Highly Complex Modeling Helps Study Epidemics

For the epidemic model, the agents are random walkers that also execute long-distance jumps, and the dimensional plane in which they move is divided into two regions where the force of infection takes different values.

The researchers demonstrated the onset of a local epidemic threshold, as well as a global one, and explained them in terms of mean-field approximations. They also defined the critical role of the agent velocity, jump probability, and density parameters in achieving the conditions for local and global outbreaks. The results were independent of the specific microscopic rules adopted for agent motion, since a similar behavior was also observed for the distribution of agent velocity based on a truncated power law, which is a model often used to fit real data on motion patterns of animals and humans.

The researchers hope the model will more accurately predict who should be vaccinated and isolated first, and what travel restrictions will be most effective in preventing different epidemics, since the actions of the infected are more relevant to the spread of an epidemic than those of healthy individuals who are avoiding a contagious area. But in some kinds of epidemics, ill people who are asymptomatic behave as if they are healthy, selfishly infecting others. The new model seeks to account for these individual reactions. The study was published on October 21, 2014, in Physical Review E.

“Even underdeveloped countries now have electronic devices that quickly spread the word about diseases, and airplanes can carry the infected everywhere nearly as quickly,” said lead author Professor of Mechanical Engineering Alessandro Rizzo, PhD, of UNICT and NYU. “Modeling thus must be improved by accounting for contagions that spread more slowly than travelers and ones that spread more quickly.”

Related Links:
University of Catania
Politecnico di Bari
New York University Polytechnic School of Engineering