Need for speed: An optimized gridding approach for spatially explicit disease simulations

Area of Science:

• Epidemiology
• Computational Biology
• Disease Modeling

Background:

• Large-scale, spatially explicit infectious disease models are crucial for public health policy.
• Computational demands of these models limit scenario exploration for effective control strategies.

Purpose of the Study:

• To introduce an efficient computational method for Susceptible-Exposed-Infectious-Removed (SEIR) models.
• To reduce simulation time for large-scale, spatially explicit disease outbreak models without compromising accuracy.

Main Methods:

• Developed a hierarchical infection process operating on groups of spatially related nodes (grid cells).
• Implemented a filtering mechanism to exclude large volumes of susceptible nodes from expensive calculations.
• Optimized grid configuration for efficient landscape partitioning and simulation.

Main Results:

• Achieved up to 500x faster calculations compared to pairwise computation.
• The method performed comparably or better than existing algorithms.
• Enabled large-scale, realistic simulations, e.g., continental USA, for foot-and-mouth disease (FMD).

Conclusions:

• The novel hierarchical method significantly enhances computational efficiency in SEIR models.
• This approach facilitates more extensive scenario analysis for infectious disease surveillance and control.
• The method supports realistic, large-scale spatial simulations for policy-making.