International Journal of Computer Science Issues

More than a traditional journal...

The ECG signals were processed in three steps: First by reconstructing their phase portraits, second by calculating their spectra and finally by computing their Largest Lyapunov exponents. This study presents a framework to assess nonlinear parameters of ECG signals that may be useful for further exploration of physiological and path-physiological significance of dynamics in electrocardiograms. This paper devotes to identify the nonlinear properties of human heart beat dynamics for 18 characteristic examples of normal subjects with normal sinus rhythm. A new phase space characterization of a class of ECG signals with normal sinus rhythm through is presented. Such approach led to find out a common visual feature namely an anchor-shaped Poincare plot resulting from an appropriate choice of the delay time. Any value of the delay time is acceptable, but the shape of the embedded time series depends critically on the choice of its value. The results reported in this study may be useful not only for the classification of ECG states, but can serve as a benchmark to which pathological cases can be compared.