Temporal-Spatial Patterns in Dynamic Functional Brain Network for Self-Paced Hand Movement.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Brain-Computer Interfaces

Background:

• Dynamic functional connectivity offers a more accurate view of brain networks than static methods.
• Brain activity fluctuations are thought to arise from shifts between distinct brain states.

Purpose of the Study:

• Investigate spatio-temporal brain interactions during self-paced hand movements using EEG.
• Develop a framework to analyze dynamic functional connectivity and identify brain states.

Main Methods:

• Calculated functional connectivity using corrected imaginary coherency to mitigate EEG volume conduction.
• Segmented brain states based on singular value decomposition (SVD) vector space distance time series.
• Extracted representative graphs for each state and applied community detection for spatial patterns.

Main Results:

• Identified distinct brain states lasting hundreds of milliseconds using SVD vector space distance and change point detection.
• Observed sudden decreases in SVD distance coinciding with early Bereitschafts potential.
• Characterized evolving functional connectivity patterns before movement, from dispersed to concentrated networks.

Conclusions:

• The study successfully characterized dynamic functional connectivity and brain state transitions during motor tasks.
• Findings highlight the predictive value of dynamic connectivity changes for movement onset.
• The proposed framework provides a robust method for analyzing complex brain dynamics in EEG data.