Predictive Artificial Intelligence Models for Long-Term Caries Risk Based on Early Childhood Oral Microbiome Patterns

Abstract

Early childhood caries is a prevalent chronic disease influenced by complex interactions within the oral microbiome. Understanding microbiome patterns offers a promising avenue for predicting long-term caries risk. This study explores the development of predictive artificial intelligence (AI) models that leverage early childhood oral microbiome profiles to forecast future caries susceptibility. Longitudinal microbiome and clinical data from a pediatric cohort were analyzed using machine learning algorithms, including feature selection and model optimization techniques, to identify key microbial signatures associated with caries progression. Results demonstrate that AI-driven models can accurately stratify children by long-term caries risk, enabling early, personalized preventive interventions. These findings highlight the potential of integrating microbiome-based AI predictions into pediatric dental care to reduce the burden of dental caries.