Исследование миокардиальной перфузии методами машинного обучения и искусственного интеллекта

Abstract

This report outlines the idea of applying machine learning methods in nuclear medicine to analyze myocardial perfusion. This work includes the classification of patients into healthy and patients with abnormalities; classification of the disease of an unhealthy patient into ischemia and myocardial infarction; segmentation of the affected area of the heart on the polar map of myocardial perfusion. This article also presents some of the results of the initial research and the results of the first stage of work.