A Mutual Information Measure of Phase-Amplitude Coupling using High Dimensional Sparse Models

Area of Science:

• Neuroscience
• Computational Biology
• Gastroenterology

Background:

• Cross-frequency coupling (CFC) in brain electrophysiology is linked to neurological disorders like Parkinson's disease and epilepsy.
• CFC has been identified in gut-brain electrophysiologic studies, highlighting its potential role in gut-brain axis disorders.
• Existing CFC detection methods do not fully capture the statistical underpinnings of the coupling phenomenon.

Purpose of the Study:

• To develop and validate a new method for quantifying phase amplitude coupling (PAC) in electrophysiological signals.
• To improve the detection of low-level PAC, particularly in the context of gut-brain communication.
• To establish a more statistically robust approach for analyzing complex neural and gut signals.

Main Methods:

• Developed an exponential generalized linear model (GLM) to estimate amplitude based on phase.
• Utilized a high-dimensional basis of von-Mises functions and l1-regularized model selection for flexible parametric modeling.
• Employed mutual information estimation to quantify the coupling between phase and amplitude.
• Validated the method using synthetically generated gut-brain coupled signals.

Main Results:

• The proposed method accurately detects low levels of phase amplitude coupling.
• Receiver operating characteristic (ROC) curve analysis demonstrated superior performance compared to existing gold-standard methods.
• The flexible parametric modeling approach effectively captures the underlying statistical relationships in coupled signals.

Conclusions:

• The novel mutual information-based method offers enhanced sensitivity and accuracy for detecting PAC.
• This approach provides a more robust tool for investigating gut-brain axis dysfunctions.
• The findings pave the way for improved diagnostic and therapeutic strategies targeting gut-brain communication.