Linking individual behavioral architecture to brain topological structure: A combined network analysis approach

Abstract

In the connectome era, understanding how the brain supports various cognitive and behavioral functions requires moving beyond isolated domains to a multimodal network-level approach. This opens a new window into modeling how the interactions between neural and behavioral facets characterize the uniqueness of the individual. This study combines two different models of graph analysis, graph theory and exploratory graph analysis, to examine how large-scale brain network topology of 379 healthy adults from the Human Connectome Project relates to the individual mental architecture. We modeled nonlinear brain-behavior relationships using Generalized Additive Models. Specifically, we investigated the extent to which the topological properties of the canonical seven resting-state networks (RSNs) contribute in explaining the individual performance across seven cognitive and behavioral domains: Mental Health, Externalizing Problems, Higher level Cognitive Functions, Core Cognitive Functions, Substance use/abuse, Delay Discounting Task, and Pain. Our results reveal domain-specific signatures of integration and segregation among RSNs and cognitive/behavioral performance. For instance, adaptive functions such as higher level and core cognitive functions were associated with greater integration of the fronto-parietal network and segregation of the default mode network. On the other hand, maladaptive domains such as substance abuse and impulsivity were linked to increased integration in sensorimotor and limbic networks. Single-node analysis further identified key hubs whose topological features aligned with distinct cognitive and behavioral profiles. Our findings highlight the advantage of using a network neuroscience approach to investigate the complex and dynamic nature of human cognition and behavior.