Simulation of osteotomy in total knee arthroplasty with femoral extra-articular deformity assisted by artificial intelligence: a study based on three-dimensional models

Abstract

Background The impact of extra-articular deformities (EADs) on lower limb alignment and collateral ligament integrity during total knee arthroplasty (TKA) poses significant challenges, increasing surgical complexity. Our study aims to evaluate the influence of EADs on mechanical axis alignment and the risk of collateral ligament injury during TKA using an AI-assisted surgical planning system, with the goal of minimizing ligament damage through precise and scientific planning. Methods A healthy volunteer underwent CT and MRI scans of the lower limbs. The scan images were imported into Mimics 20.0 software, and the reconstructed models were spatially aligned using 3-maticResearch 11.0 software. Using Unigraphics NX9.0 software, 50 three-dimensional models of femoral lateral joint deformities with varying positions and angles were created. Finally, TKA was simulated using the AI JOINT preoperative planning system. Results The larger the deformity angle and the closer it is to the knee joint, the more pronounced the deviation of the mechanical axis. During MA-aligned osteotomy, nine types of deformities can damage the collateral ligaments. After adjusting the varus/valgus of the prosthesis within a safe range of 3° and leaving a residual 3° varus/valgus in the lower limb alignment, only the 25° varus and 25° valgus deformities located at 90% of the femoral anatomical axis remain uncorrected. Conclusion For patients with osteoarthritis and concurrent EAD undergoing TKA, using reconstructed 3D models of the collateral ligaments for preoperative planning helps visually assess collateral ligament damage, providing a practical solution. Minimizing intra-articular osteotomies within a safe range and allowing some residual alignment deviation can reduce the risk of collateral ligament injury.