Spatio-temporal epidemiology of emerging viruses
Leveraging crowdsourced data and occurrence data to improve early disease detection systems
The field of Spatial Epidemiology has emerged based on the broad agreement that maps for infectious diseases of global importance are important for addressing the transmission potential, limits of transmission and underlying risk profiles, the population at risk of infection, the disease burden, and economic impact of the disease. Mapping infectious diseases can also serve to guide public health research and surveillance efforts, as well as to provide a communication platform for decision makers, enabling them to visualize risks and inform disease control efforts.
This project intends to use social media and web search data—crowdsourced data (CSD)—and official health surveillance data—occurrence data (OD)—to develop novel algorithms that will allows us to achieve higher information quality to generate more reliable conclusions and more accurate predictions about the spatio-temporal spread of diseases. These algorithms will be tested in the context of emerging pathogens of global interest, including dengue virus (DENV), chikungunya virus (CHIKV), yellow fever virus (YFV), Zika virus (ZIKV), Ebola virus (EBOV) and the novel 2019 coronavirus, recently designated as severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)
The overall goal of this study is to acquire new insights into the spatio-temporal epidemiology of emerging viruses—using DENV, ZIKV, CHIKV, YFV, EBOV, and SARS-CoV-2 as case studies—through the combined use of CSD and OD, two information-rich, albeit imperfect data sources, to achieve higher information quality. The major innovations of the project are: 1) the inclusion of the temporal dimension into spatial epidemiology, 2) the use of CSD for fine scale monitoring disease dispersal, 3) the inclusion of mobility/urban demographic and environmental covariates to derive ‘socio-ecological corridors’ that describe likely routes of disease spread after initial outbreaks, and 4) the development of ML algorithms to deal with the massive amounts of CSD and to complement and to improve rule-/model-based approaches.