Utilização de bases de dados para a vigilância epidemiológica combinando conceitos e técnicas de sistemas dinâmicos, inteligência artificial e ciência de rede.

Abstract

This study explores the application of dynamical systems, statistical approaches, artificial intelligence and network science in the context of epidemiological surveillance, with an emphasis on syndromic surveillance of respiratory infections. A first study was carried out using primary health care data from 27 immediate geographic regions, corresponding to the capitals of the states of Brazil. The integration of artificial intelligence and dynamical systems resulted in the creation of the Mixed Model of Artificial Intelligence and Next Generation, which combines different methods to improve the early detection of outbreaks from time series. Then, a second study was carried out, applying a metapopulation model and concepts from network science. Using mobility information and primary health care data from one of the largest states in Brazil, Bahia, the spatial dissemination of potential respiratory diseases was investigated by identifying propagation hubs, based on a sentinel index. This work directly contributes to the project Alert-Early System for Outbreaks with Pandemic Potential (AESOP), demonstrating the potential of new tools to mitigate the impact of emerging and re-emerging diseases in Brazil.