Measuring Phase-Amplitude Coupling Based on the Jensen-Shannon Divergence and Correlation Matrix

Area of Science:

• Neuroscience
• Computational Neuroscience
• Signal Processing

Background:

• Phase-amplitude coupling (PAC) is crucial for neural information processing and cognition.
• Existing PAC measurement methods have limitations in performance and noise resistance.

Purpose of the Study:

• To introduce a novel PAC measurement method utilizing Jensen-Shannon (JS) divergence and correlation matrices.
• To evaluate the proposed method's effectiveness, robustness, and performance compared to existing algorithms.

Main Methods:

• Constructing a correlation matrix from high-frequency oscillation (HFO) amplitude distributions within low-frequency oscillation (LFO) phase bins.
• Calculating matrix elements using JS divergence between amplitude distributions.
• Estimating PAC strength via omega complexity extracted from the correlation matrix.

Main Results:

• The proposed method accurately reflects PAC strength and shows minimal variation with data length.
• It demonstrates superior performance against additive white Gaussian and spike noise compared to five existing algorithms.
• The method effectively detects PAC across wide frequency ranges and replicates previous findings on real neural data.

Conclusions:

• The novel JS divergence-based method provides a powerful and robust tool for quantifying PAC in neural oscillations.
• This approach offers improved accuracy and noise resilience for analyzing neural information processing.
• The method is validated for analyzing real local field potential data, demonstrating its practical utility.