A neural network approach to coronary heart disease risk assessment based on short-term measurement of RR intervals

Abstract

Using short-term heart rate variability (HRV) measurements, this study investigates the relationship between respiratory sinus arrhythmia (RSA) and Coronary Heart Disease (CHD) risk in asymptomatic patients who nevertheless exhibit CHD risk factors. The aim is to train an artificial neutral network (ANN) to recognise HRV patterns related to CHD risk via a Poincare plot encoding. The ANN correctly classified 6 out of 9 'high' 6 out of 9 'medium', and 6 out of 9 'low' risk test cases. It is expected that this result can be improved by increasing the number of input neurons and by using different preprocessing techniques. This study showed that an ANN approach can be successful in detecting individuals at varying risk of CHD based on short-term HRV measurements under controlled breathing.