AN ECG CLASSIFIER DESIGNED USING MODIFIED DECISION BASED NEURAL NETWORKS

In this paper, a neural network based generalized software system is p resented for automatic analysis of electrocardiograms (ECGs). The prop osed system is capable of intuitively diagnosing the disease from the ECG using the knowledge acquired from the training. A modified decisio n based neural network which converges in a finite amount of time is e mployed. The training procedure used automatically varies the size of the network. The system is capable of being trained even without an ex pert's supervision. The physician can correct the network as and when a misclassification occurs, thus making the system less error-prone as time passes. The proposed system has been tested using an ECG data ba se representing different cardiological conditions such as bundle bran ch blocks and infarctions. The system is capable of detecting differen t types of arrhythmias also. (C) 1997 Academic Press.