Implemented The Expert System of Heart Disease by Using SVM

Heart sounds are the most significant parameter that can be used to represent the heart condition. However, the original heart sounds signal such as PCG still contained some unwanted information while the recording process happens that becomes a noise or extra sounds that will influence inside of the signal. This paper present signal processing technique and data analysis to suppress any noise in the recorded signal and classified it into two groups which are normal heart sounds and pathological heart sounds that contain Ventricular Septal Defect (VSD) inside. There are some steps in signal processing, first the signal will be processed through preprocessing technique which consists of the filter and smoothing by using IIR high pass filter in LabVIEW, secondly segmentation and feature extraction that is used to recognize the different parameter on each signal, and finally we use Support Vector Machine (SVM) to classify the heart sound signals. The result of the method capable to separate and classify the normal and pathological heart sound signals.