Nonlinear Measure of ECG Time Series: detection of cardiac diseases

Recent developments in the theory of nonlinear dynamics have paved the way for analyzing signals generated from nonlinear biological systems. The main purpose of the present work is based on the analysis of the ECG signal, ini- tially extracting the features of ECG, which are used for the detection and/or classification of ECGs. For this work, Correlation Dimension (D2), Largest Lyapunov Exponent (LLE), Ap-proximate Entropy (ApEn), Sample Entropy (SampEn) and Poincare plot methods were used from nonlinear time series analysis to characterize human ECG signals obtained from 24 hour-Holter recording. Four groups of ECG signals have been investigated. D2 and LLE are increasingly used to classify ECG signals. ECG time series were classified according to the results obtained from computation of above chaotic features. Our results, obtained from clinical data, improved the previous studies, which allow one to distinguish between healthy group and patients groups with more confidence than the standard methods for heart rate time series and gain more significant understanding of heart dynamics using Entropy features and Poincare plot along with D2 and LLE.