A study, published in the journal Proceedings of the Royal Society B, injected 16 healthy participants with either a placebo or a type of E. coli bacteria that causes flu-like symptoms. The participants were photographed just two hours after receiving the injections, and those photos were shown to 62 separate participants who were asked to determine whether the person in the photo was healthy or sick after only five seconds of viewing the photo.
Though participants could only detect a sick person accurately 52 percent of the time, they were able to point out a healthy person correctly 70 percent of the time. Facial features participants commonly associated with sickness included redder eyes, duller skin, a more swollen face, a droopier mouth and eyelids, and paler skin and lips.
“We expected that people would be better than chance at detecting sick people, but far from 100 percent since they were only allowed to see a photo for a few seconds,” John Axelsson, PhD, a co-author of the study, told Smithsonian.com. “We expect people to be a lot better when they can interact for real with someone and then also use other cues such as biological motion, smell etc.”
Other academics who study facial cues suggest this research could be used to train AI. Mark Frank, PhD, a communications professor at the University at Buffalo (N.Y.), told Smithsonian.com AI could help perform real-time analysis and decision-making to help those already “overwhelmed by too much information.”
In order for this kind of AI to be effective, though, humans have to teach the tech’s neural networks what to do, meaning humans need to know what to look for. Dr. Axelsson’s study that shows what facial changes are associated with sickness may help humans begin the teaching.
Smithsonian.com suggests airports may be a prime starting point for illness-detecting AI, since some airports around the world already use temperature scanners to weed out potentially sick travelers.
More articles on artificial intelligence:
Zocdoc CEO: Physicians will need emotional intelligence to keep up with AI
NTT Data partners with Pieces Technologies to integrate AI into clinical analytics portfolio
Survey: Tech expertise ranks as No. 1 barrier to AI development among life science pros
At the Becker's 11th Annual IT + Revenue Cycle Conference: The Future of AI & Digital Health, taking place September 14–17 in Chicago, healthcare executives and digital leaders from across the country will come together to explore how AI, interoperability, cybersecurity, and revenue cycle innovation are transforming care delivery, strengthening financial performance, and driving the next era of digital health. Apply for complimentary registration now.
