Context

The lack of operational knowledge about the spread of infectious diseases during the COVID19 crisis worsened its impact: only when it became widespread could morbidity and mortality be determined according to population segment and other environmental and sociological factors. The WHO stated in 2021 that "the COVID-19 disaster would have been avoided" if atypical pneumonia cases in Wuhan in December 2019 had been used as a risk indicator. There are now 620+ million confirmed cases and 6.5+ million deaths. The rate of spread has been different in each country; however, mitigation policies do not appear to be the only factor behind these variations (e.g. at the start of the pandemic, contagion was much larger than the official figures indicated). This was due to a lack of reliable mechanisms for effectively diagnosing and assessing spread, which remains in some cases, which implies that two confounded phenomena are being measured: increase in number of infections, and ability by health systems to detect them. To address these problems, the research community must make progress in Precision Medicine (PM) and Pandemic Resilience, making it possible to:

  • Extract epidemiological data from past pandemics (open data, statistics, articles…) and create rich semantic representations to be efficiently processed via automated AI methods. 
  • Model pandemic propagation based on past evidence, including SARS-CoV-2, using eXplainable AI (XAI); models will assist health authorities/professionals in decision-making, e.g. select spread-suppression mechanisms, dimension/prioritize medical resources…
  • Generalize the models for unknown future outbreaks of other diseases, creating a common framework that enables to build risk assessment models for potential future pandemics.