Evaluating model-estimated shoulder muscle activity during overhead work with varied task demands and exoskeleton use

Area of Science:

• Biomechanics
• Ergonomics
• Human-Computer Interaction

Background:

• Passive arm-support exoskeletons (ASEs) can mitigate shoulder stress during overhead tasks.
• Current evaluation methods for ASEs are time-consuming and resource-intensive.
• Musculoskeletal modeling offers a potential alternative for evaluating ASEs by estimating muscle activation.

Purpose of the Study:

• To assess the accuracy of a commercial biomechanical modeling system in estimating shoulder muscle activity during overhead push tasks.
• To compare model-estimated muscle activity with electromyographic (EMG) data with and without an ASE.
• To determine the influence of task height and force direction on the model's performance.

Main Methods:

• A commercial biomechanical modeling system was used to estimate shoulder muscle activities.
• Dynamic overhead push tasks were performed at varying heights and force directions, with and without an ASE.
• Model estimates were compared to normalized EMG data using pattern similarity and magnitude difference metrics.

Main Results:

• The biomechanical model demonstrated good performance at lower task heights.
• Model accuracy decreased at higher task heights.
• ASE use further reduced model performance, particularly at higher task heights.

Conclusions:

• Musculoskeletal model-estimated shoulder muscle activity can be reasonably accurate for specific overhead tasks.
• Improvements in musculoskeletal models are necessary for broader applicability in ASE evaluation.
• The study highlights the need for refined modeling techniques to account for task complexity and exoskeleton interaction.