Virtual surgery simulation for medical training using multi-resolution organ models

Abstract

Background: Real-time simulation of organ deformation is one of the biggest challenges in virtual surgery, due to the conflicting requirements of real-time interactivity and simulation realism. In this paper we propose a method to overcome this challenge by introducing a multi-resolution modelling technique.

Methods: In our approach a reasonably coarse global model is locally enhanced, using a mesh subdivision and smoothing algorithm. The global model is based on a discretization of the boundary integral representation of three-dimensional deformable objects. Local refinements are provided at the tool-tissue interaction region by a local subdivision technique.

Results: As an example, we have developed a deformable human kidney model generated from the Visible Human Dataset, with tissue properties determined from in vivo animal experiments. A mixed reality laparoscopic surgical training system has been developed, using an abdominal mannequin and force feedback devices.

Conclusions: The use of precomputation and structural re-analysis techniques results in a very rapid computation procedure. Validation of the simulator is in progress.