Validation of an automated method of three-dimensional finite element modelling of bone

This study validated an automated method of finite element modelling of bone from CT scan data. After a fresh-frozen cadaveric femur was modelled, strain gauges were attached to the bone at 11 locations and the femur was mechanically tested by applying a load to the femoral head. Linear regression analysis was used to correlate the strains predicted by the model with the experimentally measured strains. The regression results were significant (P < 0.001), indicating that the strain calculated by the FE model is a valid predictor of the measured strain. Verification of the surface strains also supports the validity of the strains and stresses predicted inside the bone. The present study provides a strong rationale for use of this modelling method as a research tool and in possible clinical applications.