Choice of tau in the estimation of ApEn and SampEn entropy of EMG, ECG and EEG signals

Electrophysiological signals have been of great interest in the area of signal processing, because they provide relevant information to medical specialists in the interpretation of pathologies. This study aims to show the effects of the choice of the tau (τ) parameter, when the Approximate Entropy (ApEn) and Sample Entropy (SampEn) algorithms are applied, to surface electromyography (EMG), electrocardiography (ECG) and electroencephalography (EEG). In this study, the values of the parameters m = 2 and r = 0.2 of the standard deviation of the signal were fixed. The values obtained from entropy, by varying the sampling frequency (Fs) and tau (τ), showed that there is a close relationship between Fs (important in the digital acquisition of the signal), τ, entropy algorithm and the type of electrophysiological signal.