Google Glass paired with face-recognition software may help children with autism spectrum disorder (ASD) recognize facial expressions, a pilot study suggests.
Superpower Glass, a machine-learning-assisted software system that runs on Google Glass and an Android smartphone, helped children with autism understand emotions conveyed in faces and improve social skills, reported Dennis Wall, PhD, of Stanford University, and co-authors in npj Digital Medicine.
“This is a viable strategy to deliver social training,” Wall said in an interview with MedPage Today. “It encourages facial contact and social interaction, but it also provides an appreciation of the salience of emotion, that there is something interesting inherently about human faces.”
“It’s exciting, fun, and functional,” he added. “In fact, the children in the study called it their superpower, so we decided to call it Superpower Glass.”
The Google Glass device, which links to a smartphone, consists of an eyeglasses-like frame with a camera to record the wearer’s field of view, a small screen, and a speaker. As a child interacts with other people, the app identifies and names their emotions through the Google Glass speaker or screen.
In this feasibility study, Wall and colleagues sent the Superpower Glass tool home with 14 families and assessed how children changed along three measures: the Social Responsiveness Scale (SRS-2; a rating scale measuring deficits in social behavior), a facial affect recognition task, and qualitative parent reports. Each family had a child between the ages of 3 and 17 with a clinically confirmed autism diagnosis.
Families completed three or more 20-minute sessions with Superpower Glass a week, for an average of 72 days. Overall, families chose evenly between structured interactive games (“Capture the Smile” and “Guess the Emotion”) and free play, although families with children who had more severe autism were more likely to choose structured game modes.
During the study, the children’s average SRS-2 score decreased by 7.38 points (P<0.001), indicating less severe autism symptoms. Six of the 14 participants had large enough drops in scores to move down one step in autism classification: four from severe to moderate, one from moderate to mild, and one from mild to normal.
Twelve of the 14 families said their children made more eye contact after receiving the treatment. Facial affect recognition scores also increased by 9.55 correct responses on average (P<0.01).
The therapy used in Superpower Glass is based on applied behavior analysis, an autism treatment that teaches emotion recognition through structured exercises, including flash cards with faces that have different emotions.
Superpower Glass could help fill a major gap in autism care, Wall said: because of a shortage of trained therapists, children now wait as long as 18 months after an autism diagnosis to receive therapy.
Still, the findings should be interpreted cautiously, since this was a small pilot study without a control arm, he said. The research team recently completed a larger, randomized controlled trial comparing Superpower Glass with standard care, and the results should be published later this year, Wall noted.
The research was supported by the National Institutes of Health, the Hartwell Foundation, the Bill and Melinda Gates Foundation, the Coulter Foundation, and the Lucile Packard Foundation, as well as by program grants from Stanford’s Precision Health and Integrated Diagnostics Center, Beckman Center, Bio-X Center, the Predictives and Diagnostics Accelerator Program, the Child Health Research Institute, and Human-Centered AI. In-kind material grants included a gift from Google (35 units of Google Glass version 1) and Amazon Web Services founder support.
The authors reported having no competing interests.
LAST UPDATED
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.