The Automatic Identification of PVC Based on Wavelet Transform and BP Neural Network
10.3969/j.issn.1005-202X.2010.02.016
- VernacularTitle:基于小波变换和BP神经网络的室性早搏(PVC)识别
- Author:
Yingjun LEI
- Publication Type:Journal Article
- Keywords:
ECG;
PVC;
BP neural network;
pattern recognition
- From:
Chinese Journal of Medical Physics
2010;27(2):1762-1765
- CountryChina
- Language:Chinese
-
Abstract:
Objective: The premature ventricular contraction (PVC) is one of the most familiar diseases of the arrhythmia. Its real-time and accurate examination has important clinical significance. To find a premature ventricular contractions (PVC) early and improve the diagnostic accuracy of premature ventricular contractions (PVC),a method called automatic identification of PVC based on wavelet transform and BP Neural Network is proposed in this paper. Methods: After selecting the characteristic parameters of ECG signal using wavelet transform,detection of the feature points, extracting characteristic parameters, constructing feature vector and recognition of PVC by BP Neural Network. The automatic identification of normal ECG and PVC was realized. Results: A 7-10-1 three-layer BP Neural Network is constructed in this paper, the neural network was proved having better automatic identification ability (average 92%) after training and testing the neural network using the ECG datum of MIT-BIH database. Conclusions: Analyzing experimental results of six groups of test data in MIT-BIH ECG database,the method could assist doctors in making better diagnosis on ECG potentially.