Advances in Myelin Water Imaging and Stack-of-Spirals MRI: Image Reconstruction and Parameter Estimation

Abstract

Myelin content plays a vital role in healthy functioning of the brain. Myelin water imaging (MWI) is a quantitative MRI technique that aims to image the myelin content in the brain. Characterizing and quantifying the exchange dynamics in myelin (between myelin water and non-myelin water) could lead to better understanding of the role of myelin loss in several neurological disorders. This dissertation introduces a method to estimate myelin water exchange in white matter, using MRI experiment design. We optimized the acquisition parameters of a set of phase-cycled bSSFP acquisitions using a cost function based on the Cramer-Rao bound, and obtained low coefficients of variation for estimating myelin water exchange in white matter (around 13.5%).

Validating the obtained myelin water exchange maps in vivo is a challenging problem. This dissertation also explores the design and use of an aqueous urea system to validate our proposed bSSFP acquisitions for estimating exchange. To potentially translate our proposed acquisitions to clinically feasible settings, undersampling is an important consideration to bring down acquisition time. This dissertation studies an undersampling scheme based on stack-of-spirals MRI. We demonstrate that the 3D NUFFT-based forward model for stack-of-spirals MRI can be simplified and sped up by exploiting the Cartesian structure in the through-plane direction, even in the presence of off-resonance effects. We show that our recommended efficient implementation is at least 22% faster than other existing 3D NUFFT-based implementations of the stack-of-spirals forward model, when run on a 20-core Intel Xeon processor.