A consistent organizational structure across multiple functional subnetworks of the human brain

Area of Science:

• Cognitive Neuroscience
• Network Neuroscience
• Computational Neuroscience

Background:

• The brain exhibits consistent subnetwork structures linked to specialized cognitive functions.
• Understanding inter-regional communication principles within subnetworks is key to cognitive process research.
• Current tools lack precision in quantifying organizational principles within brain subnetworks.

Purpose of the Study:

• To apply a novel joint modeling technique, the correlation generalized exponential random graph model (cGERGM), to precisely quantify subnetwork structure.
• To investigate the organizing principles governing communication within and across functional brain subnetworks.
• To determine if common organizational principles exist across different functional subnetworks.

Main Methods:

• Utilized the correlation generalized exponential random graph model (cGERGM) to analyze functional connectivity data.
• Modeled correlation networks based on activation motifs (coactivation patterns) and anatomical properties (e.g., Euclidean distance).
• Jointly modeled all features to assess unique variance, estimate parameters, and perform significance testing.

Main Results:

• Identified remarkably consistent organizational properties across eight functional subnetworks.
• Observed greater clustering than expected by chance in all subnetworks.
• Found lower preferential attachment (hub use) across all analyzed subnetworks.

Conclusions:

• Human functional subnetworks exhibit a fundamental, consistent organizational basis for communication.
• Subnetworks appear to follow a 'segregated highway' structure.
• Tightly clustered subcommunities develop dedicated communication channels, minimizing reliance on central hubs.