Designing an Intelligent Support System for Emergency Cardiac Care

Abstract

Electrocardiogram (ECG) is the most easily accessible bio-electric signal that provides the doctors with reasonably accurate data regarding the patient heart condition. Many of the cardiac problems are visible as distortions in the electrocardiogram. The major task in diagnosing the heart condition is analysing each heart beat and co-relating the distortions found therein with various heart diseases. In this paper the authors have focused on a particular extracted feature of the ECG signals for use with Artificial Neural Networks (ANNs). The accurate and automated detection of the R-peak value of the ECG signal is essential for efficient results. Here, the task of the ANN is to correctly classify five classes of diseases: Normal, Left bundle branch block, Right bundle branch block, premature ventricular contraction and atrial premature contraction. Further, another ANN is used to accurately determine the R-peak values of the ECG signal. The system is able to provide a rudimantary assessment of cardiac problems. This has been verified using experimental results. Here, the morphological feature extraction scheme is used for feature extraction. The simulation data are obtained from Massachusetts Institute of Technology/Beth Israel Hospital (MIT-BIH) arrhythmia database demonstrating different kinds of cardiac diseases detectable by ECG signals.