Uso de IA na modelagem preditiva de surtos de doenças infecciosas: uma revisão sistemática

Abstract

This study addresses the application of artificial intelligence in the predictive modeling of infectious disease outbreaks, highlighting the need to reconcile technical precision, socio-environmental determinants, and transparency in public health. The objective of the study is to perform a systematic literature review to analyze how scientific production characterizes the use of these technological tools in predicting epidemics. The methodology followed structured screening and eligibility guidelines, mapping publications between 2019 and 2026 in specialized databases, resulting in the selection and qualitative evaluation of eighteen final scientific articles. The results indicate the consolidation of three main paradigms: ensemble learning, physics-informed neural networks, and foundation models based on attention mechanisms. Furthermore, it was found that system validation relies on rigorous error and classification metrics, while explainable artificial intelligence emerges as an essential element to mitigate algorithmic opacity. The final considerations point out that predictive models have transcended academic experimentation, becoming strategic surveillance tools. The integration of climate and human mobility data, combined with transparent visual dashboards, provides the technical foundation and legal security for early and safe governmental decision-making.