Dengue transmission dynamics prediction by combining metapopulation networks and Kalman filter algorithm

Area of Science:

• Epidemiology
• Network Science
• Computational Biology

Background:

• Accurate prediction of infectious disease outbreaks, like dengue fever, is crucial for effective public health interventions.
• Dengue transmission is influenced by complex, nonlinear interactions between mosquito density, climate, and human movement, which are not fully understood.
• Existing models often lack the integration of these factors for precise spatiotemporal forecasting.

Purpose of the Study:

• To develop and validate a novel spatiotemporal prediction model for dengue fever outbreaks.
• To integrate human mobility patterns within a metapopulation network framework for enhanced transmission modeling.
• To improve prediction accuracy by employing data assimilation techniques.

Main Methods:

• Developed a metapopulation network model incorporating human mobility to simulate dengue's spatial diffusion.
• Utilized the ensemble adjusted Kalman filter (EAKF) data assimilation algorithm to iteratively refine model parameters with observed case data.
• Applied the integrated metapopulation network-EAKF system to retrospective forecast dengue transmission in 12 cities in Guangdong province, China.

Main Results:

• The metapopulation network-EAKF system demonstrated accurate city-level dengue transmission trajectory predictions.
• The model successfully predicted local dengue outbreak magnitude and epidemic peak timing up to 10 weeks in advance.
• Forecasts for peak time, intensity, and total cases were more accurate than isolated city-specific models.

Conclusions:

• The developed metapopulation assimilation framework provides a robust method for accurate dengue fever outbreak forecasting with improved spatiotemporal resolution.
• The system's ability to predict outbreak magnitude and temporal peaks supports better intervention strategies and public risk communication.
• This approach offers a methodological foundation for developing advanced infectious disease surveillance and prediction systems.