ECG Data Analysis Based On Chaotic Time Series

Along with economic development, people's lifestyle and dietary habits have made a vary degree change, so the cardiovascular disease's incidence rate and mortality rate have been increased gradually, which have become the number one mortal enemy of human health. While the heart is an extremely complex physiological system, look from the time series analysis's viewpoint which is a typical example of nonlinear time series. Therefore, we can not analysis ECG with conventional linear method. In view of this, this paper adopted chaotic time series analysis.In this paper, first, I make a brief presentation about the basic principles and characteristic of chaotic dynamics, and then make a detail presentation of several algorithms based on chaotic time series, such as power spectrum, phase-space reconstruction, correlation dimension, Lyapunov exponent. Use these methods to analyze healthy people, premature ventricular contraction and bundle branch block patients. Power spectral analysis show that all of the power spectrum of ECG are in commom with, which have continuous spectrum with peak, so ECG is not noise signal nor periodic signal, but a signal with some certain rules; besides phase-space reconstruction, correlation dimension and Lyapunov exponent show that the movement of heart system is chaotic, healthy heart is the most chaotic system, sick heart system's chaotic is weak, and the chaos of various pathological heart is different.The results showed that the chaotic dynamics parameters of ECG can be used as an effective indicator to evaluate heart system's health state, assisted to the early diagnosis of heart disease, hoped to supply some new methods to early clinical diagnosis.