Ecg Signal Automatic Diagnosis Based On Fuzzy Neural Network

Automatic analysis and diagnosis of ECG signal processing has been an important research topic. In this paper, to wavelet transform and fuzzy neural network as the theory background, the identification and classification of normal heart beat and premature ventricular contractions is further studied.This paper introduces the structure of ECG automatic diagnosis system and the development situation of each part, then introduces some basics of ECG, including the mechanism of ECG and the common ECG characteristics of several arrhythmia and arrhythmia diagnostic standards. On this basis, this paper makes analysis to a variety of noise and interference exist in ECG, and for noise as power line interference, electromyographic noise and baseline drift, proposes related denoising method. In the following that ECG signal is completed in the pretreatment, the paper describes two groups ECG QRS wave detection methods, that is, the peak location method and the wavelet transform. This paper has a clear focus on R wave detection, QRS wave group start and end point detection on the use of wavelet transform, on the basis makes algorithm design and software simulation on R wave detection and QRS wave group detection and then analyze the results. Using these two algorithms the paper has simulation on ECG data of MIT-BIH ECG database, and the simulation results show that the recognition rate of the two algorithms on the QRS wave group respectively reaches 99.43%and 99.84%.The problem of ECG automatic diagnosis is that automatic classification of ECG waveform makes the computer recognize a variety of abnormal waveforms. Upon completion of the following ECG feature extraction, the paper combines fuzzy logic and neural networks to the fuzzy neural network. Using the classification capability of a feedforward neural network provided with fuzzy input and output, automatic identification of normal heart beat and premature ventricular contractions of the ECG signal is realized. And then using sample data of MIT-BIH ECG database to train and test the fuzzy neural network, the paper proves that the network has high PVC automatic recognition.