Collision Detection and Tissue Modeling in a VR-Simulator for Eye Surgery

This paper gives a survey of techniques for tissue interaction and discusses their application in the context of the intra-ocular training system EyeSi. As key interaction techniques collision detection and soft tissue modeling are identified. For collision detection in EyeSi, an enhanced image-based approach for collisions between deformable surfaces and rigid objects is presented. By exploiting the computing power of graphics processing units, it achieves higher performance than existing geometry-based approaches. Deformation vectors are computed and used for the biomechanical model. A mass-spring approach is shown to be powerful enough to bridge the gap between low computational demands and a convincing tissue behavior.