Detection of coronary artery stenosis using artificial intelligence methods

Coronary heart disease is a common cardiovascular disease, which is one of the causes of high mortality in the population. Accurate and timely diagnosis is crucial for treatment. In this study, a new algorithm for detecting coronary stenosis of the heart vessels was developed based on the YOLO8m detection model. A set of X–ray images obtained by selection from video sequences of the results of performed angiography of the cardiac vessels was used to train the model. The proposed model demonstrated excellent processing speed and detection accuracy and achieved F1–score detection accuracy of 98% and mAP=98.3%, surpassing a number of known results obtained by other researchers. The proposed stenosis detection algorithm represents a significant achievement in the field of cardiovascular imaging and computer image processing algorithms.