Computer analysis of family photos could help diagnose rare genetic disorders

Computer analysis of family photos could help diagnose rare genetic disorders -

Computer analysis of photographs could help doctors to diagnose a child with what condition a rare genetic disorder, say researchers from the University of Oxford.

researchers, funded in part by the Medical Research Council (MRC), have developed a computer program that recognizes facial features in the pictures; looking for similarities with the facial structures for various conditions such as the syndrome, Angelman syndrome, Progeria or Down; and returns possible matches according to their probability index.

Using the latest in computer vision and machine learning, the algorithm learns more and more that the facial features to pay attention to and what to ignore a bank increasingly photographs of people diagnosed with different syndromes.

researchers report their findings in the journal eLife . The study was funded by the MRC, Wellcome Trust, the National Institute for Health Research (NIHR) Oxford Biomedical Research Centre (BRC) and the European Research Council (ERC VISREC).

While genetic disorders are each individually rare, collectively these conditions are thought to affect one in 17. Of these, a third party may have symptoms that significantly reduce the quality of life. However, most people fail to receive a genetic diagnosis.

"A diagnosis of a rare genetic disease can be a very important step. It can provide parents with some certainty and help with genetic counseling about risks for other children or the probability of a state must be transmitted, "says lead researcher Dr. Christoffer Nell-ker Functional Genomics Unit at MRC University of Oxford. "A diagnosis can also improve estimates of how the disease might progress, or show what symptoms are caused by genetic disease that is caused by other clinical problems that can be treated."

L the team of researchers. Oxford University included first author Quentin Ferry, a research DPhil student, and Professor Andrew Zisserman Department of Engineering Science, who provided expertise in computer vision and machine learning

Zisserman professor said: "It is good to see such inventive and beneficial use of modern methods representation of the face."

identifying a suspected developmental disorder tends to require that clinical geneticists to come to a conclusion based on facial features, monitoring of tests and their own expertise. It is believed that 30-40% of rare genetic diseases involve some form of change in the face and skull, perhaps because many genes are involved in the development of the face and skull as a baby grows in the uterus.

researchers themselves to teach a computer to do some of the same assessments objectively

They developed a program -. like Google, Picasa and other photo software - recognizes faces, photographs in all ordinary days. The program for variations in lighting, image quality, background, pose, facial expression and identity. He built a description of the facial structure by identifying the corners of the eyes, nose, mouth and other features, and compares this against what he learned other photographs entered into the system.

The researchers have developed algorithm sees patients sharing the same state automatically grouped together.

computer algorithm is better to offer a diagnosis for a picture where he has already seen many other photos of people with this syndrome, as it learns more with more data.

patients also come together when no documented diagnosis exists, which could help identify ultra-rare genetic disorders.

"a doctor should in the future, anywhere in the world, be able to take a patient's smartphone photo and run computer analysis to quickly find what troubles genetics of person could have, said Dr Nell-ker.

"This objective approach could help narrow the possible diagnoses, facilitate comparisons and allow doctors to reach a conclusion with more certainty."