Files in this item



application/pdfCLIFFORD-DISSERTATION-2019.pdf (10MB)Restricted Access
(no description provided)PDF


Title:Further development of magnetic resonance spectroscopic imaging using subspace-based models
Author(s):Clifford, Bryan
Director of Research:Liang, Zhi-Pei
Doctoral Committee Chair(s):Liang, Zhi-Pei
Doctoral Committee Member(s):Do, Minh; Sutton, Brad; Schwing, Alexander
Department / Program:Electrical & Computer Eng
Discipline:Electrical & Computer Engr
Degree Granting Institution:University of Illinois at Urbana-Champaign
Subject(s):Magnetic resonance spectroscopic imaging
subspace model
low rank
metabolic mapping
mitochondrial oxida- tive capacity
magnetic susceptibility mapping
motion correction
Abstract:Magnetic resonance spectroscopic imaging (MRSI) is a powerful, non-invasive imaging modality capable of providing in vivo measurements of chemical distributions within an object. This makes it an ideal tool for metabolic studies in clinical and research settings; yet, despite more than four decades of development, MRSI has yet to achieve clinical utility due to an inherently poor trade-off between imaging speed, resolution, and signal-to-noise-ratio (SNR). The recent integration of subspace-based models into MRSI has significantly improved this trade-off and demonstrated the potential for high-resolution, high-SNR metabolic mapping at clinically relevant imaging speeds. This work further extends subspace-based MRSI methods along two major avenues. First, we have developed a novel method for simultaneously providing tissue magnetic susceptibility information and metabolic information. By using a specialized data acquisition and processing scheme, we are able to acquire the 1H-MRSI signals from water as well as the metabolic signals of interest and use the water image for quantitative susceptibility mapping (QSM). The proposed method has been evaluated in in vivo experiments and shown capable of providing state-of-the-art MRSI and QSM of the brain at millimeter-scale resolutions from a 7 min scan. We have also developed two novel methods to improve the robustness of this method by correcting for the effects of intrascan motion. Secondly, we have developed a molecular-subspace based method for performing dynamic 31P-MRSI. Existing subspace-based MRSI methods require several minutes to generate a single image. As such, they are not suitable for studying rapid metabolic processes; however, the proposed method provides dynamic MRSI with frame-rates on the order of 1 s while maintaining sub-centimeter spatial resolutions. We have validated the method's capability to provide reproducible dynamic metabolic maps in both preclinical and clinical settings using in vivo and in vitro experiments. The method's potential for providing reliable phosphocreatine resynthesis time-constant mapping in small animal models was also evaluated using in vivo experiments and simulations, which indicated that estimates with acceptable bias and variance could be obtained over the majority of the physiological range.
Issue Date:2019-06-03
Rights Information:Copyright 2019 Bryan Clifford
Date Available in IDEALS:2019-11-26
Date Deposited:2019-08

This item appears in the following Collection(s)

Item Statistics