IDENTIFICATION OF HEALTHY MURMUR VOICE SOFTWARE USING A FAST FORIER TRANSFORMATION BASED ON NEURAL NEURAL NETWORK

  • Muhammad Jumnahdi Department of Electrical Engineering University of Bangka Belitung

Abstract

This research was to explore the merit of signal processing scheme in solving the identification problems of heart murmurs signal utilizing the available data from a number of Indonesia heart patients. The signal processings were based on an Artificial Neural Network (ANN) method preceded by the Fast Fourier Transform (FFT) analysis and directed to nine known classes of murmurs, namely the stenosis aorta valve,defect septum atrium, regurgitation mitral, defect septum ventricle, click-mid systolic, regurgitation stenosis aorta, presistolic, stenosis mitral, and ductusarteriosus paten ones. The ANN weight vectors were grouped according to their clustering patterns which represent their respective specific features, based on the corresponding training data samples. The 78.5% success of identification on other data samples is encouraging. The research resorted mainly to the available MATLAB software tools.

Keywords: Heart Murmurs, Specific Features, Artificial Neural Network, FFT, Training

Downloads

Download data is not yet available.
Published
2014-04-30
How to Cite
[1]
M. Jumnahdi, “IDENTIFICATION OF HEALTHY MURMUR VOICE SOFTWARE USING A FAST FORIER TRANSFORMATION BASED ON NEURAL NEURAL NETWORK”, Jurnal Ecotipe, vol. 1, no. 1, pp. 23-28, Apr. 2014.
Abstract viewed = 119 times
PDF (Bahasa Indonesia) downloaded = 887 times