Prediction of Cardiac Arrhythmia using Artificial Neural Network

Abstract

eart Rate Variability (HRV) is a non-invasive way of analysis for estimating the autonomic nervous system and based on measurement of the variability of heart rate signals. HRV signal is mostly noted for cardiac arrhythmia detection and classification. In this paper, artificial neural network (ANN) is used as a classifier to predict cardiac arrhythmias into five classes. The HRV data, RR interval time series is obtained using the Electrocardiogram (ECG) data from the MIT-BIH Arrhythmia Database. The linear and nonlinear parameters are extracted from RR intervals are used to train Multi-Layer Perceptron (MLP) neural network. In our study of neural network for time series i.e. RR interval time series, the prediction of Normal Sinus Rhythm (NSR), Premature Ventricular Contraction (PVC), Atrial fibrillation (AFIB), Left bundle branch block (LBBB) and Second degree block (BII) can be done using proposed algorithm. The 70% of the datasets are used to train MLP neural network. The proposed algorithm is predicted with next 30% of the datasets and satisfactory results obtained with prediction overall accuracy of 97%.