Modeling and evaluation of disease spread behaviors

Abstract

With the frequent appearance and spread of infectious diseases and their casualties in major populated areas, many researchers and organizations became interested to find different ways to forecast the spread behaviors of these diseases before they occur. This can allow them to better prepare and confine the disease in small regions and therefore reduce the loss of human lives. In this paper we study the behavior and the spread of infectious diseases and model their spread in a city. This study helps generate various prevention and control techniques to build a better and a much safer living environment. This is accomplished by using a combination of geographic information system (GIS) tools and environmental controls to create a scalable, two layered; spatio-temporal Agent-Based Model (ABM) that visualizes disease spread throughout any region by showing the interaction between agents/individuals. This specific model will be tested using data from a one thousand seven hundred kilometer square (1,700 Km2) size city with a population of over three hundred thousand individuals (300,000 people). The first layer was to formulate probabilistic models to create a population, synthesizing a population of around two hundred thousand agents that would interact in a geospatial context. The second layer uses another set of probability distributions to predict a disease spread model based on health information that has been provided by the city's health officials. This model has been created with scalability in mind whether it is for a different geographical location or a larger data set. The simulation results show that this system is very efficient and confirms that it could be used in a larger setting.