RADIAL BASIS FUNCTION NEURAL NETWORK IN THE ANALYSIS OF SEGMENTED ELECTROCARDIOSIGNALS

Abstract

Artificial neural networks (ANN) are used to model the information processing capabilities of nervous systems. Research in the field of analysis and interpretation of the cardiovascular diseases with ANN has been attracting more attention in recent years. This paper proposes the use of segmentation of electrocardiograms in time relative to R-wave. The ECG parameter is chosen as an approach based on ECG segmentation on 3 key areas that are responsible for the atria, ventricles depolarization and repolarization of the ventricles. Radial basis function neural network (RBFNN) is chosen for the recognition of abnormality in each segment.

        Results show that in the analysis of atria, ventricles depolarization and repolarization of the ventricles segments, the best spread values for each RBFNN are 1, 2.5 and 1.5 respectively. By selecting the optimal spread values for each segment the average sensitivity for all segments is 82.4 and the average   specificity is 93.7.