The establishment of finite element models to characterize mechanical properties of mouse tibiae based on micro-CT data

Abstract

FEA, based on a large amount of data from microscale image scanning and immense computing power, has shown great potential to simulate bone mechanics. This research intended to establish and validate FE models based on micro-CT data of mouse tibia to characterize the biomechanical properties.

The general steps for establishing an FE model include importing micro-CT data into Mimics software to create a 3D geometric model; creating 2D and 3D meshes based on a geometric model using 3-Matic software and Hypermesh software, respectively; assigning material properties using Mimics software; applying boundary conditions using Hypermesh software; and performing final computation as well as outcome parameter visualization using Abaqus software.

Two types of FE models of axial loading (whole tibia model and proximal tibia model) were established and validated by correlating the calculated and experimentally measured morphological and experimental parameters. However, the correlations were not significant. Later, bending stiffness was calculated and validated for the FE models simulating a three-point bending setup. The calculated and experimentally measured bending stiffness correlated positively for homogeneous (p < 0.001, R2 = 0.5852) and heterogeneous (p < 0.001, R2 = 0.9482) material assignment. However, the FE model with homogeneous material assignment significantly underestimated bending stiffness.

In conclusion, the FE model simulating three-point bending with heterogeneous material assignment based on micro-CT data of mouse tibiae showed good accuracy to simulate linear elastic behavior of mouse tibia under a three-point bending setup. Besides, this FE model, in combination with in vivo micro-CT scanning, demonstrates great potential as an alternative in silico approach to reduce the number of animals needed for experiments and to save experimental time as well as costs.