Accurate assessment of bone morphology is critical for understanding orthopedic trauma patterns, optimizing implant design, and documenting demographic variation. Traditional methods are limited by cost, access, or reproducibility, particularly for quantifying bone curvature. We introduce a novel methodology using open-source software to analyze clavicle morphology, including length, inflection point, and medial/lateral radius of curvature, modeled mathematically to enhance accessibility and precision.
Methods:Clavicle morphology was assessed from both two-dimensional (2D) photographs of physical specimens (n=10) and 3D computed tomography (CT) scans (n=82) from the Maxwell Museum of Anthropology. 3D models were processed into 2D projections using Meshmixer. Images were analyzed in Fiji (ImageJ) to generate skeletonized centerlines and extract X-Y coordinates. Polynomial modeling in Excel allowed calculation of curvature parameters. We identified the inflection point, radius of curvature of the medial arc, radius of curvature of the lateral arc, and total length of the clavicle. The inflection point is the transition point between the medial and the lateral arc. Validation was performed by comparing measurements from a subset of clavicles against models constructed using SolidWorks software. Inter-rater reliability was assessed among three independent researchers.
Results:Clavicular curvature was consistently modeled as a fourth-degree polynomial. That is, the height of the curve, y, was described by the equation y=ax4+bx3+cx2+dx+e, where x is the distance along the clavicle. We solved for the coefficients a, b ,c, d, and e. The average R² was 0.997. This R2 indicates an excellent fit. Measurements using the novel method showed excellent agreement with SolidWorks-derived values for length, inflection point location, and curvature parameters. Inter-rater reliability was high across independent researchers.
Conclusion:This study presents a reproducible approach for quantifying clavicle curvature using no-cost, widely available tools. Validation against industry-standard software supports the accuracy of this method, and its flexibility across initial image types enhances its utility. Limitations include variability in initial specimen orientation and potential future changes to software platforms. Nevertheless, this methodology broadens access to morphometric analysis and offers applications in orthopedic trauma research, implant design, and anatomical sciences. We expect to apply this approach to the morphology of long bones, such as metacarpals and bones of the forearm.
