Skeleton Segmentation Of Coronary Artery Surface Mesh Based On Mean Curvature Flow

Coronary artery is an artery that supplies blood to the heart.Coronary heart disease is a disease caused by damage to the coronary artery,it gradually jumps to the main cause of death of the population as people's living standards improve.An angiography of the organ is required prior to the arterial surgery to predict in advance whether the patient has coronary heart disease.The main research of the thesis is the method of bone extraction in the surface of the coronary artery.The cardiac coronary grid method proposed in this paper consists of two parts: Skeleton extraction of the coronary artery grid and skeleton-based bone segmentation.Traditional skeleton extraction methods generally need to calculate the fractional flow reserve to judge the degree of stenosis of blood vessels.To do this,you need to extract the centerline of the coronary mask and then build the associated mesh model.However,errors in the calculation from the mask to the mesh result in large deviations between the centerline and the mesh model.To overcome this problem,the paper studies the method of extracting bone directly from the surface mesh.By studying three basic mesh skeleton extraction algorithms(corrosion method,Laplacian smoothing method,mean curvature flow method),this paper proposes an improved mean curvature flow method to extract a relatively good skeleton.After the skeleton is obtained,it is also necessary to decompose the skeleton into a single shape(skeletal segmentation),and obtain detailed left and right veins and individual skeletons.To this end,the paper proposes an algorithm to decompose the skeleton structure step by step from the root to the tail according to the morphological characteristics of the skeleton,so as to obtain the segmental bone morphology.The experimental results show that the proposed method is effective and can extract bones and segment them.