Application on intelligent analysis for ECG signals by adopting wavelet transform

In the course of collecting ECG signal data, it is inevitable that they are interfused with noise signals, such as electromyographic interference, baseline drift and 50 Hz power line interference, which frequently cause errors in identifying the characteristics of ECG signals to different degrees. This research indicates that the dyadic wavelet transform is a set of band-pass filters taking on a different frequency band in each scale. All kinds of interferences and baseline drift can be commendably restrained by utilizing dyadic wavelet transforms in order to identify exactly the characteristics of ECG signals. The numerical analysis is verified with data collected by ECG monitor and transported by RS-232 asynchronous serial communication.