Early detection of bioterrorism: Monitoring disease diffusion through a multilayered network

Bioterrorism is a significant threat to the United States. Biological agents such as anthrax and viral hemorrhagic fever may present symptoms in the human population in a similar manner as more common illnesses such as a cold or flu. As a result, a bioterrorism attack that uses one of these biological agents would be difficult to quickly distinguish from a seasonal flu outbreak. We propose a conceptual framework for differentiating between bioterrorism and epidemic scenarios by monitoring disease diffusion in a population. We model a human population and the locations they visit as a multilayered network and simulate the spread of a disease in the two scenarios. Then, we use functional principal components analysis to characterize disease transmission. We show that our method accurately predicts a bioterrorism event for several types of populations and provides a warning of an attack when only a small proportion of the population is infected.