Research involving joint mechanics is typically conducted at the macroscopic level. However, joints and joint replacements often fail because load transfer at the microscopic level is not well understood. This gap in knowledge reduces our ability to preoperatively predict patient outcomes and assess irreversible failure modes for a variety of surgical interventions prior to clinical adoption. The present work aims to advance full-field experimental measurement techniques applied to better understand the internal load transfer of the human shoulder joint by simultaneously combining mechanical testing protocols, microCT imaging, and digital volume correlation (DVC) methods.

A CT-compatible loading apparatus was fabricated to allow for mechanical loading of cadaveric shoulder specimens within a cone beam microCT scanner. DVC was used to measure full-field displacements and strains throughout the internal structure of bone under controlled loading scenarios. Initially, the full-field experimental data was used to assess predictions generated by corresponding continuum-level finite element models (FEMs). Varying assumptions (e.g., boundary conditions, material mapping equation used, etc.) required to generate the simulations were assessed. The results of the validation efforts demonstrated that continuum-level FEMs of the shoulder can predict the experimental fullfield displacements with high accuracy if the boundary conditions are replicated correctly. Good agreement was found between the strains predicted and the experimental measurements obtained by DVC with the highest predictive errors found in locations that experienced the highest magnitude of experimental strain.

The full-field experimental methods were further applied to evaluate the magnitude of fullfield strain that trabecular bone within the shoulder can withstand prior to fracture. An experimental workflow which involved stepwise compressive loading with microCT images captured at each loading step was performed until macroscopic failure occurred. Internal strains throughout the trabecular structure were resolved using DVC. Bone density measurements and trabecular morphometric parameters were compared to outcome measures such as apparent strength and the local strain measured by DVC. The experimental data collected provides fundamental knowledge for future studies implementing DVC and lays the foundation for future validation studies that utilize fullfield experimental measures to assess the predictive accuracy of FEMs of the musculoskeletal system.