Utilizing Artificial Intelligence Models for Early Detection and Personalized Intervention Strategies in Children with Attention-Deficit/Hyperactivity Disorder: A Narrative Review

Background: Attention-deficit/hyperactivity disorder (ADHD) is a prevalent neurodevelopmental condition characterized by heterogeneous clinical presentation and reliance on subjective diagnostic approaches. Emerging artificial intelligence (AI) models offer potential to improve diagnostic accuracy and support individualized care by integrating multidimensional data sources. Objective: To synthesize current evidence on the application of AI models for early detection of ADHD and their potential role in supporting personalized intervention strategies in pediatric populations through a narrative review approach. Methods: A narrative review was conducted using literature from PubMed, Scopus, Web of Science, and Google Scholar published between 2020 and 2026. Studies examining AI applications in ADHD-related assessment, prediction, monitoring, or intervention were included. Evidence was synthesized thematically across data modalities, model types, clinical applications, and translational considerations. Results: The evidence base was predominantly concentrated in conceptual development (n=12) and diagnostic support (n=9), with fewer studies addressing intervention personalization (n=5). Multimodal AI approaches integrating behavioral, cognitive, and digital data demonstrated greater potential than single-source models. Wearable and ecological monitoring methods showed promise for real-time assessment, while EEG and MRI approaches provided biologically informed insights but faced implementation barriers. Intervention-oriented AI applications, including digital therapeutics and adaptive systems, remain in early stages of development. Conclusion: AI models show strong potential as supportive tools for early ADHD detection and emerging applications in personalized intervention planning. However, current evidence favors diagnostic augmentation over clinical implementation. Future research should prioritize interpretability, validation, and integration into real-world pediatric care pathways.