self-conistency metric. Inference in vivo showed modest correction in the prefrontal cortex and similar estimated field map compared to the acquired ground truth. |
| 3429 | Computer 21
|
Optimal control RF pulses for dual-angle T1 measurements in 31P magnetization transfer spectroscopy |
| Clemens Diwoky1, Christina Graf2, Armin Rund3, and Rudolf Stollberger2 | ||
1Institute of Molecular Biosciences, University of Graz, Graz, Austria, 2Institute of Biomedical Imaging, Graz University of Technology, Graz, Austria, 3Institute for Mathematics and Scientific Computing, University of Graz, Graz, Austria |
||
Keywords: Pulse Sequence Design, CEST & MT Herein, we show for the first time the application of optimal control (OC) to the design of RF pulses for dual-angle T1 measurements as needed for 31P magnetization transfer spectroscopy. Phantom measurements highlight that OC RF pulses are able to produce precise T1 maps within the relevant coil's sensitivity range. The resulting enlarged spin ensemble leads to high SNR and excellent T1 prediction if used in non-selective spectroscopic application. In addition, OC pulses were optimized for operation at a large bandwidth suitable for 31P spectroscopy (± 15 ppm), which is often a problem when adiabatic BIR-4 pulses are used. |
| 3430 | Computer 22
|
Magnetic Resonance Spectroscopy Spectral Registration with Unsupervised Deep Learning |
| David Jing Ma1, Yanting Yang1, Natalia Harguindeguy1, Ye Tian1, Scott A. Small2,3,4, Feng Liu2,5, Douglas L. Rothman6, and Jia Guo2,7 | ||
1Department of Biomedical Engineering, Columbia University, New York, NY, United States, 2Department of Psychiatry, Columbia University, New York, NY, United States, 3Department of Neurology, Columbia University, New York, NY, United States, 4Taub Institute Research on Alzheimer's Disease and the Aging Brain, Columbia University, New York, NY, United States, 5New York State Psychiatric Institute, Columbia University, New York, NY, United States, 6Radiology and Biomedical Imaging of Biomedical Engineering, Yale University, New Haven, CT, United States, 7Mortimer B. Zuckerman Mind Brain Behavior Institute, Columbia University, New York, NY, United States |
||
Keywords: Data Processing, Machine Learning/Artificial Intelligence A deep learning-based registration method has been a successful image processing tool adopted in medical image registration but there is a lack of learning-based registration tools for spectral registration protocols. A novel CNN-based unsupervised deep learning spectral registration model was developed and trained on a simulation dataset. The model was then further evaluated on a simulated test set with more extreme conditions and on an in vivo dataset and was compared performances to published frequency-and-phase correction models. An unsupervised deep learning-based spectral registration approach was found to demonstrate state-of-the-art performance in frequency-and-phase correction. |
| 3431 | Computer 23
|
Association of Ki-67 With Metabolite and Lipid Levels in High Grade Breast Cancer Patients based on Correlated MR Spectroscopic Imaging and Biopsy |
| Ajin Joy1, Andres Saucedo1, Melissa Joines1, Stephanie Lee-Felker1, Manoj K Sarma1, James Sayre1, Maggie Dinome2, and M. Albert Thomas1 | ||
1Radiological Sciences, University of California, Los Angeles, Los Angeles, CA, United States, 2Surgery, University of California, Los Angeles, Los Angeles, CA, United States |
||
Keywords: Data Analysis, Breast Five-dimensional (5D) echo-planar correlated spectroscopic imaging (EP-COSI) combines 2 spectral and 3 spatial dimensions to record two dimensional (2D) correlated spectroscopy (COSY) spectra in multiple regions in multiple slices. In this study, multiple 2D COSY spectra were recorded from breast cancer patients by 5D EP-COSI within practical scan time durations using non-uniform undersampling of one spectral and two spatial dimensions, and compressed sensing-based reconstruction. Different metabolite and lipid ratios were quantified and its association with Ki-67 metric was studied. Findings of this study showed statistically significant association of metabolite and lipid levels with Ki-67 measures in breast cancer patients. |
| 3432 | Computer 24
|
Glutamate Predicts Post-Concussion Symptoms in Collegiate Athletes |
| Annelise Lemaire1,2, Hui Jun Liao1, Inga Koerte3, David Howell4, Alexander Lin1,5, and Katherine Breedlove1,5 | ||
1Brigham and Women’s Hospital, Boston, MA, United States, 2University of Michigan, Ann Arbor, MI, United States, 3Ludwig-Maximilians-Universitat, Munich, Germany, 4University of Colorado-Anschutz, Aurora, CO, United States, 5Harvard Medical School, Boston, MA, United States |
||
Keywords: Data Analysis, Contrast Mechanisms, Traumatic Brain Injury A population of collegiate athletes, along with matched age and gender-matched student-athlete controls, were evaluated post-concussion, for neurometabolic levels in the PCG by magnetic resonance spectroscopy (MRS) and Post-Concussion Symptom Scale (PCSS) following injury. Evaluations were conducted 72 hours post-diagnosis. For the purpose of this work, correlations between total Glutamate (Glu+Gln), total Glutamate to total Creatine ratios, and symptoms on the PCS scale associated with cognitive fatigue - such as mental fog, drowsiness, fatigue, and feeling slowed - were explored. |
| 3433 | Computer 25
|
MRS Quantification using Deep Learning Frameworks: an Accuracy and Efficiency Study |
| Federico Turco1 and Johannes Slotboom2 | ||
1Support Center for Advanced Neuroimaging (SCAN), University of Bern, Bern, Switzerland, 2Support Center for Advanced Neuroimaging (SCAN), Institute of Diagnostic and Interventional Neuroradiology, Bern, Switzerland |
||
Keywords: Data Processing, Spectroscopy, Deep learning frameworks, Deep learning, Optimization We implemented a model fitting algorithm for magnetic resonance spectroscopy quantification using Pytorch. This network fit the spectra by minimizing the error between the spectrum modeled by the newtwork and the desired target. The implementation can fit multiple spectras in parallel using Pytorch GPU acceleration. We compared the results with a parallel version of TDFDFit and found it to be up to 12 time faster fitting 2048 spectras in 50 seconds. |
| 3434 | Computer 26
|
Performance of Multipool Lorentzian fitting for CEST-MRI using joint spatial regularization |
| Markus Huemer1, Oliver Maier1, Moritz Simon Fabian2, Manuel Schmidt2, Arnd Dörfler2, Kristian Bredies3, Moritz Zaiss2,4, and Rudolf Stollberger1,5 | ||
1Institute of Biomedical Imaging, Graz University of Technology, Graz, Austria, 2Institute of Neuroradiology, Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU), University Hospital Erlangen, Erlangen, Germany, 3Institute of Mathematics and Scientific Computing, University of Graz, Graz, Austria, 4High-Field Magnetic Resonance Center, Max-Planck Institute for Biological Cybernetics, Tübingen, Germany, 5BioTechMed Graz, Graz, Austria |
||
Keywords: Data Analysis, CEST & MT This work investigates an improved quantification algorithm for CEST spectra using a joint Frobenius TGV regularization. The performance is assessed with simulations, phantom and in-vivo measurement with respect to systematic errors and robustness. For comparison, a pixel-wise fit using the Levenberg-Marquardt algorithm and the IDEAL algorithm were implemented. The $$$TGV_j$$$ algorithm shows the best performance in the main parameter fitting stability and parameter SNR. |
| 3435 | Computer 27
|
Improved Bloch fitting and machine learning methods that analyze acidoCEST MRI |
| Tianzhe Li1,2, Julio Cardenas-Rodriguez3, and Marty David Pagel4 | ||
1Cancer Systems Imaging, UT MD Anderson Cancer Center, Houston, TX, United States, 2Medical Physics Program, UT Health, Houston, TX, United States, 3Data Translators LLC, Oro Valley, AZ, United States, 4Cancer Systems Imaging, MD Anderson Cancer Center, Houston, TX, United States |
||
Keywords: Data Analysis, CEST & MT, pH imaging AcidoCEST MRI can measure the extracellular pH of the tumor microenvironment. We have further refined our “Bloch fitting” method, and shown that this analysis method can accurately and precisely measure pH without additional MRI information, with an accuracy of 0.03 pH units. In addition, we have developed a machine learning method that can classify pH as > 7.0 or < 6.5 pH units (PPV=0.94, NPV=096), and a machine learning regression method that can estimate pH with a mean absolute error of 0.031 pH units. |
| 3436 | Computer 28
|
Demonstration of multi-pool quantitative CEST MRI with quasi-steady-state reconstruction – Insight into T1-independent CEST quantification |
| Lauren Gao1 and Phillip Sun1 | ||
1Emory University, Atlanta, GA, United States |
||
Keywords: Data Analysis, CEST & MT CEST quantification is challenging because the measurement depends on the scan protocols. Also, T1 normalization is not straightforward under non-equilibrium conditions. Using multi-pool simulations and phantom experiments, our study evaluated quasi-steady-state(QUASS) algorithm-boosted CEST quantification. The 3-pool CEST simulation showed significant T1 dependencies due to complex interactions among Ts, TR, and T1. Such dependencies were corrected with QUASS reconstruction. In addition, a multi-T1 phantom was used to evaluate the quantification. Whereas the apparent CEST MRI showed significant dependence on Ts, TR, and T1, accurate CEST quantification was demonstrated with the spinlock-model-based fitting of QUASS CEST MRI. |
| 3437 | Computer 29
|
The Influence of the MT effect on the Accuracy of T1 Quantification by the Dual Flip Angle Method |
| Zhen Hu1, Dan Zhu2,3, and Qin Qin2,3 | ||
1Department of Biomedical Engineering, Johns Hopkins University School of Medicine, Baltimore, MD, United States, 2Russell H. Morgan Department of Radiology and Radiological Science, Johns Hopkins University School of Medicine, Baltimore, MD, United States, 3Kirby Center, Kennedy Krieger Institute, Baltimore, MD, United States |
||
Keywords: Signal Modeling, Brain, Relaxometry The dual flip angle (DFA) method is widely used for brain T1 quantification. In addition to the well-known RF field effect, this study investigated the overlooked magnetization transfer (MT) effect for the accuracy of in-vivo T1 mapping. Based on the two-pool Bloch simulation, the DFA derived T1 estimation revealed FA and TR dependence, which was confirmed by the in-vivo brain T1 mapping using the same set of sequence parameters. Longer TR (30 ms) reduced the variation between FA pairs and were closer to the literature values, but at the cost of longer acquisition time. |
| 3438 | Computer 30
|
Downfield Proton MRSI with whole-brain coverage at 3T |
| İpek Özdemir1, Sandeep Ganji2, Joseph Gillen1,3, Michal Považan4, and Peter B Barker1,3 | ||
1Russell H. Morgan Department of Radiology and Radiological Science, Johns Hopkins University, Baltimore, MD, United States, 2Philips Healthcare, Best, Netherlands, 3F.M. Kennedy Krieger Institute, Baltimore, MD, United States, 4Danish Research Centre for MR, Copenhagen, Denmark |
||
Keywords: Data Acquisition, Spectroscopy, Downfield MRSI A 3D downfield-MRSI sequence with whole brain coverage, 22-minute scan time, and 0.7 cm3 nominal spatial resolution has been developed at 3T. The sequence was tested in 5 normal volunteers; LCModel analysis showed CRLB average values of 3-4% for protein amide resonances in 3 selected gray and white matter regions. |
| 3439 | Computer 31
|
In vivo Whole-brain Phosphoethanolamine Mapping using SLOW-EPSI at 7T |
| Guodong Weng1,2, Piotr Radojewski1,2, Federico Turco1,2, and Johannes Slotboom1,2 | ||
1Support Center for Advanced Neuroimaging (SCAN), University of Bern, Bern, Switzerland, 2Translational Imaging Center, sitem-insel, Bern, Switzerland |
||
Keywords: Visualization, Spectroscopy, Spectral editing In vivo detection of phosphoethanolamine (PE) using spectral editing has recently been shown to be feasible. This study shows whole-brain PE maps using SLOW-EPSI in a healthy subject with a nominal resolution of 4.3*4.3*10 mm. The result shows that the PE level is high in the cerebral cortex and low in the white matter. |
| 3440 | Computer 32
|
EigenMRS: A computationally cheap data-driven approach to MR spectroscopic imaging denoising |
| Amirmohammad Shamaei1,2, Jana Starcukova1, and Zenon Starcuk Jr1 | ||
1Institute of Scientific Instruments of the Czech Academy of Sciences Research institute in Brno, Brno, Czech Republic, 2Department of Biomedical Engineering, Brno University of Technology, Brno, Czech Republic |
||
Keywords: Data Processing, Spectroscopy, Singular value decomposition, MR spectroscopic imaging, Denoising The utility of MR spectroscopic imaging (MRSI) can be limited by a low signal-to-noise ratio (SNR) in practice. Averaging multiple coherent repetitions increases the SNR, but at the cost of time-consuming acquisition. Several computationally expensive approaches based on low-rank matrix approximation for denoising MRSI data have been proposed, which do not take advantage of previously acquired spectra. |
| 3441 | Computer 33
|
Revealing hidden phosphomonoester signals under phosphoethanolamine and phosphocholine resonances: A brain 31P MRS study at 7T |
| Jimin Ren1,2 | ||
1Advanced Imaging Research Center, University of Texas Southwestern Medical Center, Dallas, TX, United States, 2Department of Radiology, University of Texas Southwestern Medical Center, Dallas, TX, United States |
||
Keywords: Data Analysis, Spectroscopy, brain, phosphocholine, phosphoethanolamine, phosphomonoester It has been a common practice to quantify brain phosphomonoester (PME) 31P signals using a two-component model composed of phosphoethanolamine (PE) and phosphocholine (PC). This study demonstrates spectral evidence of the presence of a hidden broad PME (h-PME) signal underneath PE and PC resonances, characterized by a short T1 and potentially contributed by RBC 2,3-DPG in brain blood vessels, though other sources of signal contribution cannot be fully ruled out. The results have implication in using PE and PC as biomarkers of altered phospholipid metabolism in brain pathologies. |
| 3442 | Computer 34
|
Analysis of short-TE STEAM data at 7T: Effects of a macromolecular basis set and baseline parameters on GABA, glutamate and their ratio. |
| Tomohisa Okada1, Hideto Kuribayashi2, Yuta Urushibata2, Thai Akasaka1, Ravi Teja Seethamraju3, Sinyeob Ahn3, and Tadashi Isa1 | ||
1Human Brain Research Center, Kyoto University, Kyoto, Japan, 2Siemens Healthcare, Japan, Tokyo, Japan, 3Siemens Medical Solutions, Malvern, PA, United States |
||
Keywords: Data Analysis, Spectroscopy The effect of baseline flexibility LCModel parameter DKNTMN (0.15, 0.3, 0.6 and 1) on a 7T short-TE STEAM MRS measurement of GABA, glutamate and excitatory-inhibitory ratio (EIR) was investigated using both simulated and measured macromolecular basis sets. Mean (SD) of GABA/tCr was highest, 0.23 (0.02), using simulated MM basis. Using measured MM basis, the ratios decreased from 0.18 (0.04) to 0.12 (0.03) by the increase of DKNTMN values. The GABA/tCr ratio and EIR of a former multi-center study was 0.19 and 8.2 after T2 decay correction. Analysis using DKNTMN of 0.3 conformed best and considered to be the choice. |
| 3443 | Computer 35
|
Parametrized macromolecule models in the brain 1H MRS quantification: Effect of the baseline spline on quantification uncertainty and precision |
| Andrea Dell'Orco1,2, Layla Tabea Riemann2, Stephen L. R. Ellison3, Bernd Ittermann2, Anna Tietze1, Michael Scheel1, and Ariane Fillmer2 | ||
1Charité – Universitätsmedizin Berlin, corporate member of Freie Universität Berlin and Humboldt- Universität zu Berlin, Department of Neuroradiology, Berlin, Germany, Berlin, Germany, 2Physikalisch-Technische Bundesanstalt (PTB), Braunschweig und Berlin, Germany., Berlin, Germany, 3LGC Limited Registered, Teddington, Middlesex, United Kingdom |
||
Keywords: Data Analysis, Spectroscopy, macromolecules, single-voxel 1HMRS Parametric macromolecule (MM) models can improve the quantification of brain 1H-MRS. We quantified a publicly available repeated-acquisition dataset with LCModel and applied a parametric MM model in conjunction with four different settings for spline node distance and compared the quantification results. Coefficients of variation, average Cramér-Rao lower bound, repeatability, and reproducibility were estimated. A moderate effect of the baseline spline was shown. |
| 3444 | Computer 36
|
3D Seiffert spirals for efficient k-space sampling in 23Na-MRI: initial phantom simulations |
| Samuel Rot1,2, Matthew Clemence3, Bhavana S Solanky1,4, and Claudia AM Gandini Wheeler-Kingshott1,5,6 | ||
1NMR Research Unit, Queen Square MS Centre, Department of Neuroinflammation, UCL Queen Square Institute of Neurology, Faculty of Brain Sciences, University College London, London, United Kingdom, 2Department of Medical Physics and Biomedical Engineering, University College London, London, United Kingdom, 3Philips Healthcare, Best, Netherlands, 4Quantitative Imaging Group, Department of Medical Physics and Biomedical Engineering, University College London, London, United Kingdom, 5Department of Brain & Behavioural Sciences, University of Pavia, Pavia, Italy, 6Brain Connectivity Centre Research Department, IRCCS Mondino Foundation, Pavia, Italy |
||
Keywords: New Trajectories & Spatial Encoding Methods, Non-Proton Modern sodium MRI (23Na-MRI) uses efficient, non-Cartesian k-space sampling strategies to acquire 3D volumes in clinically feasible scan times. 3D Seiffert spirals are a novel k-space sampling scheme with improved efficiency to existing methods. 23Na-MRI, and particularly temporally resolved functional 23Na-MRI, could be an ideal application for Seiffert spirals. We present initial, in silico simulations of 3D Seiffert spiral k-space sampling for a highly undersampled 1.8 second functional 23Na-MRI protocol. Seiffert spirals generate images of improved quality and SNR in direct comparison to 3D-cones. Essential further work will involve compressed sensing reconstruction, further trajectory optimisation and in vivo imaging. |
| 3445 | Computer 37
|
Grading and assessment of intra-tumor heterogeneity of gliomas using 3D CEST imaging with compressed sensing and sensitivity encoding |
| Tatsuhiro Wada1,2, Osamu Togao3, Chiaki Tokunaga1, Kazufumi Kikuchi4, Koji Yamashita5, Masami Yoneyama6, Masahiro Oga1, Koji Kobayashi1, Toyoyuki Kato1, Kousei Ishigami4, and Hidetake Yabuuchi7 | ||
1Division of Radiology, Department of Medical Technology, Kyushu university hospital, Fukuoka, Japan, 2Department of Health Sciences, Graduate school of Medical Sciences, Kyushu University, Fukuoka, Japan, 3Department of Molecular Imaging & Diagnosis, Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan, 4Department of Clinical Radiology, Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan, 5Department of Radiology Informatics & Network, Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan, 6Philips Japan, Tokyo, Japan, 7Department of Health Sciences, Faculty of Medical Sciences, Kyushu University, Fukuoka, Japan |
||
Keywords: Tumors, CEST & MT Glioma grading using chemical exchange saturation transfer (CEST) imaging is often performed in a single cross-section. However, CEST imaging of multiple cross-sections is desirable for intra-tumor heterogeneity. Compressed sensing and sensitivity encoding (CS-SENSE) was applied to CEST imaging to obtain multi-slice CEST imaging in a clinically appropriate scan time. The diagnostic performance of three-dimensional (3D) CEST imaging was comparable to that of a two-dimensional CEST imaging. The evaluation of the entire tumors by multi-slice CEST imaging was important in the gliomas' grading because the signal intensities differed among the tumor slices. |
| 3446 | Computer 38
|
Theoretical justification for fluid-suppressed APTw imaging based on spillover correction principles |
| Maria Sedykh1, Stefano Casagranda2, Patrick Liebig3, Christos Papageorgakis2, Laura Mancini4,5, Sotirios Bisdas4,5, Manuel Schmidt1, Arnd Doerfler1, and Moritz Zaiss1 | ||
1Institut of Neuroradiology, University Clinic Erlangen, Friedrich-Alexander Universität Erlangen-Nürnberg, Erlangen, Germany, 2Department of R&D Advanced Applications, Olea Medical, La Ciotat, France, 3Siemens Healthcare GmbH, Erlangen, Germany, 4Lysholm Department of Neuroradiology, University College of London Hospitals NHS Foundation Trust, London, United Kingdom, 5Institute of Neurology UCL, London, United Kingdom |
||
Keywords: Tumors, CEST & MT Fluid suppression has an inestimable value in improving the readability of Amide Proton Transfer weighted imaging in the neuro-oncological field, by removing contamination effects in the fluid compartments. With this work we wanted to justify the derivation of this metric and to show different clinical examples. Our new fluid-suppressed APTw metric for 3T MR imaging, based on the principles of Spillover Correction for CEST imaging, can be then derived through the Bloch McConnell Equations. |
| 3447 | Computer 39
|
Amide proton transfer weighted (APTw) imaging-based radiomics in the molecular subtypes prediction and grading of adult-type diffuse gliomas |
| Minghao Wu1, Tongling Jiang2, Cong Xie3, Xianchang Zhang4, Yi Zhang5, and Yaou Liu1 | ||
1Department of Radiology, Beijing Tiantan Hospital, Capital Medical University, Beijing, China, beijing, China, 2Department of Biomedical Engineering, College of Biomedical Engineering & Instrument Science, Zhejiang University, Hangzhou, Zhejiang, China, Hangzhou, China, 3Beijing Tiantan Hospital, Capital Medical University, Beijing, China, beijing, China, 4MR Collaboration, Siemens Healthineers Ltd., Beijing, China, beijing, China, 5College of Biomedical Engineering & Instrument Science, Zhejiang University, Hangzhou, Zhejiang, China, Hangzhou, China |
||
Keywords: Radiomics, Brain, amide proton transfer weighted (APTw) imaging The fifth edition of the 2021 WHO Classification of Tumors of the Central Nervous System (WHO CNS5) has introduced significant changes in classifying glioma subtypes based on molecular profiles. This study investigated the feasibility of amide proton transfer weighted (APTw)-based radiomics for predicting molecular subtypes and WHO grade of adult-type diffuse gliomas. APTw-based radiomics achieved satisfactory performance in distinguishing three molecular subtypes and WHO grade (High (Ⅳ grade)/Low (Ⅱ-Ⅲ grade)). This application may be an effective and promising quantitative approach for better noninvasive characterization and classification of gliomas. |
| 3585 | Computer 1
|
Ultra rapid volumetric imaging for diagnosis in memory clinic: assessment of white matter hyperintensities quantification |
| Carole Helene Sudre1,2, Haroon R Chughtai2, David L Thomas3,4, David Cash4, Miguel Rosa-Grilo4, Millie Beament4, Frederik Barkhof2,4,5, Daniel Alexander2, Cath Mummery4, Nick Fox4, and Geoff JM Parker2 | ||
1MRC Unit for Lifelong Health and Ageing at UCL, University College London, London, United Kingdom, 2Centre for Medical Image Computing, University College London, London, United Kingdom, 3Department of Brain Repair and Rehabilitation, UCL Queen Square Institute of Neurology, University College London, London, United Kingdom, 4Dementia Research Centre, UCL Queen Square Institute of Neurology, University College London, London, United Kingdom, 5Department of Radiology & Nuclear Medicine, Amsterdam UMC, Amsterdam, Netherlands |
||
Keywords: Segmentation, Dementia Acquisition time is a crucial factor in the tolerability and availability of MRI examinations for patients attending memory clinics. The suitability of using prototype ultra-rapid sequences for the quantification of white matter hyperintensities was assessed by comparing segmentation outputs from standard clinical and ultra-rapid sequences. Despite a slightly lower sensitivity to smaller lesions, quantification of white matter hyperintensities when using ultra-rapid sequences was very highly correlated with lesion volumes obtained from standard sequences. |
| 3586 | Computer 2
|
Fully Automated Hippocampus Segmentation Pipeline using Deep Convolutional Neural Networks |
| Maximilian Sackl1, Christian Tinauer1, Christian Enzinger1, Reinhold Schmidt1, and Stefan Ropele1 | ||
1Department of Neurology, Medical University Graz, Graz, Austria |
||
Keywords: Segmentation, Alzheimer's Disease, Multi-Contrast, AI Segmentation of the hippocampus on T1-weighted structural MR images is required to quantify the neurodegenerative effects in Alzheimer’s disease studies. In this work, we propose an automated artificial intelligence-based pipeline for hippocampus segmentation combined with manual ground truth (GT) data that originates from high-resolution T2-weighted MR images. Results are evaluated against the manual GT-labels and compared to the segmentation results from FreeSurfer v732. Our deep learning-based segmentation outperforms FreeSurfer in terms of accuracy and speed, while reference experiments using the T2-based GT-labels yield the best results. Thus, using T2-weighted images for ground truth generation can improve automated HC segmentation. |
| 3587 | Computer 3
|
Improved Brain Lesion Detection by Integrating Subspace-based Deep Learning of Normal Intensity Distributions and Bayesian Hypothesis Testing |
| Huixiang Zhuang1, Yue Guan1, Yi Ding1, Yuhao Ma1, Yunpeng Zhang1, Ziyu Meng1, Ruihao Liu1,2, Zhi-Pei Liang2,3, and Yao Li1 | ||
1School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China, 2Beckman Institute for Advanced Science and Technology, University of Illinois at Urbana-Champaign, Urbana, IL, United States, 3Department of Electrical and Computer Engineering, University of Illinois at Urbana-Champaign, Urbana, IL, United States |
||
Keywords: Segmentation, Machine Learning/Artificial Intelligence, Unsupervised learning; Lesion segmentation Unsupervised segmentation of brain lesions is desirable in many applications and has been investigated extensively. In this work, we proposed a new method for brain lesion segmentation, which effectively learns the spatial-intensity distribution of normal brain tissues and then treats lesion segmentation as an anomaly detection problem. We overcame the high-dimensional distribution learning problem using a subspace-assisted generative network. With the learned distribution, the anomaly detection problem was solved using Bayesian hypothesis testing. Our method has been validated using simulated and real brain MR images with stroke and tumor lesions, and produced significantly improved results than several state-of-the-art methods. |
| 3588 | Computer 4
|
Human brain thalamic nuclei segmentation: compressed-sensing effects on 3D MPRAGE variants |
| Sebastian Hübner1, Stefano Tambalo1, Tobias Kober2,3,4, and Jorge Jovicich1 | ||
1Center for Mind/Brain Sciences, University of Trento, Rovereto (Trento), Italy, 2Advanced Clinical Imaging Technology, Siemens Healthineers International AG, Lausanne, Switzerland, 3Department of Radiology, Lausanne University Hospital and University of Lausanne, Lausanne, Switzerland, 4LTS5, École Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland |
||
Keywords: Segmentation, Brain, Thalamus Accurate segmentation of thalamic nuclei can significantly support clinical decisions, especially in diseases such as Parkinson’s. At the same time, highly accelerated T1w acquisition techniques find wider clinical acceptance as they not only increase radiological efficiency and patient comfort, but also reduce motion artefacts. Here, we study how compressed sensing affects thalamic nuclei volumetry based on MPRAGE or MP2RAGE in healthy subjects with respect to clinical baseline. Segmentation results are promising and agree with expectations: acceleration to about 1 min is possible maintaining overall good image quality, contrast and thalamic segmentations, with acceleration-related biases. |
| 3589 | Computer 5
|
Comparison of image-based and statistical approaches for brain volumetry harmonization |
| Yuan-Chiao Lu1,2, Blake E Dewey3, Yi-Yu Chou1,2, Danielle Greenman1,2, Russell T Shinohara4, Daniel S Reich5, Jerry L Prince6, John A Butman2, and Dzung L Pham2,7 | ||
1Center for Neuroscience and Regenerative Medicine, Henry M Jackson Foundation for the Advancement of Military Medicine, Bethesda, MD, United States, 2Radiology and Imaging Sciences, National Institutes of Health, Bethesda, MD, United States, 3Neurology, Johns Hopkins University, Baltimore, MD, United States, 4Department of Biostatistics, Epidemiology and Informatics, University of Pennsylvania, Philadelphia, PA, United States, 5Translational Neuroradiology Section, National Institute of Neurological Disorders and Stroke, Bethesda, MD, United States, 6Electrical and Computer Engineering, Johns Hopkins University, Baltimore, MD, United States, 7Radiology and Radiological Sciences, Uniformed Services University, Bethesda, MD, United States |
||
Keywords: Segmentation, Brain, Harmonization Image harmonization approaches have been proposed for reducing the variation of brain image measurements in studies involving acquisitions from multiple scanner protocols and hardware. This study compares two harmonization methods, DeepHarmony, a deep learning-based image synthesis approach, and ComBat, a statistical batch correction tool, based on their ability to yield consistent brain volume measurements from two different T1-weighted acquisitions. Our study showed that DeepHarmony outperformed the ComBat approach, although both approaches significantly improved consistency when compared to unharmonized images. |
| 3590 | Computer 6
|
A new cascaded fully convolutional neural network for the simultaneous segmentation of parasagittal dural space and arachnoid granulations |
| Kilian Hett1, Colin D. McKnight2, Jennifer S. Lindsey2, Melanie Leguizamon1, Jarrod Eisma1, Alexander K. Song1, Jason Elenberger1, Ciaran M. Considine1, Daniel O. Claassen1, and Manus J. Donahue1 | ||
1Department of Neurology, Vanderbilt University Medical Center, Nashville, TN, United States, 2Department of Radiology, Vanderbilt University Medical Center, Nashville, TN, United States |
||
Keywords: Segmentation, Neurofluids The overarching goal of this work is to develop and validate novel deep learning algorithms for segmenting the peri-sinus space, including parasagittal dural (PSD) space, which has been hypothesized to harbor cerebral lymphatic channels, and intra-veinous arachnoid granulations, which has been long-hypothesized as a site a CSF egress, from standard non-contrast anatomical imaging. The new segmentation method is based on cascaded neural networks using non-contrasted 3D T2-weighted MRI; the method is method in a mixed cohort of adults with and without neurodegeneration. |
| 3591 | Computer 7
|
Robust thalamic nuclei segmentation from T1 MRI data |
| Julie P. Vidal1,2, Lola Danet2,3, Patrice Péran2, Jérémie Pariente2,3, Dolf Pfefferbaum4, Edith Sullivan4, Emmanuel J. Barbeau1, and Manojkumar Saranathan5 | ||
1CNRS, CerCo (Brain and Cognition Research Center) - Université Paul Sabatier, Toulouse, France, 2INSERM, ToNiC (Toulouse NeuroImaging Center) - Université Paul Sabatier, Toulouse, France, 3Hôpital Purpan, Centre Hospitalier Universitaire de Toulouse, Département de Neurologie, Toulouse, France, 4Stanford University School of Medicine, Department of Psychiatry & Behavioral Sciences, Stanford, CA, United States, 5UMass Chan Medical School, Department of Radiology, Worcester, MA, United States |
||
Keywords: Segmentation, Data Processing, WMn synthesis, HIPS This work presents a methodology for thalamic nuclei segmentation from T1w MRI. Thalamus Optimized Multi-Atlas Segmentation (THOMAS), which was originally developed for white-matter nulled (WMn) MPRAGE, is adapted for standard T1 MRI by employing synthesis techniques to make the T1 images closer to WMn contrast. Robustness is tested across image contrast, MRI manufacturer, and field strength. |
| 3592 | Computer 8
|
Improving Early Post-Operative Glioblastoma Segmentation With Semi-Supervised Deep Learning |
| Lidia Luque1,2,3, Karoline Skogen4, Bradley J MacIntosh3,5,6, Kyrre Eeg Emblem1, Christopher Larsson3,7, Einar O Vik-Mo7, and Atle Bjørnerud1,3 | ||
1Department of Physics and Computational Radiology, Oslo University Hospital, Oslo, Norway, 2Department of Physics, University of Oslo, Oslo, Norway, 3Computational Radiology and Artificial Intelligence (CRAI), Oslo University Hospital, Oslo, Norway, 4Department of Radiology and Nuclear Medicine, Oslo University Hospital, Oslo, Norway, 5Department of Medical Biophysics, University of Toronto, Toronto, ON, Canada, 6Sandra E Black Centre for Brain Resilience and Recovery, Sunnybrook Research Institute, Toronto, ON, Canada, 7Department of Neurosurgery, Oslo University Hospital, Oslo, Norway |
||
Keywords: Segmentation, Machine Learning/Artificial Intelligence, Semi-supervision Improving automatic segmentation of glioblastoma on early post-operative MRI is key to study the effect of resection volumes on patient outcomes. We curate a dataset of over 700 MRI examinations, of which 87 include annotations, and train a supervised and a semi-supervised deep-learning model. Semi-supervision improves the segmentation of the high-intensity FLAIR signal with 3% to a Dice score of 0.83 (p=0.031), while the segmentation of the enhancing tumor increases with 9% to 0.55 (p=0.056). However, enhancing tumor segmentations show high variability, possibly due to imperfect annotations. Segmentation of enhancing tumor on early post-operative MRI remains a challenging task. |
| 3593 | Computer 9
|
Imaging The Thalamic Reticular Nucleus at 7T |
| ross shaw1, Penny Gowland1, and Richard Bowtell2 | ||
1University of Nottingham, Nottingham, United Kingdom, 2Physics, University of Nottingham, Nottingham, United Kingdom |
||
Keywords: Segmentation, Contrast Mechanisms, Thalamus The Thalamic Reticular Nucleus is a micro-structure surrounding the dorsal edge of the thalamus that is of fundamental significance to understanding many neurological disorders in the brain. Here we present methods to visualise the TRN using various imaging modalities to explore different sources of contrast at the edge of the thalamus. |
| 3594 | Computer 10
|
High quality brain segmentation from multi-parameter mapping at Ultra-High Field MRI |
| Marc-Antoine Fortin1, Yannik Völzke2, Rüdiger Stirnberg2, Siya Sherif3, Laurent Lamalle3, Tony Stöcker2,4, and Pål Erik Goa1 | ||
1Department of Physics, Norwegian University of Science and Technology (NTNU), Trondheim, Norway, 2German Center for Neurodegenerative Diseases (DZNE), Bonn, Germany, 3GIGA Cyclotron Research Centre (GIGA-CRC) in vivo imaging, University of Liège, Liège, Belgium, 4Department of Physics and Astronomy, University of Bonn, Bonn, Germany |
||
Keywords: Segmentation, High-Field MRI, Multi-Parameter Mapping, Quantitative MRI, 7T Very few segmentation techniques are optimized for 7T images. Due to the lack of approaches and the high level of inhomogeneity observed with Ultra-High Field images, neuroscientists typically limit the analysis to the upper half of the brain. A novel resolution- and contrast-agnostic technique, called SynthSeg, may remedy this limitation. A multi-parameter mapping protocol and MPRAGE images were acquired to test SynthSeg against FreeSurfer on 7T sub-millimeter brain images. Our results showed that SynthSeg can surpass FreeSurfer, especially for the cerebellum and temporal lobes. Moreover, we showed that quantitative maps can be great surrogates to high-resolution T1w images like MPRAGE. |
| 3595 | Computer 11
|
Validation of a Semi-Automated Method to Quantify Lesion Volume Changes in Multiple Sclerosis Using Subtraction Images |
| Rozemarijn M. Mattiesing1, Serena Stel1, Alysha S. Mangroe1, Iman Brouwer1, Adriaan Versteeg1, Ronald A. van Schijndel1, Bernard M.J. Uitdehaag2, Frederik Barkhof1,3, Hugo Vrenken1, and Joost P.A. Kuijer1 | ||
1MS Center Amsterdam, Radiology and Nuclear Medicine, Amsterdam Neuroscience, Amsterdam UMC location VUmc, Amsterdam, Netherlands, 2MS Center Amsterdam, Neurology, Amsterdam Neuroscience, Amsterdam UMC location VUmc, Amsterdam, Netherlands, 3UCL London, Institutes of Neurology and Healthcare Engineering, London, United Kingdom |
||
Keywords: Segmentation, Multiple Sclerosis Monitoring changes in white matter lesions with MRI is important to evaluate the effects of treatment in multiple sclerosis. In this study a validation of a semi-automated method to quantify lesion volume changes based on 2D proton-density-weighted images and image subtraction was performed. With this method new and enlarging but also disappearing and shrinking lesion activity can be quantified. As assessed with the intraclass correlation coefficient for absolute agreement, we found that the reproducibility was excellent and the accuracy was good overall. This semi-automated subtraction method can reliably quantify lesion volume changes in patients with (early) multiple sclerosis. |
| 3596 | Computer 12
|
Differential patterns of automatic segmentation of 3T and 7T MRI: Implications in neuropsychiatric disease |
| Gaurav Verma1, Ki-Sueng Choi1, Helen Mayberg2, and Priti Balchandani1 | ||
1Radiology, Icahn School of Medicine at Mount Sinai, New York, NY, United States, 2Neurology, Icahn School of Medicine at Mount Sinai, New York, NY, United States |
||
Keywords: Segmentation, High-Field MRI, FreeSurfer, MDD Automatic segmentation was performed on T1-MPRAGE structural MRI data acquired at 3T and 7T from 37 and 69 distinct healthy controls, respectively. Additionally, segmentation was performed on imaging acquired from 215 major depressive disorder (MDD) patients at 3T and 40 MDD patients at 7T. Of 259 segmentation-derived imaging features evaluated, 120 showed significant 3T vs. 7T differences among controls, and 153 among patients. 7T imaging metrics showed consistently lower cortical thickness and cortical gray/white matter ratios. Subcortical and cortical volumes measured at 7T were more mixed, with 7T images showing greater frontal lobe volume, but lower cortical volumes elsewhere. |
| 3597 | Computer 13
|
Reliability of brain volume measurements on 3D-FLAIR is similar to 3D-T1 in multiple sclerosis |
| David Rudolf van Nederpelt1, Samantha Noteboom2, Eva M.M. Strijbis3, Iman Brouwer1, Bastiaan Moraal1, Bas M.S. Jasperse1, Henk-Jan M.M. Mutsaerts1, Menno M. Schoonheim2, Frederik Barkhof1,4, Hugo Vrenken1, and Joost P.A. Kuijer1 | ||
1Radiology and nuclear medicine, MS Center Amsterdam, Radiology and Nuclear Medicine, Vrije Universiteit Amsterdam, Amsterdam Neuroscience, Amsterdam UMC location VUmc, Amsterdam, The Netherlands, Amsterdam, Netherlands, 2Anatomy and Neurosciences, MS Center Amsterdam, Anatomy and Neurosciences, Vrije Universiteit Amsterdam, Amsterdam Neuroscience, Amsterdam UMC location VUmc, Amsterdam, The Netherlands, Amsterdam, Netherlands, 3MS Center Amsterdam, Neurology, Vrije Universiteit Amsterdam, Amsterdam Neuroscience, Amsterdam UMC location VUmc, Amsterdam, The Netherlands, Amsterdam, Netherlands, 4Queen Square Institute of Neurology and Centre for Medical Image Computing, University College London, UK, London, United Kingdom |
||
Keywords: Segmentation, Multiple Sclerosis New methods have enabled segmentation on sequences other than T1 for brain volume measurements in multiple sclerosis (MS), such as those more readily available in clinic as FLAIR and PD. However, reliability studies have not been performed yet in MS, limiting their use. Here, we assessed within-scanner and between-scanner reliability of FLAIR segmentations in 30 people with MS on three different MR scanners. We found similar reliability compared to T1w scans, with high intra-scanner reliability but systematic differences between scanners. This suggests that within-scanner volume measurements of FLAIR are possible, but standardization is needed for between-scanner single patient measurements. |
| 3598 | Computer 14
|
Comparison of whole brain parenchyma fraction (BPF) from T1W and diffusion MRI in the assessment of CLN3, a neurodegenerative disease |
| Amritha Nayak1,2, An N Dang Do3, Audrey E Thurm4, Ariane Soldatos5, Forbes D Porter3, and Carlo Pierpaoli1 | ||
1Laboratory on Quantitative Medical Imaging, NIBIB, NIH, Bethesda, MD, United States, 2Henry Jackson Foundation for Advancement of Military Medicine, Bethesda, MD, United States, 3Division of Translational Medicine, NICHD,NIH, Bethesda, MD, United States, 4Office of Clinical Director, NIMH,NIH, Bethesda, MD, United States, 5Office of Clinical Director, NINDS,NIH, Bethesda, MD, United States |
||
Keywords: Segmentation, Quantitative Imaging, Neurodegeneration, Neurodegeneration diseases, aging, T1-weighted segmentation, T1W, DTI, Diffusion Tensor Imaging, brain parenchyma volume, brain atrophy In this work, we have compared if whole brain parenchyma fraction (BPF) measured from conventionally used T1-weighted (T1W) based brain segmentation method is comparable to signal fraction attributable to parenchymal water (Par-SF) measured from method using diffusion MRI, in assessing the overall disease state of participants with CLN3, a pediatric neurodegenerative disease. |
| 3599 | Computer 15
|
Tract-specific Evaluation of White Matter Hyperintensity Segmentation |
| Aaron Sinclair1, Ross Callaghan2, and Hui Zhang1 | ||
1Centre for Medical Image Computing and Department of Computer Science, University College London, London, United Kingdom, 2AINOSTICS Ltd., Manchester, United Kingdom |
||
Keywords: Segmentation, Neuroinflammation, White Matter Hyperintensity Segmentation We propose a tract-specific white matter hyperintensity (WMH) segmentation evaluation as a new way to assess WMH segmentation techniques. Currently, WMH segmentation is assessed globally or between periventricular and deep white matter (WM) regions, which has no obvious functional relevance. WM tracts have known functional relevance and associations to neurological conditions. We demonstrate that this new approach allows for functionally-relevant assessment of techniques. Two WMH segmentation techniques are compared to highlight the performance differences across a key tract associated with processing speed. We also found significant performance differences across this tract, which would be missed by standard evaluation methods. |
| 3600 | Computer 16
|
Intensity-Based Segmentation of Tumor on Multiparametric MRI to Aid Response Assessment of High-Grade Gliomas Treated with Immunoradiotherapy |
| Harshan Ravi1, Samuel H. Hawkins1,2, Olya Stringfield3, Malesa Pereira1,4,5, Heiko Enderling6,7, H-H Michael Yu7,8, John A. Arrington5,7, Solmaz Sahebjam7,9,10, and Natarajan Raghunand1,7 | ||
1Moffitt Cancer Center and Research Center, Tampa, FL, United States, 2Department of Computer Science, Bradley University, Peoria, IL, United States, 3Quantitative Imaging Core, Moffitt Cancer Center and Research Institute, Tampa,, FL, United States, 4Behavioral and Community Health Sciences,, LSU Health School of Public Health, New Orleans, LA, United States, 5Department of Radiology, Moffitt Cancer Center and Research Center, Tampa,, FL, United States, 6Department of Integrated Mathematical Oncology, Moffitt Cancer Center and Research Institute, Tampa,, FL, United States, 7Department of Oncologic Sciences, University of South Florida, Tampa, FL, United States, 8Department of Radiation Oncology, Moffitt Cancer Center and Research Center, Tampa,, FL, United States, 9National Cancer Institute, National Institutes of Health, Bethesda, MD, United States, 10Department of Neuro-Oncology, Moffitt Cancer Center and Research Center, Tampa,, FL, United States |
||
Keywords: Segmentation, Machine Learning/Artificial Intelligence, Glioblastoma, immunotherapy, radiotherapy Confounding appearance of radiographic changes in recurrent high grade glioma (HGG) patients treated with multimodality immunotherapy presents a challenge to the neuro-radiologist. A clinical need exists to improve upon conventional criteria for assessment of GBM on standard-of-care (SOC; T1w, T2w, FLAIR, and T1w-enhanced) MRI to help distinguish treatment-related effects from true disease progression. We have investigated the feasibility of intensity-based segmentation of tumor tissue types on multiparametric MRI (mpMRI) to inform response assessment in HGG patients treated with bevacizumab, hypofractionated stereotactic radiotherapy, and pembrolizumab. |
| 3601 | Computer 17
|
A general framework for analyzing the various contributions to reproducibility in brain morphometry |
| Ruifeng Dong1, Amritha Nayak1,2, Leighton Chan3, and Carlo Pierpaoli1 | ||
1Laboratory on Quantitative Medical Imaging, National Institute of Biomedical Imaging and Bioengineering, National Institutes of Health, Bethesda, MD, United States, 2Henry Jackson Foundation for Advancement of Military Medicine, Bethesda, MD, United States, 3Rehabilitation Medicine Department, National Institutes of Health, Bethesda, MD, United States |
||
Keywords: Segmentation, Software Tools We propose a framework for differentiating the contributions to the reproducibility in brain morphometry from the true inter-individual differences, experimental procedures, and data-processing methods. As an application, we build a linear mixed-effect model to evaluate and compare two segmentation software tools, Freesurfer and vol2Brain, in their reproducibility in measuring volumes of 32 regions of interest. For both software and for most structures under study, our approach successfully reveals the dominance of inter-subject variability over noise. Vol2Brain introduces less noise than Freesurfer for all subcortical nuclei while Freesurfer shows better performance for gray matter, cortex, cerebral white matter, and cerebellum cortex. |
| 3602 | Computer 18
|
Globally optimal region thresholding based on bimodal Gaussian distribution with physiological constraints : head MR example. |
| Artem Mikheev1 and Henry Rusinek1 | ||
1Radiology, NYU Grossman School of Medicine, New York, NY, United States |
||
Keywords: Segmentation, Segmentation, Thresholding In medical image processing thresholding of the image region is one of the most basic operations and preprocessing. While there are well-proven histogram-based partitioning methods including Otsu and Gaussian Mixture Model ) it is challenging to combine those methods with physiological constraints such as tissue volume ratio or average signal ratio to avoid anatomically invalid segmentation. We propose a new method GOTC when the distribution of the signal over the ROI is described as a bimodal Gaussian of two tissues. We describe the algorithm and validate it on brain MRI segmentation into the White Matter and Gray Matter. |
| 3603 | Computer 19
|
CEREBELLAR SURFACE PARCELLATION BASED ON DEFORMABLE SPHERICAL TRANSFORMER |
| Jiale Cheng1,2, Fenqiang Zhao2, Zhengwang Wu2, Ya Wang2, Xinrui Yuan2, Weili Lin2, Li Wang2, Xin Zhang1, and Gang Li2 | ||
1School of Electronic and Information Engineering, South China University of Technology, Guangzhou, China, 2Department of Radiology and BRIC, University of North Carolina at Chapel Hill, Chapel Hill, NC, United States |
||
Keywords: Segmentation, Brain Accurate parcellation of the extremely folded cerebellar cortex is of immense importance for both brain structural and functional studies. Manual parcellation is time-consuming and expertise dependent, which motivates us to propose a novel end-to-end deep learning-based method for cerebellar cortical surface parcellation. Leveraging the spherical topology of the cerebellar surface, we propose the Deformable Spherical Transformer, which combines the advantages of the Spherical Transformer to extract the long-range dependency and the deformable attention mechanism to adaptively focus on the critical regions. Its superior performance has been validated by comparing with advanced algorithms with an average Dice ratio of 86.40%. |
| 3604 | Computer 20
|
Deep-learning auto-segmentation and subtype classification of pituitary adenoma based on MRI radiomics |
| Bing-Fong Lin1, Dao-Chen Lin2,3,4, and Chia-Feng Lu1 | ||
1Department of Biomedical Imaging and Radiological Sciences, National Yang Ming Chiao Tung University, Taipei, Taiwan, 2Department of Radiology, Taipei Veterans General Hospital, Taipei, Taiwan, Taipei, Taiwan, 3Division of Endocrine and Metabolism, Department of Medicine, Taipei Veterans General Hospital, Taipei, Taiwan, 4School of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan |
||
Keywords: Radiomics, Quantitative Imaging, Pituitary adenoma Pituitary adenoma (PA) accounts for approximately 15% in intracranial neoplasms. The classification of PA generally based on the hormone level of blood as the gold standard test, while the analysis of hormone condition using neuroimaging biomarkers was less explored. Accordingly, our study developed a model to automatically segment PA and further used the quantitative and non-invasive MRI technique as image biomarkers to classify the three types of hormone pattern, focusing on corticortroph, gonadotroph, and plurihormonal type. We aimed to provide a feasible classification model based MRI to benefit the clinical management of patients with PA. |
| 3605 | Computer 21
|
Hybrid-space reconstruction with add-on distortion correction for simultaneous multi-slab DWI |
| Jieying Zhang1, Simin Liu1, Yishi Wang2, and Hua Guo1 | ||
1Center for Biomedical Imaging Research, Department of Biomedical Engineering, School of Medicine, Tsinghua University, Beijing, China, 2Philips Healthcare, Beijing, China |
||
Keywords: Image Reconstruction, Diffusion Tensor Imaging 3D simultaneous multi-slab imaging (SMSlab) can achieve high-resolution DWI with high SNR efficiency. Recently, we integrated SMSlab DWI with blipped-CAIPI gradients (blipped-SMSlab) and proposed a hybrid-space reconstruction algorithm, REACH. In this study, REACH is extended for distortion correction, which is called DC-REACH. It can correct the image distortions and the phase interferences introduced by the blipped-CAIPI gradients simultaneously. It also distinctly reduces the g-factor penalty via the joint reconstruction of the blip-up/down data. |
| 3606 | Computer 22
|
Self-Calibrating Aliasing-Controlled Simultaneous Multi-Slice Reconstruction for Diffusion MRI |
| Eun Ji Lim1, Hyunkyung Maeng1, and Jaeseok Park1 | ||
1Department of Intelligent Precision Healthcare Convergence, Sungkyunkwan University, Suwon, Korea, Republic of |
||
Keywords: Image Reconstruction, Diffusion Tensor Imaging To jointly resolve inter-slice leakages and in-plane aliasing for combined in-plane- and slice-accelerated diffusion MRI data, we proposed a novel, one-step solution for SMS-reconstruction optimally exploiting self-calibrating data from generalized 3D Fourier encoding perspective. To this end, we propose a generalized SMS forward signal model with an extended controlled aliasing and an extended self-calibration. Aliasing artifacts are jointly resolved in ky-and kz-directions by balancing null space consistency with a low rank prior while enforcing data fidelity in 3D k-space. We demonstrated the proposed method outperforms competing methods for diffusion MRI at SMS=3, R=2. |
| 3607 | Computer 23
|
Self-navigated high-resolution 3D diffusion MRI using an extended blipped-CAIPI sampling and structured low-rank reconstruction |
| Ziyu Li1, Xi Chen1, Mark Chiew1,2,3, Karla L. Miller1, and Wenchuan Wu1 | ||
1Wellcome Centre for Integrative Neuroimaging, FMRIB, Nuffield Department of Clinical Neurosciences, University of Oxford, Oxford, United Kingdom, 2Physical Sciences, Sunnybrook Research Institute, Toronto, ON, Canada, 3Department of Medical Biophysics, University of Toronto, Toronto, ON, Canada |
||
Keywords: Image Reconstruction, Diffusion/other diffusion imaging techniques, Diffusion Acquisition & Reconstruction 3D multi-shot multi-slab imaging can provide superior SNR efficiency for high-resolution dMRI but suffers from motion-induced shot-to-shot phase variations. We propose a highly-efficient, self-navigated method to correct for phase variations in 3D multi-slab dMRI. The sampling of each shot is designed to intersect with the kz=0 plane. These intersections are used to reconstruct a 2D phase map for each shot using a structured low-rank reconstruction that leverages the redundancy in shot and coil dimensions. The phase maps are used to eliminate the phase inconsistency in the final 3D multi-shot reconstruction. We demonstrate the method’s efficacy using highly realistic simulations. |
| 3608 | Computer 24
|
Pre-Excitation Gradients for Eddy Current-Nulled Convex Optimized Diffusion Encoding to Mitigate Distortion in 2D Diffusion Weighted Imaging |
| Matthew J. Middione1, Michael Loecher1, Xiaozhi Cao1, Kawin Setsompop1,2, and Daniel B. Ennis1,3 | ||
1Department of Radiology, Stanford University, Stanford, CA, United States, 2Department of Electrical Engineering, Stanford University, Stanford, CA, United States, 3Cardiovascular Institute, Stanford University, Stanford, CA, United States |
||
Keywords: Pulse Sequence Design, Diffusion/other diffusion imaging techniques, Eddy Currents Diffusion encoding gradients produce eddy currents that cause image distortions in DWI. Twice refocused spin-echo (TRSE) and eddy current-nulled convex optimized diffusion encoding (ENCODE) mitigate eddy current-induced image distortions in DWI, but at the expense of extending the TE. Herein, we revise the original ENCODE method by playing additional pre-excitation gradient lobes (Pre-ENCODE). Using simulations, phantom experiments, and in vivo imaging we demonstrate that Pre-ENCODE mitigates eddy current-induced image distortions in DWI with a shorter TE than TRSE and ENCODE. |
| 3609 | Computer 25
|
Impact of gradient spoiling for diffusion-weighted Double-Echo Steady-State sequences |
| Ulrich Katscher1, Jakob Meineke1, and Jochen Keupp1 | ||
1Philips Research Europe, Hamburg, Germany |
||
Keywords: Pulse Sequence Design, Diffusion/other diffusion imaging techniques Double-Echo Steady-State (DESS) sequences are a promising candidate for diffusion weighted imaging (DWI) free of geometric distortions. While diffusion weighted DESS (dwDESS) sequences were originally introduced with a unipolar diffusion weighting gradient GD, a bipolar GD is required to obtain a motion robust fully balanced sequence. The inevitable banding artefacts occurring for bipolar GD can be handled via different techniques like gradient spoiling (thus deviating from the fully balanced sequence) or phase cycling. This study compares these techniques to optimize SNR for a given scan time and given diffusion weighting. |
| 3610 | Computer 26
|
Motion robust and high resolution DWI utilizing the combination of Multi-shot Reduced FOV imaging with motion-compensated diffusion gradients |
| Zhigang Wu1, Yajing Zhang2, Guillaume Gilbert3, Wengu Su4, Yan Zhao5, and Jiazheng Wang6 | ||
1Philips Healthcare, Shenzhen, Ltd., Shenzhen, China, 2Philips Health Technology, Suzhou, China, 3MR Clinical Science, Philips Healthcare, Mississauga, ON, Canada, 4BU MR Application, Philips Health Technology, Suzhou, China, 5BU MR R&D, Philips Health Technology, Suzhou, China, 6Philips Healthcare, Beijing, China |
||
Keywords: Pulse Sequence Design, Diffusion/other diffusion imaging techniques, Diffusion, Motion compensated diffusion gradients, Reduced FOV imaging, Multi-shot DWI Reduced FOV imaging (rFOV) and multi-shot DWI both are very useful techniques to improve spatial resolution for detection of pancreatic lesion. However, respiratory and cardiovascular motion will introduce severe artifacts and ADC bias. Motion compensated diffusion gradients (MOCO) could be used to improve the image quality. In this study, a new sequence named MOCO-rFOV IRIS was developed, which combines the advantages of rFOV, MOCO and image reconstruction using image-space sampling function based multi-shot DWI (IRIS). Results from in vivo data demonstrated that the proposed method could be used to realize motion robust and high resolution DWI for pancreas |
| 3611 | Computer 27
|
DISCUS: Diffusion MRI Signal Reconstruction with Continuous Sampling |
| Christian Ewert1, David Kügler1, Anastasia Yendiki2,3, and Martin Reuter1,2,3 | ||
1AI in Medical Imaging, German Center for Neurodegenerative Diseases (DZNE), Bonn, Germany, 2Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Boston, MA, United States, 3Department of Radiology, Harvard Medical School, Boston, MA, United States |
||
Keywords: Image Reconstruction, Diffusion/other diffusion imaging techniques, q-space, denoising DISCUS addresses two challenges currently limiting the analysis potential of diffusion MRI: sparsity of measurements and variability in q-space sampling schemes. Our method combines the advantages of model-fit approaches with continuous sampling (spherical harmonics, SHORE) and rigid, discrete learning-based methods. DISCUS can be initialized from any acquisition scheme and permits signal prediction for an arbitrary q-vector. Despite the added flexibility, DISCUS performs on par with other, far less flexible learning methods, while outperforming model-fit methods. DISCUS-derived signals translate to higher-quality FA estimates promising accurate analyses even from very short acquisitions. |
| 3612 | Computer 28
|
Orthogonal diffusion encoding gradient sequence (ODEG) improves time-dependency measurements in the human brain |
| Qinfeng Zhu1, Haotian Li1, Yi-Cheng Hsu2, Yi Sun2, and Dan Wu1 | ||
1zhejiang University, Hangzhou, China, 2Siemens Healthcare China, Shanghai, China |
||
Keywords: Pulse Sequence Design, Diffusion/other diffusion imaging techniques Oscillatory gradient dMRI is used to access restricted diffusion at short diffusion times (td). But its clinical application is limited due to the limited gradient strength, leading low b-value and low resolution, and thus is subject to contamination from microcirculation and CSF partial volume. Here we proposed an orthogonal diffusion encoding gradient (ODEG) sequence to improve td–dependency measurements in human brain, by applying a pulse gradient orthogonal to the oscillating gradient to suppress the fast diffusion from microcirculation and free water. The results showed that td-dependency was significantly improved by the ODEG sequence in the hippocampus and cortical gray matter. |
| 3613 | Computer 29
|
Cardiac DTI with spiral readouts with third order k-space correction but without inner volume excitation |
| Lars Mueller1, Sam Coveney1, Maryam Afzali1,2, Irvin Teh1, Fabrizio Fasano3,4, Filip Szczepankiewicz5, Erica Dall’Armellina1, Christopher Nguyen6, Derek K Jones2, and Jurgen E Schneider1 | ||
1Leeds Institute of Cardiovascular and Metabolic Medicine, University of Leeds, Leeds, United Kingdom, 2Cardiff University Brain Research Imaging Centre (CUBRIC), Cardiff University, Cardiff, United Kingdom, 3Siemens Healthcare Ltd, Camberly, United Kingdom, 4Siemens Healthcare GmbH, Erlangen, Germany, 5Medical Radiation Physics, Clinical Sciences Lund, Lund University, Lund, Sweden, 6Cleveland Clinic, Cleveland, OH, United States |
||
Keywords: Pulse Sequence Design, Diffusion Tensor Imaging, Spiral Using spiral trajectories instead of EPI can reduce the echo time in diffusion weighted MRI. The feasibility of spiral trajectories for cardiac DTI has recently been demonstrated, but the larger object size (i.e torso vs. for example skull) may require an inner volume excitation to keep the readout at an useable duration. Here we examine the use of an undersampled spiral with SENSE reconstruction and standard slice selective pulses. We show that MD and FA derived from the slice selective and inner volume excitation yield comparable values with the MD being higher than the one measured with EPI. |
| 3614 | Computer 30
|
Feasibility of diffusion imaging using SMS-spiral acquisition with corrections on gradient waveform and field inhomogeneities |
| Zhe Wu1, Alexander Jaffray2, Lars Kasper1, and Kamil Uludag1,3 | ||
1Techna Institute, University Health Network, Toronto, ON, Canada, 2University of British Columbia, Vancouver, BC, Canada, 3Department of Medical Biophysics, University of Toronto, Toronto, ON, Canada |
||
Keywords: Pulse Sequence Design, Diffusion/other diffusion imaging techniques, Spiral Imaging, field inhomogeneity correction, gradient correction We propose a short-TE signal-to-noise ratio (SNR) enhanced diffusion imaging method using simultaneous multi-slice (SMS) accelerated spiral acquisition. The correction of field inhomogeneity and gradient waveforms are introduced in the reconstruction without any assistance of external hardware (e.g. field camera). Results showing the feasibility of this method on both phantom and human subjects, and the corrections of B0 and gradient waveforms are essential to improve the image quality. |
| 3615 | Computer 31
|
A deep learning method for diffusion tensor imaging using spherical harmonic representation |
| Yunwei Chen1 and Xinyuan Zhang1 | ||
1School of Biomedical Engineering, Southern Medical University, Guangzhou, China |
||
Keywords: Machine Learning/Artificial Intelligence, Diffusion Tensor Imaging Deep learning methods have been demonstrated state-of-the-art performance in Diffusion Magnetic Resonance Imaging (dMRI) denoising and parameter estimation. However, existing deep learning methods for dMRI are limited to the specific acquisition scheme. To solve the limitation, we proposed to use spherical harmonic coefficients as the deep learning network’s input. Our results have shown that the proposed method has a high performance in denoising and parameter estimation for DTI with a strong generalization ability. |
| 3616 | Computer 32
|
Gauge equivariant convolutional neural networks for diffusion MRI |
| Uzair Hussain1 and Ali Khan1,2,3 | ||
1Robarts Research Institute, Centre for Functional and Metabolic Mapping, London, ON, Canada, 2Department of Medical Biophysics, Western University, London, ON, Canada, 3Western Institute for Neuroscience, Western University, London, ON, Canada |
||
Keywords: Machine Learning/Artificial Intelligence, Diffusion Tensor Imaging One shortcoming of diffusion MRI (dMRI) is long scan times as numerous images have to be acquired to achieve a reliable angular resolution of diffusion gradient directions. In this work we introduce gauge equivariant convolutional neural network (gCNN) layers that overcome the challenges associated with the dMRI signal being acquired on a sphere instead of a rectangular grid. We apply this method to upsample angular resolution to predict diffusion tensor imaging (DTI) parameters from just six diffusion gradient directions. Additionally, gCNNs are able to train with fewer subjects and are general enough to be applied to other dMRI related problems. |
| 3617 | Computer 33
|
Accelerating High Resolution Diffusion Tensor Imaging Using Intra- and Inter-image Correlation |
| Zhongbiao Xu1, Rongli Zhang2, Wei Huang1, Junying Cheng3, Yingjie Mei4, Yihao Guo5, Hengwen Sun1, Yaohui Wang6, and Zhifeng Chen7 | ||
1Department of Radiotherapy, Guangdong Provincial People's Hospital, Guangzhou, China, 2Department of Imaging and Interventional radiology, The Chinese University of Hong Kong, HongKong, China, 3Department of MRI, The first Affiliated Hospital of Zhengzhou University, zhengzhou, China, 4School of Biomedical Engineering, Southern Medical University, guangzhou, China, 5Hainan General Hospital, hainan, China, 6Institute of Electrical Engineering, Chinese Academy of Sciences, beijing, China, 7Monash Biomedical Imaging, Department of Data Science and AI, Monash University, Clayton, Australia |
||
Keywords: Image Reconstruction, Diffusion Tensor Imaging DTI is challenged by the prolonged scan time in frontier studies and clinical applications. Parallel imaging can reduce the scan time, but with the SNR loss and the limitation of acceleration factor. In this work, we combined SENSE with self-supervised BM4D reconstruction model to improve image quality. The in vivo experiments demonstrated that the proposed method can obtain greatly improved image quality even with high acceleration factor of 5, compared to conventional methods. |
| 3618 | Computer 34
|
Evaluation of the impact of protocol settings on the variability of model estimations in multicenter diffusion MRI studies |
| Qiqi Tong1, Jinsong Li1,2, Jianhui Zhong3,4, and Hongjian He4,5 | ||
1Research Center for Healthcare Data Science, Zhejiang Lab, Hangzhou, China, 2College of Biomedical Engineering and Instrument Science, Zhejiang University, Hangzhou, China, 3Department of Imaging Sciences, University of Rochester, Rochester, NY, United States, 4Center for Brain Imaging Science and Technology, College of Biomedical Engineering and Instrument Science, Zhejiang University, Hangzhou, China, 5School of physics, Zhejiang University, Hangzhou, China |
||
Keywords: Data Analysis, Diffusion Tensor Imaging, multicenter A multicenter diffusion magnetic resonance imaging (dMRI) study was designed with different b-table schemes, non-diffusion b0 number, and varied echo time (TE) in two 3T scanners of different vendors. Global sensitivity analysis of 6 traveling subjects was conducted to evaluate the impact of imaging protocol setting on the observed cross-scan variability of diffusion metrics. |
| 3619 | Computer 35
|
Patch-CNN provides high-fidelity directional & scalar parameter estimation from 6-directional DWI robust to pathology unseen during training |
| Tobias Goodwin-Allcock1, Guglielmo Genovese2,3,4, Belen Zaid3,5, Stéphane Lehericy2,3, Charlotte Rosso3,5, Ting Gong1, Robert Gray6, Parashkev Nachev6, Marco Palombo7,8, and Hui Zhang1 | ||
1Department of Computer Science and Centre for Medical Image Computing, UCL, London, United Kingdom, 2Centre de NeuroImagerie de Recherche - CENIR, Paris Brain Institute - ICM, Paris, France, 3UMR S 1127, Inserm U 1127, CNRS UMR 7225, ICM, F-75013, Sorbonne Université, Paris, France, 4Center for Magnetic Resonance Research, Department of Radiology, University of Minnesota, Minneapolis, MN, United States, 5Paris Brain Institute - ICM, Centre de NeuroImagerie de Recherche - CENIR, Paris, France, 6University College London Queen Square Institute of Neurology, London, United Kingdom, 7Cardiff University Brain Research Imaging Centre (CUBRIC), School of Psychology, Cardiff University, Cardiff, United Kingdom, 8School of Computer Science and Informatics, Cardiff University, Cardiff, United Kingdom |
||
Keywords: Data Processing, Diffusion Tensor Imaging, Machine Learning This work evaluates the clinical viability of Patch-CNN for estimating diffusion MRI (dMRI) parameters from only 6 diffusion-weighted images (DWIs). Machine learning (ML) has been proposed to improve fitting from 6-directional DWIs. However, directional measures, e.g. primary fibre orientation, have only been estimated using CNNs. CNNs have not yet been validated on pathology that is not contained within the training dataset. As pathological diversity is difficult to capture in typical applications, ML methods are clinically viable only if they can generalise to unseen pathology. We show that Patch-CNN may generalise to unseen pathology and estimate directional measures. |
| 3620 | Computer 36
|
Unsupervised Susceptibility Artifact Correction in DTI Using a Deep Learning Forward-Distortion Model |
| Abdallah Zaid Alkilani1,2, Tolga Çukur1,2,3, and Emine Ulku Saritas1,2,3 | ||
1Deparment of Electrical and Electronics Engineering, Bilkent University, Ankara, Turkey, 2National Magnetic Resonance Research Center (UMRAM), Bilkent University, Ankara, Turkey, 3Neuroscience Graduate Program, Bilkent University, Ankara, Turkey |
||
Keywords: Data Analysis, Diffusion Tensor Imaging, Susceptibility, Machine Learning/Artificial Intelligence, Brain, Artifacts Diffusion weighted imaging (DWI) requires correction of susceptibility artifacts before conducting quantitative analyses. Correction is typically performed by acquiring DWI images in reversed phase-encode directions, which are used to estimate and correct for the effects of susceptibility-induced field. In this work, we propose a Forward-Distortion Network (FD-Net) for correcting susceptibility artifacts at multiple b-values. We evaluate the quality of the corrected DWI images and Diffusion Tensor Imaging (DTI) metrics, using FSL’s TOPUP as a reference classical method. In addition to rapid execution times, FD-Net exhibits high-fidelity performance for both DWI images and DTI metrics. |
| 3621 | Computer 37
|
High-resolution Diffusion Tensor Imaging with Deep Learning Reconstruction: Preliminary Results in Sub-cortical Fiber Tracking |
| Zhangxuan Hu1, Xiaocheng Wei1, Jie Lu2, and Bing Wu1 | ||
1GE Healthcare, Beijing, China, 2Department of Radiology and Nuclear Medicine, Xuanwu Hospital, Capital Medical University, Beijing, China |
||
Keywords: Machine Learning/Artificial Intelligence, Diffusion Tensor Imaging Diffusion tensor imaging (DTI) is a well-established tool for providing insights into brain structural connectivity and detecting brain microstructure. High spatial resolution diffusion MRI can provide improved resolvability of fibers with high-curvature (u-fibers). Segmented k-space methods such as Multiplexed sensitivity-encoding (MUSE) are often used to achieve high resolution diffusion images, however the shortcomings, such as prolonged scan time and low signal-noise-ratio (SNR), still exist. In this study, we aim to further improve the image quality of high-resolution diffusion images acquired with MUSE by combing with a deep learning based reconstruction method and thus to improve the sub-cortical fiber tracking accuracy. |
| 3622 | Computer 38
|
Segmented thick-slab 3D DWI with first and second order motion-compensated diffusion gradients |
| Jens Johansson1, Kerstin Lagerstrand2,3, Hanna Hebelka1,4, and Stephan E Maier1,5 | ||
1Radiology, Clinical Sciences, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden, 2Medical Radiation Sciences, Clinical Sciences, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden, 3Medical Physics and Biomedical Engineering, Sahlgrenska University Hospital, Gothenburg, Sweden, 4Radiology, Sahlgrenska University Hospital, Gothenburg, Sweden, 5Radiology, Brigham and women's hospital, Boston, MA, United States |
||
Keywords: Data Acquisition, Diffusion/other diffusion imaging techniques, Motion correction, 3D imaging Routine clinical diffusion imaging is generally performed with 2D echo planar sequences. A single thick-slab 3D approach could offer higher signal-to-noise ratio and better slice resolution, but has not been adopted due to the difficulty to avoid motion-induced phase errors that interfere with multi-shot spatial encoding. A new approach to enable 3D DWI is introduced here: rather than relying on navigator echoes for phase correction, first and second order motion-compensated diffusion encoding gradients are used to minimize phase variations at the source. |
| 3623 | Computer 39
|
Time-efficient Relaxation-Incorporated IVIM Diffusion Imaging |
| Kaibao Sun1, Guangyu Dan1,2, Qingfei Luo1, and Xiaohong Joe Zhou1,2,3 | ||
1Center for MR Research, University of Illinois at Chicago, Chicago, IL, United States, 2Department of Biomedical Engineering, University of Illinois at Chicago, Chicago, IL, United States, 3Departments of Radiology and Neurosurgery, University of Illinois at Chicago, Chicago, IL, United States |
||
Keywords: Data Acquisition, Diffusion/other diffusion imaging techniques Overestimation of perfusion volume fraction has been reported in conventional IVIM imaging. In a model known as extended T2-IVIM, compartmentalized relaxation times are incorporated to improve quantification of perfusion volume fraction, which requires acquisition of additional images at different TEs. A main challenge is that the scan time is lengthened, decreasing the efficiency while increasing vulnerability to motion. We herein introduce a novel sequence for time-efficient, relaxation-incorporated (TERI) IVIM imaging to address the aforementioned issues. TERI IVIM imaging uses multiple EPI readouts at different TEs in a single shot. The proposed technique has been demonstrated in the human brain. |
| 3624 | Computer 40
|
Super-resolution diffusion tensor imaging at 64 mT |
| Alix Plumley1, Mara Cercignani1, Álvaro Planchuelo-Gómez1,2, James Gholam1, and Derek K Jones1 | ||
1Cardiff University, Cardiff, United Kingdom, 2University of Valladolid, Valladolid, Spain |
||
Keywords: Software Tools, Diffusion Tensor Imaging, Low-field A super-resolution approach was used to create 2mm isotropic diffusion tensor images (DTI) from diffusion-weighted imaging data acquired on a low field, portable system. Mean diffusivity, fractional anisotropy and principal eigenvector orientation maps are shown. This work extends the very recently implemented capability of performing DTI on a 64mT system, and shows substantial improvement due to the increased through-plane resolution achieved with super-resolution. |
| 3760 | Computer 1
|
A deep neural network framework of few-shot learning with domain adaptation for automatic meniscus segmentation in 3D Fast Spin Echo MRI |
| Siyue Li1, Shutian Zhao1, Fan Xiao2, Dόnal G. Cahil3, Kevin Ki-Wai Ho4, Michael Tim-Yin Ong4, Queeie Chan5, James F Griffith3, Jin Hong6, and Weitian Chen1 | ||
1CU Lab for AI in Radiology (CLAIR), Department of Imaging and Interventional Radiology, The Chinese University of Hong Kong, Hong Kong, China, 2Department of Radiology, Shanghai Sixth People's Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, Shang hai, China, 3Department of Imaging and Interventional Radiology, The Chinese University of Hong Kong, Hong Kong, China, 4Department of Orthopaedics & Traumatology, The Chinese University of Hong Kong, Hong Kong, China, 5Philips Healthcare, Hong Kong, China, 6Guangdong Provincial People's Hospital, Guangdong Academy of Medical Sciences, Guang Zhou, China |
||
Keywords: Segmentation, Machine Learning/Artificial Intelligence Automatic meniscus segmentation is highly desirable for quantitative analysis of knee joint diseases. As three-dimensional Fast Spin Echo (3D FSE) is a promising MR (magnetic resonance) imaging technique to evaluate the tissues of the knee joint. In this study, we explore meniscal segmentation on 3D FSE MRI. Manually annotating 3D knee images is challenging since it is time-consuming and requires clinical expertise. In this study, we propose a domain adaption-based few-shot learning method for meniscal segmentation on 3D FSE images using only one annotated MRI data. We demonstrate that the proposed method outperformed the fully supervised segmentation model. |
| 3761 | Computer 2
|
Cardiac MR Image Segmentation in the Presence of Respiratory Motion Artifacts |
| Yasmina Al Khalil1, Sina Amirrajab1, Josien Pluim1, Marcel Breeuwer1,2, and Cian Scannell1 | ||
1Biomedical Engineering Department, Eindhoven University of Technology, Eindhoven, Netherlands, 2MR R&D - Clinical Science, Philips Healthcare, Best, Netherlands |
||
Keywords: Segmentation, Segmentation Object motion during the acquisition of magnetic resonance images can negatively impact image quality by introducing inconsistencies in the k-space data and in turn, blurring and ghosting artifacts on the acquired images. Such artifacts represent significant challenges in the clinical deployment of automated segmentation algorithms. In this work, we present an ensemble of approaches aimed at developing a robust and generalizable segmentation model, particularly tailored to handle the appearance of respiratory motion artifacts. We achieve this by introducing k-space based simulation for augmentation, as well as by reducing common errors across basal and apical slices with a region-focused segmentation approach.
|
| 3762 | Computer 3
|
A-Eye: Towards a large-scale MRI-based model of the complete eye |
| Jaime Barranco1,2, Hamza Kebiri1,2, Óscar Esteban3, Raphael Sznitman4, Oliver Stachs5, Sönke Langner6,7, Benedetta Franceschiello8,9,10, and Meritxell Bach Cuadra1,2,10 | ||
1CIBM Center for Biomedical Imaging, Lausanne, Switzerland, 2Radiology, Lausanne University Hospital and University of Lausanne, Lausanne, Switzerland, 3Lausanne University Hospital and University of Lausanne, Lausanne, Switzerland, 4ARTORG Center for Biomedical Engineering, University of Bern, Bern, Switzerland, 5Ophthalmology, Rostock University Medical Center, Rostock, Germany, 6Institute for Diagnostic and Interventional Radiology, Rostock University Medical Center, Rostock, Germany, 7Diagnostic Radiology and Neuroradiology, University of Greifswald, Greifswald, Germany, 8School of Engineering, Institute of Systems Engineering, HES-SO Valais-Wallis, Sion, Switzerland, 9The Sense Innovation and Research Center, Lausanne and Sion, Switzerland, 10These authors provided equal last-authorship contribution, Lausanne, Switzerland |
||
Keywords: Segmentation, Neuro, Eye, ophthalmology This work comparatively evaluates two approaches for the automated segmentation of eye structures from 3D T1-weighted MRI data of the whole human head (N=1210). Quantitative results on a validation sub-set with manual annotations provide accurate results for lens and globe and set the first median Dice (DSC) benchmarks for optic-nerve (0.91), muscles (0.58 to 0.76) and fat (0.67 and 0.75). The ability of our framework to automatically extract state-of-the-art measurements, such as the axial length, paves the way to accurately identify and compute new biomarkers of the eye via MRI. |
| 3763 | Computer 4
|
Automatic segmentation of myocardial 3D whole-heart T1 and T2 maps using a nnU-Net. |
| Carlos Velasco1, Roman Jakubicek2, Alina Hua1, Anastasia Fotaki1, René M. Botnar1,3, and Claudia Prieto1,3 | ||
1School of Biomedical Engineering and Imaging Sciences, King's College London, London, United Kingdom, 2Department of Biomedical Engineering, Brno University of Technology, Brno, Czech Republic, 3Institute for Biological and Medical Engineering, Pontificia Universidad Católica de Chile, Santiago, Chile |
||
Keywords: Segmentation, Myocardium The high amount of data obtained from a single 3D whole heart multiparametric scan (up to ~40 slices per parametric map) increases considerably the time required to segment and analyse the quantitative maps. Thus, an automated segmentation tool for these maps is desirable to perform this otherwise prohibitively laborious task. In this work, we leverage the potential of nnU-Net to perform fast, automated segmentation of 3D whole-heart simultaneous T1 and T2 maps and show its feasibility to predict segmentation masks with comparable quality while shortening the segmentation and analysis time by ~100x. |
| 3764 | Computer 5
|
Automated Segmentation of Knee Cartilage from Ultra-High Resolution 7 Tesla 3D bSSFP MRI Using Transfer Learning |
| Luxuan Guo1, Simran Kukran1, Krithika Balaji1, and Neal Bangerter1 | ||
1Imperial College London, London, United Kingdom |
||
Keywords: Segmentation, Cartilage, Osteoarthritis 7T 3D knee MRI shows great promise for the quantification of cartilage volume and thickness to assess osteoarthritis, but manual segmentation is time consuming. Segmented 7T 3D knee datasets to train machine learning techniques are limited. Here, we trained a network on a larger dataset of segmented 3T 3D knee MRI scans and used transfer learning to create an automatic segmentation network of 7T MRI knee cartilage using a small number of segmented 7T images. The resulting network constructed demonstrated vastly improved automatic segmentation of the knee cartilage at 7T compared to the network trained with limited 7T data only. |
| 3765 | Computer 6
|
Breast Segmentation of MRI Based on U-Net and Multi-Headed Self-Attention Mechanism |
| Hang Yu1, Yuru Guo1, Lizhi Xie2, Zhiheng Liu1, Zichuan Xie3, Chenyang Li1, and Suiping Zhou1 | ||
1School of Aerospace Science and Technology，Xidian university, xi'an, China, 2GE Healthcare, Beijing, China, 3Guangzhou institute of technology，Xidian University, Guangzhou, China |
||
Keywords: Segmentation, Breast In breast MRI, overall breast segmentation is a key step in performing breast cancer risk assessment.To achieve automatic and accurate breast segmentation in breast MR images, we propose a breast segmentation model based on U-Net and multi-head self-attention mechanism, which adds global information through multi-head self-attention module and changes the cascade structure in U-Net network to pixel-by-pixel summation.On the breast MRI dataset, the proposed model can achieve accurate, effective and fast breast segmentation with an average DSC and an average MIOU of 97.28 % and 92.01 %, respectively, which are 4.7 % and 5.56 % higher compared to U-Net, respectively. |
| 3766 | Computer 7
|
The effectiveness of the volume transfer constant (Ktrans) as an aid to regional segmentation of nasopharyngeal tumors |
| Junhui Huang1,2, Zhou Liu3, Liyan Zou3, Yingying Chen3, Zhanli Hu4, Dong Liang4, Xin Liu4, Hairong Zheng4, and Na Zhang4 | ||
1Paul C. Lauterbur Research Center for Biomedical Imaging, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, ShenZhen, China, 2Faculty of Information Engineering and Automation, Kunming University of Science and Technology, Kunming, China, 3Department of Radiology, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital & Shenzhen Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Shenzhen, China, 4Paul C. Lauterbur Research Center for Biomedical Imaging, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China |
||
Keywords: Segmentation, Machine Learning/Artificial Intelligence Accurate segmentation of nasopharyngeal tumor lesions from dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) facilitates subsequent diagnosis and treatment. However, current segmentation methods do not incorporate the pathological properties of the tumor. Therefore, this paper proposes a multimodal DCE-MRI segmentation method that uses the pharmacokinetic features of NPC, Ktrans, as modal information to assist nasopharyngeal tumor segmentation. We validated our method in several classical deep learning segmentation networks, and DCE-MRI with fused Ktrans eigenmodes had higher Dice coefficients than DCE-MRI with a single modality. The best segmentation results were obtained by this method on the ResUNet model (dice=74.39). |
| 3767 | Computer 8
|
Complementarity-aware multi-parametric MR image feature fusion for abdominal multi-organ segmentation |
| Cheng Li1,2, Yousuf Babiker M. Osman1,3, Weijian Huang1,3,4, Zhenzhen Xue1,2, Hua Han1,3, Hairong Zheng1, and Shanshan Wang1,2,4 | ||
1Paul C. Lauterbur Research Center for Biomedical Imaging, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China, 2Guangdong Provincial Key Laboratory of Artificial Intelligence in Medical Image Analysis and Application, Guangzhou, China, 3University of Chinese Academy of Sciences, Beijing, China, 4Peng Cheng Laboratory, Shenzhen, China |
||
Keywords: Segmentation, Body T1-weighted in-phase and opposed-phase gradient-echo imaging is a routine component in abdominal MR imaging. Organ segmentation with the acquired images plays an important role in identifying various diseases and making treatment plans. Despite the promising performance achieved by existing deep learning models, further investigation is still needed to effectively exploit the information provided by different imaging parameters. Here, we propose a complementarity-aware multi-parametric MR image feature fusion network to extract and fuse the information of paired in-phase and opposed-phase MR images for enhanced abdominal multi-organ segmentation. Extensive experiments are conducted, and better results are achieved when compared to existing methods. |
| 3768 | Computer 9
|
Extraction of the Utero-Placental and Fetal Vasculature using 2D Time-of-Flight Imaging |
| Karthikeyan Subramanian1, Pavan Kumar Jella1, Feifei Qu1, Tinnakorn Chaiworapongsa2,3, and Mark E Haacke1 | ||
1Department of Radiology, Wayne State University, Detroit, MI, United States, 2Perinatology Research Branch, NICHD/NIH/DHHS, Detroit, MI, United States, 3Department of Obstetrics and Gynecology, Wayne State University, Detroit, MI, United States |
||
Keywords: Segmentation, Fetus Mapping the vasculature, flow and tissue properties of the placenta and umbilical cord can serve as a means to study placental and fetal health. The goal of this work is to use a rapid, multi-echo, interleaved GRE sequence to minimize motion artifacts and cover the entire abdomen of the mother in a few minutes. To better map out the vasculature, we propose to both separate the arteries and veins in the umbilical cord using an R2* mapping thresholding technique and use vessel tracking to create 3D renderings. We have successfully done this in a series of 10 fetuses. |
| 3769 | Computer 10
|
Enhancing Vessel Continuity in Deep Learning based Segmentation using Maximum Intensity Projection as Loss |
| Soumick Chatterjee1,2,3, Karthikesh Varma Chintalapati1, Chethan Radhakrishna1, Sri Chandana Hudukula Ram Kumar1, Raviteja Sutrave1, Hendrik Mattern3, Oliver Speck3,4,5, and Andreas Nürnberger1,2,5 | ||
1Faculty of Computer Science, Otto von Guericke University Magdeburg, Magdeburg, Germany, 2Data and Knowledge Engineering Group, Otto von Guericke University Magdeburg, Magdeburg, Germany, 3Department of Biomedical Magnetic Resonance, Otto von Guericke University Magdeburg, Magdeburg, Germany, 4German Center for Neurodegenerative Disease, Magdeburg, Germany, 5Center for Behavioral Brain Sciences, Magdeburg, Germany |
||
Keywords: Segmentation, Blood vessels Vessel Segmentation with deep learning is a challenging task that involves not only learning high-level feature representations but also the spatial continuity of the features across dimensions. Semi-supervised patch-based approaches have been effective in identifying small vessels of 1-2 voxels in diameter but failed to maintain vessel continuity. This study focuses on improving the segmentation quality by considering the spatial correlation of the features using the maximum intensity projection (MIP) as an additional loss criterion. It was observed that the proposed method quantitatively improves the segmentation while also improving vessel continuity, as evident in the visual examinations of ROIs. |
| 3770 | Computer 11
|
Full spine 3D T2-weighted MRI with improved stitching for segmentation of the intrathecal space |
| Catarina Rua1, Mari Lambrechts1, Mark Tanner1, James Davies1, Ali Ghayoor2, Howard Dobson2, and Lino Becerra2 | ||
1Invicro LLC, A Konica Minolta Company, London, United Kingdom, 2Invicro LLC, A Konica Minolta Company, Needham, MA, United States |
||
Keywords: Segmentation, Neurofluids, CSF segmentation spine Computational fluid dynamics (CFD) has been used to model the behavior of cerebral-spinal-fluid (CSF) flow along the spine in research of intrathecal drug delivery. However, the model geometry highly impacts the CFD-based prediction of CSF flow. Enhanced images of the spinal CSF can be obtained with T2-weighted MRI, typically collected in 3-4 stations. Here we present a method to stitch spine images for CSF segmentation and compare it to the scanner’s stitching technique. Our offline approach appears more robust to movement and field-of-view adjustments while scanning and presents acceptable CSF signal intensity homogeneity across the spine for tissue segmentation. |
| 3771 | Computer 12
|
Deep learning-based auto-segmentation of neck nodal metastases on longitudinal MR images using self-distilled masked image transformer |
| Jue Jiang1, Ramesh Paudyal1, Bill H. Diplas2, James Han2, Nadeem Riaz2, Vaios Hatzoglou 3, Nancy Lee2, Joseph Deasy 1, Amita Shukla-Dave 1,3, and Harini Veeraraghavan 1 | ||
1Medical Physics, Memorial Sloan Kettering Cancer Center, New York, NY, United States, 2Radiation Oncology, Memorial Sloan Kettering Cancer Center, New York, NY, United States, 3Radiology, Memorial Sloan Kettering Cancer Center, New York, NY, United States |
||
Keywords: Segmentation, Cancer Manual segmentation of normal and tumor tissues on MRI is a traditional approach that is still used, but it is a very challenging and time-consuming method requiring a high level of precision and has shown inter-reader contouring variability. Therefore, semi- or fully automated segmentation algorithms are essential to segment tumors such as neck nodal metastases. The present study aimed to apply the previously developed deep learning-based self-distilled masked image transformer method for auto-segmenting neck nodal metastases on longitudinal T2-weighted MR images. |
| 3772 | Computer 13
|
Assessing the Impact of Upstream Reconstruction Models on Downstream Image Analysis: A Workflow-Centric Evaluation |
| Ben Viggiano1, Aashna Desai2, Elka Rubin3, Andrew Schmidt3, Robert Boutin3, Kathryn J Stevens3, Garry E Gold3, Christopher Ré4, Akshay S Chaudhari1,3, and Arjun D Desai3,5 | ||
1Biomedical Data Science, Stanford University, Stanford, CA, United States, 2Department of Neuroscience, University of California Berkeley, Berkeley, CA, United States, 3Department of Radiology, Stanford University, Stanford, CA, United States, 4Department of Computer Science, Stanford University, Stanford, CA, United States, 5Department of Electrical Engineering, Stanford University, Stanford, CA, United States |
||
Keywords: Machine Learning/Artificial Intelligence, Image Reconstruction, Segmentation, Classification Deep learning (DL) techniques have shown promise for both reconstruction and image analysis stages of MRI workflows. However, traditional benchmarking methods evaluate each stage separately. As a result, the impact of reconstruction on downstream image analysis tasks and biomarker quantification remains unknown. In this study, we explore how changing aspects of upstream reconstruction affects the downstream analysis. We find that insights from evaluating reconstruction models as a component of a broader end-to-end workflow do not correlate with conventional, task-specific image quality metrics. We use these findings to motivate the discussion of evaluating DL methods at the workflow level. |
| 3773 | Computer 14
|
Age Related Changes of Organs and Muscles Using a Deep Learning Based Whole Body Segmentation Technique Applied to a Healthy Adult Population |
| Ahmed Gouda1, Saqib Basar1, Yosef Chodakiewitz2, Rajpaul Attariwala1, Sean London2, and Sam Hashemi1 | ||
1Voxelwise Imaging Technology Inc., Vancouver, BC, Canada, 2Prenuvo, Vancouver, BC, Canada |
||
Keywords: Data Analysis, Aging, Volumetric Analysis, Organs, Visceral Fat, Skeletal Muscles One of the vital indicators of normal functioning physiology and anatomy, is the volume and size of body organs and muscles. Overtime due to aging and chronic illnesses, the volume of organs and muscles decreases, and quantifying this reduction would help us in predicting the rate of decline, and making efforts to improve this rate. Artificial intelligence 3D segmentation techniques allow us to measure the average age-related volume loss over a large population. They provide statistical norms which help radiologists to identify abnormal volume changes, and establish a normal age related rate of volumetric decline. |
| 3774 | Computer 15
|
An Adaptive Weighted Active Contour Segmentation Model for 3T/5T MRI from the Same Person |
| Zhenxing Huang1, Mengxiao Geng1, Liyun Zheng2, Yongming Dai3, Na Zhang1, Dong Liang1, Hairong Zheng1, and Zhanli Hu1 | ||
1Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China, 2Shenzhen United Imaging Research Institute of Innovative Medical Equipment, Shenzhen, China, 3Central Reasearch Institute, United Imaging Healthcare, Shanghai, China |
||
Keywords: Data Analysis, Segmentation, 3T/5T MRI Image segmentation is a complex and core technique in the medical image domain. However, low-quality images, such as images with weak edges, may bring considerable challenges for radiologists. In this paper, we propose an adaptive weighted curvature-based active contour model by coupling heat kernel convolution and adaptively weighted high-order total variation to improve diagnosis effectiveness. The numerical experimental results on 3T/5T MRI datasets demonstrate that the proposed model is quite efficient and robust compared with several traditional segmentation methods, which would exert great value in quantitative image evaluation of MRI diagnosis for the same person. |
| 3775 | Computer 16
|
HYBRID MULTI-LEVEL GRAPH NEURAL NETWORK FOR CARDIAC MAGNETIC RESONANCE SEGMENTATION |
| Xiaodi Li1, Peng Li1, and Yue Hu1 | ||
1Harbin Institute of Technology, Harbin, China |
||
Keywords: Data Processing, Segmentation Due to the inherent locality of convolutional operations, convolution neural network (CNN) often exhibits limitations in explicitly modeling long-distance dependencies. In this paper, we propose a novel hybrid multi-level graph neural (HMGN) network that combines the CNN and graph neural network to capture both local and non-local image features at multiple scales. With the proposed patch graph attention module, the HMGN network can capture image features over a large receptive field, resulting in more accurate segmentation of cardiac structures. Experiments on two public datasets show the proposed method obtains improved segmentation performance over the state-of-the-art methods. |
| 3776 | Computer 17
|
Deep Learning-based Prostate Lesion Segmentation and Classification Using Haralick Texture Maps on MR images |
| Dang Bich Thuy Le1, Ram Narayanan1, Meredith Sadinski1, Aleksandar Nacev1, and Srirama Venkataraman1 | ||
1Research and Design, Promaxo Inc, Oakland, CA, United States |
||
Keywords: Radiomics, Cancer, Prostate By quantifying pixel relationships from frequencies of local signal intensity spatial variations, Haralick texture features have shown promise for prostate cancer detection. In this study, axial, T2-weighted MR images combined with extracted Haralick texture feature maps were used in a deep learning framework to identify lesion locations and predict Gleason Grade. Results demonstrate potential of Haralick texture features to segment and classify prostate lesions with AUC/sensitivity/specificity of 0.87/0.923/0.776 on patient-level evaluation. |
| 3777 | Computer 18
|
Lifelong collaborative learning improves the performance of complex muscle MR image segmentation tasks |
| Francesco Santini1,2, Jakob Wasserthal2, Abramo Agosti3, Xeni Deligianni1,2, Kevin R Keene4, Hermien E Kan5, Stefan Sommer6,7,8, Christoph Stuprich9, Fengdan Wang10, Claudia Weidensteiner1,11, Giulia Manco12, Valentina Mazzoli13, Arjun Desai14, and Anna Pichiecchio12,15 | ||
1Basel Muscle MRI, Department of Biomedical Engineering, University of Basel, Basel, Switzerland, 2Research Coordination Team, Department of Radiology, University Hospital Basel, Basel, Switzerland, 3Department of Mathematics, University of Pavia, Pavia, Italy, 4Department of Neurology, Leiden University Medical Center, Leiden, Netherlands, 5C.J. Gorter MRI Centre, Department of Radiology, Leiden University Medical Center, Leiden, Netherlands, 6Siemens Healthineers International AG, Zurich, Switzerland, 7Swiss Center for Musculoskeletal Imaging (SCMI), Balgrist Campus, Zurich, Switzerland, 8Advanced Clinical Imaging Technology (ACIT), Siemens Healthineers International AG, Lausanne, Switzerland, 9University Hospital Erlangen, Erlangen, Germany, 10Peking Union Medical College, Beijing, China, 11Radiological Physics, Department of Radiology, University Hospital Basel, Basel, Switzerland, 12Advanced Imaging and Radiomics Center, IRCCS Mondino Foundation, Pavia, Italy, 13Department of Radiology, Stanford University, Stanford, CA, United States, 14Departments of Electrical Engineering & Radiology, Stanford University, Stanford, CA, United States, 15Department of Brain and Behavioural Sciences, University of Pavia, Pavia, Italy |
||
Keywords: Software Tools, Machine Learning/Artificial Intelligence An open-source, federated-learning-based segmentation software termed Dafne (Deep Anatomical Federated Network) is presented. This software continuously adapts the deep learning models used for the segmentation (currently for the muscles of the leg and thigh) based on the input of the users, who are in multiple institutions. This software was validated through data usage statistics of more than 50 users and through a retrospective study on 38 datasets of patients with suspected myositis, showing that the continuous learning approach is able to improve and generalize the performance of the original models. |
| 3778 | Computer 19
|
Deep Learning Based 3D Whole Liver and Spleen Segmentation for Quantitatively Reproducible Liver Fat and Iron Deposition Grading |
| Ahmed Gouda1, Saqib Basar1, Yosef Chodakiewitz2, Rajpaul Attariwala1, Sean London2, and Sam Hashemi1 | ||
1Voxelwise Imaging Technology Inc., Vancouver, BC, Canada, 2Prenuvo, Vancouver, BC, Canada |
||
Keywords: Liver, Fat, Hepatic Steatosis Quantification, Iron Deposition Detection, Dixon, FSF, Dual-echo MRI Over the years, Magnetic Resonance Imaging (MRI) has become the optimal noninvasive method to quantify liver steatosis and to detect hepatic iron deposition. The conventional manual sampling technique of liver fat quantification at multiple regions is complex, inefficient and a time-consuming process. In addition, it may produce varying results for the heterogeneous fat deposition, which is prone to radiologist’s subjectivity. In this research, we propose a fully automated artificial intelligence (AI) based method for hepatic steatosis quantification and hepatic iron deposition detection using whole liver and spleen volume segmentation. |
| 3779 | Computer 20
|
Automatic enhancing objects detection and segmentation in breast DCE-MRI |
| Adam DESCARPENTRIES1, Florence FERET2, and Julien ROUYER1 | ||
1Research and Innovation Department, Olea Medical, La Ciotat, France, 2Clinical solution department, Olea Medical, La Ciotat, France |
||
Keywords: Breast, Cancer, CAD The aim of our work is to detect any enhancing object of interest for reporting purposes. A deep learning approach combined with a multi-constructor and multi-centric database enabled to initiate the development of a versatile tool in line with clinical real life. The detection problem was addressed using a two-stage three-dimensional cascaded U-Net architecture. A total of 610 single-breast images were used for the model development. Results present interesting score in term of Dice similarity index (0.83) which agree well with the recent literature. Discussion section focuses on the potential benefit in the use of a recently reported loss function. |
| 3780 | Computer 21
|
Q-space trajectory imaging with positivity constraints: a machine learning approach |
| Deneb Boito1,2 and Evren Özarslan1,2 | ||
1Biomedical Engineering, Linköping University, Linköping, Sweden, 2Center for Medical Image Science and Visualization, Linköping University, Linköping, Sweden |
||
Keywords: Data Processing, Diffusion/other diffusion imaging techniques Q-space trajectory imaging (QTI) is a diffusion MRI framework which access features of the microstructure through the statistical moments of the diffusion tensor distribution. To overcome unreliable estimates obtained with standard fitting methods, a constrained estimation framework named QTI+ was recently proposed. Constrained optimization however typically requires sophisticated fitting routines which introduce a heavy computational burden. In this work we thus explore the possibility of speeding up the QTI parameter estimation, while retaining strict positivity constraints, using artificial intelligence. Results are shown on synthetic datasets as well as for healthy subjects and data from brain tumor patients. |
| 3781 | Computer 22
|
Using ‘P-scores’: a novel percentile-based normalization method to accurately assess individual deviation in heavily skewed neuroimaging data |
| Rakibul Hafiz1, Amritha Nayak1,2, M. Okan Irfanoglu1, Leighton Chan3, and Carlo Pierpaoli1 | ||
1National Institute of Biomedical Imaging and Bioengineering (NIBIB), National Institutes of Health (NIH), Bethesda, MD, United States, 2Henry Jackson Foundation for Advancement of Military Medicine, Bethesda, MD, United States, 3Rehabilitation Medicine Department, National Institutes of Health (NIH), Bethesda, MD, United States |
||
Keywords: Data Analysis, Diffusion/other diffusion imaging techniques, Quantitative Medical Imaging We propose a novel quantity to correctly assess the extent individuals deviate from the median of a heavy-tailed distribution. We compute a percentile-based score, we call ‘P-score’, which normalizes the deviation of an individual from the sample median by incorporating the individual’s position in the left/right tail of the sample distribution and the corresponding length between the sample median and the 5th/95th percentile edge values of the sample distribution, respectively. We demonstrate the skewness present in diffusion MRI data and the bias introduced when Z-scores are used and further show the control of this bias using the proposed ‘P-scores’ approach. |
| 3782 | Computer 23
|
Fast mean kurtosis measurements using navigator-free 3D multi-slab DWI |
| Chu-Yu Lee1 and Merry Mani1 | ||
1Department of Radiology, The University of Iowa, Iowa City, IA, United States |
||
Keywords: Data Acquisition, Diffusion/other diffusion imaging techniques, Kurtosis; Multi-slab; 3D The fast mean kurtosis measurements using 2D DWI has reduced the scan time substantially but remains subject to low SNR, particularly at b-value = 2500 s/mm2. A larger voxel (≥ 2 mm3) is normally used to increase SNR. This study presents the navigator-free 3D multi-slab DWI for fast mean kurtosis measurements with a resolution of 1.5 mm3. By using a short TR of 2.5 sec, the work demonstrates the promise of using the navigator-free 3D multi-slab DWI for whole-brain fast mean kurtosis measurements with an improved SNR efficiency. |
| 3783 | Computer 24
|
tDKI-Net: a joint q-t space learning network for diffusion-time-dependent kurtosis imaging and Karger’s model fitting |
| Tianshu Zheng1, Ruicheng Ba1, Xiaoli Wang2, Xizhen Wang3, Chuyang Ye4, and Dan Wu1 | ||
1Department of Biomedical Engineering, College of Biomedical Engineering & Instrument Science, Zhejiang University, Hangzhou, Zhejiang, China, Hangzhou, China, 2School of Medical Imaging, Weifang Medical University, Weifang, Shandong, China, Weifang, China, 3Medical imaging center, Affiliated hospital of Weifang Medical University, Weifang, Shandong, China, Weifang, China, 4School of Integrated Circuits and Electronics, Beijing Institute of Technology, Beijing, China, Beijing, China |
||
Keywords: Machine Learning/Artificial Intelligence, Diffusion/other diffusion imaging techniques Time-dependent diffusion magnetic resonance imaging (TDDMRI) is useful for non-invasive characterization of tissue microstructure. The TDDMRI models require both densely sampled q-space (b-value and diffusion direction) and t-space (diffusion time) data for microstructural fitting, leading to very time-consuming acquisition protocols. In this work, we presented a tDKI-Net to estimate diffusion kurtosis at multiple diffusion times, which was fed into the Karger model to obtain K0 and transmembrane exchange time, using downsampled q-space and t-space data. We tested the proposed network in the normal rat brains, as well as those in a rat model of Middle Cerebral Artery Occlusion. |
| 3784 | Computer 25
|
Simultaneous Multi-slice Single-shot Spiral Acquisitions for Accelerated Diffusion-weighted Imaging |
| Guangqi Li1, Yuancheng Jiang1, and Hua Guo1 | ||
1Center for Biomedical Imaging Research, Department of Biomedical Engineering, School of Medicine, Tsinghua University, Beijing, China |
||
Keywords: Data Acquisition, Diffusion Tensor Imaging Single-shot acquisition techniques are commonly used to acquire diffusion-weighted images due to their high sampling efficiency. Single-shot uniform-density spiral sampling combined with in-plane under-sampling has been used to achieve diffusion imaging. In this study, we investigated simultaneous multi-slice (SMS) -accelerated single-shot spiral imaging (SMS-SSH-Spiral) to further improve the scan efficiency of DWI. The in vivo results demonstrate the feasibility of SMS-SSH-Spiral acquisitions for diffusion tensor imaging with 2-fold slice acceleration and 3-fold in-plane acceleration. Diffusion data with 1.5mm isotropic resolution is acquired using our multi-band single-shot acquisition strategy. |
| 3785 | Computer 26
|
Hybrid-space reconstruction for simultaneous multi-slab DWI with blipped-CAIPI |
| Jieying Zhang1, Simin Liu1, and Hua Guo1 | ||
1Center for Biomedical Imaging Research, Department of Biomedical Engineering, School of Medicine, Tsinghua University, Beijing, China |
||
Keywords: Image Reconstruction, Diffusion Tensor Imaging 3D simultaneous multi-slab imaging (SMSlab) can achieve high-resolution DWI with high SNR efficiency. Multi-band acceleration can also achieve less SNR reduction. Recently, we integrated SMSlab DWI with blipped-CAIPI gradients to reduce the g-factor penalty and proposed a 4D k-space framework (kx-ky-kz-km) to model the signal encoding, with km representing the multi-band encoding. Because the blipped-CAIPI gradients are applied along the slice direction, they introduce kz deviations from the nominal k-space location. This study proposed a hybrid-space reconstruction algorithm, REACH, to solve the phase interferences introduced by the blipped-CAIPI gradients. |
| 3786 | Computer 27
|
Should We Harmonize or Denoise in Diffusion MRI? |
| Benjamin Ades-Aron1, Santiago Coelho1, Jelle Veraart1, Timothy M. Shepherd1, Dmitry S. Novikov1, and Els Fieremans1 | ||
1Radiology, Center for biomedical imaging, NYU Grossman School of Medicine, New York, NY, United States |
||
Keywords: Data Processing, Diffusion/other diffusion imaging techniques, Harmonization Denoising To address the ongoing reproducibility crisis in quantitative diffusion MRI (dMRI), efforts are underway to harmonize and improve precision of diffusion parameter estimation. Using inter- and intra-scanner test-retest higher order dMRI, we compare the reproducibility of two popular harmonization methods, ComBat and linear-RISH, to that of denoising using MPPCA on complex-valued dMRI. We find that denoising combined with harmonization improves voxel-wise test-retest ICC by up to 60% compared to harmonization alone. Using dMRI at different voxel sizes, we find that denoising reduces the bias due to varying noise floors more accurately than harmonization. Denoising appears essential to harmonize dMRI datasets. |
| 3787 | Computer 28
|
DIMOND：DIffusion Model OptimizatioN with Deep learning |
| Zihan Li1, Berkin Bilgic2,3, Hong-Hsi Lee2,3, Kui Ying4, Hongen Liao1, Susie Huang2,3, and Qiyuan Tian2,3 | ||
1Department of biomedical engineering, Tsinghua University, Beijing, China, 2Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Charlestown, MA, United States, 3Harvard Medical School, Boston, MA, United States, 4Department of Engineering Physics, Tsinghua University, Beijing, China |
||
Keywords: Machine Learning/Artificial Intelligence, Diffusion/other diffusion imaging techniques The accurate estimation of diffusion model parameter values using non-linear optimization is time-consuming. Supervised learning methods using neural networks (NNs) are faster and more accurate but require external ground-truth data for training. A unified and self-supervised learning-based diffusion model estimation method DIMOND is proposed. DIMOND maps diffusion data to model parameter values using NNs, synthesizes the input data from the predictions using the forward model, and minimizes the difference between the raw and synthetic data. DIMOND outperforms conventional ordinary least square regression (OLS) and has a high potential to improve and accelerate data fitting for more complicated diffusion models. |
| 3788 | Computer 29
|
Learning diffusion MRI fiber orientation distribution functions of the developing human brain |
| Hamza Kebiri1,2, Ali Gholipour2, Davood Karimi2, and Meritxell Bach Cuadra1 | ||
1Center for Biomedical Imaging & Lausanne University Hospital, Lausanne, Switzerland, 2Harvard Medical School & Boston Children's Hospital, Boston, MA, United States |
||
Keywords: Machine Learning/Artificial Intelligence, Diffusion/other diffusion imaging techniques, Fiber Orientation Distribution functions Diffusion MRI of fetal and newborn brains is constrained by short scanning time allowing only a small number of diffusion measurements to be acquired. Methods going beyond the diffusion tensor model require multi-shell and multiple gradient directions in order to unveil more accurate white matter properties. We propose a learning based framework to reconstruct fiber orientation distribution functions from only six diffusion measurements by leveraging existing high-quality datasets. Quantitative evaluation on 15 newborn subjects show that our framework achieves competitive results with state-of-the-art methods. Qualitative evaluation on a fetus shows the model ability to translate to this challenging population. |
| 3789 | Computer 30
|
Spatiotemporal signal drift in diffusion MRI of the brain and ways to correct for it: effects on estimates of ADC and IVIM f |
| Oscar Jalnefjord1,2, Amina Warsame1, and Louise Rosenqvist1 | ||
1Department of Medical Sciences, Institute of Clinical Sciences, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden, 2Department of Medical Physics and Biomedical Engineering, Sahlgrenska University Hospital, Gothenburg, Sweden |
||
Keywords: Data Processing, Diffusion/other diffusion imaging techniques Signal drift has been identified as a confounding factor in diffusion MRI (dMRI), causing increases variance and potential bias in derived parameter maps. In this work, we show that temporal signal drift is spatially dependent in human brain images for dMRI, thus calling for spatiotemporal corrections. We also show that signal drift can have a substantial effect on short-term repeatability of ADC and IVIM f obtained from data acquired with a protocol ordered by b-value as is commonly done in clinical practice. |
| 3790 | Computer 31
|
Chronic liver disease: The role of multiple diffusion-weighted models using the Bayesian shrinkage method for liver fibrosis assessment |
| Jiqing Huang1, Benjamin Leporq1, Olivier Beuf1, and Hélène Ratiney1 | ||
1Univ Lyon, INSA Lyon, CNRS, Inserm, CREATIS UMR 5220, U1206, F-69621, Lyon, Villeurbanne, France |
||
Keywords: Data Processing, Diffusion/other diffusion imaging techniques Liver fibrosis is one of the leading features in chronic liver disease (CLD) since it conditions the prognosis and guides the treatment strategy. In this work, estimated parameters from various diffusion-weighted MRI models fitted by the Bayesian method were analyzed for the relationship with liver fibrosis through spearman’s correlation and t-test. Four parameters (Ds, σ, D*_F, Dapp) were selected for fibrosis classification and achieved the best result based on the decision tree. Our result suggested that the statistical model and a hybrid IVIM-DKI model are promising models and confirmed the confounding effect of fat for diffusivity to assess liver fibrosis. |
| 3791 | Computer 32
|
Recovering high quality FODs from reduced number of diffusion weighted images using a model-driven deep learning architecture |
| Joseph Bartlett1,2,3, Catherine Davey1,2, Leigh Johnston1,2, and Jinming Duan3,4 | ||
1Department of Biomedical Engineering, University of Melbourne, Melbourne, Australia, 2Melbourne Brain Centre Imaging Unit, University of Melbourne, Melbourne, Australia, 3School of Computer Science, University of Birmingham, Birmingham, United Kingdom, 4Alan Turing Institute, London, United Kingdom |
||
Keywords: Machine Learning/Artificial Intelligence, Diffusion/other diffusion imaging techniques SDNet utilises the learning ability of deep neural networks with the robustness of model-based approaches to produce high quality fibre orientation distributions (FODs) from a reduced set of multi-shell diffusion weighted images (DWI). The cascaded architecture, with data consistency layers throughout, makes use of model based prior knowledge and spatial correlations within the DWI signal to achieve state-of-the-art performance in both sum of squared errors and angular correlation coefficient. Our model also shows competitive results with respect to apparent fibre density error and peak amplitude error over a range of regions of interest. |
| 3792 | Computer 33
|
3D Isotropic and high resolution whole brain DWI with single slab and single shot EPI utilizing motion compensated diffusion gradients |
| Zhigang Wu1, Yajing Zhang2, Gilbert Guillaume3, Wengu Su4, Yan Zhao5, and Jiazheng Wang6 | ||
1Philips Healthcare, Shenzhen, Ltd., Shenzhen, China, 2Philips Health Technology, Suzhou, China, 3MR Clinical Science, Philips Healthcare, Mississauga, ON, Canada, 4BU MR Application, Philips Health Technology, Suzhou, China, 5BU MR R&D, Philips Health Technology, Suzhou, China, 6Philips Healthcare, Beijing, China |
||
Keywords: Pulse Sequence Design, Diffusion/other diffusion imaging techniques, Pulse Sequence Design, Diffusion/other diffusion imaging techniques, Diffusion, 3D 3D Diffusion imaging (3D DWI) has showed great potential in probing tissue microstructure and brain structural connectivity. However, motion-induced phase errors introduced by diffusion gradients will cause severe artifacts in 3D DWI. Multi-slab can be used to overcome this limitation, but it will introduce slab boundary artifacts. We propose a method which utilizes motion-compensated diffusion gradients for 3D DWI to mitigate the phase error between shots. Results from in vivo data demonstrate the proposed method can improve image quality and realize an isotropic high resolution and whole brain 3D DWI in a single slab. |
| 3793 | Computer 34
|
Sub-millimeter Diffusion Tensor Imaging using Single-Shot Spiral Acquisitions with a Large Acceleration Factor |
| Guangqi Li1, Xinyu Ye1, Yuan Lian1, Yajing Zhang2, and Hua Guo1 | ||
1Center for Biomedical Imaging Research, Department of Biomedical Engineering, School of Medicine, Tsinghua University, Beijing, China, 2MR Clinical Science, Philips Health Technology (China), Beijing, China |
||
Keywords: Image Reconstruction, Diffusion Tensor Imaging Single-shot spiral acquisitions allow shorter TE, thus provide higher SNR compared to EPI acquisitions for DWI. However, spiral acquisitions are sensitive to field inhomogeneity. Parallel imaging techniques can be used to alleviate static B0 off-resonance effects. In this study, single-shot spiral acquisitions with a large acceleration factor of 5 or 6 were used to achieve sub-millimeter diffusion tensor imaging at 3T. The in vivo results demonstrate that the single-shot spiral sampling strategy can be adopted to achieve whole-brain diffusion tensor imaging with an in-plane resolution of 0.77×0.77mm2. |
| 3794 | Computer 35
|
Higher-order diffusion MRI data acquired in clinical settings: what are the pitfalls, and how to correct them? |
| Jenny Chen1, Benjamin Ades-Aron1, Saurabh Maithani1, Yvonne Lui1, Dmitry S. Novikov1, Jelle Veraart1, and Els Fieremans1 | ||
1New York University Grossman School of Medicine, New York, NY, United States |
||
Keywords: Data Acquisition, Diffusion/other diffusion imaging techniques, Clinical Translation, DKI Multi-shell diffusion MRI (dMRI) opens up the opportunity for microstructure mapping, but is not routinely acquired in clinics. This study retrospectively analyzed a large diffusion kurtosis imaging (DKI) dataset (N=7984) acquired as multi-series in the clinic and proposes a workflow to inspect the data before preprocessing. Automated analysis of dicom headers shows about 36% with incomplete acquisition, likely due to time constraints and patient discomfort, and 32% with image discrepancies. Finally, we show that signal variations between series occur and need to be corrected for in preprocessing. Our results indicate higher-order dMRI in the clinical setting is feasible. |
| 3795 | Computer 36
|
Iterative model-based Rician bias correction and its application to denoising in diffusion MRI |
| J-Donald Tournier1,2, Ben Jeurissen3,4, and Daan Christiaens5,6 | ||
1Biomedical Engineering, School of Biomedical Engineering and Imaging Sciences, King's College London, London, United Kingdom, 2Centre for the Developing Brain, School of Biomedical Engineering and Imaging Sciences, King's College London, London, United Kingdom, 3imec-Vision Lab, Department of Physics, University of Antwerp, Antwerp, Belgium, 4Lab for Equilibrium Investigations and Aerospace, Department of Physics, University of Antwerp, Antwerp, Belgium, 5Medical Imaging Research Center, KU Leuven, Leuven, Belgium, 6Department of Electrical Engineering, ESAT/PSI, KU Leuven, Leuven, Belgium |
||
Keywords: Data Analysis, Diffusion/other diffusion imaging techniques Rician bias is a long-standing problem in diffusion MRI. We propose a simple iterative strategy to mitigate the problem when a voxel-wise model of the signal is available, and demonstrate its efficacy when fitting spherical harmonics and when included in a denoising strategy. |
| 3796 | Computer 37
|
Noise and distortion reduction for Reduced FOV diffusion utilizing Compressed SENSE framework for single shot EPI (EPICS) |
| Zhigang Wu1, Yajing Zhang2, Wengu Su3, Masami Yoneyama4, Peng Sun5, Jing Zhang6, Yan Zhao7, and Jiazheng Wang5 | ||
1Philips Healthcare, Shenzhen, Ltd., Shenzhen, China, 2Philips Health Technology, Suzhou, China, 3BU MR Application, Philips Health Technology, Suzhou, China, 4BU MR Clinical Science, Philips Healthcare, Tokyo, Japan, 5Philips Healthcare, Beijing, China, 6MR Clinical Application, Philips Healthcare, Beijing, China, 7BU MR R&D, Philips Health Technology, Suzhou, China |
||
Keywords: Data Acquisition, Diffusion/other diffusion imaging techniques, Reduced FOV, Compressed SENSE DWI is very important for MRI examination, but it has limited resolution due to distortion, blurring, and signal loss caused by B0 inhomogeneity. Reduced FOV imaging could decrease these impacts. However, due to the coil geometry penalty, it’s hard to combine it with parallel imaging to further improve the image quality, it will suffer from noise breakthrough issues and unfolding artifacts. We propose a framework that combines reduced FOV imaging, Compressed SENSE framework simultaneously to overcome these issues. This framework allows a new solution for reduced FOV based diffusion imaging with high resolution, low distortion, and without noise breakthrough issue. |
| 3797 | Computer 38
|
Value of Reversed Polarity Gradients and Multiplexed Sensitivity Encoding in improving diffusion weighted imaging quality of uterus |
| Wenjuan Wang1, Wenjing Zhao1, and Dmytro Pylypenko2 | ||
1Department of Radiology, Weifang People’s Hospital, Weifang, China, 2MR Research China, GE Healthcare, Beijing, China |
||
Keywords: Artifacts, Diffusion/other diffusion imaging techniques, multiplexed sensitivity encoding, reversed polarity gradients The aim of this study is to investigate the application of multiplexed sensitivity encoding diffusion weighted imaging with reversed polarity gradients in improving the image quality of uterine tumors. 10 patients with uterine lesions in our hospital were enrolled. Conventional SS-EPI DWI, MUSE DWI, and RPG-MUSE DWI sequences were employed. The subjective image quality assessment and objective data measurement of uterine lesions scanned by different DWI sequence was evaluated by three radiologists using double-blind method. The results of this study showed that RPG-MUSE DWI can greatly improve the image quality of the uterus. |
| 3798 | Computer 39
|
Denoising Enables Faster 1 mm Isotropic Diffusion Tensor Imaging of the Human Hippocampus at 3T |
| Pablo Stack-Sanchez1, Donald Gross2, and Christian Beaulieu1 | ||
1Biomedical Engineering, University of Alberta, Edmonton, AB, Canada, 2Neurology, University of Alberta, Edmonton, AB, Canada |
||
Keywords: Data Processing, Diffusion Tensor Imaging A hippocampus-focused, 1 mm isotropic, diffusion tensor imaging 5.5 minutes 3T protocol has yielded high-quality diffusion images and maps. Post-processing with denoising can potentially reduce the scan time considerably. Compared to the previously used 10 direction/10 average protocol, denoising enabled the use of 4 averages for a scan time of only 2.2 minutes to yield acceptable 1 mm isotropic diffusion image quality and mean diffusivity maps (MD) of the whole hippocampus in healthy controls. This rapid protocol was then able to delineate focal and small hippocampal lesions with elevated MD in temporal lobe epilepsy patients. |
| 3799 | Computer 40
|
Application of machine learning with Tract-Based Spatial Statistics in the diagnosis of pediatric autism |
| Xiongpeng He1, Xiaoan Zhang1, Xin Zhao1, Yongbin Sun2, Pengfei Geng1, and Kaiyu Wang3 | ||
1Third Affiliated Hospital of Zhengzhou University, zhengzhou, China, 2Zhengzhou University People’s Hospital, zhengzhou, China, 3MR Research China, GE Healthcare, Beijing 100000, PR China, Beijing, China |
||
Keywords: Machine Learning/Artificial Intelligence, Diffusion/other diffusion imaging techniques, Tract-Based Spatial Statistics Early and accurate diagnosis of pediatric autism was difficult for clinicians, which hindered the timely treatment of patients.This study aimed to explore the applicability of machine learning models of diffusion kurtosis imaging (DKI) based on Tract-Based Spatial Statistics to diagnose pediatric autism. Results showed that DKI parameter was potential for differentiating early autism from the normal. And the machine learning models can be used for early detection of pediatric autism with high accuracy and sensitivity.
|
| 3934 | Computer 1
|
Rejecting or retaining motion corrupted transients in diffusion-weighted MRS: high b-values and multiple directions |
| Clémence Ligneul1, Jesper Andersson1, Saad Jbabdi1, Jason Lerch1,2,3, and William T Clarke1 | ||
1Wellcome Centre for Integrative Neuroimaging, FMRIB, Nuffield Department of Clinical Neurosciences, University of Oxford, Oxford, United Kingdom, 2Mouse Imaging Centre, The Hospital for Sick Children, Toronto, ON, Canada, 3Department of Medical Biophysics, University of Toronto, Toronto, ON, Canada |
||
Keywords: Data Processing, Spectroscopy, Diffusion-weighted spectroscopy Robust processing is crucial for a reliable microstructural interpretation of advanced diffusion-weighted magnetic resonance spectroscopy (DW-MRS) methods. The spectral signal intensity depends on the direction when using strong diffusion-gradients and multiple directions, even more if the voxel content is very anisotropic (e.g. contains a white matter tract). In this study we propose a model-based way of identifying motion corrupted averages that accounts for the voxel anisotropy. |
| 3935 | Computer 2
|
Zero filling does not inherently improve precision of real relative to complex-domain fitting in 1H-MR spectra with physiological baselines |
| Leonardo Campos1, Kelley M. Swanberg1, and Christoph Juchem1,2 | ||
1Biomedical Engineering, Columbia University, New York, NY, United States, 2Radiology, Columbia University, New York, NY, United States |
||
Keywords: Data Processing, Spectroscopy There remains controversy surrounding the use of zero filling during spectral quantification of in vivo proton magnetic resonance spectra (1H-MRS) using linear combination model (LCM) fitting. We examine the potential mixing of real and imaginary information theorized with zero filling, and whether this is demonstrated by a comparable change in accuracy and precision provided by complex fitting. We show that application of zero filling does improve fitting precision when a baseline is not present; with an imperfectly modeled unknown in vivo baseline, however, zero filling does not necessarily reduce error in real relative to complex fits. |
| 3936 | Computer 3
|
Simultaneous in vivo Detection of Cystathionine and 2HG at 7T and Simulation of MEGA and SLOW-editing performance |
| Guodong Weng1,2, Piotr Radojewski1,2, and Johannes Slotboom1,2 | ||
1Support Center for Advanced Neuroimaging (SCAN), University of Bern, Bern, Switzerland, 2Translational Imaging Center, sitem-insel, Bern, Switzerland |
||
Keywords: Pulse Sequence Design, Spectroscopy, Spectral editing 2-Hydroxyglutarate (2HG) is a biomarker for IDH-mutant glioma, and previous studies have shown that cystathionine (Cys) can be a potential biomarker for 1q/19q co-deletion of glioma. Previous studies used SVS based MEGA-editing MRS to detect 2HG and co-edited Cys, but editing efficiency of Cys is not optimal due to the narrow bandwidth of the applied MEGA editing-pulses. This work shows that SLOW-editing is able to detect both 2HG and Cys in an optimal way with whole-brain coverage. |
| 3937 | Computer 4
|
Compressed Sensing MPRAGE accelerates automated voxel placement for 1-H MRS of the human brain |
| Stefano Tambalo1, Sebastian Hübner1, Francesca Saviola1, Tobias Kober2,3,4, and Jorge Jovicich1 | ||
1University of Trento, Trento, Italy, 2Advanced Clinical Imaging Technology, Siemens Healthineers International AG, Lausanne, Switzerland, 3Department of Radiology, Lausanne University Hospital and University of Lausanne, Lausanne, Switzerland, 4LTS5, École Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland |
||
Keywords: Data Processing, Spectroscopy The automated voxel placement (AVP) framework has been proposed to optimize accuracy and reproducibility in single-voxel brain proton spectroscopy. AVP is based on the affine transformation of brain coordinates from standard to subject-specific space defined on a 3D T1-weighted structural scan. Here we evaluate the robustness of the AVP approach by evaluating the displacement of voxel center coordinates across 3D T1w MPRAGE sequence variants, including compressed sensing (CS) accelerated protocols. We show that AVP gave small but significantly higher voxel displacements for MP2RAGE and CS-MP2RAGE. There were no significant differences between multi-echo MPRAGE (6 min) and CS-MPRAGE (1 min). |
| 3938 | Computer 5
|
Development of a 2D MRSI sequence with Chemical-Shift Selective Adiabatic Pulses (2𝜋-CSAP) using Pulseq at 7T |
| Kyung Min Nam1,2,3, Guodong Weng2,3, Nam G Lee4, Edwin Versteeg1, Yeong-Jae Jeon5,6, Arjan Hendriks1, Jannie Wijnen1, Alex Bhogal1, Dennis Klomp1, Maxim Zaitsev7, and Johannes Slotboom2,3 | ||
1Department of High Field MR, Centre for Image Sciences, University of Medical Centre Utrecht, Utrecht, Netherlands, 2Institute for Diagnostic and Interventional Neuroradiology, Support Center for Advanced Neuroimaging (SCAN), University of Bern, Bern, Switzerland, 3Translational Imaging Center, sitem-insel AG, Bern, Switzerland, 4Department of Biomedical Engineering, University of Southern California, Los Angeles, CA, United States, 5Lee Gil Ya Cancer & Diabetes Institute, Gachon University, Incheon, Korea, Republic of, 6Department of Health Sciences and Technology, GAIHST, Gachon University, Incheon, Korea, Republic of, 7Department of Radiology, Division of Medical Physics, University Medical Center Freiburg, Freiburg, Germany |
||
Keywords: Data Acquisition, Spectroscopy An MRSI sequence using a pair of chemically selective adiabatic 2𝜋 refocus pulses, referred to as 2𝜋-CSAP, was implemented using the open-source Pulseq framework. This sequence enables full coverage of the frequency spectrum of interest through a shift in carrier frequency, which has been a limitation in rapid MRSI sequences. In addition, this shifted frequency spectrum also minimizes signal contaminations from residual water signal after water suppression or unsuppressed water signal and strong lipid signals, possibly eliminating the need for additional water suppression pre-pulses. This base sequence can potentially be combined with accelerated acquisition techniques for better scan efficiency. |
| 3939 | Computer 6
|
31P MRSI coil combination using 23Na sensitivity information acquired with the same loop array at 7T: preliminary verification |
| Jiying Dai1,2, Mark Gosselink1, Alexander J. E. Raaijmakers1,3, and Dennis W. J. Klomp1 | ||
1UMC Utrecht, Utrecht, Netherlands, 2Tesla Dynamic Coils B.V., Zaltbommel, Netherlands, 3Eindhoven University of Technology, Eindhoven, Netherlands |
||
Keywords: Image Reconstruction, Spectroscopy We utilized a quintuple-tuned RF head coil array by using the high-SNR 23Na signals from the brain to optimize the weighting for combining signals from the same coil array elements for the low-concentrated 31P metabolites. 23Na-weighted Roemer combination of 31P MRSI signals is verified on EM simulations and MR experiments. Comparing to 31P self-weighted combination, 23Na-weighted combination shows higher SNR and better-combined spectra in regions with low intrinsic SNR. It also shows potential of mitigating the signal contaminations when using 31P-self-weights for 31P data acquired with large voxels. |
| 3940 | Computer 7
|
A Case Study Analysis on Monte Carlo-Simulated Uncertainty Propagation in Absolute Quantification for in vivo 1H-MRS of the Human Brain |
| Ronald Instrella1 and Christoph Juchem1,2 | ||
1Biomedical Engineering, Columbia University, New York, NY, United States, 2Radiology, Columbia University, New York, NY, United States |
||
Keywords: Data Analysis, Spectroscopy, Brain, 1H-MRS, Quantification The effect of uncertainty propagation for absolute quantification performed on individual single-voxel in vivo 1H-MRS experiments has yet to be examined. In this case study, we conduct an uncertainty analysis using Monte Carlo simulations on two previously published studies with reported absolute concentrations of metabolites and macromolecules, each employing different field strengths, relative linewidths and internal references (e.g. total creatine, tissue-specific water). The uncertainty from 8 in vivo spectra is simulated and compared to CRLBs from Linear Combination Modeling. A multifold uncertainty increase is observed in both metabolites and macromolecules across datasets, demonstrating that the CRLB systematically overestimates quantification precision. |
| 3941 | Computer 8
|
Predicting Uncertainty of Metabolite Quantification in Magnetic Resonance Spectroscopy with Applications for Adaptive Ensembling |
| Julian P. Merkofer1, Sina Amirrajab1, Johan S. van den Brink2, Mitko Veta1, Jacobus F. A. Jansen1,3, Marcel Breeuwer1,2, and Ruud J. G. van Sloun1,4 | ||
1Eindhoven University of Technology, Eindhoven, Netherlands, 2Philips Healthcare, Best, Netherlands, 3Maastricht University Medical Center, Maastricht, Netherlands, 4Philips Research, Eindhoven, Netherlands |
||
Keywords: Machine Learning/Artificial Intelligence, Spectroscopy, Deep Learning, Uncertainty Prediction, Adaptive Ensemble Current deep learning methods for metabolite quantification in magnetic resonance spectroscopy do not offer reliable measures for uncertainty. Having a widely applicable measure can aid with the identification of fitting errors and enable uncertainty-based adaptive ensembling of model-based quantification and neural network predictions. In this abstract, we propose a training strategy based on a log-likelihood cost that allows joint optimization of concentration and uncertainty estimation for each individual metabolite. We show that the predicted uncertainties correlate well with the actual estimation errors and that uncertainty-based adaptive ensembling outperforms the individual estimators as well as standard ensembling. |
| 3942 | Computer 9
|
2D spectral-temporal fitting of synthetic fMRS data improves the precision of fitted glutamate temporal dynamics parameters three-fold |
| Yiling Liu1,2, Zhiyong Zhang1, and Assaf Tal2 | ||
1School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China, 2Department of Chemical and Biological Physics, Weizmann Institute of Science, Rehovot, Israel |
||
Keywords: Data Analysis, Spectroscopy It has been suggested that fitting dynamic MRS data in tandem (2D fitting) should be more precise than conventional 1D fitting, without impairing its accuracy. Functional MRS (fMRS) is a dynamic method used for detecting endogenous metabolic changes in the brain. In this work, we implemented a 2D spectral-temporal fitting framework for the synthetic fMRS data. Preliminary experiments confirm that 2D fitting improves precision approximately three-fold compared to the conventional 1D fitting in terms of fMRS data. |
| 3943 | Computer 10
|
Optimization and comparison of coil combination and spectral registration strategies for in-vivo DW-MRS |
| Ke Zhou1, Ziyan Wang1, Dingyi Lin1, Yi-Cheng Hsu2, and Min Wang1 | ||
1College of Biomedical Engineering & Instrument Science, Zhejiang University, Hangzhou, China, 2Siemens Healthcare Ltd, Shanghai, China |
||
Keywords: Data Processing, Spectroscopy, Spectra Registration Diffusion-weighted MRS (DW-MRS) suffers from low signal intensity and large phase errors when applying strong diffusion weighting. Coil combination and spectral registration are vital as spectra with high b-values are usually poorly reconstructed. This work aims at the optimization of post-processing pipeline with singular value decomposition (SVD)-based spectral registration and the comparison of different regimes of coil combination and spectral registration for in-vivo DW-MRS post-processing. The SNR of post-processed spectra and the metabolite diffusivities are assessed. The statistical outcomes suggest that the SVD-based spectral registration could improve the SNR of DW-MRS signal at high b-values and therefore alleviate diffusivity overestimations. |
| 3944 | Computer 11
|
Comparison of FAST(EST)MAP and BOLERO shimming for single voxel spectroscopy at 7T |
| Jullie W Pan1, Melissa Terpstra1, Junghwan Kim1, and Hoby P Hetherington2 | ||
1Radiology, University of Missouri Columbia, columbia, MO, United States, 2Resonance Research Inc., Billerica, MA, United States |
||
Keywords: Data Acquisition, Spectroscopy, shimming Because of the small volumes (2 to 15cc) used in single voxel spectroscopy (SVS), field inhomogeneities over such voxels are thought to be well managed by low order shim methods, e.g., FAST(EST)MAP using 1st-2nd spherical harmonic shim terms. In this report performed at 7T (Siemens Terra) we evaluate SVS from a key region of interest (prefrontal cortex PFC) with FAST(EST)MAP in comparison with Bolero shimming, a method that specifically manages high order shim terms. High accuracy field maps and LCModel analysis of STEAM spectroscopy are evaluated. |
| 3945 | Computer 12
|
Improved reproducibility of GABA measurement in short-TE 1H MRS by linewidth-matched basis sets |
| Ying Xiao1,2, Bernard Lanz1, Songi Lim1,2, and Lijing Xin1 | ||
1Animal imaging and technology, CIBM, EPFL, Lausanne, Switzerland, 2Laboratory for Functional and Metabolic Imaging (LIFMET), EPFL, Lausanne, Switzerland |
||
Keywords: Data Processing, Spectroscopy, LCModel This study investigated the spectral linewidth effect on GABA quantification, as well as its correction approach in LCModel with both simulated and in vivo acquired 1H MR spectra. Using the same basis set to fit spectra with different linewidths results in elevated variations between measurements. An effective method to reduce the variability caused by the linewidth difference effect and more accurately quantify GABA is to use spectral-linewidth-matched basis sets. This study shows the validity of this approach for most of the Voigt-type spectra. |
| 3946 | Computer 13
|
Correlations of 31P-containing metabolites in human brain at 3 and 7 Tesla |
| Sungtak Hong1 and Jun Shen1 | ||
1National Institute of Mental Health, National Institutes of Health, Bethesda, MD, United States |
||
Keywords: Data Analysis, Spectroscopy Numerical Monte Carlo analysis was performed to quantify metabolite-metabolite correlations of spectral origin in 31P MRS spectra acquired from human brain at both 3 and 7 Tesla without any confounding biological correlations. Significant correlations were found for many 31P-containing metabolite pairs. In particular, the 3 and 7 Tesla NAD+-NADH correlations are significantly different because of the field strength difference. These results demonstrate that it is necessary to incorporate metabolite-metabolite correlations originating from spectral overlap into statistical models that correlate MRS measurements with clinical parameters when overlapping 31P signals are of clinical interest. |
| 3947 | Computer 14
|
Can NMR phytometabolomics play a role in prevention and management of obesity? |
| Ankita Singh1, Aruna Singh1, Dushyant Kumar1, and Rama Jayasundar2 | ||
1NMR, All India Institute of Medical Sciences, New Delhi, India, 2Department of NMR, All India Institute of Medical Sciences, New Delhi, India |
||
Keywords: Data Acquisition, Spectroscopy, proton NMR spectroscopy, in vitro spectroscopy, obesity Obesity, a disorder of lipid metabolism, has become a serious health issue globally. Recently, research on pungent phytochemicals and plants with nutritional and medicinal values has gained interest in obesity management. In this study, phytochemicals (n=21) and medicinal plants (n=38) from pungent and non-pungent groups were studied using proton NMR phytometabolomics for their anti-obesity properties. Multivariate analysis of NMR data demonstrated the potential of proton NMR metabolomics in differentiating medicinal plants and their active phytochemicals with anti-obesity properties. These were further confirmed with anti-lipase assays and alkaloid analysis (indicating presence of pungent molecules) of the medicinal plants. |
| 3948 | Computer 15
|
Alterations in cerebral metabolism from wakefulness to non-REM sleep |
| Jing Xu1, Michael C Langham1, Hengyi Rao2, Marianne Nabbout1, Alessandra S Caporale1,3,4, Alexander M Barclay1, John A Detre2, and Felix W Wehrli1 | ||
1Department of Radiology, University of Pennsylvania, Philadelphia, PA, United States, 2Department of Neurology, University of Pennsylvania, Philadelphia, PA, United States, 3Department of Neurosciences, ‘G. d’Annunzio University’ of Chieti-Pescara, Chieti, Italy, 4Institute for Advanced Biomedical Technologies (ITAB), ‘G. d’Annunzio University’ of Chieti-Pescara, Chieti, Italy |
||
Keywords: Data Analysis, Metabolism Sleep is fundamental to human health and function. Cerebral metabolism and blood supply are key physiological parameters of brain function, but the manner in which they change during different sleep stages is still largely unknown. In this study, we collected wakefulness and sleep data with concurrent EEG-MRI. We measured CBF, SvO2 and CMRO2 with radial OxFlow MRI. Our data show that CMRO2 is lower during non-REM sleep than during wakefulness and declines progressively as sleep stages become deeper. CBF decreases during non-REM sleep compared with wakefulness, while SvO2 gradually increases from wakefulness to slow wave sleep. |
| 3949 | Computer 16
|
Apparent diffusion coefficients of 31P metabolites in the human calf muscle at 7T |
| Zhiwei Huang1, Giulio Gambarota2, Ying Xiao1, Daniel Wenz1, and Lijing Xin1 | ||
1Animal imaging and technology core (AIT), Center for Biomedical Imaging (CIBM), Ecole Polytechnique Fédérale de Lausanne, Ecublens, Switzerland, 2Faculty of Pharmacy, University of Rennes, Rennes, France |
||
Keywords: Data Acquisition, Spectroscopy, Diffusion 31P diffusion magnetic resonance spectroscopy could assess the diffusion properties of high energy metabolites. In this study, a diffusion weighted (DW) STEAM sequence was implemented, and spectra were acquired in the human calf muscle of six healthy volunteers. Frequency and phase alignments were applied prior to spectral averaging. The ADC of phosphocreatine (PCr), adenosine triphosphate (ATP), inorganic phosphate (Pi) and glycerol phosphorylcholine (GPC) were (0.24±0.02, 0.15±0.04, 0.43±0.14, 0.40±0.09)×10-3 mm2/s. To the best of our knowledge, this is the first study reporting the ADCs of ATP, Pi, and GPC, and the second study reporting the ADC of PCr in human. |
| 3950 | Computer 17
|
Long-term balance training enhances sensorimotor GABA levels in older adults: A 7 T longitudinal magnetic resonance spectroscopy study |
| Xinyu Liu1,2,3, Selin Scherrer4, Sven Egger4, Song-I Lim1,3, Benedikt Lauber4, Wolfgang Taube4, and Lijing Xin1,3 | ||
1Animal Imaging and Technology, Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland, 2Laboratory for functional and metabolic imaging (LIFMET), Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland, 3Center for Biomedical Imaging (CIBM), Lausanne, Switzerland, 4Department of Neurosciences and Movement Science, University of Fribourg, Fribourg, Switzerland |
||
Keywords: Data Acquisition, Spectroscopy, Data analysis, motor training, GABA, 7 Tesla This study aims to examine the modulatory effect of long-term balance training on sensorimotor cortex GABA level in an older population. In-vivo GABA levels were accessed using in-vivo MR spectroscopy (MRS) at 7T. GABA level in sensorimotor cortex was measured in sixteen healthy older adults using MEGA-sSPECIAL sequence, and short echo time semi-sSPECIAL sequence was used for metabolite profiling, before and after a three-months period of balance training. Using edited MEGA-sSPECIAL we detected a significant increase in sensorimotor GABA level after training, indicating potentially enhanced motor inhibition by coordinative balance learning. |
| 3951 | Computer 18
|
Two Dimensional MRSI of Human Calf Muscle at 3T: Decoupling and NOE effects on PDE and total NAD |
| Rajakumar Nagarajan1, Jane A Kent2, and Gwenael Layec2 | ||
1Human Magnetic Resonance Center, Institute for Applied Life Sciences, University of Massachusetts, Amherst MA, Amherst, MA, United States, 2Kinesiology, University of Massachusetts, Amherst MA, Amherst, MA, United States |
||
Keywords: Data Acquisition, Spectroscopy Using two-dimensional 31-phosphorus spectroscopic imaging (2D-MRSI) in human skeletal muscle, we have demonstrated that proton decoupling and nuclear Overhauser effect sequences improve the coefficient of variation and enhance the resolution of key metabolites in vivo. |
| 3952 | Computer 19
|
GABA spectroscopic imaging at 9.4T to localize the epileptogenic zone in an animal model of focal epilepsy |
| Alicia Plaindoux1, Yann Le Fur2, Jia Guo3, Clothilde Courivaud1, Julien Valette4, Vasile Stupar1,5, and Florence Fauvelle1,5 | ||
1Grenoble Institute Neurosciences, INSERM, U1216, University Grenoble Alpes, Grenoble, France, 2CRMBM, Aix-Marseille University, Marseille, France, 3Departement of Psychiatry, Columbia University, New York, NY, United States, 4Université Paris-Saclay, CEA, CNRS, MIRCen, Laboratoire des Maladies Neurodégénératives, Fontenay-aux-Roses, France, 5IRMaGe, INSERM, US17, CNRS, UMS 3552, CHU Grenoble Alpes, University Grenoble Alpes, Grenoble, France |
||
Keywords: Data Acquisition, Spectroscopy, MEGA-LASER, CSI, edition A surgical resection of the epileptogenic zone (EZ) can be proposed to drug-resistant epileptic patients, mainly suffering from focal epilepsy, in order to free them from their seizures. However, the mean efficiency of surgery is around 50 to 80%, indicating a possible miss-delimitation of the EZ. In recent studies, GABA was found to be a specific biomarker of the EZ in a mouse model of mesio-temporal lobe epilepsy (MTLE). Therefore, in this project, we propose a GABA-edited spectroscopic imaging method to improve the spatial localization of the EZ in this mouse model. |
| 3953 | Computer 20
|
Accelerated Rosette Spectroscopic Imaging with semi-LASER Localization |
| Ajin Joy1, Uzay Emir2,3, Paul M. Macey4, and M. Albert Thomas1 | ||
1Radiological Sciences, University of California, Los Angeles, Los Angeles, CA, United States, 2School of Health Sciences, Purdue University, West Lafayette, IN, United States, 3Weldon School of Biomedical Engineering, Purdue University, West Lafayette, IN, United States, 4School of Nursing and Brain Research Institute, University of California, Los Angeles, Los Angeles, CA, United States |
||
Keywords: Data Processing, Brain Limiting the total data acquisition time to a clinically feasible runtime has been a major challenge in MR spectroscopic imaging. Recently rosette based non-cartesian encoding of k-space has been used for spectroscopic imaging due to their fast encoding speed and lower gradient/slew rate requirements. While rosette spectroscopic imaging has been attempted for 2D and 3D spectroscopic imaging, feasibility of undersampling the petals in rosette is not shown. In this study, we implemented a rosette 2D and 3D spectroscopic imaging sequence and shown the feasibility of acceleration factors up to 8x using compressed sensing reconstruction. |
| 3954 | Computer 21
|
Comparison of uniform-density, variable-density and dual-density spiral samplings for multi-shot diffusion-weighted imaging |
| Guangqi Li1, Xiaodong Ma2, Sisi Li1, Xinyu Ye1, Peter Börnert3,4, Xiaohong Joe Zhou5, and Hua Guo1 | ||
1Center for Biomedical Imaging Research, Department of Biomedical Engineering, School of Medicine, Tsinghua University, Beijing, China, 2Department of Radiology and Imaging Sciences, University of Utah, Salt Lake City, UT, United States, 3Radiology, C.J. Gorter Center for High-Field MRI, Leiden University Medical Center, Leiden, Netherlands, 4Philips Research, Hamburg, Germany, 5Center for MR Research and Departments of Radiology, Neurosurgery, and Biomedical Engineering, University of Illinois College of Medicine at Chicago, Chicago, IL, United States |
||
Keywords: Data Acquisition, Diffusion Tensor Imaging Different multi-shot spiral sampling schemes have been developed for high-resolution DWI. However, the performances of these sampling strategies such as variable-density spiral (VDS), dual-density spiral (DDS) and uniform-density spiral (UDS) have not been compared comprehensively. In this study, we compare multi-shot UDS, VDS and DDS in brain DWI in terms of inter-shot phase error correction, overall image quality and SNR performance. Both theoretical analysis and in-vivo results demonstrate that UDS exhibits the best off-resonance performance among the three spiral sampling patterns. Additionally, UDS achieves the highest SNR in diffusion imaging over the VDS and DDS acquisitions. |
| 3955 | Computer 22
|
Deep Learning Based Self-Navigated Diffusion Weighted Multi-Shot EPI with Supervised Denoising |
| Yiming Dong1, Kirsten Koolstra2, Laurens Beljaards3, Marius Staring3, Matthias J.P. van Osch4, and Peter Börnert4,5 | ||
1LUMC, Leiden, Netherlands, 2Philips, Best, Netherlands, 3Division of Image Processing, Department of Radiology, LUMC, Leiden, Netherlands, 4C.J. Gorter MRI Center, Department of Radiology, LUMC, Leiden, Netherlands, 5Philips Research, Hamburg, Germany |
||
Keywords: Machine Learning/Artificial Intelligence, Diffusion/other diffusion imaging techniques Advanced diffusion weighted self-navigated multi-shot MRI can run at high scan efficiencies resulting in good image quality. However, the model-based image reconstruction is rather time consuming. Deep learning-based reconstruction approaches could function as a faster alternative. Tailored network architectures with appropriately set physical model constraints can help to shorten reconstruction times, resulting in good image quality with reduced noise propagation. |
| 3956 | Computer 23
|
A comparison of navigator-free multi-shot spiral and EPI in high-resolution DWI |
| Guangqi Li1, Sisi Li1, and Hua Guo1 | ||
1Center for Biomedical Imaging Research, Department of Biomedical Engineering, School of Medicine, Tsinghua University, Beijing, China |
||
Keywords: Data Acquisition, Diffusion/other diffusion imaging techniques Navigator-free multi-shot spiral and EPI acquisitions have been developed to achieve 2D high-resolution diffusion imaging. In this study, we investigated the off-resonance effects of spiral and iEPI samplings with different partial Fourier factors based on a signal model. Moreover, the SNR performances of the two multi-shot acquisitions at different resolutions and TEs were also explored. In summary, compared with iEPI, spiral provides superior SNR in multi-shot navigator-free DWI at various resolutions and TEs, even when the TE of spiral acquisition is slightly longer than that of iEPI. However, off-resonance correction for spiral sampling with ultra-long readout durations is more challenging. |
| 3957 | Computer 24
|
Feasibility of high resolution Readout-segmented echo planar imaging with simultaneous multi-slice in the assessment of rectal cancer |
| Mi zhou1, Meining Chen2, and Yuting Wang1 | ||
1Sichuan Provincial People's Hospital, chengdu, China, 2MR Scientific Marketing, Siemens Healthcare, Shanghai, China |
||
Keywords: Data Acquisition, Diffusion/other diffusion imaging techniques DWI has an important role in the staging and treatment response assessment of patients with rectal cancer, but its resolution is much lower than dynamic contrast-enhanced MRI and T2W images. We achieved a high-resolution scan of the rectal cancer using SMS HR rs-EPI sequences with a resolution of 1.1×1.1×2 mm3. The SMS HR rs-EPI provided a significantly better image quality and more valuable ADC than conventional HR rs-EPI when assessing rectal cancer. The pretreatment ADC values of HR rs-EPI could be utilized to distinguish well and poorly differentiated rectal cancer. |
| 3958 | Computer 25
|
AcceleraTed deep-LeArning for model-free and multi-Shell (ATLAS) DWI |
| Phillip Andrew Martin1, Maria Altbach2,3, and Ali Bilgin1,2,3,4 | ||
1Department of Electrical and Computer Engineering, University of Arizona, Tucson, AZ, United States, 2Department of Biomedical Engineering, University of Arizona, Tucson, AZ, United States, 3Department of Medical Imaging, University of Arizona, Tucson, AZ, United States, 4Department of Applied Mathematics, University of Arizona, Tucson, AZ, United States |
||
Keywords: Machine Learning/Artificial Intelligence, Diffusion/other diffusion imaging techniques In this work, we aim to accelerate diffusion weighted MRI (dMRI) by predicting diffusion -weighted images (DWIs) across different shells using deep learning (DL), while remaining independent of a diffusion-model constraint. The proposed approach enables the predictions of unacquired DWIs in multiple shells from a small set of acquired DWIs from a given shell. This relaxes the need for applying multiple diffusion gradient weightings for obtaining a fully-acquired dataset over multiple shells. Without the constraint of a diffusion model, accurate diffusion metrics over multiple diffusion models can potentially be obtained by acquiring a small number of DWIs.
|
| 3959 | Computer 26
|
Self-Calibrated Subspace Reconstruction using Temporally Local Matrix Completion for Multidimensional MR Fingerprinting |
| Zhilang Qiu1, Siyuan Hu1, Walter Zhao1, Ken Sakaie2, Mark A. Griswold3, Derek K. Jones4, and Dan Ma1 | ||
1Biomedical Engineering, Case Western Reserve University, Cleveland, OH, United States, 2Imaging Institute, Cleveland Clinic, Cleveland, OH, United States, 3Radiology, Case Western Reserve University, Cleveland, OH, United States, 4Cardiff University Brain Research Imaging Centre (CUBRIC), School of Psychology, Cardiff University, Cardiff, United Kingdom |
||
Keywords: Image Reconstruction, Diffusion/other diffusion imaging techniques We propose a new reconstruction method, named self-calibrated subspace reconstruction, for multidimensional MR fingerprinting (mdMRF), in order to correct the artifacts due to inter-shot (segment) magnitude and phase variations, without the need for extra navigator or calibration data. Different options for utilizing the low-rank property and the signal structure of mdMRF data are investigated and the optimal scheme is determined. Such that aliasing-free high-resolution image reconstruction and high-quality quantification, can be achieved. |
| 3960 | Computer 27
|
Extended multi-shell diffusion acceleration with Gaussian processes estimated reconstruction (ems-DAGER) |
| Xinyu Ye1, Karla Miller1, and Wenchuan Wu1 | ||
1Wellcome Centre for Integrative Neuroimaging, FMRIB, Nuffield Department of Clinical Neurosciences, University of Oxford, UK, Oxford, United Kingdom |
||
Keywords: Image Reconstruction, Diffusion/other diffusion imaging techniques, Signal modelling Diffusion-weighted MRI suffers from relatively long acquisition time, especially for high spatial- resolution and/or high angular- resolution acquisitions. Thus, methods to increase the acquisition speed are urgently needed. Recently, increasing attention has been paid to utilize the relations between k- and q-space points for further acceleration. Here, we extend the Diffusion Acceleration with Gaussian process Estimated Reconstruction (DAGER) to leverage shared information in multi-shell acquisitions and incorporate eddy-induced distortion correction. |
| 3961 | Computer 28
|
In Vivo Diffusion MRI at 7 T: High Spatial-Angular-Temporal Resolution Pursuit |
| Frank Z Tan1, Patrick Alexander Liebig2, Robin Martin Heidemann2, Frederik Bernd Laun3, and Florian Knoll1 | ||
1Artificial Intelligence in Biomedical Engineering, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany, 2Siemens Healthcare GmbH, Erlangen, Germany, 3Institute of Radiology, University Hospital Erlangen, Erlangen, Germany |
||
Keywords: Image Reconstruction, Diffusion/other diffusion imaging techniques, Ultra high field, multi-shell, q-space, crossing fiber The pursuit of high-spatial-angular-temporal resolution for in vivo diffusion MRI at 7T is challenging, but also receives continuous interest. We hereby propose shift-encoded interleaved EPI and a joint reconstruction technique with LLR regularization. Preliminary results achieve up to 8.7-fold acceleration per shot in 2-shot EPI acquisition with 1.4 mm isotropic nominal resolution. Moreover, with the integrated joint reconstruction for noise reduction, high-quality diffusion-weighted images render more spatially-continuous fiber anisotropy maps and clearer fiber crossing in the fiber orientation distribution function. |
| 3962 | Computer 29
|
Integrated shimming technique can improve ultrahigh-b-value DWI in laryngeal and hypopharyngeal squamous cell carcinoma: comparison with conventional volume shimming |
| Yan Wen1, Liling Long1, Chenhui Li1, Huiting Zhang2, and Alto Stemmer3 | ||
1Department of Radiology, The First Affiliated Hospital of Guangxi Medical University, Nanning, China, 2MR Scientific Marketing, Siemens Healthineers Ltd., Wuhan, China, 3MR Application Predevelopment, Siemens Healthcare GmbH, Erlangen, Germany |
||
Keywords: Data Analysis, Diffusion/other diffusion imaging techniques This study aimed to investigate the value of integrated shimming (iShim) technique in the ultrahigh-b-value DWI images in laryngeal and hypopharyngeal squamous cell carcinoma (SCC) by comparing with the conventional volume shimming technique. Our results showed that the ultrahigh-b-value DWI images with iShim technique had significantly higher image quality based on subjective (edge, artifact, and confidence) and objective (signal-to-noise ratio, contrast, and contrast-to-noise ratio) assessments compared with conventional single-shot-EPI DWI images with volume shimming. Therefore, ultrahigh-b-value DWI images with iShim technique can be well applied in laryngeal and hypopharyngeal SCC examination. |
| 3963 | Computer 30
|
Robust Complex Signal Averaging for Diffusion Weighted Imaging |
| Xinzeng Wang1, Daniel Litwiller2, Arnaud Guidon3, Patricia Lan4, and Tim Sprenger5 | ||
1GE Healthcare, Houston, TX, United States, 2GE Healthcare, Denver, CO, United States, 3GE Healthcare, Boston, MA, United States, 4GE Healthcare, Menlo Park, CA, United States, 5GE Healthcare, Stockholm, Sweden |
||
Keywords: Data Processing, Diffusion/other diffusion imaging techniques In the past decade, complex signal averaging has been investigated for diffusion weighted imaging to address the well-known noise floor issue. The robustness of complex signal averaging highly depends on the performance of phase correction to remove the shot-to-shot background phase variations. To achieve optimal phase correction, parameters (kernel size, regularization terms, etc.) need to be tuned for different anatomies, SNR levels and/or image size. In this work, we evaluated a deep-learning based phase correction method for various DWI applications, including brain, liver, prostate and showed improved complex signal averaging with lower noise floor and less artifacts. |
| 3964 | Computer 31
|
Whole Brain Simultaneous MultiSlice Filter EXchange Imaging in less than 15 minutes: Initial Results |
| Frederik Testud1, Arthur Chakwizira2, and Markus Nilsson2 | ||
1Siemens Healthcare AB, Malmö, Sweden, 2Department of Medical Radiation Physics, Lund University, Lund, Sweden |
||
Keywords: Data Acquisition, Diffusion/other diffusion imaging techniques, Neuro, Microstructure Filter EXchange Imaging (FEXI) is sensitive to the rate of diffusional water exchange, which depends among other factors on the cell membrane permeability. FEXI maps the so-called apparent exchange rate (AXR), however, previous FEXI protocols were limited to few slices because of very long acquisition times. In this proof-of-concept work, Simultaneous MultiSlice was included in the FEXI sequence and with protocol optimization allowed for a 12-minute whole-brain scan. FEXI accelerated with or without SMS showed similar AXR values. |
| 3965 | Computer 32
|
BUAN 2.0, streamlines as functions, nonlinear registration, and subdivision of bundles for advanced tractometry |
| Bramsh Qamar Chandio1,2, Sophia I Thomopoulos1, Paul M Thompson1, Jaroslaw Harezlak3, and Eleftherios Garyfallidis2 | ||
1Mark and Mary Stevens Neuroimaging and Informatics Institute, Keck School of Medicine, University of Southern California, Marina del Rey, CA, United States, 2Department of Intelligent Systems Engineering, Luddy School of Informatics, Computing, and Engineering, Indiana University Bloomington, Bloomington, IN, United States, 3Department of Epidemiology and Biostatistics, Indiana University Bloomington, Bloomington, IN, United States |
||
Keywords: Software Tools, Diffusion/other diffusion imaging techniques, Tractography, tractometry, bundle segmentation, white matter tracts, group analysis We propose BUAN 2.0, which adds new advancements to its predecessor BUAN. It sub-segments bundles with varying substructures in them. It uses nonlinear registration of bundles to find accurate correspondences among bundle segments across subjects. The number of horizontal segments is decided based on the bundle and data specifications. More importantly, in BUAN 2.0, instead of treating each point on the streamline as an independent observation, we treat each streamline as a complete entity, as a function. Streamlines are analyzed by deploying functional data analysis (FDA) methods for studying group differences in populations along the length of the tracts. |
| 3966 | Computer 33
|
Black Blood Cardiac Diffusion Imaging using Second Order Motion Compensation and Double Inversion Recovery |
| Yishi Wang1, Zhen Zhang2, Rui Wang2, Xiuzheng Yue1, Fang Wang2, Rongrong Zhu2, and Ruoshui Ha2 | ||
1Philips Healthcare, Beijing, China, 2Medical Imaging Center, People's Hospital of Ningxia Hui Autonomous Region, Yinchuan, China |
||
Keywords: Pulse Sequence Design, Diffusion/other diffusion imaging techniques, Cardiac Diffusion, Black Blood, Motion Compensation Second-order motion-compensated diffusion imaging is a robust solution for cardiac diffusion imaging but prone to bright blood signals due to motion compensation itself. In this work, we incorporated black blood module into second-order motion-compensated sequence to achieve cardiac diffusion imaging with blood signal suppressed. |
| 3967 | Computer 34
|
Investigation of motion probing gradient pulse for uniform collection of temporal and spatial diffusion information |
| Yuichi Suzuki1, Katsutoshi Murata2, Hideyuki Iwanaga1, and Osamu Abe1 | ||
1Radiology Center, The University of Tokyo Hospital, Tokyo, Japan, 2Siemens Healthineers, Tokyo, Japan |
||
Keywords: Data Acquisition, Diffusion/other diffusion imaging techniques We examined the application order of MPG pulses (called reordered MPG) that can uniformly collect spatial diffusion information even if the examination is interrupted, compare it with the an electrostatic repulsion method (conventional method, called original MPG), and verify its usefulness. In conclusion, The reordered MPG, which was rearranged based on the MPG direction of the original, could collect diffusion information more uniformly than the original MPG even if the inspection examination was interrupted. |
| 3968 | Computer 35
|
Accelerated High Resolution Diffusion Imaging using 3D multi-shot EPI and Model-based Reconstruction |
| Chu-Yu Lee1 and Merry Mani1 | ||
1University of Iowa, Iowa City, IA, United States |
||
Keywords: Artifacts, Diffusion/other diffusion imaging techniques Diffusion weighted images are usually acquired using 2D EPI methods. This technique has the limitation of low spatial resolution and low SNR. 3D diffusion weighted images are attractive option for improved SNR, however the motion-induced phase inconsistencies pose a challenge for the reconstruction of such data. We present a navigator-free phase-compensated reconstruction, which can be implemented as a regular parallel imaging framework. This two-step method involves first estimating a low-resolution phase from the data itself and then integrating the phase information in the forward encoding operator. We show good phase compensation from accelerated datasets using the proposed approach. |
| 3969 | Computer 36
|
Readout-segmented EPI using 2D spatially-selective RF excitation pulses for DWI with reduced FOV |
| Wei Liu1, Thomas Benkert1, and Elisabeth Weiland1 | ||
1MR Application Predevelopment, Siemens Healthcare GmbH, Erlangen, Germany |
||
Keywords: Pulse Sequence Design, Diffusion/other diffusion imaging techniques In this work, we propose to further minimize the distortion and T2* blurring in DWI by the combination of readout-segmented EPI and 2D spatially-selective RF excitation. We demonstrate its application to DWI of the temporal lobe and the uterus. The experimental results based on volunteer scans show substantial distortion reduction in the proposed method, compared to the conventional 2DRF based single-shot EPI with reduced FOV. |
| 3970 | Computer 37
|
Volume isotropic thin-slice high-quality brain DWI with no fat-suppression pre-pulse |
| Takayuki SAKAI1, Masami Yoneyama2, Diachi Murayama1, Iain Ball3, Tosiaki MIYATI4, Shigehiro Ochi1, and Atsuya Watanabe5,6 | ||
1Radiology, Eastern Chiba Medical Center, Chiba, Japan, 2Philips Japan, Tokyo, Japan, 3Philips Australia & New Zealand, North Ryde, Australia, 4Faculty of Health Sciences, Institute of Medical, Pharmaceutical and Health Sciences, Kanazawa University, Kanazawa, Japan, 5General Medical Services, Chiba University Graduate School of Medicine, Chiba, Japan, 6Orthopaedic Surgery, Eastern Chiba Medical Center, Chiba, Japan |
||
Keywords: Data Acquisition, Diffusion/other diffusion imaging techniques We hypothesized that LIPO-only (LION) DWI might be one of the best solutions to improve the image quality of volumetric thin-slice DWI if it is further optimized to increase the robustness of fat suppression. We investigated the clinical usefulness of proposed LION fat suppression combined with thin-slice DWI for brain volumetric DWI in patients with acute stroke. LION-DWI has improved SNR compared to conventional DWI and has superior visual lesion detectability in patients with acute stroke. Therefore, it is possible to improve the image quality and shorten the imaging time of thin-slice DWI by changing the fat suppression to LION. |
| 3971 | Computer 38
|
The Trouble with Free-Water Elimination using Single-Shell Diffusion MRI Data: A Case-Study in Ageing |
| Marta M. Correia1, Stefan Winzeck2,3, Marc Golub4, and Rita G. Nunes4 | ||
1MRC Cognition and Brain Sciences Unit, University of Cambridge, Cambridge, United Kingdom, 2BioMedIA Group, Department of Computing, Imperial College London, London, United Kingdom, 3Division of Anaesthesia, Department of Medicine, University of Cambridge, Cambridge, United Kingdom, 4Institute for Systems and Robotics - Lisboa and Department of Bioengineering, Instituto Superior Técnico, Universidade de Lisboa, Lisbon, Portugal |
||
Keywords: Data Analysis, Diffusion/other diffusion imaging techniques, Ageing, Free-Water Two different algorithms for free-water elimination (FWE) were applied to single-shell and multi-shell diffusion MRI data. Positive correlations were found between age and fractional anisotropy (FA) estimated with the FWE algorithm for single-shell, but this was not replicated with multi-shell diffusion data. Because only the multi-shell FWE algorithm is well-posed, we postulated that the positive correlations between age and FA must be a false positive finding, resulting from inappropriate fitting using single-shell data. FWE estimates from single-shell modelling have been shown to be biologically plausible, but this does not imply specificity, and more work is required to validate these approaches. |
| 3972 | Computer 39
|
New Rapid 3D diffusion weighted imaging using 3D hybrid Radial-Echo Planar Imaging (RAZER) sequence |
| Seong-Eun Kim1, Henrick Carl Axel Odeen 1, John A Roberts1, and Dennis L Parker1 | ||
1UCAIR, Department of Radiology and Imaging Sciences, University of Utah, Salt Lake City, UT, United States |
||
Keywords: Pulse Sequence Design, Diffusion/other diffusion imaging techniques, non-cartesian trajectory Non-Cartesian sampling schemes offer advantages over Cartesian schemes. Recently 2D radial and 3D hybrid radial-Cartesian approaches such as 3D stack of stars have gained popularity. Despite of many advantages, the current hybrid trajectory is presently limited by drawbacks including the degradation of sampling efficiency and the k-space coverage per unit time for diffusion application. For more flexible, rapid, and motion-robust 3D DWI, we implemented a 3D hybrid radial-EPI(RZAER) sequence and demonstrated the feasibility of 3D diffusion weighted RAZER sequence for carotid vessel wall.
|
| 3973 | Computer 40
|
Incorporating Tissue and Blood Relaxometry For Estimating IVIM Parameters: A Simulation Study |
| Yousef Mazaheri1 | ||
1Memorial Sloan Kettering Cancer Center, New York, NY, United States |
||
Keywords: Signal Representations, Diffusion/other diffusion imaging techniques, IVIM The aim of this aim of this study is to use numerical simulation to compare the performance of the standard bi-exponential IVIM model to an extended model which incorporates tissue and blood relaxometry for the estimation of IVIM parameters. |
| 4091 | Computer 1
|
Interleaved Whole Brain 23Na MRI and 31P MRSI Acquisitions at 7T |
| Zahra Shams1, Jiying Dai1, Mark Gosselink1, Wybe W.J.M. van der Kemp1, Fredy Visser1, Dennis W.J. Klomp1, Hans J.M. Hoogduin1, Tijl A. van der Velden1, Giel Mens1, Jannie P. Wijnen1, and Evita C. Wiegers1 | ||
1Department of Radiology and Nuclear Medicine, University Medical Center Utrecht, Utrecht, Netherlands |
||
Keywords: Data Acquisition, Data Acquisition The purpose of this study is to assess the feasibility of an interleaved acquisition of 31P MRSI and 23Na MRI. Triple-nuclei [metabolic] imaging of the brain was performed in one scan session with an interleaved acquisition of 3D 31P MRSI and 3D radial UTE 23Na MRI at 7 Tesla. The results of the interleaved 23Na-31P were compared with those acquired from non-interleaved runs, concluding no influence of each nucleus on the SNR or spectra quality. In conclusion, we showed that these two nuclei pools do not interfere with each other during interleaved acquisition. |
| 4092 | Computer 2
|
Interleaved multinuclear Na+-H+-CEST-EPI metabolic MRI for simultaneously quantifying salinity and acidity |
| Chencai Wang1,2, Xiaodong Zhong3, Alfredo L. Lopez Kolkovsky4, Sonoko Oshima1,2, Saneel Khairnar2,5, and Benjamin M Ellingson1,2,6,7 | ||
1Department of Radiological Sciences, UNIVERSITY OF CALIFORNIA, LOS ANGELES, Los Angeles, CA, United States, 2Brain Tumor Imaging Laboratory, University of California, Los Angeles, Los Angeles, CA, United States, 3MR R&D Collaborations, Siemens Medical Solutions USA, Los Angeles, CA, United States, 4Institute of Myology, Paris, France, 5Department of Molecular, Cell, and Developmental Biology, UNIVERSITY OF CALIFORNIA, LOS ANGELES, Los Angeles, CA, United States, 6Department of Neurosurgery, University of California, Los Angeles, Los Angeles, CA, United States, 7Department of Psychiatry and Biobehavioral Sciences, University of California, Los Angeles, Los Angeles, CA, United States |
||
Keywords: Pulse Sequence Design, Non-Proton, Sodium, 2-Nuclei Extensive evidence suggests abnormal metabolism, sodium homeostasis, and tumor acidity are interconnected and play critical roles in brain tumor formation, progression, seizure activity, treatment resistance, and immune suppression. While Na+ and advanced H+ -based MRI techniques can provide this information, Na+ and H+ images are traditionally acquired sequentially, resulting in a total scan time exceeding what is clinically reasonable. In the current study, we demonstrate utility of a new interleaved sodium gradient echo and proton-based pH-sensitive amine chemical exchange saturation transfer echoplanar imaging (Na+-GRE/H+-CEST-EPI) sequence in 15 minutes, making it feasible to study patients with brain tumors. |
| 4093 | Computer 3
|
Improved efficiency and quantitative accuracy in hyperpolarized 129Xe ventilation imaging using 2D spiral acquisition |
| Riaz Hussain1, Joseph W Plummer1,2, Abdullah S Bdaiwi1,2, Matthew M. Willmering1, Laura L Walkup1,2,3,4, and Zackary I Cleveland1,2,3,4 | ||
1Center for Pulmonary Imaging Research, Division of Pulmonary Medicine, Cincinnati Children's Hospital Medical Center, Cincinnati, OH, United States, 2Department of Biomedical Engineering, University of Cincinnati, Cincinnati, OH, United States, 3Imaging Research Center, Department of Radiology, Cincinnati Children’s Hospital Medical Center, Cincinnati, OH, United States, 4Department of Pediatrics, University of Cincinnati, Cincinnati, OH, United States |
||
Keywords: Data Acquisition, Hyperpolarized MR (Gas) Despite yielding high signal-to-noise ratio, conventional gradient recalled echo (GRE) sequence requires a relatively long breath-hold for 129Xe ventilation imaging. 2D-spiral sequence enables faster imaging and offers possibility to correct regional B1/flip-angle inhomogeneities and signal decay using keyhole reconstruction without making assumptions about bias texture that can obscure the physiology of interest. Here, GRE and 2D-spiral showed comparable image quality and revealed regional ventilation impairment in cystic fibrosis. Furthermore, flip-angle correction preserved signal variation due to underlying lung pathophysiology, including gravitation ventilation gradients and subtle defects. Thus, flip-angle corrections in 2D-spiral sequence may detect early and reversible disease-induced ventilation impairment. |
| 4094 | Computer 4
|
3D Pulmonary Dynamic Ventilation Imaging with High Spatial and Temporal Resolution Using Hyperpolarized 129Xe MRI |
| Hongchuang Li1,2, Haidong Li1,2, Ming Zhang1,2, Xiaoling Liu1,2, Xiuchao Zhao1,2, Yeqing Han1,2, Chaohui Ye1,2, and Xin Zhou1,2 | ||
1Key Laboratory of Magnetic Resonance in Biological Systems, State Key Laboratory of Magnetic Resonance and Atomic and Molecular Physics, National Center for Magnetic Resonance in Wuhan, Wuhan Institute of Physics and Mathematics, Innovation Academy for Precision Measurement Science and Technology, Chinese Academy of Sciences- Wuhan National Laboratory for Optoelectronics, Wuhan, China, 2University of Chinese Academy of Sciences, Beijing, China |
||
Keywords: Pulse Sequence Design, Hyperpolarized MR (Gas) The mutual restriction of temporal and spatial resolution is a challenge for hyperpolarized gas dynamic MRI, especially for 3D acquisition. Herein, we proposed a methodology to image the pulmonary dynamic ventilation with high temporal and spatial resolution using hyperpolarized gas MRI based on a 3D gradient echo sequence. Furthermore, we observed the difference of pulmonary dynamic ventilation in the posterior-anterior direction. |
| 4095 | Computer 5
|
Rapid Radiofrequency Coil Prototyping for Hyperpolarized Carbon-13 Imaging |
| Alexander Isaac Zavriyev1, John Choi1, Bukola Yetunde Adebesin1, Molly M Sheehan1, David J Tischfield1, and Terence P. F. Gade1 | ||
1University of Pennsylvania, Philadelphia, PA, United States |
||
Keywords: Data Acquisition, Non-Array RF Coils, Antennas & Waveguides Dynamic nuclear polarization magnetic resonance spectroscopic imaging (DNP-MRSI) of hyperpolarized carbon-13 (HP 13C) is an exciting new imaging technique that provides valuable information about the metabolism of disease states. Though there are many hyperpolarized substrates that can be used to investigate metabolism in a variety of models, the application of this technology often requires custom radiofrequency (RF) coil design. Furthermore, different coils within the same project may provide different results. In this study, we utilized 3D printing molds to generate inductors along with printed circuit board (PCB) designs to create a high reproducibility (<1% variation) pipeline for MR coils. |
| 4096 | Computer 6
|
Hyperpolarized 13C Metabolic Imaging of the Human Abdomen with Spatiotemporal Denoising |
| Tanner M. Nickles1,2, Yaewon Kim 1, Philip M. Lee1, Hsin-Yu Chen1, Michael Ohliger1, Peder E. Z. Larson1,2, Zhen J Wang1, Daniel B. Vigneron 1,2, and Jermey W. Gordon1 | ||
1Department of Radiology and Biomedical Imaging, University of California San Francisco, San Francisco, CA, United States, 2UC Berkeley-UCSF Bioengineering Program, University of California San Francisco, San Francisco, CA, United States |
||
Keywords: Data Processing, Hyperpolarized MR (Non-Gas), Pancreas, Abdomen, Cancer A substantial challenge in hyperpolarized (HP) 13C MRI is the limited signal-to-noise ratio (SNR) of downstream metabolites, which restricts the achievable spatial resolution. To overcome this for large coverage abdominal studies, a patch-based spatiotemporal denoising approach was applied to denoise dynamic imaging data in [1-13C]pyruvate echo-planar imaging (EPI) human datasets. With denoising, a 11.4 ± 1.8 and 8.7 ± 2.4 fold sensitivity gain was achieved for [1-13C]alanine and [1-13C]lactate, along with improved spatial coverage. These results support the potential of spatiotemporal denoising to improve quantification in HP 13C MRI for normal and cancer studies. |
| 4097 | Computer 7
|
Effects of patch size on 3D patch-based super-resolution reconstruction of hyperpolarized 13C cardiac MRI |
| Sung-Han Lin1, Junjie Ma2, and JaeMo Park1,3 | ||
1Advanced Imaging Research Center, UT Southwestern Medical Center, Dallas, TX, United States, 2GE Healthcare, New York, NY, United States, 3Radiology, UT Southwestern Medical Center, Dallas, TX, United States |
||
Keywords: Data Processing, Cardiomyopathy The volumetric patch-based super-resolution method can reconstruct a single-slice low-resolution hyperpolarized 13C MRI to multiple 13C slices with enhanced spatial resolution by exploiting the corresponding high-resolution structural 1H MRI. In this study, the overall performance of this method and the effects of patch size were evaluated using a simulated digital phantom and an anthropomorphic cardiac 13C MR phantom. The optimal patch size for this reconstruction method was also applied to in-vivo human 13C cardiac images acquired with an injection of hyperpolarized [1-13C]pyruvate. |
| 4098 | Computer 8
|
Lung Multi-Breath Wash in/out MRI with 19F with Sub 0.5 Second Scan Time |
| Sang Hun Chung1, Khoi Minh Huynh1, Yong Chen2, Pew-Thian Yap1, Jennifer Goralski1, Scott Donaldson1, and Yueh Z Lee1 | ||
1University of North Carolina Chapel Hill, chapel hill, NC, United States, 2Case Western Reserve University, Cleveland, OH, United States |
||
Keywords: Data Acquisition, Lung, 19F fluorine We studied the feasibility of multi breath wash-in/out lung acquisition with 19F gas and a sub 0.5 second acquisition spiral sequence. We compared the sub 0.5 seconds acquired data ability to measure wash-in/out time constants to a regular breath-hold 19F lung imaging and report correlation as well as differences in mean. Our results yielded high correlations (r > 0.88) for both wash-in and wash-out time constant calculations. The results indicate our proposed method yield similar calculated time wash-in/out time constants without the need for breath-holds and show promise for the use of free breathing 19F gas lung MRI. |
| 4099 | Computer 9
|
Improved Dynamic In Vivo pH Imaging Using Hyperpolarized 13C-bicarbonate |
| Xiaoxi Liu1, Changhua Mu1, Ying-Chieh Lai1,2, Robert A. Bok1, Romelyn Delos Santos1, Yaewon Kim1, Avantika Sinha1, Jeremy Gordon1, Robert Flavell1, and Peder Larson1 | ||
1Department of Radiology & Biomedical Imaging, University of California, San Francisco, San Francisco, CA, United States, 2Department of Medical Imaging and Intervention, Chang Gung Memorial Hospital at Linkou, Taiwan, Taiwan |
||
Keywords: Data Acquisition, Hyperpolarized MR (Non-Gas) We present a dynamic pH measurement using metabolite-specific gradient-echo spiral sequence with flow suppression to improve the SNR of bicarbonate and CO2 with short T2* and used bipolar gradients to suppress the high bicarbonate signal in the vessel. With the improved image quality and high temporal resolution, it is possible to measure the dynamic pH in vivo. Future studies will work to increase the sample size to confirm these results. |
| 4100 | Computer 10
|
Estimation of Metabolite Signals for In Vitro MRI of Glycolytic Flow in Hyperpolarized Carbon-13 |
| Ching-Yi Hsieh1,2, Ying-Chieh Lai3, Kuan-Ying Lu3, and Gigin Lin1,2,3 | ||
1Institute for Radiological Research, Chang Gung University, Taoyuan, Taiwan, 2Clinical Metabolomics Core Laboratory, Chang Gung Memorial Hospital, Linkou, Taoyuan, Taiwan, 3Department of Medical Imaging and Intervention, Chang Gung Hospital, Linkou, Taoyuan, Taiwan |
||
Keywords: Data Processing, Hyperpolarized MR (Non-Gas), metabolites; apparent exchange rate; kinetic model Downstream metabolites may have a poor SNR in hyperpolarized carbon-13 MRI, generating apparent exchange rate constant discrepancies when studying glycolytic flow in vivo or in vitro dynamically in real-time. We developed a method for estimating metabolite signal. This approach estimates metabolite signals using kinetic modeling and noise. The process was evaluated using simulations and in vitro studies. In vitro findings were also compared to 13C NMR cell media AUC. Comparing in vitro data from our method and NMR, both demonstrated consistency when uncertainty was included, suggesting that our method can accurately quantify metabolite signals and indicate how glycolytic flow changes. |
| 4101 | Computer 11
|
Assessment of Pulmonary Morphometry using Hyperpolarized 129Xe DWI with Variable-Sampling-Ratio Compressed Sensing Patterns |
| Qian Zhou1,2, Haidong Li1,2, Qiuchen Rao1, Ming Zhang1,2, Xiuchao Zhao1,2, Luyang Shen1, Yuan Fang1, Hongchuang Li1,2, Xiaoling Liu1,2, Sa Xiao1,2, Lei Shi1,2, Yeqing Han1,2, Chaohui Ye1,2, and Xin Zhou1,2 | ||
1Key Laboratory of Magnetic Resonance in Biological Systems, State Key Laboratory of Magnetic Resonance and Atomic and Molecular Physics, National Center for Magnetic Resonance in Wuhan, Wuhan Institute of Physics and Mathematics, Innovation Academy for Precision Measurement Science and Technology, Chinese Academy of Sciences- Wuhan National Laboratory for Optoelectronics, Wuhan, China, 2University of Chinese Academy of Sciences, Beijing, China |
||
Keywords: Data Acquisition, Hyperpolarized MR (Gas) The long acquisition time of hyperpolarized 129Xe multiple b-values DWI made it hard to apply in patients with severe pulmonary diseases. Herein, we proposed a method of variable-sampling-ratio compressed sensing patterns for accelerating hyperpolarized 129Xe DWI. A four-fold reduction in acquisition time was achieved using the proposed method while preserving good image quality. Meanwhile, the method can be used for evaluating pulmonary injuries caused by cigarette smoking. |
| 4102 | Computer 12
|
Measuring Air Flow Dynamics in the Lungs Using Hyperpolarized 129Xe MRI |
| Mostafa K Ismail1, Steve Kadlecek1, Hooman Hamedani1, Faraz Amzajerdian1, Ryan Baron1, Luis Loza1, Madeline Boyes1, Klaus Hopster1, Rachel Hilliard1, Thomas Schaer1, Benjamin Sinder2, Patrick Cahill2, Brian Snyder3, Kai Ruppert1, and Rahim Rizi1 | ||
1University of Pennsylvania, Philadelphia, PA, United States, 2Children’s Hospital of Philadelphia, Philadelphia, PA, United States, 3Boston Children’s Hospital, Boston, MA, United States |
||
Keywords: Data Analysis, Hyperpolarized MR (Gas), Xenon The ability to non-invasively measure air flow in the lungs could enable the early detection and assessment of abnormal air flow resulting from various obstructive and restrictive diseases. Here, we present a novel method for both obtaining and analyzing ventilation maps using hyperpolarized 129Xe MRI (HXe). In addition to assessing tidal volume and fractional ventilation, we introduce a novel sigmoid function describing inhalation and exhalation that enables the assessment of how fast air flows in the lungs and apply this method in a preclinical model of thoracic insufficiency syndrome (TIS) as well as three human lung transplant (LTx) recipients. |
| 4103 | Computer 13
|
Multiple-Slice, Multiple-Breath Washout Hyperpolarized 129Xe Ventilation Mapping of the Lung using Accelerated Data Acquisition |
| Faiyza Shoaib Alam1,2, Marcus John Couch3, Brandon Zanette2, Daniel Li2, Felix Ratjen2,4, and Giles Santyr1,2 | ||
1Medical Biophysics, University of Toronto, Toronto, ON, Canada, 2Translational Medicine, The Hospital for Sick Children, Toronto, ON, Canada, 3Siemens Healthcare Limited, Montreal, QC, Canada, 4Division of Respiratory Medicine, The Hospital for Sick Children, Toronto, ON, Canada |
||
Keywords: Parallel Imaging, Hyperpolarized MR (Gas), multiple-breath washout, multiple-slice, fractional ventilation Multiple-breath washout hyperpolarized 129Xe MRI (MBW Xe-MRI) results in a regional map of fractional ventilation (FV), measuring percent gas clearance/breath. One limitation of previous work from our group is the single thick coronal slice (200mm) centered on the chest cavity results in significant partial volume effects. Acquiring multiple slices using parallel imaging adds spatial resolution in the slice direction. Feasibility of multiple-slice FV mapping using parallel acceleration is investigated in one adult. Median [IQR] FV was 0.35 [0.30 0.36]. Gravitational dependence was observed in posterior-anterior direction (-0.0122 cm-1). Multiple-slice MBW Xe-MRI is feasible in healthy adults using parallel image acquisition. |
| 4104 | Computer 14
|
Highly accelerated 3D hyperpolarized MRI at ultra-high fields using Seiffert Spirals |
| Tobias Speidel1, Stephan Knecht2, Klaus Kolb1, Martin Gierse2, and Volker Rasche1 | ||
1Internal Medicine II, Ulm University Medical Center, Ulm, Germany, 2NVision Imaging Technologies GmbH, Ulm, Germany |
||
Keywords: New Trajectories & Spatial Encoding Methods, Data Acquisition Isotropic 3D imaging of hyperpolarized media is generally challenging due to the extremely limited lifetime of the hyperpolarized phase. Fortunately, images of hyperpolarized media are often intrinsically sparse, resulting in an excellent precondition for non-linear reconstruction techniques such as Compressed Sensing. To preserve this condition, a 3D Seiffert Spiral trajectory was implemented at 11.7 T, yielding low-coherent sampling properties for an arbitrary number of excitations, until depletion of the hyperpolarized signal. We show that qualitative images with a resolution of <500 $$$\mu$$$m can be reconstructed from continuous data that was acquired in less than one second. |
| 4105 | Computer 15
|
Optimizing Hyperpolarized 129Xe MRI of cardiopulmonary oscillations using a Digital Phantom |
| Junlan Lu1, Elianna Bier2, Suphachart Leewiwatwong2, David Mummy3, Sakib Kabir3, Fawaz Alanezi4, Sudarshan Rajagopal4, Scott Haile Robertson5, Peter J Niedbalski6, and Bastiaan Driehuys3 | ||
1Medical Physics, Duke University, Durham, NC, United States, 2Biomedical Engineering, Duke University, Durham, NC, United States, 3Radiology, Duke University, Durham, NC, United States, 4Cardiology, Duke University, Durham, NC, United States, 5Clinical Imaging Physics Group, Duke University, Durham, NC, United States, 6Pulmonary, Critical Care, and Sleep Medicine, University of Kansas Medical Center, Kansas City, KS, United States |
||
Keywords: Data Analysis, Hyperpolarized MR (Gas), Keyhole reconstruction, Simulation Cardiogenic red blood cell (RBC) signal oscillations in 129Xe whole-lung dynamic spectroscopy provide a promising biomarker for identifying patients with pulmonary hypertension (PH). However, to detect more complex and heterogeneous diseases, it is necessary to move from simple global metrics to spatially resolved mapping. This has been demonstrated using keyhole reconstruction methods but little work has yet been done to optimize the reconstruction and visualization of these maps. Here, we introduce a digital phantom to investigate the effects of radial views, key radius, and SNR. From these simulations, we deduce a key radius of 9 points is optimal for minimizing radial undersampling-based heterogeneity and maximizing sensitivity. |
| 4106 | Computer 16
|
Three-dimensional radial echo-planar spectroscopic imaging for in vivo hyperpolarized 13C MRSI at 3 T |
| Marcel Awenius1,2,3, Philipp Biegger1, Christin Glowa4, Melanie Müller1, Renate Bangert1, Helen Abeln1,3,5, Dominik Ludwig1, Tristan A. Kuder1, Mark E. Ladd1,2,6, Peter Bachert1,2, and Andreas Korzowski1 | ||
1Division of Medical Physics in Radiology, German Cancer Research Center (DKFZ), Heidelberg, Germany, 2Faculty of Physics and Astronomy, University of Heidelberg, Heidelberg, Germany, 3Max-Planck-Institute for Nuclear Physics, Heidelberg, Germany, 4Division of Medical Physics in Radiation Oncology, German Cancer Research Center (DKFZ), Heidelberg, Germany, 5Faculty of Chemistry and Earth Sciences, University of Heidelberg, Heidelberg, Germany, 6Faculty of Medicine, University of Heidelberg, Heidelberg, Germany |
||
Keywords: Pulse Sequence Design, Hyperpolarized MR (Non-Gas), Carbon-13 In this study, a novel radially-sampled echo-planar spectroscopic imaging (rEPSI) sequence was implemented at a clinical 3T MR scanner to enable the non-invasive investigation of metabolic processes via hyperpolarized 13C substrates in real-time. Customized data analysis pipelines yielded high quality spectra and volumetric intensity maps for in vivo experiments using hyperpolarized [1-13C]pyruvate. Data extracted from k-space center points enabled a non-localized quantification of T1 values. |
| 4107 | Computer 17
|
Robust and Automated Processing of Deuterium Metabolic Imaging (DMI) using Spatial Prior Knowledge |
| Robin A de Graaf1, Yanning Liu1, Zachary A Corbin1, and Henk M De Feyter1 | ||
1Yale University, New Haven, CT, United States |
||
Keywords: Image Reconstruction, Deuterium Deuterium Metabolic Imaging (DMI) is a novel method to generate spatial maps of dynamic metabolism. While DMI acquisition methods are simple and robust, DMI processing still requires expert user interaction, for example in the removal of extracranial natural abundance 2H lipid signals that interfere with metabolism-linked 2H-lactate formation. Here we pursue the use of MRI-based spatial prior knowledge to provide automated and objective lipid removal. Adequate lipid suppression without perturbation of brain voxels is achieved, thereby allowing the generation of distinct and reliable metabolic maps on patients with brain tumors. |
| 4108 | Computer 18
|
Regional 17O T2* in Human Brain at 3T |
| Hao Song1, Hamidreza Saeidi1, Johannes Fisher1, Ali Caglar Özen1, Stefan Schumann2, and Michael Bock1 | ||
1Dept. of Radiology, Medical Physics, Medical Center University of Freiburg, Faculty of Medicine, University of Freiburg, Freiburg, Germany, 2Dept. of Anesthesiology and Critical Care, Medical Center University of Freiburg, Faculty of Medicine, University of Freiburg, Freiburg, Germany |
||
Keywords: Data Analysis, Non-Proton Information about T2* relaxation time plays an important role in data acquisition for 17O-MRI. In this work, we investigated the in vivo T2* relaxation time constants of 17O in regional brain tissue at 3T. Overall, regional T2* relaxation times are in a range of 1.5 - 2ms (except for CSF). Additionally, a T2* variation of 0.12ms was found from superficial to deep layers in the insular cortex. Our results indicate that short-TE acquisition strategies are favorable for 17O-MRI of the brain. |
| 4109 | Computer 19
|
Signal-to-noise Ratio Enhancement of 31P Magnetic Resonance Spectroscopy using a Pre-trained Deep Learning Model |
| Yeong-Jae Jeon1,2, Kyung Min Nam3,4,5, Alex Bhogal3, and Hyeon-Man Baek1,2 | ||
1Department of Health Sciences and Technology, GAIHST, Gachon University, Incheon, Korea, Republic of, 2Lee Gil Ya Cancer & Diabetes Institute, Gachon University, Incheon, Korea, Republic of, 3Department of Radiology, University Medical Centre Utrecht, Utrecht, Netherlands, 4Institute for Diagnostic and Interventional Neuroradiology, Support Center for Advanced Neuroimaging (SCAN), University of Bern, Bern, Swaziland, 5Translational Imaging Ceter, sitem-insel AG, Bern, Switzerland |
||
Keywords: Data Processing, Spectroscopy, Non-Proton, Animals, Brain, Precision & Accuracy We demonstrate the feasibility of a novel denoising approach utilizing a pretrained deep learning model with multiscale local polynomial smoothing for single voxel 31P MRS data in the mice brain at 9.4T. We evaluated the low-rank denoising, one of the popular methods and the proposed method using LCModel to compare their performance. Both methods resulted in improved signal-to-noise ratio and decreased uncertainty (Cramer-Rao Lower Bounds). In this work, the suggested method outperformed in signal-to-noise ratio enhancement. |
| 4596 | Computer 1
|
EPG-based optimization of the SNR of the phase-cycled bSSFP in phase-based electrical conductivity mapping |
| Julien Lamy1, Flavy Savigny1, and Paulo Loureiro de Sousa1 | ||
1ICube, Université de Strasbourg-CNRS, Strasbourg, France |
||
Keywords: Signal Modeling, Simulations Phase-based electrical conductivity mapping requires high SNR due to the instability of the reconstruction. Electrical conductivity can be measured by the phase-cycled bSSFP, a high-SNR sequence which is also weighted by other biophysical parameters but for which no complete analytical model exist. In this work, we use EPG-based simulations to investigate how the bSSFP signal varies with the flip angle, the phase step and the pulse duration in a diffusive, two-pools model of the brain white matter at 3 T. We show that our simulations agree with in-vivo acquisitions and establish guidelines to optimize the SNR of the phase-cycled bSSFP. |
| 4597 | Computer 2
|
Estimation of the Spatial Gradient of the MR Image from the Diffusion Profile |
| Iman Aganj1,2, Matthew Vera1, Thorsten Feiweier3, Andre J. van der Kouwe1,2, John E. Kirsch1,2, and Bruce R. Fischl1,2 | ||
1Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Boston, MA, United States, 2Radiology Department, Harvard Medical School, Boston, MA, United States, 3Siemens Healthcare GmbH, Erlangen, Germany |
||
Keywords: Signal Modeling, Diffusion Tensor Imaging In the course of diffusion, water molecules experience varying values for the relaxation-time property of the underlying tissue, a factor that has not been accounted for in diffusion MRI (dMRI) modeling. Accordingly, we derive a relationship between the diffusion profile measured by dMRI and the spatial gradient of the image, and subsequently estimate the latter from the former. We test our hypothesized relationship via dMRI of the human brain (a public in vivo image and an acquired ex vivo stimulated-echo image), showing statistically significant results that may be due to our model and/or the confounding factor of “fiber continuity”. |
| 4598 | Computer 3
|
Multi-Continent Analysis of Geometric Distortion in Low-Field MRI |
| Stephen E. Ogier1,2, Annabel Sorby-Adams3, Kalina Jordanova1, Andrew M. Dienstfrey4, Zydrunas Gimbutas4, Jennifer Guo3, John Kirsch5, Steven Schiff6, Ronald Mulondo7, Johnes Obungoloch8, Megan E Poorman9, John Pitts9, W. Taylor Kimberly3, and Kathryn E. Keenan1 | ||
1Magnetic Imaging Group, National Institute of Standards and Technology, Boulder, CO, United States, 2Department of Physics, University of Colorado Boulder, Boulder, CO, United States, 3Department of Neurology and the Center for Genomic Medicine, Massachusetts General Hospital and Harvard Medical School, Boston, MA, United States, 4Mathematical Analysis and Modeling Group, National Institute of Standards and Technology, Boulder, CO, United States, 5A. A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Charlestown, MA, United States, 6Department of Neurosurgery, Yale University School of Medicine, New Haven, CT, United States, 7CURE Children's Hospital of Uganda, Mbale, Uganda, 8Faculty of Applied Sciences and Technology, Mbarara University of Science and Technology, Mbarara, Uganda, 9Hyperfine, Inc., Guilford, CT, United States |
||
Keywords: Artifacts, Low-Field MRI The assessment of brain morphology requires geometric accuracy, and the monitoring of pathologies like dementia requires geometric stability over time and between systems. We aim to assess the geometric accuracy and stability of a head-only 64 mT system by imaging the cylindrical phantom provided by the manufacturer, positioned with a 3D-printed holder for stability, and analyzing the geometric fidelity of the images produced by a variety of product sequences. The initial assessment examines the perimeter of the phantom and how it varies along the S-I direction. |
| 4599 | Computer 4
|
Simultaneous numerical modeling of 13C-MRI enzyme kinetics, diffusion effects, and RF refocusing |
| Dylan Archer Dingwell1,2 and Charles H Cunningham1,2 | ||
1Medical Biophysics, University of Toronto, Toronto, ON, Canada, 2Physical Sciences Platform, Sunnybrook Research Institute, Toronto, ON, Canada |
||
Keywords: Signal Modeling, Hyperpolarized MR (Non-Gas), 13C Signal Dynamics, 13C MRI, Enzyme Kinetics, Reaction Kinetics, Metabolism In this study, we used a particle-based Brownian dynamics simulator with concurrent MR modeling to characterize diffusion-related effects on 13C signal dynamics. We conducted a spectrophotometric lactate dehydrogenase (LDH) activity assay with and without the presence of glass beads to measure the influence of diffusion constraints and spatial compartmentalization on the reaction kinetics of LDH-mediated conversion of pyruvate to lactate. We also investigated the relationship between diffusion and RF refocusing in a typical fast GRE sequence by comparing simulated signal intensities for RF spoiled and non-spoiled GRE sequences applied under varying diffusion conditions. |
| 4600 | Computer 5
|
A dual multi-dimensional integration (dMDI) method for reliable noise masking |
| Yongquan Ye1 | ||
1United Imaging, Houston, TX, United States |
||
Keywords: Signal Modeling, Data Processing A dual multi-dimensional integration (dMDI) method was proposed for reliable image noise masking. By adopting a state-of-the-art MDI method, a signal ratio and a reciprocal signal ratio were constructed in the proposed dMDI method. The product of the two signal ratios display a highly reliable dependency of the intrinsic SNR of the signals, which can serve as a voxel-wise noise mask. |
| 4601 | Computer 6
|
Impact of free-water correction on white matter changes measured by diffusion tensor imaging in migraine |
| Irene Guadilla1,2, Ana Fouto1, Álvaro Planchuelo-Gómez3, Antonio Tristán-Vega3, Amparo Ruiz-Tagle1, Inês Esteves1, Gina Caetano1, Nuno Silva4, Pedro Vilela5, Raquel Gil-Gouveia6,7, Santiago Aja-Fernández3, Patrícia Figueiredo1, and Rita Nunes1 | ||
1Institute for Systems and Robotics - Lisboa and Department of Bioengineering, Instituto Superior Técnico, Universidade de Lisboa, Lisbon, Portugal, 2Universidad Autónoma de Madrid, Madrid, Spain, 3Laboratorio de Procesado de Imagen, Universidad de Valladolid, Valladolid, Spain, 4Learning Health, Hospital da Luz, Lisbon, Portugal, 5Imaging Department, Hospital da Luz, Lisbon, Portugal, 6Neurology Department, Hospital da Luz, Lisbon, Portugal, 7Center for Interdisciplinary Research in Health, Universidade Católica Portuguesa, Lisbon, Portugal |
||
Keywords: Signal Modeling, Diffusion/other diffusion imaging techniques Menstrual migraine affects about 25% of female migraine patients. However, the diagnosis of migraine is particularly difficult because the brain changes associated with migraine are challenging to detect with imaging techniques. Diffusion-weighted MRI (dMRI) permits the detection of alterations in the microenvironment of the brain tissues. We investigate whether removing the contribution of the free water component from the diffusion-signal can provide increased sensitivity to identify white matter changes in migraine using diffusion tensor metrics. |
| 4602 | Computer 7
|
A Simple Analytical Model of B1+ for MRI Measurement of RF Currents in Wire-like Medical Devices |
| Chiara Hartmann1, Mélina Bouldi2, and Jan M Warnking1 | ||
1U1216, Grenoble Institut Neurosciences, Univ. Grenoble Alpes, Inserm, Grenoble, France, 2ZMT Zurich MedTech AG, Zürich, Switzerland |
||
Keywords: Signal Modeling, Modelling Safe access to MRI for patients with active implants may be possible by measuring RF currents in implants in individual patients with low-SAR sequences. We present a simple theoretical model of the B1+-field close to a locally straight wire at any angle to B0. MRI signals fitted by our model closely match magnitude and complex signals from electromagnetic simulations with a wire parallel to B0 or tilted at 45° and magnitude experimental signals with parallel wire. RF currents reconstructed from simulated data match ground-truth values to within 6% and current profiles reconstructed from experimental signals follow simulated currents in shape. |
| 4603 | Computer 8
|
Self-Supervised Model Fitting of VERDICT MRI in the Prostate |
| Snigdha Sen1, Saurabh Singh2, Hayley Pye3, Caroline Moore4, Hayley Whitaker3, Shonit Punwani2, David Atkinson2, Eleftheria Panagiotaki1, and Paddy J Slator1 | ||
1Centre for Medical Image Computing, University College London, London, United Kingdom, 2Centre for Medical Imaging, University College London, London, United Kingdom, 3Molecular Diagnostics and Therapeutics Group, University College London, London, United Kingdom, 4Department of Urology, University College London Hospitals NHS Foundation Trust, London, United Kingdom |
||
Keywords: Signal Modeling, Microstructure Microstructure models are traditionally fitted via computationally expensive non-linear least squares. Recent model fitting techniques use supervised deep learning algorithms trained on synthetic datasets, however the training data distribution affects parameter estimates. Self-supervised learning can address this by extracting training labels directly from the input data. We introduce a self-supervised machine learning algorithm for fitting the VERDICT MRI model for prostate to diffusion-weighted MRI. On simulated data, our approach improves parameter estimation compared to non-linear least squares and supervised machine learning. We also reveal plausible tumour contrast on in-vivo prostate data. |
| 4604 | Computer 9
|
Partial Volume Averaging of Diffusion Tensor Imaging in the Fornix |
| Ken Sakaie1, Ajay Nemani2, Mark Lowe2, and Katherine Koenig2 | ||
1The Cleveland Clinic, Cleveland, OH, United States, 2Imaging Institute, The Cleveland Clinic, Cleveland, OH, United States |
||
Keywords: Signal Modeling, Multiple Sclerosis Diffusion Tensor Imaging of the fornix promises to serve as an imaging biomarker for cognitive decline in multiple sclerosis and Alzheimer’s disease. As the fornix is surrounded by cerebrospinal fluid in the lateral ventricles, partial volume averaging can bias measures of tissue microstructure. Here, we use a high spatial resolution, multishell acquisition to test the methods for combating partial volume averaging in the fornix. |
| 4605 | Computer 10
|
Denoising Magnetic Resonance Spectroscopy Signals with Stack Auto-encoder Networks |
| Jing Wang1, Bing Ji1, Yang Lei1, Tian Liu2, Hui Mao1, and Xiaofeng Yang1 | ||
1Emory University, Atlanta, GA, United States, 2Icahn School of Medicine at Mount Sinai, New York, NY, United States |
||
Keywords: Signal Modeling, Spectroscopy This abstract reports a novel and efficient self-supervised deep learning autoencoder network for denoising MRS data and improving signal-to-noise ratio (SNR) of spectra, which may enable rapid MRS data acquisition and improving its clinical workflow and applications. |
| 4606 | Computer 11
|
Distinguish Injured from Intact Axons in Coherent Fiber Bundles Using Network Estimation Operator (NEO) |
| Jacob Blum1, Tsen-Hsuan Lin1, Donsub Rim2, and Sheng-Kwei Song 1 | ||
1Radiology, Washington University School of Medicine in Saint Louis, Saint Louis, MO, United States, 2Department of Mathematics and Statistics, Washington University in Saint Louis, Saint Louis, MO, United States |
||
Keywords: Signal Modeling, Diffusion Tensor Imaging Deep Neural Network based Network Estimation Operator (NEO) was developed to differentiate and quantify parallel fibers of different axial diffusivities. Data from both Monte Carlo simulations and an In-Silico imaging phantom were analyzed by both DBSI and NEO. Results reveal that NEO distinguished parallel fibers of different axial diffusivity while DBSI only identified a single fiber. NEO can detect and quantify injured vs. non-injured axons in a coherent fiber bundle. Thus, NEO can quantify the extent of axonal injury and loss more accurately than DBSI and other advanced diffusion MRI models. |
| 4607 | Computer 12
|
Calibration of R2' relaxation rate and hepatic iron concentration by Monte Carlo simulations |
| Xiaoben Li1, Mengyuan Ma1, Jinyang Wang1, Junying Cheng2, Wenjun Yao3, and Changqing Wang1 | ||
1School of Biomedical Engineering, Anhui Medical University, Hefei, China, 2Department of MRI, The First Affiliated Hospital of Zhengzhou University, Zhengzhou, China, 3Department of Radiology, The Second Affiliated Hospital of Anhui Medical University, Hefei, China |
||
Keywords: Signal Modeling, Liver The R2' relaxation rate is expected to become a more specific biomarker in hepatic iron overload because it is not affected by pathological factors. However, calibrating the relationship between R2' relaxation rate and hepatic iron concentration (HIC) at different field strengths is time-consuming. In this work, we develop Monte Carlo simulations to investigate relationship between R2' relaxation rate and HIC in hepatic iron overload at both 1.5T and 3.0T. Results show strong linear relationships between R2' predictions and HIC. The Monte Carlo simulations may greatly shorten the clinical calibration time of MRI relaxation rates and HIC. |
| 4608 | Computer 13
|
Toward an accurate MR measure of the axon radii by accounting for myelin-specific attenuation of the diffusion MRI signal in the brain |
| Stefania Oliviero1, Cosimo Del Gratta1, Andrada Costantina Traeba2, Tommaso Boccato3, Caterina Mainero2, and Nicola Toschi2,3 | ||
1Neuroscience, Imaging and Clinical Sciences, and Institute for Advanced Biomedical Technologies, University of Chieti-Pescara G. D'Annunzio, Chieti, Italy, 2A.A. Martinos Center for Biomedical Imaging and Harvard Medical School, Boston, MA, United States, 3Biomedicine and Prevention, University of Rome Tor Vergata, Roma, Italy |
||
Keywords: Signal Modeling, White Matter, myelin, diffusion, in silico, MR An accurate, noninvasive measure of axonal radii could play a crucial role in the understanding of healthy and diseased neural processes. Several diffusion-weighted MRI (dMRI) methods report on the axon radii distribution or mean axon radii. However, these measurements systematically overestimate histologically derived values. In this preliminary study, we address this limitation by formulating a model which explicitly includes the myelin contribution to the cerebral dMR signal. In detail, we introduce a myelinic compartment in both the AxCaliber and ActiveAx models and demonstrate a significant improvement of axonal radius estimation using synthetic data simulation. |
| 4609 | Computer 14
|
Data-Driven Discovery of mechanical models directly from k-space data: initial results with an actively forced linear elastic system. |
| David G.J. Heesterbeek1, Max H.C. van Riel1, Cornelis A.T. van den Berg1, and Alessandro Sbrizzi1 | ||
1Department of Radiotherapy, Computational Imaging Group for MR Therapy and Diagnostics, University Medical Center Utrecht, Utrecht, Netherlands |
||
Keywords: Signal Representations, Elastography, Motion estimation, Dynamic system We propose a framework for data-driven discovery of the governing biomechanical equations for soft tissue directly from k-space data. This framework is based on principles from Spectro-dynamics, a novel method to infer dynamical information directly from severely undersampled spectral data with high spatiotemporal resolution, and sparse regression. As a proof-of-concept, the underlying equations of motion are identified for an actively forced linear elastic system using experimentally obtained k-space data. The systems studied in this abstract are successfully identified with errors of less than 2.5% on the system parameters, and their respective displacement fields are reconstructed accurately. |
| 4610 | Computer 15
|
Research on Denoising of Magnetic Resonance Spectrum Based on Exponential Decomposition Constraint |
| Jinyu Wu1, Tianyu Qiu1, Zhangren Tu1, Di Guo2, and Xiaobo Qu1 | ||
1Department of Electronic Science, National Institute for Data Science in Health and Medicine, Xiamen University, Xiamen, China, 2School of Computer and Information Engineering, Xiamen University of Technology, Xiamen, China |
||
Keywords: Data Processing, PET/MR Nuclear Magnetic Resonance (NMR) has been a frequently-used analytical tool in many areas of modern biology, chemistry and medicine for decades. However, it is usually limited by a low Signal-to-Noise ratio (SNR). In practical applications, Signal Averaging (SA) with repeated samplings is required to improve the signal-to-noise ratio, which greatly increases the scanning time. In this paper, based on the characteristic that NMR time-domain signals can be decomposed exponentially, a model for denoising NMR spectroscopy based on exponential decomposition constraints is proposed. It can effectively improve the denoising ability and therefore save the scanning time. |
| 4611 | Computer 16
|
Estimate the Linear Effect of Inter-Scanner Variability: Insight from Paired Cross-Scanner T1-weighted Images |
| Chang-Le Chen1, Mahbaneh Eshaghzadeh Torbati2, Weiquan Luo3, Davneet Minhas3, Charles Laymon3, Seong Jae Hwang4, Ciprian Crainiceanu5, Pauline Maillard6, Evan Fletcher6, Charles DeCarli6, Howard Aizenstein1,7, and Dana Tudorascu7,8 | ||
1Department of Bioengineering, University of Pittsburgh, Pittsburgh, PA, United States, 2Intelligent System Program, School of Computing and Information, University of Pittsburgh, Pittsburgh, PA, United States, 3Department of Radiology, University of Pittsburgh, Pittsburgh, PA, United States, 4Department of Artificial Intelligence, Yonsei University, Seoul, Korea, Republic of, 5Department of Biostatistics, Johns Hopkins University, Baltimore, MD, United States, 6Department of Neurology, University of California Davis, Davis, CA, United States, 7Department of Psychiatry, University of Pittsburgh, Pittsburgh, PA, United States, 8Department of Biostatistics, University of Pittsburgh, Pittsburgh, PA, United States |
||
Keywords: Data Processing, Data Processing, Harmonization To broaden our knowledge of inter-scanner variability, we established a pipeline to linearly estimate the site effect by incorporating ComBat modeling and superpixel parcellation. We found that the variation prominently manifested in tissue contrast, noise level, and field inhomogeneity with respect to cross-site data. We further used the estimated parameters to harmonize images. The image quality and structural similarity of cross-scanner data can be improved after the harmonization procedure, and the variation in volumetric measures can also be reduced. This study provides further insight for the research focusing on the development of image harmonization methods. |
| 4612 | Computer 17
|
PINN with Divergence-Free Vector Potential for Velocity Fields Denoising in 4D flow MRI |
| Javier Bisbal1,2,3, Joaquin Mura4, Julio Sotelo1,3,5, Hernán Mella3,6, Cristóbal Arrieta1,2,3, Pablo Irarrazaval1,2,3,7, and Sergio Uribe1,3,8 | ||
1Biomedical Imaging Center, Pontificia Universidad Catolica de Chile, Santiago, Chile, 2Department of Electrical Engineering, Pontificia Universidad Catolica de Chile, Santiago, Chile, 3Millennium Institute for Intelligent Healthcare Engineering, iHEALTH, Santiago, Chile, 4Department of Mechanical Engineering, Universidad Técnica Federico Santa Maria, Santiago, Chile, 5School of Biomedical Engineering, Universidad de Valparaíso, Valparaíso, Chile, 6School of Electrical Engineering, Pontificia Universidad Católica de Valparaiso, Valparaíso, Chile, 7Institute for Biological and Medical Engineering, Pontificia Universidad Católica de Chile, Santiago, Chile, 8Department of Radiology, School of Medicine, Pontificia Universidad Católica de Chile, Santiago, Chile |
||
Keywords: Data Processing, Machine Learning/Artificial Intelligence 4D flow MRI suffers from different sources of noise and aliasing artifacts. Recent denoising techniques are time-consuming or dependent of parameter estimation. We developed a physics informed neural network with divergence-free vector potential as a non-parametric denoising technique for 4D flow MRI. Results from simulated pulsatile flow and CFD vascular model shows significant noise reduction and aliasing correction. Future work includes comparison with other techniques on different types of data and uncertainty quantification. |
| 4613 | Computer 18
|
Cluster Based Sparse Variational Minimization for Multi-Compartment Dictionary Fitting to bSSFP Signal Profiles |
| Berk Can Acikgoz1,2, Adèle L.C. Mackowiak1,2,3, Nils M.J. Plähn1,2, Yasaman Safarkhanloo4, Eva S. Peper1,2, Tom Hilbert3,5,6, Giulia M.C. Rossi3, and Jessica A.M. Bastiaansen1,2 | ||
1Department of Diagnostic, Interventional and Pediatric Radiology (DIPR), Inselspital, Bern University Hospital, University of Bern, Bern, Switzerland, 2Translation Imaging Center (TIC), Swiss Institute for Translational and Entrepreneurial Medicine, Bern, Switzerland, 3Department of Radiology, Lausanne University Hospital (CHUV) and University of Lausanne (UNIL), Lausanne, Switzerland, 4Advanced Clinical Imaging Technology,Department of Cardiology, Inselspital, Bern University Hospital, Bern, Switzerland, 5Siemens Healthineers International AG, Lausanne, Switzerland, 6LTS5, École Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland |
||
Keywords: Data Processing, Quantitative Imaging, phase-cycled bSSFP, fat fraction A novel cluster-based sparse total-variation minimization (CSTV) dictionary matching algorithm for multi-compartment parameter mapping applied to a phase-cycled balanced steady state free precession (PC-bSSFP) signal profile is presented. A dictionary encoding off-resonance frequency and relaxation time ratio is matched to the measured profiles with the proposed method to recover the underlying fat fraction, proton density, and water-fat-separated images simultaneously. The performance of the proposed method is tested and compared with a Laplacian regularized non-negative least squares dictionary matching method in phantom and in vivo experiments. |
| 4614 | Computer 19
|
Differential quantification of cortical and medullary perfusion through Gaussian Mixture Model based segmentation on transplanted renal MRI |
| Anne Oyarzun-Domeño1,2, Izaskun Cía 1, Rebeca Echeverria-Chasco2,3, María A. Fernández-Seara2,3, Paloma L. Martin-Moreno2,4, Nuria Garcia-Fernandez2,4, Gorka Bastarrika2,3, Javier Navallas1,2, and Arantxa Villanueva1,2,5 | ||
1Electrical, Electronics and Communications Engineering, Public University of Navarre, Pamplona, Spain, 2Health Research Insitute of Navarra, IdiSNA, Pamplona, Spain, 3Department of Radiology, Clínica Universidad de Navarra, Pamplona, Spain, 4Department of Nephrology, Clínica Universidad de Navarra, Pamplona, Spain, 5Institute of Smart Cities (ISC), Pamplona, Spain |
||
Keywords: Data Processing, Perfusion Renal perfusion quantification is of importance in the post-operative surveillance of the allograft in translated patients. Together with cortical perfusion measurement, there is a strong interest in the quantification of medullary perfusion values, which requires an additional segmentation step of renal compartments. We applied Gaussian Mixture Models over renal MRI dataset to automatically extract the labels for each compartment to separately calculate cortical and medullary perfusion values. Proposed method showed performance metrics above 85% against ground truth labels and correlation coefficient above 96% and 58% for cortical and medullary perfusion values comparing with ground truth perfusion values. |
| 4615 | Computer 20
|
The effect of AIF and tissue time-course temporal alignment on pharmacokinetic model trans-cytolemmal water exchange sensitivity |
| Xin Li1, Wei Huang1, William D. Rooney1, and Charles S Springer1 | ||
1Advanced Imaging Research Center, Oregon Health & Science University, Portland, OR, United States |
||
Keywords: Data Analysis, DSC & DCE Perfusion Using DCE modeling to quantify prostate transcytolemmal water exchange effect faces the challenges of often insufficient interstitial contrast agent concentration as well as suboptimal temporal resolution. Using ultra-high temporal resolution prostate DCE data, the goal of this work is to investigate the impact of DCE-MRI temporal alignment between the arterial input function (AIF) and the tissue time-courses on water exchange quantification. Results here show that when perfect alignment is not achievable in practice, an earlier AIF bolus arrival bias will generally reduce the capability for water-exchange quantification using DCE-MRI. |
| 4616 | Computer 21
|
Non-Linear Reconstruction for Coil Sensitivity Calibration from Cartesian and non-Cartesian Data |
| H. Christian M. Holme1 and Martin Uecker1,2,3,4 | ||
1Institute of Biomedical Imaging, Graz University of Technology, Graz, Austria, 2Institute for Diagnostic and Interventional Radiology, University Medical Center Göttingen, Göttingen, Germany, 3Partner Site Göttingen, DZHK (German Centre for Cardiovascular Research), Göttingen, Germany, 4Cluster of Excellence “Multiscale Bioimaging: from Molecular Machines to Networks of Excitable Cells” (MBExC), University of Göttingen, Göttingen, Germany |
||
Keywords: Parallel Imaging, Parallel Imaging We compare the current de-facto standard for coil sensitivity calculation, ESPIRiT, to a non-linear reconstruction method, ENLIVE. While ENLIVE normally produces both images and coil sensitivies, we focus here on using it for calibration of the coil sensitivities. We applied ESPIRiT and ENLIVE to low-resolution subsets of a 3D Cartesian and a non-Cartesian spiral acquisition, using the resoluting coil profiles in a linear parallel imaging reconstruction. We showed that ENLIVE is substantially faster for a high number of channels and provides improved sensitivities in the non-Cartesian example. |
| 4617 | Computer 22
|
Fast and robust parallel imaging algorithm for Spiral MRI with Fat-Water separation |
| Tzu Cheng Chao1, Xi Peng1, Dinghui Wang1, and James G. Pipe1 | ||
1Department of Radiology, Mayo Clinic, Rochester, MN, United States |
||
Keywords: Parallel Imaging, Parallel Imaging Three different models to reconstruct fat and water spiral images with parallel acquisition are compared in terms of computational efficiency and reconstruction quality. The best reconstruction is obtained with a full model requiring the most computational time, while the model with the fastest computational time results in some degradation of image quality. A third, preferred model preserves most of the image quality of the first model, with nearly the same speedup as the latter model. |
| 4618 | Computer 23
|
Joint K-b space reconstruction of under-sampled diffusion-weighted MRI |
| Yan Dai1, Jie Deng1, and Xun Jia2 | ||
1University of Texas Southwestern Medical Center, Dallas, TX, United States, 2Johns Hopkins University, Baltimore, MD, United States |
||
Keywords: Image Reconstruction, Diffusion/other diffusion imaging techniques We developed a joint image reconstruction method in both k-space and b-space to reconstruct the under-sampled diffusion-weighted images acquired at different b-values and generate the corresponding apparent diffusion coefficient map simultaneously by solving an optimization problem. This method improved SNR in both diffusion weighted images and apparent diffusion coefficient map compared with the conventional method and allows 50% k-space data undersampling that has the potential of reducing image distortion and shortening acquisition time. |
| 4619 | Computer 24
|
Time-Resolved MRI Kinesiology: Reconstructing force and velocity during active loading |
| Max H.C. van Riel1, Cornelis A.T. van den Berg1, and Alessandro Sbrizzi1 | ||
1Department of Radiotherapy, Computational Imaging Group for MR Diagnostics and Therapy, UMC Utrecht, Utrecht, Netherlands |
||
Keywords: Signal Modeling, MSK, Time-Resolved, Velocity, Force, Motion Force and velocity are two important quantities for studying muscles, but retrieving these quantities using MRI is challenging. Spectro-Dynamic MRI allows the characterization of dynamical systems at a high spatial and temporal resolution directly from k-space data. Using an iterative reconstruction algorithm, we can reconstruct time-resolved MR images, time-resolved motion fields, mechanical parameters, and an activation force, at a temporal resolution of 11 ms. As a proof-of-principle, MR data was acquired with a motion phantom moving like an actively driven linear elastic system. All dynamic variables could be reconstructed accurately. This method could become useful for studying muscles under dynamic loads. |
| 4620 | Computer 25
|
GRAPPA Interpolation with Unity Acceleration Enhances Signal-to-Noise Ratio in Fully Sampled Acquisitions |
| Andrew S Nencka1,2 | ||
1Radiology, Medical College of Wisconsin, Milwaukee, WI, United States, 2Center for Imaging Research, Medical College of Wisconsin, Milwaukee, WI, United States |
||
Keywords: Image Reconstruction, Data Acquisition, De-noising Significant efforts are focused on maximizing SNR in high spatial and temporal resolution image acquisitions. We present a method of utilizing GRAPPA to de-noise k-space observations made with a multi-channel array. By replacing a fully sampled observed k-space with a k-space in which each observation is inferred from a weighted average of neighboring observations across all coils, an increase of SNR on the order of 5% is achieved. |
| 4621 | Computer 26
|
Spatially Fourier Excited Acquisition and Reconstruction (SFEAR) for Improved Slice Acceleration |
| Negin Yaghmaie1,2, Warda Syeda3, Yasmin Blunck1,2, Bahman Tahayori4, Rebecca K. Glarin2,5, Bradford A. Moffat2,6, and Leigh A. Johnston1,2 | ||
1Department of Biomedical Engineering, The University of Melbourne, Melbourne, Australia, 2Melbourne Brain Centre Imaging Unit, The University of Melbourne, Melbourne, Australia, 3Melbourne Neuropsychiatry Centre, The University of Melbourne, Melbourne, Australia, 4The Florey Institute of Neuroscience and Mental Health, Melbourne, Australia, 5Department of Radiology, Royal Melbourne Hospital, Melbourne, Australia, 6Department of Medicine and Radiology, The University of Melbourne, Melbourne, Australia |
||
Keywords: New Trajectories & Spatial Encoding Methods, New Trajectories & Spatial Encoding Methods A new 3D Spatially Fourier Excited Acquisition and Reconstruction method (SFEAR) is introduced, in which double-RF pulses are used to excite sinusoidally-modulated slice profiles across a slab. Target spatial frequencies are achieved by varying the time shift between the two pulses, to construct the slice-phase-encode dimension of 3D k-space from Fourier excited acquisitions. SFEAR outperforms conventional GRAPPA in the slice-phase encode dimension, given its inherent ability to undersample sine or cosine components rather than whole planes of 3D k-space. Superior reconstruction of undersampled data and lower g-factor values are demonstrated in both 7T phantom and in vivo data. |
| 4622 | Computer 27
|
Water removal in MR spectroscopic imaging with Casorati Singular Value Decomposition |
| Amirmohammad Shamaei1, Jana Starcukova1, Jedrek Burakiewicz 2, and Zenon Starcuk 1 | ||
1Institute of Scientific Instruments of the Czech Academy of Sciences Research institute in Brno, Brno, Czech Republic, 2Tesla Dynamic Coils, Zaltbommel, Netherlands |
||
Keywords: Sparse & Low-Rank Models, Data Processing, Singular value decomposition, MR spectroscopic imaging, Water removal Removing residual water from the MRSI datasets using the SVD-based algorithms is computationally demanding. We present a novel algorithm to reduce the computing time required for water removal in MRSI data. Our proposed method exploits low-rank structures that exist in MRSI data. It arranges the MRSI data in the Casorati matrix form, applies singular value decomposition, and removes residual water from the most prominent left-singular vectors. We compared our proposed method with the HLSVDPRO method, and we achieved 20x acceleration while improving effectiveness. Our proposed method is publicly available as a pip-installable Python tool. |
| 4623 | Computer 28
|
In Vivo xSPEN Imaging with a Model-Based Reconstruction for Efficient Spatial Encoding in a Single-Sided Prostate MRI Scanner |
| Muller De Matos Gomes1, Meredith Sadinski1, Alek Nacev1, and William Allyn Grissom2 | ||
1Promaxo, Oakland, CA, United States, 2Biomedical Engineering, Vanderbilt University, Nashville, TN, United States |
||
Keywords: New Trajectories & Spatial Encoding Methods, Image Reconstruction, Low-Field MRI, Spatiotemporal Encoding Single-sided low-field MRI scanners can provide image guidance during interventions such as prostate biopsies without restricting surgical access, but efficient spatial encoding is a challenge. In this work we show how spatiotemporal encoding using the xSPEN method combined with a model-based image reconstruction enables swapping a conventionally phase encoded but large-matrix-size image dimension with a conventionally frequency encoded but small-matrix-size image dimension. The model-based reconstruction yields images free of distortions due to gradient non-linearity and shows overall improved in vivo male pelvic image quality. |
| 4624 | Computer 29
|
Accelerated 3D Multi-Echo Spin-Echo sequence for T2 mapping of mouse brain metastases using a subspace-based reconstruction |
| Aurélien J. TROTIER1, Nadège Corbin1, Sylvain Miraux1, and Emeline J. Ribot2 | ||
1Centre de Résonance Magnétique et Systèmes Biologiques, UMR5536, CNRS, University of Bordeaux, Bordeaux, France, 2CRMSB - University of Bordeaux / CNRS, Bordeaux, France |
||
Keywords: Pulse Sequence Design, Sparse & Low-Rank Models T2 mapping is an important biomarker for pre-clinical imaging to characterize the grows of metastasis but suffers from long acquisition time when performed in 3D. Undersampling of the k-space combined to advanced iterative reconstruction using the temporal redondancy of the data along the echo train could be beneficial to reach a reasonable scan time. This works demonstrates that a subspace-based reconstruction for T2 mapping can be used to accelerate a 3D Multi-Echo Spin-Echo acquisition for in-vivo mouse brain metastasis quantification with an acceleration factor up to 10 while keeping an isotropic spatial resolution of 156µm. |
| 4625 | Computer 30
|
New Theory and Faster Computations for Subspace-Based Sensitivity Map Estimation |
| Rodrigo A. Lobos1, Chin-Cheng Chan1, and Justin P. Haldar1 | ||
1University of Southern California, Los Angeles, CA, United States |
||
Keywords: Parallel Imaging, Sparse & Low-Rank Models Sensitivity map estimation is important in many multichannel MRI applications. Subspace-based sensitivity map estimation methods like ESPIRiT are popular and perform well, though can be computationally expensive and their theoretical principles can be nontrivial to understand. In this work, we derive a new theoretical framework for sensitivity map estimation from a structured low-rank modeling perspective. This results in an estimation approach that is equivalent to ESPIRiT, but with theory that may be more intuitive for some readers. In addition, we propose a set of computational acceleration techniques that enable substantial (~25-fold) improvements in computational time for subspace-based sensitivity map estimation. |
| 4626 | Computer 31
|
Accelerated Calibrationless MR Parametric Mapping by a Space-Contrast-Coil-Domain Locally Low-Rank Tensor Constraint |
| Juan Gao1, Sha Hua2, Xin Tang1, Haiyang Chen1, Yixin Emu1, and Chenxi Hu1 | ||
1Institute of Medical Imaging Technology, School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China, 2Department of Cardiovascular Medicine, Ruijin Hospital Lu Wan Branch, Shanghai Jiao Tong University School of Medicine, Shanghai, China |
||
Keywords: Sparse & Low-Rank Models, Quantitative Imaging Parametric mapping is routinely used in cardiac MR, yet its resolution is relatively low due to the single-shot acquisition. Most existing acceleration methods exploit the space-contrast-domain and coil-domain information redundancy separately e.g. by combining LLR and SENSE. Here we propose a novel calibrationless parametric mapping acceleration technique based on a Locally Low-Rank Tensor (LLRT) modeling of the signal in the space-contrast-coil domain, which exploits the information redundancy over all 3 dimensions jointly. In vivo studies show that the method generates more accurate reconstructions than the LLR-based algorithm. Moreover, a nonuniform LLRT penalty further improves the reconstruction quality by reducing blurring. |
| 4627 | Computer 32
|
Cramér-Rao Bound Optimized Linear Bases for Low-Rank Subspace Reconstruction |
| Andrew Mao1,2,3, Sebastian Flassbeck1,2, Cem Gultekin4, and Jakob Asslaender1,2 | ||
1Center for Biomedical Imaging, New York University Grossman School of Medicine, New York, NY, United States, 2Center for Advanced Imaging Innovation and Research, New York University Grossman School of Medicine, New York, NY, United States, 3Vilcek Institute of Graduate Biomedical Sciences, New York University Grossman School of Medicine, New York, NY, United States, 4Courant Institute of Mathematical Sciences, New York University, New York, NY, United States |
||
Keywords: Sparse & Low-Rank Models, Magnetization transfer, MR Fingerprinting, Hybrid State, Cramer-Rao bound, Quantitative Imaging, Low-Rank Reconstruction This works extends the traditional framework for estimating low-rank bases that maximize preserved signal energy to additionally preserve the Cramér-Rao bound of the biophysical parameters in quantitative imaging. To this end, we orthogonalize the signal's derivatives wrt. the model parameters and incorporate them into the basis estimation process. We demonstrate in silico an improvement in the Cramér-Rao bound of all biophysical parameters with negligible cost to signal energy preservation, which translates to improved image quality and SNR in vivo. |
| 4628 | Computer 33
|
Improving Spiral Deblurring with Square Kernels and Low-Pass Preconditioning |
| Dinghui Wang1, Tzu Cheng Chao1, and James G Pipe1 | ||
1Department of Radiology, Mayo Clinic, Rochester, MN, United States |
||
Keywords: Image Reconstruction, Artifacts, spiral imaging, deblurring, off-resonance Advantages of spiral imaging include fast scan speed and high SNR efficiency. Success implementation of spiral imaging relies on utilization of efficient deblurring methods. The goal of this work is to improve the performance of a previous deblurring method, using square kernels and low-pass preconditioning. Data from phantom and volunteers demonstrate that artifacts can be reduced while computational demand is also reduced by the proposed kernels.
|
| 4629 | Computer 34
|
The Problem of Hidden Noise in MR Image Reconstruction |
| Jiayang Wang1, Di An1, and Justin Haldar1 | ||
1Signal and Image Processing Institute, Ming Hsieh Department of Electrical and Computer Engineering, University of Southern California, Los Angeles, CA, United States |
||
Keywords: Image Reconstruction, Machine Learning/Artificial Intelligence The performance of modern image reconstruction methods is commonly judged using quantitative error metrics like mean squared-error and the structural similarity index, where these error metrics are calculated by comparing a reconstruction against fully-sampled reference data. In practice, this reference data contains noise and is not a true gold standard. In this work, we demonstrate that this “hidden noise” can confound performance assessment methods, leading to image quality degradations when typical error metrics are used to tune image reconstruction performance. We also demonstrate that a new error metric, based on the non-central chi distribution, helps resolve this issue. |
| 4630 | Computer 35
|
K-space Based Motion Estimation for polar fMRI using Transfer learning |
| Faeze makhsousi1, Vahid Ghodrati2, Morteza Homayounfar1, sina ghaffarzadeh1, and abbas Nasiraei-Moghaddam3 | ||
1Biomedical Engineering, Amirkabir University of Technology (Tehran Polytechnic), Tehran, Iran (Islamic Republic of), 2University of California, Los Angeles, Los Angeles, CA, United States, 3Amirkabir University of Technology (Tehran Polytechnic), Tehran, Iran (Islamic Republic of) |
||
Keywords: Motion Correction, Brain, Data Analysis The motion of the head during functional MRI is an unavoidable issue that adversely affects brain mapping. Radial reading of the k-space reduces the problem to some extent but not completely. Residual motion, even at a partial pixel level, has a measurable effect on the spatial frequencies and so can be estimated directly from the k-space data. This work uses a transfer learning-based approach to estimate the head motion from radially acquired k-space information. Results showed a good agreement with the statistical parametric mapping (SPM) package. |
| 4631 | Computer 36
|
Concatenated multi-contrast wavelet-based compressed sensing reconstruction. |
| Gabriel Varela-Mattatall1,2, Jaejin Cho3,4, Omer Oran5, Corey A. Baron1,2, Berkin Bilgic3,4, and Ravi S. Menon1,2 | ||
1Centre for Functional and Metabolic Mapping (CFMM), Robarts Research Institute, Western University, London, ON, Canada, 2Department of Medical Biophysics, Schulich School of Medicine and Dentistry, Western University, London, ON, Canada, 3Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital & Harvard Medical School, Boston, MA, United States, 4Harvard-MIT Division of Health Sciences and Technology, Massachusetts Institute of Technology, Boston, MA, United States, 5Siemens Healthcare Limited, Oakville, ON, Canada |
||
Keywords: Image Reconstruction, Image Reconstruction There is a family of multi-contrast sequences such as MEGRE, MP2RAGE and 3D-QALAS which produce a limited number of contrasts, and their corresponding reconstructions in highly accelerated acquisitions are not optimally represented by either compressed sensing, low rank, or both reconstruction styles together. In this work, we explore a novel way to perform compressed sensing for simultaneous multi-contrast reconstruction. By using a concatenation operator in the regularization term, we can perform a single compressed sensing reconstruction with automatic parameter selection to reconstruct all contrasts at once and provide an implicit form of parallelization for multi-contrast reconstruction. |
| 4632 | Computer 37
|
Comparison of coil combination technique performance for phase preservation |
| Krithika Balaji1, Peter J Lally2,3, Zimu Huo4, Michael Mendoza4, Michael N Hoff5, and Neal K Bangerter4 | ||
1Bioengineering, Imperial College London, London, United Kingdom, 2Department of Brain Sciences, Imperial College London, London, United Kingdom, 3UK Dementia Research Institute Centre for Care Research and Technology, London, United Kingdom, 4Department of Bioengineering, Imperial College London, London, United Kingdom, 5Department of Radiology and Biomedical Imaging, University of California, San Francisco, San Francisco, CA, United States |
||
Keywords: Data Processing, Cartilage, Comparisons, bSSFP MRI images are typically acquired using multiple receive coils, which introduce spatially varying phase offsets to each coil image. Many techniques have been developed to combine these coil images. For a variety of reconstruction techniques, phase preservation is necessary post-combination but it is unclear which method best achieves this. This work compared ESPIRiT, Simple Phase RCC, Full Phase RCC, 3T Siemens’ Adaptive Combine and IMPA using phase-cycled bSSFP images. Phase preservation was evaluated using the elliptical signal model theory and characteristic bSSFP phase plots. ESPIRiT consistently produced combined images with phase characteristics most similar to those from a single coil. |
| 4633 | Computer 38
|
Standardization of Containerized “MRD Apps” for Reproducible and Deployable Research |
| Kelvin Chow1, Tess Wallace2, Alexander Fyrdahl3, Peter Kellman4, Hui Xue4, Florian Knoll5, and Adrienne E Campbell-Washburn4 | ||
1Siemens Medical Solutions USA, Inc., Chicago, IL, United States, 2Boston Children's Hospital, Boston, MA, United States, 3Karolinska Institutet, Karolinska, Sweden, 4National Institutes of Health, Bethesda, MD, United States, 5Friedrich-Alexander Universität, Erlangen, Germany |
||
Keywords: Software Tools, Reproductive Reproducibility and translation of advancements in MR image reconstruction and analysis algorithms have been limited by incompatibilities between heterogeneous software environments. Containerization of “MRD Apps” packages software together with all dependent libraries into a single ready-to-use container. Standardization of the container format enables broad inter-compatibility and source code examples are provided for Python, MATLAB, and C++ through Gadgetron. A neural network was implemented in the MRD Apps format using Python and run directly on the scanner with vendor-provided integration. MRD Apps can be published alongside manuscripts as reference implementations or used to streamline image reconstruction and analysis challenge contests. |
| 4634 | Computer 39
|
Exploiting diffusion MRI data redundancy on a denoising framework - application on MR images (OGSE, μFA protocols) of the Marmoset Brain (in-vivo) |
| Vinicius P. Campos1, Tales Santini2, Corey Baron2, and Marcelo A. C. Vieira1 | ||
1Computer and Electrical Engineering, University of São Paulo, São Carlos, Brazil, 2Robarts Research Institute, Western University, London, ON, Canada |
||
Keywords: Data Processing, Data Analysis, Denoise Oscillating gradients spin-echo (OGSE) and microscopic fractional anisotropy (μFA) are diffusion MRI advanced techniques able to provide additional information of the microstructures of the brain1-5, when compared to traditional diffusion MRI. However, high-resolution DWI images present low signal to noise ratio (SNR). In this work, we presented a different framework, named VST_dMRI_BM4D, for denoising dMRI data by exploiting data redundancy and using the variance stabilization transformation (VST) concept7-9. Results show the proposed method is comparable to, and in some cases superior than, MPPCA6, potentially making it a useful tool to be used in the dMRI field. |
| 4635 | Computer 41
|
Fat/water separated & SMS accelerated pseudo 3D PROPELLER |
| Ola Norbeck1,2, Henric Rydén1,2, Adam van Niekerk1, Enrico Avventi1,2, Tim Sprenger3, Stefan Skare1,2, Johan Berglund4, and Mikael Skorpil 1,2 | ||
1Karolinska Instituet, Stockholm, Sweden, 2Karolinska University Hospital, Stockholm, Sweden, 3MR Applied Science Laboratory Europe, GE Healthcare, Munich, Germany, 4Uppsala University Hospital, Uppsala, Sweden |
||
Keywords: Data Acquisition, Data Acquisition, MSK, Dixon, SMS By combining thin slices, PROPELLER acquisition, SMS acceleration and chemical shift encoding with asymmetrical gradients reformattable pseudo 3D volumes can be acquired. These volumes are, in contrast to 3D RARE volumes, efficiently chemical shift encoded, less susceptible to T2-blurring and less sensitive to motion. |
| 4636 | Computer 42
|
FSE Dixon with bandwidth-matched asymmetric readouts tailored for real-valued fat/water estimates |
| Henric Rydén1, Mikael Skorpil2, Matea Borbas3, and Adam van Niekerk1 | ||
1Clinical Neuroscience, Karolinska Institutet, Stockholm, Sweden, 2Karolinska Institutet, Stockholm, Sweden, 3Karolinska University Hospital, Stockholm, Sweden |
||
Keywords: Pulse Sequence Design, Fat, Dixon A novel asymmetric readout waveform for Dixon FSE imaging is presented for efficient sampling, tailored for real-valued estimation of fat/water estimates. |
| 4637 | Computer 43
|
SuperMRF: Deep Robust Acceleration for MR Fingerprinting |
| Hongyu Li1, Brendan L. Eck2, Mingrui Yang2, Jeehun Kim2, Ruiying Liu1, Peizhou Huang3, Dong Liang4, Xiaojuan Li2, and Leslie Ying1,3 | ||
1Department of Electrical Engineering, University at Buffalo, State University of New York, Buffalo, NY, United States, 2Department of Biomedical Engineering, Program of Advanced Musculoskeletal Imaging (PAMI), Cleveland Clinic, Cleveland, OH, United States, 3Department of Biomedical Engineering, University at Buffalo, State University of New York, Buffalo, NY, United States, 4Paul C. Lauterbur Research Center for Biomedical Imaging, Medical AI research center, SIAT, CAS, Shenzhen, China |
||
Keywords: MR Fingerprinting/Synthetic MR, Image Reconstruction We propose a novel, deep learning-based method, “SuperMRF”, for the reconstruction of MR Fingerprinting (MRF) parametric maps that enables rapid image reconstruction. Built upon a convolutional neural network, SuperMRF uses three loss functions to incorporate additional information from the Bloch equations, estimated maps, de-aliasing, and data consistency losses. We investigate the use of SuperMRF for further acceleration of data acquisition by reducing the number of MRF time frames. Our results demonstrate that proposed SuperMRF is robust to noise and can achieve a 20x reduction in acquired MRF time frames. Tissue property maps can be reconstructed in less than one second. |
| 4638 | Computer 44
|
Accelerated MRI using intelligent protocolling and subject-specific denoising |
| Keerthi Sravan Ravi1,2, Gautham Nandakumar3, Nikita Thomas3, Mason Lim3, Enlin Qian2,4, Marina Manso Jimeno2,4, Pavan Poojar5, Zhezhen Jin6, Patrick Quarterman7, Maggie Fung7, Girish Srinivasan3, John Thomas Vaughan Jr.2, and Sairam Geethanath5 | ||
1Biomedical Engineering, Columbia University, New York, NY, United States, 2Columbia Magnetic Resonance Research Center, Columbia University, New York, NY, United States, 3PhenoMX, Chicago, IL, United States, 4Department of Biomedical Engineering, Columbia University, New York, NY, United States, 5Accessible MR Laboratory, Biomedical Engineering and Imaging Institute, Dept. of Diagnostic, Molecular and Interventional Radiology,, Mt. Sinai, New York, NY, United States, 6Department of Biostatistics, Columbia University Irving Medical Center, New York, NY, United States, 7MR Clinical Solutions, GE, New York, NY, United States |
||
Keywords: Data Acquisition, Data Acquisition The routine brain screen protocol employed at our institution was accelerated using Look Up Tables to achieve a 1.94x gain in imaging throughput. Image-denoising was performed on accelerated data by leveraging deep learning models trained on contrast-specific publicly available datasets. These were corrupted by native noise during forward modeling. In addition, subject-specific denoising was demonstrated. The superior performance of denoised data on automated volumetry of Alzheimer’s Disease (AD) relevant brain anatomies on T1w data demonstrated potential for accelerated AD imaging. |
| 4639 | Computer 45
|
The Impact of Acceleration Factors of Compressed Sensing on the Image Quality of the Fast Spin Echo Diffusion Weighted Imaging for the Skull Base |
| Haonan Zhang1, Qingwei Song1, Jiazheng Wang2, and Ailian Liu1 | ||
1Department of Radiology, the First Affiliated Hospital of Dalian Medical University, Dalian, Dalian, China, 2Philips Healthcare, Beijing, China, Beijing, China |
||
Keywords: Brain Connectivity, Brain Compared with echo planar imaging diffusion weighted iamging (EPI-DWI), turbo spin echo diffusion weighted imaging (TSE-DWI) can significantly reduce magnetic sensitivity artifacts in skull base imaging. However, the longer scan time limits its clinical promotion. The purpose of this study is to investigate the effect of the compression sensing acceleration factor on the image quality of TSE-DWI in the skull base area. |
| 4640 | Computer 46
|
Highly accelerated acquisition of MP2RAGE using Compressed SENSE |
| Yang Zhao1, Yishi Wang2, Guangbin Wang1, and Weibo Chen2 | ||
1Department of Radiology, Shandong Provincial Hospital Affiliated to Shandong First Medical University, Jinan, China, 2Philips Healthcare, Beijing, China |
||
Keywords: Data Acquisition, Data Acquisition, compressed sensing The magnetization-prepared 2 rapid acquisition gradient echo (MP2RAGE) sequence provides morphological T1-weighted images and quantitative T1 maps. Compressed sensing (CS) acceleration technique has been drawing extensive attention with its high acquisition efficiency for MRI sequence. In this study, volume measurement and ROI analysis were used to assess the images obtained from CS-MP2RAGE sequence under different acceleration factors. We find that CS-MP2RAGE show high similarity in brain structure volumes measurement and ROI analysis with traditional SENSE-MP2RAGE. Those findings prove that CS-MP2RAGE can be considered as an alternative to the SENSE-MP2RAGE, though the specific acceleration factor still needs to be further studied. |
| 4641 | Computer 47
|
The effect of acceleration factor on brain magnetic resonance imaging based on artificial intelligence compressed sensing technology |
| shuai hu1, haonan zhang1, nan wang1, qingwei song1, and ailian liu1 | ||
1the First Affiliated Hospital of Dalian Medical University, dalian, China |
||
Keywords: New Devices, Brain ACS based cranial MRI has good feasibility in clinical use. Scanning time is reduced by 37%, 40%, and 71% for recommended AF in T1-FLAIR, T2WI, and T2-FLAIR, while the image quality meets the requirement of diagnosis. |
| 4642 | Computer 48
|
Effect of regularization parameter on model-based deep learning framework for accelerated MRI |
| Sampada Bhave1, Saurav Sajib1, Aniket Pramanik2, Mathews Jacob2, and Samir Sharma1 | ||
1Canon Medical Research USA Inc, Mayfield, OH, United States, 2University of Iowa, Iowa City, IA, United States |
||
Keywords: Image Reconstruction, Image Reconstruction Model-based deep learning algorithms offer high quality reconstructions for accelerated acquisitions. Training the regularization parameter λ can lead to instabilities during training. In this work, we evaluated effect of fixing λ parameter while training. We observed no difference in image quality when the network was trained with a fixed λ parameter when the fixed value was equal to the value learned from training. We observed that IQ is dependent on the fixed λ values used during training. Furthermore, we observed that tuning the λ parameter during inference adapts the framework to the SNR of the testing dataset, yielding improved performance. |
| 4643 | Computer 49
|
Comparison of Compressed Sensing Accelerated Rosette UTE and Conventional 31P 3D MRSI at 3T in Leg Muscle |
| Brian Bozymski1, Xin Shen2, Ali Caglar Ozen3, Serhat Ilbey3, Albert Thomas4, Mark Chiew5, William Clarke5, Ulrike Dydak1,6, and Uzay Emir1,7 | ||
1School of Health Sciences, Purdue University, West Lafayette, IN, United States, 2Department of Radiology and Biomedical Imaging, University of California San Francisco, San Francisco, CA, United States, 3Department of Radiology and Radiotherapy, Medical Center - University of Freiburg, Freiburg, Germany, 4Department of Radiology, University of California, Los Angeles, CA, United States, 5Wellcome Centre for Integrative Neuroimaging, University of Oxford, Oxford, United Kingdom, 6Department of Radiology and Imaging Sciences, Indiana University School of Medicine, Indianapolis, IN, United States, 7Weldon School of Biomedical Engineering, Purdue University, West Lafayette, IN, United States |
||
Keywords: Data Acquisition, Muscle Quantitative comparison of novel, rosette trajectory UTE (70 μs) and conventional weighted Cartesian 3D 31P MRSI sequences is performed at 3T in quadriceps muscle. After previous validation, five healthy volunteers were scanned by both sequences without interruption. Fitted metabolite maps and SNR calculations of PCr and ATP signals across selected slices and voxels displayed competitive performance between each acquisition. Retrospective compressed sensing (CS) acceleration results suggest this UTE MRSI may enable faster, higher resolution 31P metabolite mapping in leg muscle and other organs of interest. |
| 4644 | Computer 50
|
Accelerating myelin water fraction imaging using compressed sensing and BMC-mcDESPOT |
| Maryam Alsameen1, Zhaoyuan Gong1, John Laporte1, Mary Faulkner1, Mohammad Akhonda1, and Mustapha Bouhrara1 | ||
1Laboratory of Clinical Investigations, National Institiute on Aging, National Institutes of Health, Baltimore, MD, United States |
||
Keywords: Data Analysis, Brain We demonstrated the feasibility of compressed sensing (CS) to accelerate myelin water fraction (MWF) imaging using the BMC-mcDESPOT method. Our results showed that derived MWF maps using CS were similar to those derived without CS. Findings from this study indicate that whole brain, high resolution, MWF map can be derived within a few minutes. |
| 4645 | Computer 51
|
Acceleration of Phase-contrast Magnetic Resonance Venogram by Compressed SENSE |
| Yukun Zhang1, Na Liu1, Geli Hu2, Liangjie Lin2, Yanwei Miao1, and Qingwei Song1 | ||
1the First Affiliated Hospital of Dalian Medical University, Dalian, China, 2Clinical and Technical Support, Philips Healthcare, beijing, China |
||
Keywords: Parallel Imaging, Vessels, compressed sensitivity encoding The clinical application of phase-contrast magnetic resonance venogram (PC-MRV) is limited by the long scan time. This study aims to accelerate the acquisition of PC-MRV by the compressed sensitivity encoding (CS-SENSE) technique and find the optimal acceleration factor. Results show that the image quality of fast PC-MRV based on CS-SENSE was significantly higher than that based on conventional sensitivity encoding (SENSE) technology, and the CS-SENSE factor 9 was recommended for mild to moderate patients, and CS AF=12 for critical patients. |
| 4646 | Computer 52
|
Accelerated parameter mapping in the k-p domain via nonconvex low rank constraint |
| Kang Yan1 and Craig H Meyer1,2 | ||
1Biomedical Engineering, University of Virginia, Charlottesville, VA, United States, 2Radiology & Medical Imaging, University of Virginia, Charlottesville, VA, United States |
||
Keywords: Image Reconstruction, Relaxometry A nonconvex low rank regularization (NLR) was proposed to accelerate parameter mapping in the k-p domain. The NLR uses weighted nuclear norm minimization (WNNM) to obtain an optimized solution by differently penalizing singular values, in comparison to traditional low rank methods. The performance of the proposed algorithm was demonstrated for T2 mapping of the kidney. Our study demonstrated that the proposed algorithm outperformed k-p domain-based compressed sensing and L&S algorithms. |
| 4647 | Computer 53
|
Accelerated MRI near metallic implants at 0.55T using hexagonal sampling |
| Bahadir Alp Barlas1, Kubra Keskin1, Brian A Hargreaves2,3,4, and Krishna S Nayak1,5 | ||
1Electrical and Computer Engineering, University of Southern California, Los Angeles, CA, United States, 2Radiology, Stanford University, Stanford, CA, United States, 3Electrical Engineering, Stanford University, Stanford, CA, United States, 4Bioengineering, Stanford University, Stanford, CA, United States, 5Biomedical Engineering, University of Southern California, Los Angeles, CA, United States |
||
Keywords: Data Acquisition, Low-Field MRI Metallic implants cause severe distortions in the magnetic field that are best mitigated using multi-spectral acquisitions that require longer scan time. At conventional field strengths, this is compensated by aggressive use of parallel imaging. However, at low-field strengths (<1.5T), available parallel imaging factors are reduced because body noise dominance prevents the use of dense arrays with smaller elements. Hexagonal sampling (in ky,kz space) was previously proposed as a way to achieve 2-fold reduction in imaging time without introducing additional artifacts. In this work, we demonstrate applicability of the hexagonal sampling for 0.55T multi-spectral imaging near metal. |
| 4648 | Computer 54
|
Development of a self-supervised machine learning algorithm for automatic MRI sequence-type classification |
| Seongwon Na1, Yousun Ko2, Su Jung Ham1, Mi-Hyun Kim1, Youngbin Shin1, Yu Sub Sung1, Jimi Huh3, Seung Chai Jung1, and Kyung Won Kim1 | ||
1Asan Medical Center, Seoul, Korea, Republic of, 2University of Ulsan College of Medicine, Seoul, Korea, Republic of, 3Ajou University Hospital, Suwon, Korea, Republic of |
||
Keywords: Machine Learning/Artificial Intelligence, Machine Learning/Artificial Intelligence, Sequence; Classification For automatic sequence-type classification of brain MRI, we developed the self-supervised machine learning (ML) algorithm, named ImageSort-net, using a rule-based labeling system based on metadata of Digital Imaging and Communications in Medicine (DICOM) image files. Our rule-based labeling system and ImageSort-net showed high classification performance to predict brain MRI sequence type. ImageSort-net showed reliable performance by appending a new dataset to an existing dataset and without human labeling of the whole dataset. This result indicates that sustainable self-learning ML algorithms using the rule-based virtual label in the new datasets are feasible. |
| 4649 | Computer 55
|
SOC-GRAPPA: An MPSoC based Accelerator for GRAPPA |
| Abdul Basit1, Omair Inam1, and Hammad Omer1 | ||
1Electrical Engineering, MIPRG, Comsats University Islamabad, Islamabad, Pakistan |
||
Keywords: Parallel Imaging, Cardiovascular, Accelerator, Real-time GRAPPA is a cartesian pMRI method widely adopted for high-speed image reconstruction in many clinical applications e.g. real-time cardiac MRI. However, general-purpose computers have limited processing capabilities to address the computational complexity of GRAPPA in real-time MRI. Recently, multi-processor system-on-chip (MPSoC) has emerged as a potential candidate to meet the rising computational demands of GRAPPA for real-time image reconstruction. In this paper, design and implementation details of a novel MPSoC based GRAPPA accelerator i.e. SOC-GRAPPA are presented. The experimental results of 18-coil cardiac dataset show that the proposed accelerator is capable of reconstructing 30 frames/second in real-time cardiac MRI. |
| 4650 | Computer 56
|
Validation of a deep-learning-based method for accelerating susceptibility-weighted imaging in clinical subjects |
| Xiao Wu1, Shan Xu2, Yao Zhang2, Jianzhong Sun2, and Peiyu Huang2 | ||
1Radiology, The Second Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, China, 2The Second Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, China |
||
Keywords: Data Acquisition, Machine Learning/Artificial Intelligence, deep-learning ,susceptibility-weighted imaging,magnetic resonance imaging In this study, we validated a deep-learning-based method for accelerating susceptibility-weighted imaging (SWI) in 31 clinical subjects. Compared to the fully sampled images, the accelerated SWI images had less noise and imaging artifacts. Although the images had decreased sharpness, the anatomical details of the lesions were mostly kept, and we had not observed false negative\positive lesions. This method could be useful for clinical situations that need timely imaging results. |
| 4651 | Computer 57
|
Accelerated Four-Dimensional Rosette J-resolved Spectroscopic Imaging (4D ROSE-JRESI) with semi-LASER Localization: A pilot study |
| Ajin Joy1, Uzay Emir2,3, Paul M. Macey4, and M. Albert Thomas1 | ||
1Radiological Sciences, University of California, Los Angeles, Los Angeles, CA, United States, 2School of Health Sciences, Purdue University, West Lafayette, IN, United States, 3Weldon School of Biomedical Engineering, Purdue University, West Lafayette, IN, United States, 4School of Nursing and Brain Research Institute, University of California, Los Angeles, Los Angeles, CA, United States |
||
Keywords: Data Acquisition, Brain Undersampling spatial and spectral dimensions is essential to achieve clinically feasible scan times in multi-dimensional spectroscopic imaging. Sampling pattern of rosette trajectory has higher incoherence than that of the other non-Cartesian trajectories like spiral and radial, and can achieve higher compressed-sensing reconstruction performance. While rosette spectroscopic imaging has been attempted for 2D (2 spatial+1 spectral) and 3D (3 spatial+1 spectral) spectroscopic imaging, it has thus far not been shown for J-resolved spectroscopic imaging. In this pilot study, we implemented a rosette echo-planar J-resolved spectroscopic imaging (ROSE-JRESI) sequence and studied the feasibility of high acceleration factors for clinically feasible runtimes. |
| 4652 | Computer 58
|
In-vivo image acceleration with an 8-channel local B0 coil array and parallel imaging in a 9.4T human MR scanner |
| Rui Tian1, Theodor Steffen1, and Klaus Scheffler1,2 | ||
1High-Field MR center, Max Planck Institute for Biological Cybernetics, Tuebingen, Germany, 2Department for Biomedical Magnetic Resonance, University of Tübingen, Tuebingen, Germany |
||
Keywords: New Trajectories & Spatial Encoding Methods, New Trajectories & Spatial Encoding Methods, nonlinear gradient encoding We further developed a recent idea called spread spectrum MRI to reduce the sampling time by rapidly modulating spins with localized magnetic fields during signal readout. Given phantom experiments tested and safety evaluation for human subjects performed, this time, we started in-vivo measurements of human head with multi-slice FLASH sequence accelerated by local B0 coil modulations, and examined the reconstructed image quality. Sinusoidal modulation schemes in various phase offset patterns and frequencies given different scanner bandwidth were tested and compared, which were shown to boost the image acceleration from 6-fold (i.e., SENSE only) to about 8-fold in one phase-encoded dimension. |
| 4653 | Computer 59
|
Improved Fast Whole-Brain High Resolution Diffusion Imaging on 7 Tesla MRI |
| Merry Mani1, Xinzeng Wang2, and Baolian Yang3 | ||
1University of Iowa, Iowa City, IA, United States, 2GE Healthcare, Houston, TX, United States, 3GE Healthcare, Waukesha, WI, United States |
||
Keywords: Brain Connectivity, High-Field MRI Achieving high resolution in diffusion MRI is challenging due to its inherently low SNR, vulnerability to motion and other EPI-related artifacts. On higher field-strengths, the SNR advantage can be exploited to push the resolution if the echo-time can be effectively reduced and the increased number of slices can be efficiently acquired. Combining acceleration techniques such as multi-band and parallel imaging is critical for this approach. Here we combine two deep-learned reconstruction priors, one pertaining to the q-space and another pertaining to image artifact removal, into a model-based iterative reconstruction framework to improve the quality of highly accelerated high-resolution 7T DWIs. |
| 4654 | Computer 60
|
Fast SMWI via denoising for nigral hyperintensity detection in Parkinson’s disease |
| Jonghyo Youn1, Juhyung Park1, Sooyeon Ji1, Hongjun An1, Hwan Heo2, MyeongOh Lee2, Soohwa Song2, Eung Yeop Kim3, and Jongho Lee1 | ||
1Seoul National University, Seoul, Korea, Republic of, 2Heuron Co.Ltd., Incheon, Korea, Republic of, 3Radiology, Samsung Medical Center, Seoul, Korea, Republic of |
||
Keywords: Image Reconstruction, Parkinson's Disease Nigral hyperintensity detection in substantia nigra is a potential biomarker for PD. An advanced SWI method, SMWI, has demonstrated reliable detection of the hyperintensity at 3T but it requires a 4 m 15 s scan protocol which is too long for PD patients, suffering from motion artifacts. In this study, we developed a new 2 m 56 s protocol by reducing phase FOV and applying deep learning-powered denoising to compensate for SNR loss from the shorter scan. The new protocol was validated using both simulated and real data via an automated tool and an expert radiologist. |
| 4771 | Computer 1
|
Parallel imaging reconstruction using iterative CNN-based denoising in image domain |
| Tomoki Amemiya1, Atsuro Suzuki1, Yukio Kaneko1, Suguru Yokosawa1, and Toru Shirai1 | ||
1Imaging Technology Center, FUJIFILM Corporation, Tokyo, Japan |
||
Keywords: Parallel Imaging, Image Reconstruction We propose an iterative reconstruction method of parallel imaging using convolutional neural network (CNN)-based denoising in the image domain and data-consistency processing in k-space. The proposed method reduces the noise and artifacts of a reconstructed image compared with the iterative method using sparsity of wavelet transform, suggesting that using CNN-based denoising in iterative reconstruction is effective in reducing noise in parallel imaging. |
| 4772 | Computer 2
|
Deep learning-based Lorentzian fitting of WASSR Z-spectra |
| Sajad Mohammed Ali1, Nirbhay Yadav2, Ronnie Wirestam1, Munendra Singh3, Hye-Young Heo3, Peter van Zijl2, and Linda Knutsson4 | ||
1Medical Physics, Lund University, Lund, Sweden, 2Radiology, F.M. Kirby Research Center, Johns Hopkins University, Kennedy Krieger Institute, Baltimore, MD, United States, 3Radiology, Johns Hopkins University, Baltimore, MD, United States, 4F.M. Kirby Research Center, Radiology, Medical Radiation Physics, Kennedy Krieger Institute, Johns Hopkins University, Lund University, Baltimore, MD, United States |
||
Keywords: Data Analysis, Machine Learning/Artificial Intelligence, Lorentzian curve fitting Water saturation shift referencing (WASSR) Z-spectra can be used to correct shifts due to B0-field inhomogeneities, for magnetic susceptibility mapping and analysis of relaxation effects. The spectra follow a Lorentzian shape with discrete values. Hence, a Lorentzian fit to retrieve the shape parameters (amplitude A, line width LW and frequency shift ΔfH2O ) simplifies analysis. Conventionally, the least-squares (LS) method is used for such fitting despite being time consuming and sensitive to the unavoidable noise in vivo. We propose a deep learning-based Lorentzian-fitting neural network (LoFNet) that demonstrated improved robustness against noise and sampling density in combination with reduced time consumption. |
| 4773 | Computer 3
|
Accuracy of Deep Learning-based Signal-to-noise Measurements using Air Recon DL |
| Evan McNabb1, Véronique Fortier1,2,3,4, and Ives R. Levesque3,4 | ||
1Medical Imaging, McGill University Health Centre, Montreal, QC, Canada, 2Diagnostic Radiology, McGill University, Montreal, QC, Canada, 3Gerald Bronfman Department of Oncology, McGill University, Montreal, QC, Canada, 4Medical Physics Unit, McGill University, Montreal, QC, Canada |
||
Keywords: Data Acquisition, Data Analysis Noise estimates in Deep Learning-based image reconstruction (DLR) from background regions lead to artificially low noise estimates and thus inaccurate SNR. Noise estimate accuracy was improved using the difference between two identical acquisitions. SNR increased in DLR compared to standard reconstruction in clinical sequences over a range of acquired signal averages and voxel sizes. With DLR, varying signal averages had little effect on the measured SNR in fast spin echo acquisitions, while SNR increased by more than three-fold in single-shot fast spin echo. However, with DLR, SNR no longer follows predicable behavior, therefore sequence optimizations need to be performed experimentally. |
| 4774 | Computer 4
|
Deep Learning Prediction of Multi-channel ESPIRiT Maps for Calibrationless MR Image Reconstruction |
| Junhao Zhang1,2, Zheyuan Yi1,2,3, Yujiao Zhao1,2, Linfang Xiao1,2, Jiahao Hu1,2,3, Vick Lau1,2, Fei Chen3, Alex T.L.Leong1,2, and Ed X. Wu1,2 | ||
1Laboratory of Biomedical Imaging and Signal Processing, the University of Hong Kong, HongKong, China, 2Department of Electrical and Electronic Engineering, the University of Hong Kong, HongKong, China, 3Department of Electrical and Electronic Engineering, Southern University of Science and Technology, Shenzhen, China |
||
Keywords: Parallel Imaging, Data Acquisition, Brain reconstruction, Cardiac reconstruction We present a U-Net based deep learning model to estimate the multi-channel ESPIRiT maps directly from uniformly-undersampled multi-channel multi-slice MR data. The model is trained with a hybrid loss function using fully-sampled multi-slice axial brain datasets from the same MR receiving coil system. The proposed model robustly predicted ESPIRiT maps from uniformly-undersampled k-space brain and cardiac MR data, yielding highly comparable performance to reconstruction using to acquired reference ESPIRiT maps. Our proposed method presents a general strategy for calibrationless parallel imaging reconstruction through learning from coil and protocol specific data. |
| 4775 | Computer 5
|
CS-MRI Reconstruction Using an Improved Generative Adversarial Network with Channel Attention Mechanism |
| Xia Li1, Hui Zhang2, and Tie-Qiang Li3 | ||
1Information Engineering, China Jiliang University, Hangzhou, China, 2Information Engineering, China Jiliang University, HANGZHOU, China, 3Karolinska Institute, Stockholm, Sweden |
||
Keywords: Image Reconstruction, Brain, Compressed sensing MRI Generative adversarial network (GAN) has emerged as one of the most prominent approaches for fast CS-MRI reconstruction. However, most deep-learning models achieve performance by increasing the depth and width of the networks, leading to prolonged reconstruction time and difficulty to train. We have developed an improved GAN-based model to achieve quality performance without increasing complexity by implementing the following: 1) dilated-residual structure with different dilation rates at different depth of the networks; 2) CAM to adjust the allocation of network resources; 3) multi-scale information fusion module to achieve feature fusion. Experiment data have confirmed the validity for the modules. |
| 4776 | Computer 6
|
A Stochastic Approach for Joint Learning of a Neural Network Reconstruction and Sampling Pattern in Cartesian 3D Parallel MRI |
| Marcelo V. W. Zibetti1 and Ravinder R. Regatte1 | ||
1Radiology, NYU Grossman School of Medicine, New York, NY, United States |
||
Keywords: Data Acquisition, Image Reconstruction This work proposes a stochastic variation of the bias-accelerated subset selection (BASS) algorithm to learn an efficient sampling pattern (SP) for accelerated MRI. This algorithm is used in the joint learning of an SP and neural network reconstruction. We apply the proposed approach to two different 3D Cartesian parallel MRI problems. The proposed stochastic approach, when used for joint learning, improves the learning speed from 2.5X to 5X, obtaining SPs with similar properties as the non-stochastic approach with nearly the same RMSE and SSIM. |
| 4777 | Computer 7
|
Analysis of Deep Learning-based Reconstruction Models for Highly Accelerated MR Cholangiopancreatography: to Fine-tune or not to Fine-tune |
| Jinho Kim1,2, Thomas Benkert2, Bruno Riemenschneider1, Marcel Dominik Nickel2, and Florian Knoll1 | ||
1Artificial Intelligence in Biomedical Engineering, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany, 2MR Application Pre-development, Siemens Healthcare GmbH, Erlangen, Germany |
||
Keywords: Machine Learning/Artificial Intelligence, Image Reconstruction, Magnetic Resonance Cholangiopancreatography MR cholangiopancreatography (MRCP) is a special MRI technique to visualize the biliary systems. Deep Learning-based (DL) reconstruction models have shown to reduce scan time from many anatomical regions. However, they generally require large training datasets. This is challenging for applications like MRCP, where public datasets are not available. This work analyzes two approaches to training a DL model for highly accelerated MRCP reconstruction: training from scratch using a small MRCP dataset and fine-tuning a model pretrained on a public knee dataset. Results show that despite of the substantial data domain shift between training and testing, fine-tuning outperformed training from scratch. |
| 4778 | Computer 8
|
Improvement of CNN Denoising Performance Using Noise Control of Input Image and Application to Parallelized Image Denoising |
| Satoshi ITO1 and Keitaro TAKAHASHI1 | ||
1Graduate Program in Information, Electrical and Electronic Systems Engineering, Utsunomiya University, Utsunomiya, Japan |
||
Keywords: Data Processing, Data Processing In noise reduction, the lower the amount of noise, the lower the image degradation associated with the noise reduction. Using this property, we propose a new denoising scheme that can improve denoising performance using trained CNN. Noise suppressed image by low-pass filter is inputted to denoising CNN and then output image is enhanced by high-pass filter. Experimental results show that noise suppression of the input image improves both image structure preservation and noise processing performance. The proposed method was applied to a parallel blind image denoising method. As a results, further improvement in performance was shown. |
| 4779 | Computer 9
|
Fast spatio-temporal subspace reconstruction of 3D-MRF with B0 correction and deep-learning-initialized compressed sensing (Deli-CS) |
| Natthanan Ruengchaijatuporn1,2, Siddharth Srinivasan Iyer3,4,5, Sophie Schauman3,4, Quan Chen3,4, Xiaozhi Cao3,4, Itthi Chatnuntawech6, and Kawin Setsompop3,4 | ||
1Center of Excellence in Computational Molecular Biology, Faculty of Medicine, Chulalongkorn University, Bangkok, Thailand, 2Center for Artificial Intelligence in Medicine, Faculty of Medicine, Chulalongkorn University, Bangkok, Thailand, 3Department of Radiology, Stanford University, Stanford, CA, United States, 4Department of Electrical Engineering, Stanford University, Stanford, CA, United States, 5Department of Electrical Engineering and Computer Science, MIT, Cambridge, MA, United States, 6National Nanotechnology Center, National Science and Technology Development Agency, Pathum Thani, Thailand |
||
Keywords: Sparse & Low-Rank Models, Machine Learning/Artificial Intelligence Recent advances in spatio-temporal subspace reconstruction has enabled accurate reconstruction from highly accelerated scans. Nevertheless, such methods suffer from being computationally intensive due to their iterative nature coupled with the large dimensionality of the problem, especially when imperfection correction is incorporated into the formulation. This abstract proposes deep-learning-initialized compressed sensing (Deli-CS) to accelerate such spatio-temporal reconstruction by providing it with a deep-learning-reconstructed initial solution, reducing the number of iterations required. Using MRF as an example, Deli-CS reconstructs data from a rapid 1-mm isotropic whole-brain TGAS-SPI-MRF with time-segmented B0 correction at 3x faster speed compared to FISTA. |
| 4780 | Computer 10
|
Deep Image Prior with Structured Sparsity (DISCUS) for Dynamic MRI Reconstruction |
| Muhammad Ahmad Sultan1, Chong Chen1, Yuchi Han1, and Rizwan Ahmad1 | ||
1The Ohio State University, Columbus, OH, United States |
||
Keywords: Image Reconstruction, Data Processing, Image Reconstruction, Unsupervised Learning, Deep Image Prior We propose an unsupervised learning method for dynamic MRI reconstruction. Our method, Deep Image prior with StruCtUred Sparsity (DISCUS), is an extension of Deep Image Prior (DIP) and employs joint optimization of network parameters and latent code vectors to recover image series. We enforce group sparsity on code vectors to reveal the underlying low-dimensional manifold of the image series. Using two sets of in vivo measurements and digital phantom simulation, we show that our approach significantly improves the reconstruction quality in terms of Normalized Mean Square Error (NMSE) and Structure Similarity Index Measure (SSIM) compared to compressed sensing and DIP. |
| 4781 | Computer 11
|
Deep Attention Unfolding CNN Architecture for Parallel MRI |
| Muhammad Shafique1,2, Sohaib Ayyaz Qazi3, Faisal Najeeb1, and Hammad Omer1 | ||
1Electrical and Computer Engineering, COMSATS University Islamabad, Islamabad, Pakistan, 2Electrical Engineering, University of Poonch Rawalakot, Rawalakot AJK, Pakistan, 3Center for Medical Image Science and Visualization, Linköping University, Linköping, Sweden |
||
Keywords: Image Reconstruction, Artifacts, Parallel MRI, Deep Learning, Attention Mechanism Deep learning has made significant progress in recent years, however the local receptive field in CNN raises questions about signal synthesis and artefact compensation. This paper proposes a deep learning Spatial-Channel Attention U-Net (SCA-U-Net) to solve the folding problems arising in MR images because of undersampling. The main idea is to enhance local features in the images and restrain the irrelevant features at the spatial and channel levels. The output from SCA-U-Net is further refined by adding a small number of originally acquired low-frequency k-space data. Experimental results show a better performance of the SCA-U-Net model than classical U-Net model. |
| 4782 | Computer 12
|
Improved Clinical Diffusion Weighted Imaging by Combining Deep Learning Reconstruction, Partial Fourier, and Super Resolution |
| Thomas Benkert1, Elisabeth Weiland1, Simon Arberet2, Majd Helo1, Fasil Gadjimuradov1,3, Karl Engelhard4, Gregor Thoermer1, and Dominik Nickel1 | ||
1MR Application Predevelopment, Siemens Healthcare GmbH, Erlangen, Germany, 2Digital Technology & Innovation, Siemens Medical Solutions USA, Princeton, NJ, United States, 3Pattern Recognition Lab, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany, 4Institute of Radiology, Martha-Maria Hospital, Nuremberg, Germany |
||
Keywords: Diffusion/other diffusion imaging techniques, Translational Studies Diffusion weighted imaging (DWI) has found widespread use in daily clinical routine but can still be limited by long acquisition times and low spatial resolution. In this work, combining deep learning-based k-space to image reconstruction with super resolution processing tailored to support partial Fourier acquisitions is demonstrated to efficiently mitigate these obstacles. The approach is shown for various applications, including liver, breast, prostate, and brain DWI at 0.55T, 1.5T, and 3T. |
| 4783 | Computer 13
|
Clinical feasibility of artificial intelligence-assisted compressed sensing for accelerated MR imaging in nasopharyngeal carcinoma |
| Qin Zhao1, Song Zhang1, Liyun Zheng2, and Yongming Dai3 | ||
1Department of Radiology, Sun Yat-sen University Cancer Center, State Key Laboratory of Oncology in South China, Collaborative Innovation Center for Cancer Medicine, Guangdong Key Laboratory of Nasopharyngeal Carcinoma Diagnosis and Therapy, Guangzhou, China, 2Shenzhen United Imaging Research Institute of Innovative Medical Equipment, Shenzhen, China, 3MR Collaboration, Central Research Institute, United Imaging Healthcare, Shanghai, China |
||
Keywords: Data Acquisition, Machine Learning/Artificial Intelligence To improve the scanning efficiency of magnetic resonance imaging (MRI) for nasopharyngeal carcinoma, this study investigated the value of accelerating technique, artificial intelligence-assisted compressed sensing (ACS), in comparison to conventional sequences without accelerating technique and accelerating MRI using parallel imaging (PI). Eleven patients diagnosed with nasopharyngeal carcinoma were prospectively enrolled. As a result, ACS achieved the shortest acquisition time, with similar or even better image quality and SNR than conventional sequences. ACS has the potential to provide sufficient image quality for T1- and T2-weighted imaging in nasopharyngeal carcinoma and could be an alternative to conventional sequences in clinical practice. |
| 4784 | Computer 14
|
USING DEEP LEARNING WITH AN ACTIVE LEARNING APPROACH TO CORRECT WATER-FAT MIS-LABELING IN MR THORACIC SPINE IMAGES |
| Siddhartha Satpathi1, Jacinta E. Browne1, AbdulRahman Alfayad1, Jared T. Verdoorn1, and James G. Pipe1 | ||
1Mayo Clinic, Rochester, MN, United States |
||
Keywords: Artifacts, Spinal Cord Upon investigating a dataset of 804 studies for spinal fractures from two major vendors, the authors observed that 11% of the water-fat images in the studies are mis-labelled. This motivated the development of an automated algorithm to correct the mis-labelling. We used a 2D CNN based deep learning model to classify the images correctly with the aim of reducing error and fatigue in clinical diagnosis caused by such mis-labeling, as well as providing correct labels for further AI workflow. We also demonstrated the use of active learning in this problem by achieving the same test-error with fewer labels than using the entire training data. |
| 4785 | Computer 15
|
Lightweight encoder-decoder architecture for Fat-Water separation in MRI using biophysical model-guided deep learning method |
| Ganeshkumar M1, Devasenathipathy Kandasamy2, Raju Sharma2, and Amit Mehndiratta1,3 | ||
1Centre for Biomedical Engineering, Indian Institute of Technology Delhi, New Delhi, India, 2Department of Radio Diagnosis, All India Institute of Medical Sciences, New Delhi, India, 3Department of Biomedical Engineering, All India Institute of Medical Sciences, New Delhi, India |
||
Keywords: Machine Learning/Artificial Intelligence, Image Reconstruction, Deep Learning, Fat-Water seperation In this study, we propose a novel lightweight encoder-decoder architecture for deep learning based Fat-Water separation in multi-echo MRI data. The architecture's performance is evaluated in the biophysical model-guided deep learning-based Fat-Water separation task and compared against the widely used U-Net. This biophysical model-guided deep learning-based Fat-Water separation requires no training data and ground truths, but it involves time-consuming loss minimization for thousands of epochs. Despite having significantly fewer training parameters, the proposed architecture performed equally well in generating the Fat-Water maps compared to the U-Net. So, our proposed architecture aids in the faster generation of Fat-Water maps. |
| 4786 | Computer 16
|
Fast MR imaging with distribution convergence modeling |
| Jing Cheng1, Zhuo-Xu Cui1, Qingyong Zhu1, and Dong Liang1 | ||
1Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China |
||
Keywords: Machine Learning/Artificial Intelligence, Image Reconstruction Existing deep learning-based methods for MR reconstruction mainly use MSE as loss function to train the network under the assumption that MR images follow the sub-Gaussian distribution, without considering the real distribution of the images. In this work, we propose a new DL-based method that models the image distribution with equilibrium Langevin dynamic to converge the distribution, and trains the network with Wasserstein distance to approach the real distribution. Experimental results on highly undersampled MR data demonstrate the superior performance of the proposed method. |
| 4787 | Computer 17
|
Transferable Deep Learning for Fast MR Imaging |
| Yuxiang Zhou1, Riti Paul2, Pak Lun Kevin Ding2, Leland Hu1, Ameet C Patel1, and Baoxin Li2 | ||
1Radiology, Mayo Clinic at Arizona, Phoenix, AZ, United States, 2CIDSE, Arizona State University, Tempe, AZ, United States |
||
Keywords: Machine Learning/Artificial Intelligence, Data Processing, Reconstruction Artificial intelligence (AI) applications in the field of magnetic resonance imaging have been implemented in routine clinical practice. However, MRI still faces a practical and persistent challenge: its long acquisition time. This has led to two prominent issues in health care: high cost and poor patient experience. Long acquisition time is also a source of degraded imaging quality (e.g., motion artifacts). In the study, we propose to develop novel Deep Learning (DL) architectures combined with transfer learning capabilities to address the above challenge and apply this newly developed AI technique for image reconstruction in MRI with very fast imaging acquisition. |
| 4788 | Computer 18
|
MONAI Recon: An Open Source Tool for Deep Learning Based Accelerated MRI Reconstruction |
| Mohammad Zalbagi Darestani1, Vishwesh Nath2, Wenqi Li2, Yufan He2, Holger Reinhard Roth2, Ziyue Xu2, Daguang Xu2, Reinhard Heckel1,3, and Can Zhao2 | ||
1Rice University, Houston, TX, United States, 2NVIDIA, Santa Clara, CA, United States, 3Technical University of Munich, Munich, Germany |
||
Keywords: Software Tools, Machine Learning/Artificial Intelligence, MRI Reconstruction Deep learning models outperform traditional methods in terms of quality and speed for numerous medical imaging applications. A critical application is the acceleration of magnetic resonance imaging (MRI) reconstruction, where a deep learning model reconstructs a high-quality MR image from a set of undersampled measurements. For this application, we present the MONAI Recon Module to facilitate fast prototyping of deep-learning-based models for MRI reconstruction. Our free and open-source software is pre-equipped with a baseline and a state-of-the-art deep-learning-based reconstruction model and contains the necessary tools to develop new models. The developed open-source software covers the entire MRI reconstruction pipeline. |
| 4789 | Computer 19
|
Investigation of DWI with Deep Learning-based Reconstruction in the Differentiation of Benign and Malignant Breast Lesions |
| Tiebao Meng1, Huiming Liu1, Chuanmiao Xie1, Jialu Zhang2, and Long Qian2 | ||
1Department of Radiology, Sun Yat-Sen University Cancer Center, Guangzhou, China, 2MR Research, GE Healthcare, Beijing, China |
||
Keywords: Machine Learning/Artificial Intelligence, Diffusion/other diffusion imaging techniques As the most common malignant tumor in women, the differentiation of benign lesion from malignant breast tumors is the most essential step in early diagnosis. Recent study demonstrated that DWI could offer great help in differential diagnosis of breast tumors. The novel deep learning-based reconstruction (DLR) technique is able to increase SNR of MRI images. Using DLR, DWI images can be acquired with less NEX (fast DWI) and still maintain the image quality. This study indicated the feasibility of fast DWI protocol with DLR in the differentiation of breast benign lesions and malignant tumors. |
| 4790 | Computer 20
|
Clinical Viability of AI-enabled 0.5T MRI Scanner to Improve Access |
| Arjun Narula1, Uday Patil2, Anand SH3, Harikrishnan Raveendran4, Shailaja Muniraj5, Pallavi Rao6, Anurita Menon6, Allison Nicole Garza7, Santosh Kumar7, Gautam Kumar7, Srinivas NR7, Syam Babu7, Ravi Jaiswal7, Rajagopalan Sundaresan7, Ashok Kumar Reddy7, Sajith Rajamani7, Rajdeep Das7, Nitin Jain7, Sudhir Ramanna7, Sundar V7, Florintina Charlaas7, Sudhanya Chatterjee7, Rohan Patil7, Megha Goel7, Dattesh Shanbhag7, Vikas Kumar Anand7, Abhishek Galagali7, Sathish KV7, Preetham Shankpal7, Harsh Kumar Agarwal7, Suresh Emmanuel Joel7, and Ramesh Venkatesan7 | ||
1Narula Diagnostics, Rohtak, India, 2Manipal Hospitals, Bangalore, India, 3Jivaa Diagnostics, Tumkur, India, 4Core Clinico PET Imaging, Thane, India, 5Prima Diagnostics, Bangalore, India, 6Image Core Lab, Bangalore, India, 7GE Healthcare, Bangalore, India |
||
Keywords: Machine Learning/Artificial Intelligence, Low-Field MRI, Accessible MRI We validate the clinical viability of a 0.5T scanner to reduce cost and improve access to quality MRI using AI based IQ enhancement to compensate for IQ reduction due to lower field and other lower hardware specifications. We obtained data from 65 patients from the re-ramped 0.5T and a commercially available 1.5T MRI system for brain and cervical spine. Radiologists compared image quality between the two and rated the image on the ability to perform diagnosis. We observed that more than 90% of the images were rated to be above diagnostic levels and AI reconstruction significantly improved the image quality. |
| 4791 | Computer 21
|
Real-time MRI with Simultaneous Multi-slice and Compressed Sensing |
| Isaac Watson1,2,3, Martin Trefzer1, David Mitchell2, Angelika Sebald2, and Aneurin Kennerley3,4 | ||
1School of Physics, Engineering and Technology, University of York, York, United Kingdom, 2York Cross-disciplinary Centre for Systems Analysis, University of York, York, United Kingdom, 3York Neuroimaging Centre, University of York, York, United Kingdom, 4Institute of Sport, Manchester Metropolitan University, Manchester, United Kingdom |
||
Keywords: Data Acquisition, Data Acquisition Real time (rt) MRI is currently predominantly a single slice acquisition technique. We show that, by combining compressed sensing, simultaneous multi-slice acquisition and interleaved data ordering with a highly undersampled radial trajectory, multi-slice rtMRI images with a high temporal resolution can be acquired. This offers new opportunities, for example, the non-invasive monitoring of oral and maxillo facial dynamics.
|
| 4792 | Computer 22
|
Radial-RIM: accelerated radial 4D MRI using the recurrent inference machine |
| Chaoping Zhang1, Matthan W.A. Caan2,3,4, Robin Navest1, Jonas Teuwen1,5,6, and Jan-Jakob Sonke1 | ||
1Netherlands Cancer Institute, Amsterdam, Netherlands, 2Amsterdam UMC location University of Amsterdam, Amsterdam, Netherlands, 3Amsterdam Neuroscience, Amsterdam, Netherlands, 4Oslo University Hospital, Oslo, Norway, 5University of Amsterdam, Amsterdam, Netherlands, 6Radboud University Medical Center, Nijmegen, Netherlands |
||
Keywords: Image Reconstruction, Image Reconstruction, radial, radiotherapy, liver, lung, respiratory motion To accelerate the acquisition and reconstruction of the respiratory correlated (4D) radial MRI for image guided radiotherapy guidance, we propose a 3D radial-RIM network by adapting the forward model for Cartesian k-space in the regular RIM to radial k-space. Experiments show that the radial-RIM reconstructs images with better quality than the XD-GRASP compressed sensing for 3-4 times accelerated 4D radial free-breathing lung/liver acquisitions, with reconstruction time per slice reduced from 2 minutes to 1.9 seconds. Signal correlation exploited by the radial-RIM over the extra motion-state dimension contributes substantially to the image quality. |
| 4793 | Computer 23
|
Free-Breathing Gadoxetic Acid-Enhanced Hepatobiliary Phase Imaging Using Stack-of-Stars Radial Sampling and Compressed SENSE |
| Tetsuro Kaga1, Yoshifumi Noda1, Nobuyuki Kawai1, Kimihiro Kajita2, Yu Ueda3, Masatoshi Honda3, Fuminori Hyodo4, Hiroki Kato1, and Masayuki Matsuo1 | ||
1Department of radiology, Gifu university, Gifu, Japan, 2Gifu university hospital, Gifu, Japan, 3Philips Japan, Tokyo, Japan, 4Institute for Advanced Study, Gifu university, Gifu, Japan |
||
Keywords: Data Acquisition, Liver, Free breathing, Radial sampling The gadoxetic acid-enhanced hepatobiliary phase imaging is an effective sequence for detecting hepatic lesions; however, degraded image quality is often observed due to poor breath holding. Free-breathing sequence using radial stack-of-stars acquisition (3D VANE) has been introduced and can provide diagnosable image quality. Recently, Compressed SENSE (CS), that is one of the acceleration techniques, has been applicable in 3D VANE. In this study, we evaluated the feasibility of 3D VANE with CS in hepatobiliary phase imaging. Our results showed that 3D VANE with CS at section thickness of 2 mm had almost comparable image quality as conventional breath-holding cartesian sampling. |
| 4794 | Computer 24
|
Navistar: Golden-Angle Stack-of-Stars Sampling with Embedded 2D Navigators for Highly-Accelerated 4D Dynamic MRI |
| Ding Xia1, Sera Saju1, Kai Tobias Block2, and Li Feng1 | ||
1BioMedical Engineering and Imaging Institute (BMEII), Icahn School of Medicine at Mount Sinai, New York, NY, United States, 2Center for Advanced Imaging Innovation and Research (CAI2R), NYU Grossman School of Medicine, New York, NY, United States |
||
Keywords: Data Acquisition, Motion Correction This work proposes a new radial sampling scheme called Navistar for highly-accelerated 4D dynamic MRI. Navistar periodically inserts 2D navigators into a golden-angle radial stack-of-stars acquisition. The embedded 2D navigators provide spatially-resolved information that can be used for different purposes, such as improved temporal-basis estimation for image reconstruction or respiratory-motion estimation. In our experiments, we have shown that Navistar sampling can be combined with low-rank subspace-constrained image reconstruction for highly-accelerated free-breathing 4D DCE-MRI of the liver with sub-second temporal resolution, which helps reduce intraframe respiratory blurring and eliminates the need for explicit respiratory-motion compensation. |
| 4795 | Computer 25
|
Full-brain multi-pathway relaxometry using a golden-angle radial scheme and sparsely-sampled radial spokes |
| Li-Ya Shao1, Zih-Ci Wang2, Bruno Madore3, and Cheng-Chieh Cheng2 | ||
1Department of Electrical Engineering, National Sun Yat-sen University, Kaohsiung, Taiwan, 2Department of Computer Science and Engineering, National Sun Yat-sen University, Kaohsiung, Taiwan, 3Department of Radiology, Brigham and Women's Hospital, Harvard Medical School, Boston, MA, United States |
||
Keywords: Data Acquisition, Relaxometry Quantitative MRI is emerging as a powerful diagnostic tool for neuroimaging applications. A motion-robust, golden-angle radial acquisition version of the ‘triple-echo steady-state’ (TESS) relaxometry method was implemented here, to mitigate the artifacts induced by the flowing motion of the cerebrospinal fluid (CSF). To improve the scan efficiency, a probability density function-based sparse sampling scheme was introduced in each radial spoke. Close agreement was obtained between reference and TESS scans for both T1 and T2 values, using a multi-compartment phantom. In vivo whole-brain results were further obtained (3D T1 and T2 maps, 1-mm isotropic resolution, whole-brain coverage, 7.5-minute scan). |
| 4796 | Computer 26
|
Radial Magnetic Resonance Fingerprinting for quantification of T1 in metastatic breast tumor. |
| Chetan B. Dhakan1, Jorge De La Cerda1, William Schuler1, Christina J. MacAskill2, Chris A. Flask2, Mark D Pagel1, and Gary V. Martinez3 | ||
1Cancer Systems Imaging, MD Anderson Cancer Center, Houston, TX, United States, 2Department of Radiology, Case Western Reserve University, Cleveland, OH, United States, 3Imaging Physics, MD Anderson Cancer Center, Houston, TX, United States |
||
Keywords: MR Fingerprinting/Synthetic MR, Cancer Quantitative Magnetic Resonance Imaging (MRI) involves pixel-wise mapping of longitudinal relaxation time T1, transverse relaxation time T2 and proton density M0 and other relevant parameters at each location in the tissue to be characterized. The goal of this study is to investigate potential benefits of Magnetic Resonance Fingerprinting in quantifying T1 relaxation times in metastatic breast tumor using radial acquisition in tumor with high temporal and spatial resolution. |
| 4797 | Computer 27
|
Electric Potential Energy Optimized 3D Radial Trajectories |
| Christopher Huynh1, Datta Singh Goolaub1, and Christopher K Macgowan1,2 | ||
1Translational Medicine, Hospital for Sick Children, Toronto, ON, Canada, 2Medical Biophysics, University of Toronto, Toronto, ON, Canada |
||
Keywords: Data Acquisition, Artifacts, Optimized 3D Trajectory The commonly used golden angle trajectory for 3D radial dynamic MRI suffers from spiral-shaped sample clustering around the z-axis, often resulting in image streaking artifact. We develop a method that corrects for this for center-out 3D radial trajectories. The electric potential energy of the trajectory is minimized through the use of repulsive forces, producing a spherically uniform distribution while maintaining the quasirandom quality of the golden angle trajectory. ELECTRic potential energy Optimized (ELECTRO) trajectories are relatively inexpensive to obtain and can produce images with lower MSE compared to the golden angle counterpart. |
| 4798 | Computer 28
|
Ultra-fast Radial MRI with Silent Oscillating Gradients |
| Serhat Ilbey1, Sebastian Littin1, Feng Jia1, Niklas Wehkamp1, Philipp Amrein1, Maxim Zaitsev1, Michael Bock1, and Ali Caglar Özen1 | ||
1Division of Medical Physics, Department of Radiology, University Medical Center Freiburg, Freiburg, Germany |
||
Keywords: New Trajectories & Spatial Encoding Methods, New Trajectories & Spatial Encoding Methods The sampling efficiency of radial center-out MRI was increased by using additional oscillating readout gradients, which lead to an accelerated image acquisition. The idea of radial sampling along an oscillating spoke was demonstrated using a clinical MRI system with ultra-fast UTE MRI. |
| 4799 | Computer 29
|
A Physics-informed Conditional Wasserstein Autoencoder to Quantify Uncertainties in Accelerated 2D Dynamic Radial MRI |
| Sherine Brahma1, Tobias Schaeffter1,2,3, Christoph Kolbitsch1,3, and Andreas Kofler1 | ||
1Physikalisch-Technische Bundesanstalt (PTB), Braunschweig and Berlin, Germany, 2Department of Biomedical Engineering, Technical University of Berlin, Berlin, Germany, 3School of Imaging Sciences and Biomedical Engineering, King’s College London, London, United Kingdom |
||
Keywords: Image Reconstruction, Cardiovascular, Radial Acquisition Uncertainty quantification (UQ) can provide important information about deep learning algorithms and help interpret the obtained results. UQ for multi-coil dynamic MRI is challenging due to the large scale of the problem and scarce training data. We approach these issues by learning distributions in a lower dimensional latent space using a conditional Wasserstein autoencoder while utilizing the MR data acquisition model and by exploiting spatio-temporal correlations of the cine MR images. Our results indicate excellent image quality accompanied with uncertainty maps that correlate well with estimation errors. |
| 4800 | Computer 30
|
T1 Relaxation-Enhanced Steady-State (T1RESS) Imaging with Radial Sampling for Robust Brain Examination at 3T and 0.55T |
| Ruoxun Zi1, Kai Tobias Block1, Ioannis Koktzoglou2, Mahesh Keerthivasan3, Thomas Benkert4, Jakob Asslaender1, Daniel K Sodickson1, and Robert R Edelman2 | ||
1The Bernard and Irene Schwartz Center for Biomedical Imaging, Department of Radiology, New York University Grossman School of Medicine, New York, NY, United States, 2Radiology, Northshore University HealthSystem, Evanston, IL, United States, 3Siemens Medical Solutions, New York, NY, United States, 4Application Development, Siemens Healthcare GmbH, Erlangen, Germany |
||
Keywords: Data Acquisition, Brain A new class of sequences, named T1 relaxation-enhanced steady-state (T1RESS), has been recently proposed and showed improved lesion conspicuity in contrast-enhanced exams. T1RESS sequences combine steady-state acquisition, either balanced or unbalanced with weak gradient spoiling, with periodic contrast-modifying preparation pulses to create T1 weighting of the contrast. Here, we present radial versions of T1RESS sequences using the stack-of-stars trajectory, which offer improved motion robustness and enable dynamic imaging through use of advanced reconstructions such as GRASP. Volunteer scans acquired at 3T are shown for both sequences. Moreover, the balanced version is demonstrated at 0.55T to highlight its SNR efficiency. |
| 4801 | Computer 31
|
Stochastic optimization of 3D non-Cartesian sampling trajectory (SNOPY) |
| Guanhua Wang1, Jon-Fredrik Nielsen1, Jeffrey A. Fessler2, and Douglas C. Noll1 | ||
1Biomedical Engineering, University of Michigan, Ann Arbor, MI, United States, 2EECS, University of Michigan, Ann Arbor, MI, United States |
||
Keywords: New Trajectories & Spatial Encoding Methods, Machine Learning/Artificial Intelligence Efficient k-space trajectories are crucial for accelerated MRI. SNOPY proposes a generalized gradient-based method for optimizing 3D non-Cartesian sampling patterns. The algorithm can simultaneously tune multiple properties of sampling patterns, including image quality, hardware constraints (maximum slew rate and gradient strength), reduced peripheral nerve stimulation (PNS), and parameter-weighted contrast. The proposed method applies to various scenarios, such as optimizing gradient waveforms or optimizing rotation angles of radial/spiral trajectories. We adopted several computational strategies to address this non-convex and large-scale problem. Various simulated and in-vivo experiments demonstrated the effectiveness of SNOPY. |
| 4802 | Computer 32
|
Cartesian Spiral acquisitions for radiotherapy on an MR-Linac |
| Bastien Lecoeur1,2, Prashant Nair1, Rosie Goodburn1, Tom Bruijnen3, Uwe Oelfke1, Wayne Luk2, and Andreas Wetscherek1 | ||
1Joint Department of Physics, The Institue of Cancer Research, Sutton, United Kingdom, 2Department of Computing, Imperial College London, London, United Kingdom, 3Department of Radiotherapy, University Medical Center Utrecht, Utrecht, Netherlands |
||
Keywords: Image Reconstruction, Radiotherapy
Respiratory-resolved
4D-MRIs could measure the extent of respiratory motion for online MR-guided
radiotherapy on MR-Linacs. Cartesian spiral (CASPR) trajectories are an
alternative to radial sampling for self-gated 4D-MRIs, because they could avoid
prohibitively long image reconstruction times related to the non-uniform
Fourier Transform. We studied the effect of different sampling parameters on
image quality for CASPR-based sequences on an MR-Linac and reconstructed abdominal
4D-MRI, we found that increasing the number of points per spiral arm improved the
reconstructed image quality, creating a trade-off between spatial and temporal
resolutions.
|
| 4803 | Computer 33
|
Radial vs. Spiral – A Comparison of Stack-of-stars and Stack-of-spirals Spatial Encoding Schemes in Multiparametric Body MRI with QTI |
| Carolin M Pirkl1, Rolf F Schulte1, Pablo García-Polo2, Matteo Cencini3, Sebastian Endt1,4,5, Laura Biagi3, Michela Tosetti3, and Marion I Menzel1,4,5 | ||
1GE Healthcare, Munich, Germany, 2GE Healthcare, Madrid, Spain, 3IRCCS Stella Maris, Pisa, Italy, 4Technische Hochschule Ingolstadt, Ingolstadt, Germany, 5Technical University of Munich, Munich, Germany |
||
Keywords: MR Fingerprinting/Synthetic MR, Data Acquisition Highly accelerated multiparametric MRI techniques are more and more demonstrating their diagnostic potential. To achieve short scan times, these techniques rely on non-Cartesian (under-) sampling. Initially developed for MRI applications in the brain, these sequences are now also applied in body MRI. In this work, we present a phantom and in vivo study to demonstrate and evaluate the strengths and pitfalls of the two most common non-Cartesian – spiral and radial – sampling schemes for the application of rapid, free-breathing 3D T1, T2 and proton density (PD) mapping with quantitative transient-state imaging (QTI) in the prostate gland. |
| 4804 | Computer 34
|
Single-shot 2D Radial Echo Planar Imaging using a KWIC Filter and Model-based Reconstruction Approach |
| Christoph Rettenmeier1, Zidan Yu2, and V. Andrew Stenger3 | ||
1Medicine, University of Hawaii, Honolulu, HI, United States, 2Univeristy of Hawaii, Honolulu, HI, United States, 3University of Hawaii, Honolulu, HI, United States |
||
Keywords: Data Acquisition, Brain High quality brain images at spatial resolutions of 2x2x3 mm3 and 3.3x3.3x3 mm3 were obtained using single-shot 2D rEPI. The approach is based on an R2*, B0 and coil sensitivity informed model-based reconstruction in combination with k-space weighted image contrast (KWIC) filtering. Signal loss due to phase inconsistencies are starkly reduced by the linear phase model and R2* contrast was modifiable by adjusting the target echo time in the model of the reconstruction and KWIC filter. Bold activation maps were generated from fMRI datasets at 2x2 mm2 resolution showing activation in the visual cortex. |
| 4805 | Computer 35
|
Values of golden-angle radial-VIBE in DCE-MRI in laryngeal and hypopharyngeal squamous cell carcinoma: comparison with conventional VIBE |
| Yan Wen1, Liling Long1, Chenhui Li1, Huiting Zhang2, and Xiaodong Zhong3 | ||
1Department of Radiology, The First Affiliated Hospital of Guangxi Medical University, Nanning, China, 2MR Scientific Marketing, Siemens Healthineers Ltd., Wuhan, China, 3MR R&D Collaborations, Siemens Medical Solutions, Los Angeles, CA, United States |
||
Keywords: Image Reconstruction, Head & Neck/ENT This study aimed to investigate the value of a dynamic contrast-enhanced (DCE)-MRI with golden-angle radial volumetric‑interpolated breath‑hold examination (radial-VIBE) research application sequence by comparing with conventional VIBE in patients with laryngeal and hypopharyngeal squamous cell carcinoma (SCC) under free breathing conditions. Our results showed that compared with conventional contrast-enhanced VIBE images with single phase, the morphological radial-VIBE images of multitemporal synthesis had significantly higher image quality based on subjective (edge, artifact, and confidence) and objective (signal-to-noise ratio, contrast, and contrast-to-noise ratio) assessments. Therefore, radial-VIBE can be used in DCE‑MRI examination in laryngeal and hypopharyngeal SCC in the free breathing state. |
| 4806 | Computer 36
|
Evaluation of perfusion on Golden-Angle Radial Sparse Parallel (GRASP) MRI in discriminating benign and malignant liver tumors |
| Yu Wang1, Xiaohui Duan1, Mengzhu Wang2, Lingjie Yang1, Zhuoheng Yan1, and Huijun Hu1 | ||
1Department of Radiology, Sun Yat-Sen Memorial Hospital, Sun Yat-Sen University, Guangzhou, China, 2MR Scientific Marketing, Siemens Healthineers Ltd., Guangzhou, China |
||
Keywords: Data Analysis, Cancer This study investigated the utility of quantitative parameters of golden-angle radial sparse parallel (GRASP) dynamic contrast enhanced magnetic resonance imaging (DCE-MRI) for evaluation of perfusion and differentiation of benign and malignant liver tumors. The results showed the quantitative parameters in GRASP DCE-MRI can effectively evaluate the perfusion characteristics and differentiation in liver tumors with good diagnostic performance. This indicated that the GRASP quantitative parameters may be useful to evaluate and predict the pathological stage of liver lesions. |
| 4807 | Computer 37
|
A Convergence Analysis for Projected Fast Iterative Soft-thresholding Algorithm under Radial Sampling MRI |
| Biao Qu1, Zuwen Zhang2, Yewei Chen2, Chen Qian2, Taishan Kang3, Jianzhong Lin3, Lihua Chen4, Zhigang Wu5, Jiazheng Wang5, Gaofeng Zheng1, and Xiaobo Qu2 | ||
1Department of Instrumental and Electrical Engineering, Xiamen University, Xiamen, China, 2Department of Electronic Science, Biomedical Intelligent Cloud R&D Center, Fujian Provincial Key Laboratory of Plasma and Magnetic Resonance, Xiamen University, Xiamen, China, 3Department of Radiology, Zhongshan Hospital of Xiamen University, School of Medicine, Xiamen University,, Xiamen, China, 4Department of Radiology, Chongqing University Cancer Hospital, School of Medicine, Chongqing University, Chongqing, China, 5Philips, Beijing, China |
||
Keywords: Image Reconstruction, Image Reconstruction Radial sampling is a fast magnetic resonance imaging technique. The projected fast iterative soft-thresholding algorithm (pFISTA) has shown the advantage to solved tight frame sparse reconstruction model for removing undersampling image artifacts. However, the convergence of this algorithm under radial sampling has not been clearly set up. In this work, the authors derived a theoretical convergence condition for this algorithm and an optimal step size was further suggested to allow the fastest convergence. Verifications were made in vivo data of static brain imaging and dynamic contrast-enhanced (DCE) liver imaging, demonstrating that the recommended parameter allowed fast convergence in radial MRI. |
| 4808 | Computer 38
|
Machine learning for detecting sensorineural hearing loss utilizing functional imaging with a combination of static and dynamic brain features |
| Xiao-Min Xu1 and Yu-Chen Chen1 | ||
1Radiology, Nanjing First Hospital, Nanjing, China |
||
Keywords: Head & Neck/ENT, fMRI (resting state) Alterations of static and dynamic brain function have been found in sensorineural hearing loss (SNHL). The combination of data-driven machine learning based classifiers and multiple imaging features can identify SNHL and healthy controls automatically. The spearman rank correlation with radial basis functional kernel support vector machine (SVM) and sigmoid SVM provides promising neural biomarkers for clinical classifier of SNHL. |
| 4809 | Computer 39
|
Reducing scan-time for 3D imaging with undersampled cartesian-radial phase encoding on a point-of-care 46 mT Halbach MRI scanner |
| Chloé Najac1, Kirsten Koolstra2, Tom O’Reilly1, and Andrew Webb1 | ||
1C.J. Gorter MRI Center, Department of Radiology, Leiden University Medical Center, Leiden, Netherlands, 2Philips, Best, Netherlands |
||
Keywords: Low-Field MRI, New Trajectories & Spatial Encoding Methods Low-field MRI systems (with B0<0.1T) for point-of-care applications are becoming increasingly widespread. Imaging at low-field remains challenging due to the low intrinsic SNR. We evaluated speeding up 3D imaging using radial-based Cartesian undersampled phase-encodings (PEs). In phantoms, we tested different undersampling schemes and compared them to full in-out Cartesian PE in terms of peak SNR (PSNR) to take account of spatial resolution and image SNR. Results suggests that a radial-based Cartesian PE trajectory with an overall acceleration factor of two can be implemented while preserving image quality (PSNR~69dB with R=2 vs. ~70dB with R=1.3 in a discrete spatial resolution phantom). |
| 4810 | Computer 40
|
Fast Deep Learning Reconstruction of Interventional MRI Data with Radial, Undersampling k-Space Trajectories |
| Johanna Topalis1, Balthasar Schachtner1, Andreas Mittermeier1, Philipp Wesp1, Tobias Weber1,2, Anna Theresa Stüber1,2, Max Seidensticker1, Jens Ricke1, Katia Parodi3, Michael Ingrisch1, and Olaf Dietrich1 | ||
1Department of Radiology, University Hospital, LMU Munich, Munich, Germany, 2Department of Statistics, LMU Munich, Munich, Germany, 3Department of Medical Physics, Faculty of Physics, LMU Munich, Munich, Germany |
||
Keywords: Image Reconstruction, Machine Learning/Artificial Intelligence, Radial Acquisition; Undersampled; MR-Guided Interventions To reduce the data acquisition time and increase frame rates for image guidance during percutaneous needle interventions in the liver, k-space data can be acquired with a radial, undersampling acquisition scheme. The purpose of this work was to optimize a deep learning model for the fast reconstruction of k-space data in this context. The proposed deep learning model reconstructed artificial data with a better image quality compared to conventional reconstruction. Successful reconstruction of interventional phantom data suggests its potential for application during percutaneous needle interventions in the liver. |
| 4947 | Computer 1
|
Scaling nuFFT Memory-Overhead Down to Zero: Computational Trade-Offs and Memory-Efficient PICS-Reconstructions with BART |
| Moritz Blumenthal1,2, H. Christian M. Holme2, and Martin Uecker1,2 | ||
1Institute for Diagnostic and Interventional Radiology, University Medical Center Göttingen, Göttingen, Germany, 2Institute of Biomedical Imaging, Graz University of Technology, Graz, Austria |
||
Keywords: Image Reconstruction, Software Tools We propose a new decomposition of the nuFFT algorithm, allowing for a zero-memory-overhead implementation of the (adjoint-)nuFFT. With one grid-sized buffer, the decomposition allows for a memory efficient nuFFT with negligible computational overhead compared to a two-fold oversampled conventional nuFFT - to the prize of less efficient parallelization on many-threads CPU systems. We reduce memory requirements of 3D non-Cartesian PICS-reconstructions in BART by a factor up to ten, allowing for GPU acceleration of reconstructions with eight coils and matrix size 256x256x256 on a 4GB consumer-level GPU. |
| 4948 | Computer 2
|
A k-space insight of aliasing effects and their removal of SPEN MRI |
| Sijie Zhong1,2, Minjia Chen3, Ke Dai1,4, Hao Chen1,2, Lucio Frydman4, and Zhiyong Zhang1,2 | ||
1School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China, 2Institute of Medical Robotics, Shanghai Jiao Tong University, Shanghai, China, 3Department of Engineering, University of Cambridge, Cambridgeshire, United Kingdom, 4Department of Chemical and Biological Physics, Weizmann Institute of Science, Rehovot, Israel |
||
Keywords: Image Reconstruction, Artifacts, SPEN The reconstruction theory of Spatio-temporal encoding MRI is not complete at present. Aliasing artifacts in SPEN MRI reconstructions can be traced to image contributions corresponding to high-frequency k-space signals. The k-space picture provides the spatial displacements, phase offsets and linear amplitude modulations associated to these artifacts, as well as routes to removing these from the reconstruction results. These new ways to estimate the artifact priors were applied to reduce SPEN reconstruction artifacts on simulated, phantom and human brain MRI data. |
| 4949 | Computer 3
|
Locally high rank reconstruction through Partial Separability model (PS-LHR) with regional optimized temporal basis (ROT) of dynamic speech MRI |
| Riwei Jin1, Yudu Li2, Fangxu Xing3, Imani Gilbert4, Jamie Perry4, Jonghye Woo3, Zhi-pei Liang5, and Brad Sutton1 | ||
1Department of Bioengineering, University of Illinois Urbana-Champaign, Champaign, IL, United States, 2National Center for Supercomputing Applications, Champaign, IL, United States, 3Gordon Center for Medical Imaging, Department of Radiology, Massachusetts General Hospital/Harvard Medical School, Boston, MA, United States, 4Department of Communication Sciences and Disorders, East Carolina University, Greenville, NC, United States, 56 Department of Electrical and Computer Engineering, University of Illinois Urbana-Champaign, Champaign, IL, United States |
||
Keywords: Image Reconstruction, Sparse & Low-Rank Models To optimize the reconstruction quality of isotropic 3D dynamic speech magnetic resonance imaging with large scan volume, we applied two novel methods based on the Partial Separability model theory: 1. Locally High-Rank reconstruction through Partial Separability model (PS-LHR) which enables higher rank to be devoted to the dynamic speech region. 2. Implementation of Regional-Optimized Temporal basis (ROT) to focus the temporal navigator information on the speech region. The improvement in reconstruction quality was seen to decrease the noise of regions of the image outside the area of interest and increase dynamic smoothness in the speech region. |
| 4950 | Computer 4
|
Efficient correction for MRI artifacts due to nonlinear variations in magnetic field using principal component analysis |
| Daniel I Gendin1 and Jiaen Liu2 | ||
1Advanced Imaging Research Center, UT Southwestern Medical Center, Dallas, TX, United States, 2Advanced Imaging Research Center and Radiology, UT Southwestern Medical Center, Dallas, TX, United States |
||
Keywords: Image Reconstruction, Brain Head motion can lead to artifacts in MR images. One source of motion artifacts is dynamic nonlinear variation in the background magnetic field (B0). Fully accounting for this variation is computationally intractable. We proposed an efficient method of performing this correction by making use of principal component analysis and considering only a few dominant modes. We applied our method to measured data and compared the results to the case with no correction as well correction using a previous K-means based method. |
| 4951 | Computer 5
|
MUSIC-MRI: unleash the resolving power of k-space beyond the Fourier transform |
| Dongbiao Sun1,2, Yan Zhuo1,2, Lin Chen2,3, and Zihao Zhang1,3 | ||
1State Key Laboratory of Brain and Cognitive Science, Beijing MR Center for Brain Research, Institute of Biophysics, Chinese Academy of Sciences, Beijing, China, 2University of Chinese Academy of Sciences, Beijing, China, 3Institute of Artificial Intelligence, Hefei Comprehensive National Science Center, Hefei, China |
||
Keywords: Image Reconstruction, Image Reconstruction, Non-Fourier transform reconstruction Modern MRI reconstructs images by performing Fourier transform (FT) on k-space to decompose the signal on an orthogonal basis composed of trigonometric functions. We are inspired by the problem of estimating the Direction of Arrival in radar theory and propose a novel route for MRI reconstruction based on multiple signal classification (MUSIC). MUSIC-MRI overcomes the sidelobe problem of FT and significantly promotes the actual resolution in the meaning of full width at half maximum (FWHM). Our phantom experiments show that the FWHMs are 0.45mm by MUSIC-MRI and 1.50mm by FT, while the nominal resolution of the k-space data is 0.94mm. |
| 4952 | Computer 6
|
Modified Homodyne reconstruction using a high-resolution phase in magnetic resonance images |
| Xinzeng Wang1, Daniel Litwiller2, Arnaud Guidon3, Tim Sprenger4, and Robert Marc Lebel5 | ||
1GE Healthcare, Houston, TX, United States, 2GE Healthcare, Denver, CO, United States, 3GE Healthcare, Boston, MA, United States, 4GE Healthcare, Stockholm, Sweden, 5GE Healthcare, Calgary, AB, Canada |
||
Keywords: Image Reconstruction, Artifacts, Partial Fourier A partial Fourier acquisition has been widely used for fast MR imaging. To reduce the truncation artifacts in partial Fourier image, Homodyne reconstruction is often used, and it exploits the conjugate symmetry in real-valued signal to recover the full k-space. However, the MR signal is complex-valued. Artifacts are commonly observed in Homodyne images in the regions of rapid phase change due to the interference of imaginary components of adjacent pixels. In this work, we proposed a modified Homodyne reconstruction to reduce the conventional Homodyne artifacts and truncation artifacts by using a high-resolution phase from a pre-trained deep-learning network. |
| 4953 | Computer 7
|
Complex Quasi-Newton Proximal Methods for the Image Reconstruction in Compressed Sensing MRI |
| Tao Hong1, Jeffrey A. Fessler2, and Luis Hernandez-Garcia 1 | ||
1Department of Radiology, University of Michigan, Ann Arbor, MI, United States, 2Department of Electrical and Computer Engineering, University of Michigan, Ann Arbor, MI, United States |
||
Keywords: Image Reconstruction, Data Processing, Fast reconstruction algorithm This work studies a complex quasi-Newton proximal method (CQNPM) for MRI reconstruction using wavelets or total variation (TV) based regularization. Our experiments show that our method is faster than the accelerated proximal method [1,2] in terms of iteration and CPU time. |
| 4954 | Computer 8
|
High-resolution single-shot spiral diffusion-weighted imaging at 7T using the expanded encoding model and compressed sensing |
| Paul I. Dubovan*1,2, Gabriel Varela-Mattatall*1,2, Tales Santini1,2, Kyle M. Gilbert1,2, Ravi S. Menon1,2, and Corey A. Baron1,2 | ||
1Centre for Functional and Metabolic Mapping (CFMM), Robarts Research Institute, Western University, London, ON, Canada, 2Department of Medical Biophysics, Schulich School of Medicine and Dentistry, Western University, London, ON, Canada |
||
Keywords: Image Reconstruction, Image Reconstruction Reconstruction problems involving the expanded encoding model and field monitoring are typically solved using least squares optimization with the conjugate gradient method and early stopping as an implicit form of regularization. However, this is likely a suboptimal strategy for low SNR acquisitions, such as accelerated or high-resolution diffusion MRI. Hence, in this work we present an extension to the matMRI reconstruction toolbox that incorporates compressed sensing regularization. Results demonstrate that the expanded encoding model and compressed sensing regularization are complementary tools that mitigate artifacts from phase perturbations while permitting lower SNR conditions for high-resolution single-shot spiral diffusion-weighted imaging. |
| 4955 | Computer 9
|
Complex-Valued Fourier Primal-Dual: Undersampled MRI Reconstruction in Hybrid-space |
| Soumick Chatterjee1,2,3, Philipp Ernst1,2, Oliver Speck3,4,5, and Andreas Nürnberger1,2,5 | ||
1Faculty of Computer Science, Otto von Guericke University Magdeburg, Magdeburg, Germany, 2Data and Knowledge Engineering Group, Otto von Guericke University Magdeburg, Magdeburg, Germany, 3Department of Biomedical Magnetic Resonance, Otto von Guericke University Magdeburg, Magdeburg, Germany, 4German Center for Neurodegenerative Disease, Magdeburg, Germany, 5Center for Behavioral Brain Sciences, Magdeburg, Germany |
||
Keywords: Image Reconstruction, Artifacts Iterative undersampled MRI reconstructions, such as compressed sensing, can reconstruct undersampled MRIs - but due to their slow execution speed, they are not suitable for real-time applications. Several deep learning approaches have been proposed, mostly working in image space. Some of the approaches, which work on the k-space or in a mix of spaces, employ real-valued convolutions splitting the complex k-space into real and imaginary parts for processing - destroying the geometric relationship within the data. This research proposes Fourier-PD and Fourier-PDUNet models using complex-valued convolutions, which attempt to predict missing k-space frequencies and also to reduce artefacts in the image space. |
| 4956 | Computer 10
|
Statistical Approach to Single-Point Water-Fat Imaging with Independent Component Analysis (ICA) |
| Qing-San Xiang1 | ||
1Radiology, University of British Columbia, Vancouver, BC, Canada |
||
Keywords: Image Reconstruction, Image Reconstruction, water-fat imaging Water-Fat Imaging with only a single acquisition is desirable for its pulse sequence simplicity and scan time efficiency. However, robust reconstruction is more challenging due to limited available information. In this work, it is discovered that some previously unused information can be found by statistical analysis, leading to effective reconstruction of water and fat images. In particular, Independent Component Analysis (ICA) was used to determine the phase error. Well separated water and fat images were subsequently obtained after phase correction. |
| 4957 | Computer 11
|
Feasibility study of a simulTaneous multi-relaXation-time Imaging (TXI) method in Intervertebral Disc |
| Jie Yang1, Yujian liu1,2, Wenjia Hang1, Meining Chen3, Jianqi Li4, Yinqiao Yi4, Haodong Zhong4, and Xu Yan3 | ||
1Department of Radiology, Zigong First People's Hospital, Zigong, China, 2Sichuan Vocational College of Health and Rehabilitation, Zigong, China, 3MR Scientific Marketing, Siemens Healthineers, Shanghai, China, 4Shanghai Key Laboratory of Magnetic Resonance, East China Normal University, Shanghai, China |
||
Keywords: Image Reconstruction, MSK Quantitative imaging of intervertebral disc degeneration is important for the diagnosis of lower back pain. In our study, a simulTaneous multi-relaXation-time Imaging (TXI) method was evaluated, which generated bone marrow fat fraction (PDFF), R2* and R1 mapping within a single scan. The results showed that the BMFF, T2*, T1 mapping calculated from TXI method were close to the values reported in other studies for lumbar discs. |
| 4958 | Computer 12
|
Retrospective and prospective accelerated T1ρ/T2 mapping with Compressed Sensing: high resolution T1ρ mapping and simultaneous T1ρ/T2 mapping |
| Jeehun Kim1,2, Zhiyuan Zhang1,3, Ruiying Liu4, Brendan Eck1, Mingrui Yang1, Hongyu Li4, Mei Li1, Richard Lartey1, Carl S. Winalski1,5, Leslie Ying4,6, and Xiaojuan Li1,5 | ||
1Department of Biomedical Engineering, Program of Advanced Musculoskeletal Imaging (PAMI), Cleveland Clinic, Cleveland, OH, United States, 2Department of Electrical, Computer, and Systems Engineering, Case Western Reserve University, Cleveland, OH, United States, 3Department of Biomedical Engineering, Case Western Reserve University, Cleveland, OH, United States, 4Electrical Engineering, University at Buffalo, State University of New York, Buffalo, NY, United States, 5Department of Diagnostic Radiology, Imaging Institute, Cleveland Clinic, Cleveland, OH, United States, 6Biomedical Engineering, University at Buffalo, State University of New York, Buffalo, NY, United States |
||
Keywords: Image Reconstruction, Cartilage Quantitative MR T1rho T2 imaging shows promising results on detecting early-stage osteoarthritis, but long scan time limits the spatial resolution, making it vulnerable to partial volume averaging. Such effect reduces the sensitivity to small focal degeneration. In this study, compressed sensing reconstruction with spatio-temporal finite difference regularization was used to accelerate high-resolution (slice thickness < 2mm) T1rho imaging and standard-resolution simultaneous T1rho and T2 imaging, and evaluated the result comparing with reference imaging, retrospective and prospective reconstruction, and scan-rescan repeatability. |
| 4959 | Computer 13
|
Accelerating Multi-Contrast Imaging Near Metallic Implants with Variable Resolution Sampling and Joint Reconstruction |
| Nikolai J Mickevicius1, Andrew S Nencka1, and Kevin M Koch1 | ||
1Medical College of Wisconsin, Milwaukee, WI, United States |
||
Keywords: Image Reconstruction, Image Reconstruction In multi-spectral imaging near metallic implants, redundant information is present in neighboring spectral bins and contrasts due to the overlapping and smooth spectral profiles. This redundancy is exploited here to reduce scan time when multiple image contrasts are desired. |
| 4960 | Computer 14
|
Model-based phase-difference reconstruction for accelerated phase-based T2 mapping |
| Xiaoqing Wang1,2, Jaejin Cho1,2, Yohan Jun1,2, Borjan Gagoski3, and Berkin Bilgic1,2 | ||
1Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Charlestown, MA, United States, 2Department of Radiology, Harvard Medical School, Boston, MA, United States, 3Fetal-Neonatal Neuroimaging and Developmental Science Center, Boston Children's Hospital, Harvard Medical School, Boston, MA, United States |
||
Keywords: Image Reconstruction, Brain, model-based recontruction Phase-based T2 mapping acquires two GRE volumes with opposite RF phase increments, then computes the phase difference and relates this to T2 values through Bloch modeling. We exploit the redundancy in these two acquisitions by posing this as a nonlinear inverse problem, and directly estimate the phase difference from k-space. This reduces the problem to the estimation of a complex M0 volume, and a real-valued phase difference, thus reducing the number of real-valued unknowns from 4 to 3 per voxel. The proposed reconstruction flexibly admits additional regularization, and yields higher quality images from undersampled acquisitions for rapid T2 mapping. |
| 4961 | Computer 15
|
Evaluation of a model-based image reconstruction technique for accelerated point spread function encoded echo planar imaging |
| Nolan Meyer1,2, Myung-Ho In1, David F Black1, Norbert G Campeau1, Kirk M Welker1, John Huston III1, Matt A Bernstein1, and Joshua D Trzasko1 | ||
1Radiology, Mayo Clinic, Rochester, MN, United States, 2Mayo Clinic Graduate School of Biomedical Sciences, Rochester, MN, United States |
||
Keywords: Image Reconstruction, Neuro, Clinical, radiologist, evaluation A model-based image reconstruction (MBIR) platform was developed for accelerated point spread function encoded echo planar imaging. Beginning with retrospectively truncated data, images were reconstructed with the enhanced MBIR framework. These were compared by a team of neuroradiologists with images obtained using standard reconstructions from non-truncated datasets. Scores were assigned for evaluation criteria including high-contrast resolution (HCR), low-contrast detectability (LCD), signal to noise ratio (SNR), artifacts, and overall preference. Nonparametric statistical test results indicate preference for MBIR in HCR, LCD, and overall preference, with comparable results for SNR and artifacts. MBIR thus facilitates PSF-EPI acceleration, enhancing clinical practicality. |
| 4962 | Computer 16
|
Fast $$$T_{2}^{*}$$$ and QSM using locally low rank methods for complex-valued, multi-echo GRE reconstruction |
| Charles Iglehart1, Ali Bilgin1,2,3, and Manojkumar Saranathan4,5 | ||
1Electrical and Computer Engineering, The University of Arizona, Tucson, AZ, United States, 2Biomedical Engineering, The University of Arizona, Tucson, AZ, United States, 3Medical Imaging, The University of Arizona, Tucson, AZ, United States, 4Radiology, UMass Chan Medical School, Worcester, MA, United States, 5Neuroscience, Morningside Graduate School of Biomedical Sciences, Worcester, MA, United States |
||
Keywords: Image Reconstruction, Brain $$$T_{2}^{*}$$$ and quantitative susceptibility mapping are of increasing clinical interest, but typically require lengthy ME-GRE scans. To accelerate such acquisitions, we propose and implement a volumetric reconstruction method employing locally low rank (LLR) techniques to recover complex-valued ME images from highly undersampled k-space data that is suitable for parameter mapping. We present results for reconstructed magnitude and phase across multiple datasets and acceleration ratios, as well as computed $$$R_{2}^{*}$$$ and QSM from reconstructed images. |
| 4963 | Computer 17
|
Low Rank Subspace-Constrained Compressed Sensing Reconstruction of Highly Accelerated Phase-Cycled bSSFP MRI for Fat Fraction Quantification |
| Eva S Peper1,2, Adèle LC Mackowiak1,2,3, Berk C Açıkgöz1,2, Nils Plähn1,2, Yasaman Safarkhanlo4, Li Feng5, and Jessica AM Bastiaansen1,2 | ||
1Department of Diagnostic, Interventional and Pediatric Radiology (DIPR), Inselspital, Bern University Hospital, University of Bern, Bern, Switzerland, 2Translation Imaging Center (TIC), Swiss Institute for Translational and Entrepreneurial Medicine, Bern, Switzerland, 3Department of Diagnostic and Interventional Radiology, Lausanne University Hospital (CHUV) and University of Lausanne (UNIL), Lausanne, Switzerland, 4Department of Cardiology, University Hospital Bern, Bern, Switzerland, 5BioMedical Engineering and Imaging Institute (BMEII), Department of Radiology, Icahn School of Medicine at Mount Sinai, New York, NY, United States |
||
Keywords: Image Reconstruction, Sparse & Low-Rank Models, bSSFP, phase-cycled bSSFP, subspace reconstruction Fat fraction (FF) maps can be estimated from phase-cycled (PC) balanced steady-state free precession (bSSFP) imaging using a dictionary fitting approach (SPARCQ). Three-dimensional PC bSSFP imaging, however, is time consuming, since around 16 PCs must be sampled to fully describe a voxels’ PC profile. Compressed sensing (CS) has been proven to accelerate MRI scans. This study has two aims 1) to evaluate the performance of two CS algorithms (total variation (TV)- vs. subspace-constrained) taking regularisation along the PC dimension into account and 2) to investigate the upper limits of acceleration using these techniques on FF quantification using SPARCQ. |
| 4964 | Computer 18
|
Constrained Reconstruction of White Noise (CROWN) for Strategically Acquired Gradient Echo (STAGE) Imaging |
| Paul Kokeny1, Qiuyun Xu1, Sara Gharabaghi1, Sean Sethi1,2, Yu Liu3, Youmin Zhang3, Peng Liu3, Naying He3, Fuhua Yan3, and E. Mark Haacke1,2,4 | ||
1SpinTech MRI, Bingham Farms, MI, United States, 2Radiology, Wayne State University, Detroit, MI, United States, 3Shanghai Jiao Tong University School of Medicine, Ruijin Hospital, Shanghai, China, 4Neurology, Wayne State University, Detroit, MI, United States |
||
Keywords: Image Reconstruction, Quantitative Imaging By using the inherent relationship between proton spin density (PSD) and T1, we propose a new image-processing approach to reduce noise called CROWN (Constrained Reconstruction of White Noise). Firstly, we established a linear relationship between these two parameters, then applied a cost function to constrain simulated Strategically Acquired Gradient Echo (STAGE) PSD map and T1 map data in the presence of noise. Secondly, we applied this approach to in vivo STAGE images to reduce noise and improve SNR without the loss of image detail. CROWN has the potential to make higher resolution or faster imaging viable with improved SNR. |
| 4965 | Computer 19
|
Diffusion Weighted MR Imaging Using Low Rank Reconstruction for Multi-shot Variable Auto-Calibrating (vARC) Sampling with Volume Coil |
| Nitin Jain1, Ashok Kumar P Reddy1, Rajdeep Das1, Sajith Rajamani1, Rajagopalan Sundaresan1, Harsh Kumar Agarwal1, M Ramasubba Reddy2, and Ramesh Venkatesan1 | ||
1GE Healthcare, Bangalore, India, 2Indian Institute of Technology Madras, Chennai, India |
||
Keywords: Image Reconstruction, Body Diffusion weighted MR imaging (DWI) is key to pathology detection in anatomies such as brain, abdomen and prostate. Echo planar imaging (EPI) provides a rapid means to acquire DWI. EPI with variable k-space sampling scheme and an auto-calibrating image reconstruction technique, vARC, has recently been shown to reduce distortion in DWI and improve the image quality in single channel volume coil/body coil acquisitions. Here, we propose a new low rank reconstruction technique for robust reconstruction and improved image quality for DWI acquired using vARC’s EPI multi-shot acquisition scheme with single channel body coil. |
| 4966 | Computer 20
|
Joint K-TE reconstruction for T2 mapping with the weighted linear regression fitting |
| Yan Dai1, Jie Deng1, and Xun Jia2 | ||
1University of Texas Southwestern Medical Center, Dallas, TX, United States, 2Johns Hopkins University, Baltimore, MD, United States |
||
Keywords: Image Reconstruction, Quantitative Imaging We developed a joint quantitative imaging reconstruction algorithm to reconstruct T2-weighted (T2W) images and T2 mapping simultaneously through an iterative optimization process, which integrated the weighted linear regression fitting. The proposed method achieved accurate T2 measurements and improved signal-to-noise (SNR) of both T2W images and the T2 map. |
| 5103 | Computer 1
|
Synthesizing CT Image from Single-echo UTE-MRI using Multi-Task Framework |
| Zhuoyao Xin1, Vishwanatha Mitnala Rao2, Dong Liu3, Yanting Yang2, Ye Tian2, Chenghao Zhang2, Andrew F. Laine2, and Jia Guo2 | ||
1Biomedical Engineering, Columbia University, New York City, NY, United States, 2Columbia University, New York City, NY, United States, 3Neuroscience, Columbia University, New York City, NY, United States |
||
Keywords: Image Reconstruction, Multimodal This abstract proposed a cross-modality conversion method from UTE MRI to CT images. Using the TABS and ResidualAttentionU-net model in a processing framework combining image segmentation and prediction, the skull structure can be extracted from UTE MRI with high similarity of CT based skull. Five UTE-CT image pairs of mouse brains were used in the study. And a 3D-patch based training strategy was adopted, which took the advantage of structural continuity between slices in very limited datasets. The results show that the proposed combined image segmentation and prediction framework can achieve higher accuracy in medical image synthesizing for cross-modality conversion. |
| 5104 | Computer 2
|
Direct Imaging and T1 and T2* Quantification of Semi-Solid Red Blood Cell Membrane Lipid via Ultrashort Echo Time (UTE) Sequences |
| Soo Hyun Shin1, Dina Moazamian1, Arya Suprana1,2, Xiaojun Chen1, Jiyo S Athertya1, Chun Zeng1, Michael Carl3, Yajun Ma1, Hyungseok Jang1, and Jiang Du1,2,4 | ||
1Department of Radiology, University of California, San Diego, La Jolla, CA, United States, 2Department of Bioengineering, University of California, San Diego, La Jolla, CA, United States, 3GE Helathcare, San Diego, CA, United States, 4Radiology Service, Veterans Affairs San Diego Healthcare System, La Jolla, CA, United States |
||
Keywords: Relaxometry, Relaxometry Direct imaging of semi-solid lipids, such as cell membranes and myelin, is of great interest. Yet, the short T2 of semi-solid lipid protons hampers direct detection through conventional pulse sequences. In this study, we examined whether an ultrashort echo time (UTE) sequence can directly acquire signals from membrane lipids. Cell membranes from red blood cells were subject to D2O exchange and freeze-drying for complete water removal. High MR signals were detected with the UTE sequence, which showed an ultrashort T2* of ~77-252 µs and a short T1 of 189 ms for semi-solid membrane lipids. |
| 5105 | Computer 3
|
Motion Corrected Multi-scale Low Rank Reconstructions for Highly Accelerated 3D Dynamic Acquisitions |
| Zachary Miller1, Luis A Torres1, Sean Fain2, and Kevin M Johnson1 | ||
1University of Wisconsin-Madison, Madison, WI, United States, 2University of Iowa, Iowa City, IA, United States |
||
Keywords: Image Reconstruction, Motion Correction This work develops a method for large-scale motion compensated 4D reconstructions of fully 3D radial acquisitions at near 1mm isotropic spatial resolution and sub-second temporal resolution. We combine and extend multi-scale image reconstruction (Extreme MRI) with k-space based motion field estimation (MOTUS) to solve for a multi-scale low rank representations of both images and motion fields. This combined approach outperforms the multi-scale low rank reconstruction without motion fields. |
| 5106 | Computer 4
|
Slice-selective Zero Echo Time imaging of ultra-short 𝑇2 tissues |
| Jose Borreguero1, Fernando Galve2, Jose Miguel Algarín2, Jose María Benlloch2, and Joseba Alonso2 | ||
1Tesoro Imaging S.L., Valencia, Spain, 2Institute for Molecular Imaging and Instrumentation, Spanish National Research Council & Universitat Politècnica de València, Valencia, Spain |
||
Keywords: New Trajectories & Spatial Encoding Methods, New Signal Preparation Schemes Here we provide an MRI sequence which allows slice selection and 2D-imaging of hard tissues with T2 as short as 275 μs within clinically acceptable scan times even at fields as low as 260 mT. Our proposed sequence combines slice selection through spin-locking, which suffers a much more benign decay (T1ρ>>T2), and the fastest imaging sequence (ZTE), providing a new and robust tool for slice selection of the shortest-lived tissue signals in the body. |
| 5107 | Computer 5
|
Low-Rank Plus Sparse Accelerated Proton Resonance Frequency Shift and T1-mapping with a Dual-Echo 3D Spiral Ultra-Short Echo Time Sequence |
| Sheng Chen1, Zhixing Wang1, Yekaterina K. Gilbo2, Helen L. Sporkin1, Samuel W. Fielden3, Steven P. Allen4, John P. Mugler III5, G. Wilson Miller5, and Craig H. Meyer1,5 | ||
1Biomedical Engineering, University of Virginia, Charlottesville, VA, United States, 2Netflix, Los Gatos, CA, United States, 3U.S. Food and Drug Administration, Silver Spring, MD, United States, 4Brigham Young University, Provo, UT, United States, 5Radiology and Medical Imaging, University of Virginia, Charlottesville, VA, United States |
||
Keywords: Sparse & Low-Rank Models, Image Reconstruction, UTE, Spiral, Focused Ultrasound Unintended heating of the skull and nearby brain tissue during MR-guided Focused Ultrasound (MRgFUS) treatment is not standardly monitored. A method for addressing this issue, which combines variable flip angle (VFA) T1 mapping and proton resonance frequency (PRF) shift thermometry based on a dual-echo 3D spiral ultra-short echo time (UTE) acquisition, was accelerated with a low-rank plus sparse model in image reconstruction and validated using retrospectively undersampled data in vitro. |
| 5108 | Computer 6
|
DUDE: Diffusion tensor imaging with Ultra-short echo time Double Echo steady state – a proof of concept |
| Stefan Sommer1,2,3, Constantin von Deuster1,2,3, Tom Hilbert3,4,5, and Daniel Nanz2 | ||
1Siemens Healthineers International AG, Zurich, Switzerland, 2Swiss Center for Musculoskeletal Imaging (SCMI), Balgrist Campus, Zurich, Switzerland, 3Advanced Clinical Imaging Technology (ACIT), Siemens Healthineers International AG, Lausanne, Switzerland, 4Department of Radiology, Lausanne University Hospital and University of Lausanne, Lausanne, Switzerland, 5LTS5, École Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland |
||
Keywords: Data Acquisition, Pulse Sequence Design, DESS, UTE, UTE-DESS, DW-DESS, DW-UTE-DESS, ZTE The quality of echo-planar diffusion images can suffer from severe distortions and signal dropouts, especially in regions with high susceptibility gradients and rapidly decaying (short T2*) signal. To overcome these problems, we implemented a diffusion sensitive UTE-DESS sequence. The acquisition along six distinct diffusion directions allows for fitting of a diffusion tensor. We show the feasibility of Diffusion tensor imaging with Ultra-short echo time Double Echo steady state (DUDE) at 3T and present preliminary results in the brain of a healthy volunteer. |
| 5109 | Computer 7
|
Adapted rosette trajectory for functional 2D lung imaging |
| Hanna Frantz1, Patrick Metze1, and Volker Rasche1 | ||
1Ulm University Hospital, Ulm, Germany |
||
Keywords: New Trajectories & Spatial Encoding Methods, Lung, Function Lung MRI is a steadily evolving field of research, especially concerning the evaluation of chronic lung diseases such as cystic fibrosis or COPD. The major limitation of lung MRI is the ultrashort T2* relaxation time of lung parenchyma, which demands ultrashort echo time sequences. Sufficient SNR values in the parenchyma are crucial for clinical evaluations and the assessment of physiological parameters. This abstract presents an adapted rosette trajectory for UTE imaging, yielding higher SNR per unit time in comparison to radial UTE sampling approaches. |
| 5110 | Computer 8
|
Multi-contrast multi-resolution UTE for simultaneous quantitative and high-resolution MRI |
| Serhat Ilbey1, Antonia Susnjar2, Uzay Emir2,3, Michael Bock1, and Ali Caglar Özen1 | ||
1Division of Medical Physics, Department of Radiology, University Medical Center Freiburg, Freiburg, Germany, 2Weldon School of Biomedical Engineering, Purdue University, West Lafayette, IN, United States, 3School of Health Science Department, Purdue University, West Lafayette, IN, United States |
||
Keywords: New Trajectories & Spatial Encoding Methods, New Trajectories & Spatial Encoding Methods A multi-contrast multi-resolution UTE sequence was developed with a dual-echo acquisition between 1-1 binomial water excitation pulses and a conventional UTE readout. This acquisition scheme provides 3 images with different echoes, flip angles, spatial resolution, and contrast (water excitation). T2* was quantified in the whole brain in TA<15min, and UTE images of the brain were reconstructed with 0.5mm isotropic resolution without additional scan time. |
| 5111 | Computer 9
|
On the elimination of the fat saturation effect in ultrashort echo-time imaging of Achilles tendons. |
| Peter Latta1, Vladimir Juras2, Zenon Starcuk3, Martin Kojan1, Ivan Rektor1, Pavol Szomolanyi2, and Siegfried Trattnig2 | ||
1Masaryk University, Brno, Czech Republic, 2Medical University of Vienna, Vienna, Austria, 3Institute of Scientific Instruments, Czech Academy of Sciences, Brno, Czech Republic |
||
Keywords: Tendon/Ligament, Skeletal, Achilles tendon, Bi-exponential . T2* We investigated tri‑component analysis of 2D-UTE in-vivo measurements on a clinical 3T scanner and tested its ability to eliminate the effect of fat suppression on short T2* values in Achilles tendons. We found that fat-suppressed acquisition with two-component analysis provides ratios of short-T2* to long-T2* component intensities that are substantially lower (by a factor 0.95 to 0.55) than those resulting from fat-unsuppressed acquisition and tri-component fitting. Hence it seems that the combination of 2D-UTE acquisition with tri-component analysis might contribute to further improvement of T2* estimation accuracy. |
| 5112 | Computer 10
|
Quantification of Hemosiderin Deposition in Hemophilic Arthropathy Using Ultrashort Echo Time Quantitative Susceptibility Mapping (UTE-QSM) |
| Sam Sedaghat1, James V Luck2, Annette von Drygalski3, Scott T Ball4, Soo Hyun Shin1, Eddie Fu1, Eric Y Chang1,5, Jiang Du1,5,6, and Hyungseok Jang1 | ||
1Department of Radiology, University of California San Diego, San Diego, CA, United States, 2Department of Orthopaedic Surgery, University of California Los Angeles, Los Angeles, CA, United States, 3Department of Medicine, University of California San Diego, San Diego, CA, United States, 4Department of Orthopaedic Surgery, University of California San Diego, San Diego, CA, United States, 5Radiology Service, Veterans Affairs San Diego Healthcare System, San Diego, CA, United States, 6Department of Bioengineering, University of California San Diego, San Diego, CA, United States |
||
Keywords: MSK, Hematologic We investigated the quantitative capability of the UTE-QSM technique in the assessment of hemosiderin compared to histology. UTE-QSM and histology were performed in knee synovium tissues from hemophilia patients to visualize hemosiderin. The susceptibility maps from UTE-QSM showed a visually high similarity to the corresponding histological findings for all tissues. There was a significant difference in tissues with low and high iron load (p<0.001) in UTE-QSM. The susceptibility values on UTE-QSM showed a significant strong positive correlation to the blue signal in histology (R=0.908; p<0.001). We showed that UTE-QSM has excellent performance for detecting and quantifying hemosiderin in HA. |
| 5113 | Computer 11
|
Quantitative Ultrashort Echo-Time MRI to Assess In Vivo Rotator Cuff Tendon Degeneration |
| Misung Han1, Peder EZ Larson1,2, Thomas M Link1, Drew A Lansdown3, Brian T Feeley3, and Sharmila Majumdar1,2 | ||
1Radiology and Biomedical Imaging, University of California, San Francisco, San Francisco, CA, United States, 2UCSF-UC Berkely Graduate Program in Bioengineering, University of California, San Francisco, San Francisco, CA, United States, 3Department of Orthopaedic Surgery, San Francisco, CA, United States |
||
Keywords: Tendon/Ligament, Quantitative Imaging Rotator cuff tears are a common source of pain and disability. Surgical treatment is normally considered for patients who fail in non-operative management, but a failure of tendon healing after surgery can be dependent on tendon quality. In this work, we applied two ultrashort-echo time (UTE) quantification imaging methods, UTE T2* quantification as well as UTE magnetization transfer (MT) quantification, on healthy subjects and patients with rotator cuff tears, and evaluated their ability to assess tendon degeneration. Significant difference was observed for quantified T2* and MT parameters over segmented supraspinatus tendon between the healthy control and patient groups. |
| 5114 | Computer 12
|
Robust assessment of macromolecular fraction in muscle with differing fat fraction using ultrashort echo time magnetization transfer modeling |
| Saeed Jerban1, Yajun Ma1, Qingbo Tang2, Eddie Fu2, Nikolaus Szeverenyi1, Hyungseok Jang1, Christine B Chung1, Jiang Du1, and Eric Y Chang1,2 | ||
1Radiology, University of California, San Diego, San Diego, CA, United States, 2Radiology Service, Veterans Affairs San Diego Healthcare System, San Diego, La Jolla, CA, USA, San Diego, CA, United States |
||
Keywords: Muscle, Magnetization transfer, Collagen Content; Fat Content Magnetization transfer (MT) MR imaging can indirectly characterize the spatial distribution of the relative contents of the macromolecular and water proton pools (MMF) of skeletal muscles. Fat presence in muscle has always been a source of concern in MMF calculation where some studies have reported significant underestimation of MMF for muscles with a considerable fat fraction (FF). We investigated the impact of FF on MMF in muscle/fat phantoms using an ultrashort echo time (UTE) MT model after T1 compensation. MMF demonstrated a relatively robust value with under 5% and 20% changes for FF increases up to 30 and 45%, respectively. |
| 5115 | Computer 13
|
Ultrashort Echo Time Magnetization Transfer Imaging of Knee Cartilage After Long-Distance Running |
| Dantian Zhu1, Yijie Fang1, Wenhao Wu1, Shaolin Li1, Long Qian2, and Yajun Ma3 | ||
1Department of Radiology,Fifth Affiliated Hospital, Sun Yat-Sen University, Zhuhai, China, 2MR Research, GE Healthcare, Beijing, China, 3University of California, San Diego, Department of Radiology, San Diego, CA, United States |
||
Keywords: Cartilage, Cartilage To assess the detection of changes in the knee cartilage of amateur marathon runners before and after long-distance running. We recruited 23 amateur marathon runners prospectively. MRI scans using UTE-MT and UTE-T2* sequences. UTE-MTR and UTE-T2* were measured for knee cartilage. The UTE-MTR values in lateral tibial plateau, central medial femoral condyle, and medial tibial plateau showed a significant decrease at 2 days post-race compared to the other two time points (P < 0.05). But no significant UTE-T2* changes were found for any cartilage subregions. UTE-MTR is a promising method for the detection of dynamic changes in knee cartilage. |
| 5116 | Computer 14
|
Feasibility of detecting subvoxel deformation in intervertebral disc using quantitative ultrashort echo time (UTE) techniques |
| Saeed Jerban1,2,3, Dina Moazamian1, Sheronda Statum1, Hyungseok Jang1,2, Eric Y Chang1,2, Jiang Du1,2, Christine B Chung1,2, and Yajun Ma1 | ||
1Department of Radiology, University of California, San Diego, La Jolla, CA, USA, San Diego, CA, United States, 2Radiology Service, Veterans Affairs San Diego Healthcare System, San Diego, La Jolla, CA, USA, San Diego, CA, United States, 3Department of Orthopedic Surgery, University of California, San Diego, La Jolla, CA, USA, San Diego, CA, United States |
||
Keywords: Skeletal, Magnetization transfer, Intervertebral disc Quantitative ultrashort echo time (UTE) MRI can be used for quantitative assessment of intervertebral discs (IVDs). It is hypothesized that the investigation of quantitative UTE MRI properties of IVD under mechanical loading may highlight the affected regions of IVD by diseases and injuries. We investigated the feasibility of using UTE-T1, UTE-Adiab-T1ρ, and UTE-MT measures for detecting the IVD deformation under loading. T1 and T1 ρ decreased in IVDs under loading while MMF from UTE-MT modeling as an index for collagen content increased by loading. This study highlights the potential of UTE-MRI to detect subvoxel deformations in IVDs caused by loading. |
| 5117 | Computer 15
|
Progress toward a ZTE-based silent and motion-robust protocol for pediatric neuroimaging |
| James H Holmes1, Vincent A Magnota1, Mathews Jacob2, Yan Chen2, Joshua Hanson3, Paul A DiCamillo1, and Curtis A Corum3 | ||
1Radiology, University of Iowa, Iowa City, IA, United States, 2Electrical and Computer Engineering, University of Iowa, Iowa City, IA, United States, 3Champaign Imaging, LLC, Shoreview, MN, United States |
||
Keywords: Neuro, Pulse Sequence Design, silent imaging In this work, we present progress in developing a ZTE-based silent and motion-robust neuroimaging protocol using intermittent magnetization preparation for generation of standard imaging contrasts including T1w and T2w. |
| 5118 | Computer 16
|
Density-Controllable Spiral Trajectory Design Combined with CS-UTE Improves High-Resolution Single Breath-Hold Lung Imaging |
| Yupeng Cao1, Jun Zhao1, Weinan Tang2, Wentao Liu1, and Dong Han1 | ||
1National Center for Nanoscience and Technology, Beijing, China, 2Wandong Medical Inc, Beijing, China, Beijing, China |
||
Keywords: New Trajectories & Spatial Encoding Methods, Lung Single breath-hold high-resolution lung imaging is challenging due to the short T2* and scan efficiency. The stack-of-spirals UTE enables fast lung imaging in a single breath hold. However, fast scanning of stack-of-spirals results in a long readout time, introducing adverse effects of the short T2* of the lung. Herein, we proposed a density-controllable spiral trajectory (DCST) design method to design a short readout time trajectory concurrently satisfying the criteria of optimal compressed sensing (CS), reducing the adverse effect of the short T2*. The short readout time trajectory combined with CS-UTE improves the single breath-hold high-resolution lung imaging. |
| 5119 | Computer 17
|
Multi-Echo FLORET UTE MRI |
| Matthew Willmering1, Guruprasad Krishnamoorthy2,3, Joseph Plummer1,4, Abdullah Bdaiwi1, Laura Walkup1,4,5, Zackary Cleveland1,4,5, James Pipe3, and Jason Woods1,4,5 | ||
1Cincinnati Children’s Hospital Medical Center, Cincinnati, OH, United States, 2Philips Healthcare, Rochester, MN, United States, 3Mayo Clinic, Rochester, MN, United States, 4University of Cincinnati, Cincinnati, OH, United States, 5University of Cincinnati Medical Center, Cincinnati, OH, United States |
||
Keywords: Data Acquisition, Data Acquisition FLORET UTE MRI has been used for many applications including lung, brain, musculoskeletal, and multinuclear imaging. The appeal of the pulse sequence comes from its high SNR and sampling efficiencies. With recent improvements minimizing artifacts, high-quality FLORET images are obtained routinely. However, acquisition of multiple echoes has not been investigated. Here, we determined the feasibility and optimal pulse-sequence design of multi-echo FLORET. Collecting echoes independently, but interleaved, produces the highest quality images, especially for ≥3 echoes. Furthermore, additional echoes provide improvements to image quality and add clinically useful contrast like T2*. |
| 5120 | Computer 18
|
Assessing the clinical value of PETRA sequence for detection of subsolid pulmonary nodules：comparison with CT |
| Hui Feng1, Li Yang1, Chen Zhang2, Hui Liu1, Ning Zhang1, and Gaofeng Shi1 | ||
1The Fourth Hospital of Hebei Medical University, Shijiazhuang, China, 2Siemens Healthineers, Beijing, China |
||
Keywords: Visualization, Lung The PETRA technique had high sensitivity for the detection of subsolid pulmonary nodules and can accurately assess their diameter and morphologic characteristics. |
| 5121 | Computer 19
|
Myelin Content Imaging Using Ultra-Short TE MRI with Variable Flip Angles |
| Kwan-Jin Jung1, Ryan Larsen2, Laurie Rund2, and Andrew Steelman3,4,5 | ||
1Biomedical Imaging Center, Beckman Institute, University of Illinois at Urbana-Champaign, Urbana, IL, United States, 2Department of Animal Sciences, University of Illinois at Urbana-Champaign, Urbana, IL, United States, 3Neuroscience Program, 2325/21, University of Illinois at Urbana-Champaign, Urbana, IL, United States, 4Carl R. Woese Institute for Genomic Biology, University of Illinois at Urbana-Champaign, Urbana, IL, United States, 5Division of Nutritional Sciences, University of Illinois at Urbana-Champaign, Urbana, IL, United States |
||
Keywords: Relaxometry, Neuro, Myelin, Piglet Myelin content was measured from the fast T1 relaxation component using bi-exponential T1 relaxation regression. The data was collected using UTE with variable flip angles to detect short T2 signal of myelin and to avoid magnetic susceptibility corruption by T2*-based myelin contrast methods. The estimated myelin content was influenced by CSF, which was suppressed by use of the slow T1 relaxation time. The estimated myelin content was higher in white matter than other brain regions. However, the myelin content was increased in the anterior pole and low in motor areas in this in-vivo piglet data. |
| 5122 | Computer 20
|
Validation of open-source minimal TE sequences across three independent sites |
| Gehua Tong1, Andreia S. Gaspar2, Rita G. Nunes2, Jon-Fredrik Nielsen3, John Thomas Vaughan, Jr.1,4, and Sairam Geethanath4,5 | ||
1Biomedical Engineering, Columbia University, New York, NY, United States, 2Institute for Systems and Robotics (ISR-Lisboa)/LaRSyS, Department of Bioengineering, Instituto Superior Técnico, Universidade de Lisboa, Lisbon, Portugal, 3Biomedical Engineering, University of Michigan, Ann Arbor, MI, United States, 4Columbia Magnetic Resonance Research Center, Columbia University, New York, NY, United States, 5Accessible MR Laboratory, BioMedical Engineering and Imaging Institute, Dept. of Diagnostic, Molecular and Interventional Radiology, Icahn School of Medicine at Mt. Sinai, New York, NY, United States |
||
Keywords: Validation, Pulse Sequence Design We validated five open-source PyPulseq minimal TE sequences (half/full pulse 2D/3D UTEs and COncurrent Dephasing and Excitation (CODE)) at three sites with different scanner models. The ISMRM/NIST T2 array and a porcine muscle and bone sample were imaged. SNR, contrast-to-noise ratio (CNR), and T2 contrast curves were calculated and compared. We found that the sequences were able to recover more SNR from bone tissue compared to vendor GRE (an additional 50% - 1060% for 3T and 34% - 474% for 1.5T) and higher signal for short-T2 spheres but intersite differences existed that were consistent with field strengths. |