Artificial Intelligence could soon be used to spot signs of depression in humans who are too young to articulate their emotions.
While experts estimate approximately one in five children suffer from “internalizing disorders” such as anxiety and depression, many of those kids lack treatment due to long appointment waitlists, overlooked symptoms and problems with insurance.
“The majority of kids under 8 are undiagnosed,” says Ellen McGinnis, a clinical psychologist and the lead author of a new study on a “machine learning algorithm,” published in the Journal of Biomedical and Health Informatics.
McGinnis hopes the algorithm can streamline the diagnostic process and get more kids the help they need — especially as their young brains are still developing. Researchers say untreated children with anxiety and depression are at greater risk for substance abuse and suicide later in life.
The new AI detection method was tested on 71 children aged 3 to 8 who were asked to come up with a three-minute story, and told they’d be timed and critiqued by judges. The kids were not aware that the judge had been instructed to give only neutral and negative feedback — shoring up anxiety among the young participants.
“The task is designed to be stressful, and to put them in the mindset that someone was judging them,” says McGinnis.
In a matter of seconds, the algorithm was able to analyze with 80 percent accuracy which children — who were also diagnosed by a human — had internalizing disorders. The system was designed to play close attention to subtle symptoms of anxiety and depression, such as speaking in a low-pitched, monotone voice and being repetitious.