Various Networks used for ECG Signals, Heart Beats and ECG Feature’s Classification

Abstract

ECG is a graphical representation of heart’s electrical activity such as electrical reploarization and depolarization of heart. It is an important non- stationary signal which contains the necessary information about the heart functioning so that it can be used to identify different abnormalities in heart beats and also to identify different diseases of human beings. Classification is an important process in ECG signal analysis and cardiac diseases diagnosis process. Different ECG signals as well as ECG parameters such as heart beats, features can be classified according to requirement. In this paper different classification networks have studied. SVM classifier with empirical mode decomposition represented the maximum accuracy of 99.54%. Any optimization technique can be used to increase the accuracy of SVM classifier with suitable decomposition method such as variatinal mode decomposition.