Simulation of tetrahedral mesh based organ deformation using parameter optimized spring constants

Virtual reality based surgical simulators are becoming increasingly popular for training doctors on minimally invasive surgery. For these simulators, it is necessary to generate 3D organ models and create virtual environments depicting visually realistic deformation response by tissues and organs to manipulations by surgical tools. For creating a 3D model, several different approaches have been used by the computer graphics community, among which polygonal/polyhedral meshes have gained popularity for real time virtual reality applications.;This work investigates two problems in 3D organ modeling for virtual reality simulators. Firstly, we consider the use of unstructured 3D meshes for generation of organ models. We investigate a class of algorithms for simplification of meshes, that is, for reduction in the number of elements, and also for improving the quality of generated mesh. We propose a new algorithm for mesh simplification by building up on existing approaches. This algorithm has better and quicker simplification performance. At the same time we ensure that the quality of simplified mesh elements is not degraded beyond a certain user-specified limit. We compare the results obtained by our algorithm to existing rapid procedures.;The second problem investigated in this work deals with mass-spring-damper models, as applied to the tetrahedral mesh for modeling the physically deformable nature of the organs. We extend the work reported by Wang and Devarajan in 2007 [1] and apply the suggested parameter optimization to the more general case of irregularly shaped organs with a higher deformation tendency. We optimize the Hooke's constant for the springs used in the Mass-Spring-Damper (MSD) model in the mesh based on some constraints to generate a more realistic and visually appealing physically-based deformation response. We set up a stand-alone framework for modeling the deformation response of MSD model based organs and determine all relevant environment parameters for a stable simulation.