MRI techniques for quantitative and microstructure imaging

The ability to map brain microstructures such as axonal fibers and myelin content is critical in improving our understanding of brain organization and neurological diseases. Quantitative MRI has been proved with its high sensitivity to microstructures, providing a safe, non-invasive approach for human in vivo imaging. However, quantitative MRI typically requires the acquisition of multiple images for biophysical model fitting, which leads to long scan time that consequently causes low SNR, low spatial resolution, image artifacts and vulnerability to motion. This thesis aims at overcoming these challenges by developing novel MRI acquisition and reconstruction methods that exploit the strength of spatiotemporal encoding, recent hardware innovations and low-rank signal priors, to provide efficient microstructure imaging of the human brain with higher speed, SNR, resolution, and motion robustness. Three quantitative MRI contrasts were studied in this thesis, including diffusion MRI, myelin water imaging, and MR relaxometry. These contrasts provide sensitivity to axonal fibers, myelin concentration, and iron contents for the study of brain microstructures. Specifically, in the first part of this thesis, a fast and high-fidelity diffusion imaging method was developed that achieves 30-40% higher SNR-efficiency than the current state-of-the-art method. This technique was also shown to be robust to physiological motion and field variations, as well as the capability to resolve multi-echo images that are free from image distortions and artifacts. The second part of this thesis presents a novel acquisition method for myelin water imaging with >10× acceleration compared to current approaches, that can potentially be used for a fast examination of demyelination diseases. The work also demonstrates the first submillimeter myelin water imaging in vivo at 600-um isotropic resolution to study cortical myeloarchitecture. The third part of this thesis shows an ultra-fast MR relaxometry method with navigated motion correction, which provides fast, repeatable, and motion-robust quantitative imaging of the human brain.Ph.D