A simple way of distinguishing chaotic characteristics in ECG signals

Accurate identification of chaotic behavior from a deterministic periodic process is a difficult but significant issue as underlying dynamics in the data determines the sequent analysis techniques. Since both chaotic and periodic time series can have the similar waveform and spectrum the commonly used approaches for detecting chaotic behavior have limitation. So in this paper, we present an alternative method to observe the prediction on the trajectories of the given system against variable prediction time lags by using the nonparametric model. While applying the method to numerical data and three types of electrocardiogram (ECG) data, the results demonstrate that the approach we adopted can reveal the distinction between the chaos and periodic process, and is a convenient and effective tool for practical application.