Intelligent Cardiac Arrhythmia Detection Using Wavelet Network

Nowadays, cardiac arrhythmias are one of the most important life-threatening factors in the world. Electrocardiogram (ECG) is an exclusive non-invasive tool for cardiac arrhythmias detection (CAD). Intelligent methods of detecting cardiac arrhythmias are the best and assured ways for accurate and real-time detection methods to satisfy this purpose. Artificial neural networks (ANNs) are efficient and intelligent tools for solving diverse problems in pattern recognition tasks such as detection and classification. These networks are the simple models of human nervous system having powerful capability in modeling and parallel processing in addition to having simple structures. In this research, a novel method of detecting cardiac arrhythmias is proposed by the use of a wavelet network (WN). This network is an innovative tool, which incorporates the advantages of ANNs in learning, predicting in, predicting isار)ای تفکیک بیماری، دانشگاه تربیت مدرس تهران addition to the high accuracy and multi-resolution analysis of wavelet transform (WT). Wavelet network also has the capability of activation function selection in respect to the given signal and distinctive algorithms for network architecture optimization. In this study, the data from MIT-BIH database were used to implement this novel method. The obtained result has an accuracy of 98. 8% which is an improvement in comparison with the similar works. Discussion and The result of this work shows a higher accuracy in compared to the one obtained from other methods such as neural networks and wavelet transform.