Study of the Lyapunov exponents of ECG signals from MIT-BIH database

A preliminary analysis of the Lyapunov spectrum determination from a time-series with the Jacobian-based approach is presented. The method has been tested on the Lorenz attractor obtaining good Lyapunov exponents estimate, then it has been applied to the study of ECG signals coming from MIT-BIH Arrhythmia Database. We obtained almost always a positive first exponent confirming the chaotic nature of the electrical heart activity measured from surface ECG. However, Lyapunov spectrum estimate is often influenced by the choice of some working parameters. The main purpose of this work is to gain insights on the influence of such working parameters on results.