From many years ago, doctors actually had their own method for recognizing signs of genetic disorders suffered by both children and adults, namely through observation of appearance.
This method uses an examination of facial patterns that are deemed 'odd'. This is also supported by research results which show that 40 percent of rare disabilities cause sufferers to have different faces from normal people, Gizmodo (24/6).
Christoffer Nellaker and Andrew Zisserman, two researchers from the University of Oxford, then took a new step by creating a computer software based on a special algorithm that was able to learn automatically and was able to recognize the faces of people with disabilities from ordinary photos, including even family photos.
The software system is built with 1,363 photos of people with eight behavioral disabilities, from Down syndrome to progeria. To increase the prediction result up to 30 times from manual examination, there are about 36 markings on the face that are used to determine the type of disability.
