To begin searching the abstracts, please use the search feature above.
| 3338 | Quantitative abdominal imaging using 3D motion-robust MR fingerprinting | |
| Max van Riel1,2, Zidan Yu2,3, Shota Hodono2,3, Ding Xia2, Hersh Chandarana2, and Martijn A. Cloos2,3 | ||
1Department of Biomedical Engineering, Eindhoven University of Technology, Eindhoven, Netherlands, 2Center for Advanced Imaging Innovation and Research and Bernard and Irene Schwartz Center for Biomedical Imaging, Department of Radiology, New York University School of Medicine, New York, NY, United States, 3Sackler Institute of Graduate Biomedical Sciences, New York University School of Medicine, New York, NY, United States |
||
We demonstrate a 3D MR fingerprinting sequence for quantitative abdominal imaging. The sequence was made robust to motion by modifying the order of acquisition to allow for free-breathing imaging. Furthermore, the flip angle pattern was optimized using the Cramér-Rao Lower Bound to increase the efficiency of the sequence. A phantom was used to validate the new sequence and it was shown that the motion-robust ordering reduced motion-related artefacts. In vivo results showed reduced artefacts as well. We conclude that it is possible to generate B1-robust quantitative T1 and proton density maps at a clinically usable resolution within 5 minutes. |
| 3339 | Motion Correction of T2SE Multi-Pass Acquisitions for Super-Resolution in the Slice Direction | |
| Eric A. Borisch1, Soudabeh Kargar1, Akira Kawashima2, and Stephen J. Riederer1 | ||
1Radiology, Mayo Clinic, Rochester, MN, United States, 2Radiology, Mayo Clinic, Phoenix, AZ, United States |
||
A new method for retrospectively correcting subtle transverse slice-to-slice motion in a set of overlapped axial slices is introduced. Such motion can be problematic for high through-plane resolution imaging of the pelvis due to peristalsis of the rectum. The method exploits the known slice-to-slice correlation to determine offsets as well as characteristics of the multi-pass acquisition. When applied to 16 T2-weighted fast spin-echo exams of the prostate the method shows significant improvement in prostate-rectum margin clarity and overall image quality as assessed on sagittal reformats from axial images. |
| 3340
|
Non-rigid “image” registration in k-space | |
| Thomas Küstner1,2,3, Christopher Gilliam4, Thierry Blu5, Martin Schwartz3,6, Gastao Cruz1, Jiazhen Pan3, Christian Würslin2, Nina F Schwenzer7, Holger Schmidt7, Bin Yang3, Konstantin Nikolaou8, René M Botnar1, Claudia Prieto1, and Sergios Gatidis2,8 | ||
1Biomedical Engineering Department, School of Biomedical Engineering and Imaging Sciences, King's College London, London, United Kingdom, 2Medical Image and Data Analysis (MIDAS), University Hospital Tübingen, Tübingen, Germany, 3Institute of Signal Processing and System Theory, University of Stuttgart, Stuttgart, Germany, 4RMIT University, Melbourne, Australia, 5Chinese University Hong Kong, Hong Kong, Hong Kong, 6Section on Experimental Radiology, University Hospital Tübingen, Tübingen, Germany, 7University of Tübingen, Tübingen, Germany, 8Department of Radiology, University Hospital Tübingen, Tübingen, Germany |
||
Non-rigid motion estimation is an important task in correction of respiratory and cardiac motion. Usually this problem is formulated in image space via diffusion, parametric-spline or optical flow methods. For these applications non-rigid motion commonly needs to be estimated from cardiac or respiratory states which are highly subsampled in k-space. Image-based registration can be impaired by aliasing artefacts or by estimation from low-resolution images. In this work, we propose a novel non-rigid registration technique directly in k-space based on optical flow. The proposed method is compared against image-based registrations and tested on fully-sampled and highly-accelerated 3D motion-resolved MR imaging. |
| 3341 | Self-navigated abrupt Motion Correction for SNAP with 3D golden angle radial k-space sampling (GOAL-SNAP) in Carotid Artery T1 mapping | |
| Jiaqi Dou1, Yajie Wang1, Chunyao Wang1, Haikun Qi2, Yu Wang1, and Huijun Chen1 | ||
1CBIR, Center for Biomedical Imaging Reserch, Tsinghua University, Beijing, China, 2School of Biomedical Engineering and Imaging Sciences, King's College London, London, United Kingdom |
||
GOAL-SANP is a promising technique for carotid atherosclerotic plaque T1 mapping using 3D golden angle radial acquisition. However, sever patient abrupt motion is still a challenge for this sequence. In this study, a self-navigated retrospective motion correction scheme based on independent components analysis (ICA) and peak filtering was developed for GOAL-SNAP utilizing the advantage of its 3D golden angle radial trajectory. The results validated the feasibility of the proposed method to detect the abrupt motion (cough) and correct the related artifacts for carotid vessel wall imaging. |
| 3342 | Impact of cardiac function on objective evaluation of motion correction on cardiovascular magnetic resonance T1 mapping | |
| Sixian Hu1, Wanlin Peng1, Huayan Xu2, Xiaoyue Zhou3, Chunchao Xia1, and Zhenlin Li1 | ||
1Department of Radiology, West China Hospital of Sichuan University, Chengdu 610041, China, Chengdu, China, 2Department of Radiology, Key Laboratory of Obstetric & Gynecologic and Pediatric Diseases and Birth Defects of Ministry of Education, West China Second University Hospital, China, Chengdu, China, 3MR Collaboration, Siemens Healthineers Ltd., Shanghai, China, Shanghai, China |
||
Quantitative evaluation of whether cardiac function affects the motion correction effect on cardiovascular magnetic resonance T1 mapping was conducted. Pre- and post-contrast T1 values, ECV values, SNR, and CNR were measured and compared between MOCO and non-MOCO images in groups with either preserved left ventricular function or impaired left ventricular function. Motion-corrected images showed good T1 and ECV value agreement compared to images without correction. The group with preserved left ventricular function showed greater improvement after motion correction than the impaired left ventricular function group. |
| 3343
|
Evaluation of motion correction capability in retrospective motion correction with MoCo MedGAN | |
| Thomas Küstner1,2,3, Friederike Gänzle3, Tobias Hepp2, Martin Schwartz3,4, Konstantin Nikolaou5, Bin Yang3, Karim Armanious2,3, and Sergios Gatidis2,5 | ||
1Biomedical Engineering Department, School of Biomedical Engineering and Imaging Sciences, King's College London, London, United Kingdom, 2Medical Image and Data Analysis (MIDAS), University Hospital Tübingen, Tübingen, Germany, 3Institute of Signal Processing and System Theory, University of Stuttgart, Stuttgart, Germany, 4Section on Experimental Radiology, University Hospital Tübingen, Tübingen, Germany, 5Department of Radiology, University Hospital Tübingen, Tübingen, Germany |
||
Motion is the main extrinsic source for imaging artifacts which can strongly deteriorate image quality and thus impair diagnostic accuracy. Numerous motion correction strategies have been proposed to mitigate or capture the artifacts. These methods require some a-priori knowledge about the expected motion type and appearance. We have recently proposed a deep neural network (MoCo MedGAN) to perform retrospective motion correction in a reference-free setting, i.e. not requiring any a-priori motion information. In this work, we propose a confidence-check and evaluate the correction capability of MoCo MedGAN with respect to different motion patterns in healthy subjects and patients. |
| 3344 | High efficiency Free-Breathing 3D Thoracic Aorta Imaging with Self-Navigated Image Reconstruction | |
| Caiyun Shi1,2, Congcong Liu1, Shi Su1, Haifeng Wang1, Xin Liu1, Hairong Zheng1, Dong Liang1, and Yanjie Zhu1 | ||
1Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China, 2Shenzhen College of Advanced Technology, University of Chinese Academy of Sciences, Shenzhen, China |
||
Self-gating techniques can be used to solve and compensate for cardiac or respiratory motion during MRI with free-breathing. In this work, a self-gating based motion correction scheme is proposed, and combined with a 3D variable-flip-angle TSE sequence for high-efficiency thoracic aorta imaging. Specifically, the slab-selective SPACE sequence is modified to acquire self-gating signals, which are used for detecting the respiratory motion. The image data is subsequently corrected based on the binning motion correction and image registration approach. The comparison was conducted on healthy volunteers and compared against a conventional diaphragmatic navigator-gated acquisition to assess the feasibility of the proposed scheme. |
| 3345 | Machine learning based Magnetic tracking technique using a Mapping of the Magnetic Gradient Fields on a 3T MRI scanner | |
| Benjamin Roussel1,2, Joris Pascal3, Nicolas Weber1,2, Philip Keller4, Antoine Daridon4, Jacques Felblinger1,2, and Julien Oster1,2 | ||
1IADI, U1254, INSERM, Nancy, France, 2Université de Lorraine, Nancy, France, 3FHNW, University of Applied Sciences and Arts Northwestern Switzerland, Muttenz, Switzerland, 4Metrolab Technology SA, Plan-les-Ouates, Switzerland |
||
This paper presents a magnetic tracking method for the localization of a three-axis magnetometer within an MRI bore (Prisma, Siemens, Erlangen, Germany). The unique relationship between the Magnetic Gradient Fields (MGF) and the position within the bore allows locating the sensor through the use of a mapping of the MGF. This mapping is used for the training of a multi-layer perceptron neural network, which estimates the position of the sensor when measuring the MGF. Our technique was experimentally validated by moving the sensors within the MRI bore while playing a customized pulse sequence and by reconstructing the movements during post-processing. |
| 3346 | A Synthetic Combination of Accurate De-enhanced Registration and Dynamic Artificial Sparsity for Robust High-Resolution Liver DCE-MRI | |
| Zhifeng Chen1, Yujia Zhou1, Xinyuan Zhang1, Peiwei Yi1, Zhongbiao Xu2, Jian Gong1, Zhenguo Yuan3, Xia Kong4, Yaohui Wang5, Ling Xia6, Wufan Chen1, Yanqiu Feng1, and Feng Liu7 | ||
1School of Biomedical Engineering, Guangdong Provincial Key Laboratory of Medical Image Processing, Southern Medical University, Guangzhou, China, 2Department of Radiotherapy, Cancer Center, Guangdong Provincial People's Hospital & Guangdong Academy of Medical Science, Guangzhou, China, 3Shandong Medical Imaging Research Institute, Shandong University, Jinan, China, 4School of Computer and Information Science, Hubei Engineering University, Wuhan, China, 5Division of Superconducting Magnet Science and Technology, Institute of Electrical Engineering, Chinese Academy of Sciences, Beijing, China, 6Department of Biomedical Engineering, Zhejiang University, Hangzhou, China, 7School of Information Technology and Electrical Engineering, The University of Queensland, Brisbane, Australia |
||
High spatiotemporal DCE-MRI is a valuable tool in liver disease diagnoses and treatments. Recently, there is a growing research trend which focuses on the motion-robustness of liver DCE-MRI. However, current techniques cannot simultaneously solve the motion problem when pursuing high spatiotemporal resolution. In this work, we propose to combine an accurate registration technique with dynamic artificial sparsity for high spatiotemporal resolution DCE-MRI of liver. The experiments indicated that the proposed framework results in better image quality than iGRASP due to de-enhanced image registration. Compared to motion-sorting techniques, the proposed framework generates better temporal resolution. |
| 3347 | MR-Based Cardio-Respiratory Motion Correction of Simultaneously Acquired PET Images Through Coil Fingerprints | |
| David Rigie1, Thomas Vahle2, Tiejun Zhao3, Klaus Schaffers4, and Fernando Boada5 | ||
1New York University, New York, NY, United States, 2Siemens Healthineers, Erlangen, Germany, 3Siemens Medical Solutions, New York, NY, United States, 4European Institute for Molecular Imaging, Munster, Germany, 5Radiology, New York University, New York, NY, United States |
||
We present of a real-time, motion correction methodology for MR/PET acquisitions. Our approach utilizes self-refocused navigators that could be integrated into any pulse sequence for providing cardio-respiratory motion information after each excitation. We demonstrate that this approach improves small lesion detection sensitivity and quantitative accuracy for simultaneously acquired MR/PET scans. |
| 3348 | MR compatible 4D Ultrasound based validation of respiratory motion compensation strategies | |
| Zachary Miller1, James Holmes1, Sydney Jupitz1, Ty Cashen2, Frank Ong3, Michael Lustig4, Peder Larson5, Bryan Bednarz6, and Kevin Johnson1 | ||
1Department of Medical Physics, University of Wisconsin-Madison, Madison, WI, United States, 2GE Healthcare, University of Wisconsin-Madison, Madison, WI, United States, 3Department of Electrical Engineering, Stanford University, Stanford, CA, United States, 4Department of Electrical Engineering, University of California-Berkely, Berkeley, CA, United States, 5Department of Radiology and Biomedical Imaging, University of California San Francisco, San Francisco, CA, United States, 6Department of Engineering Physics, University of Wisconsin-Madison, Madison, WI, United States |
||
Imaging during respiratory and cardiac motion remains a major challenge for body and cardiac MR. A number of motion correction approaches have been proposed to address this challenge including use of external gating signals, navigator acquisitions, and motion estimation from the image data itself. However, the accuracy of these motion estimation techniques has not been validated in-vivo. In this work, we use a recently developed MR compatible 4D ultrasound probe combined with feature tracking to evaluate the performance of MR based navigation algorithms in the setting of free-breathing 3D pulmonary imaging. |
| 3349 | Quantitative Assessment MR-Assisted PET Respiratory Motion Correction in Colorectal Liver Metastases using PET/MR | |
| Sihao Chen1,2, Cihat Eldeniz2, Yasheng Chen3, and Hongyu An2 | ||
1Biomedical Engineering, Washington University in St. Louis, Saint Louis, MO, United States, 2Mallinckrodt Institute of Radiology, Washington University in St. Louis, Saint Louis, MO, United States, 3Department of Neurology, Washington University in St. Louis, Saint Louis, MO, United States |
||
Respiratory motion causes signal blurring and reduced tumor-to-background ratio (TBR). Simultaneous PET/MR imaging uniquely allows for MR-assisted motion correction in PET imaging, potentially leading to improved PET images for detection of lesions. In this study, we implemented several MR-assisted PET motion correction methods, including gated reconstruction (gated MoCo), reconstruct-transform-average (RTA) and motion-compensated image reconstruction (MCIR), to FDG imaging of colorectal liver metastases. We quantitatively compared TBR and CNR in FDG avid liver lesions. Our results demonstrated improvement of TBR and CNR using MCIR. |
| 3350 | Non-rigid Respiratory Motion Estimation of Coronary MR Angiography using Unsupervised Fully Convolutional Neural Network | |
| Haikun Qi1, Gastao Cruz1, Thomas Kuestner1, Niccolo Fuin1, René Botnar1, and Claudia Prieto1 | ||
1School of Biomedical Engineering and Imaging Sciences, King's College London, London, United Kingdom |
||
Non-rigid motion corrected coronary MR angiography (CMRA) in combination with 2D image-based navigators has been proposed to account for the complex respiratory-induced motion of the heart in undersampled acquisitions. However, this framework requires the efficient and accurate estimation of non-rigid bin-to-bin motion from undersampled respiratory-resolved images. In this study, we aim to investigate the feasibility of using an unsupervised fully convolutional network to estimate non-rigid motion from undersampled respiratory-resolved CMRA. The performance of the proposed approach was evaluated on 5-fold accelerated free-breathing CMRA and validated against a widely used conventional non-rigid registration method. |
| 3351 | Effective Flip Angle of Enhanced Navigator-Gated 3D Spoiled Gradient-Recalled Echo Sequence | |
| Yuji Iwadate1, Atsushi Nozaki1, Keisuke Sato2, Kengo Yoshimitsu2, Ryotaro Jingu3, Ryuji Nakamuta3, and Hiroyuki Kabasawa1 | ||
1Global MR Applications and Workflow, GE Healthcare Japan, Hino, Japan, 2Department of Radiology, Fukuoka university, Fukuoka, Japan, 3Radiology Center, Fukuoka University Hospital, Fukuoka, Japan |
||
An enhanced navigator-gated 3D-SPGR (eNAV-3D-SPGR) enables free-breathing T1-weighted abdominal MRI with navigator echo signal enhancement. In this study, we calculated effective flip angle of eNav-3D-SPGR in simulation and examined its validity for obtaining desired contrast. Simulation shows that the effective flip angle of 30° was achieved with eNav-3D-SPGR when the flip angle was set to 36.8°. Actual scan resulted in the similar signal ratios between the conventional method with flip angle 30° and eNav-3D-SPGR at flip angle of 37°. eNav-3D-SPGR with the desired effective flip angle can be useful for accurate motion detection when the liver MRI signal is low. |
| 3352 | Segmented Twin-resolution Acquisition of Cones Trajectory (STACY) and Coalesced Image Navigator for Free-breathing Cardiac MRI | |
| Kwang Eun Jang1,2, Dwight G Nishimura3, and Shreyas S Vasanawala4 | ||
1Magnetic Resonance Systems Research Lab (MRSRL), Stanford University, Stanford, CA, United States, 2Department of Bioengineering, Stanford University, Stanford, CA, United States, 3Magnetic Resonance Systems Research Lab (MRSRL), Department of Electrical Engineering, Stanford University, Stanford, CA, United States, 4Department of Radiology, Stanford University, Stanford, CA, United States |
||
3D image-based navigators (iNAVs) are effective for tracking motion in free-breathing cardiac MRI, though they require a long acquisition window. In this work, we present Segmented Twin-resolution Acquisition of Cones trajectorY (STACY) and the concept of coalesced image navigators. In STACY, different segments of a 3D low-resolution cones trajectory are collected across heartbeats as navigators. Coalesced image navigators are reconstructed by merging data from similarly translated heartbeatsh, leading to better conditioned images than iNAVs from undersampled data collected over individual heartbeats. |
| 3353 | Reduced Motion Artifact in Super Resolution T2 FSE Multislice MRI: Application to Prostate | |
| Soudabeh Kargar1, Eric A Borisch2, Adam T Froemming2, Roger C Grimm2, Akira Kawashima3, Bernard F King2, Eric G Stinson2, and Stephen J Riederer2 | ||
1Mayo Graduate School, Rochester, MN, United States, 2Mayo Clinic, Rochester, MN, United States, 3Mayo Clinic, Scottsdale, AZ, United States |
||
T2-weighted fast spin echo is acquired in virtually all clinical MRI exams; however, it has coarse through-plane vs. in-plane resolution and is thus often acquired in multiple orientations. This extends the exam time. We have shown improved through-plane resolution in T2 FSE prostate MRI by acquiring overlapped slices and accounting for the slice profile in the reconstruction. Due to acquisition of overlapped slices in several passes, an artifact was observed in the reformatted images, due to subtle peristaltic motion. We propose acquiring the phase encoding lines in several segments to suppress the motion artifact observed in the reformatted images. |
| 3354 | Motion Correction for a Multi-Contrast Brain MRI using a Multi-Input Neural Network | |
| Jongyeon Lee1, Byungjai Kim1, Namho Jeong1, and Hyunwook Park1 | ||
1Korea Advanced Institute of Science and Technology, Daejeon, Republic of Korea |
||
Numerous motion correction methods have been developed to reduce motion artifacts and improve image quality in MRI. Conventional techniques utilizing motion measurement required a prolonged scan time or intensive computational costs. Deep learning methods have opened up a new way for motion correction without motion information. A proposed method using a multi-input neural network with the structural similarity loss takes an advantage of a common clinical setting of multi-contrast acquisition to clearly correct motion artifacts in brain imaging. Motion artifacts can be fully retrospectively and greatly reduced without any motion measurement by the proposed method. |
| 3355 | Head motion estimation and correction using slab-selective FIDnavs | |
| Nurten Ceren Askin Incebacak1, Tess E. Wallace2, Tobias Kober3,4,5, Francois Lazeyras1, Simon K. Warfield2, and Onur Afacan2 | ||
1Department of Radiology and CIBM, University of Geneva, Geneva, Switzerland, 2Department of Radiology, Boston Children's Hospital and Harvard Medical School, Boston, MA, United States, 3Advanced Clinical Imaging Technology, Siemens Healthcare, Lausanne, Switzerland, 4Department of Radiology, UNIL and CHUV, Laussanne, Switzerland, 5LTS5, EPFL, Lausanne, Switzerland |
||
In this work, we propose to measure and correct rigid body head motion using slab-selective FID navigators. It is possible to estimate motion without relying on any |
| 3356 | Prospectively corrected dual-echo EPI using optical tracking | |
| Onur Afacan1, W. Scott Hoge2, Tess E. Wallace1, Tobias Kober3,4,5, Daniel Nicolas Splitthoff6, Ali Gholipour1, Sila Kurugol1, Camilo J. Cobos1, Richard L. Robertson1, and Simon K. Warfield1 | ||
1Department of Radiology, Boston Children's Hospital and Harvard Medical School, Boston, MA, United States, 2Brigham and Women's Hospital and Harvard Medical School, Boston, MA, United States, 3Advanced Clinical Imaging Technology, Siemens Healthcare, Lausanne, Switzerland, 4Department of Radiology, Lausanne University Hospital and University of Lausanne, Lausanne, Switzerland, 5École Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland, 6Siemens Healthcare, Erlangen, Germany |
||
Large head motion induces different local magnetic-field inhomogeneities, even if the field of view is corrected prospectively. In this work, we implemented and evaluated a diffusion-weighted dual-echo EPI sequence that prospectively corrects for motion using real-time measurements from an optical tracker and uses echoes acquired with reversed phase-encoding to correct for the distortions resulting from the induced local magnetic field inhomogeneities. We evaluated our motion and distortion correction framework in volunteer experiments undergoing controlled motion, and in pediatric patients undergoing routine MRI. Prospective motion correction using our proposed method produced high-quality diffusion parameter maps in all volunteer and patient scans. |
| 3357 | Prospective motion correction using real-time FID navigator motion measurements | |
| Tess E. Wallace1,2, Onur Afacan1,2, Tobias Kober3,4,5, and Simon K. Warfield1,2 | ||
1Computational Radiology Laboratory, Boston Children's Hospital, Boston, MA, United States, 2Harvard Medical School, Boston, MA, United States, 3Advanced Clinical Imaging Technology, Siemens Healthcare, Lausanne, Switzerland, 4Radiology, Lausanne University Hospital and University of Lausanne, Lausanne, Switzerland, 5LTS5, École Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland |
||
FIDnavs can be acquired extremely rapidly using standard scanner hardware and are consequently an attractive tracking strategy for prospective motion correction (PMC), which requires accurate pose updates to be passed to the sequence with minimal delay. In this work, we demonstrate for the first time the efficacy of PMC using FIDnav motion estimates in a moving phantom and in volunteers performing deliberate head motion. Real-time pose updates from measured FIDnavs enabled substantial improvements in image quality in structural scans acquired with motion. FID-navigated PMC is a promising method for motion-robust imaging of patients who have difficulty staying still during imaging. |
| 3358 | Clinical Utility of Deep Learning Motion Correction for Neuroimaging | |
| Kamlesh Pawar1, Jarrel Seah2, Meng Law3, Tom Close4, Zhaolin Chen1, and Gary Egan1 | ||
1Monash Biomedical Imaging, Monash University, Melbourne, Australia, 2Department of Neuroscience, Monash University, Melbourne, Australia, 3Radiology and Nuclear Medicine, Alfred Health, Melbourne, Australia, 4Monash Biomedical Imging, Monash University, Melbourne, Australia |
||
The deep learning techniques have been shown to reduce the motion artifact in simulated motion scenarios and a few volunteer scans, the validation of it during routine clinical scans remains an unanswered question. In this study, we focus on evaluating the quality of images from the DL motion correction approach on a cohort of 27 actual patient motion cases that were obtained from the routine clinical scans. Two board-certified radiologists evaluated 9 anatomical regions in the 3D MPRAGE brain images and rated them on the 3-point scale. |
| 3359 | Motion Correction Strategy for Multi-Contrast based 3D parametric imaging: Application to Inhomogeneous Magnetization Transfer (ihMT) | |
| Lucas Soustelle1,2, Julien Lamy3, Arnaud Le Troter1,2, Patrick Viout1,2, Lauriane Pini1,2, Claire Costes1,2, Jean-Philippe Ranjeva1,2, Maxime Guye1,2, Adil Maarouf1,2,4, Bertrand Audoin1,2,4, Clémence Boutière1,2,4, Audrey Rico1,2,4, Gopal Varma5, David C Alsop5, Jean Pelletier1,2,4, Olivier M Girard1,2, and Guillaume Duhamel1,2 | ||
1Aix-Marseille Univ, CNRS, CRMBM, Marseille, France, 2APHM, Hôpital Universitaire Timone, CEMEREM, Marseille, France, 3Université de Strasbourg, CNRS, ICube, FMTS, Strasbourg, France, 4APHM, Hôpital Universitaire Timone, Service de neurologie, Marseille, France, 5Division of MR Research, Radiology, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, United States |
||
Multi-contrast-based parametric MRI techniques provide parametric maps intrinsically sensitive to voxel misalignment upon images combination. In this work, we propose an efficient retrospective motion correction method adapted to this problem, with an application to inhomogeneous Magnetization Transfer (ihMT) imaging. The proposed method is confronted with the MCFLIRT (FSL) method used as reference, and residual motion is quantified and analysed across native and motion corrected images. |
| 3360 | MoCo Cycle-MedGAN: Unsupervised correction of rigid MR motion artifacts | |
| Tobias Hepp1, Karim Armanious1,2, Aastha Tanwar2, Sherif Abdulatif2, Thomas Küstner1, Bin Yang2, and Sergios Gatidis1 | ||
1University Hospital Tübingen, Tübingen, Germany, 2University of Stuttgart, Stuttgart, Germany |
||
Motion is one of the main sources for artifacts in magnetic resonance (MR) imaging and can affect the diagnostic quality of MR images significantly. Previously, supervised adversarial approaches have been suggested for the correction of MR motion artifacts. However,supervised approaches require paired and co-registered datasets for training, which are often hard or impossible to acquire. We introduced a new adversarial framework for the unsupervised correction of severe rigid motion artifacts in the brain region. Quantitative and qualitative comparisons with other supervised and unsupervised translation approaches showed the enhanced performance of the introduced framework. |
| 3361 | Modelling RF coil-sensitivity induced artefacts in prospective motion-corrected accelerated 3D-EPI fMRI | |
| Nadine N Graedel1, Yael Balbastre1, Nadège Corbin1, Oliver Josephs1, and Martina F Callaghan1 | ||
1Wellcome Centre for Human Neuroimaging, UCL Queen Square Institute of Neurology, University College London, London, United Kingdom |
||
The increased SNR offered by 3D EPI makes it well-suited for high-resolution fMRI but this benefit needs to be weighed against the increased motion sensitivity. Even when prospective motion correction is used to create a consistent excitation volume and image encoding, residual artefacts remain, caused for example by the apparent motion of coil sensitivities relative to the head after PMC. Using simulated fMRI data we examined the impact of this effect on fMRI and a potential correction using motion-adjusted GRAPPA calibration combined with volume-wise bias field removal which in simulations achieved improved sensitivity and specificity of detection of activation. |
| 3362 | The impact of through-plane motion on 2D FISP-MR Fingerprinting: a simulation study based on realistic patient and healthy volunteer movements | |
| Pedro Lima Cardoso1, Gregor Körzdörfer2, Eva Hečková1, Mathias Nittka2, Siegfried Trattnig1,3, and Wolfgang Bogner1,3 | ||
1High Field MR Centre, Department of Biomedical Imaging and Image-guided Therapy, Medical University of Vienna, Vienna, Austria, 2Siemens Healthcare GmbH, Erlangen, Germany, 3Christian Doppler Laboratory for Clinical Molecular MR Imaging, MOLIMA, Vienna, Austria |
||
Motion-induced artifacts in quantitative MRI may not be visually detectable. MRF has shown to be robust to in-plane movement, and methods to mitigate it were made available. However, the impact of clinically prevalent motion patterns has not yet been investigated. The effect of realistic motion data from patients with neurodegenerative disorders and young/elderly controls on 2D FISP-MRF is therefore investigated in this simulation study. Results are validated with in-vivo motion-tracked measurements. Although through-plane motions were shown to significantly impact MRF parametric maps (particularly T2), MRF may preserve its sensitivity in clinical cases in which expected lesion-related differences are large enough. |
| 3363 | Using NMR field probes to flag significant head movement at 7T | |
| Laura Bortolotti1 and Richard Bowtell1 | ||
1Physics, Sir Peter Mansfield Imaging Centre, University of Nottingham, Nottingham, United Kingdom |
||
A fixed array of NMR field probes has proved to be a valid tool for monitoring the effects of a wide range of head movements in a 7T scanner. Here, we use this set up to detect significant field changes in order to give early feedback on head motion. This information could be used in deciding to stop a scan or to reacquire all or part of the k-space data in order to avoid acquisition of motion corrupted image data. This method can be implemented without requiring image sequence modification nor rigid coupling of a motion marker to the head. |
| 3364 | A deep network for continuous motion detection during MRI scanning | |
| Isabelle Heukensfeldt Jansen1, Sangtae Ahn1, Rafi Brada2, Michael Rotman2, and Christopher J. Hardy1 | ||
1GE Global Research Center, Niskayuna, NY, United States, 2GE Global Research Center, Herzliya, Israel |
||
We introduce ISHMAPS, a method for detecting and adapting to patient motion in real time during an MR scan. The method uses a neural network trained on motion-corrupted data to detect and score motion using as little as 6% of k-space. Once motion is detected, multiple separate complex sub-images from different motion states can be reconstructed and combined into a motion-free image, or the scan can adaptively re-acquire sections of k-space taken before motion occurred. |
| 3365 | Optical prospective motion correction enables improved quantitative susceptibility mapping (QSM) of the brain at 7T | |
| Phillip DiGiacomo1, Mackenzie Carlson1, Julian Maclaren1, Murat Aksoy1, Yi Wang2, Pascal Spincemaille2, Brian Burns3, Roland Bammer1, Brian Rutt1, and Michael Zeineh1 | ||
1Stanford University, Stanford, CA, United States, 2Cornell University, Ithaca, NY, United States, 3GE Healthcare, San Francisco, CA, United States |
||
Recent literature has shown the potential of high‐resolution quantitative susceptibility mapping (QSM) with ultra‐high field (7T) MRI for investigating the magnetostatic properties of brain structures and disease pathology. Higher spatial resolutions, however, require longer scans resulting in a higher likelihood of subject movement. Here, we apply a novel prospective real-time optical motion tracking and correction system using a camera integrated between the Tx and Rx coils of a commercial 7T head coil to demonstrate the feasibility of acquiring high-resolution R2* and QSM images robust to subject motion. |
| 3366 | Towards Optimal Design of Orbital K-Space Navigators for 3D Rigid-Body Motion Estimation | |
| Thomas Ulrich1, Franz Patzig1, Bertram Jakob Wilm1, and Klaas Paul Pruessmann1 | ||
1Institute for Biomedical Engineering, ETH Zurich and University of Zurich, Zurich, Switzerland |
||
A single-shot k-space navigator trajectory and a corresponding motion estimation algorithm is proposed. They allow for 3D rigid-body motion estimation. Their performance in terms of accuracy and precision is studied and a cross-validation experiment is conducted to show that the method is suitable for in-vivo use. The accuracy and precision of the method depend on the orbital radius of the navigator. A simulation study is conducted to determine the best choice of the navigator radius. |
| 3367 | Real-time brain masking algorithm improves motion tracking accuracy in scans with volumetric navigators (vNavs) | |
| Malte Hoffmann1,2, Robert Frost1,2, David Salat1,2, M Dylan Tisdall3, Jonathan Polimeni1,2, and André van der Kouwe1,2 | ||
1Athinoula A. Martinos Center for Biomedical Imaging, Charlestown, MA, United States, 2Department of Radiology, Harvard Medical School, Boston, MA, United States, 3Radiology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, United States |
||
Volumetric navigators (vNavs) interleaved within longer MRI sequences are an effective method for dynamically detecting and correcting head motion during the acquisition. However, motion is estimated by registering each vNav back to the first, and bias can be introduced by non-rigid deformations of, e.g., the mouth, since the head is taken to be a rigid body. We demonstrate that this bias introduces correction-related artifacts. We present robust real-time brain extraction (0.02 s per vNav) and demonstrate offline that the bias and associated artifacts can be removed by using motion tracks from brain-masked registration. |
| 3368 | Sensitivities Constrained Phase Update (sCPU) for Ghost Artifacts Reduction | |
| Hai Luo1, Meining Chen1, Ziyue Wu2, Bei Lv1, Fei Peng1, Shijie Wang1, Wenkui Hou1, Weiqian Wang1, and Gaojie Zhu1 | ||
1AllTech Medical Systems, Chengdu, China, 2Marvel Stone Healthcare, Wuxi, China |
||
Compared with magnitude value, phase of MRI signal is more prone to be influenced by motion. A novel method, sensitivity constrained phase update (sCPU) was proposed for robust and efficient ghost artifacts reduction. Using coil sensitivities as constraints, a synthetic image can be generated in which the ghost is reduced due to phase cancelation. Phase error was first estimated from the raw image and the synthetic image, and then was used to update the phase of raw k-space. The results with simulated and in-vivo data show that the ghost artifacts can be efficiently reduced after several iterations. |
| 3369 | The geometric solution to banding mitigates motion artifact in bSSFP imaging | |
| Michael Nicholas Hoff1, Nathan M Cross2, Qing-San Xiang3, Daniel S Hippe2, Charles G Colip2, and Jalal B Andre2 | ||
1Radiology, University of Washington, Seattle, WA, United States, 2Radiology, University of Washington, SEATTLE, WA, United States, 3Radiology, University of British Columbia, Vancouver, BC, Canada |
||
Widespread use of phase-cycled bSSFP imaging is hampered by problematic sensitivity to artifacts caused by susceptibility and motion. The geometric solution (GS), which can eliminate susceptibility-related banding and signal modulation, has previously show relative insensitivity to motion. Here, GS-bSSFP is evaluated using a phantom and in humans, imaging the skull base. The GS mitigates motion artifact in both paradigms and is particularly resilient when one of its four phase cycles is corrupted by motion. |
| 3370 | Evaluation of Deep Learning Techniques for Motion Artifacts Removal | |
| Alessandro Sciarra1,2, Soumick Chatterjee2,3, Max Dünnwald1,4, Oliver Speck2,5,6,7, and Steffen Oeltze-Jafra1,5 | ||
1MedDigit, Department of Neurology, Medical Faculty, Otto von Guericke University, Magdeburg, Germany, 2BMMR, Biomedical Magnetic Resonance, Otto von Guericke University, Magdeburg, Germany, 3Data & Knowledge Engineering Group, Otto von Guericke University, Magdeburg, Germany, 4Faculty of Computer Science, Otto von Guericke University, Magdeburg, Germany, 5Center for Behavioral Brain Sciences (CBBS), Magdeburg, Germany, 6German Center for Neurodegenerative Disease, Magdeburg, Germany, 7Leibniz Institute for Neurobiology, Magdeburg, Germany |
||
Removing motion artifacts in MR images remains a challenging task. In this work, we employed 2 convolutional neural networks, a conditional generative adversarial network (c-GAN), also known as pix2pix, as well as a network based on the residual network (ResNet) architecture, to remove synthetic motion artifacts for phantom images and T1-w brain images. The corrected images were compared with the ground-truth ones in order to assess the performance of the chosen neural networks quantitatively and qualitatively. |
| 3371 | QUIQI – using a QUality Index for the analysis of Quantitative Imaging data | |
| Giulia Di Domenicantonio1, Nadège Corbin2, John Ashburner2, Martina Callaghan2, and Antoine Lutti1 | ||
1Department for Clinical Neuroscience, Lausanne University Hospital and University of Lausanne, Lausanne, Switzerland, 2Wellcome Centre for Human Neuroimaging, Queen Square Institute of Neurology, UCL, London, London, United Kingdom |
||
Image degradation due to head motion is ubiquitous in MRI, reduces sensitivity, hinders clinical diagnosis and increases the risk of spurious findings. The few existing objective measures of degradation are often used sub-optimally, to remove the most degraded datasets from analysis. Using a large dataset (N~1400) we show how to incorporate an validated index of degradation into the analysis of group studies. The benefit is demonstrated for the case of healthy age-related difference in brain relaxometry data using the SPM software. However, the proposed framework is flexible with broad potential, including the analysis of other metrics and body regions. |
| 3372 | Remotely controllable phantom rotation device for cross-calibration at 7T | |
| Hendrik Mattern1, Robert Odenbach2, Niklas Thoma3, Frank Godenschweger1, and Oliver Speck1,4,5,6 | ||
1Biomedical Magnetic Resonance, Otto-von-Guericke University, Magdeburg, Germany, 2Institute for Medical Engineering, Otto-von-Guericke University, Magdeburg, Germany, 3Department of Mechanical Engineering, Otto-von-Guericke University, Magdeburg, Germany, 4German Center for Neurodegenerative Disease, Magdeburg, Germany, 5Center for Behavioral Brain Sciences, Magdeburg, Germany, 6Leibniz Institute for Neurobiology, Magdeburg, Germany |
||
A low-cost rotation device was designed and built to enable remotely controllable phantom movements. With the device, motions are highly reproducible. For cross-calibrations of external tracking systems, it prevents table movement or the need to move the phantom by hand from inside the scanner during the calibration, reduces the overall calibration duration, and provides similar calibration performance compared to the freehand approach performed by an expert. The CAD model was made publically available. |
| 3373 | Relaxation-Time Selective Imaging Using the Inverse Z-transform | |
| Seong-Min Kim1, Phuong Anh Chu Dang1, Sehong Oh2,3, and Jang-Yeon Park1,4 | ||
1Department of Biomedical Engineering, Sungkyunkwan University, Suwon, Korea, Republic of, 2Division of Biomedical Engineering, Hankuk University of Foregn Studies, Yongin, Korea, Republic of, 3Imaing Institute, Cleveland Clinic Foundation, Cleveland, OH, United States, 4Center for Neuroscience Imaging Research, Suwon, Korea, Republic of |
||
There have been many attempts to separate mixed relaxation-time decays. Among them, some approaches do not assume the number of exponential decays a priori for the analysis of multiple-component decay signals. However, they are sensitive to ill conditions and have a poor resolving ability in terms of decay constants. In this study, a new method is proposed that can analyze multiple T2 decays with high resolution using the inverse Z-transform, which was demonstrated in simulation and in vivo human brain experiment. T2-selective images were also presented and used for myelin-water fraction mapping and deep brain-tissue segmentation. |
| 3374 | Noise Reduction and Ghosting Alleviation in ASL Perfusion Measurement using Multi-dimensional Integration (MDI) | |
| Yichen Hu1, Qing Wei2, Zhongyang Zhou2, Jun Xie2, and Yongquan Ye1 | ||
1UIH America, Inc., Houston, TX, United States, 2United Imaging Healthcare, Shanghai, China |
||
3D GRASE (GRAdient and Spin Echo) pulse sequence has been widely employed as the readout in arterial spin labeling (ASL) applications, given its efficient acquisition and relatively long lasting signal intensities. However, the inherent weakness of GRASE, such as vulnerability to motions, can induce ghosting artifacts to perfusion imaging and quantitative cerebral blood perfusion (CBF) maps. Herein, we propose applying the Multi-Dimensional Integration (MDI) algorithm to processing the perfusion and CBF maps, by which method noticeable alleviation of motion ghosting can be obtained, and the imaging noise is reduced as well. |
| 3375 | Highly-accelerated Compressed Sensing 4D Flow MRI for Quantification of Whole Heart Hemodynamics | |
| Daniel Gordon1, Allison Blake1, Liliana Ma1, Yoshihiro Tanaka1, Kelvin Chow1,2, Ning Jin3, Philip Greenland1, Rod Passman1, Daniel Kim1, and Michael Markl1 | ||
1Northwestern University, Chicago, IL, United States, 2Siemens Medical Solutions USA, Inc., Chicago, IL, United States, 3Siemens Medical Solutions USA, Inc., Cleveland, OH, United States |
||
The purpose of this study was to evaluate an efficient, free-breathing, whole-heart 4D flow CS protocol in a cohort of 21 healthy controls. The CS technique was evaluated for internal self-consistency by comparing net flow through various cardiac structures. In addition, cine imaging was performed and used to calculate left ventricular (LV) stroke volume (SV) for comparison to net flow in the ascending aorta. CS whole-heart 4D flow was acquired in 5:23 ± 0:51 minutes. Input and output flow, and LV stroke volume and ascending aorta net volume with 4D flow were significantly (p<0.05) correlated for all comparisons. |
| 3376 | Methods and challenges for body fat composition and hepatic fat fraction measurement in a multi-site study using MRI | |
| SHILPY CHOWDHURY1, SPENCER LOONG1, JOAN SABATE1, and SAMUEL BARNES1 | ||
1LOMA LINDA UNIVERSITY, LOMA LINDA, CA, United States |
||
Running a large longitudinal multi-center study to measure visceral fat and hepatic fat fraction present unique challenges. We scanned fat/water and ex vivo liver phantoms across all sites and over time to show consistency. SliceOmatic and ImageJ were used for quantification of visceral fat and LCModel for hepatic fat fraction. Visceral fat showed a good correlation with hepatic fat and demographics. The use of phantoms and careful design of the protocol can help address challenges of a longitudinal study and help maintain consistency across sites. |
| 3377 | Brain regional hemodynamic alterations in obstructive sleep apnea using dynamic susceptibility contrast MRI | |
| Ruyi Zhang1, Hea Ree Park2, Hosung Kim3, Gele Qin1, Eun Yeon Joo4, and Lirong Yan3 | ||
1Department of Electric Engineering, University of Southern California, Los Angeles, CA, United States, 2Inje University College of Medicine, GoyangSouth, Korea, Republic of, 3Stevens Neuroimaging and Informatics Institute, Keck School of Medicine, University of Southern California, Los Angeles, CA, United States, 4Sungkyunkwan University School of Medicine, Seoul, Korea, Republic of |
||
In this study, we comprehensively studied the regional hemodynamic measures including CBF, cerebral blood volume (CBV), and time-related hemodynamic parameters in OSA patients using Gadolinium contrast agent (Gd)-base dynamic susceptibility contrast MRI. |
| 3378 | Performance of Automatic Cerebral Arterial Segmentation of MRA Images Improves at Ultra-high Field. | |
| Alexander Saunders1,2, Tales Santini3, Tiago Martins3, Howard Aizenstein3, John C. Wood1, Matthew Borzage (co-corresponding author)1, and Tamer S. Ibrahim (co-corresponding author)3 | ||
1Children's Hospital Los Angeles, University of Southern California, Los Angeles, CA, United States, 2Rudi Schulte Research Institute, Santa Barbara, CA, United States, 3Swanson School of Engineering and School of Medicine, University of Pittsburgh, Pittsburgh, PA, United States |
||
Ultra-high field time of flight MRA can generate images with higher signal and resolution but image quality may suffer from increased field inhomogeneity. Because we would like to extract arteries for further analysis, we sought to evaluate automatic segmentation performance compared with standard field strength. Five segmentation algorithms were applied to two MRA images (one 3T, one 7T) and performance was measured against manually segmented ground truth data. We found that automatic segmentation performs better in 7T images but confounds in acquisition and image processing need to be further investigated. |
| 3379 | Hyperpolarized 13C imaging with multi-shot multi-echo echo planar imaging | |
| Kofi Deh1, Kristin Granlund1, Roozbeh Eskandari1, Arsen Mamakhanyan1, Nathaniel Kim1, and Kayvan Keshari,1 | ||
1Memorial Sloan Kettering Cancer Center, New York, NY, United States |
||
We demonstrate the use of a use of a broadband RF excitation with a multi-shot EPI readout for robust separation of metabolites in hyperpolarized 13C imaging. The approach places less demands on scanner hardware and compensates for local B0 inhomogeneity, making it possible to obtain high resolution time-resolved multi-slice imaging over a large region of interest which can be reformatted into 3D volumetric images to facilitate the study diseases such as cancer metastasis using HP probes. |
| 3380 | Validation of Small Motion Quantification by aMRI Using a Digital Phantom | |
| Zhiyue J Wang1,2 and Youngseob Seo3 | ||
1University of Texas Southwestern Medical Center, Dallas, TX, United States, 2Children's Health, Dallas, TX, United States, 3Korea Research Institute of Standards & Science, Daejeon, Republic of Korea |
||
In MRI, small signal intensity changes not perceivable by naked eyes are routinely analyzed to extract valuable information, such as in fMRI. However, a small movement was difficult to detect. Recently, image amplification methods have been introduced for visualizing small sub-pixel motions that are too small to be discerned by naked eyes, based on methods developed for video processing. We use computer simulations to explore the range of parameters for the technique to work optimally. |
| 3381 | Filter-pipeline based algorithm to find the AIF in DCE-MRI images for perfusion calculation of rectal cancer. | |
| Christian Tönnes1, Sonja Janssen2, Alena-Kathrin Schnurr1, Tanja Uhrig1, Khanlian Chung1, Lothar R. Schad1, and Frank Gerrit Zöllner1 | ||
1Computer Assisted Clinical Medicine, Mannheim Institute for Intelligent Systems in Medicine, Medical Faculty Mannheim, Heidelberg University, Mannheim, Germany, 2Institute of Clinical Radiology and Nuclear Medicine, Medical Faculty Mannheim, Heidelberg University, Mannheim, Germany |
||
Perfusion calculation is highly dependent on the expertise and input of the physician and therefore reproduction of the results is difficult. A filter pipeline based algorithm for selecting the AIF ROI was implemented and evaluated. Our results show that such an algorithm can be used to determine the AIF ROI for perfusion calculation in dynamic contrast enhanced MRI images with a high accuracy. A fully automatic and deterministic algorithm removes the need for manual interaction and standardizes clinical perfusion measurements while simultaneously reducing the time requirement for the ROI determination by 86%. |
| 3382 | Design of deep neural network in time-phase encoding plane for compressed sensing cardiovascular CINE MRI | |
| Chang-Beom Ahn1 and Seong-Jae Park1 | ||
1Kwangwoon University, Seoul, Republic of Korea |
||
We build a deep neural network in time-phase encoding plane (t-y) for compressed sensing cardiovascular CINE MRI. Previously neural networks were developed in cross-sectional image planes (x-y). A hierarchical convolutional neural network (CNN), known as U-net is used. By adopting the t-y plane, instead of the x-y plane, simultaneous restoration in time and space is effectively achieved. By computer simulation, the proposed deep neural network based on the t-y plane shows better cross-sectional images, clearer temporal profiles, and less normalized mean square errors compared to that of the x-y plane. |
| 3383 | Clinical Image Quality Evaluation For Field Strength and Contrast Independence of Deep Learning Reconstruction | |
| Erin J Kelly1, Hung P Do2, Dawn M Berkeley2, and Jonathan K Furuyama2 | ||
1Canon Medical Systems USA, Inc, Tustin, CA, United States, 2Canon Medical Systems USA, Inc., Tustin, CA, United States |
||
dDLR algorithms with two path CNN architecture that are designed to be noise adaptive are robust against differences in image contrast and field strength. This study shows clinical image quality improvement on 1.5T brain and knee datasets using an algorithm trained on 3T data. |
| 3384 | MRI Scanner Characterization with the ISMRM/NIST System Phantom | |
| Stephen E. Russek1, Michael A. Boss2, H. Cecil Charles3, Andrew M. Dienstfrey1, Jeffrey L. Evelhoch4, Jeffrey L. Gunter5, Derek L. G. Hill6, Edward F. Jackson7, Kathryn E. Keenan1, Guoying Liu8, Michele Martin1, Nikki S. Rentz1, Karl F. Stupic1, Chun Yuan9, and Zydrunas Gimbutas1 | ||
1National Institute of Standards and Technology, Boulder, CO, United States, 2American College of Radiology, Philadelphia, PA, United States, 3Duke Image Analysis Laboratory, Durham, NC, United States, 4Merck Research Laboratories, West Point, PA, United States, 5Mayo Clinic, Rochester, MN, United States, 6University College London, London, United Kingdom, 7University of Wisconsin, Madison, WI, United States, 8National Institute of Biomedical Imaging and Bioengineering, Bethesda, MD, United States, 9University of Washington, Seattle, WA, United States |
||
We describe basic scanner characterization and determination of MR-parameter measurement accuracy using the ISMRM/NIST system phantom. The phantom provides a convenient method to measure geometric distortion; the efficacy of non-linear gradient corrections; slice profiles and associated measurement uncertainties; protocol and system dependent resolution contributions; SNR according to NEMA protocols; accuracy of relaxation time; and proton density measurements. The phantom is unique in having SI-traceability, a high level of precision, long-term stability, and monitoring by a national metrology institute. |
| 3385 | Characterization and elimination of X-nuclear eddy current artifacts on clinical MR systems | |
| Mary A McLean1,2, Scott Hinks3, Joshua Kaggie1, Ramona Woitek1, Frank Riemer4, Martin Graves1, Dominick McIntyre2, Ferdia Gallagher1, and Rolf Schulte5 | ||
1Dept of Radiology, University of Cambridge, Cambridge, United Kingdom, 2Cancer Research UK Cambridge Institute, University of Cambridge, Cambridge, United Kingdom, 3GE Healthcare, Waukesha, WI, United States, 4MMIV, Dept of Radiology, Haukeland University Hospital, Bergen, Norway, 5GE Healthcare, Munich, Germany |
||
Using pulse-acquire spectroscopy, line-shape distortions characteristic of eddy currents were demonstrated for X-nuclei, which were not seen for 1H, on systems from multiple vendors. The severity of these appeared correlated with the amplitude of the f0 eddy current frequency compensation term applied by the system along the axis of the applied spoiler gradient. A proposed correction to eddy current compensation taking account of the X-nuclear gyromagnetic ratio was shown to dramatically reduce these distortions on a GE system. The same correction was also shown to improve the quality of non-Cartesian imaging (spirals and cones). |
| 3386 | Optimized gradient spoiling for B1 mapping with UTE AFI that is less sensitive to the RF-spoiling phase increment | |
| Marta Brigid Maggioni1, Martin Krämer1, and Jürgen R. Reichenbach1,2,3,4 | ||
1Medical Physics Group, Institute of Diagnostic and Interventional Radiology, Jena University Hospital - Friedrich Schiller University, Jena, Germany, 2Michael Stifel Center for Data-driven and Simulation Science Jena, Friedrich Schiller University, Jena, Germany, 3Abbe School of Photonics, Friedrich Schiller University, Jena, Germany, 4Center of Medical Optics and Photonics, Friedrich Schiller University, Jena, Germany |
||
B1 mapping is a major challenge for reliable T1 quantification, and it is especially difficult for UTE sequences, because most conventional B1 mapping methods are unable to perform on short T2* tissues. Recently, an implementation of the actual flip angle imaging (AFI) method was proposed for UTE sequences. The AFI method requires ideal spoiling of residual transverse magnetization, which is typically achieved by fine-tuning RF-spoiling phase increments as well as the strength of spoiler gradients. In this work, we propose an improved gradient spoiling scheme, which reduces the effect of the RF-spoiling phase increment on the AFI results. |
| 3387 | Trajectory correction in high-resolution gated golden-angle radial Dixon imaging using the gradient impulse response function | |
| Christoph Zöllner1, Sophia Kronthaler1, Stefan Ruschke1, Jürgen Rahmer2, Johannes M. Peeters3, Holger Eggers2, Peter Börnert2, Rickmer F. Braren1, and Dimitrios C. Karampinos1 | ||
1Department of Diagnostic and Interventional Radiology, Technical University of Munich, München, Germany, 2Philips Research Laboratory, Hamburg, Germany, 3Philips Healthcare, Best, Netherlands |
||
Stack-of-stars-type radial k-space trajectories employing golden-angle ordering have been becoming popular for either free breathing or navigator-gated volumetric T1-weighted imaging of the abdomen and heart. Most methods for compensating radial k-space trajectory errors induced by eddy currents and system delays are based either on the acquisition of calibration lines with opposite polarity or on the processing of approximately anti‐parallel spokes from the actual radial acquisition. This work shows that a trajectory correction based on a gradient system impulse response function improves image quality in high-resolution gated golden-angle radial Dixon imaging. |
| 3388 | Spiral Gradient Waveform Correction Using Multipoint Gradient Anchoring (MGA) | |
| Qi Liu1, Yuan Zheng1, Yu Ding1, Jian Xu1, and Weiguo Zhang1 | ||
1UIH America, Inc., Houston, TX, United States |
||
A new comprehensive spiral gradient waveform correction strategy is proposed that features multipoint gradient anchoring (MGA). By measuring multiple gradient delays in spiral waveforms at different rotation angles, it effectively ‘anchors’ the waveform at a series of locations and improves gradient waveform fidelity. Application of this innovative design on phantom and volunteer imaging indicates it is an effective and promising technique. |
| 3389 | Replaceable field probe holder for the Nova coil on a 7 Tesla Siemens scanner | |
| Sascha Brunheim1, Christian Mirkes2, Benjamin E Dietrich2, Jolanda M Schwarz1, Rüdiger Stirnberg1, Sebastian Ismar3, Alexander Christen3, and Tony Stöcker1,3 | ||
1MR Physics, German Center for Neurodegenerative Diseases (DZNE), Bonn, Germany, 2Skope Magnetic Resonance Technologies Inc., Zurich, Switzerland, 3Department of Physics and Astronomy, University of Bonn, Bonn, Germany |
||
An optimized holder for fast mounting of the Skope clip-on field camera into the common Nova Tx/Rx head coil on a 7 Tesla Siemens scanner is presented. Field probe positioning was based on transceiver coil architecture constraints and conditioning for best approximation of field dynamics. The conditioning results for the proposed probe holder positions show very good agreement with the corresponding simulation. They approximate the optimal spherical probe arrangement as well as possible. In comparison, the probe holder outperforms the manual mounting situation conditioning-wise, which corresponds to arbitrary placement onto the receive part of the head coil. |
| 3390 | Data-Driven non-Cartesian trajectory correction based on Cartesian reference data | |
| Felix Horger1, Mathias Nittka1, and Gregor Körzdörfer1 | ||
1Siemens Healthcare, Erlangen, Germany |
||
Trajectory deformations in non-Cartesian scans can lead to significant image artifacts, while in comparison the effect on Cartesian scans is evidentially minor. We propose a novel approach for trajectory correction relying on the minimization of a suitable measure between a non-Cartesian image and a Cartesian reference image, given a trajectory deformation model. Our proof of concept is applied on simulated as well as measured data. The approach is fast, accurate and easily extendable. Given the right use-case, no extra scan time is required (e.g. mixed spiral-Cartesian MR Fingerprinting). |
| 3391 | The impact of gradient non-linearity on Maxwell compensation when using asymmetric gradient waveforms for tensor-valued diffusion encoding | |
| Filip Szczepankiewicz1,2,3, Cornelius Eichner4, Alfred Anwander4, Carl-Fredrik Westin1,2, and Michael Paquette4 | ||
1Harvard Medical School, Boston, MA, United States, 2Radiology, Brigham and Women's Hospital, Boston, MA, United States, 3Clinical Sciences Lund, Lund University, Lund, Sweden, 4Neuropsychology, Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig, Germany |
||
Gradient non-linearity distorts the shape of the gradient waveform used for diffuison encoding. This distortion also compromises Maxwell compensation in asymmetric gradient waveforms. We show that one of two strategies for Maxwell compensation fails under non-linear gradiets (K-nulling) whereas M-nulling is immune to this effect. |
| 3392 | Formulating EPI ghost correction as a convolution: Insights into Dual-Polarity GRAPPA by examining kernels of limited extent | |
| W Scott Hoge1,2,3 and Jonathan R Polimeni2,4,5 | ||
1Radiology, Brigham and Women's Hospital, Boston, MA, United States, 2Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Charlestown, MA, United States, 3Harvard Medical School, Boston, MA, United States, 4Department of Radiology, Harvard Medical School, Boston, MA, United States, 5Massachusetts Institute of Technology, Cambridge, MA, United States |
||
We examine the relationship between traditional 1D linear phase-error correction methods and Dual-Polarity GRAPPA (DPG) for EPI ghost correction. We first show that conventional ghost correction methods based on 1D linear phase errors can be implemented with convolution kernels in k-space, and then generalized in several steps to replicate a typical DPG kernel. We identify that employing multiple phase-encoded lines is important for ghost correction kernel calibration, while incorporating data from multiple channels is critical. Surprisingly, having a kernel extent along the phase-encoded direction is less critical, demonstrating the utility of DPG kernels with limited extent for ghost correction. |
| 3393 | Correcting Gradient Delays in Multi-Echo Rosette Trajectories with RING | |
| Volkert Roeloffs1, Adam Michael Bush2, Suma Anand1, and Michael Lustig1 | ||
1EECS, UC Berkeley, Berkeley, CA, United States, 2Radiology, Stanford, Stanford, CA, United States |
||
Rosette trajectories are sensitive to gradient delays as caused by eddy currents. Here, we use the RING method recently proposed for radial trajectories to correct for these delays. To this end, we approximate the center portion of the Rosette trajectory to be radial-like, determine delay constants, and correct the entire trajectory in a second step. First phantom and patient studies show that this simple correction method improves image quality in multi-echo Rosette imaging and leads to prolonged T2* values in derived maps. The small amount of required data renders echo-adaptive corrections feasible. |
| 3394 | Comparison of three B1 correction methods for RARE imaging with transceive surface RF coils | |
| Paula Ramos Delgado1, Andre Kuehne2, Jason M. Millward1, Joao Periquito1, Thoralf Niendorf1,3, Sonia Waiczies1, and Andreas Pohlmann1 | ||
1Berlin Ultrahigh Field Facility (B.U.F.F), Max Delbrück Center for Molecular Medicine in the Helmholtz Association, Berlin, Germany, 2MRI.tools GmbH, Berlin, Germany, 3Experimental and Clinical Research Center, a joint cooperation between the Charité Medical Faculty and the Max Delbrück Center for Molecular Medicine in the Helmholtz Association, Berlin, Germany |
||
Fluorine MR methods support quantification but have inherently low SNR. To enhance sensitivity, SNR-efficient imaging techniques such as RARE and cryogenically-cooled surface RF coils are used. However, transceive surface RF coils show variation in the excitation field, impairing quantification and T1 contrast. We previously showed improved homogeneity using a novel B1 correction method developed for 1H imaging that uses experimental data acquired with a volume resonator. Here, we compared it to a sensitivity-only correction and a combination of both, to evaluate which approach yields the most accurate contrast and concentration quantification. This work is applicable to different nuclei. |
| 3395 | Feasibility of Monitoring Geometric Distortion for Radiation Therapy Planning using ACR Weekly QA | |
| Chen Lin1, Stephanie Tensfeldt1, and Christopher Serago2 | ||
1Radiology, Mayo Clinic, Jacksonville, FL, United States, 2Radiation Oncology, Mayo Clinic, Jacksonville, FL, United States |
||
To demonstrate the distortion measurements based on automatic analysis of American College of Radiology (ACR) weekly quality assurance (QA) scans is adequate for monitoring the geometric distortion of a MR scanner for radiation therapy planning by comparing with distortion measurements using a 3D grid phantom. |
| 3396 | Noise Analysis for Simultaneous Transmission and Reception Enabled MRI Scanner | |
| Bilal Tasdelen1,2, Alireza Sadeghi-Tarakameh1,2, Ugur Yilmaz2, and Ergin Atalar1,2 | ||
1Department of Electrical and Electronics Engineering, Bilkent University, Ankara, Turkey, 2National Magnetic Resonance Research Center (UMRAM), Ankara, Turkey |
||
In simultaneous transmission and reception (STAR), injection of noise and spurs stemming from transmit chain is a problem as well as the leak power. Investigation of noise and spurs are important in order to determine minimum necessary isolation to achieve similar noise levels compared to conventional imaging. Noise sources during STAR are investigated and compared with experiments. It has been shown that using an active cancellation method where dominant noise source is transmit chain, with combination of passive cancellation can reduce noise and spurs. Necessary isolation is shown to depend on input noise and total gain in transmit chain. |
| 3397 | ASCII: Acquisition with Self-Correction for Image Inhomogeneity at 7T | |
| Edward M Green1,2, James C Korte1,3, Bahman Tahayori1,4,5, Yasmin Blunck1,2, and Leigh A Johnston1,2 | ||
1Department of Biomedical Engineering, University of Melbourne, Parkville, Australia, 2Melbourne Brain Centre Imaging Unit, University of Melbourne, Parkville, Australia, 3Department of Physical Sciences, Peter MacCallum Cancer Centre, Parkville, Australia, 4Department of Medical Physics and Biomedical Engineering, Shiraz University of Medical Sciences, Shiraz, Iran (Islamic Republic of), 5Center for Neuromodulation and Pain, Shiraz, Iran (Islamic Republic of) |
||
Inhomogeneity of the RF transmit field results in undesirable shading and brightening in high field imaging due to range of flip angles across the imaged object. A new method is proposed whereby two image volumes with coarse resolution in the slice direction are acquired at different flip angles and combined to produce an image volume at higher resolution and free of B1 inhomogeneity. Reconstruction quality is comparable to conventional super-resolution imaging, while image inhomogeneity caused by flip angle error is reduced. This is a flexible method applicable to a range of MRI sequences to correct B1 inhomogeneity. |
| 3398 | Phantom Development for the Standardization of fMRI Data across Multiple Imaging Sites and Scanners | |
| Daisuke Kokuryo1, Chika Sato2, Takashi Itahashi3, Shigeyoshi Saito4, Hiroyuki Ueda1, Etsuko Kumamoto1,5, Toshiya Kaihara1, Nobutada Fujii1, Noriaki Yahata2,6, and Ichio Aoki2,6 | ||
1Graduate School of System Informatics, Kobe University, Kobe, Japan, 2National Institute of Radiological Sciences, National Institutes for Quantum and Radiological Science and Technology, Chiba, Japan, 3Medical Institute of Developmental Disabilities Research, Showa University, Tokyo, Japan, 4Graduate School of Medicine, Osaka University, Suita, Japan, 5Information Science and Technology Center, Kobe University, Kobe, Japan, 6Institute for Quantum Life Science, National Institutes for Quantum and Radiological Science and Technology, Chiba, Japan |
||
Establishing data harmonization in fMRI “big data” is necessary to ensure the reliability and reproducibility of fMRI research. Our group recently developed two distinct phantoms: a brain-shaped phantom to correct field heterogeneity and a functional image-specific phantom to calibrate shape distortion and signal irregularity. A preliminary experiment confirmed a shape distortion and signal irregularity caused by field inhomogeneity when using EPI signals, and these were corrected using an affine transformation. Therefore, our developed phantoms have the potential to contribute to achieving fMRI data standardization. |
| 3399 | A semi-autofocus method to improve two-point Dixon spiral imaging | |
| Tzu-Cheng Chao1, Dinghui Wang1, Guruprasad Krishnamoorthy2, and James G. Pipe1 | ||
1Department of Radiology, Mayo Clinic, Rochester, MN, United States, 2Philips Healthcare, Gainesville, FL, United States |
||
Deblurring with water/fat separation plays a critical role in spiral imaging. The reconstruction quality depends on good off-resonant frequency estimation. A field map before each spiral imaging exam is often acquired to approximate the B0 of the scanned subject. However, the estimation errors occur due to nonconcurrent collection of the imaging data. The proposed work introduces an iterative method to update the field map based on the acquired images to improve spiral image reconstruction. The results have shown that the blurring can be reduced along with improved fat/water separation after updating the reference B0 used for reconstruction. |
| 3400 | Off-Resonance Self-Correction for Single-Shot Imaging | |
| Franz Patzig1, Bertram Wilm1, Simon Gross1, David Brunner1, and Klaas Paul Pruessmann1 | ||
1Institute for Biomedical Engineering, ETH Zurich and University of Zurich, Zurich, Switzerland |
||
We present a novel algorithm capable of determining a B0 map and an off-resonance corrected reconstructed image from a single-shot acquisition. The functionality of the algorithm was shown in-vivo for multiple imaging geometries using regular undersampled single-shot EPI and Spiral readouts showing stable convergence. In addition, the method is tested for the common situation of inter-scan subject movement, in which pre-acquired field maps would be rendered invalid. |
| 3401 | Optimal Coil Combination with Relatively Aligned Phase (RAP) for Phase-Cycled bSSFP Imaging | |
| Qing-San Xiang1 and Michael Nicholas Hoff2 | ||
1Radiology, University of British Columbia, Vancouver, BC, Canada, 2Radiology, University of Washington, Seattle, WA, United States |
||
Phase-Cycled (PC) bSSFP imaging is useful for qualitative and quantitative studies of tissue. But the images generated should be properly combined when they are acquired with multiple receiver coils; in particular, phase information in the complex images must be preserved. We describe a straightforward yet optimal coil combination algorithm for PC-bSSFP applications, where the phase is preserved through relative phase alignment along the measurement dimension. The method is demonstrated with experimental results from phantom and volunteer scans. |
| 3402 | Model-Based Iterative Reconstruction for Echo Planar Imaging: A Preliminary Evaluation for Neuro MRI | |
| Uten Yarach1,2, Matthew Bernstein1, John Huston III1, Norbert Campeau1, Petrice Cogswell1, Daehun Kang1, Myung-Ho In1, Yunhong Shu1, Nolan Meyer1, Erin Gray1, and Joshua Trzasko1 | ||
1Radiology, Mayo Clinic, Rochester, MN, United States, 2Department of Radiologic Technology, Faculty of Associated Medical Sciences, Chiang Mai University, Chiang Mai, Thailand |
||
A recently proposed model-based iterative reconstruction (MBIR) method enables to prospectively manage B0-inhomogeneity and gradient-nonlinearity in EPI, and has been demonstrated to generate images with high SNR and minimal spatial blurring and geometric distortion. However, the clinical significance and degree of benefit remains undetermined. In this study, using commodity EPI acquisition protocols for whole-brain imaging, MBIR was radiologically compared against standard scanner-generated EPI images and post-processed versions. The results show that significant advantage of MBIR (p<0.05) over both the standard scanner-generated EPI results and post-processed variants was observed in three categories: SNR, geometric accuracy, and overall image quality. |
| 3403 | Model-Based Nyquist Ghost Correction for Reversed Readout Echo Planar Imaging | |
| Uten Yarach1,2, Matthew Bernstein1, John Huston III1, Myung-Ho In1, Yi Sui1, Daehun Kang1, Yunhong Shu1, Erin Gray1, Nolan Meyer1, and Joshua Trzasko1 | ||
1Radiology, Mayo Clinic, Rochester, MN, United States, 2Department of Radiologic Technology, Faculty of Associated Medical Sciences, Chiang Mai University, Chiang Mai, Thailand |
||
Single-polarity reference scan-based paradigms are routinely used to combat Nyquist ghosting artifacts, they often fail to the fully suppress them because of statistical biases in estimated correction coefficients that result from noise and off-resonance effects. Prior work has shown that use of dual-polarity reference scans can mitigate the latter. In this work, we propose that this concept can be generalized to enable robust Nyquist ghost correction (NGC) directly from two reversed-readout EPI acquisitions – without explicit reference scanning. In-vivo results show the similar trends as in phantom examples, with the proposed-NGC mitigating aliasing artifacts representing up to 20% intensity errors. |
| 3404 | Reverse-Encoding MRI Near Metallic Implants Using Missing Pulse Steady-State Free Precession | |
| Michael Mullen1 and Michael Garwood1 | ||
1Center for Magnetic Resonance Research, Department of Radiology, University of Minnesota, Minneapolis, MN, United States |
||
Clinical MRI sequences for imaging near metallic implants are mainly multi-spectral approaches, with fast spin-echo acquisitions to achieve clinically relevant scan times. Due to the large bandwidths necessary for refocusing all off-resonance spins near metallic implants, usually the refocusing pulse flip angles must be limited to a sub-optimal value to permit a safe specific absorption rate. Here, a low peak amplitude, radiofrequency refocused sequence is used to limit energy deposition. Reversed encoding is employed, where two acquisitions are collected with opposite frequency-encoding gradient polarity. An estimate of the displacement field is found using a B-spline basis and the magnitude images. |
| 3405 | Mitigating artifacts from eye-motion in interleaved multi-shot DWI | |
| Naveen Bajaj1, Jaladhar Neelavalli1, Rupesh VK1, Karthik Gopalakrishnan1, and Marc Van Cauteren2 | ||
1Philips Healthcare, Bangalore, India, 2Philips Healthcare, Tokyo, Japan |
||
A novel approach for mitigating artifacts from localized motion of the eye, during interleaved multi-shot diffusion weighted imaging, is presented. A 2D cylindrical spatial saturation pulse is used to suppress signal from the orbit region, thereby significantly suppressing the ghost artifacts from eye movement in low b-value images. The saturation being localized to the orbit region does not influence the surrounding tissues. Placement of such a cylindrical pulse on the orbit region is relatively straight forward. In addition, the total scan time is not significantly affected by the cylindrical pulse saturation. |
| 3406 | Mitigation of Blurring due to View-Angle Tilting in Multispectral Imaging | |
| Alexander R Toews1,2, Philip K Lee1,2, Daehyun Yoon1, and Brian A Hargreaves1,2,3 | ||
1Radiology, Stanford University, Stanford, CA, United States, 2Electrical Engineering, Stanford University, Stanford, CA, United States, 3Bioengineering, Stanford University, Stanford, CA, United States |
||
Imaging near metal is challenging due to severe metal-induced field inhomogeneities. Multispectral imaging sequences employ view-angle tilting (VAT) to correct in-plane distortions due to nearby metal at the expense of blurring the entire image. We exploit the additional phase-encoding dimension acquired in multispectral imaging in order to reduce the VAT-induced blur. The method primarily consists of regridding k-space measurements to avoid the VAT-induced shear. Results on a grid phantom demonstrate the ability of this method to reduce readout blur. This work highlights an important limitation of VAT and demonstrates strategies for mitigating this effect. |
| 3407 | BMART-Enabled Field-Map Combination of Phase-Cycled Projection-Reconstruction Cardiac Cine for Banding-Free Balanced SSFP | |
| Anjali Datta1, Dwight Nishimura1, and Corey Baron2 | ||
1Electrical Engineering, Stanford University, Stanford, CA, United States, 2Medical Biophysics, Western University, London, ON, Canada |
||
For banding-free cardiac cine bSSFP imaging, we present a non-Cartesian accelerated frequency-modulated sequence that acquires three phase-cycles within a short breath-hold. A projection-reconstruction trajectory is used and data is also acquired on the rewinds, enabling generation of a B0 field map using BMART. This facilitates field-map combination of the phase-cycle images, which results in more homogeneous blood pool and reduced signal contributions from out-of-slice spins than root-sum-of-squares. |
| 3408 | Interleaved reversed polarity MPRAGE for in vivo B0 readout distortion correction | |
| Robert Frost1,2, Divya Varadarajan1,2, Jesper Andersson3, Jonathan R Polimeni1,2, Bruce Fischl1,2,4, and André J. W. van der Kouwe1,2 | ||
1Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Charlestown, MA, United States, 2Department of Radiology, Harvard Medical School, Boston, MA, United States, 3Wellcome Centre for Integrative Neuroimaging, University of Oxford, Oxford, United Kingdom, 4Computer Science and Artificial Intelligence Laboratory, Massachusetts Institute of Technology, Cambridge, MA, United States |
||
Readout distortion can affect structural imaging with MPRAGE particularly in temporal regions. Interleaved readout polarity MPRAGE allows for estimation of distortion without motion confound between oppositely distorted images. This approach could be useful for single echo MPRAGE imaging with ~200 Hz/px. |
| 3409 | Fast library-driven approach for the two-step iterative calculation of the voxel spread function | |
| Jie Wen1, Dmitriy A Yablonskiy2, Alexander L Sukstanskii2, Feiyan Zeng1, and Ying Liu1 | ||
1Radiology, The First Affiliated Hospital of USTC, Hefei, China, 2Radiology, Washington University in St. Louis, St. Louis, MO, United States |
||
Macroscopic B0 magnetic field inhomogeneity adversely affects Gradient-Recalled-Echo (GRE) MRI images. A previously introduced voxel spread function (VSF) method has been proposed to correct for these adverse effects if data are acquired using multi-GRE (mGRE) sequences. The method accounts for both, through-slice and in-slice field inhomogeneities and accounts for between-voxel signal propagation effects. The direct voxel-wise VSF calculation requires long computation time. We propose a library-driven approach for the VSF calculation that significantly reduces computation time, thus providing a platform for the applications of mGRE-based quantitative techniques in clinical practices. |
| 3410 | Improving 7T Cervical Spine (C-spine) MRI using the Uniform Combined Reconstruction (UNICORN) Algorithm | |
| Venkata Veerendranadh Chebrolu1,2, Stefan Sommer3,4, Constantin von Deuster3,4, and Julien Galley5 | ||
1Siemens Medical Solutions USA, Inc., Rochester, MN, United States, 2Department of Radiology, Mayo Clinic, Rochester, MN, United States, 3Siemens Healthcare AG, Zurich, Switzerland, 4SCMI, Swiss Center for Musculoskeletal Imaging, Zurich, Switzerland, 5Department of Radiology, Balgrist University Hospital, University of Zurich, Zurich, Switzerland |
||
Ultra-high field (UHF) 7T MRI provides high spatial resolution with significantly better signal-to-noise ratio compared to 3T and C-spine imaging at 7T is being pursued at a few research sites across the globe. MRI of the C-spine at 7T is challenging due to multiple factors. Recently, an algorithm ‘Uniform Combined Reconstruction’ (UNICORN) was used to correct receive-coil intensity inhomogeneity and improve the uniformity and overall image quality of 7T knee MRI. In this work, we present preliminary results from the application of UNICORN to C-spine imaging and demonstrate the utility of UNICORN in improving uniformity of C-spine MRI at 7T. |
| 3411 | Imaging the Internal Acoustic Meatus (IAMs) of patients with cochlear implants in-situ using Slice Encoded Metal Artefact Reduction. | |
| Simon Shah1, Fran Padormo 1, Kristine Knott1, Suki Thomson1, Steve Conner1, Geoff Charles-Edwards1, and Phil Touska1 | ||
1Guy's and St Thomas' NHS Foundation Trust, London, United Kingdom |
||
Neurofibromatosis type 2 (NF2) patients commonly have bilateral vestibular schwannomas, which typically require regular monitoring with MRI. However these patients also have cochlear implants which makes MRI challenging. We have used Slice Encoded Metal Artefact Correction (SEMAC) to improve the imaging of IAMs in patients with cochlear implants in-situ. The implementation of SEMAC imaging reduced the geometric distortion and the volume of the signal void around the implants by over half. |
| 3412 | MRI Near Metal Using Multi-Spectral 3D Variable Flip-Angle TSE | |
| Sina Tafti1, John P Mugler III2, Kevin Vu1, William J Garrison3, and G Wilson Miller2 | ||
1Department of Physics, University of Virginia, Charlottesville, VA, United States, 2Department of Radiology and Medical Imaging, University of Virginia, Charlottesville, VA, United States, 3Department of Biomedical Engineering, University of Virginia, Charlottesville, VA, United States |
||
Multi-spectral imaging techniques such as MAVRIC, SEMAC, and MAVRIC SL based on 3D TSE sequence significantly reduce metal artifacts. MAVRIC and MAVRIC-SL techniques traditionally employ Gaussian RF pulses to achieve an optimal sum of squares composite image while SEMAC uses windowed-sinc RF pulses. In this work, we implement multi-spectral acquisition into the variable flip-angle TSE with hard RF pulses which yields a desirable sum of squares composite image while achieving high turbo-factor and minimal echo spacing in imaging near metallic screws in phantom and an ex vivo lamb leg. |
| 3413 | Local B1-Enhanced 3D Spin-Echo Imaging for Metal Artifact Reduction | |
| Kilian Stumpf1, Andreas Horneff1, Jan-Bernd Hövener2, and Volker Rasche1 | ||
1Department of Internal Medicine II, University Medical Center Ulm, Ulm, Germany, 2Department of Radiology and Neuroradiology, University Medical Center Kiel, Kiel, Germany |
||
Artifacts due to the presence of metallic dental materials often limit the application of MRI. These materials, such as dental implants or fillings, often cause substantial artifacts e.g. in the oral cavity impairing diagnostic accuracy. In this contribution, we present an approach for conducting spin-echo based sequences with significantly reduced flip angles and high excitation bandwidths of up to 17 kHz by using an inductively coupled local coil, which in combination with single-point methods enables almost completely artifact-free local imaging of e.g. dental titanium implants. |
| 3414 | Unsupervised Deep Learning Method for EPI Distortion Correction using Dual-Polarity Phase-Encoding Gradients | |
| Jee Won Kim1, Kinam Kwon2, Byungjai Kim1, Sunho Kim1, and Hyunwook Park1 | ||
1Korea Advanced Institute of Science and Technology (KAIST), Daejeon, Korea, Republic of, 2Samsung Advanced Institute of Technology (SAIT), Suwon, Korea, Republic of |
||
We propose a new scheme for EPI distortion correction, which implements unsupervised learning, trained with readily available images, such as ImageNet2012 dataset. The distortion-corrected image is obtained by the MR image generation function using the input distorted images and the frequency-shift maps that are the outputs of the network. Two distorted images obtained with dual-polarity phase-encoding gradients are the inputs of the neural network. The neural network estimates the frequency-shift maps from the distorted images. To train the neural network, unsupervised learning was conducted by minimizing the L1 loss between input distorted images and the estimated distorted images. |
| 3415 | Compensating Diffusion Bias of Quantitative T2 mapping on High-Field MRI Scanners | |
| Natalie Bnaiahu1, Ella Wilczynski1, Shir Levy2, Noam Omer1, Tamar Blumenfeld-Katzir1, and Noam Ben-Eliezer1,3,4 | ||
1Biomedical Engineering, Tel Aviv University, Tel Aviv, Israel, 2School of Chemistry, Tel Aviv University, Tel Aviv, Israel, 3Sagol school of Neuroscience, Tel Aviv University, Tel Aviv, Israel, 4Center for Advanced Imaging Innovation and Research, New York, NY, United States |
||
On high-field scanners, strong imaging gradients induce spurious diffusion signal decay, which distorts qT2 calculation. By modeling the pulse sequence gradients and coherence pathways, an effective b value was calculated to assess signal attenuation caused by diffusion. ADC values were measured, extensive phantom scans and in-vivo experiments were performed using the MSME protocol with varied voxel sizes to examine the effect and the solution. Before correction T2 values were short with high variability across different voxel sizes, after correction T2 values increased and became more accurate and consistent. Tissue properties also influenced the correction as their ADC varies. |
| 3416 | Reduction of Static Off-Resonance Artifacts in T2*-Weighted Gradient-Echo MRI | |
| Jakob Meineke1 and Tim Nielsen1 | ||
1Philips Research, Hamburg, Germany |
||
A reconstruction method is presented to reduce off-resonance induced artifacts in gradient-echo MRI. The method reconstructs the magnetization without the off-resonance induced phase and in this way reduces the effects of intra-voxel dephasing. The method is applied to simulations and volunteer scans, showing reduced signal voids and improved homogeneity in R2*-maps. |
| 3417 | Diffusion spoiling for fast steady-state gradient echo imaging of the human brain using a 300 mT/m gradient system | |
| Kerrin J Pine1, Luke J Edwards1, and Nikolaus Weiskopf1 | ||
1Department of Neurophysics, Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig, Germany |
||
In gradient echo imaging sequences, transverse coherences persisting across repetitions must be managed to avoid large variations in the steady state signal. Ultra-high amplitudes for gradient spoiling can be effective, however, conventional human MR scanners do not achieve the necessary gradient amplitudes. By using a 3T Connectom MRI scanner with a 300 mT/m gradient amplitude, we demonstrate that by operating in the diffusion spoiling regime, it is now possible to eliminate the contribution of unwanted transverse magnetization to the steady state by gradient spoiling alone. It can serve as a reference standard for validating spoiling schemes and for R1 quantification. |
| 3418 | 3DANCE VDS: Three-points Dixon golden Angle with off-resoNance CorrEction Variable Density Spiral | |
| Marina Manso Jimeno1,2, Sairam Geethanath1,2, and John Thomas Vaughan Jr.1,2 | ||
1Columbia University, New York, NY, United States, 2Columbia Magnetic Resonance Research Center (CMRRC), New York, NY, United States |
||
3DANCE is a sequence designed to be a good candidate for a whole-body sequence for multiple applicaitons. Phantom and in vivo experiments demonstrated a good spatial resolution and consistent ability to separate water and fat tissues. Off-resonance correction of the water and fat images have showed good results in vitro but needs more testing for in vivo datasets. Future steps include improving the robustness of the methods and testing for motion tolerance. |
| 3419 | Two-Dimensional Coil-signature-based Phase Cycled Reconstruction for Inherent Correction of Echo-Planar Imaging Nyquist Ghost Artifacts | |
| Silu Han1, Mahesh Bharath Keerthivasan2, Chidi Patrick Ugonna1, and Nan-kuei Chen1 | ||
1Biomedical Engineering Department, The University of Arizona, Tucson, AZ, United States, 2Medical Imaging Department, The University of Arizona, Tucson, AZ, United States |
||
A coil-signature-based phase cycled correction method has been developed for Nyquist artifact removal in echo planar imaging (EPI). In our study, we investigated a novel coil-signature-based Nyquist artifact correction method that can be applied to EPI integrating various acceleration schemes including through-plane Multi-band Imaging (MB) and in-plane parallel SENSitivity Encoding Imaging (SENSE). Our method uses a coil-signature-based phase-cycled reconstruction without requiring EPI reference scans, and can inherently correct 2D phase errors. Our results show that the developed method can effectively reduce EPI Nyquist artifacts under different acceleration approaches. |
| 3420 | Scrambled readout polarities in 3D-encoded EPI markedly reduce the coherence of N/2 ghosts | |
| M. Dylan Tisdall1 | ||
1Radiology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, United States |
||
N/2 ghosts in EPI result from consistent discrepancies between readouts with positive and negative polarities. We present 3D EPI with scrambled readout polarities as a strategy to reduce the coherence of N/2 ghosts in 4D x-f space and reduce their impact on quantitative analyses (e.g., motion tracking or resting state fMRI). We demonstrate that the method can produce either individual volumes with N/2 ghost energy spread out in the partition direction, or complete 4D x-f data with ghosts spread in both partition and frequency directions. |
| 3421 | Susceptibility effects of uHDC material upon static field homogeneity | |
| Christopher T Sica1, Navid Pourramzan Gandji2, and Qing X Yang2 | ||
1Radiology, PennState University College of Medicine, Hershey, PA, United States, 2Neurosurgery, PennState University College of Medicine, Hershey, PA, United States |
||
We performed a study to assess the effects of ultra-high dielectric constant material (uHDC) on the static field homogeneity at 3T. We characterized these effects on both a spherical and body phantom with 3D off-resonance field mapping and balanced SSFP imaging. The static field experienced shifts of up to 100 Hz in the presence of uHDC monolithic blocks. Shimming was able to partially compensate for these shifts. Future studies incorporating these materials should take their susceptibility effects into account. |
| 3422 | Quantitative Assessment of Metal Artefact Reduction Sequence Used In Imaging in Patients with Orthopaedic Total Hip Replacement Implants | |
| Simon Shah1, Danoob Dalili2, Jan Fritz 3, Geoff Charles-Edwards1, and Amanda Isaac1 | ||
1Guy's and St Thomas' NHS Foundation Trust, London, United Kingdom, 2King's College London, London, United Kingdom, 3Johns Hopkins University, Baltimore, MD, United States |
||
Musculoskeletal (MSK) imaging of patients with hip implants can be challenging due to metal artefacts. We have quantitatively assessed a number of metal artefact reduction sequences (MARS) on commonly implanted hip implants. The MARS methods used were: high receiver bandwidth, high transmit bandwidth, view angle titling (VAT) and slice encoded metal artefact correction (SEMAC). Signal pile-up and voids were most reduced across implants when SEMAC was employed and also artefacts at the bone metal interface were reduced. |
| 3423 | Improved dual inversion recovery T2-weighted black-blood image on myocardial MRI using zoomed technique | |
| Kohei Yuda1, Takashige Yoshida1, Masami Yoneyama2, Yuki Furukawa1, and Nobuo kawauchi1 | ||
1Tokyo Metropolitan Police Hospital, Nakano, Japan, 2Philips healthcare, Tokyo, Japan, minato, Japan |
||
The conventional Dual Inversion Recovery T2-weighted Black Blood (c-DIR-BB) on myocardial MRI is useful for the clinical diagnosis. One of the problems of c-DIR-BB is aliasing artifact, resulting in degradation of image quality. The zoomed method for improved aliasing artifact, orthogonal selection (i.e. phase / slice direction) excites to suppress tissue signals outside with a smaller FOV in phase encoding direction. Hence the DIR-BB using zoomed methods was improved image quality by orthogonal select excitation RF pulse. |
| 3424 | Fast, Automatic MRI Quality Assurance: Validation and Comparison with Manual Analysis using the ACR phantom | |
| Hamzeh Ahmad Mohammad Al Masri1,2,3, Tonima Ali1, Katie McMahon4, and Markus Barth1,3,5 | ||
1Centre for Advanced Imaging, The University of Queensland, Brisbane, Australia, 2Medical Imaging Department, The Hashemite University, Al-Zarqa, Jordan, 3ARC Training Centre for Innovation in Biomedical Imaging Technology, The University of Queensland, Brisbane, Australia, 4Royal Brisbane & Women's Hospital, Queensland University of Technology, Brisbane, Australia, 5School of Information Technology and Electrical Engineering, The University of Queensland, Brisbane, Australia |
||
Quality assurance (QA) is mandatory to ensure the stable performance of MR scanners over time. Automated analysis of QA tests can be useful to increase operator efficiency and overcome the manual processing issues such as time constraints and human bias. In this paper, we compare the manual and automated analysis approaches of the QA image datasets that have been collected from 3T MRI scanners using the American College of Radiology (ACR) accreditation phantom. We found that the automated method can significantly reduce the QA analysis time and the results of both methods were in agreement with each other. |
| 3425 | Estimation of B1 Map from a Single MR Image Using a Self-Attention Deep Neural Network | |
| Yan Wu1, Yajun Ma2, Jiang Du2, and Lei Xing1 | ||
1Radiation Oncology, Stanford University, Stanford, CA, United States, 2Radiology, University of California San Diego, La Jolla, CA, United States |
||
Inhomogeneity of the radiofrequency field (B1) is one of the main problems in quantitative MRI. Leveraging from the unique ability of deep learning, we propose a data driven strategy to derive quantitative B1 map from a single qualitative MR image without specific requirements on the weighting of the input image. B1 estimation is accomplished using a self-attention deep convolutional neural network, which makes efficient use of local and non-local information. Without additional data acquisition, an accurate estimation of B1 map is achieved, which is useful for the compensation of field inhomogeneity in T1 mapping as well as for other applications. |
| 3426 | EPI Phase correction using model-based deep learning reconstruction | |
| Lili Wang1, Fanwen Wang1, Yinghua Chu2, Xucheng Yu1, Chengyan Wang3, and He Wang1,3 | ||
1Institute of Science and Technology for Brain-Inspired Intelligence, Fudan University, Shanghai, China, 2MR Collaboration, Siemens Healthcare Ltd., Shanghai, China, 3Human Phenome Institute, Fudan University, Shanghai, China |
||
The commonly used approach of Nyquist ghost correction in echo planar imaging (EPI) include linear phase correction and model-free 2D phase correction. The recent proposed method termed ‘PEC-SENSE’ incorporates 2D phase error correction with parallel imaging can robustly eliminate Nyquist ghost for EPI data，while does not act well when a distortion mismatch exsisted between the calibration data and image data. The proposed model-based deep learning method can obtain more robust phase maps than PEC-SENSE to remove image ghost and preserve the image SNR in low or high-accelerated EPI data. |
| 3427 | The impact of multi-coil combination techniques on multi-baseline temperature imaging methods | |
| John Austin Roberts1, Henrik Odeen1, Sunil Patil2, Waqas Majeed2, Bradley Bolster3, Himanshu Bhat4, and Dennis L Parker1 | ||
1Radiology and Imaging Sciences, University of Utah, Salt lake City, UT, United States, 2Siemens Medical Solutions USA, Inc, Baltimore, MD, United States, 3Siemens Medical Solutions USA, Inc, Salt Lake City, UT, United States, 4Siemens Medical Solutions USA, Inc, Boston, MA, United States |
||
In magnetic resonance thermometry based on proton resonance frequency shift, temperature is computed based on changes in phase as measured: with respect to previously acquired images (baselines); between regions within the same image (referenceless); or some hybrid of the two approaches. In phase reconstruction algorithms, how multi-coil data is combined affects both computational efficiency and temperature measurement precision. We modify a hybrid thermometry phase reconstruction algorithm to test several coil combination methods for computational efficiency and to evaluate their effect on temperature measurement precision. Results suggest applying coil-combination prior to phase extraction both for efficiency and SNR. |
| 3428 | Fast radial T2 mapping of the liver | |
| Diana Bencikova1,2, Stephan Kannengiesser3, Gert Reiter4, Ahmed Ba-Ssalamah1, Siegfried Trattnig1,2, and Martin Krššák2,5 | ||
1Department of Radiology, Medical University Vienna, Vienna, Austria, 2Christian Doppler Laboratory for Clinical Molecular Imaging, MOLIMA, MUW, Vienna, Austria, 3Siemens Healthcare GmbH, Erlangen, Germany, 4Siemens Siemens Healthcare Diagnostics GmbH, Graz, Austria, 5Division of Endocrinology and Metabolism, Department of Medicine III, Medical University of Vienna, Vienna, Austria |
||
Radial T2 mapping of the liver is fast and motion insensitive, which is ideal for clinical applications. Here we tested a prototype radial turbo-spin-echo sequence for T2 mapping of the liver in phantoms and in-vivo in patients. In the phantoms we compared it to conventional Cartesian multi-SE sequence and tested the effect of fat suppression by comparing fat-suppressed and fat-unsuppressed values. From the patient evaluation, comparing it to single-voxel multi-echo MRS, we proved it to be suitable for clinical applications and a promising tool for characterization of diffuse liver disorders, but there are systematic differences between different methods. |
| 3429 | Spin-echo sequence with interleaved spiral acquisition avoids pulsation artifact for sagittal T1-weighted brain imaging at 3.0T | |
| Ke Jiang1, Jiazheng Wang1, and Xiaofang Xu1 | ||
1Philips Healthcare Greater China, Beijing, China |
||
T1-weighted imaging is a necessary component of clinical brain MR to provide anatomical information at high spatial resolution. T1w brain image in clinical practice is regularly acquired with two-dimensional turbo-spin-echo sequence, which suffers from the pulsation artifacts in the phase-encode direction from superior sagittal sinus, the straight sinus, and the sigmoid sinus. This study exploits the spiral trajectory for the spin-echo sequence to achieve the T1 weighting and to avoid the pulsation artifacts. |
| 3430 | On Coil Combination with Optimal SNR for Linear Multichannel k-Space Reconstruction Methods | |
| Daeun Kim1, Jonathan Polimeni2, Kawin Setsompop2, and Justin Haldar1 | ||
1Electrical and Computer Engineering, University of Southern California, Los Angeles, CA, United States, 2Athinoula A. Martinos Center for Biomedical Imaging, Department of Radiology, Harvard Medical School, Massachusetts General Hospital, Charlestown, MA, United States |
||
Noise correlations exist in multi-channel k-space data, and methods to optimally account for this correlation have been used for a long time in image-domain parallel imaging methods like SENSE. However, methods to address noise are not widely-utilized for Fourier-domain parallel imaging methods like GRAPPA, SPIRiT, and AC-LORAKS. In this work, we demonstrate that properly accounting for spatially-varying noise correlation can substantially reduce the noise level of coil-combined images. Further, we demonstrate the existence of previously-unknown correlations between the real and imaginary parts of the noise in reconstructed images. Accounting for this extra correlation can reduce the noise level even further. |
| 3431 | Transform-Domain g-Factor Maps | |
| Jiayang Wang1 and Justin P. Haldar1 | ||
1Department of Electrical and Computer Engineering, University of Southern California, Los Angeles, CA, United States |
||
The g-factor is commonly used for quantifying the noise amplification associated with accelerated data acquisition and linear image reconstruction, and is frequently used to compare different k-space sampling strategies. While previous work computes g-factors in the image domain, we observe in this work that g-factors can also be used to quantify uncertainty in various transform domains (e.g., the wavelet domain and the multi-channel Fourier domain). These transform-domain g-factor maps provide complementary information to conventional image-domain g-factor maps, and are potentially useful for k-space sampling pattern design. |
| 3432 | Compressed Sensing Parallel Imaging Without Calibration | |
| Nicholas Dwork1, Corey A. Baron2, Ethan M. I. Johnson3, Daniel O'Connor4, Adam B. Kerr5, Peder E. Z. Larson1, Jeremy Gordon1, and John M. Pauly6 | ||
1Radiology and Biomedical Imaging, University of California in San Francisco, San Francisco, CA, United States, 2Medical Biophysics, Western University, London, ON, Canada, 3Biomedical Engineering, Northwestern University, Evanston, IL, United States, 4Mathematics and Statistics, University of San Francisco, San Francisco, CA, United States, 5Center for Cognitive and Neurobiological Imaging, Stanford University, Stanford, CA, United States, 6Electrical Engineering, Stanford University, Stanford, CA, United States |
||
In this paper, the reconstructed image is the result of a compressed sensing optimization problem that includes constraints based on fundamental physics. The problem is solved using an alternating minimization approach: two convex optimization problems are alternately solved, one with the Fast Iterative Shrinkage Threshold Algorithm (FISTA) and the other with the Primal-Dual Hybrid Gradient method. Results show improved detail when compared to conventional SENSE results. |
| 3433 | pISTA-SENSE-ResNet for Parallel MRI Reconstruction | |
| Tieyuan Lu1, Xinlin Zhang1, Yihui Huang1, Di Guo2, and Xiaobo Qu1 | ||
1Department of Electronic Science, Xiamen University, Xiamen, China, 2School of Computer and Information Engineering, Xiamen University of Technology, Xiamen, China |
||
Magnetic resonance imaging has been widely applied in clinical diagnosis, however, is limited by its long data acquisition time. Although imaging can be accelerated by sparse sampling and parallel imaging, achieving promising reconstruction images with a fast reconstruction speed remains a challenge. In this work, we design the neural network structure from the perspective of sparse iterative reconstruction. The experimental results of a public knee dataset show that compared with the optimization-based method and the latest deep learning parallel imaging methods, the proposed network has less error in reconstruction and is more stable under different acceleration factors. |
| 3434 | Accelerating silent MRI using a gradient axis at 20 kHz and generalized conjugate gradient SENSE reconstruction | |
| Edwin Versteeg1, Tijl Van der Velden1, Jeroen Hendrikse1, Dennis Klomp1, and Jeroen Siero1,2 | ||
1Radiology, University Medical Center Utrecht, Utrecht, Netherlands, 2Spinoza Centre for Neuroimaging Amsterdam, Amsterdam, Netherlands |
||
A silent gradient axis driven at 20 kHz can be used to decrease acoustic noise and increase patient comfort in MR-exams. However, the speed of this silent acquisition is determined by the amount of data needed for artifact free reconstructions. In this work, we present a framework for increasing the time efficiency while maintaining image quality of silent imaging with a silent gradient axis. We show that, by using a generalized conjugate gradient SENSE reconstruction, a 2-3-fold decrease in scan time is feasible both on a phantom and in-vivo. |
| 3435 | Scan-specific, Parameter-free Artifact Reduction in K-space (SPARK) | |
| Onur Beker1,2, Congyu Liao1,3, Jaejin Cho1,3, Zijing Zhang1,4, Kawin Setsompop1,3,5, and Berkin Bilgic1,3,5 | ||
1Martinos Center for Biomedical Imaging, Charlestown, MA, United States, 2Electrical and Electronics Engineering, Bogazici University, Istanbul, Turkey, 3Harvard Medical School, Boston, MA, United States, 4College of Optical Science and Engineering, Zhejiang University, Hangzhou, China, 5Harvard/MIT Health Sciences and Technology, Cambridge, MA, United States |
||
We propose a convolutional neural network (CNN) approach that works synergistically with physics-based reconstruction methods to reduce artifacts in accelerated MRI. Given reconstructed coil k-spaces, our network predicts a k-space correction term for each coil. This is done by matching the difference between the acquired autocalibration lines and their erroneous reconstructions, and generalizing this error term over the entire k-space. Application of this approach on existing reconstruction methods show that SPARK suppresses reconstruction artifacts at high acceleration, while preserving and improving on detail in moderate acceleration rates where existing reconstruction algorithms already perform well; indicating robustness. |
| 3436 | Model based Deep Learning for Calibrationless Parallel MRI recovery | |
| Aniket Pramanik1, Hemant Kumar Aggarwal1, and Mathews Jacob1 | ||
1Electrical and Computer Engineering, The University of Iowa, Iowa City, IA, United States |
||
We introduce a fast model based deep learning approach for calibrationless parallel MRI reconstruction. The proposed scheme is a non-linear generalization of structured low-rank (SLR) methods that self learn linear annihilation filters. It pre-learns non-linear annihilation relations in the Fourier domain from exemplar data which significantly reduces the computational complexity, making it three orders of magnitude faster than SLR schemes. It allows incorporation of spatial domain prior that offers improved performance over calibrated image domain MoDL approach. The calibrationless strategy minimizes potential mismatches between calibration data and the main scan, while eliminating the need for a fully sampled calibration region. |
| 3437 | Advanced New Linear Predictive Reconstruction Methods for Simultaneous Multislice Imaging | |
| Rodrigo A. Lobos1, Tae Hyung Kim1, Kawin Setsompop2, and Justin P. Haldar1 | ||
1Electrical and Computer Engineering, University of Southern California, Los Angeles, CA, United States, 2Martinos Center for Biomedical Imaging, Charlestown, MA, United States |
||
Many autocalibrated parallel imaging reconstruction methods are based on linear-predictive/autoregressive principles, including noniterative GRAPPA-type interpolation methods, iterative SPIRiT-type annihilation methods, and structured low-rank matrix methods like PRUNO and Autocalibrated LORAKS. In principle, all of these approaches could be adapted for simultaneous multislice (SMS) reconstruction. However, in practice, GRAPPA-type SMS methods are popular, but there has been limited exploration of more advanced annihilation-based or structured low-rank matrix SMS methods. In this work, we adapt and evaluate these advanced approaches for SMS reconstruction. Results demonstrate that these advanced approaches can offer substantial improvements over simpler GRAPPA-type methods when applied to SMS. |
| 3438 | FPGA-based Coprocessor for Real-time GRAPPA Reconstruction (GR-co-RECON) | |
| Abdul Basit1,2, Omair Inam1, Mahmood Qureshi1, and Hammad Omer1 | ||
1Electrical and Computer Engineering, Comsats University Islamabad, islamabad, Pakistan, 2Computer Engineering, Khwaja Fareed University of Engineering and Information Technology, Rahim Yar Khan, Pakistan |
||
Data acquisition and reconstruction speed both are crucial for real-time MRI. However, MR image reconstruction speed is highly dependent on the processing capabilities of the hardware platforms (e.g. CPUs, GPUs). Recently, it has been observed that Field Programmable Gate Arrays (FPGA) are a potential candidate to address the computational demands of parallel MRI algorithms. This paper presents the first design effort to implement high performance 32-bit floating-point FPGA-based coprocessor for real-time GRAPPA reconstruction. In-vivo results of 12-channel, 3.0T human-head dataset show that the proposed system speeds up the image reconstruction time up to 106x without compromising image quality. |
| 3439 | MR Image Reconstruction using GRAPPA with Total Generalized Variation | |
| Rida Zainab1, Muhammad Haseeb Hassan1, Omair Inam1, Ibtisam Aslam1,2, and Hammad Omer1 | ||
1Department of Electrical and Computer Engineering, COMSATS University Islamabad, Islamabad, Pakistan, 2Department of Radiology and Medical Informatics, Hospital University of Geneva, Geneva, Switzerland |
||
GRAPPA reconstructed images may exhibit noise modulated by the receiver coil sensitivities. Total variation (TV) regularization has been recently used to solve the image de-noising problem. However, conventional TV fails to remove staircase artifacts in the reconstructed MR images due to inhomogeneities in field strength and receiver coils. In this abstract, total generalized variation (TGV) regularization is used to de-noise the GRAPPA reconstructed images, while eliminating the limitations posed by TV. Experiments are performed on 8-channel in-vivo human-head data set. The results show that the proposed method successfully removes the noise and preserve fine details in the GRAPPA reconstructed images. |
| 3440 | Parallel MRI Reconstruction via Residual UNet with Joint Consideration of k-Space and Image Space | |
| Xiaoxia Zhang1, Xiaopeng Zong1, Yong Chen1, Zhenghan Fang1, and Pew-Thian Yap1 | ||
1Department of Radiology, Department of Radiology and Biomedical Research Imaging Center (BRIC), University of North Carolina at Chapel Hill, Chapel Hill, NC, United States |
||
We propose a strategy based on residual U-Net to reconstruct MR images from undersampled multichannel data by considering both k-space and image space. Our method first imputes the missing data points in k-space by utilizing the intrinsic relationships among channels. Then, the image reconstructed from the imputed k-space data is fed to another network for spatial detail refinement. Our method does not necessarily require auto-calibration signal (ACS) and is hence less susceptible to motion-induced inconsistency between the ACS and the undersampled data. Comprehensive evaluation indicates that our method yields images with superior perceptual details. |
| 3441 | Calibration-less pMRI for the Reconstruction of Radially Encoded data using GROG based CS | |
| Husnain Javid Bhatti1,2, Fariha Aamir1, Ibtisam Aslam1,3, Khan Afsar1,4, and Hammad Omer1 | ||
1Department of Electrical and Computer Engineering, COMSATS University Islamabad (CUI), Islamabad, Pakistan, 2School of Electrical Engineering and Computers Science, National University of Science and Technology (NUST), Islamabad, Pakistan, 3Department of Radiology and Medical Informatics, Hospital University of Geneva, GENEVA, Switzerland, 4Department of Electronic science and Technology, Xiamen University, Xiamen, China |
||
To reduce MRI scan time, under-sampled non-Cartesian trajectories are used which lead to artifacts. This work proposes a new method ‘GROG with calibration-less pMRI for CS based p-thresholding’ to reconstruct MR images from the under-sampled radial k-space data. The proposed method is validated on the phantom and 1.5T human head data and provides significant improvement both visually and in terms of quantifying parameters (AP, RMSE & PSNR) e.g. 77% and 86% improvement in AP, 5% and 32% improvement in RMSE, 7% and 11% improvement in PSNR at AF=4 for the phantom data than POCS and pseudo Cartesian GRAPPA, respectively. |
| 3442 | High-frequency Transform Guided Denoising Autoencoding Prior for Parallel MR Imaging | |
| Yuanyuan Hu1, Zhuonan He2, Jinjie Zhou2, Minghui Zhang2, Qiegen Liu2, and Dong Liang1 | ||
1Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen 518055, China, Shenzhen, China, 2the Department of Electronic Information Engineering, Nanchang University, Nanchang 330031, China, Nanchang, China |
||
Ill-posed inverse problems embodied in parallel imaging remain an active research topic in several decades, with new approaches constantly emerging. Built on the observation that both dictionary learning and conventional sparse coding extract high-frequency component to model, we derived a novel strategy named HDAEP to explore the prior on high-frequency domain on the basis of denoising autoencoding. After the prior is learned from the trained network, the iteratively Gauss-Newton method is employed to jointly estimating the images and coil sensitivities. Experimental results show that the proposed method can achieve superior performances on parallel MRI reconstruction compared to state-of-the-arts. |
| 3443 | Robust Simultaneous Multi-Slice MRI Exploiting Hankel Subspace Learning with Self-Calibration and Self-Referencing Magnitude Prior | |
| Eun Ji Lim1, Guobin Li2, Chaohong Wang2, Zhaopeng Li2, Shasha Yang2, and Jaeseok Park1 | ||
1Sungkyunkwan University, Suwon, Republic of Korea, 2United Imaging Healthcare, Shanghai, China |
||
SMS methods typically consist of the following two steps: kernel calibration and reconstruction. However, discrepancies between calibration and imaging occur due to either different image contrasts or subject motions, resulting in residual aliasing artifacts. To tackle these problems, in this work we propose a robust SMS technique exploiting Hankel subspace learning with self-calibration and self-referencing magnitude prior. An SMS filter is designed to strictly control pass-bands and stop-bands to reduce the dependence of image contrast on reconstruction. Both external and internal calibrating signals are included in the calibration step, while a self-referencing magnitude prior is imposed in the reconstruction step. |
| 3444 | Phase Reconstruction using Iterative Multi-shot ESPIRiT (PRIME) | |
| Stephen Cauley1,2, Bryan Clifford3, Steffen Bollmann3, Thorsten Feiweier4, Berkin Bilgic1,2, Kawin Setsompop1,2,5, and Lawrence L. Wald1,2,5 | ||
1Department of Radiology, A. A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Charlestown, MA, United States, 2Harvard Medical School, Boston, MA, United States, 3Siemens Medical Solutions USA, Boston, MA, United States, 4Siemens Healthcare GmbH, Erlangen, Germany, 5Harvard-MIT Health Sciences and Technology, Massachusetts Institute of Technology, Cambridge, MA, United States |
||
We employ a compact phase modeling strategy for accurate multi-shot echo-planar imaging (msEPI) reconstruction. As an alternative to approaches that perform msEPI reconstruction using strict low-rank constraints, we recast the problem as an iterative relative phase estimation problem. This allows for us to utilize existing techniques such as ESPIRiT, which are formulated for determining relative magnitude and phase differences between multi-coil receive arrays. Through an iterative search we jointly estimate an artifact-free combined image and the smooth relative phase between each msEPI shot. We demonstrate the benefits of our approach for clinical and highly-accelerated multi-shot diffusion-weighted acquisitions. |
| 3445 | A rapid quantitative Multi-inversion SAGE-EPI brain protocol with subspace reconstruction and navigation-free shot-to-shot phase correction | |
| Mary Kate Manhard1,2, Zijing Dong1,3, Congyu Liao1,2, Merlin Fair1,2, Fuyixue Wang1,4, Berkin Bilgic1,2, and Kawin Setsompop1,2,4 | ||
1A.A. Martinos Center for Biomedical Imaging, Boston, MA, United States, 2Department of Radiology, Harvard Medical School, Boston, MA, United States, 3Department of Electrical Engineering and Computer Science, Massachusetts Institution of Technology, Cambridge, MA, United States, 4Harvard-MIT Division of Health Sciences and Technology, Cambridge, MA, United States |
||
Recently, multi-contrast EPI approaches have been proposed for a fast screening brain protocol. With EPI-encoding, multiple contrasts can be acquired quickly for robust, quantitative mapping as well as creation of synthetic weighted images. Here, a spatiotemporal subspace reconstruction is developed to jointly reconstruct multi-contrast multishot-EPI data from a multi-inversion Spin and Gradient echo EPI (MI-SAGE-EPI) acquisition. An approach to estimate and incorporate shot-to-shot phase corruption into the reconstruction was also developed. This navigation-free subspace reconstruction achieves good reconstruction for MI-SAGE-EPI at a high EPI-acceleration, thus enabling a rapid quantitative protocol at high in-plane resolution with minimal distortion and blurring. |
| 3446 | Highly accelerated distortion free diffusion imaging using joint k/q-space reconstruction. | |
| Wonil Lee1, Seohee So1, Jaejein Cho2, Congyu Liao2, Qiyuan Tian2, Hyunwook Park1, Elfar Adalsteinsson3, Kawin Setsompop4, and Berkin Bilgic2 | ||
1Electrical engineering, KAIST, Daejeon, Korea, Republic of, 2Martinos Center, Charlestown, MA, United States, 3MIT, Cambridge, MA, United States, 4Martinos Center, Charlestown, China |
||
We propose k/q Blip-Up and -Down Acquisition (k/q-BUDA) for distortion-free diffusion MRI (dMRI). We acquire different diffusion directions with alternating phase-encoding polarities, and incorporate $$$B_{0}$$$ information in the forward model of our reconstruction. Multiple images on neighboring q-space points are jointly reconstructed using Hankel structured low-rank regularization in k-space. This allows us to exploit similarities across q-space at the parallel imaging level, thus enabling k/q-space joint reconstruction. We select four neighboring q-space samples, providing us with two blip-up and two blip-down phase-encoded shots, who provide complementary $$$B_{0}$$$ encoding to eliminate distortion in the reconstructed volumes. |
| 3447 | Weighted-nuclear-norm minimization for low-rank approximation -- application to multi-shot diffusion-weighted MRI reconstruction | |
| Yuxin Hu1,2, Qiyuan Tian1,2, Yunyingying Xu2, Philip K. Lee1,2, Bruce L. Daniel1,3, and Brian A. Hargreaves1,2,3 | ||
1Department of Radiology, Stanford University, Stanford, CA, United States, 2Department of Electrical Engineering, Stanford University, Stanford, CA, United States, 3Department of Bioengineering, Stanford University, Stanford, CA, United States |
||
Low rankness of image/k-space data has been exploited in many different MRI applications. In this work, we introduce weighted-nuclear-norm minimization for MRI low-rank reconstruction, in which smaller thresholds are used for larger eigenvalues to reduce information loss. Our simulation results demonstrate that weighted nuclear norm could serve as a better rank approximation compared to (unweighted) nuclear norm. With this technique, we achieve 10-shot DWI and high-fidelity half-millimetre DWI reconstruction with significantly reduced ghosting artifacts. |
| 3448 | A Novel Compressed Sensing Multi-Slice Cartesian MRI via Alternating Phase Encoding Directions for Acceleration and Flexible Undersampling | |
| Zheyuan Yi1,2,3, Yilong Liu1,2, Fei Chen3, and Ed X. Wu1,2 | ||
1Department of Electrical and Electronic Engineering, The University of Hong Kong, Hong Kong, China, 2Laboratory of Biomedical Imaging and Signal Processing, The University of Hong Kong, Hong Kong, China, 3Department of Electrical and Electronic Engineering, Southern University of Science and Technology, Shenzhen, China |
||
In conventional compressed sensing (CS) multi-slice Cartesian 2D imaging, the undersampling is performed along phase-encoding direction only, leading to coherent 1D aliasing that significantly limits the effectivess of CS for acceleration. This study proposes a multi-slice CS reconstruction method to take advantage of extremely augmented sampling incoherence created by orthogonally alternating phase-encoding directions among adjacent slices. The multi-slice CS approach was evaluated with single-channel brain T1W and T2W datasets. The results demonstrate significant improvements with both pseudo-random and uniform undersampling. This new method will also greatly augment the existing 2D CS parallel imaging technqiues for very high acceleration. |
| 3449 | Multi-Component Quantitative T2 Shuffling: Accelerating Myelin Water Imaging by 20-30x | |
| Adam V. Dvorak1,2, Guillaume Gilbert3, Alex L. MacKay1,4, and Shannon H. Kolind1,2,4,5 | ||
1Physics and Astronomy, University of British Columbia, Vancouver, BC, Canada, 2International Collaboration on Repair Discoveries (ICORD), Vancouver, BC, Canada, 3MR Clinical Science, Philips Canada, Markham, ON, Canada, 4Radiology, University of British Columbia, Vancouver, BC, Canada, 5Medicine (Neurology), University of British Columbia, Vancouver, BC, Canada |
||
Multi-component quantitative T2 mapping can provide a range of valuable, in-vivo biomarkers, but is limited by lengthy acquisition times. Here we introduce multi-component quantitative T2 shuffling, a subspace-constrained CS method for reconstructing highly under-sampled multi-component relaxation mapping data using principal components of a temporal basis. We demonstrated reconstruction of myelin water imaging data with simulated under-sampling acceleration factors of ~20-30, which could provide accurate images, higher resolution and fit-to-noise ratios, and improved metric maps in a fraction of the acquisition time. |
| 3450 | Multiscale Low Rank Matrix Decomposition for Reconstruction of Accelerated Cardiac CEST MRI | |
| Ilya Chugunov1, Wissam AlGhuraibawi2, Kevin Godines2, Bonnie Lam2, Frank Ong3, Jonathan Tamir1,4, and Moriel Vandsburger2 | ||
1Electrical Engineering and Computer Science, University of California, Berkeley, Berkeley, CA, United States, 2Bioengineering, University of California, Berkeley, Berkeley, CA, United States, 3Electrical Engineering, Stanford University, Stanford, CA, United States, 4Electrical and Computer Engineering, University of Texas at Austin, Austin, TX, United States |
||
Multiscale low rank reconstruction has been demonstrated to efficiently reconstruct non-gated dynamic MRI by leveraging sparsity in the time domain. This abstract demonstrates its ability to reconstruct 4-fold accelerated CEST imaging of the heart via similarly exploiting sparsity in the Z-spectrum domain. This reconstruction outperforms zero-filled IFFT for quantization of magnetization transfer, nuclear overhauser, and CEST effects as derived from Lorentzian-line-fit analysis. Extension to volumetric or motion inclusive CEST imaging and development of a new regularization function may enable further acceleration. |
| 3451 | MRI Below the Noise Floor | |
| Gregory Lemberskiy1,2, Steven Baete1, Jelle Veraart1, Timothy M. Shepherd1, Els Fieremans1, and Dmitry S Novikov1 | ||
1New York University School of Medicine, New York, NY, United States, 2Microstructure Imaging INC, New York, NY, United States |
||
We develop random matrix theory (RMT)-based MRI image reconstruction able to increase SNR by up to 10-fold, and to radically increase resolution for routine clinical acquisitions. RMT offers an objective criterion for separating signal from noise across all coils, voxels and MRI contrasts, by utilizing the redundancy in MRI measurements. We demonstrate RMT on a 0.8x0.8x0.8 mm3 neuro exam that includes a series of multiple T2w, T1w, diffusion, and fMRI images on a 3T clinical scanner. RMT can serve as a paradigm for reconstructing multiple contrasts, enhancing image quality for low-field scanners, increasing MR value, and improving biomarker precision. |
| 3452 | Highly accelerated T1ρ imaging with Kernel-based low-rank compressed sensing: evaluation of retrospective and prospective undersampling | |
| Jeehun Kim1,2, Chaoyi Zhang3, Mingrui Yang1,2, Hongyu Li3, Mei Li1,2, Richard Lartey1,2, Leslie Ying3,4, and Xiaojuan Li1,2,5 | ||
1Biomedical Engineering, Cleveland Clinic, Cleveland, OH, United States, 2Program of Advanced Musculoskeletal Imaging (PAMI), Cleveland Clinic, Cleveland, OH, United States, 3Electrical Engineering, University at Buffalo, the State University of New York, Buffalo, NY, United States, 4Biomedical Engineering, University at Buffalo, the State University of New York, Buffalo, NY, United States, 5Department of Diagnostic Radiology, Cleveland Clinic, Cleveland, OH, United States |
||
Quantitative T1ρ MRI provides valuable information on compositional changes in cartilage, but requires longer scan time compared to conventional imaging. In this work, kernel-based low-rank compressed sensing reconstruction was used to accelerate T1ρ imaging and the retrospective and prospective undersampled results from four subjects were compared to reference T1ρ values. |
| 3453 | Versatile Parameter-Free Compressed Sensing MRI with Approximate Message Passing | |
| Charles Millard1,2, Aaron T Hess2, Boris Mailhé3, and Jared Tanner1 | ||
1Mathematical Institute, University of Oxford, Oxford, United Kingdom, 2Oxford Centre for Clinical Magnetic Resonance, University of Oxford, Oxford, United Kingdom, 3Siemens Healthineers, Princeton, NJ, United States |
||
We consider two algorithmic challenges for the compressed sensing MRI community: (1) the difficulty of tuning free model parameters and (2) the need to converge quickly. The authors have developed a parameter-free approach to reconstruction which accommodates structurally rich regularizers that can be automatically adapted to near-optimality, removing the need for manual adjustment between images or sampling schemes. We evaluate the algorithm’s performance on three test images of varying type and dimension and find that it converges faster and to a lower mean-squared error than its competitors, even when they are optimally tuned. |
| 3454 | Accelerated 4D MRI using joint higher degree total variation and local low-rank constraints (HDTV-LLR) | |
| Yue Hu1, Disi Lin1, and Dong Nan2 | ||
1Harbin Institute of Technology, Harbin, China, 2The First Affiliated Hospital of Harbin Medical University, Harbin, China |
||
Four-dimensional (4D) MRI can provide 3D tissue properties and the temporal profiles at the same time. However, further applications of 4D MRI is limited by the long acquisition time and motion artifacts. We propose a new 4D MRI reconstruction scheme, named HDTV-LLR, by integrating the three-dimensional higher degree total variation and the local low-rank penalties. We demonstrate the benefits of the proposed method using 4D cardiac MRI dataset with undersampling factors of 12 and 16. The proposed method is compared with iGRASP, and schemes using either low-rank or sparsity constraint alone. Results show improved image quality and reduced artifacts. |
| 3455 | Learning-Based Sampling Pattern for Compressed Sensing and Low Rank Reconstructions using Multicoil MR Images of Human Knee Joint | |
| Marcelo V. W. Zibetti1, Gabor T. Herman2, and Ravinder R. Regatte1 | ||
1Radiology, NYU, New York, NY, United States, 2The Graduate Center, CUNY, New York, NY, United States |
||
Compressed Sensing (CS) and parallel MRI (pMRI) have been successfully applied to accelerate MRI-data acquisition. CS requires incoherence, usually achieved by random undersampling the data, but pMRI does not. Combined, these methods allow even higher acceleration rates. However, it is unknown how the sampling pattern (SP) should be selected. It is also unknown if the SP is dependent on the reconstruction method. Here we demonstrate, using a new algorithm, that the SP can be learned from given data and reconstruction method. Our results show that the learned SP is superior to others such as Poisson disk and variable density. |
| 3456 | Combining Supervised and semi-Blind Dictionary (Super-BReD) Learning for MRI Reconstruction | |
| Anish Lahiri1, Saiprasad Ravishankar2, and Jeffrey A Fessler1 | ||
1Electrical and Computer Engineering, University of Michigan, Ann Arbor, MI, United States, 2Computational Mathematics, Science and Engineering, and Biomedical Engineering, Michigan State University, East Lansing, MI, United States |
||
Regularization in MRI reconstruction often involves sparse representation of signals using linear combinations of dictionary atoms. In 'blind' settings, these dictionaries are learned during reconstruction from the corrupt/aliased images, using no training data. In contrast, 'Fully supervised' dictionary learning (DL) requires uncorrupted/fully sampled training images, and the learned dictionary is used to regularize image reconstruction from undersampled data. We combine the aforementioned DL frameworks to learn two separate dictionaries in a residual fashion to jointly reconstruct an undersampled image. Our algorithm, Super-BReD Learning, shows promising results on reconstruction from retrospectively undersampled data, and outperforms recent DL schemes. |
| 3457 | Comparison of fixed vs. low rank reconstruction methods for acceleration of quantitative T2 mapping | |
| Daniel Grzeda1, Meirav Galun2, Noam Omer1, Tamar Blumenfeld-Katzir1, Dvir Radunsky1, Ricardo Otazo3, and Noam Ben-Eliezer1,4,5 | ||
1Department of Biomedical Engineering, Tel Aviv University, Tel Aviv, Israel, 2Department of Computer Science and Applied Mathematics, Weizmann Institute of Science, Rehovot, Israel, 3Departments of Medical Physics and Radiology, Memorial Sloan Kettering Cancer Center, New York, NY, United States, 4Center for Advanced Imaging Innovation and Research (CAI2R), New-York University Langone Medical Center, New York, New York, NY, United States, 5Sagol School of Neuroscience, Tel-Aviv University, Tel Aviv, Israel |
||
Quantification of T2 values is valuable for a wide range of research and clinical applications. Multi-echo spin echo protocols allow mapping T2 values, yet, at the cost of strong contamination from stimulated echoes. The echo-modulation-curve algorithm can efficiently overcome these limitations to produce accurate T2 values. Still, integration into clinical routine requires further acceleration in scan time. In this work we present a fixed-rank and sparse algorithm (SPARK) for accelerating the acquisition of T2 values, and compare it against standard L+S and GRAPPA. SPARK was found to improve robustness to reconstruction parameters, and achieve superior accuracy at high acceleration factors. |
| 3458 | Low-Rank Tensor Completion for Accelerated Magnetic Resonance Imaging | |
| Shen Zhao1, Lee C. Potter1, and Rizwan Ahmad2 | ||
1Department of Electrical and Computer Engineering, The Ohio State University, Columbus, OH, United States, 2Department of Biomedical Engineering, The Ohio State University, Columbus, OH, United States |
||
We present a method for calibration-less, accelerated Magnetic Resonance Imaging (MRI) via canonical polyadic decomposition (CPD) based low-rank tensor completion (LRTC). LRTC exploits the higher dimensional structure inherent in MRI. Preliminary results show that LRTC can outperform structured low-rank matrix completion methods for 2D and compressed sensing-based methods for dynamic applications. |
| 3459 | Data-Driven Model Consistency Condition Reconstruction Performance for Radial Acquisition in Breast DCE-MRI | |
| Ping N Wang1, Julia V Velikina1, Leah C Henze Bancroft2, Alexey A Samsonov2, and James H Holmes2 | ||
1Department of Medical Physics, University of Wisconsin Madison, Madison, WI, United States, 2Department of Radiology, University of Wisconsin Madison, Madison, WI, United States |
||
MOCCO reconstruction enables high spatial and temporal resolution DCE imaging for breast, which was proposed to be less sensitive to modeling error. However, the impact of noise level and lesion size on reconstruction performance is remained unknown. In this work we use a digital reference object phantom with the ability to adjust both spatial and temporal features for breast pharmacokinetic simulation. Further, we perform G-factor analysis of SNR performance, sharpness assessment, nRMSE and PK analysis for evaluating the temporal fidelity. |
| 3460 | Sparse-Coding Regularized QSM Reconstruction for Suppressing Motion Artifacts | |
| Jingjia Chen1 and Chunlei Liu1,2 | ||
1EECS, University of California, Berkeley, Berkeley, CA, United States, 2Helen Wills Neuroscience Institute, University of California, Berkeley, Berkeley, CA, United States |
||
Subject motion downgrades QSM image quality and accuracy and can even nullify the image for diagnostic purposes in clinical settings. While QSM plays an emerging role in evaluating neurodegenerative diseases, motion artifact reduction is crucial for its adoption by researchers and clinicians. In this project, we develop a sparse-coding regularized QSM reconstruction algorithm to mitigate motion artifacts and noise. In vivo experiments suggest that the proposed method can alleviate motion artifacts to a certain extent while preserving sharp structures. This regularization technique can be applied jointly with other regularizations to achieve a desired susceptibility map. |
| 3461 | Quantitative comparison of a brain protocol with and without FRONSAC enhancement | |
| E. H. Bhuiyan1, Y. Rodriguez1, R. Todd Constable1, and G. Galiana1 | ||
1Radiology and Biomedical Imaging, Yale University, New Haven, CT, United States |
||
FRONSAC has previously been demonstrated as a highly effective approach to using nonlinear gradients to reduce undersampling artifacts, but previous results were limited to proof-of-principle experiments in individual subjects. Here we report quantitative comparisons for a full protocol of standard brain imaging sequences (2D GRE, 2D TSE, 2D T2w-FLAIR, 3D GRE and 3D MP-RAGE) acquired in a cohort of healthy subjects, comparing standard Cartesian to FRONSAC-enhanced acquisitions. Preliminary results confirm that FRONSAC significantly improves undersampling artifacts, measured as RMSE relative to a fully sampled Cartesian reference, while retaining the contrast, SNR, and reliability of standard Cartesian sequences. |
| 3462 | Contrast prediction-based regularization for iterative reconstructions (PROSIT) | |
| Hendrik Mattern1, Alessandro Sciarra1,2, Max Dünnwald2,3, Soumick Chatterjee1,3,4, Ursula Müller1, Steffen Oeltze-Jafra2,5, and Oliver Speck1,5,6,7 | ||
1Biomedical Magnetic Resonance, Otto-von-Guericke University, Magdeburg, Germany, 2Medicine and Digitalization, Otto-von-Guericke University, Magdeburg, Germany, 3Faculty of Computer Science, Otto-von-Guericke University, Magdeburg, Germany, 4Data & Knowledge Engineering Group, Otto-von-Guericke-University, Magdeburg, Germany, 5Center for Behavioral Brain Sciences, Magdeburg, Germany, 6German Center for Neurodegenerative Disease, Magdeburg, Germany, 7Leibniz Institute for Neurobiology, Magdeburg, Germany |
||
In this study, contrast prediction is used as an auxiliary tool to regularize underdetermined image reconstructions. This novel regularization strategy enables to share information across individual reconstructions and outperforms state of the art regularizations for high acceleration factors. |
| 3463 | Highly Accelerated Free Breathing Whole Heart Real-time Cine using AuTO-calibrated Multiband Imaging and Compressed Sensing (ATOMICS) | |
| Yu Ding1, Lele Zhao2, Zhongqi Zhang2, Qi Liu1, Jian Xu1, and Yuan Zheng1 | ||
1UIH America, Houston, TX, United States, 2United Imaging Healthcare, Shanghai, China |
||
Conventional whole heart cine scan requires multiple breath-holds, and the prolonged scan time may limit its clinical use. We propose a highly accelerated whole heart real-time cine technique that combines AuTO-calibrated Multiband (MB) Imaging and Compressed Sensing (ATOMICS). With specially designed temporal phase modulation and in-plane undersampling pattern, single-band reference images can be extracted from the multiband data themselves, and cine images can be reconstructed using the CS framework. The proposed technique was demonstrated on a healthy volunteer with 16-fold acceleration (MB=2, undersampling factor=8), and free-breathing whole heart cine with diagnostic quality was acquired in about 12 seconds. |
| 3464 | Improving cartesian single-shot 2D T2 shuffling and reducing radial streaking artifacts with golden angle radial T2 shuffling | |
| Yamin Arefeen1, Fei Han2, Borjan Gagoski3,4, Jacob White1, and Elfar Adalsteinsson1,5,6 | ||
1Electrical Engineering and Computer Science, Massachusetts Institute of Technology, Cambridge, MA, United States, 2MR R&D and Collaborations, Siemens Healthineers, Los Angeles, CA, United States, 3Fetal-Neonatal Neuroimaging and Developmental Science Center, Boston Children’s Hospital, Boston, MA, United States, 4Harvard Medical School, Boston, MA, United States, 5Harvard-MIT Health Sciences and Technology, Cambridge, MA, United States, 6Institute for Medical Engineering and Science, Cambridge, MA, United States |
||
T2 shuffling reduces blurring in 3D Fast Spin Echo (FSE) by incorporating a low rank representation of signal evolution in the forward model. However, extending T2 shuffling to 2D-FSE is challenging due to the limited spatiotemporal incoherence achievable with cartesian sampling. Radial trajectories offer increased incoherence in multiple dimensions in comparison to cartesian acquisitions. We combine golden angle radial sampling with the T2 shuffling forward model to improve image quality in comparison to 2D cartesian T2 shuffling and to reduce streaking artifacts in radial single-shot FSE images. Simulation and phantom experiments illustrate the advantages of radial T2 shuffling. |
| 3465 | Enhancing GRE magnitude image using rank reduction of structured matrix | |
| Sreekanth Madhusoodhanan1, Vazim Ibrahim1, Chandrasekharan Kesavadas2, and Joseph Suresh Paul1 | ||
1Indian Institute of Information Technology and Management-Kerala, Trivandrum, India, 2Sree Chitra Tirunal Institute for Medical Science and Technology, Trivandrum, India |
||
The noise and phase dispersion in the multi-slice multi-echo Gradient Echo (GRE) acquisition causes non-exponential signal decay along the temporal dimension. Due to this, signal dropouts are observed in the air tissue interface. Construction of structured block Hankel matrix and rank minimization with a fidelity for data consistency reduces the signal loss and improves the visibility of venous structures in the GRE magnitude image. |
| 3466 | Can an Adapted SEM Accurately Deliver 129Xe MRI-based Lung Morphometry Estimates | |
| Elise Noelle Woodward1, Matthew S Fox1,2, David G McCormack3, Grace Parraga3,4,5, and Alexei Ouriadov1,2 | ||
1Physics and Astronomy, Western University, London, ON, Canada, 2Lawson Health Research Centre, London, ON, Canada, 3Department of Medicine, Respiratory, Western University, London, ON, Canada, 4Robarts Research institute, London, ON, Canada, 5Medical Biophysics, Western University, London, ON, Canada |
||
We hypothesize that the SEM equation can be adapted for fitting the spin-density dependence of the MR signal similar to fitting the time or b-value dependences: Signal=exp[-(nr)^β], where 0<β<1, n is an image number and r is the fractional spin-density. Such an approach permits consideration of the signal intensity variation as reflection of the underlying spin-density variation and hence, reconstruction of the under-sampled k-space using the adapted SEM equation. In this proof-of-concept evaluation, we have demonstrated the feasibility of this approach in a small group of patients. Lung SV/ADC/morphometry/T2* maps have been generated using reconstructed images and their corresponding weighing. |
| 3467 | Model-Based Iterative Reconstruction for diffusion Data with Phase Correction and Penalty Term | |
| Thomas Hüfken1, Jannik Arbogast1, Anna-Katinka Bracher2, Meinrad Beer2, Henning Neubauer2, and Volker Rasche 1 | ||
1Internal Medicine II, University Ulm Medical Center, Ulm, Germany, 2Department of Radiology, University Ulm Medical Center, Ulm, Germany |
||
A diffusion model is used to generate a synthetic k-space which is compared with acquired k-space data. Minimization of the cost function with an additional regularization term in form of a total variation is performed by a nonlinear conjugated gradient solver. Phase alterations due to strong diffusion gradients are considered by introduction of a phase operator. Data of nine human knee joints were acquired with a multi-shot diffusion weighted EPI sequence. Undersampling factors between R=1.5 and R=3.57 yielded excellent results for magnitude and ADC data. |
| 3468 | Direct Estimation of pre-contrast T1 for DCE-MRI | |
| Zhibo Zhu1, Yannick Bliesener1, R. Marc Lebel2,3, and Krishna Shrinivas Nayak1 | ||
1Electrical and Computer Engineering, University of Southern California, Los Angeles, CA, United States, 2Global MR Applications & Workflow, GE Healthcare, Calgary, AB, Canada, 3Radiology, University of Calgary, Calgary, AB, Canada |
||
Quantitative DCE-MRI requires pre-contrast T1 mapping with matching resolution and coverage. Recent studies have shown that sparse sampling and constrained reconstruction can be applied to both DCE-MRI and pre-contrast T1 mapping with the variable flip angle (VFA) approach. Here, we demonstrate and evaluate direct estimation of T1 from sparsely sampled VFA raw data. In healthy subjects, we demonstrate superior T1 estimation compared with prior methods at subsampling factors >10. |
| 3469 | Near bias-free robust MP2RAGE reconstruction using a signal-dependent regularization function | |
| Yi-Cheng Hsu1, He Wang2,3, and Ying-Hua Chu1 | ||
1MR Collaboration, Siemens Healthcare Ltd., Shanghai, China, 2Institute of Science and Technology for Brain-Inspired Intelligence, Fudan University, Shanghai, China, 3Human Phenome Institute, Fudan University, Shanghai, China |
||
MP2RAGE is a B1-insensitive T1-weighted imaging and T1 mapping method with strong background noise. Introducing a regularization parameter into the reconstruction can suppress the noise at the cost of lowering the signal intensity and T1 value. We propose a robust MP2RAGE using a signal-dependent regularization function. The reconstructed images are nearly bias-free with minimal artifact and reduced background noise. |
| 3470 | Online reconstruction of GRE Fat Navigators with Gadgetron on Siemens Terra 7T Scanner | |
| Ayan Sengupta1,2, Iulius Dragonu3, and Christopher T. Rodgers2 | ||
1Department of Psychology, Royal Holloway, University of London, London, United Kingdom, 2Wolfson Brain Imaging Centre, University of Cambridge, Cambridge, United Kingdom, 3Siemens Healhineers, London, United Kingdom |
||
High resolution Ultra High Field 7T imaging is highly prone towards involuntary motion artifacts. Fat Navigator based motion correction provides a robust solution but it is a retrospective correction method. In this study we present an online reconstruction method 3D GRE FatNavs with open-source reconstruction tool, Gadgetron. We improved the performance of the GRAPPA reconstruction pipeline in Gadgetron for fast online reconstruct FatNav. We also implemented a Python Gadget to perform fsl Flirt based co-registration through NiPype to produce motion parameters from the FatNav. |
| 3471 | High Resolution Sodium MRI Through Manifold Approximation at 3T | |
| Fernando Boada1, Hugh Wang1, Yongxian Qian1, Georg Schramm2, and Johan Nuyts2 | ||
1Radiology, New York University, New York, NY, United States, 2Katholieke Universiteit Leuven, Leuven, Belgium |
||
We demonstrate the use of manifold approximation for implementing anatomically guided reconstructions for the removal of T2-related blurring from sodium images acquired at 3T. |
| 3472 | Prime Factorizations for Accelerated Radial SC-GROG | |
| Nicholas McKibben1,2 and Edward V. DiBella1,2 | ||
1Biomedical Engineering, University of Utah, Salt Lake City, UT, United States, 2Radiology and Imaging Sciences, University of Utah, Salt Lake City, UT, United States |
||
GROG is an attractive alternative to convolutional gridding and non-uniform DFT methods because of comparatively low cost and no density correction. However, for large multicoil datasets, many fractional matrix powers must be performed which scale with the cube of the number of channels. For SC-GROG and real-time SC-GROG, time and memory requirements can be significantly lowered for precomputation and updates of fractional powers by decomposing required powers into smaller, composable pieces. This is an NP-hard combinatorial change-making problem. We propose a simple solution based on prime factorization which leads to significant computational and memory savings with little performance degradation. |
| 3473 | Calibration-free Highly Accelerated Multi-slice MRI via Alternating Phase Encoding Directions and Low-rank Tensor Completion Reconstruction | |
| Yujiao Zhao1,2, Zheyuan Yi1,2,3, Yilong Liu1,2, Fei Chen3, Yanqiu Feng4, and Ed X. Wu1,2 | ||
1Laboratory of Biomedical Imaging and Signal Processing, The University of Hong Kong, Hong Kong, China, 2Department of Electrical and Electronic Engineering, The University of Hong Kong, Hong Kong, China, 3Department of Electrical and Electronic Engineering, Southern University of Science and Technology, Shenzhen, China, 4School of Biomedical Engineering, Southern Medical University, Guangzhou, China |
||
2D multi-slice MR data share strong correlations in coil sensitivities and image contents over adjacent slices. Here we’ve developed a new acquisition and reconstruction strategy for calibration-free multi-slice MRI. In brief, k-space data for each slice are uniformly undersampled along one phase encoding direction, while undersampled data for next adjacent slice are acquired with phase encoding along an orthogonal direction. Multiple images are then jointly reconstructed using a low-rank Hankel tensor completion approach. This method maximizes the incoherence of aliasing artifacts, and utilizes the coil sensitivity and image content correlations across adjacent slices, leading to high accelerations with uniform undersampling. |
| 3474 | Sparse sampling reconstruction and noise removal of 0.55T brain MRI using a learned variational network | |
| Patricia M. Johnson1, Zhang Le2, David Grodzki3, and Florian Knoll1 | ||
1Center for Biomedical Imaging, New York University, New york, NY, United States, 2Siemens Shenzhen Magnetic Resonance Ltd., Shenzhen, China, 3Siemens Healthcare GmbH, Erlangen, Germany |
||
Most clinical MRI scanners operate at high magnetic field, however low-field MRI offers many advantages and promises to improve the value of MRI. The main drawback is low SNR; several signal averages are often required, which may result in prohibitively long scans. We can look to deep learning (DL) to facilitate accelerated low-field imaging through both denoising and sparse sampling. In this work, we use a variational network for both denoising and under-sampled reconstruction of brain images acquired on a 0.55T prototype system, demonstrating that low-field MRI paired with DL can produce high-quality images in very short scan times. |
| 3475 | Accelerated Free-breathing 3D Whole-heart Black-blood TSE Imaging | |
| Giovanna Nordio1, Radhouene Neji2, Karl Kunze2, Rene Botnar1, and Claudia Prieto1 | ||
1Biomedical Engineering Department, School of Biomedical Engineering and Imaging Sciences, King's College London, London, United Kingdom, 2MR Research Collaborations, Siemens Healthcare Limited, Frimley, United Kingdom |
||
In this study we investigate the feasibility of a 3D turbo spin-echo imaging technique with variable flip angle combined with an undersampled Cartesian variable density trajectory (3D TSE-VDCASPR) for efficient whole-heart black-blood imaging. Data from five healthy subjects are acquired with the proposed 3D TSE-VDCASPR and the conventional 3D TSE with Cartesian GRAPPA. The proposed imaging 3D TSE-VDCASPR imaging technique allows for comparable black-blood imaging, with the advantage of reduce nominal scan time (4.6±1 vs 8.1±1.6 minutes ± seconds). Future work will investigate more advanced motion compensation and image reconstruction techniques in order to achieve predictable and fast scan time. |
| 3476 | Optimal data sampling and image reconstruction for Cartesian bSSFP ASL-CMR | |
| Erum Mushtaq1, Ahsan Javed1, and Krishna S. Nayak1 | ||
1University of Southern California (USC), Los Angeles, CA, United States |
||
Arterial spin labeled cardiovascular magnetic resonance (ASL-CMR) is a non-contrast myocardial perfusion imaging technique, which can detect angiographically significant coronary artery disease. Sensitivity of ASL-CMR can be improved by shortening the imaging window with parallel imaging which reduces physiological noise from cardiac motion. In this work, we explore the optimal Cartesian sampling pattern and parallel imaging reconstruction strategy for bSSFP ASL-CMR. We consider different under-sampling masks, acceleration factors, and reconstruction techniques to minimize the imaging window. The optimal setting is selected based on the bias in MP measurements. |
| WITHDRAWN |
| 3477 | Accelerated MR-STAT Algorithm: Achieving 10-minute High-Resolution Reconstructions on a Desktop PC | |
| Hongyan Liu1,2, Oscar van der Heide1,2, 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 |
||
MR-STAT is a framework for simultaneous mapping of quantitative MR parameters from a single short scan. Since MR-STAT involves the solution of a large scale nonlinear optimization problem, the reconstruction time has always been one main concern. In the current work, we develop an accelerated MR-STAT algorithm, which achieves two order of magnitude acceleration in reconstruction times. High-resolution 2D dataset can be reconstructed within 10 minutes on a desktop PC thereby drastically facilitating the application of MR-STAT in the clinical work-flow. |
| 3478 | Automatic segmentation for liver fat quantification | |
| Elisabeth Sarah Pickles1,2, Alexandre Bagur1,2, Ged Ridgway2, Benjamin Irving2, Daniel Bulte1, and Michael Brady2,3 | ||
1Institute of Biomedical Engineering, Oxford University, Oxford, United Kingdom, 2Perspectum Diagnostics, Oxford, United Kingdom, 3Department of Oncology, Oxford University, Oxford, United Kingdom |
||
By segmenting the liver on an MRI Proton Density Fat Fraction (PDFF) map a median PDFF value is obtained, indicating the amount of fat in the liver. Automatic segmentation is desirable, as manual segmentation is time consuming. We investigated a direct PDFF automatic segmentation method using a U-Net model and compared it to a T1-based PDFF segmentation. We show that the median values obtained are comparable, and the Dice scores are relatively good, although not as high as desired. Visually the direct PDFF segmentation is not always optimal. We suggest that improvement of the model is desirable. |
| 3479 | Isotropic MRI Reconstruction with 3D Convolutional Neural Network | |
| Xiaole Zhao1, Tian He1, Ying Liao1, Yun Qin1, Tao Zhang1,2,3, and Mark Zou1,2,3 | ||
1School of Life Science and Technology, University of Electronic Science and Technology of China, Chengdu, China, 2High Field Magnetic Resonance Brain Imaging Laboratory of Sichuan, Chengdu, China, 3Key Laboratory for NeuroInformation of Ministry of Education, Chengdu, China |
||
Typical magnetic resonance imaging (MRI) usually shows distinct anisotropic spatial resolution in imaging plane and slice-select direction. Image super-resolution (SR) techniques are widely used as an alternative method to isotropic MRI reconstruction. In this work, we propose to reconstruct isotropic magnetic resonance (MR) volumes via 3D convolutional neural network in an end-to-end manner. 3D SRCNN is utilized to preliminarily validate the idea and it produces quantitative and qualitative results significantly superior to traditional methods, such as Cube-Avg and NLM methods. |
| 3480 | Retrospective Artifact Correction of Pediatric MRI via Disentangled Cycle-Consistency Adversarial Networks | |
| Siyuan Liu1, Kim-Han Thung1, Weili Lin1, Pew-Thian Yap1, and the UNC/UMN Baby Connectome Project Consortium2 | ||
1Biomedical Research Imaging Center, University of North Carolina at Chapel Hill, Chapel Hill, NC, United States, 2University of North Carolina at Chapel Hill, Chapel Hill, NC, United States |
||
Retrospective artifact removal using supervised learning requires explicit generation of artifact-corrupted images and is impractical since generating the wide variety of potential artifacts can be challenging. Using unsupervised learning, we show how artifacts can be disentangled with remarkable efficacy from artifact-corrupted images to recover the artifact-free counterparts, without requiring explicit artifact generation. |
| 3481 | Deep Cerebellar Nuclei Segmentation via Semi-Supervised Deep Context-Aware Learning from 7T Diffusion Image | |
| Jinyoung Kim1, Rémi Patriat1, Jordan Kaplan1, and Noam Harel1 | ||
1Center for Magnetic Resonance Research, University of Minnesota, Minneapolis, MN, United States |
||
In this study, we proposed the first deep learning and 7T MR imaging based dentate and interposed nuclei segmentation framework. We introduce dilated dense blocks to effectively encode contextual information on different receptive fields in an encoder-decoder network. Training of the proposed network is optimized with a multi-class hybrid segmentation loss, handling a class imbalance problem. Moreover, a self-training strategy facilitates the training of the proposed network by exploiting auxiliary labels. The proposed framework significantly outperforms an atlas-based deep cerebellar nuclei segmentation tool and state-of-the-art deep neural networks in terms of accuracy and consistency. |
| 3482 | PROJECTION GAN: Highly accelerated projection reconstruction using generative adversarial neural network | |
| Tetiana Dadakova1, Jian Wu1, Hyun-Kyung Chung1, Brian Anhalt1, Dmitry Tkach1, Alexander Graff1, Natalie Marie Schenker-Ahmed1, David Karow1, and Christine Leon Swisher1 | ||
1Human Longevity, Inc., San Diego, CA, United States |
||
Many clinical MRI applications in chest and abdomen require low sensitivity to motion. In addition, high acquisition speed is necessary for imaging in non-cooperative patients or those unable to perform breath holds. These applications would benefit from the highly accelerated radial acquisition. Deep learning has been shown to provide good results for image reconstruction from highly under-sampled k-space data. Here we introduce a Projection GAN - a generative adversarial neural network, which is trained to reconstruct highly accelerated MR images from uniformly rotated projections. Our results show that even with aggressive under-sampling the reconstruction has great overall performance. |
| 3483 | Quantitative 3-D SWIFT with compressed sensing | |
| Antti Paajanen1, Olli Nykänen1, Matti Hanhela1, Nina Hänninen1,2, Swetha Pala1, Ville Kolehmainen1, and Mikko J Nissi1 | ||
1Department of Applied Physics, University of Eastern Finland, Kuopio, Finland, 2Research Unit of Medical Imaging, University of Oulu, Oulu, Finland |
||
Despite the extra information offered by the quantitative magnetic resonance imaging (qMRI), these methods are not widely used due to their long acquisition times. Our approach for fast qMRI was built on minimized data acquisition with ultra-short echo time imaging, and compressed sensing image reconstruction. To validate the approach, a set of variable flip angle images of equine osteochondral specimens were acquired with the Multi-Band-SWIFT sequence. The 4-D image stack was reconstructed with CS-framework utilizing spatial and contrast regularizations. Results with 81% reduced data showed comparable image quality while maintaining correct contrast modulation for quantitative parameter (T1 relaxation time) estimation. |
| 3484 | Domain adaptation for prostate lesion segmentation on VERDICT-MRI | |
| Eleni Chiou1,2, Francesco Giganti3,4, Shonit Punwani5, Iasonas Kokkinos2, and Eleftheria Panagiotaki1,2 | ||
1Centre of Medical Imaging Computing, University College London, London, United Kingdom, 2Department of Computer Science, University College London, London, United Kingdom, 3Department of Radiology, UCLH NHS Foundation Trust, University College London, London, United Kingdom, 4Division of Surgery & Interventional Science, University College London, London, United Kingdom, 5Centre for Medical Imaging, Division of Medicine, University College London, London, United Kingdom |
||
The successful adoption of convolutional neural networks (CNNs) for improved diagnosis can be hindered for pathologies and clinical settings where the amount of labelled training data is limited. In such cases, domain adaptation provides a viable alternative. In this work we propose domain adaptation to enhance the performance of prostate lesion segmentation on VERDICT-MRI utilising diffusion weighted (DW)-MRI data from multi-parametric (mp)-MRI acquisitions. Experimental results show that domain adaptation significantly improves the segmentation performance on VERDICT-MRI. |
| 3485 | Rapid 3D Navigator-Triggered MRCP With SPACE Sequence at 3T: Only One-third Acquisition Time of Conventional 3D Navigator-Triggered MRCP. | |
| Zhiyong Chen1, bin sun1, yunjing xue1, zhongshuai zhang2, and guijin li3 | ||
1Radiology, Union Hospital, Fujian Medical University, Fuzhou, China, 2Diagnostic imaging, Siemens Healthcare, Shanghai, China., shanghai, China, 3MR application, Siemens Healthineers Ltd，Guangzhou，China, guangzhou, China |
||
The SPACE sequence based rapid MRCP protocol proposed in this study, which reduced acquisition time without deteriorating the image quality, yielded significantly higher overall image quality and better visualization of the pancreaticobiliary tree compared with the conventional MRCP. On the basis of our findings, we suggest that the rapid 3D SPACE technique could improve the clinical throughput of MRCP and show a trend toward wider clinical availability for MRCP studies. |
| 3487 | Comparing the performance of optimized support vector machine model in predicting efficacy of chemo-radiotherapy for advanced rectal cancer with and without a paired-difference up-sampling strategy under small sample size | |
| Jie Kuang1, QingLei Shi2, Gaofeng Shi1, Xu Yan3, and LI Yang1 | ||
1The Fourth Hospital of Hebei Medical University, Shijiazhuang, China, 2Siemens Healthcare, MR Scientific Marketing, Beijing, China, 3Siemens Healthcare, MR Scientific Marketing, Shanghai, China |
||
In this study, we evaluated the effect of paired-difference analysis (PDA) up-sampling strategy on the performance of optimized support vector machine model (SVM) in predicting efficacy of chemo-radiotherapy for advanced rectal cancer. A higher accuracy and robustness was gained for the model adopted the PDA method in predicting the efficacy of treatment, which means that the PDA method can be used as an up-sampling strategy in improving the performance of some machine learning models. |
| 3488 | Predicting the efficacy of neoadjuvant chemo-radiotherapy for advanced rectal cancer using random forest model and a paired-difference up-sampling strategy under small sample size | |
| Jie Kuang1, QingLei Shi2, Gaofeng Shi1, Xu Yan3, and LI Yang1 | ||
1The Fourth Hospital of Hebei Medical University, Shijiazhuang, China, 2Siemens Healthcare, MR Scientific Marketing, Beijing, China, 3Siemens Healthcare, MR Scientific Marketing, Shanghai, China |
||
In this study, we adopted a paired-difference strategy, which can improve training efficiency of random forest (RF) model with a small sample size. Through optimizing in normalization, dimensional reduction, and features selection steps, a higher accuracy was achieved in predicting the efficacy of chemo-radiotherapy for advanced rectal cancer. |
| 3489 | Predicting the efficacy of chemo-radiotherapy for advanced rectal cancer using support vector machine model and a paired-difference up-sampling strategy under small sample size | |
| Jie Kuang1, Xu Yan2, Gaofeng Shi1, QingLei Shi3, and LI Yang1 | ||
1The Fourth Hospital of Hebei Medical University, Shijiazhuang, China, 2Siemens Healthcare, MR Scientific Marketing, Shanghai, China, 3Siemens Healthcare, MR Scientific Marketing, Beijing, China |
||
In this study, we proposed a paired-difference analysis (PDA) method to up-sample the training data, which can improve training efficiency of support vector machine (SVM) model with a small sample size. Through optimizing in normalization, dimensional reduction, and features selection steps, a high accuracy was achieved in predicting the efficacy of chemo-radiotherapy for advanced rectal cancer, which means that valuable information may be provided and showed in this clinical situation. |
| 3490 | Low-rank based compressed sesning in low-field MRI for stroke | |
| Fangge Chen1, Zheng Xu1, Yucheng He1, and Liang Tan2 | ||
1State Key Laboratory of Power Transmission Equipment and System Security and New Technology, Chongqing University, Chongqing, China, 2Department of Neurosurgery, Southwest Hospital, Chongqing, China |
||
While being of great worth in convenience and timely scanning for severe disease, low-field and accessible MRI suffer from long scanning time. To speed up the scanning process in low-field MRI, a low-rank based compressed sensing method with a non-convex reconstruction model is proposed and solved by weighted SVT and gradient descent method iteratively. The in vivo data acquired from a Hemorrhage patient at a 0.05T MRI scanner is used for simulation. The reconstructed image (20% sampling rate) reveals the same hemorrhage shape as CT image shows, demonstrating the ability of compressed sensing applied in low-field MRI cliniclly. |
| 3491 | Deep Learning based Accelerated MR Image Reconstruction via Transfer Learning | |
| Madiha Arshad1, Mahmood Qureshi1, and Hammad Omer1 | ||
1Department of Electrical and Computer Engineering, COMSATS University, Islamabad, Pakistan |
||
In MRI, many deep learning-based solutions often degrade when deployed in different clinical scenarios due to lack of large training datasets. Moreover, the knowledge about the reconstruction problem is constrained to the data seen during training. This paper presents a transfer learning approach to deal with the problems of data scarcity and differences in the source and target domain while reconstructing the MR images using deep learning. Experimental results show successful domain transfer between the source and target datasets in terms of change in magnetic field strength, anatomy and Acceleration Factor (AF). |
| 3492 | A Convergence Proof of Projected Fast Iterative Soft-thresholding Algorithm for Parallel Magnetic Resonance Imaging | |
| Xinlin Zhang1, Hengfa Lu1, Di Guo2, Lijun Bao1, Feng Huang3, and Xiaobo Qu1 | ||
1Department of Electronic Science, Fujian Provincial Key Laboratory of Plasma and Magnetic Resonance, Xiamen University, Xiamen, China, 2School of Computer and Information Engineering, Fujian Provincial University Key Laboratory of Internet of Things Application Technology, Xiamen University of Technology, Xiamen, China, 3Neusoft Medical System, Shanghai, China |
||
For compressed sensing magnetic resonance imaging, algorithms plays a significant role in sparse reconstruction. The Projected Fast Iterative Soft-thresholding Algorithm (pFISTA), a simple and efficient algorithm solving sparse reconstruction models, has been successfully extended to parallel imaging. However, its convergence criterion still poses as an open question, yielding no guideline for parameter setting that allows faithful results. In this work, we prove the sufficient conditions for the convergence of parallel imaging version of pFISTA. Results on in vivo data demonstrate the validity of the convergence criterion. |
| 3493 | Use of Compressed Sensing to Reduce Scan Time and Breath-holding for cine bSSFP MRI in Children and Young Adults | |
| Nivedita Naresh1, Ladonna Malone1, Takashi Fujiwara1, Emma Hulseberg-Dwyer1, Janet McGee1, Quin Lu2, Mark Twite3, Michael Dimaria4, Brian Fonseca4, Lorna P Browne1, and Alex J Barker1,5 | ||
1Radiology, Children's Hospital Colorado, Aurora, CO, United States, 2Philips Healthcare North America, San Francisco, CA, United States, 3Anesthesiology, Children's Hospital Colorado, Aurora, CO, United States, 4Pediatrics, Children's Hospital Colorado, Aurora, CO, United States, 5Bioengineering, University of Colorado Anschutz Medical Campus, Aurora, CO, United States |
||
In this study, we have a compared a vendor-optimized compressed sensing (CS) accelerated cine MRI technique with the conventional cine MRI technique. There were no significant differences in image quality and cardiac volumes between the two techniques, except left ventricular end-diastolic volume which was significantly lower with the CS technique. The accelerated technique on average reduced scan time by 50%. |
| 3494 | Accelerating Positive Contrast MRI Reconstruction with GPU-based Parallel Chambolle-Pock Algorithm | |
| Fang Cai1, Caiyun Shi1, Jing Cheng1, Guoxi Xie2, Hanwei Chen3, Xin Liu1, Hairong Zheng1, Dong Liang1, and Haifeng Wang1 | ||
1Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China, 2Department of Biomedical Engineering, Guangzhou Medical University, Guangzhou, China, 3Department of Radiology, Guangzhou Panyu Central Hospital, Guangzhou, China |
||
Positive contrast Magnetic Resonance Imaging (MRI) based on the susceptibility mapping requires very fast reconstruction for clinical applications. Modern graphics processing units (GPUs) are very efficient at manipulating computer graphics and image processing. And Chambolle-Pock(CP) algorithm is also an efficient algorithm to solve the minimization problem. To further reduce the reconstruction time of positive contrast MRI, a GPU-based parallel CP algorithm was proposed. The experimental results showed that the proposed GPU-based parallel CP method could achieve similar image results, and provide a faster reconstruction of up to 15 times than the conventional CPU-based CP method. |
| 3495 | Weakly Supervised Deep Prior Learning for Multi-coil MRI Reconstruction | |
| Haoyun Liang1, Taohui Xiao1, Chuyu Rong1, Yu Gong1, Cheng Li1, and Shanshan Wang1 | ||
1Paul C. Lauterbur Research Center for Biomedical Imaging, SIAT, Chinese Academy of Sciences, Shenzhen, P.R.China, Shenzhen, China |
||
MRI reconstruction based on supervised learning methods, have achieved remarkable success. However, because of the difficulty and high cost of the MR images collection process, it is not always easy to obtain a big dataset with strong supervised information. Therefore, weakly supervised learning will be a possible solution. In this work, we proposed a weakly deep prior learning algorithm to train a complex UNet for multi-coil MR images reconstruction without very precise labels. The result shows that our proposed algorithm can provide a competitive performance compared to the classical methods with enoucraging quantitative indicators of SSIM and PSNR. |
| 3496 | Imaging biomarker to extract features from soft tissue tumors | |
| Kazuya Kishi1, Daisuke Yoshimaru2, Yasuwo Ide3, and Keito Saitou4 | ||
1Department of radiology, Chiba medical center, Chiba, Japan, 2RIKEN Center for Brain Science, Wako, Japan, 3Department of radiology, Chiba Central Medical Center, Chiba, Japan, 4Department of Medical Technology, Tokyo Women's Medical University Yachiyo Medical Center, Yachiyo, Japan |
||
In order to evaluate the pathological alteration in soft tissue tumors, we focused on the color, distribution, and morphology of the images, based on the tumor components. We could classify soft tissue tumors by texture analysis using MRI. T2-weighted images were more useful for classifying these tumors than T1-weighted images. Texture analysis was extremely useful for diagnosis of soft tissue tumors. |
| 3497 | MoPED: Motion Parameter Estimation DenseNet for accelerating retrospective motion correction | |
| Julian Hossbach1,2,3, Daniel Nicolas Splitthoff2, Stephen Farman Cauley4, and Andreas Maier1 | ||
1Pattern Recognition Lab, Friedrich-Alexander-University Erlangen-Nuremberg, Erlangen, Germany, 2Siemens Healthcare GmbH, Erlangen, Germany, 3Erlangen Graduate School in Advanced Optical Technologies, Erlangen, Germany, 4Martinos Center for Biomedical Imaging, Charlestown, MA, United States |
||
We exploit the data redundancy and the locality of motion in k-space for an estimation of the motion parameters using a Deep Learning approach. The exploratory Motion Parameter Estimation DenseNet (MoPED) extracts the in-plane motion parameters between echo trains of a TSE sequence. As input, the network receives the center patch of the k-space from multiple coils; the network’s output can serve multiple purposes. While an image rejection/reacquisition can be triggered by the motion guess, we show that motion aware reconstruction can be accelerated using MoPED. |
| 3498 | Image Reconstruction with Low-rankness and Self-consistency of k-space Data in Parallel MRI | |
| Xinlin Zhang1, Di Guo2, Yiman Huang1, Ying Chen1, Liansheng Wang3, Feng Huang4, and Xiaobo Qu1 | ||
1Department of Electronic Science, Fujian Provincial Key Laboratory of Plasma and Magnetic Resonance, Xiamen University, Xiamen, China, 2School of Computer and Information Engineering, Fujian Provincial University Key Laboratory of Internet of Things Application Technology, Xiamen University of Technology, Xiamen, China, 3Department of Computer Science, School of Information Science and Engineering, Xiamen University, Xiamen, China, 4Neusoft Medical System, Shanghai, China |
||
Recent low-rank reconstruction methods offer encouraging image reconstruction results enabling promising acceleration of parallel magnetic resonance imaging, however, they were not originally designed to exploit the routinely acquired calibration data for performance improvement in parallel magnetic resonance imaging. In this work, we proposed an image reconstruction approach to simultaneously explore the low-rankness of the k-space data and mine the data correlation among multiple receiver coils with the use of the calibration data. The proposed method outperforms the state-of-the-art methods in terms of suppressing artifacts and achieving lowest error, and exhibits robust reconstructions even with limited auto-calibration signals. |
| 3499 | Free breath T2*-mapping for planning and monitoring the chelation therapy in children with secondary iron overload | |
| Evelina Nazarova1, Galina Tereshchenko1, and Dmitry Kupriyanov2 | ||
1Radiology, Dmitry Rogachev National Medical Research Center Of Pediatric Hematology, Oncology and Immunology, Moscow, Russian Federation, 2Philips Healthcare, Imaging Systems, Moscow, Russian Federation |
||
Iron overload represents critical health problem in children with regular transfusional therapy. Breath hold T2* mapping is widely used for the iron quantification, but it is not always feasible in pre-school children without the sedation. We purpose the free breath T2* mapping for the quick monitoring of treatment response to the chelation therapy in pediatric patients.The purposed method demonstrate that free breath T2* mapping is more convenient for the iron quantification in children according to lack of limited by breath hold technique and provides much higher quality images. |
| 3500 | Continuous T2 Distribution Analysis using Artificial Neural Networks | |
| Tristhal Parasram1, Rebecca Daoud1, and Dan Xiao1 | ||
1University of Windsor, Windsor, ON, Canada |
||
Quantitative analysis of T2 relaxation times could reveal molecular scale information, and has significance in the study of brain, spinal cord, articular cartilage, and cancer discrimination. However, it is non-trivial to recover the relaxation times from MR signals, especially as a continuous distribution, since it is an intrinsically ill-posed problem. In this work, ANNs have been trained to generate the T2 distribution spectra. The performance was evaluated across a large parameter range. In addition to superior computation speed, higher accuracy was achieved compared to the traditional method. |
| 3501 | VIOLIN: improVIng mOdel-based deep learning reconstruction by reLaxing varIable combiNations | |
| Jing Cheng1, Yiling Liu1, Qiegen Liu2, Ziwen Ke1, Haifeng Wang1, Yanjie Zhu1, Leslie Ying3, Xin Liu1, Hairong Zheng1, and Dong Liang1 | ||
1Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China, 2Nanchang University, Nanchang, China, 3University at Buffalo, The State University of New York, Buffalo, Buffalo, NY, United States |
||
Most of the unrolling-based deep learning fast MR imaging methods learn the parameters and regularization functions with the network architecture structured by the corresponding optimization algorithm. In this work, we introduce an effective strategy, VIOLIN and use the primal dual hybrid gradient (PDHG) algorithm as an example to demonstrate improved performance of the unrolled networks via breaking the variable combinations in the algorithm. Experiments on in vivo MR data demonstrate that the proposed strategy achieves superior reconstructions from highly undersampled k-space data. |
| 3502 | Improving Dynamic Paralleling Imaging Reconstruction using Multi-Dimensional Integration (MDI) | |
| Jingyuan Lyu1, Yongquan Ye1, Sen Jia2, Zheng Shi3, Zhongqi Zhang4, Lele Zhao4, Jian Xu1, and Rui Yang3 | ||
1UIH America, Inc., Houston, TX, United States, 2Paul C. Lauterbur Research Centre for Biomedical Imaging, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China, 3Henan Chest Hospital, Zhengzhou, China, 4United Imaging Healthcare, Shanghai, China |
||
We have developed and demonstrated a novel strategy for accurate estimation of coil sensitivity profiles for time-resolved data. Instead of direct temporal average to reduce noise in ACS, the proposed multi-dimensional integrated (MDI) strategy solves the least square problem for coil sensitivity profiles estimation in a new fashion. By extracting the overall calibration equations in all temporal dimensions, MDI is able to calculate coil sensitivity profiles more accurately than without this strategy. It can also be integrated with other parallel imaging methods such as SENSE and ESPIRiT. |
| 3503 | Deep Convolutional Neural Network model for sub-Anatomy specific Landmark detection on Brain MRI | |
| Sumit Sharma1 and Srinivasa Rao Kundeti1 | ||
1Philips Healthcare, Bangalore, India |
||
A Deep CNN (D-CNN) model for Brain sub-anatomy landmark detection for auto MRI scan planning. We compare D-CNN with traditional approaches like segmentation followed by image processing (AL-Net). D-CNN shows better landmark detection with Average RMSE <= 6mm (N=100) compared to AL-Net. |
| 3504 | Evaluation of point-spread-function and signal-to-noise ratio in highly-accelerated Compressed Sensing-SENSE (CS-SENSE) and SENSE MRI | |
| Di Cao1,2,3, Adrian G.Paez1,2, Xinyuan Miao1,2, Dapeng Liu1,2, Chunming Gu1,2,3, and Jun Hua1,2 | ||
1F.M. Kirby Research Center for Functional Brain Imaging, Kennedy Krieger Institute, Baltimore, MD, United States, 2The Russell H. Morgan Department of Radiology and Radiological Sciences, Johns Hopkins University School of Medicine, Baltimore, MD, United States, 3Department of Biomedical Engineering, Johns Hopkins University School of Medicine, Baltimore, MD, United States |
||
We seek to experimentally evaluate the point-spread-function (PSF) and signal-to-noise ratio (SNR) in sequences accelerated using conventional SENSE and the recently developed Compressed Sensing-SENSE (CS-SENSE) with acceleration-factors (R) ranging from 0 to 28. Both CS-SENSE and SENSE had little effect on the PSF in the tested 3D turbo-spin-echo (TSE) sequences. CS-SENSE showed preserved SNR-per-unit-time even when R=28 (compared to R=0), while SENSE reduced SNR-per-unit-time significantly when R≥4. Fold-over artifacts were seen on SENSE images with R≥8, but not on CS-SENSE images for R=0-28. Overall, CS-SENSE seems to show clear advantages compared to SENSE, especially with high acceleration-factors. |
| 3505 | Integration of Open Source Pulse Sequence Programming Toolbox, Pulseq, for use with System Agnostic Broadband Spectrometer | |
| Courtney Bauer1 and Steven M. Wright1 | ||
1Electrical and Computer Engineering, Texas A&M University, College Station, TX, United States |
||
Introduced is a new translator for the open source pulse sequence programming toolbox, Pulseq. Unique from previous translator units developed, this translator takes the Pulseq output file and generates a time-bin based, delimited pulse sequence table. This format is specific to the target system, a homebuilt broadband spectrometer. Our lab uses multiple scanners, and using Pulseq enables seamless transition between a conventional research scanner, a Varian Unity Inova or a prototype homebuilt spectrometer. Ultimately, this translator is not only a useful tool, but it introduces the possibility of facilitating a direct comparison of homebuilt systems with their commercial counterparts. |
| 3506 | Enhanced Deep-learning-based Magnetic Resonance Image Reconstruction using Subjects’ Previous Scans | |
| Roberto Souza1, Youssef Beauferris1, Wallace Loos1, Mariana Bento1, Robert Marc Lebel2, and Richard Frayne1 | ||
1University of Calgary, Calgary, AB, Canada, 2GE, Calgary, AB, Canada |
||
Magnetic resonance (MR) compressed sensing reconstruction explores image sparsity to make MR acquisition faster while still reconstructing high quality images. Modern picture archiving and communication systems allow efficient access to previous scans acquired of the same subject. In this work, we propose to use previous scans to enhance the reconstruction of follow-up scans using a deep learning model. Our model is composed of a reconstruction network that outputs an initial MR reconstruction, which is used as input to an enhancement network along with a co-registered previous scan. Our enhancement network improved quantitative metrics on average by 15%. |
| 3507 | Joint Reconstruction of Multiparameter MRI Images for Imaging Intratumoral Subpopulations | |
| Shraddha Pandey1,2, Arthur David Snider1, Wilfrido Moreno1, and Natarajan Raghunand2 | ||
1Electrical Engineering, University of South Florida, Tampa, FL, United States, 2Cancer Physiology, Moffitt Cancer Research Center, Tampa, FL, United States |
||
Multispectral analysis of multiparametric MRI (mpMRI) images is being explored as an approach to objectively quantify intratumoral heterogeneity. However, acquisition of multiple co-registered parameters maps such as T1, T2, T2* and ADC for this task can be time-consuming. We propose a joint reconstruction method that exploits shared structural information between co-registered MRI parameter maps that provides high similarity between multispectral clusters defined on parameter maps from undersampled data and the ground truth fully-sampled maps of M0, T1, T2 and T2*. |
| 3508 | A Cascaded 3D CNN Approach for Thalamic Nuclei Segmentation | |
| Lavanya Umapathy1, Mahesh Bharath Keerthivasan2,3, Natalie M Zahr4, and Manojkumar Saranathan1,2,5 | ||
1Department of Electrical and Computer Engineering, University of Arizona, Tucson, AZ, United States, 2Department of Medical Imaging, University of Arizona, Tucson, AZ, United States, 3Siemens Healthcare USA, Tucson, AZ, United States, 4Department of Psychiatry & Behavioral Sciences, Stanford University, Menlo Park, CA, United States, 5Department of Biomedical Engineering, University of Arizona, Tucson, AZ, United States |
||
We propose the use of conventional MPRAGE images to synthesize WMn-MPRAGE images for thalamic nuclei segmentation. We compare thalamic nuclei segmentation performance on the synthesized WMn images to those directly on CSFn-MPRAGE. We also validate the clinical utility of our method by analyzing differences in thalamic nuclei volumes between patients with alcohol use disorder (AUD) and age-matched healthy controls using conventional MPRAGE data. |
| 3509 | Automatic segmentation of dentate nuclei for microstructure assessment: application to temporal lobe epilepsy patients | |
| Marta Gaviraghi1, Giovanni Savini2, Gloria Castellazzi1,3, Nicolò Rolandi4, Simone Sacco5,6, Egidio D’Angelo4,7, Fulvia Palesi4, Paolo Vitali2, and Claudia A.M. Gandini Wheeler-Kingshott2,4,8 | ||
1Department of Electrical, Computer and Biomedical Engineering, University of Pavia, Pavia, Italy, 2Neuroradiology Unit, Brain MRI 3T Research Center, IRCCS Mondino Foundation, Pavia, Italy, 33Queen Square MS Centre, Department of Neuroinflammation, UCL Queen Square Institute of Neurology, Faculty of Brain Sciences, University College London, London, United Kingdom, 4Department of Brain and Behavioral Sciences, University of Pavia, Pavia, Italy, 5UCSF Weill Institute for Neurosciences, Department of Neurology, University of California, San Francisco, CA, United States, 6Department of Clinical Surgical Diagnostic and Pediatric Sciences, University of Pavia, Pavia, Italy, 7Brain Connectivity Center (BCC), IRCCS Mondino Foundation, Pavia, Italy, 8Brain MRI 3T Research Center, IRCCS Mondino Foundation, Pavia, Italy |
||
Dentate nuclei (DN) segmentation is necessary for assessing whether DN are affected by pathologies through quantitative analysis of parameter maps, e.g. calculated from diffusion weighted imaging (DWI). This study developed a fully automated segmentation method using non-DWI (b0) images. A Convolution Neural Network was optimised on heathy subjects’ data with high spatial resolution and was used to segment the DN of Temporal Lobe Epilepsy (TLE) patients, using standard DWI. Statistical comparison of microstructural metrics from DWI analysis, as well as volumes of each DN, revealed altered and lateralised changes in TLE patients compared to healthy controls. |
| 3510 | A deep neural network with convolutional LSTM for brain tumor segmentation in multi-contrast volumetric MRI | |
| Namho Jeong1, Byungjai Kim1, Jongyeon Lee1, and Hyunwook Park1 | ||
1Korea Advanced Institute of Science and Technology, Daejeon, Republic of Korea |
||
A medical image segmentation method is a key step in contouring of designs for radiotherapy planning and has been widely studied. In this work, we propose a method using inter-slice contexts to distinguish small objects such as tumor tissues in 3D volumetric MR images by adding recurrent neural network layers to existing 2D convolutional neural networks. It is necessary to apply a convolutional long-short term memory (ConvLSTM) since 3D volumetric data can be considered as a sequence of 2D slices. We verified through the analysis that the correlation between neighboring segmentation maps and the overall segmentation performance was improved. |
| 3511 | CNN-based segmentation of vessel lumen and wall in carotid arteries from T1-weighted MRI | |
| Lilli Kaufhold1,2, Axel J. Krafft3, Christoph Strecker4, Markus Huellebrand1,2, Ute Ludwig3, Jürgen Hennig3, Andreas Harloff4, and Anja Hennemuth1,2 | ||
1Charité - Universitätsmedizin Berlin, Berlin, Germany, 2Fraunhofer MEVIS, Bremen, Germany, 3Department of Radiology, Medical Center - University of Freiburg, Faculty of Medicine, Freiburg, Germany, 4Department of Neurology, Medical Center – University of Freiburg, Faculty of Medicine, Freiburg, Germany |
||
Internal carotid artery stenosis is a major source of ischemic stroke. Multi-contrast MRI can be used for assessing wall characteristics and plaque progression. The quantification of vessel wall morphology requires an accurate segmentation of the vessel wall. To reduce inter- and intra-observer variability, we aim to provide a fully automatic segmentation method. Our approach for segmenting the lumen and vessel wall of the extracranial carotid arteries in T1-weighted 3D MR images is based on a 2D convolutional neural network. Average dice coefficients were 0.947/0.859 for the lumen/vessel wall and the median Hausdorff-distance was below the voxel-size of 0.6mm for both. |
| 3512 | Test-retest repeatability of convolutional neural networks in detecting prostate cancer regions on diffusion weighted imaging in 112 patients | |
| Amogh Hiremath1, Rakesh Shiradkar1, Harri Merisaari1,2, Prateek Prasanna1, Otta Ettala3, Pekka Taimen4, Hannu J Aronen5, Peter J Boström3, Ivan Jambor2,6, and Anant Madabhushi1 | ||
1Department of Biomedical Engineering, Case Western Reserve University, Cleveland, OH, United States, 2Department of Diagnostic Radiology, University of Turku and Turku University Hospital, Turku, Finland, 3Department of Urology, University of Turku and Turku University hospital, Turku, Finland, 4Institute of Biomedicine, University of Turku and Department of Pathology, Turku University Hospital, Turku, Finland, 5Medical Imaging Centre of Southwest Finland, Turku University Hospital, Turku, Finland, 6Department of Radiology, Icahn School of Medicine at Mount Sinai, New York, NY, United States |
||
We evaluated the short-term repeatability of convolutional neural networks (CNNs) in detecting prostate cancer (PCa) using DWI collected from patients who underwent same day test-retest MRI scans. DWI was post-processed using monoexponential fit (ADCm). Two models with similar architecture were trained on test-retest scans and short-term repeatability of network predictions in terms of intra-class correlation coefficient (ICC(3,1)) was evaluated. Although the observed ICC(3,1) was high for CNN when optimized for classification performance, our results suggest that network optimization with respect to classification performance might not yield the best repeatability. Higher repeatability was observed at lower learning rates. |
| 3513 | Automated Quantification of Lung Cysts at 0.55T MRI with Image Synthesis from CT using Deep Learning | |
| Ipshita Bhattacharya1, Marcus Y Chen1, Joel Moss1, Adrienne Campbell-Washburn1, and Hui Xue1 | ||
1National Institutes of Health, Bethesda, MD, United States |
||
We propose a novel machine learning approach for segmentation of lung cystic structures using MRI. Following our recent development on improved structural lung imaging at low-field MRI we use a combination of generative adverserial networks and modified UNet for segmentation of cyst and lung tissues. This provides a non-ionizing radiation free alternative for patients with Lymphangioleiomyomatosis who are evaluated using CT imaging. We employ cross-modality image synthesis and segmentation approaches which work synergistically to take advantage of available CT data. In this work we demonstrate the potential of MRI for quantitative analysis of cystic lung . |
| 3514 | Multi-scale Entity Encoder-decoder Network Learning for Stroke Lesion Segmentation | |
| Hao Yang1, Kehan Qi1, Xin Yu2, Hairong Zheng1, and Shanshan Wang1 | ||
1Paul C. Lauterbur Research Center for Biomedical Imaging, Shenzhen Inst. of Advanced Technology, Shenzhen, China, 2Case Western Reserve University, Cleveland, OH, United States |
||
The encoder-decoder structure have demonstrated encouraging progress in biomedical image segmentation. Nevertheless, there are still many challenges related to the segmentation of stroke lesions, including dealing with diverse lesion locations, variations in lesion scales, and fuzzy lesion boundaries. In order to address these challenges, this paper proposes a deep neural network architecture denoted as the Multi-Scale Deep Fusion Network (MSDF-Net) with Atrous Spatial Pyramid Pooling (ASPP) for the feature extraction at different scales, and the inclusion of capsules to deal with complicated relative entities. Experimental results shows that the proposed model achieved a higher evaluating score compared to 5 models. |
| 3515 | Automated segmentation of midbrain structures using convolutional neural network | |
| Weiwei Zhao1, Fangfang Zhou1, Yida Wang1, Yang Song1, Gaiying Li1, Xu Yan2, Yi Wang3, Guang Yang1, and Jianqi Li1 | ||
1Shanghai Key Laboratory of Magnetic Resonance, School of Physics and Electronic Science, East China Normal University, Shanghai, China, 2MR Collaboration NE Asia, Siemens Healthcare, Shanghai, China, Shanghai, China, 3Department of Radiology, Weill Medical College of Cornell University, New York, NY, United States |
||
Accurate and automated segmentation of substantia nigra (SN), the subthalamic nucleus (STN), and the red nucleus (RN) in quantitative susceptibility mapping (QSM) images has great significance in many neuroimaging studies. In the present study, we present a novel segmentation method by using convolution neural networks (CNN) to produce automated segmentations of the SN, STN, and RN. The model was validated on manual segmentations from 21 healthy subjects. Average Dice scores were 0.82±0.02 for the SN, 0.70±0.07 for the STN and 0.85±0.04 for the RN. |
| 3516 | Multi-Task Learning: Segmentation as an auxiliary task for Survival Prediction of cancer using Deep Learning | |
| José Maria da Silva Moreira1,2, João da Silva Santinha1,2, Thomas Varsavsky3, Carole Sudre3, Jorge Cardoso3, Mário Figueiredo2, and Nickolas Papanikolaou1 | ||
1Computational Clinical Imaging Group, Champalimaud Center for the Unknown, Lisbon, Portugal, 2Instituto de Telecomunicações, Instituto Superior Técnico, Lisbon, Portugal, 3Biomedical Engineering and Imaging Sciences, King's College London, London, United Kingdom |
||
This work presents a new method for multi-task learning that aims to increase the performance of the regression task, using the support of the segmentation task. While requiring further validation to guarantee the increase in performance, the preliminary data of this study suggests that using a ”helper” function might increase performance on the main task. In our study, a better performance of the survival prediction model was observed on the validation set when using the multi-task network, compared to a simpler single-task process. |
| 3517 | Segmenting Brain Tumor Lesion from 3D FLAIR MR Images using Support Vector Machine approach | |
| Virendra Kumar Yadav1, Neha Vats1, Manish Awasthi1, Dinil Sasi1, Mamta Gupta2, Rakesh Kumar Gupta2, Sumeet Agarwal3, and Anup Singh1,4 | ||
1Center for Biomedical Engineering, Indian Institute of Technology, Delhi, India, 2Fortis Memorial Research Institute, Gurugram, India, 3Electrical Engineering, Indian Institute of Technology, Delhi, India, 4Biomedical Engineering, AIIMS, New Delhi, India |
||
Segmentation of brain tumor lesion is important for diagnosis and treatment planning. Tumor tissue and edema usually appears hyperintense on fluid-attenuated-inversion-recovery (FLAIR) MR images. FLAIR images are widely used for brain tumor localization and segmentation purpose. In this study, a Support-Vector-Machine (SVM) model was developed for segmentation of FLAIR hyper-intense region semi-automatically using BraTS 2018 dataset. The proposed approach require a minimal user involvement in selecting one region around tumor in the central slice. It was observed that proposed SVM approach segmentation results shows better dice coefficient in comparison to what reported in literature. |
| 3518 | Anisotropic Deep Learning Multi-planar Automatic Prostate Segmentation | |
| Tabea Riepe1, Matin Hosseinzadeh1, Patrick Brand1, and Henkjan Huisman1 | ||
1Diagnostic Image Analysis Group, Radboudumc, RadboudUMC, Nijmegen, Netherlands |
||
Optimized acquisition of prostate MRI for detection of clinically significant prostate cancer requires automatic prostate segmentation. State of the art automatic prostate segmentation is performed with convolution neural networks (CNNs). Exention of our previously developed anisotropic single plane CNN to handle multi-planar input is expected to decrease segmentation problems caused by the low inter-plane resolution of t2-weighted images. Data preprocessing includes volume alignment, intensity clipping and normalization. Comparing the performance to a similar axial network, the multi-stream model shows a visually relevant improvement in prostate segmentation. |
| 3519 | Simulated CMR images can improve the performance and generalization capability of deep learning-based segmentation algorithms | |
| Yasmina Al Khalil1, Sina Amirrajab1, Cristian Lorenz2, Jürgen Weese2, and Marcel Breeuwer1,3 | ||
1Biomedical Engineering Department, Eindhoven University of Technology, Eindhoven, Netherlands, 2Philips Research Laboratories, Hamburg, Germany, 3Philips Healthcare, MR R&D - Clinical Science, Best, Netherlands |
||
The generalization capability of deep learning-based segmentation algorithms across different sites and vendors, as well as MRI data with high variance in contrast, is limited. This affects the usability of such automated segmentation algorithms in clinical settings. The lack of freely accessible medical datasets additionally limits the development of stable models. In this work, we explore the benefits of adding a simulated dataset, containing realistic contrast variance, into the training procedure of the neural network for one of the most clinically important segmentation tasks, the CMR ventricular cavity segmentation. |
| 3520 | Region of Interest Localization in Large 3D Medical Volumes by Deep Voting | |
| Marc Fischer1,2, Tobias Hepp3, Ulrich Plabst2, Bin Yang2, Mike Notohamiprodjo3, and Fritz Schick1 | ||
1Section on Experimental Radiology, Department of Radiology, University Hospital Tübingen, Tübingen, Germany, 2Institute of Signal Processing and System Theory, University of Stuttgart, Stuttgart, Germany, 3Diagnostic and Interventional Radiology, University Hospital Tübingen, Tübingen, Germany |
||
Identifying Regions of Interest (ROI) such as anatomical landmarks, bounding boxes around organs, certain Field of Views or the selection of a particular body region is of increasing relevance for fully automated analysis pipelines of large cohort imaging data. In this work, a 3D Deep Voting approach based on recent advancements in the field of Deep Learning is proposed, which is able to locate ROIs including single points like anatomical landmarks as well as planes to identify region separators within 3D MRI and CT datasets. |
| 3521 | Multi-Contrast Hippocampal Subfield Segmentation for Ultra-High Field 7T MRI Data using Deep Learning | |
| Daniel Ramsing Lund1,2, Mette Tøttrup Gade1,2, Tina Jensen1,2, Thomas B Shaw2, Maciej Plocharski1, Lasse Riis Østergaard1, Steffen Bollmann2, and Markus Barth2,3 | ||
1Department of Health Science and Technology, Aalborg University, Aalborg, Denmark, 2Centre for Advanced Imaging, University of Queensland, Brisbane, Australia, 3School of Information Technology and Electrical Engineering, University of Queensland, Brisbane, Australia |
||
Ultra-high field 7T MRI and the utilization of multiple MRI contrasts potentially enable a superior hippocampal subfield segmentation. A residual-dense fully convolutional neural network based on U-net, including a dilated-convolutional-block was implemented for hippocampal subfield segmentation. Two data sets were combined for training and mean DSC of 0.7723 was obtained. DSC was higher for larger subfields, which were undersegmented, while smaller subfields were oversegmented. Results were comparable to the atlas-based method ASHS, while providing a substantially faster processing time. |
| 3522 | Automated Segmentation for Myocardial Tissue Phase Mapping Images using Deep Learning | |
| Daming Shen1,2, Ashitha Pathrose2, Justin J Baraboo1,2, Daniel Z Gordon2, Michael J Cuttica3, James C Carr1,2,3, Michael Markl1,2, and Daniel Kim1,2 | ||
1Biomedical Engineering, Northwestern University, Evanston, IL, United States, 2Radiology, Northwestern University Feinberg School of Medicine, Chicago, IL, United States, 3Medicine, Northwestern University Feinberg School of Medicine, Chicago, IL, United States |
||
Tissue phase mapping (TPM) provides regional biventricular myocardial velocities, while the slow manual segmentation process limits it use in clinic. The purpose of this study was to develop a fully automated segmentation method for TPM images with deep learning and explore the optimal method to use the magnitude and phase information. |
| 3523 | A 3D attention model based Recurrent Neural Network for Alzheimer’s Disease Diagnosis | |
| Jie Zhang1,2, Xiaojing Long1, Xin Feng2, and Dong Liang1 | ||
1Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, shenzhen, China, 2Chongqing University of Technology, ChongQing, China |
||
The early diagnosis of AD is important for patient care and disease management. However, early diagnosis of AD is still challenging. In this work, we proposed a 3D attention model based densely connected Convolution Neural Network to learn the multilevel features of MR brain images for AD classification and prediction. The proposed network was constructed with the emphasis on the interior resource utilization and introduced the attention mechanism into the classification of AD for the first time. Our results showed that the proposed model is effective for AD classification. |
| 3524 | Class activation mapping methods for interpreting deep learning models in the classification of MRI with subtypes of multiple sclerosis | |
| Jinseo Lee1, Daniel McClement2, Glen Pridham1, Olayinka Oladosu1, and Yunyan Zhang1 | ||
1University of Calgary, Calgary, AB, Canada, 2University of British Columbia, Vancouver, BC, Canada |
||
As deep learning technologies continue to advance, the availability of reliable methods to accurately interpret these models is critical. Based on a trained deep learning model (VGG19) for image classification, we have shown that methods using class activation mapping (CAM) and Grad-CAM have the potential to detect the most critical MRI feature patterns associated with relapsing remitting and secondary progressive multiple sclerosis, and healthy controls, and that these patterns seem to differentiate the two continuing subtypes of MS. This can help further understand the mechanisms of disease development and discover new biomarkers for clinical use. |
| 3525 | Prediction of prostate cancer aggressiveness using open-source machine learning tools for 5-minute prostate MRI: PRODIF CAD 1.0 | |
| Harri Merisaari1,2,3, Pekka Taimen4, Otto Ettala5, Marko Pesola1, Jani Saunavaara6, Anant Madabhushi3, Peter J Boström5, Hannu Aronen1,6, and Ivan Jambor1,7 | ||
1Department of Diagnostic Radiology, University of Turku, Turku, Finland, 2Department of Future Technologies, University of Turku, Turku, Finland, 3Department of Biomedical Engineering, Case Western Reserve University, Cleveland, OH, United States, 4Department of Pathology, Institute of Biomedicine, Turku, Finland, 5Department of Urology, Turku University Hospital, Turku, Finland, 6Medical Imaging Centre of Southwest Finland, Turku University Hospital, Turku, Finland, 7Department of Radiology, Icahn School of Medicine at Mount Sinai, New York, NY, United States |
||
Acquisition time of a routine prostate MRI can be up to 20-25 minutes leading to significant financial burden on healthcare systems as the number of prostate MRI continue to increase. We developed, validated and tested an open-source radiomics/texture tools for 5-minute biparametric prostate MRI (T2-weighed imaging and DWI obtained using 4 b-values (0, 900, 1100, 2000 s/mm2)) using whole mount prostatectomy sections of 157 men with prostate cancer, PCa, (244 PCa lesions). Best features were corner detectors with AUC (clinically insignificant vs insignificant prostate cancer) in the range of 0.82-0.89. Code and data are available at: https://github.com/haanme/ProstateFeatures and http://mrc.utu.fi/data . |
| 3526 | Differentiation of Breast Cancer Molecular Subtypes on DCE-MRI by Using Convolutional Neural Network with Transfer Learning | |
| Yang Zhang1, Yezhi Lin1,2, Siwa Chan3, Jeon-Hor Chen1,4, Jiejie Zhou2, Daniel Chow1, Peter Chang1, Meihao Wang2, and Min-Ying Su1 | ||
1Department of Radiological Science, University of California, Irvine, CA, United States, 2Department of Radiology, First Affiliate Hospital of Wenzhou Medical University, Wenzhou, China, 3Department of Medical Imaging, Taichung Tzu-Chi Hospital, Taichung, Taiwan, 4Department of Radiology, E-Da Hospital and I-Shou University, Kaohsiung, Taiwan |
||
A total of 244 patients were analyzed, 99 in Training, 83 in Testing-1 and 62 in Testing-2. Patients were classified into 3 molecular subtypes: TN, HER2+ and (HR+/HER2-). Deep learning using CNN and Convolutional Long Short Term Memory (CLSTM) were implemented. The mean accuracy in Training dataset evaluated using 10-fold cross-validation was higher using CLSTM (0.91) than CNN (0.79). When the developed model was applied to testing datasets, the accuracy was very low, 0.4-0.5. When transfer learning was applied to re-tune the model using one testing dataset, it could greatly improve accuracy in the other dataset from 0.4-0.5 to 0.8-0.9. |
| 3527 | A two-stage deep learning method for the identification of rectal cancer lesions in MR images | |
| Jiaxin Li1, Cheng Li2, Xiran Jiang1, Zhenkun Peng2, Chaohe Zhang3, Qiegen Liu4, and Shanshan Wang2 | ||
1China Medical University, Shenyang, China, 2Paul C. Lauterbur Research Center for Biomedical Imaging, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China, 3Cancer Hospital of China Medical University, Shenyang, China, 4Department of Electronic Information Engineering, Nanchang University, Nanchang, China |
||
End-to-end deep learning methods, such as the well-known U-Net, have achieved great successes in biomedical image segmentation tasks. These models are often fed with the full field of view images which may contain irrelevant organs or tissues influencing the segmentation performance. In this study, targeting at the accurate segmentation of rectal cancer lesions in T1-weighted MR images, we propose a two-stage deep learning method that is composed of a detection stage and a segmentation stage. Experimental results show that under the guidance of the detected bounding boxes, better segmentation performance is achieved. |
| 3528 | A Deep-Learning Based 3D Liver Motion Prediction for MR-guided-Radiotherapy | |
| Yihang Zhou1, Jing Yuan1, Oi Lei Wong1, Kin Yin Cheung1, and Siu Ki Yu1 | ||
1Medical Physics & Research Department, Hong Kong Sanatorium & Hospital, Hong Kong, China |
||
Respiratory induced organ motion reduces radiation delivery accuracy of radiotherapy in thorax and abdomen. MR-guided-radiotherapy (MRgRT) is capable of continuous MRI acquisition during treatment. However, the latency due to MRI acquisition and reconstruction, the detection of tumor position change, and the interaction with multileaf collimator (MLC) have been identified as the major challenges for real-time MRgRT. In this study, we proposed a deep-learning based 3D motion prediction technique to predict liver motion from volumetric dynamic MR images. Our algorithm showed promising results (< 0.3 cm prediction error on average) , suggesting its possibility of real-time motion tracking in the future MRgRT. |
| 3529 | Combined clinical parameters and Magnetic Resonance Images for prostate cancer detection | |
| Yi Zhu1, Rong Wei1, Ge Gao2, Jue Zhang1, and Xiaoying Wang2 | ||
1Peking University, Beijing, China, 2Peking University First Hospital, Beijing, China |
||
In the wake of population aging, prostate cancer has become one of the most important diseases in elderly men. The low specificity in only image-based diagnosis may lead to unnecessary biopsies. Therefore, clinicians need to consider other variables to make diagnosis, such as age, PSA, and prostate volume. In this study we developed a novel 3D CNN model which combined clinical parameters and MR images for differentiating benign and malignant prostate lesions. The area under the receiver operating characteristics (ROC) of our proposed model (0.84) is significantly higher than that of traditional prediction model (0.71, P < 0.001). |
| 3530 | Direct Pathology Detection and Characterization from MR K-Space Data Using Deep Learning | |
| Linfang Xiao1,2, Yilong Liu1,2, Peiheng Zeng1,2, Mengye Lyu1,2, Xiaodong Ma1,2, Alex T. Leong1,2, and Ed X. Wu1,2 | ||
1Laboratory of Biomedical Imaging and Signal Processing, The University of Hong Kong, Hong Kong, China, 2Department of Electrical and Electronic Engineering, The University of Hong Kong, Hong Kong, China |
||
Present MRI diagnosis comprises two steps: (i) reconstruction of multi-slice 2D or 3D images from k-space data; and (ii) pathology identification from images. In this study, we propose a strategy of direct pathology detection and characterization from MR k-space data through deep learning. This concept bypasses the traditional MR image reconstruction prior to pathology diagnosis, and presents an alternative MR diagnostic paradigm that may lead to potentially more powerful new tools for automatic and effective pathology screening, detection and characterization. Our simulation results demonstrated that this image-free strategy could detect brain tumors and their sizes/locations with high sensitivity and specificity. |
| 3531 | Analysis of Deep Learning models for Diagnostic Image Quality Assessment in Magnetic Resonance Imaging. | |
| Jeffrey Ma1, Ukash Nakarmi1, Cedric Yue Sik Kin1, Joseph Y. Cheng1, Christopher Sandino2, Ali B Syed1, Peter Wei1, John M. Pauly2, and Shreyas S Vasanawala1 | ||
1Department of Radiology, Stanford University, Stanford, CA, United States, 2Department of Electrical Engineering, Stanford University, Stanford, CA, United States |
||
In this abstract we investigate deep learning frameworks for medical image quality assessment and automatic classification of diagnostic and non-diagnostic quality images. |
| 3532 | Utility of Stockwell Transform variants for local feature extraction from MR images: evidence from multiple sclerosis lesions | |
| Glen Pridham1, Olayinka Oladosu2, and Yunyan Zhang1 | ||
1Department of Radiology, Department of Clinical Neurosciences, Hotchkiss Brain Institute, University of Calgary, Calgary, AB, Canada, 2Department of Neurosciences, Hotchkiss Brain Institute, University of Calgary, Calgary, AB, Canada |
||
The Stockwell Transform (ST) is an advanced local spectral feature estimator, that is prohibitively large for use in machine learning applications for typical MR images. We compared two memory-efficient variants: the Polar ST (PST) and the Discrete Orthogonal ST (DOST) as feature extraction steps in competing random forest classifiers, built to classify white matter regions-of-interest as: lesion or normal-appearing. The DOST failed to out-perform guessing, whereas the PST: out-performed guessing, and improved the accuracy of an intensity-based random forest, achieving 88.8% accuracy. We conclude that the PST can complement MR intensity, whereas the DOST may not. |
| 3533 | Association of MRI-derived radiomic biomarker with disease-free survival in patients with early-stage cervical cancer | |
| Jin Fang1, Bin Zhang1, Shuo Wang2, and Shuixing Zhang1 | ||
1The First Affiliated Hospital of Jinan University, Guangzhou, China, 2Key Laboratory of Molecular Imaging, Institute of Automation, Chinese Academy of Sciences, Beijing, China |
||
Radiomics is a promising methodology that automatically extracts high-dimensional features from imaging data for supplementary evaluation of prognosis. Herein, we developed radiomic signature based on pretreatment MRI, which can be used as a biomarker for risk stratification for disease-free survival (DFS) in patients with early-stage cervical cancer. This study provides a non-invasive and cost-effective preoperative predictive tool to identify the early stage cervical cancer patients with high possibility of recurrence or metastasis; and they may help to determine whether more intensive observation and aggressive treatment regimens should be administered, aim at assisting clinical treatment and healthcare decisions. |
| 3534 | Weakly Supervised Exclusion of Non-Tumoral Enhancement in Low Volume Dataset for Breast Tumor Segmentation | |
| Michael Liu1,2, Richard Ha1, Yu-Cheng Liu1, Tim Duong2, Terry Button2, Pawas Shukla1, and Sachin Jambawalikar1 | ||
1Radiology, Columbia University, New York, NY, United States, 2Stony Brook University, Stony Brook, NY, United States |
||
Quantitative measures of breast functional tumor volume are important response predictors of breast cancer undergoing chemotherapy. Automated segmentation networks have difficulty excluding non tumoral enhancing structures from their segmentations. Using a small small DCE-MRI dataset with coarse slice level labels to weakly supervised segmentation was able to exclude large portions of non tumor structures. Without manual pixel wise segmentation, our Class activation map based region proposer excluded 67% of non-tumoral voxels in a sagittal slice from downstream segmentation networks while maintaining 94% sensitivity. |
| 3535 | Informed deep convolutional neural networks | |
| R. Marc Lebel1 and Daniel Litwiller2 | ||
1Applications and Workflow, GE Healthcare, Calgary, AB, Canada, 2Applications and Workflow, GE Healthcare, New York, NY, United States |
||
Convolutional neural networks are an emerging tool in medical imaging. Conventional CNNs accept an image as input and return a task-specific output (e.g., a filtered image, a disease probability). Conventional CNNs struggle to generalize or perform poorly when image data alone is insufficient to solve the problem. We propose three ways to incorporate relevant scan information into a CNN. The value of this method is demonstrated on rSOS image denoising, a previously unstable problem. |
| 3536 | Deep PET-Prior: MR-derived Zero-Dose PET Prior for Differential Contrast Enhancement of PET | |
| Abhejit Rajagopal1, Andrew P. Leynes1, Thomas Hope1, and Peder E.Z. Larson1 | ||
1Department of Radiology and Biomedical Imaging, University of California, San Francisco, San Francisco, CA, United States |
||
We introduce a deep neural network scheme for predicting FDG-PET activity from a T1-weighted MR volume. This is useful for creating realistic anatomy-conforming synthetic PET data for prototyping of PET reconstruction algorithms, e.g. from abundant MR-only exam data. While deep networks can learn the average or nominal uptake patterns, in most cases, MR is ultimately incapable of fully predicting PET activity due to fundamental differences in the sensing modalities. We show, however, that these MR-derived “zero-dose” images can aid in differential contrast enhancement and visualization of PET by localizing and highlighting activity uniquely detected by the PET radiotracers. |
| 3537 | USR-Net: A Simple Unsupervised Single-Image Super-Resolution Method for Late Gadolinium Enhancement CMR | |
| Jin Zhu1, Guang Yang2,3, Tom Wong2,3, Raad Mohiaddin2,3, David Firmin2,3, Jennifer Keegan2,3, and Pietro Lio1 | ||
1Department of Computer Science and Technology, University of Cambridge, Cambridge, United Kingdom, 2Cardiovascular Research Centre, Royal Brompton Hospital, London, United Kingdom, 3National Heart and Lung Institute, Imperial College London, London, United Kingdom |
||
Three-dimensional late gadolinium enhanced (LGE) CMR plays an important role in scar tissue detection in patients with atrial fibrillation. Although high spatial resolution and contiguous coverage lead to a better visualization of the thin-walled left atrium and scar tissues, markedly prolonged scanning time is required for spatial resolution improvement. In this paper, we propose a convolutional neural network based unsupervised super-resolution method, namely USR-Net, to increase the apparent spatial resolution of 3D LGE data without increasing the scanning time. Our USR-Net can achieve robust and comparable performance with state-of-the-art supervised methods which require a large amount of additional training images. |
| 3538 | Empirically-Trained Image Contrast Synthesis for Intravascular MRI Endoscopy | |
| Xiaoyang Liu1,2 and Paul Bottomley1,2 | ||
1Department of Electrical and Computer Engineering, Johns Hopkins University, Baltimore, MD, United States, 2MR Research Division, Department of Radiology, Johns Hopkins University, Baltimore, MD, United States |
||
Parameterized image synthesis is desirable for region-specific image contrast and protocol optimization. While this could arguably be performed with quantitative mapping and Bloch equation simulations, it is not well-suited for real-time cine MRI or MRI endoscopy in a real-time interventional application with views changing from frame-to-frame. Here we propose to use empirical data to calibrate signal behavior patterns from a specific MRI endoscopic device and pulse sequences to fit new acquisitions from which extra images and ultimately guide automated local contrast optimization. |
| 3539 | A Multi-Stream GAN Approach for Multi-Contrast MRI Synthesis | |
| Mahmut Yurt1,2, Salman Ul Hassan Dar1,2, Aykut Erdem3, Erkut Erdem3, and Tolga Çukur1,2,4 | ||
1Department of Electrical and Electronics Engineering, Bilkent University, Ankara, Turkey, 2National Magnetic Resonance Research Center, Bilkent University, Ankara, Turkey, 3Department of Computer Engineering, Hacettepe University, Ankara, Turkey, 4Neuroscience Program, Sabuncu Brain Research Center, Bilkent University, Ankara, Turkey |
||
For synthesis of a single target contrast within a multi-contrast MRI protocol, current approaches perform either one-to-one or many-to-one mapping. One-to-one methods take as input a single source contrast and learn representations sensitive to unique features of the given source. Meanwhile, many-to-one methods take as input multiple source contrasts and learn joint representations sensitive to shared features across sources. For enhanced synthesis, we propose a novel multi-stream generative adversarial network model that adaptively integrates information across the sources via multiple one-to-one streams and a many-to-one stream. Demonstrations on neuroimaging datasets indicate superior performance of the proposed method against state-of-the-art methods. |
| 3540 | Super-Resolution with Conditional-GAN for MR Brain Images | |
| Alessandro Sciarra1,2, Max Dünnwald1,3, Hendrik Mattern2, Oliver Speck2,4,5,6, and Steffen Oeltze-Jafra1,4 | ||
1MedDigit, Department of Neurology, Medical Faculty, Otto von Guericke University, Magdeburg, Germany, 2BMMR, Biomedical Magnetic Resonance, Otto von Guericke University, Magdeburg, Germany, 3Faculty of Computer Science, Otto von Guericke University, Magdeburg, Germany, 4Center for Behavioral Brain Sciences (CBBS), Magdeburg, Germany, 5German Center for Neurodegenerative Disease, Magdeburg, Germany, 6Leibniz Institute for Neurobiology, Magdeburg, Germany |
||
In clinical routine acquisitions, the resolution in the slice direction is often worse than the in-plane resolution. Super-resolution techniques can help to retrieve the lack of information. Employing a conditional generative adversarial network (c-GAN), known as pix2pix for T1-w brain images, we were able to reconstruct downsampled images till 10-fold. The neural network was compared with the traditional bivariate interpolation method, and the results show that pix2pix is a valid alternative. |
| 3541 | A Lightweight deep learning network for MR image super-resolution using separable 3D convolutional neural networks | |
| Li Huang1, Xueming Zou1,2,3, and Tao Zhang1,2,3 | ||
1School of Life Science and Technology, University of Electronic Science and Technology of China, Chengdu, China, 2Key Laboratory for Neuroinformation, Ministry of Education, Chengdu, China, 3High Field Magnetic Resonance Brain Imaging Laboratory of Sichuan, Chengdu, China |
||
The existing deep learning networks for MR super-resolution image reconstruction using standard 3D convolutional neural networks typically require a huge amount of parameters and thus excessive computational complexity. This has restricted the development of deeper neural networks for better performance. Here we propose a lightweight separable 3D convolution neural network for MR image super-resolution. Results show that our method can not only greatly reduce the amount of parameters and computational complexity but also improve the performance of image super-resolution. |
| 3542 | QSMResGAN - Dipole inversion for quantitative susceptibilitymapping using conditional Generative Adversarial Networks | |
| Francesco Cognolato1,2, Steffen Bollmann1,3, and Markus Barth1,3 | ||
1The University of Queensland, Brisbane, Australia, 2Technical University of Munich, Munich, Germany, 3ARC Training Centre for Innovation in Biomedical Imaging Technology, The University of Queensland, Brisbane, Australia |
||
In our abstract we present QSMResGAN, a conditional Generative Adversarial Network (cGAN) with a novel architecture for the generator (ResUNet), trained only with simulated data of different shapes to solve the dipole inversion problem for quantitative susceptibility mapping (QSM). The network has been compared with other state-of-the-art QSM methods on the QSM challenge 2.0 and on in vivo data. |
| 3543 | Enhanced One-minute EPImix Brain MRI Scans with distortion correction Based on supervised GAN model | |
| Haining Wei1, Jiang Liu2, Enhao Gong3, Stefan Skare4, and Greg Zaharchuk5 | ||
1Tsinghua University, Beijing, China, 2The Johns Hopkins University, Baltimore, MD, United States, 3Subtle Medical Inc., Menlo Park, CA, United States, 4Karolinska Institutet, Stockholm, Sweden, 5Stanford University, Stanford, CA, United States |
||
EPImix is a one-minute full brain magnetic resonance exam using a multicontrast echo-planar imaging (EPI) sequence, which can generate six contrasts at a time. However, the low resolution and signal-to-noise ratio impeded its application. And EPI distortion makes it harder to improve the quality of imaging. In this study, we applied topup for distortion correction and propose a supervised deep learning model to enhance the EPImix images. We test the GRE, T2, T2FLAIR images on network and analyze the peak signal to noise ratio and structural similarity results. The results suggest that the proposed method can effectively enhance EPImix images. |
| 3544 | Data Augmentation with Conditional Generative Adversarial Networks for Improved Medical Image Segmentation | |
| Gregory Kuling1, Matt Hemsley 1,2, Geoff Klein1, Philip Boyer3, and Marzyeh Ghassemi4 | ||
1Medical Biophysics, University of Toronto, Toronto, ON, Canada, 2Physical Sciences Platform, Sunnybrook Research Institute, Toronto, ON, Canada, 3Institute of Biomaterials and Biomedical Engineering, University of Toronto, Toronto, ON, Canada, 4Computer Science and Medicine, University of Toronto, Toronto, ON, Canada |
||
Performance of machine learning models for medical image segmentation is often hindered by a lack of labeled training data. We present a method for data augmentation wherein additional training examples are synthesized using a conditional generative adversarial network (cGAN) conditioned on a ground truth segmentation mask. The mask is later used as a label during the segmentation task. Using a dataset of N=48 T2-weighted MR volumes of the prostate, our results demonstrate the mean DSC score of a U-Net prostate segmentation model increased from 0.74 to 0.76 when synthetic training images are included with real data. |
| 3545 | Multi-Contrast-Specific Objective Functions for MR Image Deep Learning - Losses for Pixelwise Error, Misregistration, and Local Variance | |
| Hanbyol Jang1, Sewon Kim1, Jinseong Jang1, Young Han Lee2, and Dosik Hwang*1 | ||
1School of Electrical and Electronic Engineering, Yonsei University, Seoul, Republic of Korea, 2Yonsei University College of Medicine, Seoul, Republic of Korea |
||
The goal of this study is to make new contrast image from multiple contrast Magnetic Resonance Image (MRI) using deep learning with loss function specialized for multiple image processing. Our contrast-conversion deep neural network (CC-DNN) is an end-to-end architecture that trains the model to create one image (STIR image) from three images (T1-weighted, T2-weighted, and GRE images). And we propose a new loss function to take into account intensity differences, misregistration, and local intensity variations. |
| 3546 | A Self-Regularized and Over-Determined Deep Network for Cranial Pseudo-CT Generation | |
| Max W.K. Law1, Oilei Wong1, Jing Yuen1, and S.K. Yu1 | ||
1Hong Kong Sanatorium & Hospital, Hong Kong, Hong Kong |
||
This study presented a hyperparameter-free deep network modal for cranial pseudo-CT generation. The model was potentially universal to various scanning machines without the need of network hyperparameter adjustment and could handle testing images from MR- and CT-simulators different from the training data. It is beneficial to perform clinical trial in institutions where multiple MR- and CT-machines are in operations, without supervision by deep learning experts. The proposed model was examined using training and testing datasets acquired from two sets of MR- and CT-simulators, showing promising accuracy, <79 mean-absolute-error and <170 root-mean-squared-error. |
| 3547 | Deep Learning-based Perfusion Parameter Mapping (DL-PPM) with Simulated Microvascular Network Data | |
| Liangdong Zhou1, Jinwei Zhang1,2, Qihao Zhang1,2, Pascal Spincemaille1, Thanh D Nguyen1, Yi Wang1,2, and Liangdong Zhou3 | ||
1Weill Medical School of Cornell University, New York, NY, United States, 2Cornell University, Ithaca, NY, United States, 3Radiology, Weill Medical School of Cornell University, New York, NY, United States |
||
Perfusion parameters, including blood flow (BF), apparent blood velocity (V), blood volume (BV) and arterial transit time (ATT) are useful for the disgnosis of many dieases. Typically, perfusion quantification methods utilize the tracer concentration (ASL, DEC, DSC, etc.) as input and blood flow map as output. We proposed a deep learning-based perfusion parameters mapping (DL-PPM), which uses 4D time-revolved tracer concentration as input and perfusion parameters (BF, V, BV, ATT) as output. We tested the propose method using simulated data and in vivo data in kidney. |
| 3548 | Visualizing intrinsic magnetic resonance imaging (MRI) dataset variations in image-space through Bayesian deep auto-encoding | |
| Andrew P. Leynes1,2, Abhejit Rajagopal1, Valentina Pedoia1,2, and Peder E.Z. Larson1,2 | ||
1Radiology and Biomedical Imaging, University of California San Francisco, San Francisco, CA, United States, 2UC Berkeley - UC San Francisco Joint Graduate Program in Bioengineering, Berkeley and San Francisco, CA, United States |
||
We investigated the use of a Bayesian deep auto-encoder to visualize intrinsic variations within a dataset in image-space. The variations were visualized by calculating a voxel-wise standard deviation over the predictions of the Bayesian deep auto-encoder. The low mutual information that was measured between the MRI and the standard deviation maps suggests that new information is contained in the standard deviation maps. This may be useful in the training of deep learning models for anomaly detection. |
| 3549 | Predicting T2 Maps from Morphological OAI Images with an ROI-Focused GAN | |
| Bragi Sveinsson1,2, Akshay Chaudhari3, Bo Zhu1,2, Neha Koonjoo1,2, and Matthew Rosen1,2 | ||
1Massachusetts General Hospital, Boston, MA, United States, 2Harvard Medical School, Boston, MA, United States, 3Stanford University, Stanford, CA, United States |
||
The osteoarthritis initiative (OAI) performed several morphological MRI scans on both knees of a large patient cohort, but only acquired T2 maps in the right knee of most patients. We train a conditional GAN to use the morphological scans acquired in both knees to predict the T2 map, using the acquired T2 map in the right knee as a training target. Post-training, we apply the network to predict T2 values in the left knee, without an acquired T2 map. |
| 3550 | PET Image Denoising Using Structural MRI with a Dilated Convolutional Neural Network | |
| Mario Serrano-Sosa1, Christine DeLorenzo1,2, and Chuan Huang1,2,3 | ||
1Biomedical Engineering, Stony Brook University, Stony Brook, NY, United States, 2Psychiatry, Stony Brook University, Stony Brook, NY, United States, 3Radiology, Stony Brook University, Stony Brook, NY, United States |
||
We developed a new PET denoising model by utilizing a dilated CNN (dNet) architecture with PET/MRI inputs (dNetPET/MRI) and compared it to three other deep learning models with objective imaging metrics Structural Similarity index (SSIM), Peak signal-to-noise ratio (PSNR) and mean absolute percent error (MAPE). The dNetPET/MRI performed the best across all metrics and performed significantly better than uNetPET/MRI (pSSIM=0.0218, pPSNR=0.0034, pMAPE=0.0305). Also, dNetPET performed significantly better than uNetPET (p<0.001 for all metrics). Trend-level improvements were found across all objective metrics in networks using PET/MRI compared to PET only inputs within similar networks (dNetPET/MRI vs. dNetPET and uNetPET/MRI vs. uNetPET). |
| 3551 | Fetal pose estimation from volumetric MRI using generative adversarial network | |
| Junshen Xu1, Molin Zhang1, Esra Turk2, Polina Golland1,3, P. Ellen Grant2,4, and Elfar Adalsteinsson1,5 | ||
1Department of Electrical Engineering & Computer Science, MIT, Cambridge, MA, United States, 2Fetal-Neonatal Neuroimaging and Developmental Science Center, Boston Children’s Hospital, Boston, MA, United States, 3Computer Science and Artificial Intelligence Laboratory (CSAIL), MIT, Cambridge, MA, United States, 4Harvard Medical School, Boston, MA, United States, 5Institute for Medical Engineering and Science, MIT, Cambridge, MA, United States |
||
Estimating fetal pose from 3D MRI has a wide range of applications including fetal motion tracking and prospective motion correction. Fetal pose estimation is challenging since fetuses may have different orientation and body configuration in utero. In this work, we propose a method for fetal pose estimation from low-resolution 3D EPI MRI using generative adversarial network. Results show that the proposed method produces a more robust estimation of fetal pose and achieves higher accuracy compared with conventional convolution neural network. |
| 3552 | Adversarial Inpainting of Arbitrary shapes in Brain MRI | |
| Karim Armanious1,2, Sherif Abdulatif1, Vijeth Kumar1, Tobias Hepp2, Bin Yang1, and Sergios Gatidis2 | ||
1University of Stuttgart, Stuttgart, Germany, 2University Hospital Tübingen, Tübingen, Germany |
||
MRI suffers from incomplete information and localized deformations due to a manifold of factors. For example, metallic hip and knee replacements yield local deformities in the resultant scans. Other factors include, selective reconstruction of data and limited fields of views. In this work, we propose a new deep generative framework, referred to as IPA-MedGAN, for the inpainting of missing or complete information in brain MR. This framework aims to enhance the performance of further post-processing tasks, such as PET-MRI attenuation correction, segmentation or classification. Quantitative and qualitative comparisons were performed to illustrate the performance of the proposed framework. |
| 3553 | Attention-Based Deep Kidney Segmentation Framework for GFR Prediction | |
| Edgar Rios Piedra1,2, Morteza Mardani1,2, and Shreyas Vasanawala1,2 | ||
1Department of Radiology, Stanford University, Stanford, CA, United States, 2Department of Electrical Engineering, Stanford University, Stanford, CA, United States |
||
Automated segmentation of kidneys and their sub-components is a challenging problem, particularly in pediatric patients and in the presence of anatomical deformations or pathology. We present an improved segmentation framework using a multi-channel U-Net with added attention block that allows for the automated segmentation of the multi-phase DCE-MRI of kidneys as well as a functional evaluation of the glomerular filtration rate. Results achieve an average Dice similarity coefficient of 0.912, 0.853, and 0.917 for kidney cortex, medulla, and collector system, respectively. |
| 3554 | Segmentation of Contrast Enhancing Brain Tumor Region using a Machine Learning Framework based upon Pre and Post contrast MR Images | |
| Neha Vats1, Virendra Kumar Yadav1, Manish Awasthi1, Dinil Sasi1, Mamta Gupta2, Rakesh Kumar Gupta2, and Anup Singh1,3 | ||
1Centre for Biomedical Engineering, Indian Institute of Technology, Delhi, New Delhi, India, 2Department of Radiology, Fortis Memorial Research Institute, Gurugram, India, 3Biomedical Engineering, AIIMS, New Delhi, India |
||
Segmentation of contrast enhancing tumor region from post-contrast T1-W MR images is sometime difficult due to low enhancement or presence of infarct tissue around or inside tumor, which exhibits similar intensity as contrast enhancement. Relative difference map obtained from pre-and post-contrast T1-weighted images can increase sensitivity to enhancement visualization as well as clearly differentiate infarct tissue from enhancing lesion. The objective of the current study was to evaluate accuracy of segmentation of contrast enhancing lesion using Support Vector Machine (SVM) classifier developed on relative difference map intensities. Optimized SVM classifier enabled accurate segmentation of contrast enhancing tumor lesion. |
| 3555 | Attention-based Semantic Segmentation of Thigh Muscle with T1-weighted Magnetic Resonance Imaging | |
| Zihao Tang1, Kain Kyle2, Michael H Barnett2,3, Ché Fornusek4, Weidong Cai1, and Chenyu Wang2,3 | ||
1School of Computer Science, University of Sydney, Sydney, Australia, 2Sydney Neuroimaging Analysis Centre, Sydney, Australia, 3Brain and Mind Centre, University of Sydney, Sydney, Australia, 4Discipline of Exercise and Sport Science, Faculty of Medicine and Health, University of Sydney, Sydney, Australia |
||
Robust and accurate MRI-based thigh muscle segmentation is critical for the study of longitudinal muscle volume change. However, the performance of traditional approaches is limited by morphological variance and often fails to exclude intramuscular fat. We propose a novel end-to-end semantic segmentation framework to automatically generate muscle masks that exclude intramuscular fat using longitudinal T1-weighted MRI scans. The architecture of the proposed U-shaped network follows the encoder-decoder network design with integrated residual blocks and attention gates to enhance performance. The proposed approach achieves a performance comparable with human imaging experts. |
| 3556 | Unsupervised learning for Abdominal MRI Segmentation using 3D Attention W-Net | |
| Dhanunjaya Mitta1, Soumick Chatterjee1,2,3, Oliver Speck2,4,5,6, and Andreas Nürnberger1,3 | ||
1Faculty of Computer Science, Otto von Guericke University, Magdeburg, Germany, 2Department of Biomedical Magnetic Resonance, Otto von Guericke University, Magdeburg, Germany, 3Data & Knowledge Engineering Group, Otto von Guericke University, Magdeburg, Germany, 4Center for Behavioral Sciences, Magdeburg, Germany, 5German Center for Neurodegenerative Disease, Magdeburg, Germany, 6Leibniz Insitute for Neurobiology, Magdeburg, Germany |
||
Image segmentation is a process of dividing an image into multiple coherent regions. Segmentation of biomedical images can assist diagnosis and decision making. Manual segmentation is time consuming and requires expert knowledge. One solution is to segment medical images by using deep neural networks, but traditional supervised approaches need a large amount of manually segmented training data. A possible solution for the above issues is unsupervised medical image segmentation using deep neural networks, which our work tries to solve with our proposed 3D Attention W-Net. |
| 3557 | Model uncertainty for MRI segmentation | |
| Andre Maximo1, Chitresh Bhushan2, Dattesh D. Shanbhag3, Radhika Madhavan2, Desmond Teck Beng yeo2, and Thomas Foo2 | ||
1GE Healthcare, Rio de Janeiro, Brazil, 2GE Research, Niskayuna, NY, United States, 3GE Healthcare, Bengaluru, India |
||
It is common practice to use dropout layers on U-net segmentation deep-learning models, and it is usually desirable to measure uncertainty of a deployed model while inferencing in clinical scenario. We present a method to convert a pre-trained model to a Bayesian model that can estimate uncertainty by posterior distribution of its trained weights. Our method uses both regular dropouts and converted Monte-Carlo dropouts to estimate uncertainty via cosine similarity of fixed and stochastic predictions. It can identify cases differing from training set by assigning high uncertainty and can be used to ask for human intervention with tough cases. |
| 3558 | MRI raw k-space mapped directly to outcomes: A study of deep-learning based segmentation and classification tasks | |
| Hariharan Ravishankar1, Chitresh Bhushan2, Arathi Sreekumari1, and Dattesh D Shanbhag1 | ||
1GE Healthcare, Bangalore, India, 2GE Global Research, Niskayuna, NY, United States |
||
Most of the advancements with deep learning have come from mapping the reconstructed MRl images to outcomes (e.g. tumor segmentation, survival rate, pathology risk map). In this work, we present methods to arrive at critical medical imaging tasks like segmentation, classification directly from raw k-space data without image reconstruction. We specifically demonstrate that from k-space MRI data, we can perform hippocampus segmentation as well as detection of motion affected scans with similar performance to that obtained from imaging data. We also demonstrate that such an approach is more resilient to localized artifacts (e.g signal loss in hippocampus due to metal). |
| 3559 | Automated Segmentation of Late Gadolinium Enhanced CMR with Deep Learning | |
| Daming Shen1,2, Justin J Baraboo1,2, Brandon C Benefield3, Daniel C Lee2,3, Michael Markl1,2, and Daniel Kim1,2 | ||
1Biomedical Engineering, Northwestern University, Evanston, IL, United States, 2Radiology, Northwestern University Feinberg School of Medicine, Chicago, IL, United States, 3Feinberg Cardiovascular Research Institute, Northwestern University Feinberg School of Medicine, Chicago, IL, United States |
||
Late gadolinium enhanced (LGE) CMR is the gold standard test for assessment of myocardial scarring. While quantifying scar volume is helpful to clinical decision making, its lengthy image segmentation time limits its use in practice. The purpose of this study is to enable fully automated LGE image segmentation using deep learning (DL) and explore a more efficient way of using annotation. |
| 3560 | Addressing the need for less MRI sequence dependent DL-based segmentation methods: model generalization to multi-site and multi-scanner data | |
| Yasmina Al Khalil1, Cristian Lorenz2, Jürgen Weese2, and Marcel Breeuwer1,3 | ||
1Biomedical Engineering Department, Eindhoven University of Technology, Eindhoven, Netherlands, 2Philips Research Laboratories, Hamburg, Germany, 3Philips Healthcare, MR R&D - Clinical Science, Best, Netherlands |
||
The versatility of MRI acquisition parameters and sequences can have a substantial impact on the design and performance of medical image segmentation algorithms. Even though recent studies report excellent results of deep-learning (DL) based algorithms for tissue segmentation, their generalization capability and sequence dependence is rarely addressed, while being crucial for inclusion in clinical settings. This study attempts to demonstrate the lack of adaptation of such algorithms to unseen data from different sites and scanners. For this purpose, we use a 3D U-Net trained for brain tumor detection and test it site-wise to evaluate how well generalization can be achieved. |
| 3561 | Whole Knee Cartilage Segmentation using Deep Convolutional Neural Networks for Quantitative 3D UTE Cones Magnetization Transfer Modeling | |
| Yanping Xue1,2, Hyungseok Jang1, Zhenyu Cai1, Hoda Shirazian1, Mei Wu1, Michal Byra1, Yajun Ma1, Eric Y Chang1,3, and Jiang Du1 | ||
1University of California, San Diego, San Diego, CA, United States, 2Beijing Chao-Yang Hospital, Beijing, China, 3VA San Diego Healthcare System, San Diego, CA, United States |
||
The existence of short T2 tissues and high ordered collagen fibers in cartilage render it “invisible” to conventional MR and sensitive to the magic angle effect. Segmentation is the first step to obtain parameters of cartilage, which is often performed manually (time-consuming and variable). Automatic segmentation and providing a biomarker that visualizes both short and long T2 tissues and insensitive to the magic angle effect is desideratum. U-Net is based on CNN to process images. The purpose of this study is to describe and evaluate the pipeline of fully-automatic segmentation of cartilage and extraction of MMF in 3D UTE-Cones-MT modeling. |
| 3562 | Intelligent Knee MRI slice placement by adapting a generalized deep learning framework | |
| Chitresh Bhushan1, Dattesh D Shanbhag2, Andre de Alm Maximo3, Arathi Sreekumari2, Dawei Gui4, Uday Patil2, Brandon Pascual4, Rakesh Mullick2, Teck Beng Desmond Yeo1, and Thomas Foo1 | ||
1GE Global Research, Niskayuna, NY, United States, 2GE Healthcare, Bangalore, India, 3GE Healthcare, Rio de Janeiro, Brazil, 4GE Healthcare, Waukesha, WI, United States |
||
We demonstrate a deep learning-based workflow for intelligent slice placement (ISP) in MR knee imaging: meniscus plane, femoral condyle plane, tibial plane, sagittal plane and ACL plane, based on standard 2D tri-planar localizer images. We leveraged a previously described generalized architecture for ISP planning in brain, with only the training data and plane definitions adapted for knee. The mean absolute distance error between GT plane and predicted plane was < 0.5 mm for all planes except tibial plane (~ 1 mm). The results indicate the generalization of deep-learning ISP framework and its suitability for ISP in any anatomy of interest. |
| 3563 | Fully automatic three dimensional prostate images registration: from T2-weighted-images to diffusion‐weighted images | |
| Rong Wei1, Yi Zhu1, Xiaoying Wang2, Ge Gao2, and Jue Zhang1 | ||
1Peking University, Beijing, China, 2Peking University First Hospital, Beijing, China |
||
In the wake of population aging, prostate cancer has become one of the most important diseases in western elderly men. T2-weighted-images (T2WIs) provides the best depiction of the prostate’s anatomy, and diffusion‐weighted images (DWIs) provides clues of cancer as prostate cancer appears as an area of high signal on DWIs. However, high b-value DWIs is affected by the susceptibility effects. In this paper，we describe and utilize a fully automatic 3D deep neutral network to correct misalignment between T2WIs and DWIs. The average increasement of Dice coefficient of before-and-after registration is 73% and 84%, which reflects the removal of susceptibility effects. |
| 3564 | Landmark detection of fetal pose in volumetric MRI via deep reinforcement learning | |
| Molin Zhang1, Junshen Xu1, Esra Turk2, Borjan Gagoski2,3, P. Ellen Grant2,3, Polina Golland1,4, and Elfar Adalsteinsson1,5 | ||
1Department of Electrical Engineering and Computer Science, Massachusetts Institute of Technology, Cambridge, MA, United States, 2Fetal-Neonatal Neuroimaging and Developmental Science Center, Boston Children’s Hospital, Boston, MA, United States, 3Harvard Medical School, Boston, MA, United States, 4Computer Science and Artificial Intelligence Laboratory (CSAIL), Massachusetts Institute of Technology, Cambridge, MA, United States, 5Institute for Medical Engineering and Science, Massachusetts Institute of Technology, Cambridge, MA, United States |
||
Fetal pose estimation could play an important role in fetal motion tracking or automatic fetal slice prescription by real-time adjustments of the prescribed imaging orientation based on fetal pose and motion patterns. In this abstract, we used a multiple image scale deep reinforcement learning method (DQN) to train an agent finding the target landmark of fetal pose by optimizing searching policy based on landmark features and its surroundings. Under an error tolerance of 15-mm, the detection accuracy reaches 58%. |
| 3565 | Retraining a Deep Learning Model to Detect Cerebral Microbleeds Using Single-Echo Stroke Data | |
| Miller Fawaz1, Saifeng Liu1, David Utriainen1, Sean Sethi1, Zhen Wu1, and E. Mark Haacke1 | ||
1Magnetic Resonance Innovations, Inc., Bingham Farms, MI, United States |
||
Automatic cerebral microbleed detection is attainable with our two step model for many disease states. We attributed previously shown lower performance in stroke data to different scenarios unique to stroke, including asymmetrically prominent cortical veins. We improved our existing pipeline for this detection by retraining the deep learning step of our model using stroke cases both in the acute and subacute stages. The results were improved performance in validation data in stroke cases as well as our previously tested data (multiple diseases). This makes our pipeline a viable and versatile real time automatic microbleed detection procedure. |
| 3566 | Retrofitting a Brain Segmentation Algorithm with Deep Learning Techniques: Validation and Experiments | |
| Punith B Venkategowda1, Asha K Kumaraswamy1,2, Jonas Richiardi3,4,5, Sanjeev Krishnan Thampi1, Tobias Kober3,4,5, Bénédicte Maréchal3,4,5, and Ricardo A. Corredor-Jerez3,4,5 | ||
1Siemens Healthcare Pvt. Ltd., Bangalore, India, 2Vidyavardhaka College of Engineering, Mysuru, India, 3Advanced Clinical Imaging Technology, Siemens Healthcare AG, Lausanne, Switzerland, 4Department of Radiology, Centre Hospitalier Universitaire Vaudois (CHUV) and University of Lausanne (UNIL), Lausanne, Switzerland, 5Signal Processing Laboratory (LTS 5), École Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland |
||
Deep learning techniques have proved their robustness in solving medical image analysis problems. This study proposes a conservative approach to benefit from the use of these methods to incrementally improve the performance of a well-established brain segmentation method. For this purpose, convolutional neural networks are trained to perform a reliable skull-stripping, based on weak labels of the original algorithm. The performance of the new pipeline is evaluated in a large cohort of dementia patients and healthy controls. The results present significant improvements in reproducibility and computation speed, while preserving accuracy and power of discrimination between groups. |
| 3567 | Oriented Object Detection Convolutional Neural Network for Automated Prescription of Oblique MRI Acquisitions | |
| Eugene Ozhinsky1, Valentina Pedoia1, and Sharmila Majumdar1 | ||
1Radiology and Biomedical Imaging, University of California San Francisco, San Francisco, CA, United States |
||
High quality scan prescription that optimally covers the area of interest with scan planes aligned to relevant anatomical structures is crucial for error-free radiologic interpretation. The goal of this project was to develop a machine learning pipeline for oblique scan prescription that could be trained on localizer images and metadata from previously acquired MR exams. To achieve that, we have developed a novel multislice rotational region-based convolutional neural network (MS-R2CNN) architecture and evaluated it on dataset of knee MRI exams. |
| 3568 | Segmenting Tumour Habitats Using Machine Learning and Saturation Transfer Imaging | |
| Wilfred W Lam1, Wendy Oakden1, Elham Karami1,2,3, Margaret M Koletar1, Leedan Murray1, Stanley K Liu1,2,4, Ali Sadeghi-Naini1,2,3,4, and Greg J Stanisz1,2 | ||
1Sunnybrook Research Institute, Toronto, ON, Canada, 2University of Toronto, Toronto, ON, Canada, 3York University, Toronto, ON, Canada, 4Sunnybrook Health Sciences Centre, Toronto, ON, Canada |
||
Saturation transfer-weighted images along with T1 and T2 maps at 7 T for 31 tumour xenografts in mice were used to automatically segment 1) tumour, 2) necrosis/apoptosis, 3) edema, and 4) muscle. Independent component analysis and Gaussian mixture modeling were used to segment these regions. Qualitatively excellent agreement was found between MRI and histopathology. An nine-image subset was identified that resulted in a 96% match in voxel labels compared to those found using the entire 24-image dataset. This subset had positive and negative predictive values of 96% and 97%, respectively, for tumour and 88% and 97%, respectively, for necrosis/apoptosis voxels. |
| 3569 | No more localizers: deep learning based slice prescription directly on calibration scans | |
| Andre de Alm Maximo1, Chitresh Bhushan2, Dawei Gui3, Uday Patil4, and Dattesh D Shanbhag4 | ||
1GE Healthcare, Rio de Janeiro, Brazil, 2GE Global Research, Niskayuna, NY, United States, 3GE Healthcare, Waukesha, WI, United States, 4GE Healthcare, Bangalore, India |
||
In this work, we demonstrate a novel automated MRI scan plane prescription workflow by making use of the pre-scan calibrations scans to generate prescription planes for knee MRI planning. Using large-FOV, low-resolution 3D calibration data, we find the meniscus plane with very-high accuracy (angle error = 0.76, distance error = 0.07 mm). The approach obviates the need to acquire any localizer images with potential benefits: (1) avoiding subsequent retakes for correct planning of plane prescription; (2) reducing total scan time; and (3) easing the MRI scanning experience for both patient and technologist by enabling single push scan. |
| 3570 | No-Reference Assessment of Perceptual Noise Level Defined by Human Calibration and Image Rulers | |
| Ke Lei1, Shreyas S. Vasanawala2, and John M. Pauly1 | ||
1Electrical Engineering, Stanford University, Stanford, CA, United States, 2Radiology, Stanford University, Stanford, CA, United States |
||
We propose accessing the MRI quality, perceptual noise level in particular, during a scan to stop it when the image is good enough. A convolutional neural network is trained to map an image to a perceptual score. The label score for training is a statistical estimation of error standard deviation calibrated with radiologist inputs. Image rulers for different scan types are used in the inference phase to determine a flexible classification threshold. Our proposed training and inference methods achieve a 89% classification accuracy. The same framework can be used to tune the regularization parameter for compressed-sensing reconstructions. |
| 3571 | Whole-brain CBF and BAT mapping in 4 minutes using deep-learning-based, multi-band MR fingerprinting (MRF) ASL | |
| Hongli Fan1,2, Pan Su1, Yang Li1, Peiying Liu1, Jay J. Pillai1,3, and Hanzhang Lu1,2,4 | ||
1The Russell H. Morgan Department of Radiology & Radiological Science, Johns Hopkins School of Medicine, Baltimore, MD, United States, 2Department of Biomedical Engineering, Johns Hopkins School of Medicine, Baltimore, MD, United States, 3Department of Neurosurgery, Johns Hopkins School of Medicine, Baltimore, MD, United States, 4F. M. Kirby Research Center for Functional Brain Imaging, Kennedy Krieger Institute, Baltimore, MD, United States |
||
Arterial-Spin-Labeling (ASL) MRI has not been used widely in clinical practice because of lower SNR and the lack of ability to resolve cerebral-blood-flow (CBF) from bolus-arrival-time (BAT) effects1. MR fingerprinting (MRF) ASL is a recently developed technique which has the potential to provide multiple parameters such as CBF, BAT, T1 and cerebral-blood-volume (CBV) in one single scan2-6. However, it still suffers from low SNR. The present work proposes a multi-band MRF-ASL in combination with deep learning, which can improve the reliability of MRF-ASL parametric maps up to 3-fold and provide whole-brain mapping of CBF and BAT in 4 minutes. |
| 3572 | Deep Learning Magnetic Resonance Fingerprinting for in vivo Brain and Abdominal MRI | |
| Peng Cao1, Di Cui1, Vince Vardhanabhuti1, and Edward S. Hui1 | ||
1Diagnostic Radiology, The University of Hong Kong, Hong Kong, China |
||
We proposed a multi-layer perceptron deep learning method to achieve 100-fold acceleration for MRF quantification. |
| 3573 | Dictionary-based convolutional neural network (CNN) for MR Fingerprinting with highly undersampled data | |
| Yong Chen1,2, Zhenghan Fang3, and Weili Lin1,2 | ||
1Radiology, University of North Carolina at Chapel Hill, Chapel Hill, NC, United States, 2Biomedical Research Imaging Center, University of North Carolina at Chapel Hill, Chapel Hill, NC, United States, 3Biomedical Engineering, University of North Carolina at Chapel Hill, Chapel Hill, NC, United States |
||
In this study, we proposed a framework to generate simulated training dataset to train a convolutional neural network, which can be applied to highly undersampled MR Fingerprinting images to extract quantitative tissue properties. This eliminates the necessity to acquire training dataset from multiple subjects and has the potential to enable wide applications of deep learning techniques in quantitative imaging using MR Fingerprinting. |
| 3574 | 3D MRI Processing Using a Video Domain Transfer Deep Learning Neural Network | |
| Jong Bum Son1, Mark D. Pagel2, and Jingfei Ma1 | ||
1Imaging Physics Department, The University of Texas MD Anderson Cancer Center, Houston, TX, United States, 2Cancer Systems Imaging Department, The University of Texas MD Anderson Cancer Center, Houston, TX, United States |
||
3D deep-learning neural networks can help ensure the slice-to-slice consistency. However, the performance of 3D networks may be degraded due to limited hardware. In this work, we developed a video domain transfer framework for 3D MRI processing to combine benefits of 2D and 3D networks with less graphical processing unit memory demands and slice-by-slice coherent outputs. Our approach consists of first translating “3D MRI images” to “a time-sequence of 2D multi-frame motion pictures,” then applying the video domain transfer to create temporally coherent multi-frame video outputs, and finally translating the output back to compose “spatially consistent volumetric MRI images.” |
| 3575 | Transfer learning framework for knee plane prescription | |
| Radhika Madhavan1, Andre Maximo2, Chitresh Bhushan1, Soumya Ghose1, Dattesh D Shanbhag3, Uday Patil3, Matthew Frick4, Kimberly K Amrami4, Desmond Teck Beng Yeo1, and Thomas K Foo1 | ||
1GE Global Research, Niskayuna, NY, United States, 2GE Healthcare, Rio de Janeiro, Brazil, 3GE Healthcare, Bangalore, India, 4Mayo Clinic, Rochester, MN, United States |
||
On model deployment, ideally deep learning models should be able learn continuously from new data, but data privacy concerns in medical imaging do not allow for ready sharing of training data. Retraining with incremental data generally leads to catastrophic forgetting. In this study, we evaluated the performance of a knee plane prescription model by retraining with incremental data from a new site. Increasing the number of incoming training data sets and transfer learning significantly improved test performance. We suggest that partial retraining and distributed learning frameworks may be more suitable for retraining of incremental data. |
| 3576 | Deep Learning Global Schedule Optimization for Chemical Exchange Saturation Transfer MR Fingerprinting (CEST-MRF) | |
| Or Perlman1, Christian T Farrar1, and Ouri Cohen2 | ||
1Radiology, Athinoula A. Martinos Center for Biomedical Imaging, Charlestown, MA, United States, 2Medical Physics, Memorial Sloan Kettering Cancer Center, New York, NY, United States |
||
Chemical exchange saturation transfer MR fingerprinting (CEST-MRF) enables quantification of multiple tissue parameters. Optimization of the acquisition schedule can improve tissue discrimination and reduce scan times but is highly challenging because of the large number of acquisition and tissue parameters. The goal of this work is to demonstrate a scalable deep learning based global optimization method that provides schedules with improved discrimination. The benefits of our approach are demonstrated in an in vivo mouse tumor model. |
| 3577 | Fetal Brain Motion Estimation to Evaluate Motion-induced Artifacts in HASTE MRI | |
| Yi Xiao1, Muheng Li1, Ruizhi Liao2,3, Tingyin Liu3, Junshen Xu2, Esra Turk4, Borjan Gagoski4,5, Karen Ying1, Polina Golland2,3, P.Ellen Grant4,5, and Elfar Adalsteinsson2,6 | ||
1Department of Engineering Physics, Tsinghua University, Beijing, China, 2Department of Electrical Engineering and Computer Science, Massachusetts Institute of Technology, Cambridge, MA, United States, 3Computer Science and Artificial Intelligence Laboratory (CSAIL), Massachusetts Institute of Technology, Cambridge, MA, United States, 4Fetal-Neonatal Neuroimaging and Developmental Science Center, Boston Children’s Hospital, Boston, MA, United States, 5Harvard Medical School, Boston, MA, United States, 6Institute for Medical Engineering and Science, Massachusetts Institute of Technology, Cambridge, MA, United States |
||
Artifacts generated by severe and unpredictable fetal and maternal movements during MRI limit the success of imaging during pregnancy. Although modern clinical applications use single-shot imaging sequences, such as HASTE, to partially mitigate this problem, inter-slice motion artifacts are unavoidable and their impact on fetal imaging is not fully characterized. In order to analyze this problem, we exploit a large repository of volumetric EPI over long duration across many pregnant women to estimate the artifact load due to inter-slice motion on single-shot fetal brain MRI. |
| 3578 | Magnetic Resonance Elastography Analysis using Convolutional Neural Networks | |
| Bogdan Dzyubak1, Joel P Felmlee1, and Richard L. Ehman1 | ||
1Radiology, Mayo Clinic, Rochester, MN, United States |
||
Magnetic Resonance Elastography (MRE) accurately predicts fibrosis by measuring liver stiffness. The subjectivity in human analysis poses the biggest challenge to stiffness measurement reproducibility, and also complicates the training of a neural network to automate the task. In this work, we present a CNN-based stiffness measurement tool, giving special attention to training and validation in context of reader subjectivity. Compared to an older automated tool used by our institution in a reader-verified workflow, the CNN reduces ROI failure rate by 50%, and has an excellent agreement in measured stiffness with reader-verified target ROIs. |
| 3579 | Rapid phase unwrapping with deep learning for shimming applications | |
| Yuhang Shi1 | ||
1Corporate Research Institute, United Imaging Healthcare, Shanghai, China |
||
The implementation of shimming relies on phase unwrapping techniques to measure the underlying field inhomogeneity. This work presents a simple and effective deep learning based phase unwrapping method to accelerate the field mapping stage for shimming applications. The method significantly reduces the compulational time and shows similar performance compared to the conventional path-following method for shimming applications. |
| 3580 | Deep Inversion Net: A Novel Neural Network Architecture for Rapid, and Accurate T2 Relaxometry Inversion | |
| Jeremy Kim1, Thanh Nguyen2, Pascal Spincemaille2, and Yi Wang2 | ||
1Hunter College High School, New York, NY, United States, 2Weill Cornell Medical College, New York, NY, United States |
||
A novel deep neural network architecture, Deep Inversion Net, and a training scheme is proposed to accurately solve the multi-compartmental T2 relaxometry inverse problem for myelin water imaging in multiple sclerosis. Multiple neural networks communicate their outputs to regularize each other — thus better handling the ill-posed nature of this inverse problem. Results in simulated T2 relaxometry data and patients with demyelination show that Deep Inversion Net outperforms conventional optimization algorithms and other neural network architectures. |
| 3581 | Learned Off-Resonance Correction for Simultaneous Radial 23Na and 1H Acquisitions at 7T | |
| Kirsten Koolstra1, Olga Dergachyova2, Zidan Yu2,3, Andrew Webb1, and Martijn Cloos2,3 | ||
1C.J. Gorter Center for High Field MRI, Department of Radiology, Leiden University Medical Center, Leiden, Netherlands, 2Center for Advanced Imaging Innovation and Research, Department of Radiology, New York University School of Medicine, New York, NY, United States, 3Sackler Institute of Graduate Biomedical Sciences, NYU Langone Health, New York, NY, United States |
||
Simultaneous proton (1H) and sodium (23Na) acquisition can provide important metabolic information. However, proton data may suffer from off-resonance artifacts due to the long dwell time required to obtain sufficient SNR for 23Na. In this work we use center outward and center inward image pairs to train a convolutional neural network that performs an off-resonance correction for the proton data without an additional measured field map. |
| 3582 | Learned Unrolled Optimization for Rapid Computation of Local RF field Enhancement Near Implants. | |
| Peter Stijnman1,2, Cornelis van den Berg1, and Alexander Raaijmakers1,2 | ||
1Computational Imaging, UMC Utrecht, Utrecht, Netherlands, 2Biomedical Engineering, Eindhoven University of Technology, Eindhoven, Netherlands |
||
The RF safety assessment of implants is a computationally demanding task. An acceleration method was presented in [1] where the local field enhancement was determined by sparse matrix inversion. In this work, we show how a model-based deep learning approach for unrolled optimization could significantly reduce the number of iterations required. The benefit of this approach is that traditional minimization is still possible afterwards, combining short computation times with high accuracy. We trained 5 iterations with 10.000 randomly generated implants. The hybrid approach finds a numerically equivalent solution in$$$\,\frac{1}{13}^{th}\,$$$of the traditional method. This approach would enable online RF safety assessment. |
| 3583 | Reconstruction of Quantitative Susceptibility Maps from the Phase of Susceptibility-Weighted Images Using a Deep Neural Network | |
| Jun Li1, Hongjian He1, Yi-Cheng Hsu2, and Jianhui Zhong1,3 | ||
1Center for Brain Imaging Science and Technology, Department of Biomedical Engineering, Key Laboratory for Biomedical Engineering of Ministry of Education, Zhejiang University, Hangzhou, China, 2MR Collaboration, Siemens Healthcare Ltd, Shanghai, China, 3Department of Imaging Sciences, University of Rochester, Rochester, NY, United States |
||
There is a need to obtain quantitative measures of tissue susceptibility in the form of susceptibility-weighted imaging (SWI). In this study, we used a deep neural network to generate QSM maps from SWI high pass (HP)–filtered phase images. Using the QSM maps reconstructed from mGRE data by iLSQR (mGRE iLSQR) as the ground truth, the QSM maps generated from SWI HP-filtered phase images by UNet (SWI UNet) resulted in lower residuals and a better performance in quantitative metrics compared with the QSM maps reconstructed from SWI HP-filtered phase images by iLSQR (SWI iLSQR). |
| 3584 | Quantification of Inter-Muscular Adipose Infiltration in Calf/Thigh MRI using Fully and Weakly Supervised Semantic Segmentation | |
| Rula Amer1, Jannette Nassar1, David Bendahan2, Hayit Greenspan1, and Noam Ben-Eliezer1,3,4 | ||
1Department of Biomedical Engineering, Tel Aviv University, Tel Aviv, Israel, 2CNRS, CRMBM, Aix Marseille University, Marseille, France, 3Center for Advanced Imaging Innovation and Research, New York University, New York, NY, United States, NY, United States, 4Sagol School of Neuroscience, Tel Aviv University, Ramat-Aviv, Israel |
||
Quantification of subcutaneous fat infiltration in diseased muscle regions holds great prognostic value as it helps monitor the progression of muscular dystrophies. To estimate the infiltration, two stages are performed. The first isolates the region of muscle from the thigh/calf anatomy using U-net architecture for fully supervised segmentation. The second stage classifies muscle into diseased/healthy pixels using a weakly supervised segmentation method, incorporating a deep convolutional auto-encoder with triplet loss, and creates two clusters in the embedded feature space using k-means. The results showed a high Dice coefficient and a strong correlation between the fat infiltration and disease severity level. |
| 3585 | Untrained Modified Deep Decoder for Joint Denoising and Parallel Imaging Reconstruction | |
| Sukrit Arora1, Volkert Roeloffs1, and Michael Lustig1 | ||
1UC Berkeley, Berkeley, CA, United States |
||
An untrained deep learning model based on a Deep Decoder was used for image denoising and parallel imaging reconstruction. The flexibility of the modified Deep Decoder to output multiple images was exploited to jointly denoise images from adjacent slices and to reconstruct multi-coil data without pre-determed coil sensitivity profiles. Higher PSNR values were achieved compared to the traditional methods of denoising (BM3D) and image reconstruction (Compressed Sensing). This untrained method is particularly attractive in scenarios where access to training data is limited, and provides a possible alternative to conventional sparsity-based image priors. |
| 3586 | Deep Dixon: Deep learning-based chemical-shift corrected water-fat separation with only simulated training data | |
| Frank Zijlstra1 and Peter R Seevinck1 | ||
1Image Sciences Institute, UMC Utrecht, Utrecht, Netherlands |
||
Deep learning has been successfully applied to Dixon reconstruction, but requires good training data, which limits clinical applicability. We propose a deep learning-based Dixon method with chemical-shift correction that is trained with only simulated data. Results on three anatomies show that the method produces equivalent or better results than conventional methods for Dixon water-fat separation with chemical-shift correction. This approach is fundamentally different from conventional linear and non-linear solvers and shows promise for extension to more complex problems. |
| 3587 | Joint Optimization of Sampling Patterns and Deep Priors for Improved Parallel MRI | |
| Hemant Kumar Aggarwal1 and Mathews Jacob1 | ||
1Electrical and Computer Engineering, University of Iowa, Iowa City, IA, United States |
||
Model-based deep learning (MoDL) frameworks, which combine deep learned priors with imaging physics, are now emerging as powerful alternatives to compressed sensing in a variety of reconstruction problems. In this work, we investigate the impact of sampling patterns on the image quality. We introduce a scheme to jointly optimize the sampling pattern and the reconstruction network parameters in MoDL scheme. Experimental results demonstrate the significant improvement in reconstruction quality with sampling optimization. The results also show that the decoupling between imaging physics and image properties in MoDL offers improved performance over direct inversion scheme in the joint optimization scheme. |
| 3588 | Real-Time Cardiac Cine MRI with Residual Convolutional Recurrent Neural Network | |
| Eric Z. Chen1, Xiao Chen1, Jingyuan Lv2, Yuan Zheng2, Terrence Chen1, Jian Xu2, and Shanhui Sun1 | ||
1United Imaging Intelligence, Cambridge, MA, United States, 2UIH America, Inc., Houston, TX, United States |
||
Real-time cardiac cine MRI does not require ECG gating in the data acquisition and is more useful for patients who can not hold their breaths or have abnormal heart rhythms. However, to achieve fast image acquisition, real-time cine commonly acquires highly undersampled data, which imposes a significant challenge for MRI image reconstruction. We propose a residual convolutional RNN for real-time cardiac cine reconstruction. To the best of our knowledge, this is the first work applying deep learning approach to Cartesian real-time cardiac cine reconstruction. Based on the evaluation from radiologists, our deep learning model shows superior performance than compressed sensing. |
| 3589 | Single-shot T2 mapping improvement through Multi-train Multiple Overlapping-Echo Detachment planar imaging sequence | |
| Xiaoyin Wang1, Qizhi Yang2, Hongjian He1, Congbo Cai2, Yi-Cheng Hsu3, and Jianhui Zhong1,4 | ||
1Center for Brain Imaging Science and Technology, Department of Biomedical Engineering, Key Laboratory for Biomedical Engineering of Ministry of Education, Zhejiang University, Hangzhou, China, 2Department of Electronic Science, Fujian Provincial Key Laboratory of Plasma and Magnetic Resonance, Xiamen University, Xiamen, China, 3MR Collaboration, Siemens Healthcare Ltd., Shanghai, China, 4Department of Imaging Sciences, University of Rochester, Rochester, NY, United States |
||
Magnetic Resonance parametric mapping can provide quantitative information to characterize tissue properties. Recently, a single-shot T2 mapping method based on Multiple Overlapping-Echo Detachment (MOLED) planar imaging was proposed. However, limited echo time ranges still affected the reconstruction accuracy of the T2 values, especially when large T2 value ranges were present. In this abstract, MOLED was expanded through multiple-echo-train acquisitions that achieved high accuracy and better texture. The deep convolution neural network was used to reconstruct T2 maps, B1 maps and spin densities in synchrony. The sequence efficiencies were demonstrated in digital-brain, phantom and human-brain experiments. |
| 3590 | Model-Free Deep MRI Reconstruction: A Robustness Study | |
| Gopal Nataraj1 and Ricardo Otazo1,2 | ||
1Medical Physics, Memorial Sloan Kettering Cancer Center, New York, NY, United States, 2Radiology, Memorial Sloan Kettering Cancer Center, New York, NY, United States |
||
Speed is often claimed as a key advantage of deep learning (DL) for undersampled parallel MRI reconstruction. However, leading DL methods require repeated application of the MR acquisition model and its adjoint, just as in conventional iterative methods. This work investigates the feasibility and robustness of model-free DL reconstruction, which has the potential to be much faster. Results in varied patient cases of increasing pathological rarity demonstrate that while model-free DL can reasonably reconstruct anatomies similar to those seen in training, performance can degrade drastically in more challenging situations. |
| 3591 | Hybrid Deep Neural Network Architectures for Multi-Coil MR Image Reconstruction | |
| Salman Ul Hassan Dar1,2, Mahmut Yurt1,2, Muzaffer Özbey1,2, and Tolga Çukur1,2,3 | ||
1Department of Electrical and Electronics Engineering, Bilkent University, Ankara, Turkey, 2National Magnetic Resonance Research Center (UMRAM), Bilkent University, Ankara, Turkey, 3Neuroscience Program, Aysel Sabuncu Brain Research Center, Bilkent University, Ankara, Turkey |
||
Two main frameworks for |
| 3592 | A Further Analysis of Deep Instability in Image Reconstruction | |
| Yue Guan1, Yudu Li2,3, Yao Li1, Yiping Du1, and Zhi-Pei Liang2,3 | ||
1Institute for Medical Imaging Technology, School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China, 2Department of Electrical and Computer Engineering, University of Illinois at Urbana-Champaign, Urbana, IL, United States, 3Beckman Institute for Advanced Science and Technology, University of Illinois at Urbana-Champaign, Urbana, IL, United States |
||
Deep learning (DL) has emerged as a new tool for solving ill-posed image reconstruction problems and generated a lot of interest in the MRI community. However, image learning is a very high-dimensional problem and deep networks, if not trained properly, would have instability problems. Building upon a recent analysis, we present a further analysis of the instability problems, highlighting: a) the overfitting problem due to limited training data, b) inaccurate density estimation, and c) inadequate sampling from a probability density function. We also present a theoretical analysis of the prediction error based on statistical learning theory. |
| 3593 | A direct MR image reconstruction from k-space via End-To-End reconstruction network using recurrent neural network (ETER-net) | |
| Changheun Oh1, yeji han2, and HyunWook Park1 | ||
1Electrical Engineering, Korea Advanced Institute of Science and Technology, Daejeon, Korea, Republic of, 2Biomedical engineering, Gachon University, Incheon, Korea, Republic of |
||
In this work, we propose a novel neural network architecture named ‘ETER-net’ as a unified solution to reconstruct an MR image directly from k-space data. The proposed image reconstruction network can be applied to k-space data that are acquired with various scanning trajectories and multi or single-channel RF coils. It also can be used for semi-supervised domain adaptation. To evaluate the performance of the proposed method, it was applied to brain MR data obtained from a 3T MRI scanner with Cartesian and radial trajectories. |
| 3594 | A deep network for reconstruction of undersampled fast-spin-echo MR images with suppressed fine-line artifact | |
| Sangtae Ahn1, Anne Menini2, and Christopher J Hardy1 | ||
1GE Global Research, Niskayuna, NY, United States, 2GE Healthcare, Menlo Park, CA, United States |
||
Fine-line artifact is suppressed in Fast Spin Echo (FSE) images reconstructed with a deep-learning network. The network is trained using many examples of fully sampled Nex=2 data. In each case the two excitations are combined to generate fully sampled ground-truth images with no fine-line artifact, which are used for comparison with the generated image in the loss function. However only one of the excitations is retrospectively undersampled and fed into the input of the network during training. In this way the network learns to remove both undersampling and fine-line artifacts. At inferencing, only Nex=1 undersampled data are acquired and reconstructed. |
| 3595 | Exploiting Coarse-Scale Image Features for Transfer Learning in Accelerated Magnetic Resonance Imaging | |
| Ukash Nakarmi1, Joseph Y. Cheng1, Edgar P. Rios1, Morteza Mardani1, John M. Pauly2, and Shreyas S Vasanawala1 | ||
1Department of Radiology, Stanford University, Stanford, CA, United States, 2Department of Electrical Engineering, Stanford University, Stanford, CA, United States |
||
This work investigates coarse-scale image features for transfer learning in accelerated magnetic resonance imaging. The model uses multi-scale unrolled CNN architecture that captures image features at coarse and fine scale to efficiently reduce the training sample size for deep learning model training. |
| 3596 | Deep-learning based motion correction for brain conductivity reconstruction | |
| Jan Hendrik Wuelbern1, Ulrich Katscher1, Karsten Sommer1, Axel Saalbach1, and Jalal B Andre2 | ||
1Philips Research Europe, Hamburg, Germany, 2University of Washington, Seattle, WA, United States |
||
Since tissue conductivity is determined by the numerical second derivative of the phase map, it is particularly susceptible to motion. This abstract investigates the application of deep-learning based methods for retrospective correction of motion artifacts to obtain suitable phase maps as input for conductivity reconstruction. Different types of motion were investigated in the framework of volunteer experiments, revealing that the applied motion correction was indeed capable of improving conductivity reconstruction. |
| 3597 | Calgary-Campinas raw k-space dataset: a benchmark for brain magnetic resonance image reconstruction | |
| Roberto Souza1, M Louis Lauzon1, Marina Salluzzi1, Letícia Rittner2, and Richard Frayne1 | ||
1University of Calgary, Calgary, AB, Canada, 2University of Campinas, Campinas, Brazil |
||
Machine learning is a new frontier for magnetic resonance (MR) image reconstruction, but progress is hampered by a lack of benchmark datasets. Our datasets provides ~200 GB of brain MR data (both raw and reconstructed data) acquired with different acquisition parameters on different scanners from different vendors and different magnetic field intensities. The fastMRI initiative (https://fastmri.org/), also provides raw data but otherwise is complementary. For instance, fastMRI provides raw k-space data corresponding to 2D acquisitions, while our dataset is composed of 3D acquisitions (i.e., with our data, you can under-sample in two directions). |
| 3598 | Attention Based Scale Recurrent Network for Under-Sampled MRI Reconstruction | |
| Gabriel della Maggiora1,2,3, Alberto Di Biase1,2,3, Carlos Castillo-Passi2,3,4, and Pablo Irarrazaval1,2,3,4 | ||
1Electrical Engineering Department, Pontificia Universidad Católica de Chile, Santiago, Chile, 2Biomedical Imaging Center, Pontificia Universidad Católica de Chile, Santiago, Chile, 3Millennium Nucleus for Cardiovascular Magnetic Resonance, Santiago, Chile, 4Institute for Biological and Medical Engineering, Pontificia Universidad Católica de Chile, Santiago, Chile |
||
We propose an Attention Based Scale Recurrent Network for reconstructing under-sampled MRI data. This network is a variation of the recently proposed Scale Recurrent Network for blind deblurring1. We treat the reconstruction problem as a deblurring problem. Thus the under-sampling pattern does not need to be known. We trained and tested our network with the NYU knee dataset available for the fastMRI challenge. The proposed model shows promising results for single-coil reconstruction outperforming both baselines. |
| 3599 | Uncertainty Quantification for Deep MRI | |
| Vineet Edupuganti1, Morteza Mardani1, Joseph Cheng1, Shreyas Vasanawala2, and John Pauly1 | ||
1Electrical Engineering, Stanford University, Stanford, CA, United States, 2Radiology, Stanford University, Stanford, CA, United States |
||
Reliable MRI reconstruction is crucial for accurate diagnosis. However, high resolution imaging leaves substantial uncertainty about the authenticity of the recovered pixels especially when using overparameterized deep learning. Leveraging variational autoencoders (VAEs), this study proposes a Bayesian imaging algorithm that distills the uncertainty in a low-dimensional latent code. One can then simply draw independent samples from the decoder to procure pixel variance maps along with the image. To further quantify the prediction risk of unseen images, we adopt Stein's Unbiased Risk Estimator (SURE), which we find correlates well with the true risk. |
| 3600 | Deep learning-based referenceless distortion correction for single-shot non-Cartesian spatiotemporally encoded MRI | |
| Wei Wang1, Jian Wu1, Qinqin Yang1, Jian Han1, Congbo Cai1, Shuhui Cai1, and Zhong Chen1 | ||
1Xiamen University, Xiamen, China |
||
Non-Cartesian spatiotemporally encoded MRI sequence shows obvious advantages in spatial Non-Cartesian spatiotemporally encoded MRI sequence shows obvious advantages in spatial selectivity and sampling efficiency. However, the resulting image is susceptible to severe distortion due to the cumulative effect of the B0 field inhomogeneity. In this work, this issue is addressed by introducing a deep learning-based method that is specifically tailored to inhomogeneous field correction. Simulation and in vivo rat brain experiments show that our method can effectively correct the image distortion and obtain the field map without extra reference scan. |
| 3601 | Rapid Region-of-Interest MRI Reconstruction Using Context-Aware Non-Local U-Net | |
| Xinwen Liu1, Jing Wang1,2, Fangfang Tang1, Hongfu Sun1, Feng Liu1, and Stuart Crozier1 | ||
1School of Information Technology and Electrical Engineering, the University of Queensland, Brisbane, Australia, 2School of Information and Communication Technology, Griffith University, Brisbane, Australia |
||
In MRI, region-of-interest (ROI) imaging is frequently used in clinical applications. The sub-sampling-based scheme is capable of accelerating the ROI-focused image reconstruction process but degrades the image quality. The degradation could be alleviated by ROI-weighted optimization; however, existing methods mainly focus on the local signal restoration and have no explicit control of the noise from the entire image. In this abstract, we propose to reconstruct the ROI using a non-local U-net method that incorporates contextual information from the whole image. The results show the proposed algorithm improves PSNR and SSIM over conventional methods. |
| 3602 | Multi-contrast MR imaging with enhanced denoising autoencoder prior network learning | |
| Xiangshun Liu1,2, Minghui Zhang1, Qiegen Liu1, Leslie Ying3, Xin Liu2, Hairong Zheng2, and Shanshan Wang2 | ||
1Department of Electronic Information Engineering, Nanchang University, Nanchang, China, 2Paul C. Lauterbur Research Center for Biomedical Imaging, SIAT, Chinese Academy of Sciences, Shenzhen, China, 3Department of Biomedical Engineering and Electrical Engineering, The State University of New York, Buffalo, New York, NY, United States |
||
This paper proposes an enhanced denoising autoencoder prior (EDAEP) network learning method for multi-contrast MR reconstruction using deep learning. Specifically, a multi-model autoencoder with various noise levels was developed to capture different features from multi-contrast images. A weighted aggregation strategy was also adopted to balance the impact of multiple model outputs. These designs empower the network to explore the correlations and similarities among multi-contrast images, handle different acceleration trajectories and avoid a lot of cumbersome retraining. Experimental results demonstrate that our method can improve the quality of reconstructed images compared to other classical methods. |
| 3603 | ODE-based Deep Network for MRI Reconstruction | |
| Ali Pour Yazdanpanah1,2, Onur Afacan1,2, and Simon K. Warfield1,2 | ||
1Harvard Medical School, Boston, MA, United States, 2Boston Children's Hospital, Boston, MA, United States |
||
Fast data acquisition in MRI is vastly in demand and scan time depends on the number of acquired k-space samples. The data-driven methods based on deep networks have resulted in promising improvements, compared to the conventional methods. The connection between deep network and Ordinary Differential Equation (ODE) has been studied recently. Here, we propose an ODE-based deep network for MRI reconstruction to enable the rapid acquisition of MR images with improved image quality. Our results with undersampled data demonstrate that our method can deliver higher quality images in comparison to the reconstruction methods based on the UNet and Residual networks. |
| 3604 | Quantitative characterization of image reconstruction training dataset complexity with Rademacher Complexity measures | |
| Bo Zhu1,2,3, Neha Koonjoo1,2,3, Bragi Sveinsson1,2,3, and Matthew S Rosen1,2,3 | ||
1Department of Radiology, A.A Martinos Center for Biomedical Imaging/MGH, Charlestown, MA, United States, 2Harvard Medical School, Boston, MA, United States, 3Department of Physics, Harvard University, Cambridge, MA, United States |
||
Here we propose to quantitatively measure the data complexity of training datasets using the Rademacher Complexity metric, and demonstrate its effectiveness in analyzing dataset composition and its effect on neural network training for image reconstruction tasks. |
| 3605 | Unsupervised Reconstruction of Continuous Dynamic Radial Acquisitions via CNN-NUFFT Self-Consistency | |
| Matthew Muckley*1, Tullie Murrell*2, Suvrat Booshan2, Hersh Chandarana1, Florian Knoll1, and Daniel K. Sodickson1 | ||
1Radiology, NYU School of Medicine, New York, NY, United States, 2Facebook AI Research, Menlo Park, CA, United States |
||
The introduction of machine learning for medical image reconstruction has opened up new opportunities for reconstruction speed and subsampling; however, acquiring ground truth data is expensive or impossible in the case of dynamic imaging. Here we investigate a technique for optimizing a CNN on continuous radial data by treating the NUFFT-CNN function as an autoencoding deep image prior. Using this method, we are able to reconstruct images that increment over time frames as short as a single spoke. The technique opens up new possibilities for dynamic image reconstruction. |
| 3606 | ISTA-nets: enhancing the performance of the unrolled deep networks for fast MR imaging | |
| Jing Cheng1, Yiling Liu1, Qiegen Liu2, Ziwen Ke1, Haifeng Wang1, Yanjie Zhu1, Leslie Ying3, Xin Liu1, Hairong Zheng1, and Dong Liang1 | ||
1Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China, 2Nanchang University, Nanchang, China, 3University at Buffalo, The State University of New York, Buffalo, Buffalo, NY, United States |
||
We introduce an effective strategy to maximize the potential of deep learning and model-based reconstruction based on the network of ISTA-net, which is the unrolled version of iterative shrinkage-thresholding algorithm for compressed sensing reconstruction. By relaxing the constraints in the reconstruction model and the algorithm, the reconstruction quality is expected to be better. The prior of the to-be-reconstructed image is obtained by the trained networks and the data consistency is also maintained through updating in k-space for the reconstruction. Brain data shows the effectiveness of the proposed strategy. |
| 3607 | A Parameter-free Plug-and-Play Method for Accelerated MRI Reconstruction | |
| Sizhuo Liu1, Ning Jin2, Philip Schniter3, and Rizwan Ahmad1 | ||
1Biomedical Engineering, Ohio State University, Columbus, OH, United States, 2Siemens Medical Solutions Inc, Columbus, OH, United States, 3Electrical and Computer Engineering, Ohio State University, Columbus, OH, United States |
||
The recently proposed Plug-and-Play (PnP) methods provide an avenue to combine physics-driven MR models with sophisticated, learned image models instantiated by image denoising subroutines. The performance of PnP methods, however, is sensitive to changes in the measurement signal-to-noise ratio (SNR) and algorithmic parameters that balance the contributions from the data fidelity and denoising terms. We propose a discrepancy-principle-based scheme that mediates the impact of the denoising subroutine, leading to more consistent performance across different measurement SNRs without manual intervention. For validation, the proposed scheme is applied to cine images collected at 3T, 1.5T, and 0.35T scanners. |
| 3608 | Efficient Phase-varied Image Reconstruction using Single Deep Convolutional Neural Network without Estimation of Phase Distribution. | |
| Shohei OUCHI1 and Satoshi ITO1 | ||
1Utsunomiya University, Utsunomiya, Japan |
||
A novel single image domain learning CNN based reconstruction method for phase-varied images is proposed in which real and imaginary part of complex image are reconstructed independently. Proposed method uses symmetrical sub-sampling which enable reconstruction for real and imaginary part of complex images independently of each other without estimating phase distribution on the image. Reconstruction experiments showed that higher PSNR images are obtained in proposed method compared to phase estimating CNN or ADMM-CSNet. Proposed method is highly practical since it is robust to phase variation and is easy for training because of its simple CNN structure |
| 3609 | Deep Learning to Produce Realistic MR Images through Fréchet Inception Distance Monitoring | |
| Sunghun Seo1, Seung Hong Choi2, and Sung-Hong Park1 | ||
1Department of Bio and Brain Engineering, Korea Advanced Institute of Science and Technology, Daejeon, Korea, Republic of, 2Department of Radiology, Seoul National University College of Medicine, Seoul, Korea, Republic of |
||
It is known that optimizing a deep learning model based on best validation loss achieves best quantitative results in image reconstruction, but resulting images are often blurry. In this study we propose an alternative way of optimization in which convolutional neural network (CNN) is trained beyond best validation loss to produce realistic MR images by monitoring Fréchet Inception Distance. The new approach generated sharper and more realistic images than the conventional optimization, providing a new insight into optimization for MR image reconstruction. |
| 3610 | Calibrationless SENSE Reconstruction with Deep Coil Sensitivity Learning | |
| Chengyan Wang1, Yan Li2, Jun Lv3, Bo Li4, Fei Dai5, Weibo Chen6, and He Wang1,5 | ||
1Human Phenome Institute, Fudan University, Shanghai, China, 2Department of Radiology, Ruijin Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China, 3Department of Computer Science, Yantai University, Yantai, China, 4The Third Affiliated Hospital of Nanchang University, Nanchang, China, 5Institute of Science and Technology for Brain-Inspired Intelligence, Fudan University, Shanghai, China, 6Philips Healthcare, Shanghai, China |
||
Conventional SENSE requires accurate estimation of coil sensitivity maps, which remains to be a challenge in practical scenarios. This study aims to apply CNN to extract coil sensitivity information from the undersampled center k-space, and use the estimated sensitivity maps for parallel imaging. Results show that no obvious residual signal can be seen in the reconstructed images for all cases, which indicates the efficacy of the proposed method. Besides, the CNN based SENSE image without ACS appears to be less noisy than conventional SENSE results with ACS, which may benefit from the denoising effect from CNN on the sensitivity maps. |
| 3611 | Parallel Imaging with a Combination of SENSE and Generative Adversarial Networks (GAN) | |
| Jun Lyu1, Peng Wang1, and Chengyan Wang2 | ||
1Yantai University, Yantai, China, 2Fudan University, Shanghai, China |
||
This study aims to use GAN architecture to remove the g-factor artifacts in SENSE reconstruction. The proposed method outperforms SENSE and ZF+GAN in terms of the measured quality metrics (decreases of NMSE and increases of PSNR and SSIM). Besides, our method performs well in preserving images details with under-sampling factor of up to 6-fold, which is promising to be applied in clinical applications. |
| 3612 | An Information Theoretical Framework for Machine Learning Based MR Image Reconstruction | |
| Yudu Li1,2, Yue Guan3, Ziyu Meng2,3, Fanyang Yu2,4, Rong Guo1,2, Yibo Zhao1,2, Tianyao Wang5, Yao Li3, and Zhi-Pei Liang1,2 | ||
1Department of Electrical and Computer Engineering, University of Illinois at Urbana-Champaign, Urbana, IL, United States, 2Beckman Institute for Advanced Science and Technology, University of Illinois at Urbana-Champaign, Urbana, IL, United States, 3Institute for Medical Imaging Technology (IMIT), School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China, 4Department of Bioengineering, University of Illinois at Urbana-Champaign, Urbana, IL, United States, 5Department of Radiology, The Fifth People's Hospital of Shanghai, Shanghai, China |
||
Machine learning (ML) based MR image reconstruction leverages the great power and flexibility of deep networks in representing complex image priors. However, ML image priors are often inaccurate due to limited training data and high dimensionality of image functions. Therefore, direct use of ML-based reconstructions or treating them as statistical priors can introduce significant biases. To address this limitation, we treat ML-based reconstruction as an initial estimate and use an information theoretical framework to incorporate it into the final reconstruction, which is optimized to capture novel image features. The proposed method may provide an effective framework for ML-based image reconstruction. |
| 3613 | Super-resolution MRI using deep convolutional neural network for adaptive MR-guided radiotherapy: a pilot study | |
| Yihang Zhou1, Hongyu Li2, Jing Yuan1, Leslie Ying2, Kin Yin Cheung1, and Siu Ki Yu1 | ||
1Medical Physics & Research Department, Hong Kong Sanatorium & Hospital, Hong Kong, China, 2Department of Biomedical Engineering, Department of Electrical Engineering, The State University of New York at Buffalo, Buffalo, NY, United States |
||
MR-guided radiotherapy (MRgRT) is creating new perspectives towards an individualized precise radiation therapy solution. However, spatial resolution of fractional MRI can be much restricted, in order to shorten scan time, by patient tolerance of immobilization, intra-fractional anatomical motion and complicated MRgRT workflow. We hypothesized that the quality of low-resolution daily MRI could be greatly restored to generate super-resolution MRI, whose quality should be comparable of high-resolution planning MRI, by applying deep learning techniques. In this study, we aimed to investigate the feasibility of deep learning super-resolution MRI generation in the head-and-neck for adaptive MRgRT purpose. |
| 3614 | Model-augmented deep learning for VFA-T1 mapping | |
| Lea Bogensperger1,2, Oliver Maier1, and Rudolf Stollberger1,3 | ||
1Institute of Medical Engineering, Graz University of Technology, Graz, Austria, 2Institute for Computer Graphics and Vision, Graz University of Technology, Graz, Austria, 3Biotechmed, Graz, Austria |
||
A deep learning approach is proposed to estimate M0 and T1 maps from undersampled variable flip angle (VFA) data to explore the potential of this method for acceleration and rapid reconstruction even without parallel imaging. A U-Net was implemented with a model consistency term containing the signal equation to ensure the physical validity and to include prior knowledge of B1+. Training is performed on numerical brain phantoms and by means of transfer-learning on retrospectively undersampled in-vivo data. Qualitative and quantitative results show the acceleration potential for both numerical and in-vivo data for acceleration factors R=1.89, 3.43, and 5.84. |
| 3615 | Relax-ADMM-Net: A Relaxed ADMM Network for Compressed Sensing MRI | |
| Yiling Liu1, Jing Cheng2, Yanjie Zhu2, Haifeng Wang2, Ziwen Ke1, Qiegen Liu3, Xin Liu2, Hairong Zheng2, Leslie Ying4, and Dong Liang1,2 | ||
1Research center for Medical AI, Shenzhen Institutes of Advanced Technology, shenzhen, China, 2Paul C. Lauterbur Research Center for Biomedical Imaging, Shenzhen Institutes of Advanced Technology, shenzhen, China, 3Department of Electronic Information Engineering, Nanchang University, Nanchang, China, 4Departments of Biomedical Engineering and Electrical Engineering, University at Buffalo,the State University of New York, Buffalo, NY, United States |
||
ADMM is a popular algorithm for Compressed sensing (CS) MRI. ADMM-based deep networks have also achieved a great success by unrolling the ADMM algorithm into deep neural networks. Nevertheless, ADMM-Nets only make the components in the regularization term learnable. In this work, we propose a relaxed version of ADMM-Net (i.e. Relax-ADMM-Net) to further improve its performance for fast MRI, where the additional data consistency term and variable combinations in the updating rules are all freely learned by the network. Experiments reveal the effectiveness of the proposed network compared with several competing model-driven networks. |
| 3616 | Deep Network Interpolation for Accelerated Parallel MR Image Reconstruction | |
| Chen Qin1, Jo Schlemper1,2, Kerstin Hammernik1, Jinming Duan3, Ronald M Summers4, and Daniel Rueckert1 | ||
1Imperial College London, London, United Kingdom, 2Hyperfine Research Inc., Guilford, CT, United States, 3School of Computer Science, University of Birmingham, Birmingham, United Kingdom, 4NIH Clinical Center, Bethesda, MD, United States |
||
We present a deep network interpolation strategy for accelerated parallel MR image reconstruction. In particular, we examine the network interpolation in parameter space between a source model that is formulated in an unrolled scheme with L1 and SSIM losses and its counterpart that is trained with an adversarial loss. We show that by interpolating between the two different models of the same network structure, the new interpolated network can model a trade-off between perceptual quality and fidelity. |
| 3617 | Unsupervised Deep Learning Reconstruction Using the MR Imaging Model | |
| Peizhou Huang1, Chaoyi Zhang2, Hongyu Li2, Sunil Kumar Gaire2, Ruiying Liu2, Xiaoliang Zhang1, Xiaojuan Li3, Liang Dong4, 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 |
||
Deep learning has been applied to MRI image reconstruction successfully. Most existing works require labeled ground-truth images to learn network parameters for image reconstruction, which is not practical in some MR applications where acquisition of fully sampled images takes too long. In this abstract, we propose a novel unsupervised deep neural network for reconstruction from undersampled data. The proposed network, named URED-net, is built upon conventional ADMM algorithm for compressed sensing reconstruction, but incorporating noise2noise, an unsupervised deep denoising network. The experimental results demonstrate proposed URED-net is superior to the standard noise2noise network with and without ground-truth images for training. |
| 3618 | Multi-objective Deep Learning for Joint Estimation and Detection Tasks in MRI | |
| Zhiyang Fu1, Maria I Altbach2, Diego R Martin2, and Ali Bilgin1,2,3 | ||
1Electrical and Computer Engineering, University of Arizona, Tucson, AZ, United States, 2Department of Medical Imaging, University of Arizona, Tucson, AZ, United States, 3Biomedical Engineering, University of Arizona, Tucson, AZ, United States |
||
MR images are often reconstructed first and then used for medical image analysis tasks such as segmentation or classification. This sequential procedure can compromise the performance of the image analysis task. In this work, we propose a multi-task learning framework that jointly reconstructs underlying images and detects multiple sclerosis lesions. This framework outperforms the conventional sequential processing pipeline. We also introduce a multi-objective optimization as an effective and automated approach to balance the trade-off among multi-task losses. Experimental results suggest that taking into account subsequent detection tasks during image reconstruction may lead to enhanced detection performance. |
| 3619 | Deep Simultaneous Optimization of Sampling and Reconstruction for Multi-contrast MRI | |
| Xinwen Liu1, Jing Wang1,2, Fangfang Tang1, Shekhar S. Chandra1, Feng Liu1, and Stuart Crozier1 | ||
1School of Information Technology and Electrical Engineering, the University of Queensland, Brisbane, Australia, 2School of Information and Communication Technology, Griffith University, Brisbane, Australia |
||
MRI images of the same subject in different contrasts contain shared information, such as the anatomical structure. Utilizing the redundant information amongst the contrasts to sub-sample and faithfully reconstruct multi-contrast images could greatly accelerate the imaging speed, improve image quality and shorten scanning protocols. We propose an algorithm that generates the optimized sampling pattern and reconstruction scheme of one contrast (e.g. T2-weighted image) when images with different contrast (e.g. T1-weighted image) have been acquired. The proposed algorithm achieves increased PSNR and SSIM with the resulting optimal sampling pattern compared to other acquisition patterns and single contrast methods. |
| 3620 | A Hybrid SENSE Reconstruction Combined with Deep Convolution Neural Network | |
| Hangfei Liu1, Jingjing Li1, Qing Tang1, and Tao Zhang1,2,3 | ||
1School of Life Science and Technology, University of Electronic Science and Technology of China, chengdu, China, 2High Field Magnetic Resonance Brain Imaging Laboratory of Sichuan, Chengdu, China, 3Key Laboratory for Neuro Information, Ministry of Education, Chengdu, China |
||
Recently parallel imaging reconstruction based on deep learning has made lots of progresses, however, there still exist several common challenges, i.e. generalization, transferability and robustness. On the contrary, SENSE reconstruction has been routinely used in clinical scans due to its high robustness and excellent image quality. A high-quality coil sensitivity map (HQCSM) is the key to achieve good SENSE reconstruction. We proposed a hybrid SENSE reconstruction frame, combining the SENSE reconstruction algorithm with a deep convolutional neural network to learn HQCSM from a few automatic calibration lines (ACS), which shows good generalization for different under-sampling ratio and enhanced robustness. |
| 3621 | Reconstructing non-Cartesian acquisitions using dAUTOMAP | |
| Maarten L Terpstra1,2, Federico d'Agata1,2,3, Bjorn Stemkens1,2, Jan JW Lagendijk1, Cornelis AT van den Berg1,2, and Rob HN Tijssen1,2 | ||
1Department of Radiotherapy, Division of Imaging & Oncology, University Medical Center Utrecht, Utrecht, Netherlands, 2Computational Imaging Group for MR diagnostics & therapy, Center for Image Sciences, University Medical Center Utrecht, Utrecht, Netherlands, 3Department of Neurosciences, University of Turin, Turin, Italy |
||
Recently, dAUTOMAP has been presented to perform deep learning-based image reconstruction. dAUTOMAP uses gridded k-space points and so far it has only been used to reconstruct Cartesian acquisitions. In this work, we demonstrate that dAUTOMAP can produce high-quality reconstructions on radial and spiral non-Cartesian acquisitions and can resolve artifacts beyond those introduced by the undersampled acquisition. |
| 3622 | End-to-End Deep Learning Reconstruction for Ultra-Short MRF | |
| Mingdong Fan1, Brendan Eck2, Nicole Seiberlich3, Michael Martens1, and Robert Brown1 | ||
1Physics, Case Western Reserve University, Cleveland, OH, United States, 2Cardiovascular and Metabolic Sciences, Cleveland Clinic, Cleveland, OH, United States, 3Radiology, University of Michigan, Ann Arbor, Ann Arbor, MI, United States |
||
There are two major challenges in MRF reconstruction, the aliasing artifacts that results from the largely under-sampled k-space, and the very long MRF sequence used in practice to improve the reconstruction accuracy. In this study, we propose an end-to-end deep learning based reconstruction model that aims to address the issue of the spatial aliasing artifacts and provide accurate reconstruction with ultra-short MRF signals. |
| 3623 | k-t CNN for Modeling Spatio-temporal Mappings and an Application to Reconstruction of k-space Data with Stack-of-spirals Trajectory | |
| Hidenori Takeshima1 and Hideaki Kutsuna2 | ||
1Advanced Technology Research Department, Research and Development Center, Canon Medical Systems Corporation, Kanagawa, Japan, 2MRI Systems Development Department, MRI Systems Division, Canon Medical Systems Corporation, Kanagawa, Japan |
||
The authors propose a new model using a convolutional neural network (CNN) named k-t CNN for approximating non-linear spatio-temporal mappings used in various applications. Existing studies imply that spatio-temporal mappings are non-linear. Most existing studies developed various methods using linear models for spatio-temporal applications. Meanwhile, the effectiveness of non-linear models was shown for spatial-domain applications. As an application of k-t CNN, the effectiveness of the proposed method is shown experimentally in the case of reconstruction of stack-of-spirals k-space data. |
| 3624 | Deep Learning-Based Adaptive Noise Reduction for Improving Image Quality of 1.5T MR Images | |
| Shigeru Kiryu1, Yasutaka Sugano2, Tomoyuki Ohta3, and Kuni Ohtomo4 | ||
1Radiology, International University of Health, School of medicine, Chiba, Japan, 2Canon Medical Systems Corporation, Kanagawa, Japan, 3International University of Health and Welfare Hospital, Tochigi, Japan, 4International University of Health and Welfare, Tochigi, Japan |
||
We assessed the performance of the Deep Learning-based Reconstruction (dDLR) technique in improving 1.5T MR images. Eleven volunteers underwent MR imaging at 3T and 1.5T on the same day with the same imaging parameters. We applied the dDLR to the 1.5T image data (dDLR-1.5T), and then compared the 1.5T and dDLR-1.5T datasets with reference to the 3T dataset. The structure similarity of dDLR-1.5T was higher than that of 1.5T and dDLR increased SNR at 1.5T. The dDLR technique improves the image quality of MR images obtained at 1.5T. |
| 3625 | Simultaneous Brain Anatomical and Arterial Imaging by 3T MRI: Reconstruction Based on a Generative Adversarial Network | |
| Wei Yu1,2, Lixin Wang1,2, Xianchang Zhang3, Zhentao Zuo4, Rong Xue4,5,6, and Tianyi Qian1,2 | ||
1Sinovation Ventures AI Institute, Beijing, China, 2QuantMind, Beijing, China, 3MR Collaboration, Siemens Healthcare, Beijing, China, 4State Key Laboratory of Brain and Cognitive Science, Beijing MRI Center for Brain Research, Institute of Biophysics, Chinese Academy of Sciences, Beijing, China, 5University of Chinese Academy of Sciences, Beijing, China, 6Beijing Institute for Brain Disorders, Beijing, China |
||
This study investigates a reconstruction method designed to generate 7T magnetic resonance images (MRI) from 3T MRI based on a generative adversarial network. Compared with current reconstruction methods, this method can simultaneously reconstruct well-defined anatomical details and salient blood vessels. This reconstruction method may be useful for increasing the efficiency of brain MRI examinations. |
| 3626 | A Recurrent Neural Network (RNN) based reconstruction of extremely undersampled neuro-interventional MRI | |
| Ruiyang Zhao1, Tao Wang2, Kang Yan1, Chengcheng Zhang3, Zhipei Liang4, Yiping Du1, Dianyou Li3, Bomin Sun3, and Yuan Feng1 | ||
1Institute for Medical Imaging Technology, Shanghai Jiao Tong University, Shanghai, China, 2Functional Neurosurgery,Ruijin Hospital affiliated to Shanghai Jiao Tong University, Shanghai, China, 3Shanghai Jiao Tong University Medical School Affiliated Ruijin Hospital, Shanghai, CHINA, Shanghai, China, 4Beckman Institute for Advanced Science & Technology, Department of Electrical & Computer Engineering，University of Illinois at Urbana-Champaign, Urbana-Champaign, IL, United States |
||
Real –time MR image-guided neurosurgery could greatly improve the surgery accuracy and outcome. However, real-time guidance requires highly accelerated imaging. In this study, we proposed a Convolutional Long Short-term Memory (Conv-LSTM) based U-net to reconstruct consecutive image frames with golden-angle sampling. The Conv-LSTM based architecture was developed to explore time coherence information. Training and test datasets were generated from MR images of patients treated with Deep Brain Stimulation (DBS). Results showed that our model could achieve an acceleration rate ~80x, which provided great potentials for application in MR-guided interventional therapy. |
| 3627 | Partial Fourier MRI Reconstruction Using Convolutional Neural Networks | |
| Peibei Cao1,2, Linfang Xiao1,2, Yilong Liu1,2, Yujiao Zhao1,2, Yanqiu Feng3, Alex T Leong1,2, and Ed X Wu1,2 | ||
1Laboratory of Biomedical Imaging and Signal Processing, The University of Hong Kong, Hong Kong, China, 2Department of Electrical and Electronic Engineering, The University of Hong Kong, Hong Kong, China, 3School of Biomedical Engineering, Southern Medical University, Guangzhou, China |
||
Convolutional neural network (CNN) has emerged as a powerful tool for medical image reconstruction. In this study, we designed and implemented a CNN model for partial Fourier MRI reconstruction, and compared its performance with the existing projection onto convex sets (POCS) method. The results demonstrated that our proposed deep learning approach could effectively recovered the high frequency components and outperformed the POCS method especially when partial Fourier fraction is close to 50%. |
| 3628 | Derivation of quantitative T1 map from a single MR image using a self-attention deep neural network | |
| Yan Wu1, Yajun Ma2, Jiang Du2, and Lei Xing1 | ||
1Radiation Oncology, Stanford University, Stanford, CA, United States, 2Radiology, University of California San Diego, La Jolla, CA, United States |
||
The application of quantitative MRI is limited by additional data acquisition for variable contrast images. Leveraging from the unique ability of deep learning, we propose a data-driven strategy to derive quantitative T1 map and proton density map from a single qualitative MR image without specific requirements on the weighting of the input image. The quantitative parametric mapping tasks are accomplished using self-attention deep convolutional neural networks, which make efficient use of local and non-local information. In this way, qualitative and quantitative MRI can be attained simultaneously without changing the existing imaging protocol. |
| 3629 | Impact of machine learning in iterative motion corrected reconstructions | |
| Rita G. Nunes1, Santiago Sanz-Estébanez2, Joseph V. Hajnal3, Lucilio Cordero-Grande3, and Carlos Alberola-López2 | ||
1ISR-Lisbon/LARSyS and Department of Bioengineering, Instituto Superior Técnico – University of Lisbon, Lisbon, Portugal, Lisbon, Portugal, 2Laboratorio de Procesado de Imagen, Universidad de Valladolid, Valladolid, Spain, Valladolid, Spain, 3Centre for the Developing Brain and Department of Biomedical Engineering, School of Biomedical Engineering and Imaging Sciences, King’s College London,London U.K, London, United Kingdom |
||
Incorporation of machine learning (ML) approaches in MR reconstruction is currently a hot topic because of the enormous potential that deep learning solutions have shown in vision and imaging communities. Recently, a procedure known as NAMER has been proposed; this procedure incorporates a ML module into an iterative reconstruction for multishot acquisitions with inter-shot motion estimation and correction (referred to as aligned reconstruction). In this abstract we provide some insight on the benefits and limitations associated to NAMER by analyzing its behavior both with a steady and a discontinued use of the ML artifact cleaning step. |
| 3630 | Assessment of the Generalization of Learned Unsupervised Deep Learning Method | |
| Ziwen Ke1,2, Yanjie Zhu3, Jing Cheng2,3, Leslie Ying4, Xin Liu3, Hairong Zheng3, and Dong Liang1,3 | ||
1Research Center for Medical AI, Shenzhen Institutes of Advanced Technology，Chinese Academy of Sciences, Shenzhen, China, 2Shenzhen College of Advanced Technology, University of Chinese Academy of Sciences, Shenzhen, China, 3Paul C. Lauterbur Research Center for Biomedical Imaging, Shenzhen Institutes of Advanced Technology，Chinese Academy of Sciences, Shenzhen, China, 4Department of Biomedical Engineering and Department of Electrical Engineering, The State University of New York, Buffalo, NY, United States |
||
In our previous work, we proposed an unsupervised deep learning method for parallel MR cardiac imaging via time interleaved sampling. The comparisons with classical methods on in vivo data have shown that this method can achieve improved reconstruction results. However, the proposed unsupervised framework is based on the time interleaved sampling scheme. Does the model trained with time interleaved undersampling pattern have good generalization to other sampling patterns? In this paper, we will explore the generalization performance of the learned unsupervised deep learning method under different sampling patterns. |
| 3631 | Deep-learning reconstruction for 3D Delayed Myocardial Enhancement | |
| Gaspar Delso1, Suryanarayanan Kaushik2, Graeme C McKinnon2, Daniel Lorenzatti3, Julián Vega3, Teresa M. Caralt3, Adelina Doltra3, José T. Ortiz-Pérez3, Rosario J. Perea3, Susanna Prat3, Marta Sitges3, and Martin A. Janich4 | ||
1ASL MR, GE Healthcare, Barcelona, Spain, 2GE Healthcare, Waukesha, WI, United States, 3Hospital Clínic de Barcelona, Barcelona, Spain, 4GE Healthcare, Munich, Germany |
||
We present the evaluation results of a novel Deep Learning reconstruction framework, applied to clinical Delayed Myocardial Enhancement datasets acquired with a 3D inversion-recovery T1-weighted gradient echo sequence. |
| 3632 | Deep learning for undersampled spiral DENSE reconstruction | |
| Samuel W Fielden1,2, Eric D Carruth1, Christopher D Nevius1, Christopher M Haggerty1, and Brandon K Fornwalt1,3 | ||
1Imaging Sciences & Innovation, Geisinger, Danville, PA, United States, 2Medical & Health Physics, Geisinger, Danville, PA, United States, 3Radiology, Geisinger, Danville, PA, United States |
||
Displacement Encoding with Stimulated Echoes (DENSE) is a powerful technique that has found great utility in accurately measuring cardiac tissue displacement. However, DENSE remains time-consuming to acquire, particularly for 3-dimensionally encoded or higher resolution schemes, and so methods to accelerate image acquisition are needed. Here, we apply the Deep Cascade of Convolutional Neural Networks (DCCNN) to the complex-valued, non-Cartesian data of DENSE to show that accelerated imaging via k-space undersampling is feasible using a deep learning-based reconstruction. |
| 3633 | Suppression of Artifact-Generating Echoes in Cine DENSE using Deep Learning | |
| Mohammad Abdishektaei1, Xue Feng1, Craig H Meyer1, and Frederick H Epstein1 | ||
1Biomedical Engineering, University of Virginia, Charlottesville, VA, United States |
||
Cine displacement encoding with stimulated echoes (DENSE) is an accurate and reproducible method of strain imaging. The stimulated echo (STE), which carries the tissue displacement information in it’s phase, is simultaneously acquired with two artifact-generating echoes. A combination of phase-cycled acquisitions and through-plane dephasing are typically used to suppress the artifact-generating echoes. The limitations of these methods are longer acquisition times, susceptibility to breathing motion and the loss of signal-to-noise ratio due to intravoxel dephasing. To potentially overcome these limitations, the use of a deep convolutional neural network to suppress the undesired echoes from a single acquisition was investigated. |
| 3634 | Learning reconstruction without ground-truth data: an unsupervised way for fast MR imaging | |
| Jing Cheng1, Ziwen Ke1, Haifeng Wang1, Yanjie Zhu1, Leslie Ying2, Xin Liu1, Hairong Zheng1, and Dong Liang1 | ||
1Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China, 2University at Buffalo, The State University of New York, Buffalo, Buffalo, NY, United States |
||
Most deep learning methods for MR reconstruction heavily rely on the large number of training data pairs to achieve best performance. In this work, we introduce a simple but effective strategy to handle the situation where collecting lots of fully sampled rawdata is impractical. By defining a CS-based loss function, the deep networks can be trained without ground-truth images or full sampled data. In such an unsupervised way, the MR image can be reconstructed through the forward process of deep networks. This approach was evaluated on in vivo MR datasets and achieved superior performance than the conventional CS method. |
| 3635 | Learning-based Optimization of the Under-sampling PattErn with Straight-Through Estimator (LOUPE-ST) for Fast MRI | |
| Jinwei Zhang1,2, Hang Zhang2,3, Cagla Deniz Bahadir4, Alan Wang3, Mert Rory Sabuncu1,2,3, Pascal Spincemaille2, Thanh D. Nguyen2, and Yi Wang1,2 | ||
1Biomedical Engineering, Cornell University, Ithaca, NY, United States, 2Weill Cornell Medicine, New York, NY, United States, 3Electrical and Computer Engineering, Cornell University, Ithaca, NY, United States, 4Siemens Healthineers, Princeton, NJ, United States |
||
In this work, we propose LOUPE-ST, which extends the previously introduced optimal k-space sampling pattern learning framework called LOUPE by employing a straight-through estimator to better handle the gradient back-propagation in the binary sampling layer and incorporating an unrolled optimization network (MoDL) to reconstruct T2w images from under-sampled k-space data with high fidelity. Our results indicate that, compared with the variable density under-sampling pattern at the same under-sampling ratio (10%), superior reconstruction performance can be achieved with LOUPE-ST optimized under-sampling pattern. This was observed for all reconstruction methods that we experimented with. |
| 3636 | Locally and Globally Concatenated Network for MR Image Reconstruction | |
| Zechen Zhou1, Christophe Schülke2, Chun Yuan3, and Peter Börnert2 | ||
1Philips Research North America, Cambridge, MA, United States, 2Philips Research Hamburg, Hamburg, Germany, 3Department of Radiology, University of Washington, Seattle, WA, United States |
||
Recently, the convolutional neural network (CNN) based reconstruction concept has emerged as a promising implementation of compressed sensing tailored for specific fast imaging applications. The reconstruction performance of such data-driven models may depend on the CNN structure which determines the feature extraction process for sparse representation. In this study, a locally and globally concatenated network is proposed and compared with the residual network as well as the traditional L1-wavelet ESPIRiT. Preliminary experiments on a public knee imaging database showed that the proposed approach provided improved fine structure (e.g. vessel wall) restoration and background noise reduction. |
| 3637 | Image Reconstruction Using Generative Adversarial Networks with MR-Specific Feature Map | |
| Ruiying Liu1, Hongyu Li1, Dong Liang2, Xiaojuan Li3, Chaoyi Zhang1, Peizhou Zhou1, Leslie Ying1, and Xiaoliang Zhang4 | ||
1Department of Biomedical Engineering, Department of Electrical Engineering, The State University of New York at Buffalo, Buffalo, NY, United States, 2Paul C. Lauterbur Research Center for Biomedical Imaging, SIAT, CAS,, Shenzhen, China, 3Program of Advanced Musculoskeletal Imaging (PAMI), Cleveland Clinic, Cleveland, OH, United States, 4Department of Biomedical Engineering, The State University of New York at Buffalo, Buffalo, NY, United States |
||
Deep learning methods have demonstrated great potential in image reconstruction due to its ability to learn the non-linearity relationship between the undersampled k-space data and the corresponding desired image. Among these methods, Generative Adversarial Networks (GANs) is known to reconstruct images that are sharper and more realistic-looking. In this abstract, we study whether an MR-specific feature map that is trained on a large number of MRI images and used in the loss function can improve the GAN-based reconstruction. We demonstrate that the MR-specific feature map is superior to the pre-trained feature map typically used for GAN-based reconstruction. |
| 3638 | A deep-learning based synthesized T2 weighted imaging with multi-modality information and k-space correction | |
| Qing Tang1, Ye Li1, Hangfei Liu1, and Tao Zhang1,2,3 | ||
1School of Life Science and Technology, University of Electronic Science and Technology of China, Chengdu, China, 2High Field Magnetic Resonance Brain Imaging Laboratory of Sichuan, Chengdu, China, 3Key Laboratory for Neuro Information, Ministry of Education, Chengdu, China |
||
T2 weighted image (T2WI) usually takes more time and thus is more vulnerable to motion artifacts. With the recent development of applying deep learning to MR imaging, many neural networks are proposed to synthesize high-quality T2 images from under-sampled T2 or other modalities (such as T1). Here we develop a Simple-ResNet network to synthesize high-quality T2 images based on multi-modality information and followed by a k-space correction module. Results show that our model is very easy to train and the synthesized T2 images can achieve comparable image quality as the fully-sampled T2 images. |
| 3639 | Quantitative and Volumetric Assessment of a Deep Cascade Network for MR Reconstruction Under Different Acceleration Factors | |
| Wallace Souza Loos1, Roberto Souza1, Mariana Bento1, Robert Marc Lebel1,2, and Richard Frayne1 | ||
1University of Calgary, Calgary, AB, Canada, 2General Electric Healthcare, Calgary, AB, Canada |
||
Magnetic resonance (MR) imaging still has a high acquisition time due to inherent sequential procedure required to fill k-space. Deep-cascade networks have been used to reconstruct MR images from an under-sampled k-space in order to reduce acquisition time. In this work we investigate a deep-cascade to reconstruct MR images of the brain. We trained the network with 14 different acceleration factors (R). Relevant brain structures were preserved until R = 7x. For R ≥ 8x, MR images presented noticeable blurring artifact. The quality of the segmentation of the brain structures were similar to the reference MR image until R=9. |
| 3640 | Deep Residual Grappa (DeepGrappa): A General Purpose Self-calibrated AI based MR Reconstruction | |
| Hui Xue1, James C Moon2, and Peter Kellman1 | ||
1NHLBI, NIH, Bethesda, MD, United States, 2Barts Heart Centre, Barts Health NHS Trust, London, United Kingdom |
||
In this abstract, we proposed a novel self-calibration AI based MR reconstruction algorithm to utilize the power of a deep neural network. Unlike most deep learning MR reconstruction, this algorithm does not require extra training data and only works on the auto-calibration kspace lines. This algorithm is integrated to run on MR scanner via the Gadgetron InlineAI toolbox. We demonstrated this algorithm on cardiac cine imaging, showing improved image quality without introduced unrealistic anatomical structures. |
| 3641 | Are deep learning MR reconstruction models robust against adversarial attacks? | |
| Taohui Xiao1, Cheng Li1, Haoyun Liang1, Hairong Zheng1, and Shanshan Wang1 | ||
1Paul C. Lauterbur Research Center for Biomedical Imaging, SIAT, Chinese Academy of Sciences, Shenzhen, P.R.China, Shenzhen, China |
||
This paper investigates the robustness of deep learning MR reconstruction models for adversarial attacks like new lesions, different anatomy and noise pollutions. Specifically, three popular MR reconstruction algorithms were selected to investigate this issue. Experimental results show that model-based deep learning MR reconstruction method is relatively more robust than end-to-end data-driven reconstruction networks when transfer to other organs or face new lesions. Data-driven approaches can achieve better results when the testing images follow similar distributions as the training images. Severe noise can be a big issue for both deep learning methods and the traditional method. |
| 3642 | Plug-and-Play Deep Learning Module for Faster Parallel MR Imaging | |
| Kamlesh Pawar1, Gary Egan1, and Zhaolin Chen1 | ||
1Monash Biomedical Imaging, Monash University, Melbourne, Australia |
||
Deep learning (DL) methods are superior to the conventional method of accelerated imaging such as parallel imaging and compressed sensing but the integration of DL methods into the MR scanners is still in its infancy. The integration of the DL methods into the MR scanner requires the design of new pulse sequences with a modified sampling patterns/trajectories and development of DL reconstruction framework within the MR scanner. In this work, we present an effective plug-and-play approach of integrating the DL reconstruction into the MR scanner that eliminates the need to modify the existing image acquisition and reconstruction pipeline. |
| 3643 | Brain MR image super-resolution using regularized deep image prior | |
| Yue Hu1, Peng Li1, and Dong Nan2 | ||
1Harbin Institute of Technology, Harbin, China, 2The First Affiliated Hospital of Harbin Medical University, Harbin, China |
||
We propose a novel algorithm for the super-resolution of brain MR images based on feature regularized DIP network, where no prior training pairs are required. We formulate the network by including the total variation (TV) term as the sparsity regularization and the Laplacian as the sharpness regularization. The network is iteratively updated using the image feature regularizations and the measured image. Numerical experiments demonstrate the improved performance offered by the proposed method. |
| 3644 | Graphical sequence visualization and development in the dynamic platform-independent framework gammaSTAR (γ*): A first prototype | |
| Simon Konstandin1, Cristoffer Cordes1, and Matthias Günther1,2 | ||
1MR Physics, Fraunhofer MEVIS, Bremen, Germany, 2MR-Imaging and Spectroscopy, Faculty 01 (Physics/Electrical Engineering), University of Bremen, Bremen, Germany |
||
MR sequence development is usually carried out by means of manufacturer-specific frameworks that do not allow an easy sequence transfer to scanners from other manufacturers. Sequence development in the recently presented framework is performed by scripting that can be time consuming and confusing for sophisticated MR sequences. Here, a module editor prototype is presented that provides a graphical sequence representation and should allow for a fast and relatively easy-to-understand MR sequence development in the future. Functionality could already be demonstrated by exchanging the rf pulse and implementing a parallel acquisition technique into a gradient echo sequence in a few steps. |
| 3645 | Reconstream: A Software Framework for Developing Image Reconstruction Algorithms in Scientific Research Activities | |
| Hidenori Takeshima1 | ||
1Advanced Technology Research Department, Research and Development Center, Canon Medical Systems Corporation, Kanagawa, Japan |
||
Existing software tools for assisting scientific research mainly focus on increasing reproducibility of primitives and/or integrated algorithms (exchangeable reproducibility). While exchangeable reproducibility is clearly important, reproducing the outputs using past configurations again is also important in daily research activities (local reproducibility). The author proposes a new software framework named reconstream. To improve local reproducibility, reconstream implemented a revision manager for reproducing past configurations easily. To improve exchangeable reproducibility, the author reviewed advantages of individual software tools and implemented them in the reconstream framework. Experimental results demonstrate several examples using various features of reconstream. |
| 3646 | Utility of Whole-Liver Histogram and Texture Analysis on T1 Maps in the Risk Stratification of Advanced Fibrosis in NAFLD | |
| Xinxin Xu1, Caixia Fu2, Robert Grimm3, Huimin Lin1, Ruokun Li1, and Fuhua Yan1 | ||
1Radiology, Ruijin Hospital affiliated to Shanghai Jiaotong University School of Medicine, Shanghai, China, 2Siemens healthcare, Shanghai, China, 3Siemens healthcare, Erlangen, Germany |
||
We aim to assess the utility of whole-liver texture analysis on T1 maps for the risk stratification of advanced fibrosis in patients with suspected NAFLD. Our experiments show that whole-liver histogram and texture parameters of T1 maps can distinguish NAFLD patients with an intermediate-to-high risk of advanced fibrosis. A combination of histogram and texture parameters as a multivariate model showed a better diagnostic performance than any sole parameter and noninvasive fibrosis test in risk stratification of advanced fibrosis. |
| 3647 | Python pulse sequence development kit for a fast MRI simulator | |
| Ryoichi Kose1 and Katsumi Kose1 | ||
1MRIsimulations,Inc., Tokyo, Japan |
||
A software tool that can efficiently describe MRI pulse sequences in Python has been developed for a fast MRI simulator. The essential part of this tool consists of application program interface (API) developed in C++, and any pulse sequence can be written by calling the API from a Python program. Using this tool, several pulse sequences including magnetic resonance fingerprinting were developed, and their usefulness was evaluated by performing MRI simulations for numerical phantoms. As a result, the tool developed in this study was shown to be very efficient in developing MRI pulse sequences. |
| 3648 | An MR guided focused ultrasound software with Gadgetron reconstruction | |
| Yangzi Qiao1, Chao Zou1, Jianhong Wen1, Sen Jia1, Xin Liu1, and Hairong Zheng1 | ||
1Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China |
||
An integrated focused ultrasound guidance software named MARFit was developed base on the framework of Gadgetron. The software realized automatically focus localization and real-time temperature change monitoring during HIFU therapy. These features has been evaluated in animal experiments. The software makes the post processing procedure easily translated between different vendors, and could be a powerful tool for MR guided focused ultrasound therapy. |
| 3649 | Cross-vendor implementation of a Stack-of-spirals PRESTO BOLD fMRI sequence using TOPPE and Pulseq | |
| Marina Manso Jimeno1,2, Sairam Geethanath1,2, Jon-Fredrik Nielsen3, and Douglas C. Noll3 | ||
1Columbia University, New York, NY, United States, 2Columbia Magnetic Resonance Research Center (CMRRC), New York, NY, United States, 3University of Michigan, Ann Arbor, MI, United States |
||
TOPPE and Pulseq are two examples of open-source, vendor-agnostic frameworks for rapid prototyping and implementation of MR sequences. Both have demonstrated their ability to easily execute a variety of highly specific and customized sequences on different vendors. Cross-vendor translation of sequences generated by these two frameworks can enable multi-site and multi-vendor evaluation of a single sequence implementation. This study demonstrates this capability by porting a spiral-based fMRI sequence implemented in TOPPE on one vendor platform to Pulseq to run on another vendor. Results show comparable in vivo SNR and image contrast for Pulseq and TOPPE implementations. |
| 3650 | Open-source Python package for spiral off-resonance correction | |
| Marina Manso Jimeno1,2, Sairam Geethanath1,2, and John Thomas Vaughan Jr.1,2 | ||
1Columbia University, New York, NY, United States, 2Columbia Magnetic Resonance Research Center (CMRRC), New York, NY, United States |
||
Off-resonance blurring is one of the challenges with long read-out trajectories such as spirals. Existent techniques deblur the images by demodulating k-space data with the conjugate term of the phase accumulated from off-resonant spins. Continuous and frequency-segmented CPR and Multi-Frequency Interpolation are some examples. This work is aimed at sharing an open-source package for off-resonance correction, demonstrating on spiral datasets. The methods have been tested on simulations, in vitro and in vivo and show promising preliminary results. We believe that the availability of this package will benefit the spiral imaging community. |
| 3651 | An open-source graphical tool for interactive slice planning: Application to pseudo-continuous Arterial Spin Labeling | |
| Jon-Fredrik Nielsen1, Luis Hernandez-Garcia1, and Douglas C. Noll1 | ||
1Biomedical Engineering, University of Michigan, Ann Arbor, MI, United States |
||
Vendor-agnostic MR pulse sequence programming tools such as Pulseq or TOPPE allow rapid prototyping of complex MR sequences, but are not compatible with existing graphical slice planning interfaces on all supported vendor platforms. We introduce a simple open-source slice planning tool that allows any number of 3D rectangular regions-of-interest (ROIs) to be defined interactively and exported to a portable data file format (HDF5). We use this tool to prescribe both a 2D labeling plane in the neck and a 3D imaging volume in the brain for a vendor-agnostic implementation of 3D stack-of-spirals fast spin-echo (FSE) pseudo-continuous Arterial Spin Labeling. |
| 3652 | Intuitive MRI Data Visualization and Reslicing Using Augmented Reality | |
| Bragi Sveinsson1,2, Neha Koonjoo1,2, and Matthew Rosen1,2 | ||
1Massachusetts General Hospital, Boston, MA, United States, 2Harvard Medical School, Boston, MA, United States |
||
We present a smartphone app to view 3D medical image data using Augmented Reality (AR). The app superimposes the imaged anatomy on the patient and allows for completely free and intuitive choice of viewing plane by rotating the phone and moving it towards and away from the patient. The app is shown to be enable free choice of viewing plane in a much easier and faster manner than standard medical image viewing software on a computer workstation. We envision this app to enable lower-cost, more mobile viewing of MRI data, potentially in combination with portable, low-field MR hardware. |
| 3653 | Development of a method for Bloch image simulation of living tissues | |
| Katsumi Kose1, Ryoichi Kose1, Yasuhiko Terada2, Daiki Tamada3, and Utaroh Motosugi3 | ||
1MRIsimulations, Tokyo, Japan, 2University of Tsukuba, Tsukuba, Japan, 3University of Yamanashi, Chuo, Japan |
||
A method for Bloch image simulation of living tissues was developed and the simulated results were compared with experiments performed for phantoms including T2* distribution and various chemical components. The Bloch image simulation for living tissues is based on a four-dimensional numerical phantom consisting of spatially defined 3D datasets of proton density, T1, and T2 with spatially homogeneous B0 (frequency) offset. The multiple gradient-echo images reconstructed from the MR signal obtained from the Bloch image simulation of the numerical phantom reproduced experimental results for samples that simulated living tissues. |
| 3654 | Development of a management system for radiology–common data model (R-CDM) and its application in the liver disease: extension of OMOP-CDM | |
| Min-Gi Pak1, Seong-Min Han2, ChungSub Lee3, SeungJin Kim1, Tae-Hoon Kim3, Chang-Won Jeong3, and Kwon-Ha Yoon3,4 | ||
1Medical Science, Wonkwang University, Iksan, Republic of Korea, 2Computer Software Engineering, Wonkwang University, Iksan, Republic of Korea, 3Medical Convergence Research Center, Wonkwang University, Iksan, Republic of Korea, 4Radiology, Wonkwang University, Iksan, Republic of Korea |
||
The Observational Medical Outcomes Partnership-Common Data Model (OMOP-CDM) used in distributed research networks has low coverage of clinical data and does not reflect the latest trends of precision medicine. Radiology data have great merits to visual and identify the lesions in specific diseases. However, radiology data should be shared to obtain the sufficient scale and diversity required to provide strong evidence for improving patient care. Our study was to develop a web-based management system for radiology-CDM (R-CDM), as an extension of the OMOP-CDM, and to evaluate the feasibility of R-CDM dataset for application of radiological image data in clinical practice. |
| 3655 | Assessing generalized multi-pool exchange tissue model MRI simulations for the modelling of an ultra-low field scanner | |
| Ignacio Xavier Partarrieu1, Elliot Fox1,2, Frederic Brochu1, Matt Cashmore1, Neha Koonjoo3, Sheng Shen3, David Sinden1, Matthew Rosen3, and Matt G. Hall1,4 | ||
1National Physical Laboratory, Teddington, United Kingdom, 2Durham University, Durham, United Kingdom, 3Harvard, Boston, MA, United States, 4University College London, London, United Kingdom |
||
The large expense and running costs of traditional 1.5T scanners have meant that as demand for scans has grown supply has not been able to keep up. Novel, ultra-low field systems aim to address this problem by producing images with some resolution losses as a trade-off for reduced costs. However, access to such devices is currently restricted, limiting the possible research output. In this work we develop and test a simulation model of 6.5 mT b-SSFP acquisitions, which we compare to actual scanner acquisitions. We find that such simulations produce qualitatively and quantitatively similar images to real scans. |
| 3656 | Development of a quantification software for body composition imaging and its assessment in the patient with sarcopenia | |
| SeungJin Kim1, Min-Gi Pak1, Tae-Hoon Kim2, Chang-Won Jeong2, and Kwon-Ha Yoon2,3 | ||
1Medical Science, Wonkwang University, Iksan, Republic of Korea, 2Medical Convergence Research Center, Wonkwang University, Iksan, Republic of Korea, 3Radiology, Wonkwang University, Iksan, Republic of Korea |
||
In 2016, sarcopenia has been classified by the international classification of diseases(ICD-10-CM), with the code(M62.84). The role of imaging techniques has rapidly increased in the field of sarcopenia. In the past decade, the importance of muscle and fat mass has been emphasized on imaging evaluation of body composition including of muscle and body fat such as visceral fat or subcutaneous fat. However, there are diverse quantification methods for assessing muscle and fat mass by imaging and thus, these methods must be standardized. This study developed a customized quantification software based on ImageJ-platform and evaluated in the patient with sarcopenia. |
| 3657 | Development and validation of a novel finite element inversion method and analysis pipeline for anisotropic MR Elastography | |
| Behzad Babaei1, Daniel Fovargue2, David Nordsletten2,3, and Lynne Bilston1,4 | ||
1Neuroscience Research Australia, Sydney, Australia, 2King's College London, London, United Kingdom, 3Biomedical Engineering, University of Michigan, Ann Arbor, MI, United States, 4Prince of Wales Clinical School, University of New South Wales, Randwick, Australia |
||
MR elastography is well established for measuring the mechanical properties of soft tissues in a broad range of clinical populations. However, unlike liver tissue, which is generally considered to be isotropic, muscle is significantly anisotropic, with stiffer mechanical behaviour along the muscle fibre directions. Identifying anisotropic properties is challenging due to the need to track fibre directions and additional numerical difficulties in estimating properties. Here, we describe a novel finite element inversion method for estimating anisotropic mechanical properties of ex vivo tissue data, and validate this method using three sets of simulated data. |
| 3658 | Automatic Longitudinal Assessment of Hippocampal Subfields using Multi-Contrast MRI | |
| Thomas B Shaw1, Steffen Bollmann1, Ashley York1, Maryam Ziaei1, and Markus Barth1 | ||
1The University of Queensland, Brisbane, Australia |
||
Characterising longitudinal changes in the subfields of the hippocampus using high-resolution MRI is important for understanding neurological disorders and diseases including Alzheimer’s. Hippocampus subfields are differentially involved in these disorders and can be used as biomarkers for disease. Here, we introduce an automatic longitudinal pipeline for measuring the subfields of the hippocampus (LASHiS). Unlike other longitudinal pipelines, LASHiS harnesses multiple MR contrasts and is therefore more sensitive to changes in these subfields. We compared LASHiS with four other established methods including Freesurfer and found that our method yields higher test-retest reliability and robust Bayesian longitudinal linear mixed effects modelling results. |
| 3659 | Rapid Opensource Minimum Spanning TreE AlgOrithm for Phase Unwrapping (ROMEO) | |
| Simon Daniel Robinson1,2,3, Korbinian Eckstein1, Siegfried Trattnig1, Karin Shmueli4, and Barbara Dymerska4 | ||
1Department of Biomedical Imaging and Image-guided Therapy, Medical University of Vienna, Vienna, Austria, 2Department of Neurology, Medical University of Graz, Graz, Austria, 3Centre for Advanced Imaging, University of Queensland, Queensland, Australia, 4University College London, London, United Kingdom |
||
We present a phase unwrapping algorithm (ROMEO), which improves on previous path-based unwrapping methods by using weights (the values of which determine the order in which voxels are unwrapped) that guide the unwrapping on paths through the object which avoid noise and a computationally efficient minimum spanning tree algorithm. ROMEO was tested against the region-growing method PRELUDE and the voxel-based method Best Path in unwrapping simulated data (a complex topography) and in-vivo GRE and EPI data acquired at 7T. ROMEO was found to be faster and more reliable than PRELUDE and Best Path. |
| 3660 | Measures of lesion load on white matter fiber bundles for damage assessment | |
| Guillem Garcia1, David Moreno-Dominguez1, Marc Ramos 1, and Matt Rowe1 | ||
1Medical Imaging, QMENTA Inc., Boston, MA, United States |
||
Measures of lesion load are useful to study how damage in the white matter present in different pathologies relate to changes in cognitive function. So far, most research focus on global or local volumetric metrics, an approach that exhibits limitations in cases where small lesions in specific places cause major damages. In this work, we propose the combined use of a set of metrics that measure different aspects of the lesions over the major white matter bundles. We expose how these metrics provide complementary information, and discuss how its usefulness could be assessed in future work. |
| 3661 | Retrospective MRI non-uniformity correction: quantitative assessment of two methods | |
| Artem Mikheev1, Louisa Bokacheva1, Heesoo Yang1, Jeremy Sobel1, Carlos Fernandez-Granda2, Hersh Chandarana1, and Henry Rusinek1 | ||
1Radiology, NYU School of Medicine, New York City, NY, United States, 2Courant Institute of Mathematical Science, New York University, New York City, NY, United States |
||
We have implemented a method (BiCal) for correction of image nonuniformity. Using objective criteria we have compared BiCal to widely used N4 algorithm in several challenging clinical MRI applications. There was a significant advantage of BiCal over N4 for 7T brain MRI and for accelerated radial GRASP of the abdomen. The performance of BiCal and N4 were comparable in breast imaging. |
| 3662 | Improving spatial normalization of functional MRI data of the spinal cord using cerebrospinal fluid segmentation | |
| Benjamin De Leener1,2, Linda Soltrand Dahlberg2, Ali Khatibi2,3, Nawal Kinany2,4, and Julien Doyon2 | ||
1Department of computer engineering and software engineering, Polytechnique Montreal, Montreal, QC, Canada, 2Montreal Neurological Institute, McGill University, Montreal, QC, Canada, 3Center of Precision Rehabilitation for Spinal Pain (CPR Spine), University of Birmingham, Birmingham, United Kingdom, 4Center for Neuroprosthetics, École Polytechnique Fédérale de Lausanne, Lausanne, Switzerland |
||
Analyzing functional MRI of the spinal cord is challenging due to large susceptibility artifacts and inter-subject variability in terms of shape and curvature. More particularly, the poor contrast between the spinal cord and cerebrospinal fluid (CSF) leads to co-registration errors when analyzing the spinal cord functional activity at the group level. This study proposes a new registration framework that leverages the contrast between the CSF and its surrounding structure. Results show an increase of 67% in the number of active voxels at the group level, with an increase of 5% and 23% for the mean and max z-score, respectively. |
| 3663 | Fully Automated Quantitative Assessment of the Lumbar Intervertebral Disk | |
| Tom Hilbert1,2,3, Marcus Raudner4, Markus Schreiner4,5, Anna Szelenyi4, Vladimir Juras4, Siegfried Trattnig4, and Tobias Kober1,2,3 | ||
1Advanced Clinical Imaging Technology, Siemens Healthcare, Lausanne, Switzerland, 2Department of Radiology, Lausanne University Hospital and University of Lausanne, Lausanne, Switzerland, 3LTS5, École Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland, 4High Field MR Centre, Department of Biomedical Imaging and Image-guided Therapy, Medical University of Vienna, Vienna, Austria, 5Department of Orthopaedics and Trauma Surgery, Medical University of Vienna, Vienna, Austria |
||
Quantitative imaging can detect subtle changes invisible to the naked eye and enables establishing healthy normative ranges. We report a comprehensive pipeline for fully automated assessment of the lumbar intervertebral disk based on a single fast T2-mapping acquisition of 2:27 min. In this proof-of-concept study, norm values are derived from 19 healthy subjects. Feasibility is shown in 19 patients: three regions are automatically segmented in each intervertebral disk and compared to corresponding norm values. Results are then compiled in a report where abnormal regions are flagged. |
| 3664 | Synthetic T2-weighted contrasts of the lumbar spine derived from fast quantitative mapping – the icing on the cake | |
| Marcus Raudner1, Markus Schreiner1,2, Tom Hilbert3,4,5, Tobias Kober3,5,6, Anna Szelenyi7, Vladimir Juras1, and Siegfried Trattnig1 | ||
1High Field MR Centre, Department of Biomedical Imaging and Image-guided Therapy, Medical University of Vienna, Vienna, Austria, 2Department of Orthopedics and Trauma Srugery, Medical University of Vienna, Vienna, Austria, 3Advanced Clinical Imaging Technology, Siemens Healthcare, Lausanne, Switzerland, 4Department of Radiology, Lausanne University Hospital and University of Lausanne, Lausanne, Switzerland, 5LTS5, Ecole Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland, 6Department of Radiology, University Hospital and University of Lausanne, Lausanne, Switzerland, 7Department of Biomedical Imaging and Image-guided Therapy, Medical University of Vienna, Vienna, Austria |
||
GRAPPATINI has already been shown to allow short acquisition times with diagnostic synthetic T2-weighted (T2w) images in the knee. This study investigates how different effective echo times influence the resulting contrast-to-noise ratio of GRAPPATINI compared to a conventional T2w turbo spin-echo (TSE) sequence in the spine. Overall, the synthetic T2w images maintained diagnostic quality while being reconstructed from highly-undersampled k-space data in 2:27 min acquisition time vs. 2:01 min for the T2w TSE with the advantage of GRAPPATINI to also offer multiple synthetic contrasts alongside robust T2 mapping. |
| 3665 | Complex semi-quantitative noise analysis for single set of MR images | |
| Yongquan Ye1, Jingyuan Lv1, Yichen Hu1, Zhongqi Zhang2, Jian Xu1, and Weiguo Zhang1 | ||
1UIH America, Inc., Houston, TX, United States, 2United Imaging Healthcare, Shanghai, China |
||
We propose a noise analysis method based on complex MRI signals. By actively inducing concomitant fluctuation into a single complex MR image, an abrupt change in a predefined signal function (e.g. signal ratio) is observed when signal intensity becomes stronger than background noise. The method is very sensitive to the existence of weak signal such as tissue-background boundaries or very low signal tissues, and can be used to semi-quantitatively distinghuish regions of signal and noise. |
| 3666 | Evaluation of a standardized quantitative method of brain Iron based on quantitative susceptibility mapping and Brainnetome Atlas | |
| Dongxue Li1, Bin Dai1, Zhenliang Xiong1, Xianchun Zeng1, Lisha Nie2, Pu-Yeh Wu2, and Rongpin Wang1 | ||
1Department of Radiology, Guizhou Provincial People's Hospital, Guiyang, China, 2GE Healthcare, MR Research, Beijing, China |
||
Delineating the region of interest manually to obtain the magnetic susceptibility values from quantitative susceptibility mapping (QSM) for diagnosis of neurological diseases is a common method. However, this approach is time-consuming and may be inaccurate due to human error. Here we introduced a standardized brain region susceptibility quantification procedure based on QSM and Brainnetome Atlas. Our results show that this method has high accuracy in measuring brain iron content. Based on this method, we found that the content of brain iron tends to increase with age. Additionally, magnetic susceptibility values of partial brain regions have sex differences should be noted. |
| 3667 | Optimizing a preprocessing pipeline for structural 7T MR analyses in FreeSurfer | |
| Giske Opheim1,2, Oula Puonti3,4, Jan Ole Pedersen5, Vincent O. Boer3, Ane Kloster1,2, Martin Prener1,2, Helle Juhl Simonsen6, Olaf B. Paulson1,2, Lars H. Pinborg1,2, and Melanie Ganz1,7 | ||
1Neurobiology Research Unit, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark, 2Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark, 3Danish Research Centre for Magnetic Resonance, Funktions- og Billeddiagnostisk Enhed, Copenhagen University Hospital Hvidovre, Copenhagen, Denmark, 4Dept. of Health Technology, Technical University of Denmark, Kongens Lyngby, Denmark, 5Philips Healthcare, Copenhagen, Denmark, 6Functional Imaging Unit, Department of Clinical Physiology, Nuclear Medicine and PET, Rigshospitalet Glostrup, Copenhagen, Denmark, 7Dept. of Computer Science, University of Copenhagen, Copenhagen, Denmark |
||
Automated cortical segmentations benefit from higher SNR and spatial resolutions on 7T MR images, but are also challenged by B1 inhomogeneities, causing faulty surface inflations primarily in the temporal lobes. We investigated how FreeSurfer outputs were affected by applying eight different preprocessing schemes prior to reconstructions of submillimeter 7T MPRAGE images. The highest segmentation robustness across subjects was obtained by setting bias-field correction FWHM to 60mm and adding light regularization, and additionally performing intensity normalization. |
| 3668 | MRI-based Body Composition Analysis – Reproducibility and Repeatability | |
| Magnus Borga1,2, André Ahlgren2, Thobias Romu2, Per Widholm2,3, Olof Dahlqvist Leinhard2,3, and Janne West2,3 | ||
1Department of Biomedical Engineering and Center for Medical Image Science and Visualization (CMIV), Linköping University, Linkoping, Sweden, 2AMRA Medical, Linköping, Sweden, 3Department of Medicine and Health and Center for Medical Image Science and Visualization (CMIV), Linköping University, Linköping, Sweden |
||
This study investigated between-scanner reproducibility and repeatability of a method for MRI-based body composition analysis. Eighteen healthy volunteers were scanned twice on five different MR scanners from three vendors including both 1.5 and 3T, using a 2-point Dixon neck-to-knee imaging protocol. Visceral adipose tissue, abdominal subcutaneous adipose tissue, thigh muscle volume, muscle fat infiltration and liver fat were quantified using AMRA® Researcher. The reproducibility coefficients for the volume measurements ranged from 7 cl (anterior thigh muscles) to 4.2 dl (abdominal subcutaneous adipose tissue). The reproducibility coefficient for muscle fat infiltration was 1.5 percentage points and for liver fat 2.4 pp. |
| 3669 | Quantification of brain age using high-resolution 7T MR imaging and implications in major depressive disorder | |
| Gaurav Verma1, Yael Jacob1, Laurel Morris2, Priti Balchandani1, and James Murrough2 | ||
1Radiology, Icahn School of Medicine at Mount Sinai, NEW YORK, NY, United States, 2Psychiatry, Icahn School of Medicine at Mount Sinai, NEW YORK, NY, United States |
||
A regression model estimating brain-age from about 250 T1-weighted imaging features was developed using data from 29 healthy controls (mean age 39.8). The model estimated brain age with average absolute error of 6.0 years. The model was applied 35 patients (mean age 38.7) diagnosed with major depressive disorder (MDD), yielding similar performance (7.6 years mean absolute error), but showed trend of over-estimation of average brain-age by 2.4 years. This technique demonstrates the feasibility of brain-age estimation using imaging features, and may help assess the differential effects of pathology like MDD in the aging process. |
| 3670 | Color deconvolution and optical density of Multiple Sclerosis lesions | |
| Jerry Kang1,2, Kelly Gillen1, and Yi Wang1 | ||
1Weill Cornell Medicine, New York, NY, United States, 2Bronx High School of Science, Bronx, NY, United States |
||
Multiple sclerosis (MS) is a chronic inflammatory disease of the central nervous system characterized by focal inflammatory demyelination. In chronic active lesions, microglia and macrophages may contain high amounts of iron and express markers of pro-inflammatory polarization, driving tissue damage and disease progression. Therefore, studying the mechanisms behind iron accumulation and microglial inflammation are of great clinical importance, but require rigorous histological characterization of autopsied brain tissue from MS patients. We developed a rapid, automated method to quantify histopathological markers in human tissue and then validated our findings with manual counting and quantitative susceptibility mapping (QSM). |
| 3671 | Differentiating Predominant Gait Disorder Parkinsonism Using Automated Geometric Indices | |
| Ling Yun Yeow1, Bhanu Prakash KN1, AJY Lee2, EK Tan3,4, and LL Chan2,4 | ||
1Signal Image Processing Group, Singapore BioImaging Consortium, A*STAR, Singapore, Singapore, 2Diagnostic Radiology, Singapore General Hospital, Singapore, Singapore, 3National Neuroscience Institute – SGH Campus, Singapore, Singapore, 4Duke-NUS Medical School, Singapore, Singapore, Singapore |
||
Gait apraxia has attributed to ventriculomegaly and periventricular leukoariaosis. Geometric quantification of ventriculomegaly using Evans’ Index (EI) and Callosal Angle (CA) have been proposed as biomarkers of normal pressure hydrocephalus (NPH). The novel Splenial Angle (SA) may aid in differentiating healthy controls (C) and Parkinson's Disease (P) patients from those with postural instability and gait difficulty (G) subtype and NPH. CA and SA are significantly lower in G patients compared to C and P while EI is significantly higher. Automation of these geometric measures is relatively accurate for EI and SA but further improvements are needed for CA. |
| 3672 | Post-acquisition image processing can reduce artifacts in UTE images | |
| Yang Xia1, Farid Badar1, and Simon Miller1 | ||
1Physics, Oakland University, Rochester, MI, United States |
||
Imperfect compensation for the eddy currents can cause artifacts in UTE images. Several post-acquisition processing methods can reduce artifacts in the images that are acquired in the radial k-space trajectories. |
| 3673 | Evaluation of quantitative MRI parameters reproducibility across a major scanner upgrade: the example of T1 | |
| Ratthaporn Boonsuth1, Marco Battiston1, Francesco Grussu1,2, Marios C. Yiannakas1, Torben Schneider3, Rebecca S. Samson1, Ferran Prados1, and Claudia A. M. Gandini Wheeler-Kingshott1,4,5 | ||
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, 2Centre for Medical Image Computing, Department of Computer Science, University College London, London, United Kingdom, 3Philips Healthcare, Guildford, Surrey, United Kingdom, 4Brain MRI 3T Research Centre, IRCCS Mondino Foundation, Pavia, Italy, 5Department of Brain and Behavioural Sciences, University of Pavia, Pavia, Italy |
||
Major MRI scanner upgrades are generally required to improve performance and image quality; however, they can also potentially introduce systematic changes in quantitative MRI (qMRI) metrics and affect their accuracy and precision. To date the evaluation of the effect of scanner upgrades has focussed mainly on volumetric measurements, whereas the effect on quantitative parametric maps remains unexplored, especially when comparing analog and digital signal pathways plus multiband. Here we report findings on quantitative T1 to assess the potential effect of scanner upgrades on qMRI. We found negligible differences, suggesting that T1 measurements remain stable following a major scanner upgrade. |
| 3674 | Quantitative MRI Metrics in Routine Clinical Practice: A Validation Study from a Large Heterogeneous Cohort of Multiple Sclerosis Patients | |
| Adrian Tsang1, Mário J Fartaria2,3,4, Rodrigo D Perea1, Ricardo Corredor-Jerez2,3,4, Shirley Liao1, Tammie L.S. Benzinger5, Maria Laura Belfari1, Peter A Calabresi6, Carrie M Hersh7, Till Huelnhagen2,3,4, Stephen E Jones8, Hagen H Kitzler9, Nicholas Levitt1, Yvonne W Lui10, Sara J Makaretz1, Robert Naismith5, Kunio Nakamura8, Dan Ontaneda8, Stephen Rao8, Alex Rovira11, Madalina E Tivarus12, James R Williams1, Richard A Rudick1, Tobias Kober2,3,4, and Elizabeth Fisher1 | ||
1Biogen, Cambridge, MA, United States, 2Advanced Clinical Imaging Technology, Siemens Healthcare AG, Lausanne, Switzerland, 3Department of Radiology, Lausanne University Hospital and University of Lausanne, Lausanne, Switzerland, 4LTS5, École Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland, 5Washington University in St. Louis, St. Louis, MO, United States, 6Department of Neurology, Johns Hopkins University School of Medicine, Baltimore, MD, United States, 7Cleveland Clinic Lou Ruvo Center for Brain Health, Cleveland, OH, United States, 8Cleveland Clinic, Cleveland, OH, United States, 9Center of Clinical Neuroscience, 'Carl Gustav Carus' University Hospital, Technische Universitaet, Dresden, Germany, 10New York University, New York, NY, United States, 11Vall d’Hebron University Hospital, Barcelona, Spain, 12University of Rochester Medical Center, Rochester, NY, United States |
||
Use of MRI metrics for routine monitoring of MS patients has been hindered by lack of standardization, imprecise measurements, and workflow hurdles. To overcome these challenges, we developed a novel prototype, MSPie, as part of the MS PATHS initiative. The goal of this study was to assess the validity and suitability of MRI metrics automatically computed by MSPie in the MS clinical workflow. |
| 3675 | Experimental demonstration of 3D Cartesian-FRONSAC with new reconstruction approach | |
| Yanitza Marie Rodriguez1, Gigi Galiana1, Enamul Buyian1, and Todd Constable1 | ||
1Radiology and Biomedical Imaging, Yale University, New Haven, CT, United States |
||
In this work we show the first experimental images, in both phantoms and humans, from 3D Cartesian Fast Rotary Nonlinear Spatial Acquisition (FRONSAC) imaging. FRONSAC adds oscillating nonlinear gradients to a standard Cartesian sequence, such as 3D GRE or MP-RAGE, both of which were acquired for this work. The additional modulation reduces undersampling artifacts and improves response to advanced reconstruction techniques, while maintaining other desirable image characteristics. In addition, we show that a PSF-based reconstruction, a generalization of the approach used in wave imaging, can be applied to FRONSAC data, yielding faster reconstructions that can be highly parallelized. |
| 3676 | Perceptual Noise-Estimation based Method for MRI denoising with deep learning | |
| Xiaorui Xu1, Siyue Li1, Shutian Zhao1, Chun Ki Franklin Au1, and Weitian Chen1 | ||
1CUHK lab of AI in radiology (CLAIR), Department of Imaging and Interventional Radiology, The Chinese University of Hong Kong, Hong Kong, Hong Kong |
||
Most of the methods in MRI denoising derive denoised images from corrupted images directly. DnCNN is a network that is used to remove Gaussian noise from natural images. The noise distribution in MRI images are often non-Gaussian due to the latest development of reconstruction algorithm and MRI hardware. In this work, we investigated the case when the noise follows Racian distribution. We utilized the idea of DnCNN and combine it with a perceptual architecture to remove Rician noise of MRI images. We demonstrated this method can generate ideal clean images. |
| 3677 | Automated MR HIC Determination using Deep Learning and Frangi Filters | |
| Ralf Berthold Loeffler1,2, M. Beth McCarville2, Aaryani Tipirneni-Sajja2,3, Jane S Hankins4, and Claudia Maria Hillenbrand1,2 | ||
1Research Imaging NSW, University of New South Wales, Sydney, Australia, 2Diagnostic Imaging, St. Jude Children's Research Hospital, Memphis, TN, United States, 3Biomedical Engineering, University of Memphis, Memphis, TN, United States, 4Hematology, St. Jude Children's Research Hospital, Memphis, TN, United States |
||
Hepatic iron content (HIC) quantification requires segmentation. Deep learning and Frangi Filtering allow to fully automate segmentation. 664 manually segmented data sets were available for training and testing a UNET. Data sets segmented by UNET were Frangi filtered for vessel exclusion, HIC was calculated using a published calibration, and correlated with HIC obtained with manual segmentation. Very good correlation (R2 > 0.99) with a correlation line close to unity was found. Fully automated HIC quantification using deep learning and Frangi filtering can lead to significant time savings in clinical practice. |
| 3678 | A Phase unwrapping Method by Phase Partition and Local Polynomial Fitting with Application to Abdominal Quantitative Susceptibility Mapping | |
| Cheng Junying1, Xu Man1, Mei Yingjie2, Liu Liang1, Feng Yanqiu3, and Cheng Jingliang1 | ||
1Magnetic Resonance Department, The First Affiliated Hospital of Zhengzhou University, Zhengzhou, China, 2Philips Healthcare, Guangzhou, China, 3School of Biomedical Engineering, Southern Medical University, Guangzhou, China |
||
The accurate recovery of underlying true phase is vital for a large number of applications, such as water/fat separation, quantitative susceptibility mapping and brain imaging. In this work, we propose a 3D phase unwrapping method based on phase partition and local polynomial fitting. The simulated and in vivo experiments demonstrate the proposed method can obtain perfect unwrapped results even in the regions with low SNR and rapidly changed phase, and can be applied to the abdominal QSM. |
| 3679 | Fast, Automated DTI-based Thalamus Nuclei Segmentation | |
| Charles Iglehart1, Adam Bernstein2, Ted Trouard3, Craig Weinkauf4, and Manoj Saranathan5 | ||
1Department of Electrical and Computer Engineering, University of Arizona, Tucson, AZ, United States, 2College of Medicine, University of Arizona, Tucson, AZ, United States, 3Department of Biomedical Engineering, University of Arizona, Tucson, AZ, United States, 4Department of Surgery, University of Arizona, Tucson, AZ, United States, 5Department of Medical Imaging, University of Arizona, Tucson, AZ, United States |
||
Fast, accurate, and automatic thalamus segmentation is critical in evaluating the roles of individual thalamic nuclei in pathology and treatment. Many segmentation techniques developed to date involve the use of Diffusion Tensor MRI and are inhibited by their reliance upon i) time-consuming processing to produce an initial mask for the entire thalamus, and ii) orientation distribution functions incapable of modeling intricate small-scale fiber tract geometries. We present a technique that addresses these issues by i) greatly accelerating the masking process via template-based registration, and ii) using constrained spherical deconvolution to produce enhanced ODFs that drive a modified k-means clustering algorithm. |
| 3680 | Real Time MRI at 70 frames per second: Establishing protocols for maxillofacial surgery applications. | |
| Aneurin J Kennerley1, Isaac J Watson2, Lloyd A.E Bollans1, David A Mitchell3, and Angelika J Sebald1,3 | ||
1Chemistry, University of York, York, United Kingdom, 2Electronic Engineering, University of York, York, United Kingdom, 3York Cross Disciplinary Centre for Systems Analysis, University of York, York, United Kingdom |
||
We showcase a 70+ frame per second Real Time MRI protocol that is robust and flexible for a number of different applications in maxillofacial surgery (including multi-plane monitoring speech and swallowing mechanisms) and associated follow up/long term monitoring for reducing depression and improving post-operative quality of life for patients. We explore the use of different food types to act as MR contrast agents to highlight changes in oral function. The project is co-created with a consulting Oral & Maxillofacial and Head & Neck Surgeon (Mr D Mitchell). |
| 3681 | Feasibility of Dynamic Inhaled Gas MRI-based Measurements using Acceleration Factors of 10 and 14 | |
| Matthew S Fox1,2, Elise Woodward3, Marcus Couch4, Tao Li5, Iain Ball6, and Alexei V Ouriadov1,2 | ||
1Lawson Health Research Institute, London, ON, Canada, 2Physics and Astronomy, The University of Western Ontario, London, ON, Canada, 3The University of Western Ontario, London, ON, Canada, 4Montrel Neuro Institute, Montreal, QC, Canada, 5Thunder Bay Regional Research Institute,, Thunder Bay, ON, Canada, 6Philips Australia & New Zealand, North Ryde, Australia |
||
We hypothesize that the SEM equation can be adapted for fitting the gas-density dependence of the MR signal similar to fitting time or b-value dependences S(n)=exp[-(nR)β], where 0<β<1, n is the image number and R is the apparent-fractional-ventilation parameter. This interpretation allows us to consider the signal-intensity variation as reflection of the underlying gas-density variation and hence, reconstruction of the under-sampled k-space using the adapted SEM equation. Lung fractional-ventilation maps have been generated using reconstructed images. We have demonstrated the feasibility of our approach using retrospective under-sampling mimicking acceleration factors of 10 and 14 in a small animal cohort. |
| 3682 | Self-Navigated MEDIC: an improved and reliable technique for cervical spine MRI | |
| Qiong Zhang1, Yong Xiao Zhang2, Yuanyuan Kang3, and Wuyi Zhao1 | ||
1Siemens Shenzhen Magnetic Resonance, Ltd, Shen Zhen, China, 2MR Collaborations, Siemens Healthcare Ltd., Shenzhen, China, 3SW, Siemens Shenzhen Magnetic Resonance, Ltd, Shen Zhen, China |
||
In this work, we developed a navigated Multi-Echo Data Imaging Combination (MEDIC) sequence for robust cervical spine (c-spine) imaging. In this sequence, a navigator echo is acquired in each repetition period to monitor the phase instability caused by respiration-induced field variations, and such instability among multiple echoes is subsequently compensated with a linear-phase evolution model. In vivo experimental results showed that the developed navigated MEDIC sequence outperformed the conventional MEDIC sequence and might be a potential technique for the diagnosis of cervical spine on 3T |
| 3683 | Radial sampling interactions in multi-dimensional sparse encoding problems using a joint decoding-reconstruction framework | |
| Sophie Schauman1, Thomas W Okell1, and Mark Chiew1 | ||
1Wellcome Centre for Integrative Neuroimaging, NDCN, University of Oxford, Oxford, United Kingdom |
||
Many physical properties cannot be directly measured with MRI, but are instead derived from a number of encoded measurements. Novel sampling methods in these regimes generally consider how to sample these encoded signals in k or k-t space, but not how to best sample across encodings. Here we present a study into how choice of encoding protocol interacts with the sampling and how these choices affect the multi-dimesnional point spread function. Simulations on vessel-encoded ASL angiography are used to study how these choices affect image reconstruction quality. We show that jointly decoding and reconstructing improves image reconstruction fidelity. |
| 3684 | Joint Compressed Sensing and Sensitivity Estimation for Free-Breathing Whole Heart CINE MRI Without ECG Gating | |
| Jingyuan Lyu1, Jiali Zhong2, Yu Ding1, Qi Liu1, Lele Zhao3, Jian Xu1, Weiguo Zhang1, and Ruchen Peng2 | ||
1UIH America, Inc., Houston, TX, United States, 2Beijing LuHe Hospital, Capital Medical University, Beijing, China, 3United Imaging Healthcare, Shanghai, China |
||
This abstract presents a new approach to compressed sensing cardiac MRI, which enables free-breathing whole heart coverage cine imaging within 30 seconds. Using a phased array coil, data was acquired continuously along Cartesian sampling trajectories using a lookup table without ECG gating. Each slice was continuously sampled for a fixed period of time, before the slice-selective RF excitation pulse switch to the next slice. In reconstruction, the approach jointly updates coil sensitivity maps and images, integrated with compressed sensing. In post-processing, virtual ECG is calculated based on unsupervised machine learning. |
| 3685 | A Combined Approach of Variable Flip Angle, Keyhole and Averaging (CAVKA) for Accelerating the Acquisition of a low SNR Image Series | |
| Hen Amit Morik1, Patrick Schuencke1, and Leif Schröder1 | ||
1Leibniz-Forschungsinstitut für Molekulare Pharmakologie (FMP), Berlin, Germany |
||
We propose a method that is a Combined Approach of Variable Flip Angle (VFA), Keyhole undersampling and Averaging (CAVKA). It is designed to optimize the use of the limited magnetization and to accelerate the acquisition in MRI series that suffer from low SNR and thus require averaging. The method is applied to the acquisition of a CEST (chemical exchange saturation transfer) image series, where the sensed nucleus is hyperpolarized 129Xe. There it provides ~4-fold SNR increase compared to conventional imaging without averaging or 7-fold acceleration with the same SNR compared to imaging with averaging. |
| 3686 | Partial Fourier Readout in Readout Segmentation of Long Variable Echo Trains-Based Imaging In Nasopharyngeal Carcinoma | |
| Zilong Yuan1, Hao Chen1, Yaoyao He1, Huiting Zhang2, Xiaofang Guo1, Zhaoxi Zhang1, Yulin Liu1, and Robert Grimm3 | ||
1Radiology, Hubei Cancer Hospital, Tongji Medical College, Huazhong University of Science and Technology, wuhan, China, 2MR Scientific Marketing, Siemens Healthcare, Wuhan, China, 3MR Application Predevelopment，Siemens Healthcare, Erlangen, Germany |
||
There is a potential influence of partial Fourier readout on readout segmentation of long variable echo trains (RESOLVE)-based intravoxel incoherent motion (IVIM) and diffusion kurtosis imaging (DKI) models. The aim of this study was to investigate the influence of partial Fourier readout on RESOLVE-based IVIM and DKI parameters in nasopharyngeal carcinoma. The results showed that partial Fourier readout has limited influence on DKI parameters, but significant influence on the perfusion parameters of IVIM. |
| 3687 | Connectivity-based Graph Convolutional Network for fMRI Data Analysis | |
| Lebo Wang1, Kaiming Li2, and Xiaoping Hu1,2 | ||
1Department of Electrical and Computer Engineering, University of California, Riverside, Riverside, CA, United States, 2Department of Bioengineering, University of California, Riverside, Riverside, CA, United States |
||
Graphs have been widely applied for ROI-based fMRI data analysis, in which the functional connectivity (FC) between all pairs of regions is thoroughly considered. Combined with convolutional neural networks, we define graphs based on FC and introduce a connectivity-based graph convolution network (cGCN) architecture for fMRI data analysis. cGCN allows us to extract spatial features within connectivity-based neighborhood for each frame and capture the temporal dynamics between frames. Our results indicate that cGCN outperforms traditional deep learning architectures on fMRI data analysis. |
| 3688 | MRI with Sub-millisecond Temporal Resolution over a Reduced Field of View | |
| Zheng Zhong1,2, Kaibao Sun2, Guangyu Dan1,2, Muge Karaman1,2, and Xiaohong Joe Zhou1,2,3,4 | ||
1Bioengineering, University of Illinois at Chicago, Chicago, IL, United States, 2CMRR, University of Illinois at Chicago, Chicago, IL, United States, 3Radiology, University of Illinois at Chicago, Chicago, IL, United States, 4Neurosurgery, University of Illinois at Chicago, Chicago, IL, United States |
||
Sub-millisecond Periodic Event Encoded Dynamic Imaging or SPEEDI (also known as SMILE) has been reported to be capable of achieving sub-millisecond temporal resolution. However, the total scan times of this technique is typically very long. In this study, we have incorporated two techniques – reduced field of view (rFOV) and echo-train acquisition – into a SPEEDI sequence to substantially reduce the total scan times. This sequence, which we call re-SPEEDI, has been demonstrated in phantom experiments for capturing rapidly changing currents in a wire loop with a temporal resolution of 0.6ms-0.8ms. |
| 3689 | Optimal Flip Angle Formula for SWIFT and CEA (Concurrent Excitation and Acquisition) MRI | |
| Serhat Ilbey1, Michael Garwood2, Michael Bock1, and Ali Caglar Özen1,3 | ||
1Dept. of Radiology, Medical Physics, Medical Center – University of Freiburg, Freiburg, Germany, 2Center for Magnetic Resonance Research and Department of Radiology, University of Minnesota, Minneapolis, MN, United States, 3German Consortium for Translational Cancer Research Partner Site Freiburg, German Cancer Research Center (DKFZ), Heidelberg, Germany |
||
MR imaging of tissue components with extremely fast transverse relaxation times require advanced MRI techniques such as continuous SWIFT (cSWIFT), Concurrent Excitation and Acquisition (CEA), and full duplex MRI. Frequency modulated pulses are used in these methods. A steady-state signal model with transverse relaxation term was formulated and solved for an arbitrary frequency and phase modulated RF pulse. An accurate expression for the optimal flip angle was derived. The performance of the new formulation was verified by a home-made Bloch equation simulator for T2 values between 10µs and 100ms. |
| 3690 | Mapping oxygen extraction fraction using a tailored parallel transmission RF pulse at 7 T | |
| Yan Tong1, Peter Jezzard1, Caitlin O'Brien1, and William T Clarke1 | ||
1Wellcome Centre for Integrative Neuroimaging, FMRIB Division, NDCN, University of Oxford, Oxford, United Kingdom |
||
T2‐relaxation under‐spin‐tagging (TRUST) is a robust spin tagging based method to quantify oxygen extraction fraction (OEF), but it lacks spatial specificity. Recently O’Brien et al. proposed a method involving multiple saturation pulses to achieve spatial specificity. Parallel transmission (pTx) provides additional degrees of freedom for spatial localisation. A pTx RF pulse design strategy based on a shells trajectory is applied to perform regional OEF measurement at 7 T. Initial in-vivo results acquired from one healthy subject showed that spatial localisation of OEF could be achieved. |
| 3691 | Simultaneous multi-slice imaging reduces sensitivity of local-SAR to patient motion at 7T | |
| Emre Kopanoglu1, Cem M. Deniz2, and Richard G. Wise1,3 | ||
1CUBRIC, School of Psychology, Cardiff University, Cardiff, United Kingdom, 2Department of Radiology, New York University Langone Health, New York, NY, United States, 3Institute for Advanced Biomedical Technologies, University of Chieti-Pescara, Chieti, Italy |
||
This study investigates the effect of within-scan patient motion on local-SAR for simultaneous multi-slice (SMS) imaging at 7T. A virtual body model was simulated at 104 different positions. 1-/2-/3-spokes pulses were designed to excite a region of 60 slices covering the cerebellum and the brain, using SMS-factors of 1 through 5. Local-SAR was observed to increase by up to 2.75-fold due to patient motion. Pulses with higher SMS-factors were up to 50% less sensitive against changes in local-SAR due to patient motion, compared to SMS:1 pulses. Pulses with higher SMS-factors yielded more consistent local-SAR throughout the scan. |
| 3692 | Designing B0 robust adiabatic multi-band inversion pulses with high time-bandwidth products and smooth slice selective gradients | |
| Christoph Stefan Aigner1 and Sebastian Schmitter1 | ||
1Physikalisch-Technische Bundesanstalt (PTB), Braunschweig and Berlin, Germany |
||
We demonstrate the design of adiabatic multi-band inversion pulses with high time-bandwidth products and acceptable RF power demands. The proposed utilization of GOIA single-band pulses results in smooth slice selective gradients with highly reduced RF power and matched multi-band RF waveforms. Phantom and in vivo experiments at 3T and 7T validate the simulations and demonstrate robust inversion of multiple slices over a wide range of both B1+ and B0 values. |
| 3693 | Inversion Pulses with B1-Robustness and Reduced Energy by Optimal Control | |
| Christina Graf1, Christoph Stefan Aigner2, Armin Rund3, Andreas Johann Lesch1, and Rudolf Stollberger1 | ||
1Institute of Medical Engineering, Graz University of Technology, Graz, Austria, 2Physikalisch-Technische Bundesanstalt (PTB), Braunschweig and Berlin, Germany, 3Institute of Mathematics and Scientific Computing, University of Graz, Graz, Austria |
||
The aim of this work is to design slice-selective inversion RF pulses that are robust among $$$B1$$$-variations while the pulse energy is reduced. For that purpose, an optimal control framework based on an ensemble formulation is introduced. The numerically optimized RF pulses showed an excellent performance compared to the target magnetization for a broad range of $$$B1$$$-scalings. Phantom measurements were performed using a 32 channel head coil for receive and a birdcage body coil for transmit and revealed excellent inversion profiles. |
| 3694 | Fully embedded parallel transmission: Stable performances using universal pulses with fast online customization | |
| Jürgen Herrler1, Patrick Liebig2, Rene Gumbrecht2, Manuel Schmidt1, Michael Uder3, Andreas Maier4, Arnd Dörfler1, and Armin Nagel3,5 | ||
1Institute of Neuroradiology, University Hospital Erlangen, Erlangen, Germany, 2SIEMENS Healthineers, Erlangen, Germany, 3Institute of Radiology, University Hospital Erlangen, Erlangen, Germany, 4Friedrich Alexander University Erlangen Nürnberg, Erlangen, Germany, 5Institute of Medical Physics, Friedrich Alexander University Erlangen Nürnberg, Erlangen, Germany |
||
To homogenize signals 7 Tesla MRI reliably and fast, a pTx pulse design process was established. Trajectory, energy regularisation parameter and universal pulse shapes were calculated by solving global optimization problems using a dataset of 12 subjects (B1+ profiles, B0, VOP). Pulse shaping is done online using the actual subject‘s data acquired during the sequence preparation phase which lasts about 90s. A fitting curve for SED is also provided which, combined with online calculated SED values, aims to use the full amount of tolerable SAR. Compared with the CP-Mode, the NRMSE could be substantially improved for all 36 volunteers. |
| 3695 | Optimization of adiabatic pulses for Pulsed ASL at 7T – Comparison with Pseudo-continuous ASL | |
| Kai Wang1, Xingfeng shao1, Lirong Yan1, Jin Jin1,2,3, and Danny Wang1 | ||
1Laboratory of FMRI Technology (LOFT), Mark & Mary Stevens Neuroimaging and Informatics Institute, Keck School of Medicine, University of Southern California (USC), Los Angeles, CA, United States, 2School of Information Technology and Electrical Engineering, The University of Queensland,, Brisbane, Australia, 3ARC Training Centre for Innovation in Biomedical Imaging Technology, The University of Queensland, Brisbane, Australia |
||
The goal of this work was to optimize and evaluate four types of adiabatic pulses for pulsed ASL (PASL) at 7T including Hyperbolic Secant (HS), WURST, FOCI and trFOCI pulses using theoretical simulation, phantom and in vivo scans, and compare them with pseudo-continuous ASL (pCASL). PASL with WURST pulse outperformed PASL with HS, FOCI and trFOCI in terms of labeling efficiency and residual signal, and also showed higher labeling efficiency compared with pCASL, and thus is recommended for 7T perfusion measurement. |
| 3696 | Tailored 3D Inner Volume Suppression Pulses for MR Corticography | |
| Jun Ma1, Xinqiang Yan2, Bernhard Gruber3, Jonathan Martin1, Zhipeng Cao2, Jason Stockmann3, Kawin Setsompop3, and William Grissom1 | ||
1Biomedical Engineering, Vanderbilt University, Nashville, TN, United States, 2Department of Radiology and Radiological Sciences, Vanderbilt University Medical Center, Nashville, TN, United States, 3Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Charlestown, MA, United States |
||
To fully utilize the potential of a new cortical imaging-optimized 7T scanner with Gmax= 200 mT/m & Smax=700 T/m/s gradients, 128 Rx channels, 16 Tx channels, and an AC/DC B0 shim array, we describe a tailored parallel transmit inner-volume-suppression (IVS) pulse design, which can enable highly accelerated functional and diffusion imaging by reducing g-factor and suppressing physiological noise from ventricle CSF. 3D IVS pulses were designed using simulated B1+ maps for the scanner’s 24-element coil. Simulation results demonstrated uniform inner volume suppression for 3D and 2D imaging. A 2D in-vivo IVS pulse design experiment demonstrated IVS’s ability to reduce g-factor. |
| 3697 | The Effects of Coupled B1 Fields in B1 Encoded TRASE MRI – A Simulation Study | |
| Pallavi Bohidar1, Hongwei Sun2, Jonathan C. Sharp2, and Gordon E. Sarty1 | ||
1Division of Biomedical Engineering, University of Saskatchewan, Saskatoon, SK, Canada, 2Department of Oncology, University of Alberta, Edmonton, AB, Canada |
||
Transmit Array Spatial Encoding (TRASE) is a novel MRI technique that achieves spatial encoding by introducing phase gradients in the transmit RF magnetic field (B1). In this study, Bloch simulations were performed to investigate and study the effects of B1 field perturbations arising from inductive coupling among RF coils for 2D TRASE imaging. Simulations show that a flip angle contribution of ~95% or higher from the primary (driven) transmit coil is required for 2D TRASE MRI. This result is of crucial importance for designers of practical TRASE transmit array systems. |
| 3698 | Motion Robust Parallel Transmission Excitation Pulse Design for Ultra-High Field MRI | |
| Luke Watkins1, Alix Plumley1, Kevin Murphy1, and Emre Kopanoglu1 | ||
1Cardiff University Brain Research Imaging Centre, Cardiff, United Kingdom |
||
Within-scan patient motion hampers the performance of parallel-transmit (pTx) pulses. In this study we developed a motion-robust pTx pulse (MRP) in the small tip angle regime to improve magnitude homogeneity over a range of pitch/roll/yaw rotations from 0° to 5°. The 4-spoke MRP was compared to a 3-spoke conventionally designed pulse (CDP). The MRP produced more homogenous magnitude profiles than the CDP for all evaluated rotations (0°,1°,2°,5° around each axis). The benefit of the method was more prominent with increased rotation angles. The proposed design shows potential for excitation pulses that maintain highly homogeneous magnitude profiles during patient head motion. |
| 3699 | Universal Parallel Transmit Pulse Design for 3-D Local-Excitation based on different sized databases of B0/B1+-maps – A 7T Study | |
| Ole Geldschläger1, Tingting Shao1, Jürgen Herrler2, Armin Nagel3,4, and Anke Henning1,5 | ||
1High-field Magnetic Resonance, Max-Planck-Institut for biolog. Cybernetics, Tübingen, Germany, 2Institute of Neuroradiology, University Hospital Erlangen, Erlangen, Germany, 3Institute of Radiology, University Hospital Erlangen, Erlangen, Germany, 4Institute of Medical Physics, Friedrich Alexander University Erlangen Nürnberg, Erlangen, Germany, 5Advanced Imaging Research Center, University of Texas Southwestern Medical Center, Dallas, TX, United States |
||
This study investigates universal parallel-transmission (pTx) radio-frequency-pulses for 3-dimensional local-excitation designed on different sized databases of B0/B1+-maps from human heads at 7T. Thus, it prospectively abandons the need for time-consuming subject specific B0/B1+-mapping and pTx-pulse calculation during the scan session. For the proposed calculation routine, the design-database does not need to include more than five heads, to achieve a pTx-pulse that excites the same 3-dimensional local-excitation target-pattern on the tested 40 different heads. The resulting universal pulses created magnetization-profiles with (in most cases) an only marginally worse Normalized-Root-Mean-Square-Error compared to the magnetization-profiles produced by pulses tailored to individual heads. |
| 3700 | Large tip-angle, motion robust pulse design for parallel transmission at 7T using composite B1 distributions. | |
| Alix Plumley1, Luke Watkins1, and Emre Kopanoglu1 | ||
1CUBRIC, School of Psychology, Cardiff University, Cardiff, United Kingdom |
||
Parallel-transmit (pTx) pulses can help overcome B1 inhomogeneity at ultra-high field, however their performance is sensitive to geometry, orientation and composition of the coil load. Head motion in pTx effectively alters the load, damaging pulse performance and reintroducing artificial contrast to images. Here, we introduce a versatile method to reduce in-plane motion sensitivity of large tip-angle, parallel transmit pulses by designing pulses based on weighted-average B1-distributions. These include B1-distributions from multiple head positions to effectively expand the area over which the pulse can function. Error in resulting inversion profiles is reliably reduced following most in-plane translations and rotations, demonstrating motion-robustness. |
| 3701 | Sub-µs gradient delay correction in tilted slab-selective bipolar multi-spoke RF pulses using trim blips | |
| Redouane Jamil1, Vincent Gras1, Franck Mauconduit1, and Nicolas Boulant1 | ||
1CEA, Université Paris-Saclay, NeuroSpin, Gif sur Yvette, France |
||
Bipolar spokes RF pulses can be employed in parallel transmission (pTX) used to mitigate the RF field inhomogeneity problem at ultra-high field in 2D. However, their performance can dramatically drop with delays between gradients and RF pulses. This work reports a new method for (anisotropic) gradient delay correction for multi-spoke RF pulses based on so-called gradient trim blips, applicable to tilted slice and slab excitations. This solution has been tested in Bloch simulation and validated on phantom at 7T with tilted slab-selective pulses. Experimental results obtained with the corrected pulses match the simulations incorporating no gradient delay. |
| 3702 | 3D Inner Volume MR Fingerprinting with Parallel Transmission | |
| Xiaoxuan He1, Naoharu Kobayashi1, Myung Kyun Woo1, Edward J. Auerbach1, Xiaoping Wu1, and Gregory J. Metzger1 | ||
1Center for Magnetic Resonance Research, University of Minnesota, Minneapolis, MN, United States |
||
In this simulation study, we explored the feasibility of incorporating parallel transmission to achieve a 3D inner volume MR fingerprinting with an aim to mitigate field inhomogeneities at UHF and reduce field of view (rFOV) for high resolution acquisitions and improved parameter estimation. Our preliminary results showed uniform and consistent contrast within the rFOV. We are currently working on the implementation of the method for further experimental validations. |
| 3703 | Rapid pre-saturated TFL transmit field mapping with an optimized 3D centric single-shot readout | |
| Dario Bosch1, Jonas Bause1, Philipp Ehses2, Moritz Zaiss1,3, and Klaus Scheffler1,4 | ||
1High Field Magnetic Resonance Center, Max Planck Institute for Biological Cybernetics, Tuebingen, Germany, 2German Center for Neurodegenerative Diseases, Bonn, Germany, 3Department of Neuroradiology, University Hospital Erlangen, Erlangen, Germany, 4Institute for Biomedical Magnetic Resonance, University Hospital Tuebingen, Tuebingen, Germany |
||
Robust and fast measurements of the transmit field strength is of great importance particularly in parallel transmit applications at ultra-high field. In this work, a pre-saturation TurboFLASH B1+ mapping sequence was optimized for 3-dimensional single-shot acquisition with spiral-centric reordering. Improved SNR and reduced artifacts were achieved by using a variable flip-angle readout. The ability of the proposed sampling scheme to perform fast whole-brain B1+ mapping with low SAR was demonstrated in phantom and in-vivo experiments at 9.4 T. The obtained B1+ maps were comparable to those obtained with the conventional 2D-SatTFL approach but required significantly less scan time. |
| 3704 | TRASE Enforced and Accelerated Nonlinear Spatial Encoding for Low-field Portable MRI and its Local k-space Analysis | |
| GONG JIA1, ZHOU WENSHEN1, YU WENWEI2, and HUANG SHAO YING1 | ||
1EPD, Singapore University of Technology, Singapore, Singapore, 2Chiba University, Chiba, Japan |
||
We propose an encoding method which combines Transmit Array Spatial Encoding (TRASE) and spatial encoding magnetic field (SEM) to improve the image quality in a permanent-magnet-array (PMA)-based low-field portable MRI system with acceleration. TRASE is used to introduce phase shift to re-arrange the signal points in local k-spaces, to gain more information quicker to increase imaging quality and speed. A significant quality improvement can be achieved in the reconstructed images, especially in the central regions, which is shown numerically. The number of rotation angles is reduced 80% for the same image quality. The experiments are being conducted for a validation. |
| 3705 | Fat Suppressed Magnetization Transfer Contrast Imaging Near Metal | |
| Philip Kenneth Lee1,2, Daehyun Yoon2, and Brian Andrew Hargreaves1,2,3 | ||
1Electrical Engineering, Stanford University, Stanford, CA, United States, 2Radiology, Stanford University, Stanford, CA, United States, 3Bioengineering, Stanford University, Stanford, CA, United States |
||
Imaging with fat suppression near metal-induced off-resonance has limited options due to the spectral ambiguity between fat and water. We propose Magnetization Transfer (MT) as an alternative for fat suppressed imaging near metal. We demonstrate that fat suppressed images can be efficiently obtained by modifying slice interleaving of conventional multispectral acquisitions that utilize fast spin echo readouts. These high RF power fast spin echo readouts act as MT preparation pulses for adjacent slices. Furthermore, we show that the image contrast can be predicted by Bloch simulations that incorporate the two-pool model and validate the simulations with in-vivo measurements. |
| 3706 | High resolution multi T1-weighted contrast with reduced B1 sensitivity using the FLAWS sequence at 7T | |
| Jeremy Beaumont1,2, Giulio Gambarota1, Herve Saint-Jalmes1, Oscar Acosta1, Jean-Christophe Ferré3,4, Parnesh Raniga2, Olivier Salvado5, and Jurgen Fripp2 | ||
1Univ Rennes, CRLCC Eugene Marquis, Inserm, LTSI-UMR 1099, F-35000 Rennes, France, 2The Australian e-Health Research Centre, CSIRO, Brisbane, Australia, 3Univ Rennes, Inria, CNRS, Inserm, IRISA, EMPENN ERL U-1228, F-35000 Rennes, France, 4CHU Rennes, Department of Neuroradiology, F35033 Rennes, France, 5Data 61, CSIRO, Brisbane, Australia |
||
Recent studies showed that the FLAWS sequence provides multiple co-registered T1-weighted contrasts of the brain that exhibit reduced B1 sensitivity. In particular, the FLAWS sequence allows for the generation of a standard anatomical contrast, a contrast with WM signal suppression and a GM-specific contrast. This study introduces a combination of FLAWS images to generate a new contrast that may be suitable for the detection of brain lesions at 7T. While the preliminary results of this study are promising, a further validation is required by imaging subjects with lesions and comparing the new FLAWS contrast with other MR sequences. |
| 3707 | High resolution T1 mapping with reduced B1+ sensitivity using the FLAWS sequence at 7T | |
| Jeremy Beaumont1,2, Giulio Gambarota1, Herve Saint-Jalmes1, Oscar Acosta1, Parnesh Raniga2, Olivier Salvado3, and Jurgen Fripp2 | ||
1Univ Rennes, CRLCC Eugene Marquis, Inserm, LTSI-UMR 1099, F-35000 Rennes, France, 2The Australian e-Health Research Centre, CSIRO, Brisbane, Australia, 3Data 61, CSIRO, Brisbane, Australia |
||
The MP2RAGE sequence can be used to provide 7T T1-weighted images and T1 maps with reduced transmitted bias field (B1+) sensitivity, at the cost of limiting the image resolution and contrast to noise ratio (CNR). The FLAWS sequence was derived from the MP2RAGE sequence to provide multiple T1-weigthed contrasts that were shown to be of interest for a wide range of clinical applications. The current study shows that the FLAWS sequence can be used to generate 7T T1-weigthed images and T1 maps with reduced B1+ sensitivity, while overcoming the resolution and CNR limitations imposed to the MP2RAGE sequence. |
| 3708 | Demonstrating the benefits of high gradients in short-T2 MRI | |
| Romain Froidevaux1, Markus Weiger1, Manuela Barbara Rösler1, David Otto Brunner1, Jonas Reber1, and Klaas Paul Pruessmann1 | ||
1Institute for Biomedical Engineering, ETH Zurich and University of Zurich, Zurich, Switzerland |
||
Recent developments on high-performance gradients allow MR signals to be encoded much faster than with clinical gradients, hence decreasing T2-related k-space apodization. As a consequence, the point spread function becomes taller and narrower, thus leading to improved actual resolution and higher signal intensity for components with rapid transverse relaxation. In the present work, these benefits are investigated with PSF calculations and demonstrated experimentally using a high-performance gradient. Data of short-T2 tissues (T2 < 1 ms) are acquired using the PETRA technique with different gradient strengths and compared using image subtraction. |
| 3709 | PASTeUR package extension with MP2RAGE for robust T1 mapping technique in parallel transmit at 7T | |
| Franck Mauconduit1, Aurélien Massire2, Vincent Gras1, Alexis Amadon1, Alexandre Vignaud1, and Nicolas Boulant1 | ||
1DRF/Joliot/Neurospin & Université Paris-Saclay, CEA, Gif-sur-Yvette, France, 2Siemens Healthcare SAS, Saint Denis, France |
||
Universal Pulses (UP) were recently proposed as a plug-and-play pTx solution. To extend the so-called PASTeUR package of anatomical sequences with a mapping technique, we focused on a MP2RAGE acquisition that has shown to be robust against receive profiles when acquired at 7T in single channel transmission. UP provide the advantage of intrinsically correcting for flip angle variations throughout the brain, leading to an increased SNR in regions usually left with low B1+ intensity. Moreover, T1 quantification can be obtained without the need for B1+ acquisition and additional post-processing. |
| 3710 | A flexible linear reorder scheme for improved fat saturation in 3D VIBE imaging | |
| Qiong Zhang1, Yong Xiao Zhang2, and Yulong Liu3 | ||
1DL, Siemens Shenzhen Magnetic Resonance Ltd, Shenzhen, China, 2MR Collaborations, Siemens Healthcare Ltd., Shenzhen, China, 3Beijing Institute of Technology, BeiJing, China |
||
The fast (‘Quick’) fat saturation technique (Q-fat sat) [1], which acquires several k-space lines after each Fat-Sat module, is widely used for suppression of signals from fat in MR imaging. In this work, we developed and evaluated a linear flexible reorder scheme for improved fat saturation VIBE imaging. In vivo experiments indicated that this technique markedly improves suppression of fat saturation signals compared with conventional Q-fat saturation imaging. Studies are underway to validate the clinical value of the technique. |
| 3711 | Using dielectric bags to increase SNR in dynamic speech imaging | |
| Bradley P. Sutton1,2, Natalie Ramsy3, Riwei Jin1, and Andrew Webb4,5 | ||
1Bioengineering, University of Illinois at Urbana-Champaign, Urbana, IL, United States, 2Beckman Institute, University of Illinois at Urbana-Champaign, Urbana, IL, United States, 3Carle-Illinois College of Medicine, University of Illinois at Urbana-Champaign, Urbana, IL, United States, 4Carle Foundation Hospital, Urbana, IL, United States, 5Leiden University, Leiden, Netherlands |
||
Dynamic speech MRI has seen increasing use in the clinic for cleft palate and oral cancer and in research for examining the articulatory patterns of speech. Although potential frame rates have increased, in order to optimize acquisitions, special RF coils may be required to match the anatomy of the population of interest. In this work, we examine the potential of low-cost dielectric pads to significantly improve the SNR of acquisitions using standard head/neck coils. We show a 20% improvement in SNR in speech articulator regions with two dielectric bags placed on either side of the face. |
| 3712 | Feasibility of High-resolution Knee Imaging with A Spiral Dixon Technique | |
| Dinghui Wang1, Francis I. Baffour1, Zhiqiang Li2, Tzu-Cheng Chao1, Guruprasad Krishnamoorthy3, and James G. Pipe1 | ||
1Department of Radiology, Mayo Clinic, Rochester, MN, United States, 2Department of Neuroradiology, Barrow Neurological Institute, Phoenix, AZ, United States, 3Philips Healthcare, Gainesville, FL, United States |
||
In this study, a spiral Dixon technique has been implemented for knee imaging at 3T with high in-plane resolution of up to 0.4 by 0.4 mm2. Volunteer scans have demonstrated that high quality proton density-weighted, T2-weighted and T1-weighted water and fat images can be obtained using spiral spin-echo and dual spin-echo Dixon methods with comparable total scan time as the conventional Cartesian fast (turbo) spin-echo sequences without Dixon. In addition, spiral radial plane imaging can shorten the scan time by 62% compared to the conventional radial plane method. |
| 3713 | Compressed sensing accelerated susceptibility-weighted imaging at 3T with SPARKLING: looking for favorable parametrization | |
| Aurélien Massire1,2, Chaithya G R3, Loubna El Gueddari3, Franck Mauconduit1, Carole Lazarus1, Mathilde Ripart1, Pierre Brugières4, Philippe Ciuciu3, and Alexandre Vignaud1 | ||
1CEA\DRF\JOLIOT\NeuroSpin\UNIRS, Gif-sur-Yvette, France, 2Siemens Healthcare SAS, Saint-Denis, France, 3CEA\DRF\JOLIOT\NeuroSpin\PARIETAL, Gif-sur-Yvette, France, 4AP-HP, Hôpital Henri Mondor, Service de Neuroradiologie, Paris, France |
||
Compressed sensing (CS) theory has been successfully employed to drastically reduce MRI acquisition time. Recently, a new optimization-driven algorithm (SPARKLING) was proposed to design optimal non-Cartesian sampling patterns for CS-MRI. This method has several advantages compared to radial or spiral non-Cartesian imaging, yet the question on how the acceleration factor should be selected to ensure satisfactory image quality should be investigated. In this work, we applied the SPARKLING method for 3D susceptibility-weighted imaging, achieving 500μm in-plane resolution and full brain coverage in 3 minutes at 3T (6-fold acceleration compared to fully-sampled Cartesian imaging). |
| 3714 | SPRING TSE: 2D T2-Weighted Brain Imaging using SPiral RING Turbo Spin-Echo | |
| Zhixing Wang1, Steven Allen1, Xue Feng1, John P. Mugler2, and Craig H. Meyer1 | ||
1Biomedical Engineering, University of Virginia, CHARLOTTESVILLE, VA, United States, 2Radiology & Medical Imaging, University of Virginia, CHARLOTTESVILLE, VA, United States |
||
2D Cartesian turbo spin-echo (TSE) is widely used in the clinical neuroimaging, yet the high specific absorption rate (SAR) induced by a large number of refocusing RF pulses limits its use in high magnetic field. Thus, this study describes a new TSE sequence with annular spiral ring acquisitions, dubbed “SPRING TSE”, for fast T2-weigthed imaging and reducing SAR. Preliminary results show that two sets of high spatial resolution images with intermediate and strong T2-weighted contrast characteristics can be obtained by the proposed method within a few seconds. |
| 3715 | 3D sector-wise golden-angle (3D-SWIG) – Improved k-space uniformity after ECG binning compared to golden-angle profile ordering | |
| Alexander Fyrdahl1, Joao G Ramos1, Martin Ugander1,2, and Andreas Sigfridsson1 | ||
1Department of Clinical Physiology, Karolinska University Hospital and Karolinska Institutet, Stockholm, Sweden, 2The Kolling Institute, Royal North Shore Hospital, and Northern Clinical School, Sydney Medical School, University of Sydney, Sydney, Australia |
||
A radial trajectory with a three-dimensional sector-wise golden-angle (3D-SWIG) profile ordering optimized for retrospective ECG-based binning is presented. Data are acquired in patches using a low-distorting and area-preserving mapping from a cube to a sphere. Within each patch, the readouts are ordered according to the double golden-angle scheme. By acquiring one patch per heartbeat, k-space uniformity is guaranteed even after ECG-binning, resulting in reduced radial streak artifacts. |
| 3716 | 4D-EPICS: Compressed Sensing EPI for highly accelerated fMRI at 7T | |
| Thomas Roos1,2, Lukas Gottwald3, Tomas Knapen1,4, Benedikt Poser5, and Wietske van der Zwaag1 | ||
1Spinoza Centre for Neuroimaging, Amsterdam, Netherlands, 2Delft University of Technology, Delft, Netherlands, 3Radiology and Nuclear Medicine, Amsterdam UMC, Amsterdam, Netherlands, 4Cognitive Psychology, Vrije Universiteit Amsterdam, Amsterdam, Netherlands, 5Maastricht Brain Imaging Centre, Faculty of Psychology and Neuroscience, Maastricht University, Maastricht, Netherlands |
||
The proposed EPICS readout takes a traditional EPI readout and makes it suitable for Compressed Sensing. This novel readout keeps the EPI direction constant and employs pseudo-spiral under-sampling on a Cartesian grid in the other two (phase encoded) directions. EPICS delivers the flexibility offered by spiral sequences with the relative ease of Cartesian CS reconstruction. Multiple spiral types and patterns with flexible TE's are compared. In-vivo brain image data shows good image quality at 7T. Future 4D-EPICS fMRI acquisitions will be able to profit from regularized CS reconstructions. |
| 3717 | Real-Time Z-Shimming for Magnetic Resonance Imaging of the Spinal Cord | |
| Eva Alonso-Ortiz1, Cyril Tous2,3, Ryan Topfer1, and Julien Cohen-Adad1,4 | ||
1NeuroPoly Lab, Ecole Polytechnique, Montreal, QC, Canada, 2Department of Radiology, Radiation-Oncology and Nuclear Medicine and Institute of Biomedical Engineering, Université de Montréal, Montreal, QC, Canada, 3Laboratory of Clinical Image Processing Le Centre de Recherche du CHUM, Centre hospitalier de l'Université de Montréal, Montreal, QC, Canada, 4Functional Neuroimaging Unit, Centre de recherche de l'Institut universitaire de gériatrie de Montréal, Montreal, QC, Canada |
||
Spinal cord MRI is notoriously challenging, in part due to the fact the spinal cord is near the lungs, causing magnetic field inhomogeneities that are changing throughout respiration. We propose a novel solution, called real-time z-shimming to address this problem. Real-time z-shimming involves compensating for magnetic field inhomogeneities that arise during image acquisition, by adapting the sequence in real-time. Our findings show that real-time z-shimming can recover signal loss, caused by respiration-induced magnetic field inhomogeneities, in the spinal cord region. |
| 3718 | Clinical valid of TOF-MRA with sparse under-sampling in evaluation of intracranial aneurysm: DSA as a reference standard as a reference standard | |
| XU XU1, Zhenlin Li1, and Wanlin Peng1 | ||
1Department of Radiology, West China Hospital, Sichuan University, Chengdu, China |
||
When compared with conventional TOF, the sparse TOF which applied compressed sensing to accelerate the acquisition time has been proved to be well-performed in evaluation of UIAs. However, the image quality and the agreement between sparse TOF and DSA need to be studied further. We therefore assessed the clinical validation of sparse TOF compared with conventional TOF in qualitative and quantitative image qualities and explored the correlation among two MRAs and DSA in evaluation of size parameters. |
| 3719 | Fast Distortion-Free DWI using PSF-EPI on A 1.5T Clinical MRI Scanner with An 8-Channel Head Coil | |
| Simin Liu1, Wenpeng Wu2, Yishi Wang1,3, Jieying Zhang1, Pengcheng Xie2, and Hua Guo1 | ||
1Center for Biomedical Imaging Research, Department of Biomedical Engineering, School of Medicine, Tsinghua University, Beijing, China, 2Shenzhen Anke High-tech Co., Ltd., Shenzhen, China, 3Philips Healthcare, Beijing, China |
||
DWI is valuable for diagnosing diseases such as stroke and brain tumor. The traditional single-shot EPI is widely used in DWI because of its high acquisition speed, but it suffers from severe geometric distortions. A recently proposed technique, point-spread-function encoded EPI (PSF-EPI) that is highly accelerated by tilted-CAIPI, can achieve fast distortion-free DWI images on 3T scanners with a 32-channel head coil. This study investigated the feasibility of tilted-CAIPI accelerated PSF-EPI on a 1.5T clinical MRI scanner with an 8-channel head coil. Images with high anatomical fidelity and satisfactory SNR were obtained in 1~2.5 min using this technique. |
| 3720 | Automated k-Space Trajectory Generation using Bayesian Reinforcement Learning for Quiet Single Shot Readout | |
| Zhenliang Lin1, Qikang Li1, Lihong Tang1, Hui Huang1, Junwei Zhao1, and Jie Luo1 | ||
1Institute of Medical Imaging Technology, School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China |
||
Acoustic noise during MR scans generated by the gradient coil vibration has been compromising for patient comfort. Single-shot echo planar imaging (EPI), ubiquitously used in functional MRI and diffusion MRI acquisitions, has a rapid switching readout gradient, which is very efficient but also very loud. In this study, we employed a model free reinforcement-learning agent to optimize 2D single shot readout gradient waveforms toward the “reward” of lowering acoustic noise. The preliminary results show that the acoustic noise of the arbitrary trajectory is 17.2 dB lower than EPI for a 2D single slice readout. |
| 3721 | Multiband accelerated Fast Interrupted Steady-State (FISS) imaging with 2D Cartesian sampling | |
| Anthony N Price1,2, Lucilio Cordero-Grande1,2, Shaihan J Malik1,2, and Joseph V Hajnal1,2 | ||
1Biomedical Engineering Department, 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 |
||
In this work we demonstrate the application of Fast Interrupted Stead-State (FISS) imaging with 2D Cartesian sampling and multiband (MB) acceleration to regain scan efficiency. MB-FISS cardiac cine imaging and a free-running sequence for vessel imaging are shown to work reliably with good fat suppression. For cine cardiac imaging FISS with readouts equal to number phase-encode lines in each cardiac segment had minimal scan time increase, while FISS (n=1) lead to breath-hold durations increase ~2.5 fold, but benefit from improved fat suppression and no additional ghosting artefacts compared to bSSFP. |
| 3722 | Localized imaging using matrix gradient coils | |
| Sebastian Littin1, Feng Jia1, Huijun Yu1, and Maxim Zaitsev1 | ||
1Department of Radiology, Medical Physics, University of Freiburg, University Medical Center, Freiburg, Germany |
||
Matrix gradient coils may be used for a non-linear phase preparation to suppress signals in unwanted regions. This allows for a reduction of the FOV smaller than the object without fold-over artifacts and can be seen as a viable alternative to selective excitation. |
| 3723 | Unbalanced SSFP for Ultra-High Field Super-Resolution MRI | |
| Peter J Lally1, Paul M Matthews1,2, and Neal K Bangerter3 | ||
1Department of Brain Sciences, Imperial College London, London, United Kingdom, 2UK Dementia Research Institute, London, United Kingdom, 3Department of Bioengineering, Imperial College London, London, United Kingdom |
||
Here we exploit the off-resonance profile of the steady-state free precession (SSFP) sequence to encode spatial information with ultra-low flip angle radiofrequency pulses, enabling a super-resolution reconstruction from a rapid series of low-resolution images. This opens up new possibilities for rapid, high-resolution and low specific absorption rate (SAR) SSFP imaging, particularly at ultra-high magnetic field strengths. |
| 3724 | Rapid whole-brain imaging with sub-mm resolution using sampling on tilted hexagonal grids (t-Hex) | |
| Maria Engel1, Lars Kasper1, Franz Patzig1, Bertram Wilm1,2, Benjamin Dietrich1, Laetitia Vionnet1, and Klaas Paul Pruessmann1 | ||
1Institute for Biomedical Engineering, ETH Zurich and University of Zurich, Zurich, Switzerland, 2Skope Magnetic Resonance Technologies, Zurich, Switzerland |
||
In this work, we show high-resolution stacks of spirals on a tilted hexagonal grid. The new scheme provides flexibility in balancing readout and scan time, thereby allowing for high-quality images in a temporal resolution regime suitable for fMRI. 0.6 mm whole-brain coverage is achieved in below 5. |
| 3725 | Dual Stage Super-Resolution in Quadratic Phase Scrambling Imaging using Iterative Signal Band Expansion and Deep CNN Super-resolution | |
| Satoshi ITO1 and Yasumichi WAKATSUKI1 | ||
1Utsunomiya University, Utsunomiya, Japan |
||
Spatial resolution of images in phase scrambling Fourier transform imaging can be improved by post-processing signal band extrapolation. However, the improvements is small in the central area of image space. In our research, super-resolution using deep convolutional neural network are applied to temporally resolution improved images through iterative reconstruction. Simulation and experimental results showed that spatial resolution was fairly improved in the central area as well as in the edge of images. Proposed method is applicable to phase varied images by using adequately estimated phase distribution map since signal under-sampling is not utilized in proposed method. |
| 3726 | Dual-Repetition Gradient-Echo Dixon Imaging with High Signal-to-Noise Ratio Efficiency | |
| Holger Eggers1, Christoph Katemann2, and Hendrik Kooijman2 | ||
1Philips Research, Hamburg, Germany, 2Philips Healthcare, Hamburg, Germany |
||
Dual-repetition gradient-echo sequences are currently not widely used for Dixon imaging, because shorter scan times and better signal-to-noise ratios are usually achieved with their dual-echo counterparts. In this work, the efficiency of dual-repetition gradient-echo sequences is optimized by introducing partial echo sampling to extend the acquisition windows and by applying compressed sensing to reduce scan times. On the example of knee imaging, this is shown to enable high-resolution water-fat imaging with good image quality in reasonable scan times. |
| 3727 | Optimizing trajectory ordering for faster small-tip center-out radial gradient echo(GRE) | |
| Xucheng Zhu1,2, Kevin Johnson3,4, and Peder Larson1,2 | ||
1Radiology and Biomedical Imaging, University of California, San Francisco, San Francisco, CA, United States, 2UC Berkeley-UCSF Graduate Program in Bioengineering, University of California, San Francisco and University of California, Berkeley, Berkeley, CA, United States, 3Medical Physics, University of Wisconsin, Madison, Madison, WI, United States, 4Radiology, University of Wisconsin, Madison, Madison, WI, United States |
||
Center-out sequence has natural spoiling effect, called “self” spoiling. By leveraging this idea, the standard large spoiler gradient could be reduced or removed without sacrificing the image quality. In this study, we investigated how 3D radial trajectory ordering affects spoiling, and proposed a reordered 2D golden angle scheme to maximize “self” spoiling effect and retain robustness to free breathing scan. We evaluated the proposed ordering scheme on both phantom and volunteer. Proposed scheme without or with small spoiler gradient can reduce up to 40% scan time to get comparable image quality. |
| 3728 | Magnetic Resonance Imaging Using Multiple-leaf Smoothed Random-like Trajectory | |
| Haifeng Wang1, Yuchou Chang2, Xin Liu1, Hairong Zheng1, and Dong Liang1 | ||
1Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China, 2University of Houston-Downtown, Houston, TX, United States |
||
Smoothed Random-like Trajectory (SRT) is a promising MR imaging method, but the data acquisitions have some hardware limitations of the gradient amplitude and slew rate. In order to realize the SRT in k-space, we proposed a new multiple-leaf SRT in k-space to reduce hardware requirements for applying the Compressed Sensing (CS) theory. To guarantee the constrains of the gradient amplitude and slew rate and reduce readout, the proposed multiple-leaf gradient waveforms were optimized by the time-optimal method for arbitrary k-space trajectories. The simulations have showed that the proposed method could greatly improve the reconstruction image quality, comparing to spiral trajectories. |
| 3729 | Single Point Imaging with Rapid Encoding (SPIRE) for MRI with Sub-millisecond Temporal Resolution | |
| Johannes Fischer1 and Michael Bock1 | ||
1Dept. of Radiology, Medical Physics, Medical Center University of Freiburg, Faculty of Medicine, University of Freiburg, Freiburg, Germany |
||
Single Point Imaging with Rapid Encoding (SPIRE) is a new imaging technique for measuring fast and repetitive motion. Here, the temporal resolution does not depend on TR but on the duration of the fast-switching phase encoding gradients. We show the abilities of SPIRE by imaging the falling of water drops and reconstruct the images using self-gating. The obtained image data has a temporal resolution below 0.5 ms and the drops oscillations frequency is measured. The good agreement with predictions shows the accuracy and possibilities of this technique. |
| 3730 | A multi echo pulse sequence with optimized excitation pulses and a 3D cone readout for hyperpolarized 13C imaging | |
| Vencel Somai1,2, Alan J Wright1, Maria Fala1, Friederike Hesse1, and Kevin M Brindle1,3 | ||
1Cancer Research UK Cambridge Institute, Cambridge, United Kingdom, 2Department of Radiology, University of Cambridge, Cambridge, United Kingdom, 3Department of Biochemistry, University of Cambridge, Cambridge, United Kingdom |
||
We describe here a single shot multi echo sequence for dynamic hyperpolarized 13C imaging with a short readout time, isotropic point spread function (PSF) and high immunity to B0 and B1 field inhomogeneity. The sequence uses numerically optimized excitation pulses and a 3D cone k-space trajectory composed of 13 cones, all fully refocused and distributed among 7 spin echoes. The maximal gradient amplitude and slew-rate were set to 4 G/cm and 20 G/cm/ms respectively to demonstrate the feasibility of clinical translation. The sequence was demonstrated with dynamic imaging of hyperpolarized [1-13pyruvate] and [1-13C]lactate in vivo. |
| 3731 | Highly Accelerated EPI with Wave Encoding and Multi-shot Simultaneous MultiSlice Imaging | |
| Jaejin Cho1,2, Congyu Liao1,2, Zijing Zhang1,3, Wei-Ching Lo4, Jinmin Xu1,3, Onur Beker1,5, Kawin Setsompop1,2,6, and Berkin Bilgic1,2,6 | ||
1Radiology, Athinoula A. Martinos Center for Biomedical Imaging, Charlestown, MA, United States, 2Harvard Medical School, Boston, MA, United States, 3College of Optical Science and Engineering, Zhejiang University, Hangzhou, China, 4Siemens Medical Solution, Boston, MA, United States, 5Electrical and Electronics Engineering, Bogazici University, Istanbul, Turkey, 6Harvard-MIT Division of Health Sciences and Technology, Cambridge, MA, United States |
||
We combine wave controlled aliasing strategy with echo-planar imaging (EPI) and simultaneous multi-slice (SMS) encoding to increase acquisition efficiency and reduce geometric distortion simultaneously for functional (fMRI) and diffusion (dMRI) imaging. Wave-EPI enables RinxRsms=4x3-fold accelerated single-shot gradient-echo (GE)-EPI by reducing the maximum g-factor noise amplification by 2-fold compared to blip-CAIPI. We extend wave-EPI to multi-shot acquisition and incorporate Hankel low-rank constraint to leverage the similarities among the shots for improved reconstruction. This allows us to create a 10-second structural scan with 12-fold reduced distortion and push the acceleration rate to RinxRsms=6x3-fold in dMRI using 3-shots with markedly improved image quality. |
| 3732 | Accelerated Multi-Echo Gradient Echo Imaging using Locally-Low Rank Regularization | |
| Mirco Grosser1,2 and Tobias Knopp1,2 | ||
1Section for Biomedical Imaging, University Medical Center Hamburg-Eppendorf, Hamburg, Germany, 2Institute for Biomedical Imaging, Technical University Hamburg, Hamburg, Germany |
||
Reconstruction methods for multi-contrast MRI often employ a low-rank constraint to reconstruct images from highly undersampled data. For multi-echo gradient echo sequences such a constraint is hard to impose due off-resonance-induced oscillations of the time signals. In this work we show that spatio-temporal correlations can be exploited efficiently using locally low rank regularization. Based on this observation we develop a locally-low rank regularized reconstruction scheme and test it on a multi-echo gradient echo dataset of a human brain. In our tests the proposed method shows significantly improved image quality compared to regular compressed sensing reconstructions. |
| 3733 | Looping Star: Revisiting echo in/out separation | |
| Florian Wiesinger1,2 and Ana Beatriz Solana1 | ||
1GE Healthcare, Munich, Germany, 2King's College London, London, United Kingdom |
||
Recently, we introduced a new MR pulse sequence, termed Looping Star, for fast and yet quiet 3D radial, multi-gradient echo MR imaging. The method is based on the 3D radial Rotating Ultra-Fast Imaging Sequence (RUFIS, aka Zero TE) extended by a time-multiplexed gradient refocusing mechanism; providing an initial free-induction-decay (FID) image followed by equidistant gradient echo (GRE) images. In its original implementation, the method was affected by overlapping echo in/out signals. Here we present a solution to this problem, allowing faster scanning and/or higher resolution. |
| 3734 | Joint optimization of sampling patterns and reconstruction for multi-contrast fast magnetic resonance imaging | |
| Jiechao Wang1, Qinqin Yang1, Qizhi Yang1, Shuhui Cai1, and Congbo Cai1 | ||
1Department of Electronic Science, Xiamen University, Xiamen, China |
||
Multi-contrast magnetic resonance imaging (MRI) is usually required in clinical diagnosis but different contrast MRI may need different scan time. To balance total scan time and reconstruction fidelity, the recovery of multi-contrast MR images relies on the collaborative acquisition of sampling patterns and reconstruction algorithm. We proposed a novel neural network that could jointly optimize sampling patterns and concurrently reconstruct multi-contrast MR images. The reconstructed multi-contrast MR images using optimized sampling patterns on a two-contrast dataset demonstrate that the average peak signal-to-noise ratio and structural similarity among contrasts improve obviously compared with reconstructed results using fixed and independent sampling patterns. |
| 3735 | Single-shot Pseudo-Centric EPI for Magnetization-Prepared Imaging | |
| Hyun-Soo Lee1, Seon-ha Hwang1, Jaeseok Park2, and Sung-Hong Park1 | ||
1Department of Bio and Brain Engineering, KAIST, Daejeon, Korea, Republic of, 2Department of Biomedical Engineering, Sungkyunkwan University, Suwon, Korea, Republic of |
||
Single-shot EPI is a famous ultra-fast MR imaging technique, but is limited to linear reordering due to its special k-space trajectory. In this study, we proposed a single-shot pseudo-centric EPI where k-space is encoded from center to periphery in a groupwise manner by utilizing grouped oscillating readout gradients, phase‑encoding blips within each group, and big phase‑encoding jumps between two consecutive groups. The concept was tested on phantoms and human brains in 3T and 7T. The proposed method enabled the significant reduction of TE, which is expected to maximize SNR of magnetization-prepared imaging and enable ultrashort-TE imaging in the Cartesian coordinate. |
| 3736 | Retrospective motion correction of three-dimensional Magnetic Resonance Fingerprinting (MRF) | |
| Jan W Kurzawski1,2, Matteo Cencini1,3, Luca Peretti1,3, Pedro A Gomez4, Rolf Schulte5, Graziella Donatelli1,6, Mirco Cosottini1,3, Paolo Cecchi6, Mauro Costagli1,7, Alessandra Retico2, Michela Tosetti1,7, and Guido Buonincontri1,7 | ||
1Imago7, Pisa, Italy, 2INFN, Pisa, Italy, 3University of Pisa, Pisa, Italy, 4Technical University of Munich, Munich, Germany, 5GE Healthcare, Munich, Germany, 6Neuroradiology, Azienda Ospedaliero-Universitaria Pisana, Pisa, Italy, 7IRCCS Stella Maris, Pisa, Italy |
||
We demonstrate a novel motion correction of 3D magnetic resonance fingerprinting (MRF) using spiral projection k-space trajectory. The motion was corrected using rigid motion parameters extracted from a whole-brain navigator obtained every 7s of imaging. Firstly, we optimized the trajectory ordering in simulation and selected the acquisition scheme that allowed the best navigator for motion correction. Secondly, we applied this scheme to invivo data in healthy subjects scanned first without motion and then while performing a motion paradigm. Our motion correction improved the correlation of motion-corrupted data with motionless data by over 20% for both T1 and T2. |
| 3737 | Generalised Low-Rank Non-rigid Motion Corrected reconstruction for 3D free breathing Liver Magnetic Resonance Fingerprinting | |
| Gastao Cruz1, Olivier Jaubert1, Haikun Qi1, Torben Schneider2, René M. Botnar1, and Claudia Prieto1 | ||
1Biomedical Engineering Department, School of Biomedical Engineering and Imaging Sciences, King's College London, London, United Kingdom, 2Philips Healthcare, Guildford, United Kingdom |
||
Magnetic Resonance Fingerprinting (MRF) has been shown to enable simultaneous T1 and T2 mapping of the liver and abdomen. 2D liver MRF requires breath holding, whereas preliminary results have been demonstrated for 3D free-breathing liver MRF using respiratory gating. However, gating approaches lead to unpredictable scan times and may impair the MRF encoding, since only data within the respiratory gating window is used for reconstruction. Here we propose a novel low-rank motion corrected approach to both resolve MRF varying contrast and perform non-rigid respiratory motion correction directly in the reconstruction, enabling 3D free-breathing liver MRF with 100% respiratory scan efficiency. |
| 3738 | Motion Tracking using Continuous Magnetic Resonance Fingerprinting | |
| Edward S. Hui1, Di Cui1, Jing Cai2, Queenie Chan3, and Peng Cao1 | ||
1Diagnostic Radiology, The University of Hong Kong, Hong Kong, Hong Kong, 2Health Technology and Informatics, The Hong Kong Polytechnic University, Hong Kong, Hong Kong, 3Philips Healthcare, Hong Kong, Hong Kong |
||
Unlike conventional contrast-weighted imaging, quantitative MR parametric maps can be obtained from MRF and should conceivably be very useful for the quantification and delineation of normal and pathologic tissues. Together with the fact that MRF is very efficient , our central hypothesis is that MRF is an ideal alternative to existing MRI motion tracking methods. In this study, we have demonstrated that it is possible to track motion by continuously performing MRF during free breathing. MR parametric maps for each respiratory phases were retrospectively estimated from the MRF snapshots that fall into a given respiratory bin. |
| 3739 | Slice Accelerated EPI-based Magnetic Resonance Fingerprinting for Simultaneous Estimation of T1, T2, T2* with Whole-brain Coverage | |
| Mahdi Khajehim1,2, Thomas Christen3, Fred Tam4, Simon Graham1,4, and J. Jean Chen1,2 | ||
1Medical Biophysics, University of Toronto, Toronto, ON, Canada, 2Rotman Research Institute, Baycrest Health Sciences, Toronto, ON, Canada, 3Grenoble Institute of Neurosciences, Inserm, Grenoble, France, 4Hurvitz Brain Sciences Research Program, Sunnybrook Research Institute, Toronto, ON, Canada |
||
Magnetic resonance fingerprinting (MRF) offers a way to quantitatively estimate multiple relaxation parameters with a single scan, but fast MRF estimation of T1, T2, and T2* with a single acquisition remains challenging. Here, a dual-stage EPI-based MRF approach with online image reconstruction is proposed for estimating T1, T2, and T2*. We achieved 1.1x1.1x3 mm3 resolution with minimal distortion, and with the use of simultaneous multi-slice acceleration, our method can provide whole-brain coverage (1.7x1.7x3 mm3) in less than 3 minutes. |
| 3740 | Accelerated High-Resolution 3D MR Fingerprinting Using a Graph Convolutional Network | |
| Feng Cheng1, Zhenghan Fang2, Xiaopeng Zong3, Weili Lin3, Yong Chen3, and Pew-Thian Yap3 | ||
1Department of Computer Science, University of North Carolina at Chapel Hill, CHAPEL Hill, NC, United States, 2CuraCloud Corporation, Seattle, WA, United States, 3Department of Radiology, University of North Carolina at Chapel Hill, CHAPEL Hill, NC, United States |
||
In this study, a k-space interpolation technique for high-resolution 3D MR Fingerprinting is proposed. We formulate the problem as a graph and apply a graph convolutional network on the graph to interpolate the missing partitions. Our preliminary results show that the proposed method can provide improved results both in reconstructed k-space data and in extracted quantitative maps and can potentially allow higher acceleration factors along the partition-encoding direction. |
| 3741 | ST-UNet: Spatio-Temporal U-Net for Accelerating Simultaneous MultiSlice Magnetic Resonance Fingerprinting | |
| Yilin Liu1, Xiaopeng Zong1, Zhenghan Fang1, Weili Lin1, Dinggang Shen1, Pew-Thian Yap1, and Yong Chen1 | ||
1University of North Carolina at Chapel Hill, Chapel Hill, NC, United States |
||
In this study, a spatio-temporal U-Net was developed for efficient and accurate T1 and T2 quantifications from multislice MR Fingerprinting acquisitions. Our preliminary results demonstrate that the proposed method outperforms the standard template matching method and deep learning methods in the literature, enabling higher multislice factors for MR Fingerprinting. |
| 3742 | Flip Angle Optimization for Continuous Flow-MRF Acquisition under Periodic Boundary Conditions | |
| Sebastian Flassbeck1, Fabian Kratzer1, Simon Schmidt1, Lisa Leroi2, Mark E. Ladd1, and Sebastian Schmitter1,2 | ||
1German Cancer Research Center (DKFZ), Heidelberg, Germany, 2Physikalisch-Technische Bundesanstalt (PTB), Braunschweig and Berlin, Germany |
||
Flow-MRF is a highly promising technique for rapid quantification of both relaxation times and time-resolved flow velocities. However, this method currently suffers from two major drawbacks, firstly a low efficiency since multiple shots are used and secondly strongly increased velocity noise for low velocities due to magnetization preparation pulses. This abstract targets both issues by presenting an optimized the flip angle (FA)-pattern for Flow-MRF, while assuming periodic boundary conditions. |
| 3743 | Optimizing signal patterns for MR vascular fingerprinting | |
| Aurelien Delphin1, Fabien Boux1,2, Clément Brossard1, Jan M Warnking1, Benjamin Lemasson1, Emmanuel Luc Barbier1, and Thomas Christen1 | ||
1Univ. Grenoble Alpes, Inserm, U1216, Grenoble Institut Neurosciences, GIN, 38000, Grenoble, France, 2Univ. Grenoble Alpes, Inria, CNRS, G-INP, 38000, Grenoble, France |
||
MR vascular fingerprinting proposes to map vascular properties such as blood volume fraction, average vessel radius or blood oxygenation saturation (SO2). The fingerprint pattern used in previous studies provides low sensitivity on SO2. We optimised signal patterns built from pre and post USPIO acquisitions. Concatenation of different echoes associated with higher dimensional dictionaries led to better estimates in both healthy and tumoral tissues. |
| 3744 | Quantum Optimization Framework for MR Fingerprinting Framework Incorporating Undersampling and Noise | |
| Siyuan Hu1, Ignacio Rozada2, Rasim Boyacioglu3, Stephen Jordan4, Sherry Huang1, Matthias Troyer4, Mark Griswold3, Debra McGivney1, and Dan Ma1 | ||
1Biomedical Engineering, Case Western Reserve University, Cleveland, OH, United States, 21Qbit, Vancouver, BC, Canada, 3Radiology, Case Western Reserve University, Cleveland, OH, United States, 4Microsoft, Redmond, WA, United States |
||
MR fingerprinting is a novel quantitative MR imaging technique that provides multiple tissue properties maps simultaneously. Designing appropriate MR fingerprinting sequence patterns is crucial to speed up data acquisition while obtaining accurate measurements. Here we propose an advanced MR fingerprinting optimization framework that incorporates undersampling artifacts and random noise in the cost function which directly compute quantitative errors in the result maps. We use quantum-inspired algorithm to solve the problem and generate optimized sequences. In both simulation and in vivo experiments, the optimized sequence showed improved image quality and measurement accuracy. |
| 3745 | A generalized combined FISP and PSIF MR Fingerprinting with improved T2 quantification | |
| Huihui Ye1,2, Qiqi Tong2, Qing Li2,3, Xiaozhi Cao2, Hongjian He2, Jianhui Zhong2, and Huafeng Liu1 | ||
1State Key Laboratory of Modern Optical Instrumentation, College of Optical Science and Engineering, Zhejiang University, Hangzhou, China, 2Center for Brain Imaging Science and Technology, Key Laboratory for Biomedical Engineering of Ministry of Education, College of Biomedical Engineering and Instrumental Science, Zhejiang University, Hangzhou, China, 3MR Collaborations, Siemens Healthcare Ltd., Shanghai, China |
||
A generalized combined FISP and PSIF MR Fingerprinting sequence is proposed where either FISP block or PSIF block can be used in each TR. An example sequence pattern shows improved T2 mapping accuracy and its immunity to slice profile imperfection. |
| 3746 | Magnetic Resonance Fingerprinting Reconstruction and Dictionary Matching with Compensation of Frequency Drifts | |
| Rasim Boyacioglu1 and Mark Griswold1 | ||
1Radiology, Case Western Reserve University, Cleveland, OH, United States |
||
Magnetic Resonance Fingerprinting (MRF) maps multiple tissue properties and system parameters simultaneously. The accuracy of MRF maps depends on the simulation of all possible system properties into the signal evolutions via the Bloch equations. We have observed frequency drifts during MRF scans, similar to those seen in fMRI scans, which might cause various artifacts if not accounted for. Here, it is shown that 2D MRF frequency drifts can be compensated with a simple dictionary update. For correction of 3D MRF frequency drifts, a novel reconstruction framework is introduced. Results show significant improvements on quantitative maps for 2D and 3D MRF. |
| 3747 | A Fast Approximation of Undersampling Artifacts in MR Fingerprinting | |
| Debra McGivney1, Rasim Boyacıoğlu2, Stephen Jordan3, Ignacio Rozado4, Sherry Huang1, Siyuan Hu1, Brad Lackey3, Matthias Troyer3, Mark Griswold2, and Dan Ma1 | ||
1Biomedical Engineering, Case Western Reserve University, Cleveland, OH, United States, 2Radiology, Case Western Reserve University, Cleveland, OH, United States, 3Microsoft, Redmond, WA, United States, 41QB Information Technologies, Vancouver, BC, Canada |
||
Iterative optimization in MRI is a large problem with many degrees of freedom. Depending on the cost function and parameters of interest, it may be beneficial to model errors from undersampling with non-Cartesian trajectories. This typically requires repeated use of the nonuniform FFT (NUFFT), which is computationally expensive. Here we propose an approximation based on a limited number of tissue types that eliminates the need for repeated NUFFTs, and allows a wide range of applications for sequence optimization in MRI and MRF. |
| 3748 | Reducing off-resonance effects by spatial decorrelation in Magnetic Resonance Fingerprinting | |
| Ronal Manuel Coronado1,2,3, Gastão Lima da Cruz4, Claudia Prieto1,2,4, and Pablo Irarrázaval1,2,3,5 | ||
1Biomedical Imaging Center, Pontificia Universidad Catolica de Chile, Santiago, Chile, 2Department of Electrical Engineering, Pontificia Universidad Católica de Chile, Santiago, Chile, 3Millennium Nucleus for Cardiovascular Magnetic Resonance, Chile, Santiago, Chile, 4School of Biomedical Engineering and Imaging Sciences, King's College London, London, United Kingdom, London, United Kingdom, 5Institute for Biological and Medical Engineering, Pontificia Universidad Católica de Chile, Santiago, Chile, Santiago, Chile |
||
Magnetic Resonance Fingerprinting (MRF) acquisitions based on balanced Steady State Free Precession (bSSFP) with spiral trajectories produces higher signal noise to ratio (SNR) in comparison to the unbalanced version, nevertheless its usage is hider because of higher sensitivity to off-resonance artifacts. These artifacts affect the quality of parametric values. Here we propose a novel method to decrease these susceptibility distortions by means of spatial decorrelation. We show promising results where off-resonance artifacts were reduced in a realistic brain simulation and standardized T1/T2 phantom acquisition. |
| 3749 | Effects of spatial encoding strategies on 2D and 3D magnetic resonance fingerprinting | |
| Matteo Cencini1,2, Pedro A Gómez3, Mohammad Golbabaee 4, Rolf F Schulte5, Giada Fallo1,2, Luca Peretti1,2, Michela Tosetti2,6, Bjoern H Menze3, and Guido Buonincontri2,6 | ||
1University of Pisa, Pisa, Italy, 2Imago7 Foundation, Pisa, Italy, 3Technical University of Munich, Munich, Germany, 4University of Bath, Bath, United Kingdom, 5GE Healthcare, Munich, Germany, 6IRCCS Stella Maris, Pisa, Italy |
||
Transient-state imaging techniques such as MR Fingerprinting allow for simultaneous quantification of tissue properties by using variable acquisition parameters in conjunction with undersampled non-Cartesian trajectories. Several implementations exist in literature, relying on two- or three-dimensional sampling readouts. Here, we studied the effect of the spatial encoding on quantification. In addition, an evaluation of the impact on parametric maps of anti-aliasing techniques (k-space weighted image constrast) was performed, both in vitro and in vivo. |
| 3750 | Optimized dimensionality reduction for parameter estimation in MR fingerprinting via deep learning | |
| Quentin Duchemin1, Kangning Liu1, Carlos Fernandez-Granda2, and Jakob Assländer3 | ||
1Center for Data Science, NYU, New York, NY, United States, 2Courant Institute of Mathematical Sciences and Center for Data Science, New York University, New York, NY, United States, 3Center for Biomedical Imaging, Department of Radiology, New York University School of Medicine, Center for Advanced Imaging Innovation and Research (CAI2R), Department of Radiology, New York University School of Medicine, New York, New York, NY, United States |
||
We propose a deep learning approach for MR fingerprinting that jointly learns a low-dimensional representation of the fingerprints and estimates biophysical parameters from this subspace. In contrast to SVD-based projections, which are agnostic to the estimation task, the learned subspace is optimized to maximize information content about the parameters of interest. Incorporating the learned basis functions in the forward imaging operator suppresses undersampling artifacts and increases computational efficiency. |
| 3751 | Accelerated 3D Magnetic Resonance Fingerprinting Using Alternating Direction Method of Multipliers | |
| Di Cui1, Xiaoxi Liu1, Peng Cao1, Queenie Chan2, and Edward S. Hui1 | ||
1Diagnostic Radiology, The University of Hong Kong, Hong Kong, China, 2Philips Healthcare, Hong Kong, Hong Kong |
||
3D magnetic resonance fingerprinting was developed for volumetric parametric quantification. In this study, an alternating direction method of multipliers (ADMM) based 3D MRF approach is proposed to jointly utilize sparsity constraint and spatial coil sensitivity information, leading to better reconstruction performance with 3-fold undersampled 3D MRF data. We demonstrated that the effective scan time for R=3 and whole-brain MRF is less than 7 minutes. |
| 3752 | Renal magnetic resonance fingerprinting for simultaneous T1 and T2* quantification | |
| Ingo Hermann1,2, Jorge Chacon-Caldera1, Irène Brumer1, Sebastian Weingärtner2, Lothar R. Schad1, and Frank G. Zöllner1 | ||
1Computer Assisted Clinical Medicine, Medical Faculty Mannheim, University Heidelberg, Mannheim, Germany, 2Magnetic Resonance Systems Lab, Department of Imaging Physics, Delft University of Technology, Delft, Netherlands |
||
We implemented and validated a MRF sequence for simultaneous T1 and T2* quantification in the kidneys covering 4 slices within one breath-hold. Therefore, we used an echo-planar-imaging readout with fully sampled k-space and 35 measuremets for the matching process. We achieved promising results in phantom and in 8 healthy volunteers compared with conventional methods as MOLLI and GRE and with reference scans. Additionally, we performed smoothing on the baseline images to further improve the image qualitiy of the MRF parametric maps. |
| 3753 | Validation of a water and fat separation framework for Liver MR Fingerprinting | |
| Cristobal Arrieta1,2, Olivier Jaubert3, Gastao Cruz3, Sergio Uribe1,2,4, Rene M Botnar3, Claudia Prieto3, and Carlos Sing-Long2,5,6 | ||
1Biomedical Imaging Center, Pontificia Universidad Catolica de Chile, Santiago, Chile, 2Millennium Nucleus for Cardiovascular Magnetic Resonance, Santiago, Chile, 3School of Biomedical Engineering and Imaging Sciences, King’s College London, London, United Kingdom, 4Radiology Department, School of Medicine, Pontificia Universidad Catolica de Chile, Santiago, Chile, 5Instituto de Ingeniería Matemática y Computacional, Pontificia Universidad Catolica de Chile, Santiago, Chile, 6Millennium Nucleus Center for the Discovery of Structures in Complex Data, Santiago, Chile |
||
In this work we validated a liver MR Fingerprinting technique which allows to simultaneously estimate T1, T2, Proton Density Fat Fraction (PDFF) and T2*. This technique is based on a novel water and fat separation optimisation scheme, which includes l2-norm of the fieldmap gradient, TV regularisation on T2* and l1-norm denoising of the water and fat concentrations. We presented results on T1/T2 and water/fat phantoms and in 11 volunteers. This approach showed to be robust and it is ready to be applied on more clinical challenging cases. |
| 3754 | Simultaneous Multi-contrast Four-dimensional Magnetic Resonance Imaging for Radiotherapy Applications | |
| Tian Li1, Di Cui2, Edward S. Hui2, and Jing Cai1 | ||
1The Hong Kong Polytechnic University, Hong Kong, Hong Kong, 2The University of Hong Kong, Hong Kong, Hong Kong |
||
Tumor motion imaging is of vital importance in managing mobile cancers in radiation therapy. However, current 4D-MRI techniques are inefficient and ineffective, potentially leading to suboptimal and inaccurate results. To solve this problem, we propose a novel four-dimensional magnetic resonance fingerprinting (4D-MRF) technique for radiation therapy applications. Our proposed method has been validated through simulations and in-vivo volunteer experiment. |
| 3755 | The effect of hormone therapy on T1 mapping-related values obtained from MR fingerprinting-derived images of prostate cancer patients | |
| Nikita Sushentsev1, Joshua D Kaggie1, Guido Buonincontri2,3, Rolf Schulte4, Vincent J Gnanapragasam5,6,7, Martin J Graves1, and Tristan Barrett1,8 | ||
1Department of Radiology, University of Cambridge, Cambridge, United Kingdom, 2IRCCS Stella Maris, Piza, Italy, 3Imago7 Foundation, Piza, Italy, 4GE Healthcare, Munich, Germany, 56) Cambridge Urology Translational Research and Clinical Trials Office, University of Cambridge, Cambridge, United Kingdom, 6Academic Urology Group, Department of Surgery & Oncology, University of Cambridge, Cambridge, United Kingdom, 7Department of Urology, University of Cambridge, Cambridge, United Kingdom, 82) CamPARI Prostate Cancer Group, Addenbrooke's Hospital, Cambridge, United Kingdom |
||
This study investigates the effect of hormone therapy for prostate cancer on T1 mapping values obtained from MR Fingerprinting. No differences were observed between tumour T1 values before and after treatment. However, lower T1 MRF values were noted in prostate lesions compared to normal peripheral and transition zones, which is consistent with current literature. Moreover, a significant decrease in T1 values was observed in normal transition zone (TZ) following treatment, which may justify the need of using T1 mapping in contrast-enhanced MRI studies aimed at calculating quantitative TZ-derived parameters. |
| 3756 | Quantitative Tracking of T1 and T2* in DCE-MRI by Dynamic MR Fingerprinting | |
| Xueying Zhao1, Ying-Hua Chu2, Qianfeng Wang1, and He Wang1,3 | ||
1Institute of Science and Technology for Brain-inspired Intelligence, Fudan University, Shanghai, China, 2MR Collaboration, Siemens Healthcare Ltd., Shanghai, China, 3Human Phenome Institute, Fudan University, Shanghai, China |
||
Precise quantification of the contrast agent uptake in DCE-MRI is still challenging. Here, we redesigned the reconstruction scheme of MR fingerprinting by introducing a sliding window into the process of dictionary matching, which allows dynamic quantification of T1 and T2* in DCE-MRI with high temporal resolution. The performance of this proposed dynamic MRF method is simulated and tested on MATLAB. The time-varying trajectories of T1 and T2* have been successfully captured with a temporal resolution of 4.5 seconds. This method may open the door of utilizing MR fingerprinting to quantify time-varying variable with adjustable temporal resolution. |
| 3757 | Validity and repeatability of Magnetic Resonance Fingerprinting in the healthy human brain at 3T | |
| Joely Smith1,2, Ben Statton3, Sarah Cardona1, Mary Elizabeth Finnegan1,2, Rebecca Abigail Quest1,2, and Matthew Grech-Sollars1,4 | ||
1Department of Imaging, Imperial College Healthcare NHS Trust, London, United Kingdom, 2Department of Bioengineering, Imperial College London, London, United Kingdom, 3MRC London Institute of Medical Sciences, Imperial College London, London, United Kingdom, 4Department of Surgery and Cancer, Imperial College London, London, United Kingdom |
||
Obtaining quantitative measurements in a 7-minute acquisition could improve the sensitivity of MR diagnosis. We investigated the validity and reproducibility of magnetic resonance fingerprinting (MRF) relaxometry in 12 tissue compartments in the human brain through comparison to standard mapping techniques: variable flip angle for T1 and multi-echo spin echo for T2. Statistically significant strong and moderate correlations were found between the MRF and standard mapping methods for T1 and T2, respectively. The MRF results were shown to be highly reproducible and in agreement with values found within the literature. However, a bias was found between MRF and standard relaxometry methods. |
| 3758 | Two-site repeatability study of Tailored Magnetic Resonance Fingerprinting (TMRF) | |
| Enlin Qian1, Amaresh Shridhar Konar2, Pavan Poojar1, Maggie Fung3, and Sairam Geethanath1 | ||
1Columbia Magnetic Resonance Research Center, New York, NY, United States, 2Memorial Sloan Kettering Cancer Center, New York, NY, United States, 3GE Healthcare Applied Sciences Laboratory East, New York, NY, United States |
||
This work evaluates repeatability of T1 and T2 estimation of the TMRF method compared to MRF and the gold standard measurements at two sites. In this work, we acquired data for 10 days on ISMRM/NIST phantom using the three methods at the two sites with the same vendor and field strength. The data was reconstructed and matched with an EPG simulated dictionary. ROI analysis was performed to extract T1 and T2 estimation for each sphere. The results and statistical analysis are presented here. |
| 3759 | Accuracy, reproducibility and temperature variability of Magnetic Resonance Fingerprinting using the ISMRM/NIST system phantom | |
| Ben Statton1,2, Joely Smith3,4, Mary E Finnegan3,4, Rebecca A Quest3,4, and Matthew Grech-Sollars3,5 | ||
1Imperial College London, London, United Kingdom, 2Medical Research Council, London Institute of Medical Sciences, London, United Kingdom, 3Department of Imaging, Imperial College Healthcare NHS Trust, London, United Kingdom, 4Department of Bioengineering, Imperial College London, London, United Kingdom, 5Department of Surgery and Cancer, Imperial College London, London, United Kingdom |
||
Magnetic Resonance Fingerprinting (MRF) is a technique which produces multiple parametric maps during a single fast acquisition. Before MRF can be adopted clinically, quantitative values derived from these maps must be proven accurate and reproducible over a range of T1 and T2 values and temperatures. The aim of this study was to investigate the accuracy and reproducibility of T1and T2 values derived from two different methods of MRF compared to conventional quantitative maps using the ISMRM/NIST system phantom. |
| 3760 | Repeatability of Magnetic Resonance Fingerprinting using ISMRM/NIST MRI Phantom in Philips 3T MRI Scanner | |
| Varut Vardhanabhuti1, Ho Ting Au2, Jie Ding1, Elaine Y Lee1, Peng Cao1, and Edward S Hui1 | ||
1Diagnostic Radiology, The University of Hong Kong, Hong Kong, Hong Kong, 2The University of Hong Kong, Hong Kong, Hong Kong |
||
The purpose of this study is to assess the repeatability of the magnetic resonance fingerprinting (MRF) sequence in a Philips 3T scanner with reference to the ISMRM/NIST MRI system phantom. The results showed a strong correlation between the T1 and T2 estimated from MRF versus the reference values (R2 = 0.999, and R2 = 0.981). A high level of repeatability was achieved over the 15 scanning sessions, although the T1 estimated from MRF has a significantly higher degree of repeatability than T2. |
| 3761 | Comparing FLASH vs GRE for 2D cardiac MR fingerprinting | |
| Gastao Cruz1, Olivier Jaubert1, Aurelien Bustin1, Torben Schneider2, Peter Koken3, Mariya Doneva3, René M. Botnar1, and Claudia Prieto1 | ||
1Biomedical Engineering Department, School of Biomedical Engineering and Imaging Sciences, King's College London, London, United Kingdom, 2Philips Healthcare, Guildford, United Kingdom, 3Philips Research Hamburg, Hamburg, Germany |
||
2D Cardiac Magnetic Resonance Fingerprinting (cMRF) has been previously proposed to provide co-registered T1/T2 maps from a single breath-hold scan. Gradient Spoiled gradient echo (GRE) readout is conventionally employed for 2D cMRF. RF-spoiled readouts (FLASH) have reduced signal-to-noise ratio but are less sensitive to field inhomogeneities that may bias the parametric maps (if not encoded in the MRF sequence). Here we compare the performance of FLASH and GRE readouts for simultaneous T1/T2 mapping in 2D cMRF. Phantom and in-vivo results showed that both sequences produce comparable maps and reproducible values, however GRE-cMRF presented larger underestimations for T1 and T2. |
| 3762 | Tissue based denoising for MR fingerprinting via long short-term memory networks | |
| Gastao Cruz1, Thomas Kuestner1, Ilkay Oksuz1, Olivier Jaubert1, Niccolo Fuin1, Andy P. King1, Julia A. Schnabel1, René M. Botnar1, and Claudia Prieto1 | ||
1Biomedical Engineering Department, School of Biomedical Engineering and Imaging Sciences, King's College London, London, United Kingdom |
||
Conventional Magnetic Resonance Fingerprinting (MRF) relies on pixelwise dictionary matching of highly undersampled time-series images. However, remaining aliasing artefacts in these images can compromise the matching step and thus affect the accuracy of the parametric maps. Dictionary-based compression has been proposed to exploit redundancies in the signal evolution dimension, however these approaches do not exploit tissue redundancies within the images. Here we propose to leverage redundant information between similar tissues and the MRF dictionary to suppress residual artefacts along time, using long short-term memory (LSTM) networks. Preliminary results indicate that proposed MRF-LSTMs can suppress aliasing in highly undersampled scenarios. |
| 3763 | Accurate estimation of multiple parameters from MRF signals using deep learning | |
| Ryoichi Sasaki1 and Yasuhiko Terada1 | ||
1Institute of Applied Physics, University of Tsukuba, Tsukuba, Japan |
||
Ideally, MRF can quantify multiple parameters at a single scan. In some cases, however, these parameters are not separable and additional scans for inseparable parameters are required, which reduces the advantage of the short scan time of MRF. This is remarkable when the large number of parameters are involved in the signal evolution process. Here, we used a pattern matching using a deep neural network called DRONE, and verified the separability of four parameters of T1, T2, B1, and ADC from MRF-FISP signals acquired at a single scan. |
| 3764 | Learning how to Clean Fingerprints -- Deep Learning based Separated Artefact Reduction and Regression for MR Fingerprinting | |
| Yiling Xu1, Elisabeth Hoppe1, Peter Speier2, Thomas Kluge2, Mathias Nittka2, Gregor Körzdörfer2, and Andreas Maier1 | ||
1Pattern Recognition Lab, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany, 2Magnetic Resonance, Siemens Healthcare, Erlangen, Germany |
||
Various deep learning approaches have been recently introduced to enable a fast MRF reconstruction compared to dictionary matching. Artefacts resulting from the strong undersampling during the acquisition often impair the reconstruction results. In this work, we introduce a deep learning artefact reduction method in order to provide clean fingerprints for the subsequent regression network. Our results achieve a decreased relative error by over 50% using our artefact reduction method compared to previously proposed deep learning regression model without prior artefact reduction. |
| 3765 | Dictionary-based Learning in MR Fingerprinting: Statistical Learning versus Deep Learning | |
| Fabien Boux1,2, Florence Forbes2, Julyan Arbel2, Aurélien Delphin1, Thomas Christen1, and Emmanuel L. Barbier1 | ||
1Univ. Grenoble Alpes, Inserm, U1216, Grenoble Institut Neurosciences, GIN, 38000, Grenoble, France, 2Univ. Grenoble Alpes, Inria, CNRS, G-INP, 38000, Grenoble, France |
||
In MR Fingerprinting, the exhaustive search in the dictionary may be bypassed by learning a mapping between fingerprints and parameter spaces. In general, the relationship between these spaces is particularly non-linear, which implies the use of advanced regression methods: deep learning frameworks but also methods based on statistical models have been proposed. In this study, we compare reconstruction time, accuracy and noise robustness of the conventional dictionary-matching method and two methods that handle the modelling of the non-linear relashionship with a neural network and a statistical inverse regression model. |
| 3766 | Natural, multi-contrast and quantitative imaging of the brain using tailored MR fingerprinting | |
| Pavan Poojar1,2, Enlin Qian1, Maggie Fung3, and Sairam Geethanath1,2 | ||
1Columbia University Magnetic Resonance Research Center, Columbia University in the City of New York, New York, NY, United States, 2Dayananda Sagar College of Engineering, Bangalore, India, 3GE Healthcare Applied Sciences Laboratory East, New York, New York, NY, United States |
||
Tailored MRF (TMRF) provides scheme to acquire non-synthetic, multi contrast images simultaneously in one sequence by tailoring the acquisition scheme derived in MRF. Natural contrast obtained from TMRF overcomes artifacts observed in MRF such as flow and errors in tissue parameter quantitation. We acquired TMRF and MRF data on the ISMRM/NIST phantom along with gold standard relaxometry data. We also acquired MRF and TMRF data for 5 volunteers and compared the natural and synthetic contrasts; and the relaxometric maps. The TMRF natural contrasts are similar to control experiments and do not contain flow artifacts found in synthetic contrasts. |
| 3767 | Fast T1 mapping with inversion recovery followed by radial acquisition and dictionary-matching reconstruction | |
| Qing Li1, Xiaoyue Zhou1, and Yi Sun1 | ||
1MR Collaborations, Siemens Healthcare Ltd., Shanghai, China |
||
A dictionary-based reconstruction method was applied to estimate T1 mapping from a series of highly undersampled images acquired with an inversion recovery–prepared FISP sequence and a radial readout. Motion robustness improved using interleaved slice acquisition and data rejection in T1 estimation. The measurement time was 4 s/slice. Phantom and in vivo knee results show an up to 50% data rejection results in less than 10% of quantification accuracy loss, compared to the T1 map reconstructed with the full dataset. |
| 3768
|
Compressed sensing applied to 1 mm isotropic multi-parametric imaging with 3D-QALAS: A phantom, volunteer, and patient study | |
| Shohei Fujita1,2, Akifumi Hagiwara1, Naoyuki Takei3, Ken-Pin Hwang4, Issei Fukunaga1, Shimpei Kato1,2, Masaaki Hori5, Ryusuke Irie1,2, Christina Andica1, Toshiaki Akashi1, Koji Kamagata1, Ukihide Tateishi6, Osamu Abe2, and Shigeki Aoki1 | ||
1Department of Radiology, Juntendo University, Tokyo, Japan, 2Department of Radiology, The University of Tokyo, Tokyo, Japan, 3MR Applications and Workflow, GE Healthcare, Tokyo, Japan, 4Department of Radiology, MD Anderson Cancer Center, Houston, TX, United States, 5Department of Radiology, Toho University Omori Medical Center, Tokyo, Japan, 6Department of Radiology, Tokyo Medical and Dental University, Tokyo, Japan |
||
Simultaneous relaxometry of T1, T2, and proton density by 3D-QALAS still requires relatively lengthy acquisition times and further acceleration is desired in clinical practice. Here, we applied compressed sensing combined with parallel imaging to 3D-QALAS to accelerate scan time. Quantitative values and tissue segmentation performance were compared with and without acceleration in phantom, healthy volunteers, and patients. With accelerated 3D-QALAS acquisition, whole brain 1 mm isotropic volume data with multiple parametric maps and co-registered tissue segmentation maps were obtained in less than 6 minutes, with comparable quality to 3D-QALAS without acceleration. |
| 3769 | Quantitative and Synthetic MRI for Breast Assessment in Clinical Acquisition Times | |
| Nina Pötsch1, Marcus Raudner1,2, Tom Hilbert3,4,5, Tobias Kober3,4,5, Elisabeth Weiland6, Panagiotis Kapetas1, and Pascal Baltzer1 | ||
1Department of Biomedical Imaging and Image-guided Therapy, Medical University of Vienna, Vienna, Austria, 2High Field MR Centre, Department of Biomedical Imaging and Image-guided Therapy, Medical University of Vienna, Vienna, Austria, 3Advanced Clinical Imaging Technology, Siemens Healthcare, Lausanne, Switzerland, 4Department of Radiology, Lausanne University Hospital and University of Lausanne, Lausanne, Switzerland, 5LTS5, Ecole Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland, 6Siemens Healthcare GmbH, Erlangen, Germany |
||
Contrast-enhanced breast MRI is the most sensitive tool for the detection of breast cancer. Contrast enhancement is however not specific to cancer: incidental benign findings regularly require further workup including ultimately avoidable follow-up scans and biopsies. Quantitative T1 and T2 measurements could be used as discriminative imaging markers to assist clinical decision-making. This feasibility study illustrates the potential clinical benefits of quantitative imaging for breast MRI by combining two fast and robust mapping techniques for high-resolution parameter mapping in a clinically feasible scan time of 7:32 min using prototype compressed sensing MP2RAGE and GRAPPATINI sequences with a harmonized protocol. |
| 3770 | Three-Dimensional Free-breathing Whole-Liver Simultaneous T1, R2*, and Fat-Fraction Quantification Using MR Multitasking | |
| Nan Wang1,2, Anthony G Christodoulou1, Yibin Xie1, Fei Han3, Xiaodong Zhong3, Sen Ma1,2, Xiaoming Bi3, Vibhas Deshpande4, and Debiao Li1 | ||
1Biomedical Imaging Research Institute, Cedars-Sinai Medical Center, Los Angeles, CA, United States, 2Bioengineering, University of California, Los Angeles, Los Angeles, CA, United States, 3Siemens Healthcare, Los Angeles, CA, United States, 4Siemens Healthcare, Austin, TX, United States |
||
Quantitative MRI has been playing a key role in the assessment and characterization of liver. However, current applications are still facing technical challenges. In this work, we proposed a novel technique based on MR Multitasking, which enables simultaneous T1, R2*, and FF quantification in one scan with free-breathing acquisition, 3D whole-liver coverage, and sufficient spatial resolution. The feasibility of the proposed method were demonstrated on the volunteer study (N = 8), showing that the T1, R2*, and FF quantification was repeatable in vivo and consistent with the results from the reference methods. |
| 3771 | Blood Flow and Permeability Imaging in a Rat Stroke Model Using 3D Compressed-Sensing DCE-MRI | |
| Radovan Jiřík1, Lucie Krátká1, Jiří Kratochvíla1, Ondřej Macíček1, Ye Tian2, Peter Scheer3, Jana Hložková3, Edward DiBella2, and Zenon Starčuk, jr.1 | ||
1Institute of Scientific Instruments of the CAS, Brno, Czech Republic, 2Utah Center for Advanced Imaging and Research, University of Utah, Salt Lake City, UT, United States, 3University of Veterinary and Pharmaceutical Sciences Brno, Brno, Czech Republic |
||
The main challenge of using DCE-MRI in stroke in combination with advanced pharmacokinetic models is the need for high temporal resolution, whole-brain coverage and a low SNR (especially in small-animal MRI) of the DCE signal (compared to DSC, due to the BBB and low fractional blood volume in brain). In this pilot study we test if compressed sensing could help solve these challenges. |
| 3772 | Snapshot Quantitative MRI: Application to Simultaneously T2 and T2* Mapping | |
| Jian Wu1, Xiaoyin Wang2, Hongjian He2, Lingceng Ma1, Shuhui Cai1, Congbo Cai1, and Jianhui Zhong3 | ||
1Department of Electronic Science, Xiamen University, Xiamen, China, 2The Center for Brain Imaging Science and Technology, Zhejiang University, Zhejiang, China, 3Department of Imaging Sciences, University of Rochester, Rochester, NY, United States |
||
Simultaneously quantifying T2 and T2* properties can provide great sensitivity and specificity to diseases. However, most of the existing methods are time-consuming. Here, we develop an overlapping-echo method for simultaneously achieving reliable T2 and T2* maps in a single shot. This method is robust to motion and the inhomogeneity of B0. Experimental results demonstrate that the resulting T2 and T2* values are in good agreement with those obtained with reference methods. |
| 3773 | CRISTINA: Cartesian Single and Triple Quantum Imaging of 23Na in the Brain | |
| Michaela A U Hoesl1, Lothar R Schad1, and Stanislas Rapacchi2 | ||
1Medical Faculty Mannheim, Heidelberg University, Mannheim, Germany, 2Aix-Marseille University, CRMBM, CNRS, Marseille, France |
||
Multi-quantum sodium imaging offers additional insights compared to standard single-quantum (SQ) images with information beyond tissue sodium concentration (TSC). Simulation including B0 inhomogeneity and stimulated echo signals of multi-quantum signals and spectra of phase cycle choices facilitated robust sequence development. The characteristic temporal evolution of the SQ and triple quantum (TQ) sodium is of interest and a multi echo sequence for SQ and TQ signal acquisition was developed, which was evaluated in phantom and 5 healthy volunteers in the brain. |
| 3774 | Phantom study on magnetic resonance imaging (MRI) T1 and T2 relaxation times measurements standardization | |
| Davide Cicolari1, Domenico Lizio2, Patrizia Pedrotti3, Monica Teresa Moioli2, Alessandro Lascialfari1, Manuel Mariani1, and Alberto Torresin2,4 | ||
1Department of Physics, Università degli Studi di Pavia, Pavia, Italy, 2Department of Medical Physics, ASST Grande Ospedale Metropolitano Niguarda, Milan, Italy, 3Department of Cardiology, ASST Grande Ospedale Metropolitano Niguarda, Milan, Italy, 4Department of Physics, Università degli Studi di Milano, Milan, Italy |
||
Relaxation times measurement standardization, a great issue in clinical inter-centre and inter-scanner applications, is studied by comparing relaxation times maps of a MnCl2 phantom, scanned with two different MRI imagers, with reference values measured with an NMR laboratory spectrometer. For this study standard sequences were used: NMR reference T1 and T2 values were obtained from IR (Inversion-Recovery) and CPMG (Carr-Purcell-Meiboom-Gill) sequences respectively; MRI maps were generated from clinical IR and SE (Spin-Echo) sequences. The MRI and NMR results agreement within the experimental error limits (5%) suggests that the estimation of the relaxation times is independent from the spectrometer/scanner utilized. |
| 3775 | Synthetic MRI with T2-based Water Suppression without Loss of Tissue SNR | |
| Tokunori Kimura1, Kousuke Yamashita1, and Kouta Fukatsu1 | ||
1Department of Radiological Science, Shizuoka College of Medicare Science, Hamamatsu-Shi, Japan |
||
We proposed a modified T2-based water suppression synthetic-MRI technique without loss of tissue SNR, which was introduced by subtraction of heavy-T2W image from standard acquired images. Our water suppression was achieved by subtracting only water portions except for tissue portions. We demonstrated both effects that CSF-PVE artifacts were dramatically suppressed and the tissue SNR was kept to before subtraction in our water suppressed quantitative maps; and thus water suppressed synthetic images of FLAIR and SE provided better gray-white matter contrasts than those for subtracting uniformly. |
| 3776 | Evaluating Compressed SENSE acceleration for multi-parametric quantitative mapping of R1, R2*, PD, and MTsat with the hMRI toolbox | |
| Ronja Berg1, Tobias Leutritz2, Stephan Kaczmarz1, Claus Zimmer1, Nikolaus Weiskopf2, and Christine Preibisch1 | ||
1School of Medicine, Department of Neuroradiology, Technical University of Munich, Munich, Germany, 2Max Planck Institute for Human Cognitive and Brain Sciences, Department of Neurophysics, Leipzig, Germany |
||
Measuring physical parameters quantitatively by magnetic resonance imaging (MRI) has a high value for diagnostic applications as it allows the detection of disease related systemic changes. However, quantitative MRI mapping methods increase the scan time significantly compared to conventional imaging. Therefore, we investigated the applicability of Compressed SENSE (CS) acceleration in a multi-parametric mapping protocol for R1, R2*, PD, and MTsat imaging. Our results demonstrate that absolute parameter values remained constant in all evaluated regions-of-interest when applying CS. Thus, CS can be used to almost halve the scan for R1, R2*, PD, and MTsat mapping without loss of fidelity. |
| 3777 | Ex vivo mapping of the cyto- and the myeloarchitecture of the human cerebral cortex using ultra-high field MRI (7T and 11.7T) | |
| Raïssa Yebga Hot1,2, Alexandros Popov1,2, Justine Beaujoin1, Gaël Perez1,3, Fabrice Poupon1,2, Igor Lima Maldonado4, Jean-François Mangin1,2, Christophe Destrieux4, and Cyril Poupon1,2 | ||
1CEA - NeuroSpin, Gif-sur-Yvette, France, 2Université Paris-Saclay, Orsay, France, 3CentraleSupélec, Gif-sur-Yvette, France, 4Imaging and Brain laboratory (iBrain), Université de Tours - INSERM, Tours, France |
||
The investigation of the human cerebral cortex at the mesoscopic scale remains challenging but promising to better understand brain pathologies associated with cortex damage. In this ex vivo study, samples of occipital cortex from both hemispheres of an unique subject have been delineated using their microstructural and myeloarchitectural information inferred from ultra-high field anatomical, quantitative and diffusion MRI. The high-resolved UHF-MRI dataset enabled to perform an automatic segmentation of the cortical layers within the primary and secondary visual cortices. The segmentations highlighted their commonalities and differences between the two hemispheres. |
| 3778 | Cross vendor validation of Hybrid Multidimensional MRI in the non-invasive measurement of prostate tissue composition | |
| Aritrick Chatterjee1, Grace Lee2, Deb Dietz2, Aytekin Oto1, and Gregory Karczmar1 | ||
1Department of Radiology, University of Chicago, Chicago, IL, United States, 2Ingalls Memorial Hospital, Flossmoor, IL, United States |
||
This study evaluates the consistency of HM-MRI for non-invasive measurement of prostate tissue composition on scanners from two MRI vendors. HM-MRI was performed on Philips Achieva 3T (with endorectal coil) and Siemens Skyra 3T (without endorectal coil) MRI scanners. HM-MRI metrics measured for cancerous and benign prostatic tissue using Philips and Siemens were similar, with slight variation due to different patients in each cohort. Diagnostic accuracy for detecting PCa using HM-MRI was similar for both MR vendors: Philips (AUC = 0.94-0.99, p<0.05) and Siemens (AUC = 0.83-0.98, p<0.05). |
| 3779 | Accuracy assessment of high-spatial resolution whole-brain tracer-kinetic parameter maps generated using dual-temporal resolution DCE-MRI | |
| Ka-Loh Li1, Daniel Lewis2, Sha Zhao1, Alan Jackson1, and Xiaoping Zhu1 | ||
1Division of Informatics, Imaging and Data Sciences, The University of Manchester, Manchester, United Kingdom, 2Manchester Centre for Clinical Neurosciences, Salford Royal NHS Foundation Trust, Manchester, United Kingdom |
||
Accurate, high-spatial resolution whole-brain pharmacokinetic maps are highly desirable in clinical neuro-oncology practice. A new dual-temporal resolution based kinetic mapping technique for this purpose, termed LEGATOS, was recently described and tested with in-vivo patient data. In this study, quantitative assessment of the accuracy of parameter estimates derived using the LEGATOS analysis procedure was evaluated through computer simulation. Structural similarity and percentage deviation of the “measured” values from the known “true” values were used for evaluation and demonstrated that the LEGATOS technique offered superior accuracy compared to the use of unreconstructed composite HT and HS curves alone. |
| 3780 | T1ρ could be a complementary tool for T2* in the assessment of rat liver iron overload | |
| Qianfeng Wang1, Hong Xiao2, He Wang1, Xuchen Yu1, and Fuhua Yan2 | ||
1Institute of Science and Technology for Brain-Inspired Intelligence, Fudan University, Shanghai, China, 2Department of Radiology, Ruijin Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China |
||
In this study, spin-lock and 2D UTE pulse sequences were developed to quantify T1ρ and T2* of 32 rat livers with iron overload at 11.7T MR system. Moreover, T2 was also acquired to compare with T1ρ and T2*. Pearson correlation analyses (two-tailed) illustrated that T1ρ were significantly associated with T2 (all p ≤ 0.002), but insignificantly related with T2* (all p > 0.05). No significant association was found between T2 and T2*. |
| 3781 | Simultaneous B1- and Fat-Corrected T1 Mapping Using Chemical-Shift Encoded MRI | |
| Nathan Tibbitts Roberts1,2, Diego Hernando1,2,3,4, Timothy Colgan1, Daiki Tamada1, and Scott B Reeder1,3,4,5,6 | ||
1Radiology, University of Wisconsin - Madison, Madison, WI, United States, 2Electrical and Computer Engineering, University of Wisconsin - Madison, Madison, WI, United States, 3Medical Physics, University of Wisconsin - Madison, Madison, WI, United States, 4Biomedical Engineering, University of Wisconsin - Madison, Madison, WI, United States, 5Medicine, University of Wisconsin - Madison, Madison, WI, United States, 6Emergency Medicine, University of Wisconsin - Madison, Madison, WI, United States |
||
Spatially varying B1 inhomogeneities and tissue fat are known confounders of quantitative T1 mapping methods that use variable flip angle techniques. Separately acquired B1 calibration maps can be used to correct flip angle errors caused by B1 inhomogeneities, but this requires an additional acquisition. In this work we propose a novel approach for simultaneous estimation of B1, T1, proton density fat-fraction and R2* using dual orthogonal RF pulses and multiple flip angles. The feasibility and noise performance of this proposed acquisition and fitting strategy are evaluated using Cramer-Rao Lower Bound analysis, simulations, and preliminary phantom experiments. |
| 3782 | Increasing robustness to fat partial volume effects in T1-mapping with MOLLI | |
| Andreia S Gaspar 1, Andreia C Freitas1, and Rita G Nunes1 | ||
1ISR-Lisboa/LARSyS and Department of Bioengineering, Instituto Superior Técnico – Universidade de Lisboa, Lisbon, Portugal |
||
Modified Look-Locker inversion recovery (MOLLI) with a balance Steady State Free Precession (bSSFP) readout is widely applied in the clinical setting but it is known to be sensitive to fat partial volume effects (PVE), as well as static B0 field inhomogeneities (ΔB0). As an alternative, the FLASH readout, more robust to ΔB0, can be applied for T1 mapping. In this work we studied the robustness of FLASH-MOLLI to fat PVE, and propose a new bi-component model to improve the accuracy of water T1 estimates in both bSSFP and FLASH MOLLI. |
| 3783 | In Vivo Validation of MRS- and MRI-Based Fatty Acid Composition Quantification against Gas Chromatography in Adipose Tissue | |
| Lena Trinh1, Pernilla Peterson1,2, Håkan Brorson3, and Sven Månsson1,4 | ||
1Medical Radiation Physics, Department of Translational Medicine, Lund University, Malmö, Sweden, 2Medical Imaging and Physiology, Skåne University Hospital, Lund, Sweden, 3Department of Plastic and Reconstructive Surgery, Skåne University Hospital, Malmö, Sweden, 4Radiation Physics, Skåne University Hospital, Malmö, Sweden |
||
Non-invasive estimation of the fatty acid composition of adipose tissue using MRI or MRS may be valuable in a number of disease scenarios. However, in vivo validation against an independent gold standard is still needed. In this work, we find that especially MRI estimation of the fraction of saturated fatty acids is strongly associated to gas chromatography analysis in the adipose tissue of lymphedema patients. The reliability of the estimated mono- and polyunsaturated fractions are reliant on the used signal model. Also MRS provided results which are associated to GC analysis, but with a lower agreement compared to MRI. |
| 3784 | Initial Experience with Inline Evaluation of Liver Stiffness using Magnetic Resonance Elastography in Hemochromatosis Patients | |
| Stephan A.R. Kannengiesser1, Cathy Cazin2, Michel Lapp2, Khalid Ambarki3, Berthold Kiefer1, and Valérie Laurent2,4 | ||
1MR Application Predevelopment, Siemens Healthcare GmbH, Erlangen, Germany, 2Department of Radiology, CHRU Nancy, Brabois Adults Hospital, Vandoeuvre-lès-Nancy, France, 3Siemens Healthcare SAS, Saint Denis, France, 4IADI, U1254, INSERM, Université de Lorraine, CHRU de Nancy Brabois, Vandoeuvre-lès-Nancy, France |
||
A simple prototype inline mean stiffness evaluation approach for MR Elastography based on anatomical liver segmentation and the confidence map was evaluated in 25 hemochromatosis patients, and compared with manual evaluation by two readers according to QIBA guidelines. Pairwise comparisons were found to be of similar statistics: mean±std relative difference -3.40±5.10% and -0.69±6.72% for reader1 vs. reader2 and manual (average of readers) vs. inline. The presented approach is promising given the particular patient cohort and limited stiffness range, but further evaluation is needed. |
| 3785 | Fast, high resolution QSM of the brainstem at 7T using super-resolution 2D EPI | |
| Andreas Ehrmann1, Beata Bachrata1,2, Korbinian Eckstein1, David Bancelin1, Pedro Lima Cardoso1, Charles Joseph Guthrie1, Siegfried Trattnig1,2, and Simon Daniel Robinson1,3,4 | ||
1High Field MR Centre, Department of Biomedical Imaging and Image-guided Therapy, Medical University of Vienna, Vienna, Austria, 2Christian Doppler Laboratory for Clinical Molecular MR Imaging, Vienna, Austria, 3Department of Neurology, Medical University of Graz, Graz, Austria, 4Centre for Advanced Imaging, University of Queensland, Queensland, Australia |
||
2D EPI-based Quantitative Susceptibility Mapping offers a means to localise brainstem structures very quickly. However, this comes at the price of limited achievable slice thickness, restricting the precision of the localisation. In this work, we develop a fast super-resolution reconstruction method for complex MRI data, allowing the reconstruction of high resolution magnitude and phase images, and hence also high resolution susceptibility maps. We demonstrate that the reconstructed super-resolution 2D EPI images significantly improve the visibility of small structures allowing for improved localisation of the brainstem and other small structures such as veins in the generated susceptibility maps. |
| 3786 | Towards robust QSM in cortical and sub-cortical regions of the human brain at 9.4T: influence of coil combination and masking strategies | |
| Gisela E Hagberg1,2, Korbinian Eckstein3, Enrique Cuna4, Simon Robinson3,5, and Klaus Scheffler1,2 | ||
1Biomedical Magnetic Resonance, University Hospital Tübingen, Tübingen, Germany, 2High Field Magnetic Resonance, Max Planck Institute for Biological Cybernetics, Tübingen, Germany, 3High Field MR Centre, Medical University of Vienna, Vienna, Austria, 4Centro Uruguayo de Imagenología Molecular (CUDIM), Montevideo, Uruguay, 5Centre for Advanced Imaging, University of Queensland, Brisbane, Australia |
||
Strong background signals leading to multiple phase wraps may hamper accurate quantification of magnetic tissue susceptibility (QSM) especially at high field strengths using long echo times to achieve a high spatial sampling. Here we show how different coil-combination, automated tissue masking and background removal techniques can be used to improve QSM quality. Performance was evaluated with regard to iron quantification in subcortical and cortical areas in the same subjects. We found a substantial improvement in accuracy and precision of QSM in high-field applications at long echo times through the use of ASPIRE and removal of areas with excessive phase evolution. |
| 3787 | The cerebral blood flow derived from multi-delay ASL MRI varies with the choice of post-labeling delay | |
| Ya-Fang Chen1, Sung-Chun Tang2, and Wen-Chau Wu3,4 | ||
1Department of Medical Imaging, National Taiwan University Hospital Hsin-Chu Branch, Hsinchu, Taiwan, 2Department of Neurology, National Taiwan University Hospital, Taipei, Taiwan, 3Institute of Medical Device and Imaging, National Taiwan University, Taipei, Taiwan, 4Graduate Institute of Clinical Medicine, National Taiwan University, Taipei, Taiwan |
||
With arterial spin labeling (ASL) MRI, the delay time between labeling and image acquisition is a critical imaging parameter for accurate flow quantification. Multi-delay ASL has been proposed to tackle the problem of unknown arterial transit time (ATT). However, it is not fully understood how the range of delay times affects the derived CBF. The present study aimed to investigate this potential issue. Results show that the CBF derived from multi-PLD ASL varies with the choice of PLD (overestimated when the PLD range does not cover ATT) as well as SNR (overestimated along with greater variability when SNR is low). |
| 3788 | To Evaluate the Effect of Compressed-SENSE on Accuracy of High-Resolution Quantitative T1-Perfusion MRI Parameters and on glioma grading at 3T | |
| Dinil Sasi S1, Rakesh K Gupta2, Indrajit Saha3, Anandh K Ramaniharan3, and Anup Singh4,5 | ||
1Indian Institute of Technology Delhi, New Delhi, India, 2Fortis Memorial Research Institute, Gurugram, India, 3Philips India Limited, Gurugram, India, 4Indian Institute of Technology Delhi, Hauz Khas, India, 5All India Institute of Medical Science, New Delhi, India |
||
T1-perfusion derived parameters have been utilized for brain tumor quantification and grading. In this study, we have analyzed the potential of Compressed SENSE (CSENSE) acceleration technique on high resolution T1-perfusion MRI for quantitative analysis of brain tumor(glioma) and compared its performance and accuracy with conventional acquisition protocol in terms of concentration-time curve, perfusion parameters and glioma grading. A prospective analysis was also carried on healthy subjects to analyze the spatial error propagation at different acceleration factors(R). All derived perfusion parameters were able to quantify and differentiate different tumor classes with improved concentration-time curve(High grade and low grade glioma). |
| 3789 | Estimation of net receive sensitivity - at 3T and 7T - for correction of inter-scan motion artefacts in R1 mapping | |
| Yaël Balbastre1, Nadège Corbin1, and Martina F Callaghan1 | ||
1Wellcome Centre for Human Neuroimaging, University College London, London, United Kingdom |
||
The longitudinal relaxation rate is a useful myelin proxy that can be estimated by combining multiple acquisitions with different nominal flip angles. Inter-scan motion can introduce large biases that can be efficiently corrected if reference data with a relatively ‘flat’ sensitivity profile, e.g. a body coil, is available. This is generally not the case at ultra-high field where highly inhomogeneous localised transmit-receive coils are used. Therefore, we propose a new method for estimating and removing the net coil sensitivity that does not require a body coil, needs only coil-wise magnitude images, yet reduces motion-induced bias in R1. |
| 3790 | Acceleration of 3D high-resolution variable-flip-angle T1 mapping based on SUPER | |
| Fan Yang1, Jun Xie2, Guobin Li2, Meng Jiang3, and Chenxi Hu4 | ||
1College of Life Science and Technology, Huazhong University of Science and Technology, Wuhan, China, 2United Imaging Healthcare Co., Ltd, Shanghai, China, 3Department of Cardiology, Renji Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai, China, 4Institute of Medical Imaging Technology, School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China |
||
Three-dimensional variable-flip-angle(VFA) T1-mapping is a rapid T1 quantification method with important clinical applications. However, the requirement of acquiring multiple 3D images greatly increases the scan time. We propose a novel application of SUPER—a recently developed framework for parametric mapping acceleration—to reduce scan time and/or improve spatial resolution of 3D VFA T1-mapping. In healthy subjects, we demonstrate that SUPER(R=2) and SUPER-SENSE(R=4 combining SUPER and parallel imaging) achieved similar accuracy, reasonable noise amplification, and similar reconstruction time compared with the non-acceleration gold standard. The whole upper-brain T1-mapping scan time was reduced from 5.33 minutes to 1.33 by employment of SUPER-SENSE. |
| 3791 | Thinking outside the black box: A fully transparent T1 mapping pipeline | |
| Agah Karakuzu1,2, Mathieu Boudreau1,2, Julien Cohen-Adad1,3, and Nikola Stikov1,2 | ||
1NeuroPoly Lab, Institute of Biomedical Engineering, Polytechnique Montreal, Montreal, QC, Canada, 2Montreal Heart Institute, Montreal, QC, Canada, 3Unité de Neuroimagerie Fonctionnelle (UNF), Centre de recherche de l’Institut Universitaire de Gériatrie de Montréal (CRIUGM), Montreal, QC, Canada |
||
Quantitative MRI is pointless if we cannot peak inside the black box that generates the numbers. This is the main motivation behind our concept for a fully transparent qMRI pipelines starting at the scanner console and extending all the way to a journal publication. In this work, we developed a simple application to demonstrate this for variable flip angle (VFA) T1 mapping. This application exemplifies the feasibility of a vendor-neutral qMRI sequence that is publicly shared under version control (https://github.com/qMRLab/pulse_sequences). Our approach opens a way towards community-driven development and improvement of qMRI applications without the need for specialized programming skills. |
| 3792 | Brain T1 relaxometry changes across the life span: a comparison between two populations | |
| Gian Franco Piredda1,2,3, Peipeng Liang4, Tom Hilbert1,2,3, Karl Egger5, Shan Yang5, Jean-Philippe Thiran2,3, Bénédicte Maréchal1,2,3, Yi Sun6, Kuncheng Li7,8, and Tobias Kober1,2,3 | ||
1Advanced Clinical Imaging Technology, Siemens Healthcare AG, Lausanne, Switzerland, 2Department of Radiology, Lausanne University Hospital and University of Lausanne, Lausanne, Switzerland, 3LTS5, École Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland, 4School of Psychology, Capital Normal University, Beijing Key Laboratory of Learning and Cognition, Beijing, China, 5Faculty of Medicine, University of Freiburg, Freiburg, Germany, 6MR Collaboration, Siemens Healthcare Ltd., Shanghai, China, 7Department of Radiology, Xuanwu Hospital, Capital Medical University, Beijing, China, 8Beijing Key Laboratory of Magnetic Resonance Imaging and Brain Informatics, Beijing, China |
||
Understanding the normal evolution of T1 values across the life span allows to disentangle ageing from degenerative pathologies. Following previous studies reporting brain anatomical and functional differences between Western and Chinese cohorts, this work investigates whether differences in the evolution of T1 values across the life span exist between these two populations using two datasets with 200 healthy subjects each. Derived trends were found to differ between the two populations in some brain structures, especially in grey matter tissues. The observed differences may indicate that norms derived from one population may not be directly applied to another without recalibration. |
| 3793 | Variable flip angle T1 mapping with MT-balanced RF pulses | |
| Nam Gyun Lee1, Zhibo Zhu2, and Krishna S. Nayak2 | ||
1Biomedical Engineering, University of Southern California, Los Angeles, CA, United States, 2Electrical and Computer Engineering, University of Southern California, Los Angeles, CA, United States |
||
Quantitative DCE requires accurate and precise pre-contrast M0/T1 maps. Variable flip angle (VFA) T1 mapping has good precision but has accuracy issues. Magnetization transfer (MT) effects is one of the major reasons for differences between VFA and reference IR-FSE T1 measurements. Here, we apply the recent framework for MT-balanced RF pulse design to high-resolution whole-brain T1 mapping. |
| 3794 | Quantitative T2 Mapping using Accelerated 3D Stack-of-Spiral GRE Acquisition | |
| Ruoxun Zi1, Dan Zhu1,2, Wenbo Li2,3, 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, 3F.M. Kirby Research Center for Functional Brain Imaging, Kennedy Krieger Institute, Baltimore, MD, United States |
||
We presented a T2 prepared stack-of-spiral gradient echo (GRE) pulse sequence with or without CSF nulling for T2 mapping of brain. T2 maps were iteratively optimized by data consistency, model consistency and spatial sparsity regularization using projected-gradients approach, which was evaluated by in vivo experiments and compared with SENSE, CS SENSE and model-based methods. The proposed reconstruction method provided better performance with normalized root mean square error (nRMSE) of the whole brain T2 estimation equal to 8.2±0.5% with an acceleration factor of 5. The T2 estimation with and without CSF nulling were consistent. |
| 3795 | Joint reconstruction and estimation for fast and accurate T2* mapping | |
| Seonyeong Shin1, Riwaj Byanju2, Seong Dae Yun1, Stefan Klein2, Dirk H. J. Poot2, and N. Jon Shah1,3,4,5 | ||
1Institute of Neuroscience and Medicine 4, INM-4, Forschungszentrum Jülich, Jülich, Germany, 2Department of Radiology and Nuclear Medicine, Erasmus MC, Rotterdam, Netherlands, 3Institute of Neuroscience and Medicine 11, INM-11, JARA, Forschungszentrum Jülich, Jülich, Germany, 4JARA - BRAIN - Translational Medicine, Aachen, Germany, 5Department of Neurology, RWTH Aachen University, Aachen, Germany |
||
Quantification of T2* is relatively time-efficient. However, still the scan times might be too long for the desired resolution. In this work, we demonstrated the ability to accelerate the acquisition by joint reconstruction and parameter estimation for T2* quantification. From retrospectively highly subsampled k-spaces, images and T2* maps were reconstructed directly. It was shown that an acceleration factor of 16 is feasible for T2* mapping. |
| 3796 | A 3D-Printed Multi-level Perfusion Phantom for Quantitative Analysis of Perfusion Models vs. Experimental Data | |
| John Morgan1 and Yi Wang1 | ||
1Cornell University, New York, NY, United States |
||
Current methods for perfusion characterization are difficult to quantify absolutely and provide only relative and qualitative information. Perfusion-phantoms that enable quantitative analysis of transport phenomena are needed to test theoretical models against experimental data. The initial design and deployment of a 3D-printed phantom with pump-driven perfusion of multi-level microvascular structure encapsulated in hydrogel is presented. It simulates vascularized tissue and enables experimental validation. A new quantitative analytical method , using voxelized constitutive convection-diffusion equations, is applied to DCE data and compared to traditional Kety’s methods. The largely qualitative and unmeasurable Kety global AIF assumption is replaced with measurable and reproducible data. |
| 3797 | Variable Spatial Resolution Dual-Venc 4D Flow MRI: Balancing Image Quality and Scan Time | |
| Maria Aristova1, Jianing Pang2, Liliana Ma1, Michael Markl1, and Susanne Schnell1 | ||
1Radiology, Northwestern University, Chicago, IL, United States, 2Siemens Healthcare, Chicago, IL, United States |
||
Dual-venc 4D flow MRI has previously been reported to achieve wide velocity dynamic range while maintaining a high velocity-to-noise ratio, which is particularly relevant for neurovascular applications. To increase flexibility in spatiotemporal resolution, we present a prototype implementation of a interleaved 8-point dual-venc 4D Flow acquisition with independently prescribed (prospectively undersampled) spatial resolution of the high venc acquisition, i.e. Variable Spatial Resolution Dual Venc (VSRDV). This allowed anti-aliasing error rates less than 15% error for a high venc spatial resolution of 70-80%. This represents 10-15% reduction of scan time. |
| 3798 | Validation of dynamic contrast-enhanced MRI analyses via virtual MRI simulation on a dynamic digital phantom | |
| Chengyue Wu1, Ty Easley2, Victor Eijkhout3, David A. Hormuth4, Federico Pineda5, Gregory S. Karczmar5, and Thomas Yankeelov1,4,6,7 | ||
1Department of Biomedical Engineering, University of Texas at Austin, Austin, TX, United States, 2Department of Biomedical Engineering, Washington University in St. Louis, St. Louis, MO, United States, 3HPC Software Tools Group, Texas Advanced Computing Center, Austin, TX, United States, 4Oden Institute for Computational Engineering and Sciences, University of Texas at Austin, Austin, TX, United States, 5Department of Radiology, University of Chicago, Chicago, IL, United States, 6Department of Diagnostic Medicine, University of Texas at Austin, Austin, TX, United States, 7Department of Oncology, University of Texas at Austin, Austin, TX, United States |
||
We have recently developed novel image processing and computational methods to characterize both the morphology and hemodynamics of tumor-associated vessels to assist in the diagnosis of suspicious breast lesions. In this contribution, we employ a dynamic digital phantom of contrast agent perfusion and extravasation to systematically evaluate the sensitivity of our methods to spatial resolution, temporal resolution, and signal-to-noise ratio (SNR). The immediate goal is to use the phantom to determine an acquisition-reconstruction protocol that can implemented in the routine clinical setting, thereby enabling quantitative hemodynamic analysis to be widely available for the characterization of cancer. |
| 3799 | Optimization of Keyhole Imaging Parameters for Glutamate Chemical Exchange Saturation Transfer (GluCEST) MRI at 7.0 T | |
| Dong-Hoon Lee1, Do-Wan Lee2, Chul-Woong Woo3, Hwon Heo4, Jae-Im Kwon3, Yeon Ji Chae4, Su Jung Ham5, Jeong Kon Kim2, Kyung Won Kim2,5, and Dong-Cheol Woo3,4 | ||
1Faculty of Health Sciences and Brain & Mind Centre, The University of Sydney, Sydney, Australia, 2Department of Radiology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of Korea, 3Convergence Medicine Research Center, Asan Institute for Life Sciences, Asan Medical Center, Seoul, Republic of Korea, 4Department of Convergence Medicine, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of Korea, 5Asan Image Research, Asan Institute for Life Sciences, Asan Medical Center, Seoul, Republic of Korea |
||
We evaluated the effects of a reference image and keyhole factor selections for high-frequency substitution on a keyhole imaging technique for application in glutamate CEST (GluCEST) imaging to reduce data acquisition time. The calculated GluCEST signals and visually inspected results from the reconstructed GluCEST maps indicated that a combination of unsaturated image as a reference image and >50% of keyhole factors showed consistent signals and image quality as opposed to the fully-sampled CEST data. Combining the keyhole imaging technique with GluCEST imaging enables stable image reconstruction and quantitative evaluation, and this approach is potentially implemented in various CEST imaging applications. |
| 3800 | Quantitative Macromolecular Proton Fraction Imaging Based On Spin-Lock | |
| Jian Hou1, Vincent Wong2, Baiyan Jiang1, Yixiang Wang1, Anthony Chan3, Winnie Chu1, and Weitian Chen1 | ||
1Department of Imaging and Interventional Radiology, The Chinese University of Hong Kong, Hong Kong, Hong Kong, 2Department of Medicine & Therapeutics, The Chinese University of Hong Kong, Hong Kong, Hong Kong, 3Department of Anatomical and Cellular Pathology, The Chinese University of Hong Kong, Hong Kong, Hong Kong |
||
Macromolecular Proton Fraction (MPF) is the relative amount of protons associated with macromolecules involved in magnetization transfer with free water protons. MPF is typically measured by quantitative magnetization transfer methods. In this work, we reported that MPF can also be measured based on spin-lock. Compared to the existing MPF methods, our method requires fewer parameter maps. We demonstrated our method using simulation, phantom, and in vivo experiments. |
| 3801 | Constraints in Estimating the Proton Density Fat Fraction | |
| Mark Bydder1, Vahid K Ghodrati1, Yu Gao1, Matthew D Robson2, Yingli Yang1, and Peng Hu1 | ||
1UCLA, Los Angeles, CA, United States, 2Perspectum Diagnostics, Oxford, United Kingdom |
||
The study evaluates four physically motivated constraints in the estimation of the proton density fat fraction (PDFF) based on the physics of magnetic resonance imaging. These were smooth fieldmap, smooth initial phase, nonnegative proton density and moderate $$$R2*$$$ values. Results show that constraints are effective at reducing standard deviation and bias. |
| 3802 | Faster Myelin water mapping by joint reconstruction and estimation from highly undersampled 3D-MSE acquisitions. | |
| Riwaj Byanju1, Stefan Klein1, and Dirk H. J. Poot1 | ||
1Department of Radiology and Nuclear Medicine, Erasmus Medical Center, Rotterdam, Netherlands |
||
Myelin water fractions (MWF) can provide biomarkers for many brain disorders. Multiple-spin-echo (MSE) is considered as gold standard for obtaining MWF. However, its clinical application is limited by its long acquisition time. Efficient undersampling and exploiting the redundancy in the contrast images of MSE can decrease the acquisition time. We propose a subspace based image reconstruction technique designed for reconstructing contrast images from highly undersampled MSE acquisition, so that it can be used for MWF mapping. We evaluate it for different signal-to-noise ratios for up to acceleration factor of 16 and show its feasibility for accelerating acquisition beyond parallel imaging. |
| 3803 | Improved T2 Quantification using Noise Corrected Stimulated Echo Compensation | |
| Fei Han1, Mahesh Bharath Keerthivasan2, and Vibhas Deshpande3 | ||
1MR R&D and Collaboration, Siemens Healthineers, Los Angeles, CA, United States, 2MR R&D and Collaboration, Siemens Healthineers, Tucson, AZ, United States, 3MR R&D and Collaboration, Siemens Healthineers, Austin, TX, United States |
||
Quantitative imaging is the key to developing reliable and reproducible imaging methods for standardized diagnostic exams. Most practical T2 mapping solutions are based on the CPMG concept, using the SEMC or TSE sequence. However, several confounding variables, such as, but not limited to the B1, RF profile, refocusing flip-angle, choice of TEs, and the imaging noise/artifacts, can lead to distorted signal and inaccurate T2 quantification. In this work, we investigated some of these confounders in CPMG-based T2 quantification and proposed a new model and fitting methods to improve the reliability, reproducibility of in-vivo T2 quantification. |
| 3804 | Method for Automatic blood vessel removal from quantitative T1-perfusion MRI maps and evaluating its impact on tumor grading | |
| Manish Awasthi1, Bansmita Kar1, Neha Vats1, Virendra Kumar Yadav1, Dinil Sasi1, Mamta Gupta2, Rakesh Kumar Gupta2, and Anup Singh1,3 | ||
1Centre for Biomedical Engineering, Indian Institute of Technology, Delhi, New Delhi, India, 2Department of Radiology, Fortis Memorial Research Institute, Gurugram, India, 3Biomedical Engineering, All India Institute of Medical Science, Delhi, New Delhi, India |
||
Presence of large blood vessels within tumor region can mislead interpretation on quantitative T1-perfusion MRI, particularly using automatic classification approaches. Purpose of this study was to develop a methodology for automatic blood vessel removal from quantitative T1-perfusion maps, compare it with previously reported methodology and finally evaluating impact of blood-vessel removal on tumor grading. In the proposed approach, signal intensity time curves characteristics, particularly contrast wash-out rate and peak value provided accurate automatic removal of blood-vessel from tumor region. Significant differences between T1-perfusion maps with and without blood-vessel removal were observed and tumor grading were also influenced. |
| 3805 | Statistical Analysis of Parameter Estimation for Multiexponential Decay in MR Relaxometry in One, Two, and Higher Dimensions | |
| Richard G Spencer1, Mustapha Bouhrara1, Kenneth W Fishbein1, and Joy You Zhuo1 | ||
1National Institute on Aging, NIH, Baltimore, MD, United States |
||
Analysis of multiexponential decay has remained a topic of active research for over 200 years, attesting to the difficulty in deriving the rate constants and component amplitudes of the underlying monoexponential decays. We have shown previously that parameter estimates from 2D relaxometry, with two distinct time variables, exhibit substantially greater accuracy and precision than those from analysis of 1D data, with a single time variable. We now present a statistical study of this remarkable phenomenon and indicate applications in 2D relaxometry and related experiments. These results provide a potential means of circumventing conventional limits on multiexponential parameter estimation. |
| 3806 | CNN-based synthesis of T1, T2 and PD parametric maps of the brain with a minimal input feeding | |
| Elisa Moya-Sáez1, Óscar Peña-Nogales1, Santiago Sanz-Estébanez1, Rodrigo de Luis-Garcia1, and Carlos Alberola-López1 | ||
1Laboratorio de Procesado de Imagen, Universidad de Valladolid, Valladolid, Spain |
||
Parametric MR maps (T1, T2 and PD) not only play a key role in quantitative imaging but they also have the capability of synthesizing any modality. However, their direct acquisition is hardly used in practice due to the need of lengthy relaxometry protocols. Synthetic MRI is a surrogate; however, no approach has been described to synthesize these maps out of a small number of customary sequences. In this work we synthesize T1, T2 and PD maps out of a T1- and a T2-weighted image using a CNN trained only with synthetic data. Our approach yields realistic maps from real data. |
| 3807 | Cross-Field Strength and Cross-Vendor Reliability of Quantitative MR-based Breast Density (MagDensity) | |
| Renee Cattell1, Shenglan Chen1, Jie Ding1,2, and Chuan Huang1,3,4 | ||
1Biomedical Engineering, Stony Brook University, Stony Brook, NY, United States, 2Diagnostic Radiology, Li Ka Shing Faculty of Medicine, The University of Hong Kong, Hong Kong, Hong Kong, 3Radiology, Stony Brook University, Stony Brook, NY, United States, 4Psychiatry, Stony Brook University, Stony Brook, NY, United States |
||
Quantitative measurement of breast density is important for personalized risk assessment and frequent, longitudinal monitoring of breast density modifying drugs. Current standard of care for assessing breast density is qualitative assessment of mammogram, which involves ionizing radiation and breast compression. MRI-derived breast density (MagDensity) using fat-water decomposition has the potential to provide a sensitive and useful tool for clinicians. This study aims to prospectively determine the reliability and reproducibility of this measure across different vendors, scanner models and field strengths |
| 3808 | Rapid 3D sub-millimeter isotropic T1ρ mapping of the knee using motion-robust interleaved spin-lock acquisition with compressed sensing | |
| Keita Nagawa1, Suzuki Masashi1, Masami Yoneyama2, Kaiji Inoue1, Eito Kozawa1, and Mamoru Niitsu1 | ||
1Saitama Medical University, Saitama, Japan, 2Philips Japan, Tokyo, Japan |
||
For T1ρ quantification, a three-dimensional (3D) acquisition is desired to obtain high-resolution images. In order to achieve a rapid and robust acquisition of 3D T1 rho mapping data, we examined here the 3D sub-millimeter isotropic resolution sequences, applying them to the assessment of knee-joint. 3D isotropic resolution sequences can reduce partial-volume artifacts through the acquisition of thin continuous sections through joints. Furthermore, the isotropic source data can be used to create multiplanar reformations (MPRs). With the optimized voxel size (0.8mm3), we could obtain motion-robust high-quality isotropic images within the time constraints of a clinical exam (<10 minutes) . |
| 3809 | CNNs improve tissue sodium concentration accuracy in white and grey matter from stroke patients at 3T 23Na MRI | |
| Anne Adlung1, Nadia K. Paschke1, Alena-Kathrin Schnurr1, Sherif Mohamed2, Victor Saase2, Melina Samartzi3, Marc Fatar3, Eva Neumaier-Probst2, and Lothar R. Schad1 | ||
1Computer Assisted Clinical Medicine, Medical Faculty Mannheim, Heidelberg University, Mannheim, Germany, 2Department of Neuroradiology, Medical Faculty Mannheim, Heidelberg University, Mannheim, Germany, 3Department of Neurology, Medical Faculty Mannheim, Heidelberg University, Mannheim, Germany |
||
This study investigates the possibility to reduce 23Na MRI measurement time for stroke patients by applying CNNs and evaluates the resulting TSC accuracy in GM and WM. Three different CNN architectures were implemented and compared. The CNNs’ performance was evaluated by calculating the TSC quantification error and with a qualitative evaluation from neuroradiologists. The implementation of TSC quantification into the clinical routine might be greately facilitated by an acceleration factor of 4 for the 23Na MRI acquisition time while keeping its TSC accuracy in WM and GM. |
| 3810 | A joint-community effort to standardize quantitative MRI data: Updates from the BIDS extension proposal | |
| Agah Karakuzu1,2, Gilles Hollander3,4, Stefan Appelhof5, Tibor Auer6, Mathieu Boudreau1,2, Franklin Feingold7, Ali R. Khan8, Alberto Lazari9, Christophe Phillips10, Nikola Stikov1,2, and Kirstie Whitaker11,12 | ||
1NeuroPoly Lab, Institute of Biomedical Engineering, Polytechnique Montreal, Montreal, QC, Canada, 2Montreal Heart Institute, Montreal, QC, Canada, 3Laboratory for Social and Neural Systems Research (SNS Lab), Department of Economics, University of Zurich, Zürich, Switzerland, 4Spinoza Centre for Neuroimaging, Amsterdam, Netherlands, 5Center for Adaptive Rationality, Max Planck Institute for Human Development, Berlin, Germany, 6University of Surrey, Guildford, United Kingdom, 7Stanford University, Stanford, CA, United States, 8Department of Medical Biophysics, Robarts Research Institute, University of Western Ontario, London, ON, Canada, 9Wellcome Centre for Integrative Neuroimaging, FMRIB, Nuffield Department of Clinical Neurosciences, University of Oxford, Oxford, United Kingdom, 10GIGA Institute, University of Liège, Liège, Belgium, 11Department of Psychiatry, University of Cambridge, Cambridge, United Kingdom, 12The Alan Turing Institute, London, United Kingdom |
||
The Brain Imaging Data Structure (BIDS) aims to develop a standard for organizing and describing neuroimaging data (https://bids.neuroimaging.io/). It has rapidly gained traction across neuroimaging disciplines through community-driven development of BIDS extension proposals (BEPs; http://bit.ly/bids_bep). Here were present such an extension proposal, that captures a wide range of structural MRI contrasts and parametric mapping protocols that are of interest to the broader ISMRM community (https:/bit.ly/bep001). This abstract reports the current state of the proposal and illustrates the developments made since the 2018 ISMRM virtual meeting Bringing BIDS closer to quantitative MRI. |
| 3811 | Magnetic Field Mapping for Transcranial Direct Current Stimulation on human brain: a preliminary study | |
| Fábio Seiji Otsuka1, Carlos Ernesto Garrido Salmon1, Iman Ghodratitoostani2,3, Zahrasadat Vazirikangolya2,4, Renan Maatsuda5, Abhishek Datta6, and Chris Thomas6 | ||
1Inbrain Lab, Department of Physics, Faculty of Phylosophy, Sciences and Letter of Ribeirão Preto (FFCLRP), University of São Paulo (USP), Ribeirão Preto, Brazil, 2Neurocognitive Engineering Laboratory (NEL), Institute of Mathematics and Computer Sciences, University of São Paulo (USP), São Carlos, Brazil, 3Reconfigurable Computing Laboratory, Institute of Mathematics and Computer Science, University of São Paulo (USP), São Carlos, Brazil, 4Department of Neuroscience and Behavioral Sciences, Ribeirão Preto Medical School, University of São Paulo (USP), Ribeirão Preto, Brazil, 5Biomag Lab, Department of Physics, Faculty of Phylosophy, Sciences and Letter of Ribeirão Preto (FFCLRP), University of São Paulo (USP), Ribeirão Preto, Brazil, 6Soterix Medical, New York, NY, United States |
||
Two methods to map the magnetic field distributions generated by tDCS’s electric currents using MRI were evaluated in this work. First method Stimulus/Rest Difference and second GLM. Phase images were preprocessed using motion parameters calculated by SPM from the magnitude images. To evaluate the results, simulated magnetic field distribution was calculated using simulated data for the electric current distribution. The first method showed similarities with simulated data except for one case, while the second method showed magnetic field variations only at occipital region. These results points toward the possibility of mapping these very low intensity magnetic fields using MRI. |
| 3812 | Evaluation of Native T1, Extracellular Volume, and apparent diffusion coefficient in Invasive Ductal Carcinoma | |
| Ryoko Yamamori1, Akihiro Kitanaka1, Masatoshi Sakai1, Shinsuke Oie1, Naoki Ohno2, Yuki Koshino2, and Ayako Katagiri3 | ||
1Department of Radiological technology, Ishikawa Prefectural Central Hospital, Kanazawa, Japan, 2Division of Health Sciences, Institute of Sciences, Kanazawa University, Kanazawa, Japan, 3Department of Radiology, Ishikawa Prefectural Central Hospital, Kanazawa, Japan |
||
We evaluated native T1 and extracellular volume (ECV) of the breast using modified Look-Locker inversion recovery (MOLLI) sequence and compared those between invasive ductal carcinoma (IDC) and fibroglandular tissue. Moreover, we investigated the relationship between ECV and apparent diffusion coefficient (ADC) in IDC. Our results showed that ECV of IDC was significantly higher than fibroglandular tissue, whereas there was no significant difference in native T1 between both tissues. In addition, there was no significant correlation between ECV and ADC in IDC. ECV measurement using MOLLI may provide new and more detailed information in breast cancer. |
| 3813 | Optimized on-the-fly reconstruction for combined T2*-weighted imaging of the human brain and cervical spinal cord | |
| Ying Chu1 and Jürgen Finsterbusch1 | ||
1Department of Systems Neuroscience, University Medical Center Hamburg-Eppendorf, Hamburg, Germany |
||
Cortico-spinal functional magnetic resonance imaging (fMRI) covering slices in the brain and cervical spinal cord usually involves specific geometry (field-of-view, voxel size) and timing parameters (bandwidth) for the two volumes. This requires a time-consuming and cumbersome retrospective reconstruction after the experiment because the standard program cannot handle the different parameter settings properly, e.g., for regridding of ramp-sampled data or correction of distortions induced by Maxwell terms. Here, a reconstruction program is presented that performs an optimized reconstruction on-the-fly including the correction of receive-coil inhomogeneities without the need of a retrospective reconstruction. It could help the applicability of cortico-spinal fMRI. |
| 3814 | Blindly trusting MRI radiomics? A study on radiomic features repeatability and reproducibility with a dedicated phantom | |
| Linda Bianchini1, João Santinha2, Nuno Loução3, Mário Figueiredo4, Francesca Botta5, Daniela Origgi5, Marta Cremonesi5, Nickolas Papanikolaou2, and Alessandro Lascialfari1 | ||
1Università degli Studi di Milano, Milan, Italy, 2Champalimaud Center for the Unknown, Lisbon, Portugal, 3Philips Healthcare, Lisbon, Portugal, 4Instituto de Telecomunicações, Instituto Superior Técnico, Lisbon, Portugal, 5Istituto Europeo di Oncologia (IEO) IRCCS, Milan, Italy |
||
The radiomic features stability when MR scanners of different vendors or magnetic fields are involved is not known and is needed to support clinical studies. A study on the radiomic features repeatability and reproducibility was carried out with a pelvis phantom designed for radiomic purposes on three MR scanners, two of same field but different vendors and two of same vendor but different fields. The crucial results include a consistent percentage loss of features repeatability after phantom repositioning, suggesting the need for a features selection in studies involving patient’s repositioning. The limited reproducibility demands attention when dealing with multicentric studies. |