Collision detection algorithm based on AABB for Minimally Invasive Surgery

Abstract

Minimally Invasive Surgery (MIS) is a specialized surgical technique that can reduce patients' pain and permit a faster recovery compared with conventional open surgery. But surgeons require extensive training to obtain the expertise due to its complexity. In the previous research, we have developed a virtual reality system to train surgeons to obtain the expertise to correctly perform the interventions. In the system we have reconstructed 3D vessel model based on CT images and designed a 3D catheter model based on finite element method. In this paper, in order to provide the surgeon with a sense of touch, we work on force feedback and collision detection algorithm. We design a force feedback which contains three kinds of force: viscous force, frictional force and collision force. The viscous force stands for the force between the catheter and blood; the frictional force means the force between catheter body and vessels; the collision force is the force between the catheter tip and vessel. Then we develop a fast collision detection algorithm based on AABB (Axis-Aligned Bounding Boxes) method. In the algorithm, there are two key areas: Area I - the area around the catheter tip and Area II - the previous collision area. Area I is a dynamic area and it will change with the position of the catheter tip. Area II is some static areas but the amount of the areas is changeable. By this way, we can reduce the collision detection time and achieve real time training.