ECG Signal Based Heart Disease Prediction System using DWT and SVM

Abstract: The prediction of heart diseases in the medical science world is quiet difficult even today. Human heart generates the electrical signal known as Electrocardiogram (ECG) signal which is used to identify the human heartbeat. It contains the valuable information about the human heart activities for identifying any abnormalities and also are used to measure the heart rate and regularity of heartbeat. In this paper, wavelet functions db8 and sym8 are used to predict the heart diseases such as bradycardia, tachycardia, first degree heart block and healthy person. In this context, the cycle P-QRS-T in the ECG signal which determines the amplitude and location of each peak using QRS complex to identify the cardiac disorder. Thus based on P-QRS-T peak values it is easy to predict the heart diseases. For feature extraction DWT (Discrete Wavelet Transform) is used and SVM (Support Vector Machine) is used for classification. The results are found to be encouraging in terms of detection heart disorder.

Keywords: Image Processing, DWT, SVM, ECG Signals.