Structural determinants of dynamic fluctuations between segregation and integration on the human connectome

Abstract

While segregation and integration of neural information in the neocortex are thought to be important for human behavior and cognition, the neural substrates enabling their dynamic fluctuations remain elusive. To tackle this problem, we aim to identify specific network features of the connectome (the complete set of structural brain connections) that are responsible for the emergence of dynamic fluctuations between segregated and integrated patterns in human resting-state fMRI functional connectivity. The contributions of network features to the dynamic fluctuations were examined by constructing randomly rewired surrogate connectome data in which network features of interest were selectively preserved, and then by assessing the magnitude of fluctuations simulated with these surrogates. Our analysis demonstrates significant contributions from spatial geometry and network topology of the connectome, as well as from localized structural connections involving visual areas. By providing a structural account of dynamic fluctuations in functional connectivity, this study offers new insights into generative mechanisms driving temporal changes in segregation and integration in the brain.