Benchmarking the predictive capability of human gait simulations

Physics-based simulation generate movement patterns based on a neuro-musculoskeletal model without relying on experimental movement data, offering a powerful approach to study how neuro-musculoskeletal properties shape locomotion. Yet, simulated gait patterns and metabolic powers do not always agree with experiments, pointing to modeling errors reflecting gaps in our understanding. Here, we systematically evaluated the predictive capability of simulations based on a 3D musculoskeletal model to predict gait mechanics, muscle activity and metabolic power across gait conditions. We simulated the effect of adding mass to body segments, variations in walking speed, incline walking, crouched walking. We chose tasks that are straightforward to model, ensuring that prediction errors stem from shortcomings in the neuro-musculoskeletal model.

The simulations predicted stride frequency and walking kinematic with reasonable accuracy but underestimated variation in metabolic power across conditions. In particular, they underestimated changes in metabolic power with respect to level walking in tasks requiring substantial positive mechanical work, such as incline walking (27% underestimation). We identified two possible errors in simulated metabolic power.

First, the Hill-type muscle model and phenomenological metabolic power model produced high maximal mechanical efficiency (average 0.58) during concentric contractions, compared to the observed 0.2–0.3 in laboratory experiments. Second, when we multiplied the mechanical work with more realistic estimates of mechanical efficiency (i.e. 0.25), simulations overestimated the metabolic power by 84%. This suggests that positive work by muscle fibers was overestimated in the simulations. This overestimation may be caused by several assumptions and errors in the musculoskeletal model including its interacting with the environment and/or its many parameters.

This study highlights the need for more accurate models of muscle mechanics, energetics, and passive elastic structures to improve the realism of human movement simulations. Validating simulations across a broad range of conditions is important to pinpoint shortcomings in neuro-musculoskeletal modeling.

Author summary: (non-technical summary of the work)

Our research focuses on understanding how humans walk by using computer simulations. These simulations are based on detailed models, i.e. mathematical descriptions, of skeleton, muscles, joints, and control system. By comparing our simulations to actual experiments where people walked under different conditions—such as carrying extra weight, walking faster or slower, or moving uphill or downhill—we evaluated how well the simulations could predict real-life movement and energy use.

We found that while the simulations performed well in predicting the walking pattern, they underestimated metabolic energy used by the body, especially in tasks like walking uphill. Errors in simulated metabolic power likely stem from two issues. First, the metabolic power model resulted in unrealistically high mechanical efficiency compared to experiment. Second, positive work (and as a result also net negative work) by muscle fibers was overestimated in the simulations.

These findings highlight the need to improve the models so they can more accurately reflect the complexity of human movement and energy use. Ultimately, better models will help us design devices like exoskeletons and prosthetics and improve treatments for people with movement difficulties.