# 4 Simulations Reveal Coronavirus from Different Perspectives

Videos show pandemic simulation, office sneezes, social distancing, AI-based CT-Scans, and respiratory digital twin modeling.

John Blyler

April 16, 2020

Data is critical to understand almost anything in engineering and science. Once you have data, you can begin to create models and simulations that are essential for understanding real-world phenomena.

While world governments rely on mathematical models and simulations to help guide health and economic decisions during the pandemic, scientists and engineers use models to understand and develop ways to fight the disease. The challenge is that much of the information about COVID-19 is still unknown and must be estimated or assumed. Naturally, any model is only as good at its inputs and assumptions, which is why the best models are constantly updated and improved.

What follows is a sample of coronavirus-related models and simulations from several leading tool vendors and organizations. The simulations highlight different types of aspects of COVID-19 spreading, testing and predictions about the future. Here’s a brief description of each one:

• The simulation of an epidemic.

• Office sneezes, which examine the spread of the virus from person-to-person in air-conditioned spaces.

• Modeling a small group of three to show the importance of social distancing.

• iCOVID Initiative, which uses CT-Scan image AI and global cloud-based collaboration.

• Digital twin modeling of respiratory product differences.

 Image Source: Dassault Systemes Simula, 4 Simulation Reveal Coronavirus from Different Perspectives

Simulating an Epidemic

How do organizations like the WHO and CDC use mathematical modelling to predict the growth of an epidemic? This video – although a bit long - does a great job of explaining the details in an easily understood video.

For those wanting a lot more math, the follow-on video introduces the Susceptible-Infected-Recovered or SIR model. It presents a simple system of differential equations to model early exponential growth and similar standard metrics in epidemiology. Real world data shows the current epidemic is well modeled by this kind of exponential growth.

Both are presented by 3blue1brown, or 3b1b for those who prefer less of a tongue-twister. This group presents math with a visuals-first approach. Rather than first deciding on a lesson then putting illustrations to it for the sake of having a video, almost all projects start with a particular visualization, with the narrative and storyline then revolving around this image.

Office Sneeze - Spreading COVID-19 from Person-To-Person in Air-Conditioned Spaces

When a person sneezes, contaminated particles, including aerosols, are projected within the room. In an office building, the particles’ spread can be enhanced by the central heating/air that is used to control the temperature and ensure the comfort of the workers. After a sneeze, the heaviest particles fall to the floor or are carried up to 1m (3 feet) by the turbulent airflow generated by the sneeze. As observed in the video, the contaminated surfaces can then be identified. The aerosols, however, which can be as small as the virus itself, remain in suspension in the air and can potentially get carried across the room or into the ventilation system by the airflow of the central air. The Dassault Systemes Simula simulation supports that viruses can indeed be aerosolized, which is backed up by published research on water droplet size distribution emitted by a sneeze, which was the model’s starting point.

Group of Three Model Shows Importance of Social Distancing

Coronavirus infections are spreading quickly with each passing day. This is putting an enormous strain on healthcare infrastructures around the world. And with no proven treatment in sight, the best way to prevent illness is to avoid being exposed to this virus ideally by staying at home but certainly by keeping your distance. WHO and CDC guidelines advise following “social distancing” measures to reduce the risk of virus transmission from person-to-person. Ansys Computational Fluid Dynamics (CFD) software, Ansys Fluent, is used to model a possible scenario of person to person transmission with a person coughing.

A CAD model of three people standing in proximity is built using SpaceClaim. Spacing of approximately 3 feet (0.9 m) is maintained in the first scenario and then increased to 6 feet (1.8 m) in the second scenario. Coughing into elbow with 3 feet (0.9 m) inter-person distance is considered in the third scenario. A CFD model is prepared by defining floor and side wall as no slip walls, while the remaining boundaries are defined as pressure outlets. Expelled cough droplets are modeled using Discrete Particle Model (DPM). Research has found that coughing generates droplets that travel between 6 and 28 m/s and that these droplets range between 50 and 400 microns in diameter. A velocity of 17 m/s and a diameter distribution of 50 to 400 microns is used to define the initial condition of cough droplets.

iCOVID Initiative – CT-Scan Image AI and Cloud-Based Collaboration

As China continues to lift the Wuhan lockdown, they are showing just how committed they are to ensure that the COVID-19 doesn’t flare up again. As reported in the Global Times, the authorities are ramping up efforts to test people and screen their health, who will be eligible to come back to work. These tests are conducted at hospitals and include nucleic tests –to detect a particular nucleic acid sequence unique to a particular virus or bacteria – and CT scans.

Such tests, especially a CT scan, hold the promise of reliably testing who has COVID-19 and who does not, thus reducing the need for social distancing. Based on a need for fast screening, an ad hoc initiative for AI-assisted, pattern recognition of CT scans has been started. Initially requested by radiologists, the initiative has been gaining more attention, support and data.

“Thanks to the collaboration of many parties, CT scans and the associated reports are sent to an AI tool via a cloudplatform,” explained Jef Vandemeulebroucke, professor of medical image analysis at ETRO, an imec research group of Vrije Universiteit Brussels (VUB), Belgium. “The more data (and annotation), the more valuable will be the result.

 Image Source: iCOVID Initiative, 4 Simulation Reveal Coronavirus from Different Perspectives

Digital twin of respiratory products

The same digital-twin modeling techniques used in support of large engineering design and manufacturing projects is being used to design breathing apparatus especially for COVID-19. Vyaire Medical uses Siemens Simcenter software to do just that.

Many elderly and infirm patients infected with COVID-19 have significant issues breathing and are struggling staying alive. Not only do such patients come in a variety of shapes and sizes (morphology), but they also breathe in a variety of ways: some people are “mouth breathers”, some are “nose breathers”, most of us breath using a combination of both. In addition, each patient has a unique breathing profile, that depends on their lung function and structure, that also needs to be accounted for.

“Traditionally we were forced to use simplified models in the development of a respiratory mask. We would have used a simplified head model for our CFD simulations, with idealized breathing holes for the nose and the mouth,” says Vyaire Medical’s Dr. Christopher Varga, Senior Engineering Fellow and Senior Director of R&D. “Today we are using scans of actual human heads – different morphologies of actual patient features.”

“When it comes to respiratory simulation, we incorporate lung structure and volume using a variety of breathing profiles,” Dr. Varga explains. “We are starting to build a library of patient morphologies, which already consists of representative patient geometries for all of the patient populations: adult, pediatric and infant.”

John Blyler is a Design News senior editor, covering the electronics and advanced manufacturing spaces. With a BS in Engineering Physics and an MS in Electrical Engineering, he has years of hardware-software-network systems experience as an editor and engineer within the advanced manufacturing, IoT and semiconductor industries. John has co-authored books related to system engineering and electronics for IEEE, Wiley, and Elsevier.