Development and Internal Validation of a Safety-Gated Artificial Intelligence System for Predicting Serious Bacterial Infection among Children in Northern Nigeria
This study developed and internally validated a safety-gated artificial intelligence framework using Nigerian pediatric clinical data that combines an ensemble prediction model with a local large language model and a deterministic gating engine to identify serious bacterial infections, achieving moderate predictive performance while withholding 20% of predictions to ensure data quality and clinical safety.