Community detection for directional neural networks inferred from EEG data

Area of Science:

• Neuroscience
• Computational Neuroscience
• Network Science

Background:

• Identifying functional brain modules from complex neural recordings is a significant challenge.
• Existing community detection methods often rely on functional connectivity, neglecting causal (effective) connectivity.

Purpose of the Study:

• To develop a novel community detection algorithm for directed networks representing effective connectivity.
• To apply this algorithm to multichannel electroencephalogram (EEG) data.
• To extend the algorithm for identifying common community structures across multiple subjects.

Main Methods:

• Developed a community detection algorithm tailored for weighted and asymmetric networks.
• Applied the algorithm to multichannel EEG recordings.
• Extended the algorithm to perform multi-subject analysis for common structure detection.

Main Results:

• The proposed algorithm effectively identifies community structures in directed networks.
• The method successfully applied to EEG data, revealing network causality.
• A common community structure was identified across multiple subjects.

Conclusions:

• The developed algorithm provides a robust approach for analyzing effective connectivity in neuroscience.
• This method advances the understanding of brain network organization and causality.
• The multi-subject extension enables the study of consistent brain network patterns.