Evaluation of predicted knee function for component malrotation in total knee arthroplasty

Soft-tissue balancing for total knee arthroplasty (TKA) remains subjective and highly dependent on surgical expertise. Pre-operative planning may support the clinician in taking decisions by integrating subject-specific computer models that predict functional outcome. However, validation of these models is essential before they can be applied in clinical practice. The aim of this study was to evaluate a knee modelling workflow by comparing experimental cadaveric measures to model-based kinematics and ligament length changes. Subject-specific models for three cadaveric knees were constructed from medical images. The implanted knees were mounted onto a mechanical rig to perform squatting, measuring kinematics and ligament length changes with optical markers and extensometers. Coronal malrotation was introduced using tibial inserts with a built-in slope. The model output agreed well with the experiment in all alignment conditions. Kinematic behaviour showed an average RMSE of less than 2.7 mm and 2.3° for translations and rotations. The average RMSE was below 2.5% for all ligaments. These results show that the presented model can quantitatively predict subject-specific knee behaviour following TKA, allowing evaluation of implant alignment in terms of kinematics and ligament length changes. In future work, the model will be used to evaluate subject-specific implant position based on ligament behaviour.