The human brain is made up of functionally and structurally connected neural elements that form a brain-wide complex network. A principal goal of network neuroscience is to understand how the organization of this network helps support cognition, evolves over the course of the human lifespan, and becomes compromised in disease and neuropsychiatric disorders. However, virtually all progress made towards addressing these questions has relied upon one particular network model for mathematically representing patterns of brain connectivity, at the expense of other models that could provide complementary or unique insight. This project aims to extend and validate an alternative edge-centric framework for representing and analyzing patterns of brain connectivity. The project will deliver new insights into the relationship of brain network organization with cognitive/behavioral phenotypes and shed light on brain network dynamics at ultra-fast timescales are paralleled by changes in subjects' cognitive states. This research will support cross-disciplinary collaboration among the brain sciences, informatics, and statistics, and will support a diverse set of trainees at all levels, from high school to postdoctoral.
This principal innovation of the edge-centric framework is a spatiotemporal decomposition of functional connections into their framewise contributions. This decomposition yields a time series of co-fluctuations for every pair of brain regions (edges in the network). The first aim investigates the novel construct of edge functional connectivity -- the correlation pattern estimated among all pairs of co-fluctuation time series. Edge connectivity will be generated for a large cohort of subjects (N > 1000) using imaging data acquired both at rest and while subjects were performing cognitively demanding tasks. Multivariate statistical methods will be used to discover robust associations between edge connectivity and subjects' behavioral, demographic, and clinical measures. The second aim analyzes co-fluctuation time series directly, taking advantage of the ultra-fast timescale at which they are estimated to investigate potential drivers of brain network reconfiguration during naturalistic viewing (movie-watching). This project advances the edge-centric framework as a viable tool for general neuroscientific discovery and will open the door for future studies to investigate brain-behavior relationships and network dynamics in applied contexts and not restricted to large-scale imaging data.
This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.
