A virtual virus spread between smartphones could help track disease transmissions in real-time
Throughout the country we have seen states lifting social-distancing measures and spring breakers crowding the beaches of Miami. What is the effect that this will have on the global pandemic? With current technologies we will have to wait until it’s too late to find out. To address this challenge, researchers developed a new scientific machine learning module that could provide real-time monitoring of COVID-19 transmission through a digital “virus” which spreads between smartphones via Bluetooth.
Raj Dandekar, a PhD candidate in the MIT Department of Civil and Environmental Engineering and Christopher Rackauckas, an instructor in the Department of Mathematics along with researchers from a number of universities are tackling the challenge of mitigating disease spread to save lives with their research project, called Safe Blues.
“Safe Blues, was developed to better understand how epidemics spread among the population, says Raj Dandekar, co-author in the research. “One of the biggest challenges in managing the coronavirus pandemic has been a lack of real-time data. The data of a patient being infected and recorded as positive can be one or two weeks.”
This time delay in having real-time data hinders the ability of public health officials to monitor the current situation to predict future viral spread and governments from initiating regulatory measures to control outbreaks.
Population behavior is also changing rapidly due to pandemic fatigue and states lifting social distancing measures and mask mandates, all challenges that becomes harder for public health and government officials to observe, model, and predict where new infections might occur.
In the Safe Blues methodology, virus-like tokens are spread between mobile devices via Bluetooth, similarly to how a biological virus spreads between people. By studying how this virtual virus evolves, we can gain real-time insights into how COVID-19 is spreading through the population.
“This method could be potentially really useful because the virus moves through social networks, says Christopher Rackauckas, Applied Mathematics Instructor in the MIT Department of Mathematics. “The network connection through people’s phones is where you can start to understand the social network effects and then start to observe in real-time where people are interacting with one another and the potential clusters in certain regions.”
Safe Blues uses digital tokens (strands) that are passed between other mobile phones that come in proximity with the virtual virus only when certain data criteria are met, such as length of time and distance. “It mimics the behavior of how a biological virus spreads between people, but in a harmless and traceable way using a machine learning supported epidemiological model,” says Dandekar.
“Safe Blues is not about predicting who’s infected, it’s about predicting population statistics and when infections are going to rise to better predict what type of measures are going to be effective,” added Rackauckas.
The application has the potential to empower policy decision-making for future epidemics to save lives and minimize the economic and social effects around the world.
The researchers have developed Safe Blues into a prototype in the form of an app on Android devices, similar to existing contact tracing apps. They are in the initial stages of launching a campus wide experimental trial at the University of Auckland City Campus.
The research on the model framework is published in Patterns Cell Press journal. Safe Blues is in its early stages and being piloted for efficacy at the University of Auckland. More information about the novel approach is available on the Safe Blues website at safeblues.org