Multimodal Human-Computer Interaction Interventions for Children and Adolescents With Autism Spectrum Disorder: A Scoping Review.

PURPOSE: Multimodal human-computer interaction (HCI) is increasingly used in interventions for children and adolescents with autism spectrum disorder (ASD), but technological and methodological heterogeneity limits cross-study comparison. This scoping review mapped recent evidence using a pathway framework defined by the dominant interaction mechanism. METHODS: Following JBI and PRISMA-ScR guidance, we searched Web of Science, Scopus, PubMed, and IEEE Xplore for peer-reviewed empirical publications from 2019 to 2025 and conducted backward and forward citation searches. Publications were classified into four mutually exclusive pathways: extended reality (XR), socially assistive robotics (SAR), exergames or embodied-interaction systems (EXG), and hybrid or adaptive systems (HYB). RESULTS: Seventy-two publications were included: XR (n = 25, 34.7%), SAR (n = 23, 31.9%), EXG (n = 13, 18.1%), and HYB (n = 11, 15.3%). Social communication appeared in 49 publications (68.1%) and cognition or executive function in 32 (44.4%), whereas daily-living or safety-related skills appeared in 5 (6.9%) and sensory processing in 3 (4.2%). Eighteen publications (25.0%) used randomized designs, and five (6.9%) used physiologically or neurally driven closed-loop adaptation. CONCLUSION: The pathway framework moves beyond device-based categories by focusing on the mechanisms that organize intervention. This framework helps align system design with intervention goals and participant needs in ASD research. Future research should strengthen pathway-specific reporting, longitudinal evaluation in real-world settings, and transparency of adaptive mechanisms.