Development of an Animal Brain FE model

Abstract

During this thesis a rat finite element (FE) model was developed and the relative methodology to operate was presented. The new rat brain FE model was developed using image processing techniques by extracting the geometry and mesh from magnetic resonance imaging (MRI). A new rat brain model was needed due to several limitations present in previous models Baumgartner D. et al (2009). Finite element simulations using a more detailed model were required to improve the understanding of the mechanisms responsible for injuries in the brain. The rat brain model developed in this thesis will be used in the future to simulate animal experiments carried out by Davidsson J. et al (2009). A method to get the geometry and the final model was developed in order to have the highest possible level of repeatability. Several limitations were found during the geometry generation due to the low quality of T1- weighted MR images and to the lack of T2- weighted MRI and computerized tomography (CT) scan. Also, a low control on the mesh generation process led to an excessive element density and tetrahedral elements concentration. A large amount of time was spent on the segmentation process having to use a wide range of methods with increasing user interaction. This has led to a low level of repeatability of the process. Several evaluations were made during this work, such as on original data, geometry result, mesh and simulations result. A correlation was found between original data (i.e. MRI, CT-scan) quality and simulation result accuracy.