Assessment of dynamic phase amplitude coupling using matching pursuit

Area of Science:

• Neuroscience
• Computational Neuroscience
• Signal Processing

Background:

• Neuronal communication relies on interactions across multiple oscillatory frequencies.
• Phase amplitude coupling (PAC) quantifies these interactions, but existing measures are limited to average coupling within fixed time windows.
• Dynamic PAC is crucial for understanding cognitive functions, yet current time-varying methods using sliding windows lack adaptability and are sensitive to arbitrary window length selection.

Purpose of the Study:

• To introduce a novel, data-driven dynamic phase amplitude coupling (PAC) measure.
• To overcome the limitations of existing sliding window approaches for estimating time-varying PAC.
• To improve the accuracy and robustness of PAC quantification in dynamic neural signals.

Main Methods:

• A new dynamic PAC measure based on matching pursuit (MP) is proposed.
• The MP approach decomposes signals into time- and frequency-localized atoms.
• Dynamic PAC is calculated by quantifying the coupling between these localized MP atoms.

Main Results:

• The proposed MP-based method accurately detects coupled frequencies and their temporal variations with high resolution in synthesized data.
• Analysis of real electroencephalogram (EEG) data revealed significant theta-gamma and alpha-gamma PAC during specific response intervals.
• The method demonstrates superior performance in capturing PAC within short time windows and exhibits increased robustness to noise compared to sliding window methods.

Conclusions:

• The proposed MP-based dynamic PAC measure effectively quantifies and tracks time-varying PAC.
• This data-driven approach offers a more robust and potentially more sensitive method for analyzing neural oscillations.
• The findings suggest improved insights into the dynamic nature of brain communication during cognitive processes.