ECG Signal Process And ECG Data Compression

Cardiac diseases are the major factor that harms the human health. The analysis of ECG waves is important for clinical diagnosis of cardiac diseases. The popular use of computers to accurate and speedy diagnose cardiac disease diagnosis has been a subject fervently pursued by both internal and external researchers. The detection of The QRS complexes is the most important task in automatic ECG signal analysis, and the first accordance to arrhythmia diagnosis. The major factor that interferes with the detection of QRS complexes is noise. With a view to raise the rale of accuracy in QRS detection, how to suppress noise is the most difficult in ECG waves analysis. On the other hand, regarding some disease patients, needs to enter president to his/her ECG the monitor record, for instance 24 hours, by now the data quantity was extremely big, in order to reduce its occupies the storage space, needed to the data which gathered to carry on the compression, by enhanced the storage space the use factor.The paper is Wavelet Transform-Based ECG Characteristics Detector. It is the more important application in signal process based Wavelet Transform (WT). The purpose of this thesis is to recognize several typical QRS wave by using wavelet transform, re-construct ECG signal with the filter bank, and to provide a new approach for ECG data compression based on wavelet neural network.Firstly, Based on two-channel perfect reconstruction condition, a biorthogonal filter bank based on quadratic spline wavelet is designed. Then referent to the previous experience about QRS detection with first difference wavelet, with the feature of WT's, the R wave is detected. Bringing forward madule max value pair's detector spanning scales and dynamic threshold arithmetic combining with body physiology and synthesis strategy, which takes duplicate detector. The result is famous.Secondly, Furthermore, before and after noise suppressed, ECG signal is re-constructed with the filter bank.