ECG Arrhythmias Detection Using a New Intelligent System Based on Neural Networks and Wavelet Transform

In this paper, Automatic electrocardiogram (ECG) arrhythmias classification is essential to timely diagnosis of dangerous electromechanical behaviors and conditions of the heart. In this paper, a new method for ECG arrhythmias classification using wavelet transform (WT) and neural networks (NN) is proposed. Here, we have used a discrete wavelet transform (DWT) for processing ECG recordings, and extracting some time frequency features to be used for training a multi-layered perceptron (MLP) neural network. In fact, the MLP NN performs the classification task. Although many algorithms have been presented for ECG arrhythmias detection over the past years, the results reported in the past, have generally been limited to relatively small set of data patterns. Here, we have used 20 recordings of the MIT-BIH arrhythmias data base for training and testing our neural network based classifier. The simulation results show that the classification accuracy of our method is 97% over 420 patterns using 20 files including four arrhythmias.