Comparing parametric and nonparametric methods for detecting phase synchronization in EEG

► We compared parametric and nonparametric methods for assessing phase synchronization in the presence of EMG artifacts. ► The parametric methods used an autoregressive approach to model the data before deriving the time-frequency transform. ► We used two simulations: simple sinusoidal model and more complex neural mass model, to which were added simulated EMG. ► The results indicate that parametric approaches are more robust to EMG noise as power in the signal increases.