Based On Matlab Ecg Wavelet Analysis

At present, the world's heart disease morbidity and mortality stays high, which has been the first threats of human health nowadays. Whereas, ECG signal is the body surface of comprehensive performance of electrical activity of muscle cells composed of human heart, to a certain extent which reflects physiological status of every part of heart. Therefore, studying ECG processing is of practical value and great significance. Over the years, by studying ECG processing to diagnose heart disease or evaluate its treatment effect, remarkable results have been achieved. However, due to the complexity of ECG, there are still some problems in its research process, which need further study and resolution.Meanwhile, wavelet analysis is applied to the field of natural science, including biomedical signal processing. Due to the good time-frequency local property of wavelet transform, various time-vary signal can be decomposed effectively to separated the signal from noise, and the purpose of de-noise is achieved. According to the existing shortage of ECG noise reduction, ECG noise reduction based on wavelet analysis was deeply studied and discussed in the paper. The main research contents are as follows:1. On the noise characteristics of the ECG are described, and the causes of noise are given. The advantages and disadvantages of several traditional processing method for ECG signal, and wavelet transform is proposed as to an ideal method to ECG de-noise.2. The Fourier transform and wavelet theory are introduced. The characteristics of wavelet transform and signal-to-noise property of wavelet analysis are analyzed, providing theoretical basis for ECG noise reduction. Finally, several wavelet generating function commonly used noise reduction are summarized.3. The steps of ECG de-noising are summarized; some methods of threshold selection and quantization are given. Algorithm of Wavelet Transform Modulus Maximum is analyzed, and wavelet threshold de-noising is mainly studied. The simulation experiment of wavelet threshold de-noising based on MATLAB7 is carried out, and the noise reduction effects in which select different threshold and threshold function are compared detailed. The effectiveness of wavelet threshold de-noising is proved.4. Based on the principle of wavelet threshold de-noising, Stationary wavelet transform is analyzed, the de-noising results based on MATLAB simulation is obtained. The de-noising effects indicated by SNR and MSE of traditional discrete wavelet transform and Stationary wavelet transform were compared, and the conclusion that stationary wavelet noise reduction is better is obtained, as well the Gibbs phenomenon is restrained.5. Due to better time-frequency local property of wavelet packet, it is applied to ECG signal de-noising. Through MATLAB simulation, it is proved that signal characteristics is better preserved by wavelet packet noise reduction. Because ECG signal frequency is low and wavelet packet is better applied to de-noising of high-frequency signals, so the de-noising effect of wavelet packet is not obvious compared with traditional wavelet de-noising.