News



COVID-19: Multiscale modelling provides new insights into the global dynamics of the disease


Background
The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), first identified in December 2019 in Wuhan, China, has since spread throughout the world resulting in more than 6.63 million reported infections and 391,000 deaths. As a vaccine is not expected before 2021, currently several non-pharmaceutical interventions such as quarantine, social-distancing, lockdown, shutdowns etc. have been implemented world-wide. However, this has led to unprecedented slowdown of economic activity. There is a lot at stake here. Within this context, there are several open questions.


  • • Are the mitigation measures helping to decelerate COVID-19?

  • • Can we use available country-specific data of reported COVID-19 cases to predict whether lockdown measures have to be extended, improved or relaxed?

  • • Will there be a second wave?

  • • What is the influence of asymptomatic cases?






Strategy
A novel multiscale mathematical model called the Lattice-SIRQL that takes into account spatial interactions, lockdown measures and quarantine procedures has been proposed for the analysis of the global dynamics of the infection. Using reported data from 105 countries, country-specific model parameters were computed using a global minimization algorithm. According to this procedure, country-specific non-pharmaceutical measures such as lockdown, quarantine and testing is implicitly taken into account. To study the influence of lockdown intensity, lockdown duration and relaxation rate at small geographic scales, an individual-based model was developed that takes into account complex interactions at the scale of a single individual.






Figure 1: Simulation and comparison of the evolution of a pandemic without mitigation (left column), with quarantine only (center column) and with lockdown (right column) using the Lattice-SIRQL model. The top most curve is the global evolution of the susceptibles and the active infecteds in the population while the 2D images are spatially resolved simulation of the fraction of active infections per cluster after 10, 20 and 30 days.


Findings
Large scale spatially resolved simulations showed that global indicators for COVID-19 can be misleading and that a possibility of a second wave cannot be ruled-out. Using a data-driven approach, country specific recommendations for lockdown continuation and lockdown relaxation have been made (see link). With regards to lockdown procedures, simulations using the individual-based model showed that inefficient lockdown procedures could simply delay the infection. Such sub-optimal measures will not reduce the number of peak infections (i.e. flatten the curve). Moreover, model predictions showed that asymptomatic cases can significantly extend the lifetime of the disease. This was recently confirmed by the WHO on 9 June. This was few days after our paper was uploaded on the medArxiv server.






Figure 2: Number of confirmed cases (coloured markers) vs model (black line). The red vertical line corresponds to 1st June 2020 and the blue vertical line correspond to 1st August 2020. For a full list of model predictions for 105 countries see Paper.




For detailed information see the paper:
Timothy, Holla & Meschke:
Multiscale Dynamics of COVID19 and Model-Based Recommendations for 105 countries
doi:10.1101/2020.06.05.20123547
https://www.medrxiv.org/content/10.1101/2020.06.05.20123547v1







Contact



Dr.-Ing. Jithender J. Timothy
timothy.jithenderjaswant@rub.de
+49 234 / 32-29070



Vijaya Holla
vijaya.holla@ruhr-uni-bochum.de



Prof. Dr. techn. Günther Meschke
guenther.meschke@rub.de
+49 234 / 32-29051




https://www.medrxiv.org/content/10.1101/2020.06.05.20123547v1

Lecture Dates...

The lecture dates for the sommer term 2024 are online.

more...

Recent publication on a DeepONet for fusing simula...

The new open access article with the title "A multi-fidelity deep operator network (DeepONet) for fu

more...
MORE NEWS