Enhancing Shoulder Joint Biofidelity in the VIVA+ Human Body Model and Evaluating its Response

Abstract

Previous versions of the VIVA+ human body models have demonstrated excessive shoulder stiffness compared to experiments conducted on postmortem human subjects. This study aims to enhance the biofidelity of the shoulder in the VIVA+ model. Data from studies involving postmortem human subject experiments were extracted and applied to two separate models of the acromioclavicular and sternoclavicular joint. Different verification tests were implemented to ensure that the mechanical properties of the joints were implemented correctly. The data from the studies were also used to construct biomechanical response corridors. The model modifications were implemented and evaluated in LS-DYNA, and the corridors were constructed in Python using principal component analysis and Bayesian regression. Initially, a force-deflection corridor was intended, but limitations in the deflection data restricted the final corridor construction to the force-time response. Analysis of the corridors showed that the choice of principal components affected the physical reasonability. The final validation of the updated model involved comparing it with the previous iteration and the response corridors. By implementing the joints in an updated model in LS-DYNA and comparing the simulated VIVA+ response with these corridors, it was possible to assess how closely the updated model was to the experimental human response. The results showed that the updated model was less stiff than the original VIVA+ model, for all model types. The comparison to the corridors showed that the modified model was closer to the corridors at lower impact speed, while the response at higher speeds still differed from the experimental range. This indicates that the updated shoulder joints improved the response of the model, but further development is still needed to fully represent the biomechanical behaviour of the shoulder.