Exploring the application of AI in the education of children with autism: a public health perspective

Le résumé fourni par la source

Introduction: Autism Spectrum Disorder (ASD) presents significant challenges in social communication and interaction, critically impacting the lives of children with ASD. Traditional interventions, such as Applied Behavior Analysis (ABA) and Social Skills Training (SST), have been widely used to address social skill deficits in these children. While these methods are effective, they often require substantial resources, long-term engagement, and specialized expertise, which limit their accessibility and adaptability to diverse social contexts. Recent advancements in artificial intelligence (Al), particularly Transformer-based models, offer a novel opportunity to enhance and personalize social skills training. Methods: This study introduces a Public Health-Driven Transformer (PHDT) model specifically designed to improve social skills in children with ASD. By integrating public health principles with state-of-the-art Al methodologies, the PHDT model creates interventions that are adaptable, accessible, and sensitive to individual needs. Leveraging multi-modal data inputs-such as text, audio, and facialcues-PHDT provides real-time social context interpretation and adaptive feedback, enabling a more naturalistic and engaging learning experience. Results and discussion: Experimental results reveal that PHDT significantly outperforms traditional methods in fostering engagement, retention, and social skill acquisition. These findings highlight PHDT's potential to improve social competencies in children with ASD and to revolutionize access to specialized support within public health frameworks. This work underscores the transformative impact of Al-driven, public health-oriented interventions in promoting equitable access to essential developmental resources and enhancing the quality of life for children with ASD.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.