The Study On Processing Ecg Signals Based On Wavelet Transform

The heart attack is a big common disease that endangers people's health and is the mankind's leading killer. So the technique of diagnosing heart attack is of great importance. After analyzing comprehensively biomedical signals processing at home and abroad, aiming at the shortage of existing processing ECG signals, this paper chooses a latest subject - the study on processing ECG signals based on wavelet transform. It can be prepared for advanced examination instruments and automatic analyst systems of ECG signals.Examination and processing methods of ECG signals include hardware methods and software methods. But all traditional and existing theories and practical methods of processing ECG signals exist shortages to some extent. Therefore it is necessary to research further in order to look for a more dependable project. Wavelet transform is a new realm that develops and strengthens from applied mathematics recently. Because multi-resolution analysis of the wavelet transform has excellent time-scale characteristic, especially it is suitable for processing unsteady picture signals. This paper takes advantage of wavelet transform to decompose the ECG signals in different scale, so characteristic information of ECG signals can be abstracted, and the possibility of this method was verified through recreation.Adopting the wavelet transform will do away with noises of ECG signals. The threshold method and max modulus method of wavelet transform was adopted to de-noise ECG signals according to the characteristic of ECG signals, and the two methods both attain to de-noising result. The threshold method is better than traditional de-noising method and can meet requests of engineering application basically, although it is not perfect. In the end, it was testify validly with an experiment and simulation in Matlab.In addition, this paper reviewed some problems that will meet in signals abstraction and processing from many aspects, such as the realization of fast...