Sistem Deteksi Fibrilasi Atrium menggunakan Fitur RR Elektokardiogram dengan Jaringan Syaraf Tiruan Eka Anzihory, Nuryani, Darmanto

Abstract

Research for Atrial Fibrillation detection at electrocardiogram (ECG) using Artificial Neural Network (ANN) with RR statistic features has been successfully implemented. This study was conducted by varying RBF NN, MLP-BP NN and LVQ NN to determine the best of ANN in detecting AF. The best results were found when seven features from RR statistic features at length 15 beats of ECG segment by using RBF NN. The best performance were 99,97%, 99,84% and 99,89% in terms of sensitivity, specificity and accuracy, respectively.