DETEKSI ARITMIA BLOKADE CABANG BERKAS KIRI PADA ELEKTROKARDIOGRAM DENGAN JARINGAN SYARAF TIRUAN BERDASARKAN FITUR INTERVAL QR DAN RS

Abstract

Left branch bundle block (LBBB) detection system have been done and tested. The system created by using Artificial Neural Network (ANN). Some ANN methods have been investigated.  Electrocardiogram features that represent the characteristic of LBBB which were QR interval and RS interval used as input. Output of the system was electrocardiogram status to detect LBBB and normal beat. LBBB detection has been done with various input with QR interval, RS interval, QR and RS interval. Detection system have been tested using chlinical data and shown that MLP method gave the best performance. Its performance shown by sensitivity, specificity, and accuracy up to 99,92%, 100%, and 99,94% .

Keywords: LBBB, artificial neural network, QR interval, RS interval