Cortical Statistical Correlation Tomography of EEG Resting State Networks

Area of Science:

• Neuroscience
• Computational Neuroscience
• Biomarkers

Background:

• Resting-state networks (RSNs) are crucial for brain function, characterized by correlated activity in distributed brain regions.
• Current methods often rely on functional magnetic resonance imaging (fMRI), limiting real-time or portable applications.

Purpose of the Study:

• To introduce a novel computational framework for reconstructing electrophysiological RSNs from human EEG data.
• To validate the framework's ability to identify consistent RSNs and detect condition-specific differences.

Main Methods:

• Utilized independent component analysis (ICA) on short-time Fourier transformed inverse source maps from EEG data.
• Applied statistical correlation analysis to generate cortical tomography of RSNs.
• Validated the framework on three EEG datasets comparing different conditions (eyes open/closed, healthy vs. balance disorder, pre/post rTMS).

Main Results:

• Successfully reconstructed five consistent RSNs across individuals and conditions with similar spatial and spectral patterns.
• EEG-derived RSN tomographic maps showed high similarity to fMRI-based RSN templates.
• Observed significant spatial and spectral differences in RSNs across compared conditions, aligning with existing literature.
• Demonstrated potential for identifying biomarkers in patient data for diagnosis and treatment evaluation.

Conclusions:

• The novel framework enables robust reconstruction of RSNs from EEG data, offering spatial and spectral characterization.
• This approach provides a new dimension for understanding RSNs and their neural mechanisms.
• The method shows promise as a tool for diagnosing and evaluating treatment efficacy in neuropsychiatric disorders.