Researchers have developed a new artificial intelligence–based modeling tool that can accurately predict how infectious diseases spread in confined environments and help identify more effective containment strategies, according to a study published in the Proceedings of the National Academy of Sciences(PNAS).
The study, conducted by Dmitry Korkin, a WPI computer science professor; Suhas Srinivasan (Stanford University); Jeffrey King (Earthshot Labs); Jacob Collins (UMass Chan Medical School); and Andres Colubri (UMass Chan Medical School), introduces the AI-GIS Infection Dynamics (AGID) model. Unlike traditional epidemic models that rely on population-level averages, the AGID model simulates the movements and behaviors of individual people and incorporates biological data specific to each pathogen, including how infections spread through shared air, surfaces, and close contact. The COVID-19 pandemic highlighted these risks, particularly following major outbreaks on cruise ships and in long-term care facilities.
Using this model, researchers were able to test and propose pathogen-specific improvements to containment measures, including changes to cleaning practices, the wearing of masks, isolation timing, and movement restrictions.
