Computer Accurately Detects Pain

Pretend to hurt to get out of work or school? Researchers led by a UC San Diego scientist have found a way to give you a real pain.

The researchers have developed a computer system that detects whether pain is faked. And it’s far more accurate than the best human observers. With more development, the system could find uses helping doctors spot those in true pain, employers nail malingerers, and even warn drivers when they are dozing off behind the wheel.

The computer system was 85 percent accurate at determining the true state of a person’s pain, according to a study published Thursday in the journal Current Biology. That’s far above the accuracy of trained observers, who were correct just 55 percent of the time.

Marian “Marni” Bartlett of UCSD’s Institute for Neural Computation was the study’s lead author. Kang Lee of the University of Toronto was senior author.  The technology could help determine whether people are lying to get prescription painkillers and to detect the intensity of pain in children who can’t articulate well, such as some children with autism, Bartlett said. She’s working with Dr. Jeannie Huang of Rady Children’s Hospital San Diego for the latter goal.

 

© The UCLA Institute for Society and Genetics. All Rights Reserved.