Abstract:
This study introduces the Surrogate Contact Modeling Toolbox (SCMT), an open-source software package designed to simplify and accelerate the development of surrogate models of elastic joint contact for biomechanical simulations. These surrogate models provide a fast alternative to computationally expensive finite element (FE) contact models, making them suitable for iterative musculoskeletal simulations. SCMT interfaces with FEBio for FE analyses and MATLAB for training artificial neural networks, ultimately enabling the creation of fast, accurate, and reusable surrogate contact models. The toolbox features a structured GUI-driven workflow including input space sampling, filtering, model training, and integration with OpenSim. Two detailed example applications are presented: first, the generation and validation of surrogate contact models for artificial tibiofemoral and patellofemoral joints; second, the use of these models within a 12-degree-of-freedom OpenSim simulation of seated knee extension-flexion incorporating ligaments and muscle actuation. SCMT significantly reduces the effort needed to incorporate detailed joint contact models into biomechanical analyses and represents a powerful tool for advancing biomechanical research involving joint mechanics, muscle-ligament interactions, and surgical or rehabilitation planning.
