Quantitative analysis of the MRI structural and functional images is a fundamental component in the assessment of brain anatomical abnormalities, in mapping functional activation onto human anatomy, in longitudinal evaluation of disease progression, and in computer-assisted neurosurgery or surgical planning. Image registration and segmentation is central in analyzing structural and functional MR brain images.
However, due to increased variability in brain morphology and age-related atrophy, traditional methods for image registration and segmentation are not suitable for analyzing MR brain images from elderly individuals. The overall goal of this dissertation is to develop algorithms to improve the registration and segmentation accuracy in the geriatric population. The specific aims of this work includes (1) to implement a fully deformable registration pipeline to allow a higher degree of spatial deformation and produce more accurate deformation field, (2) to propose and validate an optimum template selection method for atlas-based segmentation, (3) to propose and validate a multi-template strategy for image normalization, which characterizes brain structural variations in the elderly, (4) to develop an automated segmentation and localization method to access white matter integrity (WMH) in the elderly population, and finally (5) to study the default-mode network (DMN) connectivity and white matter hyperintensity in late-life depression (LLD) with the developed registration and segmentation methods.
Through a series of experiments, we have shown that the deformable registration pipeline and the template selection strategies lead to improved accuracy in the brain MR image registration and segmentation, and the automated WMH segmentation and localization method provides more specific and more accurate information about volume and spatial distribution of WMH than traditional visual grading methods. Using the developed methods, our clinical study provides evidence for altered DMN connectivity in LLD. The correlation between WMH volume and DMN connectivity emphasizes the role of vascular changes in LLD’s etiopathogenesis.
Keywords. image registration, image segmentation, template selection, resting state connectivity, late-life depression.
|Advisor:||Aizenstein, Howard J.|
|School:||University of Pittsburgh|
|School Location:||United States -- Pennsylvania|
|Source:||DAI-B 71/03, Dissertation Abstracts International|
|Subjects:||Biomedical engineering, Cognitive psychology, Computer science|
|Keywords:||Brain imaging, Depression, Image registration, Image segmentation|
Copyright in each Dissertation and Thesis is retained by the author. All Rights Reserved
The supplemental file or files you are about to download were provided to ProQuest by the author as part of a
dissertation or thesis. The supplemental files are provided "AS IS" without warranty. ProQuest is not responsible for the
content, format or impact on the supplemental file(s) on our system. in some cases, the file type may be unknown or
may be a .exe file. We recommend caution as you open such files.
Copyright of the original materials contained in the supplemental file is retained by the author and your access to the
supplemental files is subject to the ProQuest Terms and Conditions of use.
Depending on the size of the file(s) you are downloading, the system may take some time to download them. Please be