New method for automatic generation of patient specific finite element models of the distal femur

Abstract

Three-dimensional (3-D) finite element (FE) modeling in orthopedics biomechanics has gained considerable interest in the past decades. In biomedical applications, the geometry of the bodies to be meshed is complex and manual generation would be a complex and time-consuming task. In the past few years, automatic mesh generators (AMG) algorithms have become widely available and these programs can automatically generate FE meshes of an object from its given geometric description. Most commercial AMG programs are accurate enough for biomechanical studies but these programs require computer-aided design (CAD) tools to first define the geometry. A new algorithm to automatically generate patient-specific FE models of human femur directly from magnetic resonance imaging (MRI) data is presented. The emphasis is given to the development of the meshing algorithm. To implement this algorithm, a set of MRI dataset of the human distal femur is used to form FE meshes. For model validation, the cortical bone was model with four-node shell elements while eight-node hexahedral elements and six-node wedge elements were used to model the cancellous bone. The bone was assumed to have homogeneous, linearly elastic and isotropic material properties throughout and was tested under simple load configurations. In the study of axial stresses, strains and von Mises stress distribution, all three models agreed well with the analyses reported by the other authors. Although only qualitatively reasonable, it is sufficed to validate the practicability of the meshing algorithm. Further work is required to validate the results quantitatively.