Realtime Cardiac Surgery Simulation

This thesis is an investigation of the possibilities of realtime surgical simulation, specifically within the domain of congenital heart diseases. The problem domain of surgery on the heart of an infant is presented and the primary information necessary for a pre-operative surgical simulation tool is identified. Through cooperation with surgeons, domain specific abstractions and simplifications with respect to soft tissue simulation is identified and used in the investigation of models for tissue deformation. In general it is not possible to calculate the deformation of tissue precisely in realtime. We therefore need to use simplified elastic models that approximate the real behavior of tissue. The most popular models for realtime interaction and response; the Finite Element model and the Spring Mass model are presented and discussed with respect to to surgery on the heart of an infant. A number of variations and extensions of the classic Spring Mass model are investigated to identify the characteristics that can use the domain specific knowledge of tissue deformation and interaction in cardiac surgery. An elastic model and a geometrical model build specifically for cardiac surgery simulation is developed and presented in this thesis. The thesis is done in cooperation with surgeons from Århus University hospital. A surgical simulator supporting the techniques discussed in the thesis has been evaluated with surgeons for pre-operative use. Furthermore the elasticmodels under investigation have been compared and validated formally with respect to their behavior of deformation.