A data-driven mobile application for efficient, engaging, and accurate screening of ASD in toddlers

Abstract

Early detection of Autism Spectrum Disorder (ASD) followed by targeted intervention has been shown to yield meaningful improvements in outcomes for individuals with ASD. However, despite the potential to curtail developmental delays, constrained clinical resources and barriers to access for some populations prevent many families from obtaining these services. In response, we have developed a tablet-based ASD screening tool called Autoscreen that uses machine learning methods and a data-driven design with the ultimate goal of efficiently triaging toddlers with ASD concerns based on an engaging and non-technical administration procedure. The current paper describes the design of Autoscreen as well as a pilot evaluation to assess the feasibility of the novel approach. Preliminary results suggest the potential for robust risk classification (i.e., F1 score = 0.94), adequate levels of usability based on the System Usability Scale (M = 87.19, 100 point scale), and adequate levels of acceptability based on a novel instrument called ALFA-Q (M = 85.94, 100 point scale). These results, combined with participant feedback, will be used to improve Autoscreen prior to evaluation with the target population of toddlers with concerns for ASD.