Abstract
Background: Pediatric mental health disorders affect approximately 13% of children globally, yet access to specialist evaluation remains severely limited. Artificial intelligence (AI) technologies offer potential to augment screening, diagnostic assessment, and treatment monitoring in pediatric psychiatry. Objective: To systematically review evidence on AI applications in pediatric mental health across the spectrum from population screening to treatment optimization. Methods: Systematic review of studies published 2020–2025 was conducted across five databases. AI tools evaluated included computerized adaptive testing, machine learning classifiers, natural language processing, and wearable sensor analysis. Results: Fifty-eight studies met inclusion criteria. AI screening tools demonstrated sensitivity of 82–94% for depression and anxiety, 78–91% for ADHD, and 85–93% for autism spectrum disorder. Treatment response prediction achieved AUC-ROC of 0.80–0.92 across mood disorders. Conclusion: AI demonstrates meaningful efficacy across pediatric mental health applications while raising important ethical questions regarding data privacy, algorithmic bias, and the appropriate role of technology in child psychiatric care.
References
This work is licensed under a Creative Commons Attribution 4.0 International License.