Use of remote sensing and artificial intelligence to predict high risk areas for Aedes aegypti and Arbovirus infestation

• Grant number: 20/01596-8
• Support type: Regular Research Grants
• Duration: November 01, 2020 - October 31, 2022
• Field of knowledge: Health Sciences - Collective Health - Public Health
• Principal researcher: Francisco Chiaravalloti Neto
• Grantee: Francisco Chiaravalloti Neto
• Home Institution: Faculdade de Saúde Pública (FSP). Universidade de São Paulo (USP). São Paulo , SP, Brazil
• Assoc. researchers: Antonio Henrique Alves Gomes ; Gerson Laurindo Barbosa ; Jefersson Alex dos Santos ; José Alberto Quintanilha ; Marcia Caldas de Castro ; Marta Blangiardo ; Maurício Lacerda Nogueira ; Monica Pirani ; Valmir Roberto Andrade

Abstract

The current epidemiological scenario in Brazil is worrying since in the last years thousands of cases of dengue (DEN), zika (ZIK) and chikungunya (CHIK) have been reported. These arboviruses and their complications are important public health problems, and studies in the state of São Paulo are extremely fragmented, often unrelated to the triple vector x population x environment. The Aedes aegypti mosquito plays a fundamental role in the spread of all these diseases, but there is a great difficulty in identifying risk areas based only on the traditionally used entomological indicators (Breteau, Building and Containers). Our aim in this study is to develop a model to identify areas of high risk for Ae. aegypti infestation and arbovirose occurence (DEN, ZIK and CHIK) based on the quantification of adult vector females, physical, economic, social and climatic characteristics of the regions. The study will be developed in the urban area of Campinas, State of São Paulo, Br. To reach our aim we will use Artificial Intelligence and deep learning techniques to classify remote sensing images, as well as Bayesian modeling that relate the number of females of Ae. aegypti, as well DEN, ZIK and CHIK cases with socioenvironmental characteristics. Methods for identifying high-risk areas are intended to be developed and a spatial pattern of occurrence is expected to point to higher-risk areas that could provide useful information for control and surveillance activities. These methods, as well as part or all of the results obtained through these technologies, after validation, could be regularly adopted for public health management, optimizing resources and time in identifying risk areas for the occurrence of these diseases, prioritizing the application of surveillance and control measures in these regions. (AU)