Analysis of structural and functional brain networks
The brain is a representative example of a network. It consists of numerous spatially distributed regions that continuously exchange information through structural connections. In the past decade, an increasing number of studies have explored the brain network in both structural and functional aspects; they have begun to decipher complex brain wirings, as well as elucidate how the rich functionality emerges from this architecture. Based upon previous studies, this thesis addresses three critical gaps in the field. (I) Although it is known that the community structures of brain network are spatially overlapping, conventional studies have focused on grouping brain regions into communities such that each region belongs to only one community. Therefore, a recent “link community” concept was employed to disentangle those overlapping architectures. (II) Spatial independent component analysis (sICA) and k-means clustering are two representative data-driven algorithms used to analyze functional networks. However, it is still unclear how these two methods compare to each other in terms of their theoretical basis and biological relevance. Hence, the relationship between these two methods were investigated. (III) Despite the multi-scale functional organization of the brain, previous studies have primarily examined the large-scale networks of the entire brain. Complex neural activity patterns in relatively smaller spatial scales have been poorly understood. Therefore, the fine-scale spatiotemporal patterns within visual cortex were explored. The distinguishing results obtained in this study may provide new insights regarding the brain's organization, as well as a better understanding of mathematical and statistical tools for functional and structural network analysis.^
Zhongming Liu, Purdue University.
Off-Campus Purdue Users:
To access this dissertation, please log in to our