Research on automatic coronary artery extraction algorithm based on CTA image and model guidance

When doctors diagnose coronary artery by CTA data, they need to look at each section layer by layer, which consumes a lot of diagnostic time. Introducing artificial intelligence technology, using CTA image computer-aided technology to extract coronary artery automatically or semi-automatically, and assisting analysis of coronary artery, can improve the efficiency of doctors, quickly make diagnosis of patients. A priori model of coronary artery centerline based on image registration is studied. A single priori model is formed by fusing multiple cardiac CTA data, which increases the generalization ability of the model and provides more accurate priori position information of main branches for unknown CTA data. According to the composition of the model, it will be divided into two parts: the cardiac cavity model and the centerline model. The implementation of the model establishment, DMP centerline extraction and coronary analysis is introduced. The methods of central line tracing were discussed, and the clinical application of central line extraction was divided into 1) radius extraction 2) calcification and stenosis analysis. Coronary CTA (Computed Tomography Angiography) technology can not only obtain the image of coronary artery, but also image the wall of coronary artery, assisting doctors in the analysis of vascular calcification and stenosis. Coronary CTA technology adds a diagnostic approach to cardiovascular disease and makes it more acceptable to patients because of the use of non-interventional techniques.