Electrocardiogram (ECG) signal classification using s-transform, genetic algorithm and neural network

The identification of the electrocardiogram (ECG) signal into different pathological categories is a complex pattern recognition task. In this paper, a classifier model is designed to classify the beat from ECG signal of the MIT-BIB ECG database. The classifier model consists of three important stages (i) feature extraction (ii) selection of qualitative features; and (iii) determination of heartbeat classes. In the first stage, features are extracted using S-transform where as second stage uses the genetic algorithm to optimize the extracted features which represent the major information of the ECG signal. The final stage classifies the ECG arrhythmia. In this study, we have classified six types of arrhythmia such as normal (N), premature ventricular contraction (PVC), atrial premature contraction (APC), right bundle branch block (RBBB), ventricular fusion (VF) and fusion (f). The experimental results indicate that our method gives better result than earlier reported techniques.