An iPhone app that uses automatic emotion and attention analysis of young children while they watch short movies has the potential to identify autism in children, increasing access and affordability of screening, according to the results of a study published in Digital Medicine.
In the United States, 1 in 68 children have autism spectrum disorder (ASD), the most common childhood neurodevelopmental disorder. Although it is known that early intervention improves outcomes in children with ASD and that children can be diagnosed as early as age 24 months, children may wait months or years to undergo evidence-based screening because of lack of access and limited resources. In the United States, the median age at which a child is diagnosed with ASD is 4 years.
In the Autism & Beyond study, Helen L. Egger, MD, of the department of psychiatry and behavioral sciences at Duke Health in Durham, North Carolina, and colleagues examined data collected from families with children age 12 to 72 months over a period of 1 year. Data included caregiver reports and child video behavioral data. The researchers collected usable data from 87.6% of 4441 uploaded videos recorded in the children’s home environments. Caregivers completed a mean of 3.2 out of 4 possible caregiver-reported surveys. Automatic coding was able to detect significant differences in emotion and attention by age, sex, and autism risk.
The use of smartphone technology to collect data on child development using an app built on Apple’s open-source ResearchKit framework and available on the Apple App Store demonstrates the feasibility of doing research studies with caregivers and young children in the home environment. The investigators noted that enrollment in this study was 10 times that of a clinic-based study they conducted during the same time period. They also noted that many of the participants in the Autism & Beyond study enrolled because they heard about the study on social media, indicating that researchers can leverage new media and technology to further their research in novel ways.
The researchers remarked that they need to further validate tools tested in this study in representative populations of children. They call for the inclusion of children with a variety of developmental disorders to better determine not only the feasibility of this approach but also its sensitivity and specificity.
The investigators believe that although these tools are only for research at the moment, they lay the foundation for what may become accessible, affordable, and scalable tools for the early identification of autism and other childhood developmental disorders in the future.
Disclosures: GD is on the Scientific Advisory Boards of Janssen Research and Development, Akili, Inc, and Roche Pharmaceuticals, has received grant funding from Janssen Research and Development, LLC, received royalties from Guildford Press and Oxford University Press, and is a member of DASIO, LLC. GS is on the Board of Surgical Information Sciences, LLC, and is a member of DASIO, LCC. RB moved to Apple after the design was completed and the app was released. Duke University has IP and patent pending on some of the material reported in this work. The remaining authors declare no competing financial interests.
Egger HL, Dawson G, Hashemi J, et al. Automatic emotion and attention analysis of young children at home: a ResearchKit autism feasibility study [published online June 1, 2018]. Digital Med. doi:10.1038/s41746-018-0024-6