Examining the Efficacy of a Mobile Therapy for Children With Autism Spectrum Disorder

The purpose of this research is to study the effects of a novel artificial intelligence (AI) tool for automatic facial expression recognition that runs on Google Glass through an Android app to deliver social emotion cues to children with autism during social interactions. This novel device will use a camera, microphone, head motion tracker to analyze the behavior of the subject during interactions with other people. The system is designed to give participants non-interruptive social cues in real-time and will record social responses that can later be used to help aid behavioral therapy. It is hypothesized that the system's ability to provide continuous behavioral therapy during social interactions will enable faster gains in social skills.

Participants in the experimental group will receive the autism glass for 6 weeks once they are assigned to the experimental condition. Participants will be asked to use the glasses at least 3 times a week for 20 minutes sessions in addition to continuing Applied Behavior Analysis (ABA) therapy.

Participants randomized to the control arm, will continue treatment as usual (receiving ABA twice a week) while the intervention participants will receive the Autism Glass intervention (while continuing to receive ABA therapy). After 6 weeks, control participants will receive the Autism Glass intervention after which, they will be asked to come in for a second round of follow-up testing following 6 weeks of use (at week 18).

The intervention uses the outward-facing camera on the google glasses to read facial expressions and provides social cues within the child's natural environment during usual social interaction and during games accessed via the smartphone application. Participants who receive the Google Glass intervention will be asked to use it for around 20 minutes 3 times a week with their parents or during ABA therapy.

Change in socialization subscale scores of the Vineland Adaptive Behavior Scales, 2nd Edition (VABS-II) from baseline to week 6.

Vineland Adaptive Behavior Scales, 2nd edition (VABS-II) Socialization subscale will be conducted at the university by a trained research team member. Scores from the socialization domain of the VABS-II reflects one's functioning in social situations. The socialization subscale is 32 items, where raw scores are converted to IQ-type standard scores (mean: 100 sd: 15) for each domain and for the composite adaptive behavior score.

The SRS-2 is a 65-item measure where parents rate their child selecting responses on a Likert Scale. This measure will be used to measure and identify social impairment associated with Autism Spectrum Disorder (ASD) and to quantify its severity.

The NEPSY-II Affect Recognition subdomain assesses a child's social perception of facial affect recognition. It is designed to assess a child's ability to recognize 6 emotions (happy, sad, angry, fear, disgust, and neutral) from colored photographs of children's faces in four different tasks.

The Emotion Guessing Game is a novel test created for purposes of this study to evaluate the child's ability to correctly label emotions expressed by an examiner in real time. EGG is a pre-set list of 8 emotions, listed 5 times each (Happy, Sad, Angry, Afraid, Surprised, Calm, Disgust, and "Meh"/contempt). During the quick 40-question evaluation, the research coordinator first lists the various emotion choices to the child before beginning the evaluation. Then, the examiner acts out each emotion listed, in order, and waits for a guess from the child, who labels the emotion. The EGG is scored by summing the number of correct responses from the child.

Change in ASD symptoms as measured by Brief Observation of Social Communication Change (BOSCC) from baseline to week 6

The Brief Observation of Social Communication Change (BOSCC) is designed to measure change in core symptoms of children with autism. It aims to capture change in social communication, interaction, and eye contact. The BOSCC is a play based assessment that consists of two boxes filled with specific toys and takes about 12 minutes to complete.

Change in adaptive social and personal skills as measured by Vineland Adaptive Behavior Scales, 2nd edition (VABS-II) Full Scale from baseline to week 6.

The VABS-II is a robust and comprehensive measure of personal and social skills needed in everyday living. The VABS-II covers a child's conceptual, social, and practical skills and can assess children from birth to 90 years. The parent can complete the entire questionnaire in about 20-30 minutes.

Change in child's emotional, behavioral, and social problems from Baseline (week 0) to Week 6 as measured by Child Behavior Checklist (CBCL).

The CBCL is a caregiver-directed report that identifies emotional, behavioral, and social problems in children. It is a 20 item measure completed by parents.

The ABIQ assessment measures a child's IQ based on an abbreviated (10 minute) task that measures Nonverbal Fluid Reasoning and Verbal Knowledge to create a standard score for IQ. It will be completed for each child during Intake.

The MARA screens for, quantifies, and tracks the severity of core autism symptoms. Parents respond to the survey and it takes less than 5 minutes to complete. The core behavioral domains the MARA focuses on are communication, social reciprocity, and restricted, repetitive, and stereotyped behaviors. Each response to a question is run through a machine learning model that uses an alternating decision tree algorithm to generate a total score ranging from most severe, -10 to least severe, 7.

The social communication questionnaire (SCQ) screens for autism in children over 4:0 years in age. Parents are asked 40 Yes/No questions and the resulting score is out of 39 (the first question is not associated with a numerical value). The SCQ assesses a child's communication skills and social functioning. It takes less than 10 minutes to complete and is administered by clinical research coordinators to parents of children during the phone screen to participants who have already completed the online screening questionnaire. A score of 15 or above is indicative of autism and is required to be eligible in the research study.

Inclusion Criteria: the child has been diagnosed professionally with ASD the child is currently receiving ABA therapy at least twice per week at home. The child's family is willing to drive to Stanford University for up to 4 study appointments. Exclusion Criteria: The child scores less than 15 on the Social Communication Questionnaire The child's family does not speak English

Voss C, Schwartz J, Daniels J, Kline A, Haber N, Washington P, Tariq Q, Robinson TN, Desai M, Phillips JM, Feinstein C, Winograd T, Wall DP. Effect of Wearable Digital Intervention for Improving Socialization in Children With Autism Spectrum Disorder: A Randomized Clinical Trial. JAMA Pediatr. 2019 May 1;173(5):446-454. doi: 10.1001/jamapediatrics.2019.0285.