Cardiac Surgery Simulation with Active Interaction and Adaptive Physics-Based Modeling

Computer graphics (CG) and virtual reality (VR) are increasing their possibility in the medical field. Most of current clinical applications for diagnosis or surgical planning consist of the developed CG and VR techniques. On the other hand, surgeons make great efforts in learning surgical procedures and training of their skills. Their total loads are increasing due to the development of new surgical procedures produced by recent medical and engineering techniques. Careful planning is a key for successful surgery. Nevertheless, no information of surgical fields is available in current preoperative discussion. To learn surgical skills efficiently and to prevent unsuccessful cases, surgeons strongly desire more realistic planning and rehearsal of surgical procedures in the preoperative stage. This paper aims to establish a new approach to support planning of patient-specific surgical strategy using virtual reality based simulation. So far, numerous computer-assisted systems have been developed for improving preoperative planning and procedural training. However, foregoing systems are not applicable to simulate whole process of the procedures interactively. In order to concrete the proposed approach, the authors first analyze actual surgical procedures and design a desktop surgery simulation system ActiveHeart, which provides preoperative rehearsal of the following essential two-step procedures in the field of cardiovascular surgery. • Construction of surgical fields • Identification of 3D region of interest Next, the authors present an integrated simulation framework that enables haptic interaction and physics-based estimation in rehearsing surgical procedures with virtual patient models. For providing this framework, several new methods listed as below are proposed and evaluated. Active interaction with beating virtual organs is proposed, and essential factors for visual and haptic feedback of heartbeats are extracted. The authors focus on time series dataset and provide constraint-based haptic ii rendering and stress-pressure model. Characteristic experiments with cardiovascular surgeons shows both 3D shape with motion and time series internal pressure efficiently simulate realistic haptic feedback reflecting autonomous motion of cardiac muscle. Considering above results of the experiments, a deformable beating aorta model based on finite element method is proposed. Using the proposed model, the authors construct an interactive palpation environment that enables surgeons to pinch virtual beating aorta of normal and sclerosis cases. Several experiments show the developed system is useful to recognize the normal stiffness of the virtual aorta model. The haptic feedback enables grasping sclerosis regions for rehearsal of surgical palpation. In order to simulate constructing surgical fields interactively and accurately, an adaptive physics-based modeling …