Abstract
Femoral fractures, often resulting from high-energy accidents, are commonly treated with intramedullary nails. Most currently used intramedullary nails are designed based on the anatomical characteristics of the Caucasian population, which are unsuitable for Asian populations due to differences in femoral medullary canal curvature. This mismatch causes stress in the femur, which leads to nail tip penetration. This research proposes a novel multi-femur shape optimization approach to design a long intramedullary nail that reduces stress on the femur by better conforming to the Thai femoral medullary canal. 60 CT scan images of Thai femurs were analyzed, and ten representative femurs were selected to cover the geometric diversity of the femur population using k-means clustering. The nail curve was modeled using four circular arcs, with radii optimized within a 700-1200 mm boundary. The nails were simultaneously optimized in the ten representative femurs utilizing a multi-objective genetic algorithm (MOGA) in Ansys, with the objective of minimizing the maximum principal stress in all ten femurs. The multi-femur shape optimized nail can reduce stresses in the representative femurs. Finite element analysis was conducted to compare the stresses induced by the multi-femur shape optimized nail with the stress caused by the commercial nails with radii of 1275 mm, 1500 mm, and 2000 mm. The results indicate that the multi-femur shape optimized nail significantly reduced the average maximum principal stress in randomly selected femurs. This research demonstrates that a multi-femur shape optimization approach can effectively design an intramedullary nail that reduces stress in a diverse set of femurs, making it a better fit for the Thai femur population. The optimized nail design was further evaluated in a verification model, including lag/distal screws and a one-leg-stance load; the stress in the multi-femur shape optimized nail was substantially lower than the commercial designs.