3D Modelling Of Muscle Attachement Sites: The Clavicle

Background Clavicle anatomy is highly variable. This large variation is an important reason why current clavicle fracture plates don’t fit optimally to the bone. Soft tissue interference of prominent plates requires implant removal and this might also be to interference of the plates with the muscle attachments. Currently there is almost no literature which describes the muscle topography of the clavicle. Objectives Our primary objective is to develop a new 3D modelling technique which takes muscle attachment sites into account. And our secondary objective is to evaluate the differences in surface area of the different muscle attachment sites for the clavicle. Study Design and Methods We dissected 28 clavicles out of 14 human cadavers donated by the human body donation program of the university of Leuven (KU Leuven, Belgium). All soft tissues (fat, ligaments, blood vessels,…) apart from the muscles were meticulously removed from the cadavers until only the muscles remained. The specimens were then CT-scanned and the bony anatomy was reconstructed using image segmentation (Mimics, Materialise, Belgium). Then the models were optically laser scanned using a ultra-high surface scanner (Nikon, Nikon Metrology) while at the same time over 200 photos were taken to create photogrammetry models (Agisoft, Russia) which provided accurate surface information. These models were then uploaded as STL files in 3-matic (Materialise, Belgium) and as 3D pdf files. By subtracting the bone models from the muscle surface models and manually improving them using the photogrammetry models we obtained models of the different surface sizes. Results We obtained 28 high quality 3D models that could be used for more advanced analysis. We found no statistically significant length difference between left (153 +/- 12cm) and right (151 +/- 12cm) clavicles (p=0.734). There was no statistically significant length differences between male (155,8 +/- 10,8 cm) and female (148,6 +/- 12,2 cm) clavicles (p=0,114). Mean relative surface sizes ([Surface of the muscle attachment site/Total surface of the clavicle] x 100) for the Deltoid, Trapezoid, Pectoral, Sternocleidomastoid and Subclavian muscles are respectively 7,3 +/- 1,5% (Deltoid); 11 +/- 2,7% (Trapezoid); 7,4 +/- 2,2% (Pectoral); 5,9 +/- 2,2% (Sternocleidomastoid) and 5 +/- 1,9% (Subclavian). There was no statistically significant difference regarding the relative muscle surfaces between men and women or between left and right. We did see a significant difference between absolute muscle attachment surfaces men (3663 +/- 422cm²) and women (3148 +/- 453 cm²) (p=0,004). Conclusions We developed a new anatomic registration method which can be used to create 3D anatomical models that take muscle attachment size into account. These models could be used for more advanced biomechanical analysis methods. Furthermore, we are the first study that has taken the muscle attachment areas of the clavicle into account and hereby provide the first data towards muscle attachment size. Further research is needed to imply these results in to a clinically valuable model.status: accepte