Investigations into finite element modelling aspects of the human head

Fig. 1: Original Polygonal skull, contours with sections used for meshing, and finally the finite element model with brain visible. A set of Computed Tomography and Magnetic Resonance Imaging data files has been determined and are available through the Visible Human Database (NIH, 2002). A polygonal representation of the skull was constructed and from this the finite element model was formed. Contours were extracted from the polygonal model and then sections were carefully chosen (Fig. 1) across these contours in order for the model to have good hexahedral element quality and ease of mesh generation. A number of variations of the model were created. The model was skewed in such a way so as to represent the same size head as that used in Nahum’s experiment (1977). The resulting model was validated against the cadaver tests of Nahum, with good agreement being found for pressure response. A parametric High/Low study was performed, which consisted of 16 tests, to investigate the effect of the bulk and shear modulus of the brain and cerebrospinal fluid (CSF). The range of values for the bulk modulus and shear modulus of the brain chosen were 200MPa20GPa and 20kPa-500kPa respectively, with 0.25MPa-2.5GPa and 0.5kPa-50kPa for the CSF. The influence of different mesh densities on the models and the use of different element formulations to model the skull were also investigated. Mesh densities were varied from approximately 9,000 elements to 60,000 elements overall, limited only by computing resources. The skull was modelled as (a) a layered brick element skull, with the middle two bricks assigned to be trabecular bone, (b) a composite shell element skull (with and without a defined variance of thickness) and (c) as a ‘shell-brick’ element type skull where the shell elements were cortical bone and the brick elements were trabecular bone.