Desain Sistem Klasifikasi Kelainan Jantung menggunakan Learning Vector Quantization Endah Purwanti, Franky Chandra A. S., Pujiyanto, Muhammad Arief Bustomi

Abstract

Electrocardiograph (ECG) is a diagnostic tool that measures and records the heart’s electrical activity. ECG signal analysis is often used to diagnose some types of heart defects. In this study, we designed a system of artificial neural networks for image classification electrocardiogram. Image processing method used for ECG feature extraction and image classification process using learning vector quantization. Some electrocardiogram data is used as training data and testing the network classification. Three types of cardiac abnormalities can be detected by the system. Simulation results show that the accuracy of the classification algorithm is composed by 89% of the normal 9, 4 bradycardia, tachycardia 8 and 7 arrhythmias.