Cheetah anatomy: going digital is not enough
In their comments to a previous post, a couple of friends called my attention to an illustration of cheetah anatomy published in the November issue of NG magazine. I had missed that issue when it appeared, but have caught up now. These friends point to some anatomical errors in that representation, including among the most serious the following: 1) the lumbar vertebrae are in reverse (front side pointing backwards); and 2) the humerus (arm bone) is also reversed. I have found other equally alarming errors, for instance: the animal has 5 instead of 7 cervical (neck) vertebrae, with the last 3 reversed; and the shoulder blades have the wrong morphology, and they either belong to a different species or to a badly pathological individual. These mistakes go against the very anatomical adaptations that the figure captions are highlighting (an animal whose bones were assembled like this would not be able to walk, not to mention run at high speed!!), and one wonders how such errors could happen. I have read the account by NG staff of how the image was created, in an extensive blog post which you can check here:
Now this is an interesting story because it points to the huge potential of 3D imaging technologies but also to the dangers if one puts too much faith in them!
As you can gather from the NG blog post, they scanned all the bones of a cheetah skeleton, making clear their intentions to use all the necessary resources. Great!. Then they put an intern in charge of assembling the whole skeleton in 3D in the computer, and here is where the problems started. The intern got most bones roughly in their right place, but for an illustration like this one, where anatomical accuracy is the whole purpose of the thing, “most of them roughly in place” is clearly not enough.
A nice mount of a cheetah skeleton is exhibited in the Natural History museum in Paris, and it is almost exactly in the same pose as the NG illustration. This one has the bones in the proper anatomical positions, and it is a good model against which to check the accuracy of the digital image. You can check it in this link (actually a review of the beautiful book “Evolution”, where a photo of the mount was published):
The fact that the bones were scanned and not drawn or modelled by an artist probably gave the editors a sense of confidence about the accuracy of the result, and that is justified as long as the morphology of the individual bones is concerned (assuming that every bone that is in the box belongs to the same animal, which some times is not the case: curatorial errors also occur!). Such overconfidence in the results may lead to undue relaxation of the reviewing process. It is wrong to just sit back and relax, trusting that the machine has done everything right for us. We still need someone with a real knowledge of carnivore osteology checking, bone by bone, the anatomical position and accuracy of each element. It is all right to have clever machines working for us but we should remember that they are and will remain just that –machines!
For me as a cat enthusiast and a passionate student of felid adaptations for many years, the painful side of all this is that I get the impression that too little attention is paid to the real animal in the whole process of creating an image with a huge potential to disseminate the anatomy of this beautiful cat. An enviable amount of resources have been used, but I miss more real passion for the cats themselves!
Finally, a picture: a comparison between the gallop of a zebra and that of a cheetah, in a nice photo taken by my son Miguel last year in Kenya. You can be sure that the bones that support this amazing cheetah from inside were not assembled like those in the digital image commented above!!!