Asiye jafarzadeh

Determining of heart condition by cardiac auscultation is a difficult task and needs Special education and medical personnel. The presented computer techniques provide better and wider results. Hence, this study with the aim of providing Computer analysis of the heartbeat for supporting the physicians about their decisions on these issues has been conducted. In this study, the overall algorithm is based on sampling of the signals and signal processing by using short time Fourier. In order to recognize the Heart lobe signal crossing zero algorithm and to Calculate signal strength and for extracting the Cardiac signal features into Wavelet- Dabochy by Shannon coefficients and coefficients used in signal Details and for extracting the best feature signals of Wavelet- Dabochy To apply the six-step signal and Coefficients are used six times. Then, Signals are classified by support vector machine with the sensibility of 98.46 and the accuracy of 98.13 % separated and classified. In this research, the data base of Michigan University and Physionet website have been used which includes 176 pcs normal audio files and 656 pcs audio files related to systolic Disease and so on…

PCG Signals, Wavelet Dabochy adapter, SVM classifier