A silhouette of a person against a green backdrop of binary codeShare on Pinterest
New research shows that a “child-like robot” was able to detect mental health concerns in children with more accuracy than humans. Aitor Diago/Getty Images
  • The growing youth mental health crisis highlights a need for early detection and treatment of mental health disorders.
  • A new study by University of Cambridge researchers found that socially assistive robots (SARs) could serve as a potential diagnostic tool for mental health.
  • According to the researchers, the study is the first time robots have been used to asses children’s mental well-being.
  • The study shows that robots were more likely to identify cases of well-being anomalies than self-reports filled by children or reports made by their parents.
  • Still, researchers did not use robots to deliver mental health interventions, but rather, to detect and diagnose children’s mental health concerns.

Even before the pandemic, an estimated 4.4% of children (about 2.7 million) between the ages of three and 17 had been diagnosed with depression in the United States, according to the National Survey of Children’s Health. The same survey found about 9.4% (about 5.8 million) of children had been diagnosed with anxiety.

Experts believe that stress from COVID-19 has led to increased depression and anxiety in young people.

Mental health-related emergency room visits for children ages 5 to 11 increased by 24% in 2020 compared to the year prior, according to a report from the Centers for Disease Control and Prevention (CDC). Among children between the ages of 12 and 17, visits increased by 31%.

At the same time, adequate mental health care and access are still lacking in the U.S.

Nearly 91 million Americans live in regions with shortages of mental health providers, according to the Substance Abuse and Mental Health Services Administration, which estimates a minimum of 1,846 psychiatrists and 5,931 other practitioners are needed to fill the gap.

Recently, a group of researchers from the University of Cambridge studied the effects of socially assistive robots (SARs), which could potentially serve as an assessment and diagnostic tool in areas where mental health professionals are in short supply.

Their work was presented this week at the 2022 IEEE International Conference on Robots and Human Interactive Communication in Naples, Italy.

For the study, researchers selected 28 children from Cambridgeshire, England, ages 8 to 13 years old. Among the participants, 21 were female and 7 were male with an average age of 9.5.

Children who had already been diagnosed with neurological or psychological disorders were excluded from the research.

First, the participants answered about their well-being on an online questionnaire. Additionally, parents or guardians answered a questionnaire about the well-being of their children.

Later, the young participants spent 45 minutes with a Nao robot, created by SoftBank Robotics. The robot then administered the Short Mood and Feelings Questionnaire, which measures symptoms of depression, and the Revised Child Anxiety and Depression Scale.

Additionally, the robot asked the children about happy and sad memories they experienced over the last week and administered a task where children were shown pictures and then asked questions about them.

The researchers found that questionnaires conducted by robots were more likely to identify cases of well-being anomalies than the children’s online self-reports or parent or guardian reports.

Some participants shared information with the robot that they did not share via self-report.

Study co-author Prof. Hatice Gunes, Ph.D., a professor of affective intelligence and robotics and head of the Affective Intelligence and Robotics Lab at the University of Cambridge, explained to Medical News Today that among participants, “the group that might have some well-being related concerns” were more likely to provide negative response ratings during the robot-led questionnaires.

“The interesting finding here is when they interact with the robot their responses become more negative,” Prof. Gunes noted.

Socially assistive robots have previously demonstrated potential as a tool to improve the accessibility of care, the researchers explain in their paper. For instance, a 2020 study illustrated that robots may be helpful in assessing risk factors for autism spectrum disorder (ASD).

“Robots have been used for various tasks — and they’ve been shown to be effective in certain things because they have this physical embodiment, unlike a mobile phone or a virtual character or even videos,” Prof. Gunes said.

And despite the potential dangers of allowing a child too much time with an electronic device, working one-on-one with a robot is different from screentime, Prof. Gunes noted.

“This is a physical interaction, right? So, it’s not virtual. It’s not a video — they’re physically interacting with a physical entity,” she said.

Prof. Gunes also pointed to a key aspect of the study: the “child-like robot” used for the research was less than 2 feet tall.

“Here we have a child-looking and child-sounding robot. “In such situations, children actually see the robot more as a peer. So, it’s not an adult that is trying to get some information out of them.”

– Prof. Hatice Gunes, Ph.D., professor of affective intelligence and robotics at the University of Cambridge

Diane Hodge, Ph.D., LCSW, director of the school of social work at Radford University in Virginia, said she’s used puppets and dolls to help make her pediatric clients feel more at ease while she worked as a clinical social worker earlier in her career.

Robots, she told MNT, are the 21st-century equivalent of those puppets.

“I’m all about technology that really enhances and helps people,” Hodge said. “More kids today are just so used to it that they expect it.”

Hodge also pointed out that in the study, the researchers did not use a robot to deliver mental health care interventions, but rather, to assess the well-being of children. “This is just to get people access,” she said.

Hodge also highlighted how the Nao robot was able to successfully identify more “well-being-related anomalies” in children compared to those revealed by humans. “[That] shows that if we had done nothing, right, they wouldn’t have caught that,” she said.

According to Prof. Gunes, her research interests evolved after having a baby in 2018. “I think I became more sensitive to child-related issues and their well-being,” she said.

In the future, Prof. Gunes said the researchers hope to study how children respond to interacting with a diagnostic robot over video chat.

Already, researchers are preparing to conduct a study similar to the one they presented at the conference only with a more equal ratio of male and female participants, according to Prof. Gunes.

“We want to see actually if the findings are consistent across genders,” she said.