Application Of Time-Frequency Analysis Method In ECG Signal Analysis

Recording of electrocardiograms (ECG) signal is an important basis for heart disease diagnosis. For getting accurate intrinsic characteristic message of physiology activity of patient in time, ECG signal analysis and processing becomes an important problem facing medicinal researchers. ECG signal is a weak low frequency non-stationary signal, and possessing random character and noisy setting strong character. So we need to reflect local vary character during analyzing process. Time-frequency analysis method can reflect signal time-frequency local alter, and occupies important action in ECG signal analysis field.ECG signal usually is weaker, amplitude scope is only from 10μV to 4 mV, so it can be influenced by various interfere easily, so signal de-noising processing is very crucial; R wave takes up very vital position in ECG signal characteristic waves, R wave accurate position is the basic for extracting else parameters of ECG signal; heart rate variability signal contains important message of body autonomic nervous system, extracting instantaneous character message, can afford us reliable quantitative index for pathology analysis and disease diagnosis.This thesis mainly analyzing ECG signal, based on wavelet transform method and Hilbert-Huang transform method, containing signal de-noising problem, R wave detection problem and heat rate variability signal analysis research. Through large experiment results comparing analysis, proves their each excellences and flaws. Multi-resolution analysis character of wavelet transform can well reflect signal local characteristic. But the selection problem of wave base function and not single frequency component problem of wavelet decomposition are very difficult to researchers. Hilbert-Huang transform method possesses self-adaptation and time-frequency centralizing, show its advantage in ECG signal analysis filed.