Thanks to new algorithms and a massive amount of processing power, scientists from IBM, Johns Hopkins University, and the University of California San Francisco have created a powerful new prediction system for dengue fever and malaria. The new models for the open-source Spatio Temporal Epidemiological Modeler (STEM) allow scientists to quickly predict the trajectory of infectious disease outbreaks and create containment plans.
“Recognizing a need to see what the potential spread of a pandemic might be for a given country, geographic region, or world over the course of days, weeks and months, IBM Research started putting some processing muscle into the fight against world health problems by collaborating on a new age of science-based, data-centric disease modeling,” said IBM’s Ari Entin. “We’re collecting data to better understand the extent of health risk behaviors, preventive care practices, and the burden of diseases. These computational models of infectious disease can help public health officials quickly understand and forecast spread of diseases and assist policymakers responsible for strategies to contain disease epidemics.”
The new data tools for STEM, published in two separate papers in the academic journals Malaria and Theoretical Biology, are aimed at improving on existing predictive systems for malaria and dengue fever. According to IBM, the data tools allow for “models on top of models” that take into account weather patterns, human travel patterns, warfare, and more to allow for more accurate prediction of epidemics. STEM, which came out of IBM, was designed so that epidemiologists could make predictive disease models with minimal coding knowledge.