Research Area 4

Simulation and designing countermeasures against possible COVID-19 resurgence: predicting spreading of infection, estimating and verifying the effectiveness of countermeasures, and predicting demands on medical resources and optimizing their allocations

Objective: Modeling and simulation of spreading infection and suppression of SARS-CoV-2, economic impacts of countermeasures, estimation of burden on medical resources is an important policy decision support tool. This part of the research program aims to develop, deploy, and continuously improve a modeling and simulation system to simulate the spread and control of infection in response to the ever-changing dynamics of the situation, and establish methods and operations for systematic data acquisition to make the simulation meaningful.

  • (4-1)To develop a systematic data acquisition and maintenance method for the simulation, which will enable us to obtain epidemiological and medical information such as the mode of transmission of the virus, the accuracy of various tests, the linkage with actual test data reflecting the characteristics of the host response to SARS-CoV-2, the status of implementation of various countermeasures and their effects, as well as a system for the simulation of "elderly people's facilities", "medical facilities", and "night club districts", etc. To enable heterogeneous estimation of the infection situation and simulation of the spread and suppression of infection, which incorporates characteristics of specific situations, industries, and regions, and to evaluate effectiveness of countermeasures. This will incorporate a process for improving the accuracy of the model by modifying it from time to time based on the real-time information obtained from the tests, surveillance, and other data sources.
  • (4-2)Modeling and simulation need to incorporate characteristics of SARS-CoV-2 virus, such as: (a) a small percentage of infected individuals will cause many secondary infections (super spreader), (b) infected individuals will cause transmission to others over a period of time, including asymptomatic periods, and (c) depending on the post-infection period, the probability of detection of viral RNA by PCR testing and antigen and antibody tests results varies, (d) a substantial proportion of infected individuals are asymptomatic, (e) the risk of severe disease exists in the elderly and those with underlying disease, (f) genetic predisposition and factors such as cross-immunity and previous immunizations may change host response, although it has not been validated, and (g) regional age composition, policy, and social norm may greatly affect spreading of virus. It is necessary to be able to incorporate spatio-temporal factors, such as changes in the movement of people due to various factors. Essentially, it is a partially observable stochastic process problem that involves a peculiar probability distribution with a large degree of uncertainty. It is expected that the construction of the framework should be carried out without any preconceived assumptions. It should be noted that multiple medical science hypotheses exist that may affect the model parameters, and at the same time, the simulation results will provide indications as to the validity of the medical science hypotheses.
  • (4-3)In order to estimate and validate the simulation parameters, acquisition, maintenance, continuous operation, and real-time access to various systematic data are essential. For example, real-time or near-real-time access to PCR, antibodies, antigen test results, and certain clinical information is required. Solid platform and operational system should be established. In addition, if simulations and other studies present more promising policy options or important hypotheses to be tested, and access to certain tests and data is needed to test them, a framework that enables such hypothesis-driven action shall be established.
  • (4-4)Development of a realistic Individual Based Model, a multi-agent model, or even a network-based model, or even a combination of multiple methods shall be considered. A fine-grain simulation at the whole population level, using computational resources such as Fugaku HPC system, could also be considered. It would be desirable to be able to systematically compare and verify results based on multiple scenarios, considering their use in policy verification and policy making.
  • (4-5)In addition, assuming the reopening of the international air traffic, a multi-national cooperation model at a certain level of granularity may be necessary. In addition, a system to identify infected people entering the country and to prevent the spread of infection need to be developed and implemented.
  • (4-6)Comparison and evaluation of multiple mathematical models, including the SIR model, and, if effective, integration and combination of these models, and development of a system to provide informative outcome that will serve as a basis for policy decisions.
  • (4-7)Verify the effectiveness of contact tracing for clusters, and to estimate the scale of resources required for the expected spread of infection, and to derive knowledge that will contribute to the planning of more effective countermeasures? Furthermore, is it possible to predict the effectiveness of the measures to be taken against COVID-19 and to verify their effectiveness?
  • (4-8)In addition, can we derive hypotheses for effective countermeasures by using reinforcement learning and adversarial generation methods to plan effective countermeasures?