In this paper, a novel use of Kernel-Adatron learning algorithm to aid SVM (support vector machine) for ECG arrhythmias classification is proposed. The proposed pattern classifier is compared with MLP (multi-layered perceptron) using Backpropagation learning algorithm. The ECG signals taken from MIT-BIH arrhythmia database are used in training to classify 6 different arrhythmia together with normal ECG. Arrhythmias were LBBB (left bundle branch block), RBBB (right bundle branch block), PAB (premature atrial beat), PVB (premature ventricular beat), PB (Paced beat) and FB (Fusion of paced and normal beat). The MLP and SVM training and testing stages were carried out twice. They were first trained only with one ECG lead signal (MLII) and then a second ECG lead signal (V1) was added to the training and testing datasets. This was done to investigate its influence on training and testing performance (generalization ability) and time of training for both classifiers. The results indicate that SVM in comparison to MLP is much faster in training stage and nearly 7 times higher in performance, but MLP generalization ability in terms of mean square error is more than 3 times less. The proposed SVM method shows considerable improvement in relation to recently reported results obtained by (Osowski & Markiewicz, 2008).

Keywords