AutiScan: AI-Based Early Screening System for Autism Spectrum Disorder Using Machine Learning

Autism Spectrum Disorder (ASD) is a neurodevelopmental condition that affects communication, behaviour, and social interaction in early childhood. Early identification of ASD is crucial for timely intervention; however, conventional diagnostic methods are often expensive, time-consuming, and require trained professionals, making them inaccessible to many families. This paper presents AutiScan, an AI-based web platform designed to support early screening of ASD in children aged 1 to 4 years. The proposed system integrates structured questionnaires with behavioural analysis using uploaded video data to provide a preliminary risk assessment. Machine learning models, including Random Forest and Convolutional Neural Networks, are employed to analyse behavioural patterns and classify risk levels into low, moderate, and high categories. The system also incorporates explainable AI techniques to enhance transparency and user trust by highlighting key contributing factors in the assessment. Experimental evaluation demonstrates promising performance, achieving an accuracy of 89.40%, along with high precision and recall values. The platform is user-friendly, scalable, and accessible, enabling parents and caregivers to perform screening at home. AutiScan aims to reduce diagnostic delays, increase awareness, and facilitate early intervention, ultimately improving developmental outcomes for children with ASD..