Features extraction from time-varying cortical networks adopting a theoretical graph approach

Area of Science:

• Neuroscience
• Computational Neuroscience
• Signal Processing

Background:

• Understanding brain functional networks is crucial for neuroscience.
• Existing methods may not fully capture the dynamic, time-varying nature of brain connectivity.
• Time-frequency analysis offers a powerful lens for examining neural oscillations and their interactions.

Purpose of the Study:

• To propose a novel computational framework for analyzing functional brain network structure and organization.
• To capture time-varying functional connectivity patterns in the time-frequency domain.
• To demonstrate the utility of this approach using high-resolution EEG data.

Main Methods:

• Utilizing high-resolution electroencephalography (EEG) to estimate cortical electrical activity.
• Employing adaptive Partial Directed Coherence to estimate time-varying functional connectivity.
• Applying graph theory-based mathematical indexes to summarize and interpret causality patterns.

Main Results:

• The proposed method successfully estimated dynamic functional connectivity patterns from high-resolution EEG.
• Causality patterns evolving over time were identified during a motor task.
• Graph theory metrics provided a means to interpret these dynamic network changes.

Conclusions:

• The novel approach effectively captures relevant features of functional brain networks in the time-frequency domain.
• This method provides a robust way to analyze dynamic brain connectivity.
• The findings demonstrate the potential of integrating EEG, time-varying connectivity analysis, and graph theory for neuroscience research.