Denoising and arrhythmia classification using EMD based features and neural network

Computer-assisted cardiac arrhythmia detection and classification can play a major role in the management of cardiac disorders. But detecting the type of arrhythmia is tedious due to the contamination of ECG signal during acquisition. In this paper the proposed work is to remove the major noises like 50 Hz power line interference and baseline wandering from the ECG signal using Empirical Mode Decomposition. Then the QRS complex is detected from the intrinsic mode function and the different types of arrhythmias are classified using back propagation neural network. Most of the arrhythmia signals are taken from MIT-BIH arrhythmia database and some of the simulated ECG signals are also used in this work. The simulations are carried out in a MATLAB environment.