ECG SIGNAL CLASSIFICATION BASED ON STATISTICAL FEATURES WITH SVM CLASSIFICATION

Abstract

An important diagnostic technique to detect the abnormalities in the human heart is Electrocardiogram (ECG). The growing number of heart patients increases the physicians work load. To reduce their work load, a computerized automated detection system is required. In this paper, a computerized system is presented to categorize the ECG signals. MIT-BIH ECG arrhythmia database is used for analysis purpose. After de-noising the ECG signal in the preprocessing stage, the following features; mean, variance, standard deviation, and skewness are extracted in the feature extraction stage and Support Vector Machine (SVM) is developed to classify the ECG signal into two categories; normal or abnormal. Results show that the system classifies the given ECG signal with 90% of sensitivity and specificity as well.