High Precision Implicit Modeling for Patient-Specific Coronary Arteries

High precision geometric reconstruction of patient-specific coronary arteries plays a crucial role in visual diagnosis, treatment decision-making, and the evaluation of the therapeutic effect of interventions in coronary artery diseases. It is also a fundamental task and a basic requirement in the numerical simulation of coronary blood flow dynamics. In this paper, a new implicit modeling technique for the geometric reconstruction of patient-specific coronary arteries has been developed. In the proposed method, the coronary arteries geometry is reconstructed segment by segment using radial basis functions with ellipsoid constraint from the point cloud obtained with a volumetric vascular image segmentation method, and the individually reconstructed coronary branches are then combined using a shape-preserving implicit blending operation to form a complete coronary artery surface. The experiment results and validations indicate that the reconstructed vascular shapes are of high smoothness and faithfulness.