SUEIR Epidemic Forecasting Model for Policy Impact Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current epidemic forecasting models are inadequate for predicting the spread of infectious diseases in the information age, as they lack the ability to account for prevention and control policies and rely on incomplete data, particularly failing to accurately forecast confirmed cases due to missing parameters for contagion, cure, and mortality rates.
Innovation Solution
An epidemic forecasting model based on the SUEIR model, which classifies populations by infection ability and state, calculates future numbers of unconfirmed, confirmed, and isolated individuals, and uses parameter fitting to forecast confirmed cases, incorporating parameters like isolation rates and infection rates to model the impact of prevention and control policies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional kinetic models are used for epidemic forecasting, then the model structure is simple, but the model cannot account for prevention and control policies and cannot accurately forecast confirmed cases
Solution Approach 1:
The patent segments the traditional SEIR model by introducing a new compartment U (undetected infected individuals) and subdividing the I (infected) compartment into I1 (confirmed but not isolated) and I2 (isolated). This segmentation allows the model to separately track different populations and their interactions with prevention policies, thereby improving forecasting accuracy while maintaining manageable model complexity through structured compartmentalization.
Solution Approach 2:
The patent transforms fixed parameters into time-varying parameters that can reflect policy changes. Specifically, the isolation rate λ and detection rate ε are allowed to vary over time to capture the dynamic impact of prevention and control policies. This parameter flexibility enables the model to adapt to changing policy conditions while maintaining mathematical tractability for forecasting.
2Measurement precision
If only confirmed case data is available for forecasting, then data collection is straightforward, but the exact number of infected individuals cannot be obtained
Solution Approach 1:
The patent introduces U (undetected infected individuals) as an intermediary compartment that bridges the gap between confirmed cases and total infections. This intermediary allows the model to infer the hidden infected population based on observable data (confirmed cases) and policy parameters (detection rate), thereby recovering information that would otherwise be lost and improving measurement precision without requiring direct observation of all infected individuals.
Solution Approach 2:
The model incorporates feedback mechanisms where the detected confirmed cases feed back into the system to update estimates of the total infected population. By using the detection rate ε as a feedback parameter, the model continuously refines its estimate of undetected infections based on observed confirmation patterns, thereby compensating for information loss and improving detection accuracy over time.
3Adaptability or versatility
If simple contagion rates are used, then the model is easy to operate, but the impact of prevention and control policies cannot be realized
Solution Approach 1:
The patent applies local quality by assigning different rates to different compartments: σ for U→I transmission, β for I1→I2 isolation, λ for isolation rate, and ε for detection rate. Each compartment has its own specific rate parameter that reflects local conditions and policy impacts. This localized parameterization allows the model to capture heterogeneous policy effects across different population groups while maintaining operational simplicity through clear parameter definitions.
Solution Approach 2:
The model transforms static contagion rates into dynamic rates that change over time according to policy interventions. The isolation rate λ and detection rate ε are modeled as time-varying parameters that can respond to policy changes. This dynamic approach enables the model to adapt to evolving prevention strategies while maintaining ease of operation through systematic parameter updates rather than structural changes.
Data Source
AI summary
An epidemic forecasting model for policies of prevention, control and isolation is disclosed, including: establishing a SUEIR model; dividing, by the SUEIR model, population according to an ability of infection and a way of infection; classifying patients further into three states: unconfirmed, confirmed but not in isolation, and confirmed and in isolation; calculating a number of people in each state in the future by using the SUEIR model; and accumulating the number of people in the confirmed state to realize the forecasting of the number of confirmed cases. The disclosure not only accurately forecasts the number of confirmed cases, but also explains various policies of prevention and control.