Denoising for deuterium magnetic resonance spectroscopic imaging based on posterior-score-guided subspace modeling
Xu, Ziyang
This item is only available for download by members of the University of Illinois community. Students, faculty, and staff at the U of I may log in with your NetID and password to view the item. If you are trying to access an Illinois-restricted dissertation or thesis, you can request a copy through your library's Inter-Library Loan office or purchase a copy directly from ProQuest.
Permalink
https://hdl.handle.net/2142/122083
Description
Title
Denoising for deuterium magnetic resonance spectroscopic imaging based on posterior-score-guided subspace modeling
Author(s)
Xu, Ziyang
Issue Date
2023-12-08
Director of Research (if dissertation) or Advisor (if thesis)
Liang, Zhi-Pei
Department of Study
Electrical & Computer Eng
Discipline
Electrical & Computer Engr
Degree Granting Institution
University of Illinois at Urbana-Champaign
Degree Name
M.S.
Degree Level
Thesis
Keyword(s)
deuterium magnetic spectroscopic imaging
denoising
score-based diffusion model
posterior sampling
machine learning
Abstract
Deuterium magnetic resonance spectroscopic imaging (DMRSI) has recently been recognized as a powerful tool for imaging the energy metabolism and tumors of the brain. However, the low sensitivity of DMRSI has primarily limited its practical utility in research and clinical studies. In this work, we present a novel approach to improving the sensitivity of DMRSI, incorporating both physics-based subspace modeling and data-driven distribution learning to remove measurement noise. Particularly, a union-of-subspaces model is used to represent the spatial-spectral-temporal variations of the desired DMRSI signal, leading to a significant reduction in degrees of freedom. The subspace structures are pre-learned from spin physics and training data, incorporating known resonance structures and a priori experimental variations. The subspace coefficients are treated as random variables, whose joint probabilistic distributions are learned using an advanced diffusion model. With the proposed subspace model and learned signal priors, denoising is accomplished using the Langevin dynamics process. The proposed method has been evaluated using both simulated and experimental data, producing very promising results. The resulting algorithm is expected to enhance the practical utility of DMRSI and could also be useful for denoising MRSI of other nuclei.
In this thesis, background materials on subspace modeling of MRSI signals, Bayesian parameter estimation theory and diffusion generative models will be first presented. Then, a detailed description of the proposed denoising algorithm is provided. Finally, simulation and in vivo evaluation results are presented to demonstrate the efficacy of the proposed method for dynamic DMRSI denoising.
Use this login method if you
don't
have an
@illinois.edu
email address.
(Oops, I do have one)
IDEALS migrated to a new platform on June 23, 2022. If you created
your account prior to this date, you will have to reset your password
using the forgot-password link below.
