AI-Based Early Detection of Coronary Artery Disease Using Atherosclerosis Inflammatory and MicroRNA Biomarkers in Angina Patients

Coronary artery disease (CAD) is one of the most common causes of sudden cardiac arrest and is responsible for a high percentage of global mortality. Early diagnosis of CAD in patients with chest pain is challenging. Coronary angiography, an invasive procedure that is not widely available in developing countries, confirms the diagnosis of CAD. It is imperative to establish noninvasive biomarkers for the early diagnosis of CAD in patients with angina. Motivated by two recent medical diagnostic studies that used novel molecular (microRNA) and chemical (atherosclerosis-inflammatory) biomarkers for early detection of CAD, the proposed study explores the hidden potential of novel biomarkers using machine learning (ML). The study combines the biomarkers of two medical studies and presents a NUMS-NIHD dataset. From the complete dataset, the best feature set is shortlisted after an extensive evaluative search in the biomarker feature space. Ten ML classifiers are applied to the selected biomarker inputs, including clinical, chemical, and molecular cardiac biomarkers. To select optimal biomarker-classifier combinations, experiments use K-fold cross-validation. The results of K-fold cross-validation are validated against a hold-out set of data. The results are then compared to current state-of-the-art benchmark studies. The detection strategy achieved an accuracy of 90.18%, a sensitivity of 92.33 % and a specificity of 100.00%. The proposed study validated previous medical diagnostic studies and significantly improved the results. These ML biomarkers-classifier combinations can be used to exclude healthy individuals non-invasively before angiography.