International Journal of Scientific Research and Engineering Development

Autism Spectrum Disorder (ASD) is a neurodevelopmental condition affecting communication, social interaction, and behaviour in early childhood. Well-timed diagnosis significantly improves developmental outcomes, yet orthodox diagnostic tools such as the Autism Diagnostic Observation Schedule (ADOS) and the Autism Diagnostic Interview-Revised (ADI-R) are expensive, slow, and heavily dependent on trained clinical specialists. The growing inaccessibility of these tools in rural and resource inadequate settings further compounds the diagnostic gap. This survey reviews the current state of Artificial Intelligence (AI) and Computer Vision (CV) based approaches for early ASD detection in children, identifying key methodologies, datasets, behavioural signals, and research gaps. Based on the discoveries, we propose a gamified AI based mobile screening framework that provides real time facial expression analysis, gaze estimation, and interaction behaviour modelling on mobile hardware. The proposed system incorporates a dual interface design comprising a child facing gamified application for natural behavioural data capture and a clinician facing analytics dashboard providing ASD risk scores, behavioural trend graphs, and AI generated clinical insights. This work is currently in the application and development phase; full experimental validation is planned for subsequent phases. The proposed framework aims to democratize early ASD screening beyond specialist clinics, offering a scalable, non-invasive, and accessible tool for home and primary care settings.

Dr. Shantala T, Ayush Krishnan P, Avikrit Chhetri, Jeswin Mickle, "Early Detection of Autism Spectrum Disorder in Toddlers Using Computer Vision" International Journal of Scientific Research and Engineering Development, V9(5): Page(252-257) September - October 2026. ISSN: 2581-7175. www.ijsred.com. Published by Scientific and Academic Research Publishing.