A New Deformation Model of Brain Tissues for Neurosurgical Simulation

An accurate and realistic brain tissue deformation model with real-time performance is very important for virtual neurosurgical simulation. In this paper, a new finite element method (FEM) brain tissue deformation model, which is based on the optimization implicit Euler method, is introduced. Biomechanical properties of brain tissue such as anisotropy and viscoelasticity are incorporated into the model, which provides more accurate and realistic imitation of the deformation of brain tissue. A descent method with GPU-based implementation is used to solve the optimization problem, which makes it possible to achieve a high degree of computational efficiency. Simulation results show that both the anisotropic and viscoelastic behaviors are presented in the deformation model. The GPU-based implementation of the proposed model significantly improves the computational efficiency over CPU-based FEM models with the implicit integration scheme. Moreover, the result of the proposed model converges to the exact solution of implicit Euler integration after 96 iterations. The proposed model was implemented on the development of a neurosurgical simulator. A relative high degree of realistic brain tissue deformation was rendered at a refreshment rate of 32.5 frames/s on a regular PC.