Sydney researchers to develop computer models that can predict an epidemic

A team of Australian researchers will use software “agents” modelled on real-life individuals to develop high-precision computer models that can predict where or when an epidemic may strike.

By simulating interactions of these “agents” with work, study and other activities, it is possible to trace how a specific outbreak, originated at a particular point, may develop over time, according to the University of Sydney team.

The software “agents” will be constructed using information from the Australian Census, says study leader Professor Mikhail Prokopenko, who is the Complex Systems Research Group director at the university’s School of Civil Engineering. They will have attributes typical of any individual, such as age, gender, household status, occupation, and sector of employment, as well as a set of health characteristics including the susceptibility and immunity to diseases.

“We want to improve the accuracy of modern epidemiological models. We are aiming to integrate large-scale datasets and the explicit computer simulation of entire populations down to the scale of a single individual,” Prokopenko says.

The team will build the software “agents” with hypothetical scenarios to improve the forecasting of a developing epidemic. By running multiple computer simulations while varying the sources of infection, it is possible to estimate the average social and health impact, as well as focus on specific pathways and patterns of epidemics, explains Professor Pip Pattison, Deputy Vice Chancellor and co-researcher on the project.

“This is very important for a detailed understanding of how diseases may spread in varying circumstances and localities, and for identifying the best ways to locate and curtail the epidemics,” says Pattinson, the specialist in the development and application of mathematical and statistical models for social networks and network processes.

Pattison notes that a recent study estimated the cost of influenza in Australia between $828 million and $884 million per year. This project will provide valuable inputs to estimation and advance planning of required resources, such as hospital beds, vaccinations and transport, he says.

According to Pattinson, the team will specifically aim to develop a computationally efficient large-scale simulation model of disease diffusion on a national-level for Australia.

The project is part of the university’s Complex Systems, a new approach to science, engineering, health and management that studies how relationships between parts give rise to the collective emergent behaviours of the entire system, and how the system interacts with its environment.

In September 2015, a study led by researchers at Harvard T.H. Chan School of Public Health revealed that mobile phone records can be used to predict the geographical spread and timing of an epidemic, particularly dengue fever. Utilising data from a large dengue outbreak in Pakistan in 2013, the largest data set of mobile phone records ever analysed to estimate human mobility, the researchers developed an innovative model that can predict epidemics and provide critical early warning to policy makers.

Contact the writer at feedback@ibtimes.com.au or tell us what you think below.