Bispectrum Estimation of Electroencephalogram Signals During Meditation

Objective: Electroencephalogram is a reliable reflection of many physiological factors modulating the brain. The Bispectrum is very useful for analyzing non-Gaussian signals such as EEG, and detecting the quadratic phase coupling between distinct frequency components in EEG signals.The main aim of this study was to test the existence of nonlinear phase coupling within the EEG signals in a certain psycho-physiological state; meditation. Methods: Eleven meditators and four non-meditators were asked to do meditation by listening to the guidance of the master, and 10 subjects were asked to do meditation by themselves. Bispectrum estimation was applied to analyze EEG signals, before and during meditation. EEG signals were recorded using 16-channel PowerLab. ANOVA test was used to establish significant changes in Bispectrum parameters, during two different states (before and during meditation). Results: Mean Bispectrum magnitude of each channel increased during meditation. These increments of phase coupling are more obvious in occipital region (Pz channel) than frontal and central regions (Fz and Cz channels). Besides that phase coupled harmonics are shifted to the higher frequencies during meditation. Conclusion: Bispectrum methods can be useful for distinction between two states (before and during meditation).