Home
Resource
Sparsity-based methods for cardiac magnetic resonance image reconstruction and analysis
PDF
PDF format is widely accepted and good for printing.
Plug-in required
PDF-1(6.62 MB)
Citation & Export
View Usage Statistics
Staff View

Citation & Export
Hide

Simple citation

Yu, Yang. Sparsity-based methods for cardiac magnetic resonance image reconstruction and analysis. Retrieved from https://doi.org/doi:10.7282/T3W66NSP

Export

Click here for information about Citation Management Tools at Rutgers.

Statistics
Hide

Description

TitleSparsity-based methods for cardiac magnetic resonance image reconstruction and analysis
NameYu, Yang (author); Metaxas, Dimitris (chair); Elgammal, Ahmed (internal member); Bekris, Kostas (internal member); Huang, Xiaolei (outside member); Rutgers University; Graduate School - New Brunswick
Date Created2015
Other Date2015-10 (degree)
SubjectComputer Science, Magnetic resonance imaging, Heart--Imaging
Extent1 online resource (xiv, 103 p. : ill.)
DescriptionIn signal processing, sparseness means that there are only small amounts of non-zero elements. This property has been widely observed in various types of signals. However, the data sparseness is hard to be regularized due to its non-convex nature. The recent development of the compressed sensing technique builds a theoretical connection between the sparse constraint and its convex relaxation. This discovery motivates us to explore different types of sparse properties for the generation and analysis of the cardiac magnetic resonance images (MRIs). In this work, our proposed a series of sparse optimization algorithms have been applied to cardiac image reconstruction, segmentation and motion tracking problems for fast and robust analyzing the cardiac data. The cardiac imaging is a challenging problem to MRI due to its fast motion. We proposed a novel calibration-less algorithm to accelerate the generation of dynamic MR images with both compressed sensing and parallel imaging. In addition to the temporal signal, which usually provides more data redundancy than spatial signals, the strong correlations among signals from different coils are utilized to form joint sparse constraints. A general optimization framework is presented to solve the problem under different types of temporal sparse constraints efficiently. We then apply the sparse constraint to the cardiac muscle motion tracking. The 3D deformable heart model is built by simulating its motion in a cardiac cycle based on tagged MRI. The tagged MR data is widely used to reveal the internal myocardial motion. However, the automated tagging line detection results are very noisy due to the poor image quality. To alleviate this issue, we introduce a new family of sparse deformable models based on the sparseness of the detection noise. Our new models track the heart motion robustly, and the resulting strains are consistent with those calculated from manual labels.
NotePh.D.
NoteIncludes bibliographical references
Noteby Yang Yu
Genretheses, ETD doctoral
Persistent URLhttps://doi.org/doi:10.7282/T3W66NSP
Languageeng
CollectionGraduate School - New Brunswick Electronic Theses and Dissertations
Organization NameRutgers, The State University of New Jersey
RightsThe author owns the copyright to this work.
Version 8.5.5
Rutgers University Libraries - Copyright ©2024