Towards Shoulder Injury Assessment for Safer Cycling: Describing Clavicle Variability and Mechanical Response

Licentiatavhandling, 2026

Cycling is both a key mode of sustainable urban mobility and a popular sport. Yet crashes expose cyclists to severe injuries and permanent medical impairment, while imposing substantial costs on healthcare systems. Clavicle fractures are among the most common skeletal injuries sustained in bicycle crashes and are associated with a risk of permanent medical impairment. Within a Ph.D. project aimed at developing methods and tools for shoulder injury risk evaluation in bicycle accident scenarios, this licentiate thesis presents two foundational steps toward that goal.

In the first study, a statistical shape model (SSM) of the right clavicle was derived from 97 clinical computed tomography (CT) scans using principal component analysis (PCA) to describe surface geometry and cortical bone thickness. Twenty-three principal components were required to explain 80% of the total morphological variance. Multivariate linear regression was used to link the model to sex, stature, age, and body mass index (BMI). The regression model did however only explain 25% of the total variability, demonstrating that gross anthropometric descriptors are weakly correlated with individual clavicle morphology and that population sampling is necessary to capture the full range of anatomical variation.

In the second study, a morphable finite element (FE) clavicle model was developed from, and linked to, the SSM and validated in axial compression and three-point bending against published experimental tests. Sampling from the full SSM variability generated synthetic populations whose simulated response corridors captured from 60 to 160% of the experimentally observed spread, establishing anatomical variability in geometry and cortical bone thickness as driver of inter-individual differences in mechanical response.

Accounting for morphological diversity is essential for developing and designing effective and robust protective countermeasures. The framework established in this thesis provides the methodological foundation for that population-informed approach to shoulder injury assessment and will aid the design of countermeasures protecting cyclists in bicycle crashes.