Introduction

MRI is a noninvasive imaging modality that plays an increasingly significant role in the understanding of body and the management of diseases. It provides a large variety of tissue contrast with multitude of contrast generating mechanisms, including longitudinal relaxation time (T1), transverse relaxation time (T2), transverse magnetization with field inhomogeneity (T2*), fat concentration, etc. These contrasts reflect not only the tissue’s macrostructure, but also their microstructure, metabolism, and function, providing essential information for vital processes and disease developments. However, conventional MRI usually acquires a certain contrast-weighted image, which is qualitative and subject to coil positioning, and inter-scanner and inter-reader variability. In recent years, with the development of accelerated techniques, quantitative MRI becomes possible and available. There are numerous quantitative MRI techniques developed for multiple organs, playing an increasingly important role in clinical applications. However, for moving organs such as heart, lung, and abdominal organs, the appearance of physiological motions introduce severe motion artifacts which can significantly degrade the image quality. ECG gating has been used to reduce cardiac-motion-induced artifacts, which can significantly lengthen the scan time, especially for quantitative MRI which requires the acquisition at multiple contrast weightings. For respiratory motion, respiratory gating can be an option but faces the same problem as lengthened scan. The most popular way to deal with respiratory motion is a 10 to 20-second breath-hold from the subject. But this limits the resolution of the quantitative maps due to the short scan time and causes extra burden on the subjects. For actual clinical situation, patients may fail to hold the breath, causing increased artifacts and failure in diagnosis. To address the limitations for quantitative mapping on moving organs, several techniques to achieve motion-resolved parametric mapping have been proposed, which will be introduced in this talk.

MR Multitasking

MR Multitasking framework captures the physiological and bulk motion by frequently acquiring the center k-space area, which forms the training data. The contrast dynamics and motion information are extracted from the training data. The images containing the combination of all the contrasts and all motion states are reconstructed by exploiting the high image correlation along and across the multiple dynamic dimensions using low-rank tensor structure¹.

XD-GRASP

Extra-dimensional golden-angle radial sparse parallel (XD-GRASP) technique has been used for dynamic imaging in cardiac and abdominal organs². The center k-space in each spoke is used to extract the physiological motion over time. The data are sorted according to different motion states. In the reconstruction, a different sparsity constraint is enforced along each dynamic dimension to exploit the image correlation and recover the images of different motion states.

Motion-corrected MRF

MR Fingerprinting (MRF) has been an efficient parametric mapping technique and widely applied in the non-moving organs. In recent years, motion-corrected MRF has been proposed to achieve the parametric mapping for moving organs. In a recent work³, the ECG and respiratory bellow signals are used to bin the data into different motion states. An initial reconstruction for each motion state is performed and a non-rigid deformative motion operator is estimated based on the initial recon. Finally the motion operator is incorporated into the final recon using a low-rank framework.

Acknowledgements

No acknowledgement found.

References

1. Christodoulou, Anthony G., et al. "Magnetic resonance multitasking for motion-resolved quantitative cardiovascular imaging." Nature biomedical engineering 2.4 (2018): 215-226.2. Feng, Li, et al. "XD‐GRASP: golden‐angle radial MRI with reconstruction of extra motion‐state dimensions using compressed sensing." Magnetic resonance in medicine 75.2 (2016): 775-788.3. Cruz, Gastao, et al. "Generalized low‐rank nonrigid motion‐corrected reconstruction for MR fingerprinting." Magnetic Resonance in Medicine 87.2 (2022): 746-763.