Artificial intelligence to improve social skills in children with Autism Spectrum Disorder: a systematic review

Eddier Albino F.-I.; Luz Sheyla C.-C.; Heber G.-U. · Proceedings of the LACCEI international Multi-conference for Engineering, Education and Technology (2025) · DOI: 10.18687/laccei2025.1.1.1881

Resumen

This study conducts a systematic review of the literature on tools based on artificial intelligence (AI) to improve social skills in children with Autism Spectrum Disorder (ASD), covering the period 2019-2024. The information search was conducted in the Scopus database, where 40 articles were identified and selected using the PRISMA and PICO method, specifically its PIO variant. Subsequently, a bibliometric analysis was performed to obtain a deeper insight into research in this field. Tools such as VOSviewer were used to visualize cooperation between countries in related scientific production; Bibliometrix to graphically represent the countries with the highest scientific production and Power BI to generate a keyword cloud (WordCloud). The study investigates various AI methods, including analysis of brain images (MRI, fMRI), electroencephalographic (EEG) signals and behavioral data, with the aim of developing more accurate and efficient diagnostic systems. Different machine learning algorithms (SVM, CNN, neural networks, etc.) are compared and feature selection techniques are analyzed to improve the accuracy of predictive models. Finally, applications of social robotics in the interaction and learning of children with ASD are explored.

Instituciones: Universidad Tecnológica del Perú