Hybrid 4D cardiovascular modeling based on patient-specific clinical images for real-time PCI surgery simulation

In this paper, we proposed a hybrid 4D cardiovascular modeling method based on the clinical CTA and MRI to simulate the motion of the heart, as well as the patient-specific vessel structure attached to the heart model, in order to support the personalized Percutaneous Coronary Intervention (PCI) simulation which is used to train surgeons of skills and to help planning surgery. To obtain patient-specific vessel structure, a coarse segmentation with the centerlines extraction subsequently is applied to the computed tomography (CT) scans and the vessels along the centerlines is modeled using a lofted 2D segmentation method. The vessels are then combined with a template heart model to construct a cardiovascular system. For the cardiac motion, we estimate the ventricles motion from 4D Magnetic Resonance Imaging (MRI) sequences to drive the whole heart motion. And the position-based method coupling with a mass-spring model constructed with elastic spheres is used to simulate the cardiac motion cycle stably in the interactive PCI simulator. With our method, a personalized highly realistic beating motion of a whole heart is able to be created and applied to our patient-specific PCI surgery simulation system.