Quantitative measures of topographic and divergent/convergent connectivity in diffusion MRI of the human cerebral cortex

Abstract

Spatial features of connections, such as topography and divergence/convergence, reflect the information-processing mechanisms crucial for understanding and modeling the brain. However, this has not yet been comprehensively investigated. Using diffusion Magnetic Resonance Imaging (dMRI) data, we developed a topographic factor (TF) and divergence/convergence factor (DC) to quantitatively explore the spatial connectivity patterns of one entire hemisphere of the human cerebral cortex. In the analysis, the early sensory areas, which are located far away from the center of the cerebral cortex, exhibited high topographic connectivity. In contrast, the limbic system, which is located proximal to the center, showed high divergence/convergence in two types of connectivity: one region to another region at the region-to-region level, and one region to all other regions at the global level. Topography had anti-correlation with divergence/convergence over the cerebral cortex, and the two types of divergence/convergence displayed positive correlation with each other. These results suggest that topographic and divergent/convergent connectivity are spatially organized with respect to cytoarchitecture over the cerebral cortex to optimize energy efficiency and information transfer performance in the cerebral cortex.