Early Detection of ASD Risk in Children Through a Web Application with Random Forest

Ramos A.G.G.; Ccorahua C.R.R.; Junes A.N.H. · Communications in Computer and Information Science (2026) · DOI: 10.1007/978-3-032-16451-3_22

Resumen

Autism Spectrum Disorder (ASD) is a developmental disorder that affects social communication and behavior. Early detection is crucial for initiating interventions that promote child development. However, in resource-limited settings such as Peru, access to specialized diagnostic tools is scarce. This paper presents a web application designed for parents and caregivers to estimate the risk of Autism Spectrum Disorder (ASD) in children aged 12 to 36 months, using the Q-CHAT-10 questionnaire and a Random Forest machine learning model. The system was trained with a real-world, public dataset composed of 1,560 records collected through online forms and mobile applications. The model achieved an accuracy of 97.76%, a precision of 98.14%, a sensitivity of 98.60%, an F1-score of 98.37%, and an AUC-ROC of 99.79%, demonstrating a high predictive capacity. The tool allows for the generation of an automated, immediate prediction accessible from any device, offering a noninvasive alternative for early ASD risk screening. Although the results are promising, further clinical validation in real-world settings is required to support its use in public health contexts. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2026.

Instituciones: Universidad Peruana de Ciencias Aplicadas