Sign In

Communications of the ACM

ACM TechNews

Looking For Answers

View as: Print Mobile App Share:
This 16th-century painting is thought to be of a child with Angelmans syndrome.

A group of researchers at Oxford University are developing software that learns to spot syndromes (collections of co-occurring symptoms) by looking at pictures of people who have been diagnosed with them.

Credit: Getty Images

Dysmorphology is the idea that neurological and behavioral conditions cause changes in the body's shape that can be used for diagnosis, and Oxford University researchers Christoffer Nellaker and Andrew Zisserman want to make dysmorphology work better by using face-recognition technology.

They are developing software that learns to identify syndromes, or collections of co-occurring symptoms, by examining pictures of people who have been diagnosed with them. The program pays attention to features in each face critical to a diagnosis, such as the shape and position of the eyes, eyebrows, lips, and nose, after which it clusters together faces with common characteristics. The program is able to learn to ignore factors such as inconsistent lighting, background, or angle of presentation.

The researchers have tested the system on 1,400 pictures of people with eight of the most common disorders, such as Down's syndrome and progeria, or rapid aging in children. The software was able to divide the photographs spontaneously into eight clusters, which agreed 93 percent of the time with doctors' diagnoses of these disorders.

Nellaker and Zisserman now are working with teams in other countries and are launching a website to gather more pictures of such conditions to improve the system's accuracy.

From The Economist
View Full Article


Abstracts Copyright © 2015 Information Inc., Bethesda, Maryland, USA


No entries found

Sign In for Full Access
» Forgot Password? » Create an ACM Web Account