All code (with instructions) is available on my GitHub: github.com/OliverDemuth... and on Zenodo (with additional data): doi.org/10.5281/zeno.... If you have any questions I am more than happy to help and explain the workflow/pipeline! 11/11

These simulations can not only be used to measure the lengths in experimental data, but can also constrain osteological ROM measurements to receive more functionally informative ROM estimates. Importantly, they are applicable to all type of joints and even fossils! 10/11

We calculated ligaments across the whole XROMM dataset and compared them to lengths measuared through dissection. It appears that different ligaments have different amounts of elasticity. Potential differences in (primary?) functions? 9/11

Similar to the osteological ROM simulation we used a signed distance field representation of the bones to speed up the optimisation process. The minimal length of each ligament was calculated from origin to insertion without any way points intersecting the bones 8/11

We were not only interested in the osteological range of motion (ROM) but also in the soft tissues constraints surrounding the joint. We simulated the ligaments of the shoulder capsule that prevent excessive movement while enabling contact between the bones during motion 7/11

Our approach is similar to work by Lee et al. (doi.org/10.1098/rspb...) but implemented in Autodesk Maya with Python. The optimisation approach minimises a cost function that determines joint proximity (cartilage thickness) and congruency (overlap) over a signed distance field 6/11

This made simulations more complicated, as previous pipelines often assumed a static joint centre around which the distal element (e.g., humerus) moves. However our approach automatically optimises joint translations for each rotational joint orientation and it is very fast! 5/11

In the shoulder we captured both joint rotations and translations. Our results clearly show that there is no (functional) joint centre in the bird shoulder and the humerus moves dynamically over the glenoid articular cartilage. The joint does not act as a spherical joint 4/11

We captured the joint motion of Red legged partridge specimens using XROMM (X-Ray reconstruction of moving morphology) and wiggled three specimens in the biplanar X-ray to capture their joint mobility 3/11

How do joints move? The articular surfaces between the proximal and distal bones are in contact during motion. However, they are usually not neatly spherical or hinge like but move in complex patterns that include sliding and/or rolling (e.g., in the shoulder joint of birds) 2/11

More information about the project STEPS can be found here:
www.arch.cam.ac.uk/news/ashleig...

#ERCStG

More information about the project STEPS can be found here: www.arch.cam.ac.uk/news/ashleig...

#ERCStG

Also asking it to create an image in the style of "paleoart" (see SI) is deeply insulting. Rather than chosing cheap AI slop one should always prioritise palaeoartists - with whom we have been collaborating for centuries. Let's not abandon our principles and our colleagues!

Generative AI is not science. It is the opposite of objective truth. It is misinformation dressed in a fancy package.