On the aspect of feature extraction and classification of the ECG signal

Abstract

The electrocardiogram provides a physician with a view of the heart's activity through electrical signals generated during the cardiac cycle and measured with external electrodes. Because of the high mortality rates of heart diseases faithful detection and classification of ECG arrhythmias is essential for the treatment of patients in the clinics. Arrhythmia classification is one of the most important research domains of computer aided medical systems. The authors have made an exploratory investigation of the classification of ECG signal. DWT (Discrete Wavelet Transform) method has been used to determine the wavelet coefficients which were associated with five features. The features were ranked by using class separability criteria. The authors have established the Shannon Entropy as one of the most suitable features for the purpose of classification.