A new algorithm to screen potential Implantable Cardiac Defibrillator (ICD) receivers using artificial neural networks

Abstract

In this paper a new decision support system with intelligent algorithm to screen high risk patients due to sudden cardiac death (SCD) patients for implanting cardiac defibrillator (ICD) has been developed. SCD is known as one of top killer in many developed countries. It is caused by the ventricular arrhythmias such as ventricular tachycardia (VT) and ventricular fibrillation (VF). Thus, Implantable Cardiac Defibrillator (ICD) is introduced as the gold therapy for the patients who are at the high risk of VT. However, the ICD is relatively expensive to be installed for every patient and moreover there is a vague guideline on whom to be the potential ICD receiver. The proposed method consists of extracting the standard deviation of the RR interval (SDNN) and left ventricular ejection fraction (LVEF) value, obtaining clinical inputs from cardiologist and using artificial neural network (ANN) for potential ICD receiver identification process. According to the preliminary results, the average good detection rate is 93.33%. This novel algorithm not only can help medical practitioners and cardiologist as a decision support system, also will help patients with most priority to be detected and cured before any serious heart attack.