Dynamic graph and polynomial chaos based models for contact tracing data analysis and optimal testing prescription

Area of Science:

• Epidemiology and Public Health
• Computational Modeling and Simulation
• Network Science

Background:

• Disease transmission, exemplified by COVID-19, presents challenges in early warning, asymptomatic identification, and resource-limited testing.
• Existing epidemiological models struggle with accurate state estimation and uncertainty quantification under constrained testing capacities.
• Dynamic contact networks are crucial for understanding disease propagation patterns over time.

Purpose of the Study:

• To develop an early warning system for individuals exposed to infectious diseases.
• To identify asymptomatic carriers and optimize testing strategies for limited testing capacities.
• To propose a framework for optimizing resource allocation, such as testing and vaccination, to mitigate disease spread.

Main Methods:

• A dynamic-graph based SEIR (Susceptible-Exposed-Infectious-Recovered) epidemiological model incorporating time-varying individual interactions.
• A diffusion-reaction mechanism to model disease state dynamics within the network.
• Arbitrary Polynomial Chaos Expansion for quantifying state uncertainty and enabling risk assessment.

Main Results:

• The dynamic graph model effectively identifies likely exposed individuals for early warnings, including asymptomatic cases.
• Uncertainty quantification using Polynomial Chaos Expansion allows for optimized testing prescriptions to reduce overall disease uncertainty.
• Simulations demonstrate the framework's performance in managing disease spread and estimating the impact of incomplete contact tracing.

Conclusions:

• The proposed dynamic graph SEIR model with uncertainty quantification provides a robust framework for disease transmission management.
• This approach enhances early detection, asymptomatic identification, and optimizes resource allocation for testing and vaccination.
• The methodology is adaptable for various infectious diseases and can inform public health policy during pandemics.