Detection of supraventricular tachycardia by Machine Learning

Abstract

Supraventricular tachycardia (SVT) means abnormally fast heartbeat. It often affects young, healthy people and it is generally caused by the fault in electrical signals in the heart. Supraventricular tachycardia (SVT) is one of the type of arrhythmias that is basically an abnormal heartbeat. Tachycardia means a rapid increase in the heart rate which is of more than 100 beats per minute. Electrocardiogram (ECG) is one of the most important diagnostic tool used for the detection of the health of a heart. The number of growing heart patients has made necessary development in the techniques of automatic detection for detecting the various types of abnormalities or the arrhythmias of the heart to basically reduce the pressure and share the load of the physicians. ECG recordings were taken from MIT-BIH supraventricular arrhythmia database (SVDB) of the Physionet repository. Each record contains ST, N and VB rhythm and of 30 minutes length. Then using feature extraction , we extracted features for ST,N and VF and finally put into classifier like decision tree(LMT) and functions(MLP) to classify the ECG signals.