Electrical and Computer Engineering Ph.D. Public Defense
Integration of Structural and Functional Connectivity of Brain, and Clinical Applications
Supervised by Professor Jianhui Zhong and Professor Giovanni Schifitto
Monday, August 17, 2020
Join Zoom Meeting https://rochester.zoom.us/j/94462610910
MRI-based neuroimaging techniques have been shown to be valuable in the investigation of HIV-infection associated neuropathology. HIV-infected subjects have reduced cortical thickness and subcortical brain volumes, measured by T1-weight anatomical images, and reduced functional connectivity (FC) measured by resting-state functional MRI (rsfMRI). Brain injury at a microstructural level can be quantified by diffusion tensor imaging (DTI) metrics, such as decreased fractional anisotropy (FA) and increased mean diffusivity (MD). In HIV-infected subjects both FA and MD abnormalities have been previously reported. However, most of these studies investigated one modality at the time, while the integration of multi-modalities in HIV- related studies, and the relationship between the neuronal activity and the underlying white matter pathway has not been fully investigated.
To overcome this gap in this study, we firstly used graph theoretical analysis to investigate the topological changes in structural connectivity (SC) and functional connectivity (FC) respectively. Then we adopted the relaxed mean field dynamic modeling (rMFM) method to integrate SC and FC. We generated whole-brain dynamic models for the HIV-baseline, HIV- 12 week treatment, and healthy control group respectively. After optimizing the model parameters to maximize the similarity between simulated FC derived from SC and empirical FC which is directly calculated from rsfMRI, we resulted in three whole-brain dynamic models which provide good representations of these three groups. Then the microscale brain properties can be interpreted by comparing the model parameters between three models. Lastly, the full spectrum of neuropsychological assessment and the cognitive performance for each subject were also compared between cohorts, and their relations with imaging derived metrics were also investigated.