- ‘American Dirt’ controversy inspires meme about Latinx stereotypes in literature Wednesday 9:02 PM
- What is the TikTok ‘flex challenge’? Wednesday 8:03 PM
- GoFundMe to send ‘Target Tori’ on vacation raises more than $30K Wednesday 6:54 PM
- Furries stop domestic assault in viral video Wednesday 6:10 PM
- Gritty under police investigation for allegedly punching a teen fan Wednesday 6:04 PM
- Twitter users throw animal parties with emoji in new meme Wednesday 5:21 PM
- Woman who went viral supporting Soleimani killing exposed as Libyan militia lobbyist Wednesday 5:01 PM
- Jeff Bezos subtweets Saudi prince following phone hack report Wednesday 3:29 PM
- ‘Yeah, good. OK’ Bernie Sanders meme is a new way to dismiss people Wednesday 3:10 PM
- ‘Vanderpump Rules’ recap: Petty displays of affection Wednesday 2:12 PM
- Makeup artist transforms into Timothée Chalamet on TikTok Wednesday 1:54 PM
- Iguanas are falling from trees—and people are selling them online for food Wednesday 1:02 PM
- 75,000 sign petition to fire Wendy Williams after ‘cleft lip’ comment about Joaquin Phoenix Wednesday 12:30 PM
- Kim Kardashian says Kylie Jenner’s setting spray is ‘cheap sh*t’ Wednesday 11:59 AM
- Trump continues to demand Apple unlock iPhones for the government Wednesday 11:46 AM
Software can analyze faces in photos for genetic diseases
You might find more than your dad’s nose in family photos.
We already know that computers can recognize human faces, and apparently decipher secret codes left on bird eggs, but researchers at the University of Oxford have now developed software that can diagnose rare genetic diseases just from glancing at a portrait.
It all started with over 1,300 photos of people who are known to have one of eight different genetic disorders including Down’s syndrome, Angelman syndrome, and Progeria. The software analyzed the photos and made observations about various facial features for comparison. With a total of 36 data points from each photo, the system quickly learned the effect each disorder had on a person’s appearance.
When the software is fed a new photo of a person who may or may not have a genetic disorder, it scans their features and then ranks the potential diagnoses based on the likelihood that the individual pictured has it. The first tests were promising, with the system correctly predicting the specific genetic defect in 93 percent of individuals.
One particularly interesting test subject was the late Abraham Lincoln, which the smart software suggested may have suffered from Marfan syndrome, which can result in skeletal abnormalities. It was once postulated that Lincoln had the disorder, though today most geneticists believe he did not.
Since its initial testing, the system has been gotten even smarter, and is now capable of identifying 90 different disorders. The researchers hope that it will be a new first step in diagnosing genetic issues without the need for multiple doctor visits. If a computer can offer a hint at what a person may suffer from, doctors have a great starting point for more definitive testing.
Mike Wehner is a former tech editor for the Daily Dot who now writes for BGR. His work has appeared everywhere from Yahoo to CNN, and there’s a good chance his Apple Watch is dead right now.