Can a small motion sensor identify anxiety and depression in young children?
There is strong evidence that anxiety disorders begin much earlier than most people think, with as many as one in five young children showing sign of clinical anxiety. The problem is that these young children often lack the understanding and language to communicate their anxiety.
Additionally, these “internalizing disorders”, as they are often called, are particularly hard to identify from the outside. Parents, teachers, and even doctors may be unable to notice the subtle signs that a young child is already dealing with anxiety issues.
This can have huge implications for these children’s lives as well, seeing as studies have shown children with internalizing disorders are at much greater risk for substance abuse or suicide later in life.
“Because of the scale of the problem, this begs for a screening technology to identify kids early enough so they can be directed to the care they need,” explains Ryan McGinnis, a biomedical engineer at the University of Vermont.
With this in mind, McGinnis partnered with Ellen McGinnis, a clinical psychologist at UV, to create a new wearable sensor that identifies “hidden anxiety” and subtle signs of depression in younger children. The two also worked with Maria Muzik, Katherine Rosenblum, and Kate Fitzgerald from the Department of Psychiatry at the University of Michigan.
As the team reports in the latest issue of PLoS One, they used a “mood induction task” to elicit specific signs or responses to feelings of anxiety and depression in 63 children – some of whom were previously diagnosed with internalizing disorders.
Specifically, the children were led into a dimly lit room by a trained individual who would say things such as “I have something to show you” and “Let’s be quiet so it doesn’t wake up.” Then, the facilitator would unveil a terrarium which contained a fake snake. After this, the children were reassured by the facilitator and allowed to play with the fake snake.
Typically, when conducting a mood induction task such as this, researchers would review footage of the interaction to score the child’s responses and behavior during the task to identify internalizing disorders. However, in this case, the children were equipped with wearable motion sensors connected to a machine learning algorithm designed to analyze the child’s behavior and identify signs of anxiety or depression.
After reviewing the results of the sensors, the algorithm was able to detect internalizing disorders in the young children with 81% accuracy – significantly better than the commonly used parent questionnaires.
To put it simply, Ryan McGinnis says “the way that kids with internalizing disorders moved was different than those without.”
In particular, the algorithm indicated that during the time before the snake was revealed, the children with internalizing disorders tended to turn away from the potential threat considerably more than the control group. More so, there were subtle changes to the way the children turned that helped to identify the children with anxiety disorders.
As the report explains, this was expected based on existing psychological theory and understandings of internalizing disorders.
Children with internalizing disorders typically experience greater anticipatory anxiety, which is reflected by turning away in the face of a potential threat or stressor. It is also the kind of behavior researchers are trained to watch for in mood induction tasks, but the sensors and algorithm provide a faster, more objective analysis.
“Something that we usually do with weeks of training and months of coding can be done in a few minutes of processing with these instruments,” says Ellen McGinnis.
“Children with anxiety disorders need an increased level of psychological care and intervention. Our paper suggests that this instrumented mood induction task can help us identify those kids and get them to the services they need,” she concludes.