A statistical framework to assess cross-frequency coupling while accounting for confounding analysis effects

Area of Science:

• Neuroscience
• Computational Neuroscience
• Signal Processing

Background:

• Cross-frequency coupling (CFC) is a critical neural mechanism linking brain function and dysfunction.
• Numerous analysis methods exist for specific CFC types, but inappropriate method selection can reduce statistical power and introduce confounds.
• Accurate quantification of CFC is essential for understanding neural dynamics.

Purpose of the Study:

• To develop a robust statistical modeling framework for estimating cross-frequency coupling (CFC).
• To create a measure of phase-amplitude coupling that accounts for variations in low-frequency amplitude.
• To provide a more reliable method for analyzing brain activity patterns.

Main Methods:

• Proposed a statistical modeling framework to estimate high-frequency amplitude as a function of low-frequency amplitude and phase.
• Developed a measure of phase-amplitude coupling that explicitly controls for low-frequency amplitude.
• Validated the method using simulations and biologically-motivated examples.

Main Results:

• The proposed method successfully detects CFC between low-frequency phase/amplitude and high-frequency amplitude.
• The framework outperforms existing methods in detecting CFC in simulated and real-world neural data.
• Demonstrated the method's utility in analyzing in vivo data from seizure activity and response to electrical stimuli.

Conclusions:

• The developed statistical framework offers a more accurate and reliable approach to quantifying cross-frequency coupling.
• This method enhances the understanding of neural dynamics by accounting for amplitude variations.
• The findings have implications for studying brain function, dysfunction, and responses to stimuli.