FEATURE EXTRACTION AND AUTOMATIC CLASSIFICATION OF ECG SIGNAL USING WAVELET TRANSFORM

Abstract

Electrocardiogram (ECG) is a graphical representation of the electromechanical activity of the cardiac system. It provides fast and reliable information to the expert cardiologist with respect to the functional aspects of the heart. For the identification, signal analysis as well as for the diagnosis Electrocardiogram (ECG) signal feature parameters are used as the basis. These parameters can be extracted from the interval and amplitudes of the signal. Where WT (Wavelet Transform) provides efficient localization in both time and frequency and to perform classification. DWT has been used to extract relevant information from the ECG signal. In this work, we have accurately classified and differentiated Normal and abnormal heartbeats such as Right Bundle Branch Block (RBBB), Left Bundle Branch Block (LBBB), Atrial Premature Contractions (APC) and Premature Pentricular Contractions (PVC) with good accuracy using ANN. All the analysis is done using MATLAB. The detection rate of the QRS complex is about 98.8% and classification accuracy is about 98% based on MIT-Bill ECG database.