Neural Network Analysis of Electrocardiogram Signal

Improving the efficiency of diagnosis of cardiovascular diseases is one of the main tasks of modern health care system. The article considers the possibility of improving the efficiency of diagnosis of cardiovascular diseases through research and justification of the analysis of electrocardiogram (ECG) signals using neural network approach. It is shown that the consideration of direct and reciprocal signs of myocardial infarction lets determine the localization, depth and stage of myocardial infarction. It is established that the choice of neural network architecture and the neural network training algorithm determine the effectiveness of neural network analysis of electrocardiogram (ECG) signals.