Extracting Space-Time Patterns in Infectious Diseases

I discuss the usefulness of space-time data mining and 3D visualization to facilitate the discovery of new, significant space-time patterns and shapes of infectious diseases. I illustrate the concepts on a set of dengue fever cases during an outbreak in an urban environment of Colombia. Monte-Carlo simulations are used to evaluate the impact of positional uncertainty (e.g. geocoding error) and temporal inaccuracies (e.g. delay in reporting) on these patterns. A parallel spatial computing solution is used to reduce the computational effort; recursive decomposition methods mitigate the computational imbalance by adapting to the spatial structure of the data.