What if diagnosing rare genetic disorders was as easy as tagging photos on Facebook?
Researchers at Oxford University have created a computer program that uses photo software with face recognition capabilities, similar to those used on social media sites, to discern the likenesses in facial features of people who have the same genetic disorders.
The software can then compare a photograph of the patient to these prominent facial structures and generate a ranked list of possible disorders.
How can a mere photograph provide the basis for a diagnosis? Disorders such as Down’s syndrome, Angelman syndrome and Progeria are caused by genetic mutations that also affect the growth of the face and the cranium as a baby develops, allowing for recognition of characteristic facial features.
Doctors already use the analysis of facial features as part of the way they diagnose patients, but a computer program could reduce the room for error that comes with human judgment.
The computer program, an example of the newest machine learning technology, uses an algorithm that becomes better at knowing which facial features are pertinent to diagnosis with each photograph it processes. It can analyse faces despite differences in lighting, facial expression and photo quality.
Researchers are already noting the algorithm’s effectiveness. Photographs of people with the same disorder are automatically grouped together by the program.
Likewise, the program clusters patients whose facial features are similar but do not match any known diagnosis, a useful tool in the identification of new or extremely uncommon disorders.
The developers of the program foresee a future where face recognition technology aids in the quick diagnosis of a patient. Doctors will be able to snap a picture of the patient on their smartphone, run it through the program, and analyze the results.
One in 17 people have a genetic disorder, but most go undiagnosed. With this new technology, patients could be diagnosed from a younger age, receive treatment earlier and improve their quality of life.
The potential for consistency, speed and accuracy in diagnosis could also help shift research towards treatment and cures for these genetic disorders.
On social media, face recognition technology can seem nothing more than a handy feature that saves you a few minutes when tagging photos. However, innovative applications of the technology we use every day, like this photo software, could put us on track for a fuller understanding of conditions that affect millions of people.