Cliques and cavities in the human connectome

Area of Science:

• Neuroscience
• Network Science
• Computational Topology

Background:

• Human cognitive processes rely on complex, distributed brain networks.
• Understanding information flow in these networks requires analyzing higher-order interactions beyond pairwise connections.

Purpose of the Study:

• To investigate mesoscale network structures in the human brain using algebraic topology.
• To identify densely connected substructures (cliques) and topological cavities within brain networks.

Main Methods:

• Applied algebraic topology to analyze structural connectomes of 8 healthy adults.
• Detected cliques (fully connected brain region sets) and topological cavities.
• Compared findings to null networks generated via wiring minimization.

Main Results:

• Discovered more large cliques than expected in null models, suggesting rapid local processing capabilities.
• Identified consistent topological cavities across subjects, differing from null models.
• Found these cavities link evolutionarily distinct brain regions in long loops.

Conclusions:

• Algebraic topology provides novel insights into structural connectomics.
• Loop-like structures (cavities) are crucial features in the human brain's architecture.
• These structures may play a key role in controlling complex brain functions.