Deciphering the "Art" in Modeling and Simulation of the Knee Joint: Model Benchmarking.

Area of Science:

• Orthopaedic biomechanics
• Computational modeling and simulation
• Scientific reproducibility

Background:

• Computational models in orthopaedics aid understanding of joint mechanics and clinical decisions.
• Reproducibility in scientific modeling is vital for data sharing and responsible resource use.
• The KneeHub project explored variability in knee model development by multiple teams.

Purpose of the Study:

• To evaluate the predictive accuracy of calibrated knee models under novel loading scenarios.
• To assess model performance in simulating anterior cruciate ligament (ACL) resection and pivot shift loading.
• To identify limitations and sources of error in computational knee models.

Main Methods:

• Five independent teams developed and calibrated knee models using identical experimental data.
• Models were benchmarked against experimental data for scenarios not used in calibration (ACL resection, pivot shift).
• Model predictions of kinematics and kinetics were compared to experimental measurements.

Main Results:

• Models captured major trends in knee joint kinematics and kinetics but showed inter-team variability and discrepancies with experimental data.
• Internal-external rotation predictions exhibited the largest errors.
• Optimized ligament properties in models did not fully replicate experimental joint kinematics under benchmark conditions.

Conclusions:

• While calibrated knee models predict general biomechanical trends, their accuracy is limited, particularly for rotational movements.
• Benchmarking is essential for quantifying model limitations and improving the credibility and reproducibility of orthopaedic simulations.
• Establishing best practices for computational modeling is crucial for advancing the field.