Accelerated J-resolved proton magnetic resonance spectroscopic imaging

1H-MRSI has long been recognized as a powerful tool for noninvasive mapping of metabolites and neurotransmitters. A fundamental challenge of conventional 1H-MRSI is that resonances from different molecules overlap with one another in the crowded chemical shift frequency spectrum, which will lead to great difficulties in separating different resonances and accurate detection of specific molecules. J-resolved 1H-MRSI has the potential to overcome this limitation by encoding J-evolution information of different molecules thus adding another spectral dimension. The additional dimension, however, requires multiple repetitions of conventional 1H-MRSI scans with different acquisition parameters, leading to a prohibitively long scan time. In this work, we present a novel approach to accelerating J-resolved 1HMRSI by integrating semi-LASER based signal excitation, sparse data acquisition and subspace based data processing. More specifically, we first recognize that the high-dimensional J-resolved 1H-MRSI signals reside in a lowdimensional subspace, thereby significantly reducing the degrees-of-freedom of the desired image function. Based on the subspace model, a semi-LASER J-resolved 1H-MRSI pulse sequence is developed to efficiently sample sparse and limited data, whose parameters are optimally selected according to theoretic analysis. Finally, a joint subspace model is utilized to reconstruct high-quality spatiospectral distributions of metabolites and neurotransmitters, incorporating both spatial and spectral prior information through a weighted regularization and pre-trained spectral basis. Phantom and in vivo experiments were performed to demonstrate the feasibility of the proposed method, generating impressive results. The proposed method achieved unprecedented capability of high-resolution mapping of metabolites and neurotransmitters within a clinically feasible scan time. The proposed method is expected to make high-resolution J-resolved 1H-MRSI experiments more practically useful.

Name thesis:degree_name

M.S.

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

Electrical & Computer Engr

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2020

Author dc:creator

• Zhao, Yibo

Contributors dc:contributor

• Liang, Zhi-Pei

• Magnetic resonance spectroscopic imaging
• J-resolved magnetic resonance spectroscopic imaging
• neurometabolic imaging
• partial separability
• union-of-subspaces model

Statement dc:rights

• Copyright 2019 Yibo Zhao

Language dc:language

en

Handle dc:identifier

http://hdl.handle.net/2142/106452

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/106452