I just discovered this story from the Science Show on ABC Radio National from earlier this month about researcher Christoffer Nellåker. It sounds a lot to me like the kind of research that researchers in Perth, Western Australia have been busy with in the Perth Face-Space Project.
I’ve written it before and I’ll write it again; I’m bemused that so much effort and research is continuing to be put into very sophisticated efforts to create systems for recognizing genetic disorders and diseases in facial appearances, involving sophisticated computer algorithms and photography, to do a job that any super-recognizer could be trained to do with I expect little time and effort. In fact I expect that super-recognizers and people with average levels of face recognition ability naturally have unconsciously-used skills at detecting facial dysmorphology. Who couldn’t pick a case of Down’s syndrome in a crowd, regardless of age or gender? And Treacher Collins is hard to miss, and easy enough to name once you know what it is. I get the point that some of the diseases that can be detected in the face are exceedingly rare, and therefore even a medical specialist would have little or nothing in the way of familiarity with the typical appearance linked to the disease, but I would think it is also true that for any medical case a short-list of diseases will be identified during the diagnostic process, and for each of those diseases, regardless of rarity, there should be some kind of photographic record that can be accessed and studied, and compared against the appearance of the patient. Whichever way you look at it, this job needn’t be rocket science, but as is so often the case, people feel more comfortable in placing their faith in the performance computer system than the skill of a person.
Williams, Robyn Identifying rare diseases from facial images. Science Show. Radio National ABC. December 6th 2014. http://www.abc.net.au/radionational/programs/scienceshow/identifying-rare-diseases-from-facial-images/5947924
Saw, Samantha Defining normal. InkWire. April 17, 2014. http://inkwirenews.com.au/2014/04/17/defining-normal/