It is now possible to detect signs of genetic illnesses by using a computer programme which analyses photographs of faces.
Image recognition is already contributing to rapid diagnostic testing using a Google Glass application, and now it is helping to detect rare genetic illnesses, which are generally difficult to diagnose. At the Medical Research Council (MRC) Functional Genomics Unit at the University of Oxford, UK, researchers have developed an algorithm which is able to analyse photographs of children’s faces and detect signs of rare genetic disorders. Conditions such as Down's syndrome, Angelman syndrome and Progeria can be identified using the computer programme, which uses machine learning and computer vision techniques to recognise facial characteristics on patient photographs.
Today, detecting the potential onset of a rare genetic disorder first requires doctors to give an opinion based on the patient’s observed facial characteristics, followed up by further tests. The researchers estimate that 30 - 40% of all rare genetic disorders involve some form of change to the face and skull. The algorithm they have developed enables the first observational diagnostic step to be carried out in automated fashion. The programme they have created recognises faces in ordinary, everyday photographs. It then builds a description of the facial structure by identifying corners of eyes, nose, mouth and other features, and compares this against what it has learned from other photographs of people with diagnosed medical conditions that have been fed into the system. The more data it analyses, the more the algorithm will learn and the more able it will be to identify a range of genetic disorders.
Early diagnosis helps planning and treatment
“A diagnosis of a rare genetic disorder can be a very important step. It can provide parents with some certainty and help with genetic counselling on risks for other children or how likely a condition is to be passed on,” points out lead researcher Dr Christoffer Nellaker at the MRC Functional Genomics Unit at the University of Oxford. The computer programme could also enable the detection of very rare disorders, as the algorithm preserves the analyses of cases which have never been observed before in order to provide a comparison. “A doctor anywhere in the world should in future be able to take a smartphone picture of a patient and run the computer analysis to quickly find out which genetic disorder the person might have,” predicts Dr Nellaker.