There are some genetic disorders, such as Down Syndrome, that are common enough to easily diagnose. Others, however, require the expertise of certain geneticists who are trained to diagnose based on facial features. Very few medical professionals have this skill, making it hard to get a proper diagnosis.
A team at the University of Oxford have created software that can analyze digital photographs and compare them to publicly available photos of people with genetic disorders. It then learns how to identify certain features, like the size of eyes or lips, to diagnose genetic disorders. A round of testing had 93 percent accuracy.
The software is now able to identify 90 rare genetic disorders with a database of 2754 images. The program itself isn't able to make a diagnosis, but studies have shown that it makes getting a correct diagnosis 30 times more likely.
While this software would be beneficial to the first world, its real impact will be for those in developing countries, where access to doctors and medical supplies is scarce. Photos could be sent to the program from anywhere in the world to help prompt a diagnosis.
Researchers are hoping to eventually train the software to analyze profile photos as well as frontal, and pair the program with DNA analyzing software so facial and genetic features can be analyzed together.