Extended Kalman filter based on stochastic epidemiological model for COVID-19 modelling

Area of Science:

• Epidemiology
• Mathematical Biology
• Computational Science

Background:

• COVID-19 spread is influenced by factors like re-infection and social distancing.
• Traditional models often fail to capture the inherent uncertainties in disease transmission dynamics.
• Accurate modeling is crucial for effective public health interventions.

Purpose of the Study:

• To develop a novel stochastic model for COVID-19 spread analysis.
• To incorporate re-infection and social distancing into a modified SEIRD model.
• To improve the accuracy of COVID-19 spread prediction by addressing uncertainties.

Main Methods:

• A new deterministic Susceptible, Exposed, Infectious, Recovered (Re-infected), and Deceased-based Social Distancing (SEIR(R)D-SD) model was proposed.
• The deterministic model was converted into a stochastic form to handle uncertainties.
• An extended Kalman filter (EKF) was developed for parameter and state estimation.

Main Results:

• The SEIR(R)D-SD model effectively accounts for re-infection and social distancing effects.
• The stochastic approach successfully incorporated uncertainties in COVID-19 spread.
• The EKF provided accurate simultaneous estimation of model parameters and transmission states.

Conclusions:

• The proposed stochastic-based method offers improved accuracy in predicting COVID-19 spread.
• The model's ability to handle re-infection, social distancing, and uncertainty is a significant advancement.
• This approach provides a valuable tool for epidemiological analysis and intervention planning.