Study on VR-Based Medical Image Deformation for Surgical Training System

As minimally invasive surgical techniques become widely available to patients, medical training systems based on virtual reality (VR) are highly desired. These systems help surgeon trainees to acquire, practice and evaluate their surgical skills. A key component in a VR-Based training system is to simulate the dynamics that occur in surgical procedures. Tissue deformation, as a most common phenomenon during surgery, has attracted many research efforts in computer simulation. In this paper, we propose an improved mass-spring model method aiming at realistic simulation of tissue deformation and reduced computational complexity. First, virtual springs are proposed which, along with the conventional mass-spring model, to represent the tissue surface. Second, Verlet integration is adopted to calculate the position of mass points during deformation without explicit computation of their velocities. Finally, a bilinear interpolation method is employed to generate a smooth mesh to render the deformed tissue surface. The proposed method has been implemented using OpenGL which generates realistic tissue deformation images in real-time.