Accelerating the discovery of space-time patterns of infectious diseases using parallel computing.

Area of Science:

• Epidemiology
• Computational Science
• Public Health

Background:

• Infectious disease outbreaks require timely monitoring for effective public health interventions.
• Space-time statistical analysis is crucial for understanding disease transmission dynamics but is computationally intensive.
• Increasing data size and complexity necessitate efficient computational methods for outbreak analysis.

Purpose of the Study:

• To develop an adaptive space-time domain decomposition method for parallel computation of kernel density.
• To improve the efficiency of analyzing large-scale infectious disease data.
• To enhance the timely identification and communication of disease event distributions.

Main Methods:

• Developed an adaptive space-time domain decomposition approach for parallel processing.
• Applied the methodology to individual dengue case data from Cali, Colombia (2010-2011).
• Utilized high-performance computing to account for data heterogeneity.

Main Results:

• The parallel implementation achieved significant speedups compared to sequential methods.
• Visualized density values in an interactive 3D environment for intuitive analysis.
• Demonstrated effective identification of uneven space-time disease event distributions.

Conclusions:

• The proposed framework enhances the efficiency of space-time analysis for infectious diseases.
• This approach has the potential to significantly improve real-time outbreak monitoring capabilities.
• Adaptive domain decomposition is effective for parallelizing computationally demanding epidemiological analyses.