Deciphering the "Art" in Modeling and Simulation of the Knee Joint: Assessing Model Calibration Workflows and...

Area of Science:

• Biomechanics
• Orthopedics
• Computational Modeling

Background:

• Reproducibility is a key concern for the National Institutes of Health (NIH) and scientific research.
• Computational modeling is increasingly used in biomechanics and orthopedics.
• Assessing the reproducibility of modeling workflows and simulation predictions is essential.

Purpose of the Study:

• To investigate the influence of subjective modeling decisions on the reproducibility and predictive uncertainty of computational knee joint models.
• To compare different model calibration approaches and outcomes among expert teams.

Main Methods:

• Five expert teams calibrated computational knee joint models using the same specimen-specific joint mechanics dataset.
• Investigated differences in calibration targets, workflows, and modeling decisions.
• Quantified variations in post-calibration ligament properties and predicted kinematics.

Main Results:

• Significant differences were observed in the selection of calibration targets and modeling approaches among teams.
• Calibration improved model performance, but variations in ligament properties and predicted kinematics persisted.
• Model calibration proved difficult to plan and foresee in detail, even for experienced teams.

Conclusions:

• Subjective modeling decisions, or the modeler's 'art', significantly impact the reproducibility and predictive uncertainty of computational knee models.
• Standardization of best practices for data sharing and calibration is crucial for improving reliability in biomechanical modeling.
• Further research is needed to identify and implement standardized protocols for computational modeling in orthopedics.