Prediction Model for Infectious Disease Outbreak Tree in Social Contact Networks

Contact tracing is essential for identifying a population exposed to infectious diseases such as COVID-19. Often, there is a lag between the infection date and confirmatory testing, leading to uncertainties while tracing the potential of disease transmission. Our research proposes to predict the maximum likelihood infection tree across known infected nodes and potential spreaders based on partial available information. Using a social-contact network, we infer the spatiotemporal path of infection for a given pandemic. An integer linear programming method there traces back the infection tree and identifies modalities to disrupt the infection propagation. Here, the novelty lies in accurately predicting the paths of infection travel in real-time with the model leveraging incubation period, multiple contacts and time-varying contact probabilities. The results thus obtained are validated using the Susceptible-Infected-Recovered (SIR) simulation.