Statistical approach to quantify the presence of phase coupling using the bispectrum

Area of Science:

• Signal Processing
• Nonlinear Dynamics
• Biophysics

Background:

• Phase coupling detection is crucial for analyzing complex signals.
• The bicoherence index (BCI) is a common but limited method due to noise and low coupling strength issues.
• Existing methods struggle with specificity and robustness.

Purpose of the Study:

• To introduce a novel statistical approach for detecting quadratic phase coupling.
• To overcome the limitations of the bicoherence index (BCI) in noisy and low-coupling scenarios.
• To provide an unbiased and statistically rigorous method for phase coupling analysis.

Main Methods:

• Developed a statistical method combining bispectrum analysis with surrogate data.
• This approach avoids the normalization issues inherent in the BCI.
• Validated the method using simulations with varying noise and coupling levels.

Main Results:

• The proposed method demonstrated superior sensitivity and specificity compared to the BCI.
• Successfully identified significant nonlinear interactions in renal hemodynamic data.
• Observed differences in coupling magnitude and interaction peaks between normotensive and hypertensive rats.

Conclusions:

• The new statistical method offers an accurate and robust way to detect quadratic phase coupling.
• This technique is more reliable than the BCI, especially in the presence of noise.
• The findings suggest distinct nonlinear dynamics in renal hemodynamics between normotensive and hypertensive states.