Applying EEG phase synchronization measures to non-linearly coupled neural mass models

Area of Science:

• Neuroscience
• Computational Neuroscience
• Signal Processing

Background:

• Understanding information integration across brain regions is a key goal in neuroimaging.
• Detecting functional connectivity networks from electroencephalography (EEG) data presents significant challenges.

Purpose of the Study:

• To evaluate the performance of three phase-based connectivity measures in recovering simulated neural connectivity patterns.
• To compare these measures across various experimentally relevant conditions using neural mass models.

Main Methods:

• Simulated 10-second EEG trials using neural mass models with defined coupling periods (1-3s and 5-8s).
• Varied simulation parameters including oscillation frequency, power spectrum, feedforward/feedback connections, and volume conduction.
• Assessed the ability of three phase-based connectivity measures to detect simulated connectivity patterns.

Main Results:

• Successful detection of synchronization onset and offset was achieved for most of the 28 simulated configurations.
• The tested phase-based measures exhibited differential sensitivity and specificity in identifying underlying connectivity.
• Performance was evaluated across diverse conditions, extending prior work on coupled oscillator models.

Conclusions:

• No single phase synchronization measure demonstrated superior performance across all tested scenarios.
• Combining multiple phase-based measures is recommended for experimental EEG investigations.
• The choice of measures should consider the specific research question, data signal-to-noise ratio, and statistical approach.