Heart sounds recognition using multifractal detrended fluctuation analysis and support vector machine

Abstract

In this paper, a heart sound recognition algorithm is based on Multifractal detrended fluctuation analysis (MFDFA) to obtain most of the specifications of the heart sound signals, and support vector machine (SVM) to classify the features of signals of heart sound which distinguish the normal signals from the abnormal ones. The aim of this study is the development of computerized program to help physicians in the diagnosis of heart diseases. This algorithm allows us to classify the signals with accuracy percentage of 96.875%. The proposed method is evaluated using heart sound signal available in the web site of PhysioNet. They are collected from a variety of several environments from both healthy subjects and pathological patients. The recording signals include children and adults.