Untangling cross-frequency coupling in neuroscience.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Brain Dynamics

Background:

• Cross-frequency coupling (CFC) is theorized to coordinate neural activity across different spatial and temporal scales.
• Standard CFC analysis methods present significant challenges in physiological interpretation.
• Apparent CFC can arise from spectral correlations due to common non-stationarities, not true neural interactions.

Purpose of the Study:

• To address fundamental problems in standard cross-frequency coupling (CFC) analysis and interpretation.
• To provide a roadmap for improving the mechanistic understanding of CFC.
• To organize statistical and modeling approaches based on biophysical interpretability.

Main Methods:

• Review and organization of existing statistical and modeling approaches for CFC analysis.
• Categorization of methods based on their biophysical interpretability.
• Identification of common errors and challenges in CFC analysis.

Main Results:

• Apparent CFC can be generated by non-interacting neural frequency components due to shared non-stationarities.
• A framework is presented to organize methods for a better mechanistic understanding of CFC.
• Practical recommendations are provided to mitigate common analytical errors.

Conclusions:

• Standard CFC analysis requires careful interpretation to avoid spurious findings.
• Improved statistical and modeling approaches are needed for accurate understanding of neural coordination.
• Recommendations are offered to enhance the reliability and interpretability of CFC studies in neuroscience.