The main goal of this study was to build a finite element model for trabecular bone from micro-computed tomography (micro-CT) images and study elastic properties of bone. We also studied the relationship between Young’s modulus and porosity of trabecular bone. Researchers in the past have shown that porosity and apparent density were the two main factors that affect Young’s modulus of trabecular bone. However, experiments also showed that trabecular bone samples with similar porosity may have very different Young’s modulus. In this study, trabecular bone samples from porcine femoral head of 6 months old animals were scanned by Xradia micro-CT and finite element models were built and meshed by pre-FEA software ScanIP. Then, Abaqus software was used to conduct uniaxial compression test simulations and calculate Young’s modulus of the samples. Two different boundary condition settings were used during the simulations. In one of the settings, displacement was applied directly on the top surface of the bone model (Method 1), while in the other one, displacement was applied through a rigid plate parallel to the top surface (Method 2).

We compared the results from both computational methods with the experimental data we had on these samples and found that our FEA models generally accurately captured elastic properties of trabecular bone when using boundary condition setting 1. Method 2, however, gave results much stiffer than experimental data. We observed that porosity indeed is a major factor that contributes to Young’s modulus of trabecular bone. Models with higher porosity tended to have lower elastic modulus as described by former researchers. We also studied models with similar porosities that had a large range of elastic modulus values. Therefore, the technique of building FEA model from micro-CT images is suitable for porous structure such as trabecular bone. The models build in this study will be used in the future study to investigate mechanical properties of trabecular bone beyond the yield point and with account for time dependent properties.