Machine learning improves the accuracy of coronary artery disease diagnostic methods

The diagnostic process of coronary artery disease (CAD) consists of evaluation of symptoms and signs of the disease and ECG at rest, ECG during exercise, myocardial perfusion scintigraphy (MPS) and coronary angiography. Machine Learning (ML) can use all particular data in interpretation of result. The authors' goal was to predict in a group of 327 patients the results of coronary angiography obtained by ML method and compare them with the results of MPS as the highest step in the classical diagnostic procedure. The Naive Bayesian Classifier as one of the ML methods was applied. The sensitivity of MPS was 0.83 and specificity 0.85. The post-test probability for CAD was 0.75 for positive results and 0.43 for negative ones. With application of ML the authors achieved sensitivity 0.89, specificity 0.88 and the post-test probability 0.90 for positive and 0.25 for negative results.