Deep learning based evaluation of coronary artery disease and estimation of quantitative coronary angiography using myocardial perfusion imaging

Coronary artery disease is the most common type of heart disease worldwide and a leading cause of death globally every year. Early and precise diagnoses are important, and one way to diagnose is by using myocardial perfusion imaging. The goal of this project is to develop a tool based on deep learning for evaluation of coronary artery disease, by performing an automatic estimation of the probability of myocardial ischemia and the degree of coronary artery stenosis. This information is otherwise obtained by interventional methods. Making the prediction automatically from myocardial perfusion imaging would save suffering for patients and free resources within the healthcare system.