Aplikasi Jaringan Syaraf Tiruan untuk Klasifikasi Isyarat Suara Jantung Menggunakan Metode Spektral

Antonisfia, Yul (2008) Aplikasi Jaringan Syaraf Tiruan untuk Klasifikasi Isyarat Suara Jantung Menggunakan Metode Spektral. POLI REKAYASA, 4 (1). pp. 69-76. ISSN 1858-3709

[img]
Preview
PDF
507Kb

Official URL: http://www.polinpdg.ac.id

Abstract

The aim of this research is to develop a program with a back-propagation artificial neural network algorithm to classify heart sound signals. The classification is done for normal heart sound and 24 heart murmurs. Using a stethoscope with 1 kHz bandwidth amplifier, normal heart sound from some human samples are detected. For completing pathology data, heart murmurs are downloaded from internet. The signals of heart sound were analyzed it’s power spectral, for feature extraction. The results show that using feature extraction with spectral method and back-propagation artificial neural network, we can differentiate the normal heart sound and heart murmurs. The input tolerance of the system is tested using noisy data. The percentage of success of 90 % is obtained with SNR of 10 dB for 125 frequency samples. The best architecture and network parameters result 125 input neurons, 2 hidden layers with 75 and 50 neurons on each layer, 25 output neurons, 0.6 learning rate and 0.7 momentum coefficient.

Item Type:Article
Subjects:G Technology > GI Electrical engineering. Electronics Nuclear engineering
Divisions:Jurusan Teknik Elektro > Prog.Studi Teknik Elektronika-D3 > Journal, Proceeding
ID Code:546
Deposited By:Mr. OP Repo-1 Puskom
Deposited On:30 Nov 2016 15:30
Last Modified:30 Nov 2016 15:30

Repository Staff Only: item control page