Investigating brain community structure abnormalities in bipolar disorder using path length associated community...

Area of Science:

• Neuroscience
• Network Science
• Computational Psychiatry

Background:

• Understanding brain network community structure is crucial for neuroscience.
• Existing metrics like Q modularity may not fully capture complex community dynamics.
• Node-level community analysis offers deeper insights into brain organization.

Purpose of the Study:

• To present Path Length Associated Community Estimation (PLACE), a novel framework for node-level community structure analysis.
• To introduce a new metric, Ψ(PL), based on intercommunity versus intracommunity path lengths.
• To compare PLACE with traditional metrics and investigate brain network differences in bipolar disorder.

Main Methods:

• Developed PLACE framework utilizing top-down hierarchical binary trees for community extraction.
• Incorporated a novel metric Ψ(PL) measuring path length differences.
• Applied PLACE to structural brain networks of 25 bipolar I subjects and 25 healthy controls.

Main Results:

• Identified significant community structural differences in posterior default mode network regions.
• Observed left-right decoupling in the default mode network of the bipolar I group compared to controls.
• Demonstrated potential theoretical advantages of the Ψ(PL) metric over Q modularity.

Conclusions:

• PLACE provides a comprehensive framework for detailed node-level community analysis in brain networks.
• The Ψ(PL) metric offers a promising alternative for community detection with potential theoretical benefits.
• Findings suggest altered brain network organization in bipolar disorder, specifically within the default mode network.