Wearable intervention may improve socialization in children with autism
A new wearable intervention could potentially improve the social skills of children diagnosed with autism spectrum disorder (ASD), a new study has shown.
Participants were 71 children aged 6–12 years (mean age 8.38 years, 89 percent male, 24 percent East Asian/Asian American) diagnosed with ASD (96 percent with ASD, 4 percent with Asperger). They were randomized to receive standard of care (SOC) applied behavioural analysis therapy (at-home therapy; median, 15–20 hours/week) with or without the Superpower Glass (SG), an artificial intelligence-driven wearable device for behavioural intervention that “promotes facial engagement and emotion recognition by detecting facial expressions and providing reinforcing social cues” (n=40 and 31, respectively).
The SG intervention was coordinated using Google Glass – which would be worn by the participant – and a smartphone app which was managed by the caregiver, and the 20-minute “intervention” sessions were conducted at home four times/week for 6 weeks (three times/week with family members and once/week with the child’s behavioural therapist).
“The system tracks faces, classifies the emotions* of the child’s social partners, and provides two forms of cues to the child in real time,” said the researchers.
At 6 weeks, participants assigned to the intervention demonstrated a significant improvement on the socialization subscale of the Vineland Adaptive Behaviour Scale (VABS-II; a measurement of communication, socialization, adaptive behaviour, motor function, and daily living) compared with those who were assigned SOC therapy only (mean impact of treatment, 4.584 points; p=0.005). The improvement was more evident among the 52 children who completed the study (mean impact of treatment, 5.384 points; p<0.001). [JAMA Pediatr 2019;doi:10.1001/jamapediatrics.2019.0285]
Patients assigned to the intervention also demonstrated improvements in Social Responsiveness Scale (SRS-II), Emotion Guessing Game (EGG), and Developmental Neuropsychological Assessment (NEPSY-II-Affect) scores compared with those assigned to SOC only, but the findings were not significant (mean impact of treatment, -1.654; p=0.26, 2.790; p=0.05, and 0.099; p=0.86, respectively).
Six weeks after stopping the intervention, the impact of the device on the VABS-II socialization subscale was no longer significant compared with SOC (p=0.26), though the effect on SRS-II and EGG were significantly better than that of SOC (mean treatment change, -2.832; p=0.003 and 5.647; p<0.001, respectively).
According to the researchers, recommended behavioural therapy for ASD is costly and the increase in ASD prevalence has led to a long wait-list due to unavailability of behavioural therapists. [JAMA Pediatr 2014;168:721-728; Pediatr Clin North Am 2016;63:851-859] As such, there is a need for out-of-office therapy, they said.
“Overall, these results support the hypothesis that the SG intervention can improve social skills of children with ASD between the ages of 6 and 12 years as an augmentation to SOC therapy [and underscore the potential of digital home therapy to augment the current SOC],” said the researchers.
“The SG intervention has at least two potential mechanisms of action – reinforcement that faces have variation in emotion … and training on how to differentiate emotions,” said the researchers.
“[W]e hypothesize that the dual mechanisms of action may have contributed to the observed socialization gains … it is [also] possible that the learning aid generally encourages social interaction in the family around face contact and emotion, thereby increasing social acuity in the child,” they added.
Due to the lower than recommended use of the device (approximately half the recommended usage), the researchers called for further study into factors affecting device use, which could be due to lack of motivation among the parents or lack of engagement among the children. They also acknowledged that the participants had different levels of social skills at study onset which could have affected the findings.