In vivo kinematical validated knee model for preclinical testing of total knee replacement.

Background And Objective: A computational knee model facilitates efficient component design evaluations and preclinical testing under various dynamic loadings. However, the development of a highly mimicked dynamic whole knee model with specified ligament constraints that provides high predictive accuracy with in-vivo experiments remains a challenge.

Methods: In the present study, a musculoskeletal integrated force-driven explicit finite-element knee model with tibiofemoral and patellofemoral joints constrained with detailed soft tissue was developed. A proportional-integral-derivative controller was concurrently added to the knee model to track the boundary conditions. The actuations of the quadriceps and hamstrings were predicted via a subject-specific musculoskeletal model and matched with electromyography results.

Results: Compared to in-vivo fluoroscopic results in a gait cycle, the predicted results of the kinematics of the tibiofemoral joint exhibited an agreement in terms of tendency and magnitude (anterior-posterior translation: RMSE = 1.1 mm, r = 0.87; inferior-superior translation: RMSE = 0.83 mm, r = 0.84; medial-lateral translation: RMSE = 0.82 mm, r = 0.05; flexion-extension rotation: RMSE = 0.23°, r = 1; internal-external rotation: RMSE = 1.85°, r = 0.65; varus-valgus rotation: RMSE = 1.39°, r = 0.08). Contact mechanics, including the contact area, pressure, and stress, were synchronously simulated on the tibiofemoral and patellofemoral joints.

Conclusions: The study provides a calibrated knee model and a kinematical validation approach that can be widely used in preclinical testing and knee prosthesis design.