Comparative Analysis of Methods for Training Artificial Neural Networks Used in Diagnosing Coronary Heart Disease

Abstract

The widespread prevalence of coronary heart disease is one of the main causes of death in developed countries of the world. This disease is difficult to diagnose due to the large number of factors needed to be considered, and therefore it is advisable to use artificial neural networks to increase the accuracy of diagnosis. The article presents methods for training artificial neural networks used in diagnosing coronary heart disease, and their comparative analysis based on indicators of specificity, sensitivity, and accuracy of trained artificial neural networks. Finding an accessible, reliable method for diagnosing these diseases is of great medical importance.