Collision handling of deformable anatomical models for real-time surgery simulation

Abstract

Efficient collision handling of dynamically deforming anatomical models is a very challenging problem in real-time surgery simulation. In this paper, we present a new, efficient approach that detects collisions and self-collisions of deformable models. Our algorithm is integrated in a simulation framework for multiple interacting objects. This framework consists of three components. The first one computes the dynamic behavior of deformable structures using an efficient and robust approach with distance- and volume-preserving constraints. The second component detects collisions and self-collisions of deformable models. A novel image-space technique is employed that detects volumetric intersections. This allows to compute the penetration depth which is used to resolve collisions in the third component of our framework. Our system handles volumetric deformable models of up to several thousand surface triangles and tetrahedra in real-time and can be used in surgical training systems.