An infectious disease risk early warning method based on endogenous bearing force and exogenous input nonlinear coupling

Abstract

The application relates to the technical field of infectious disease risk early warning, and discloses an infectious disease risk early warning method based on endogenous bearing capacity and exogenous input nonlinear coupling, which comprises the following steps: step one: relying on Ross-Macdonald human-mosquito coupling differential equation groups, a next-generation matrix method is used to deduce a basic basic reproduction number containing mosquito external latent period, and then a benchmark reproduction number under no environmental disturbance is obtained based on benchmark environmental parameters; step two: all seven transmission parameters of the model are defined as functions of an environmental vector, an exponential correlation form of the parameters and multiple types of environmental indexes is constructed by taking a biting rate as an example, and after logarithmic transformation and simplification, a reproduction number expression dynamically changing with the environment is obtained. The application couples endogenous environmental bearing capacity in the region and exogenous case input, considers environmental instant effects and biological lag effects, improves a traditional gravity model, avoids linear superposition disadvantages, accurately identifies a high-risk window period of an epidemic, realizes fine early warning of mosquito-borne infectious diseases, and provides a quantitative basis for epidemic prevention resource allocation.

Description

Technical Field

[0001] This invention relates to the field of infectious disease risk early warning technology, specifically to an infectious disease risk early warning method based on the nonlinear coupling of endogenous carrying capacity and exogenous input. Background Technology

[0002] With the acceleration of globalization and urbanization, as well as abnormal changes in the climate and environment, the cross-regional spread of infectious diseases (especially mosquito-borne infectious diseases) is becoming increasingly frequent. Complex population flows between regions make it extremely easy for an outbreak in a single city to escalate into a cross-regional pandemic, posing a serious threat to public health security and socio-economic order. Therefore, accurately predicting the spatiotemporal transmission trends and outbreak risks of infectious diseases is beneficial for governments and disease control departments to identify high-risk areas in advance and take targeted intervention measures to maintain public health resources at the optimal cost.

[0003] Spatial epidemiological models are crucial tools for understanding and controlling the cross-regional spread of infectious diseases. Traditional risk assessment models are often limited to single nodes or assume simple linear transmission paths. When assessing regional comprehensive risk, most basic mathematical models tend to treat influencing factors such as meteorological conditions, land use, and population movement as independent variables and simply add them up. Their underlying logic assumes that "the risk of infectious diseases in a region equals the absolute sum of external inputs and local proliferation." However, this simplistic linear addition hypothesis ignores the significant differences (i.e., "spatiotemporal heterogeneity") between different cities in terms of static ecological backgrounds (such as land use and socioeconomic conditions) and dynamic climatic conditions (such as lagged fluctuations in temperature and humidity). Furthermore, traditional spatial flow models often focus only on the scale of physical inputs, failing to reveal the complex relationship between the inherent biological laws of infectious diseases and environmental carrying capacity. This can easily lead to the illusion of "homogenization" in risk assessment and fail to accurately depict the true outbreak patterns of pathogens in complex and variable environments.

[0004] Based on the above, in order to overcome the limitations of traditional spatial epidemiological models that only focus on a single node or simple linear transmission, this invention proposes a spatiotemporal risk early warning model for infectious diseases based on the nonlinear coupling of endogenous carrying capacity and exogenous input. This invention focuses on and breaks through the conventional approach of independent superposition of factors: First, from the core perspective of "dynamic-static factor interaction," it accurately quantifies the background outbreak potential of a single node (city) in a closed environment, focusing on how static background factors such as land use and medical resources regulate the health effects of virus transmission exposure, as well as the immediate and delayed effects of dynamic exposure factors such as meteorological and air quality factors. Second, for the spread of pathogens in cross-regional spatial networks, we attempt to introduce nonlinear coupling relationships. By decoupling the two independent yet highly correlated driving variables of "external input" (the potential scale of infected persons brought about by cross-regional population movement) and "local carrying capacity" (the objective ability to accept and amplify the epidemic determined by static background and dynamic meteorology), we focus on analyzing how the regional outbreak risk breaks the traditional linear growth when "high-intensity external case inflow" encounters "a local meteorological window period that is extremely suitable for mosquito breeding," producing a nonlinear multiplier effect and soaring exponentially. When applied to the prevention and control of infectious diseases in urban agglomerations under complex environments, this model is more realistic and can more accurately identify the "spatiotemporal resonance window period" that is highly likely to trigger an outbreak. It provides intuitive and quantitative decision-making basis from the management foundation for implementing targeted traffic quarantine, differentiated environmental remediation, and dynamic allocation of limited public health resources.

[0005] To address this issue, we propose an infectious disease risk early warning method based on the nonlinear coupling of endogenous carrying capacity and exogenous input. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides an infectious disease risk early warning method based on the nonlinear coupling of endogenous bearing capacity and external input, thus solving the problems mentioned in the background technology.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a method for early warning of infectious disease risks based on the nonlinear coupling of endogenous bearing capacity and exogenous input, comprising the following steps: Step 1: Based on the Ross-Macdonald human-mosquito coupled differential equations, the basic reproduction number including the mosquito's incubation period is derived using the next-generation matrix method. Then, the baseline reproduction number under no environmental disturbance is obtained based on the baseline environmental parameters. Step 2: Define all seven propagation parameters of the model as functions of the environmental vector. Taking the bite rate as an example, construct the exponential relationship between the parameters and multiple environmental indicators. After simplification by logarithmic transformation, obtain the expression for the regeneration number that changes dynamically with the environment. Step 3: Distinguish between immediate and delayed environmental effects, use variable difference to characterize the magnitude of environmental changes, construct a dynamic factor lag effect formula with the help of a distributed lag model, and combine it with the immediate environmental term correction to obtain a dynamic propagation capacity calculation formula that takes into account the lag. Step 4: Divide the influencing factors into static and dynamic indicators, construct a static two-dimensional feature matrix and a dynamic three-dimensional time series tensor respectively, obtain the standardized comprehensive factor through factor dimensionality reduction and standardization, and calculate the time series difference of each comprehensive factor. Step 5: Based on the divided static and dynamic factor sets, substitute them into the previous formula to construct the regional time-sharing effective regeneration number. Mathematical expression; combining the quantitative correlation between new cases and effective reproduction number, a panel negative binomial regression model is built, and the coefficients of each factor and lag coefficient are solved by regression using historical case data to determine the effective reproduction number of each spatiotemporal unit; Step Six: Improve the traditional spatial gravity model by introducing the effective regeneration number of the target location to represent the local environment's acceptance capacity. Combine this with the scale of outflow of infected individuals from the source and spatial traffic impedance to construct an intercity directional transmission weight formula. Then, accumulate the weights of all outflow directions to calculate the total risk of external input in a single region. Step 7: Couple the effective regeneration number, which characterizes the local endogenous carrying capacity, with the value of external input risk to complete the spatiotemporal risk quantification and early warning of regional infectious diseases.

[0008] Preferably, the Ross-Macdonald model in step one includes formulas for changes in population infection rate and mosquito infection rate, from which the baseline reproduction number of the out-of-band incubation period T is derived. The baseline regeneration number is obtained after fixing the baseline environment values. .

[0009] Preferably, in step two, an exponential function of the bite rate is constructed using vegetation coverage, population density, and the proportion of healthcare workers. After all parameters are exponentialized and logarithmically linearly integrated, the environment-dependent regeneration number is derived. .

[0010] Preferably, in step three, the environmental change is first obtained by performing a difference operation on the original environmental indicators. Then, the cumulative lag effect of the dynamic environment is calculated using a multi-order distributed lag formula. Combined with the immediate environmental index term, a formula for calculating the propagation capacity that takes into account both immediate and lag environmental impacts is obtained. .

[0011] Preferably, the static indicators in step four include land cover, socio-economic, and healthcare indicators, constituting... Static matrix; dynamic indicators include meteorological and air quality indicators, constituting... The time series tensor is factored and standardized using Z-scores. The difference between adjacent time series of the standardized factor is then calculated.

[0012] Preferably, in step five, the set of the first m static factors and the remaining dynamic factors D are distinguished, and the effective reproduction number is constructed using the standardized factor differences:

[0013] Based on the case generation pattern, a panel negative binomial regression was established:

[0014] Coefficients are obtained through regression fitting. , .

[0015] Preferably, the intercity directional propagation weights are obtained from the improved gravity model in step six. Space impedance is adopted Calculate the risk of exogenous input in a single region. .

[0016] Another technical problem to be solved by the present invention is to provide an infectious disease risk early warning system, comprising: Data acquisition module: Collects static land, economic, and medical data, as well as dynamic meteorological and air quality monitoring data for each region; Data preprocessing module: realizes the construction of indicator matrix / tensor, factor transformation, standardization and factor difference calculation; Parameter fitting module: Run panel negative binomial regression to solve for static factor coefficients and dynamic lag coefficients; Bearing capacity calculation module: Substitute fitting parameters for spatiotemporal unit calculation External risk calculation module: Substituting the improved gravity formula into the calculation of intercity propagation weight and regional exogenous input risk ; Risk warning module: Combining local endogenous carrying capacity and external input risks, it outputs the spatiotemporal outbreak warning results of regional infectious diseases.

[0017] Compared with existing technologies, this invention provides an infectious disease risk early warning method based on the nonlinear coupling of endogenous bearing capacity and exogenous input, which has the following beneficial effects: This invention couples the regional carrying capacity of the local environment with the input of exogenous cases, taking into account both the immediate effects of the environment and the delayed effects of biology. It improves the traditional gravity model, avoids the drawbacks of linear superposition, accurately identifies the high-risk window period of the epidemic, realizes the refined early warning of mosquito-borne infectious diseases, and provides a quantitative basis for the allocation of epidemic prevention resources. Detailed Implementation

[0018] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] A method for early warning of infectious disease risk based on the nonlinear coupling of endogenous carrying capacity and exogenous input includes the following steps: Step 1: Based on the Ross-Macdonald human-mosquito coupled differential equations, the basic reproduction number including the mosquito's incubation period is derived using the next-generation matrix method. Then, the baseline reproduction number under no environmental disturbance is obtained based on the baseline environmental parameters. The Ross-Macdonald model includes formulas for changes in population infection rates and mosquito infection rates, from which the baseline reproduction number of the out-of-band incubation period T is derived. The baseline regeneration number is obtained after fixing the baseline environment values. ; Step 2: Define all seven propagation parameters of the model as functions of the environmental vector. Taking the bite rate as an example, construct the exponential relationship between the parameters and multiple environmental indicators. After simplification by logarithmic transformation, obtain the expression for the regeneration number that changes dynamically with the environment. In this step, an exponential function of the bite rate is constructed using vegetation coverage, population density, and the proportion of healthcare workers. After all parameters are exponentialized and logarithmically linearly integrated, the environment-dependent regeneration number is derived. ; Step 3: Distinguish between immediate and delayed environmental effects, use variable difference to characterize the magnitude of environmental changes, construct a dynamic factor lag effect formula with the help of a distributed lag model, and combine it with the immediate environmental term correction to obtain a dynamic propagation capacity calculation formula that takes into account the lag. This step first performs a difference operation on the original environmental indicators to obtain the environmental change, then calculates the dynamic environmental cumulative lag effect using a multi-order distributed lag formula, and combines it with the immediate environmental index term to obtain a formula for calculating the propagation capacity that takes into account both immediate and lag environmental impacts. ; Step 4: Divide the influencing factors into static and dynamic indicators, construct a static two-dimensional feature matrix and a dynamic three-dimensional time series tensor respectively, obtain the standardized comprehensive factor through factor dimensionality reduction and standardization, and calculate the time series difference of each comprehensive factor. Among them, static indicators include land cover, socio-economic and health indicators, which constitute... Static matrix; dynamic indicators include meteorological and air quality indicators, constituting... The time series tensor is factored and standardized using Z-scores. The difference between adjacent time series of the standardized factor is then calculated.

[0020] Step 5: Based on the divided static and dynamic factor sets, substitute them into the previous formula to construct the regional time-sharing effective regeneration number. Mathematical expression; combining the quantitative correlation between new cases and effective reproduction number, a panel negative binomial regression model is built, and the coefficients of each factor and lag coefficient are solved by regression using historical case data to determine the effective reproduction number of each spatiotemporal unit; This step distinguishes between the first m static factors and the remaining dynamic factor set D, and constructs the effective reproduction number using standardized factor differences:

[0021] Based on the case generation pattern, a panel negative binomial regression was established:

[0022] Coefficients are obtained through regression fitting. , .

[0023] Step Six: Improve the traditional spatial gravity model by introducing the effective regeneration number of the target location to represent the local environment's acceptance capacity. Combine this with the scale of outflow of infected individuals from the source and spatial traffic impedance to construct an intercity directional transmission weight formula. Then, accumulate the weights of all outflow directions to calculate the total risk of external input in a single region. This step improves the gravity model to obtain the intercity directional propagation weights. Space impedance is adopted Calculate the risk of exogenous input in a single region. .

[0024] Step 7: Couple the effective regeneration number, which characterizes the local endogenous carrying capacity, with the value of external input risk to complete the spatiotemporal risk quantification and early warning of regional infectious diseases.

[0025] The present invention also includes: an infectious disease risk early warning system, comprising: Data acquisition module: Collects static land, economic, and medical data, as well as dynamic meteorological and air quality monitoring data for each region; Data preprocessing module: realizes the construction of indicator matrix / tensor, factor transformation, standardization and factor difference calculation; Parameter fitting module: Run panel negative binomial regression to solve for static factor coefficients and dynamic lag coefficients; Bearing capacity calculation module: Substitute fitting parameters for spatiotemporal unit calculation External risk calculation module: Substituting the improved gravity formula into the calculation of intercity propagation weight and regional exogenous input risk ; Risk warning module: Combining local endogenous carrying capacity and external input risks, it outputs the spatiotemporal outbreak warning results of regional infectious diseases.

[0026] In summary, this invention derives a baseline reproduction number based on the classic Ross-Macdonald mosquito-borne infectious disease model, constructs a dynamic effective reproduction number by combining environmental factor index correlation, environmental variable difference, and distributed lag modeling, distinguishes between static and dynamic indicators to build feature matrices and time-series tensors, and performs factor standardization. It then uses panel negative binomial regression to calibrate the model correlation coefficients to obtain regional and time-specific effective reproduction numbers to quantify the regional endogenous environmental carrying capacity. Furthermore, it improves the traditional gravity model by introducing local effective reproduction coefficients, population flow scale, and composite traffic impedance to calculate the weight of intercity directional transmission. The cumulative inflow weight across the entire region is used to obtain the risk of external input in each region. Based on the nonlinear coupling relationship between local endogenous carrying capacity and external input risk, it accurately identifies the spatiotemporal resonance window of the epidemic, overcoming the shortcomings of traditional models that simply linearly superimpose factors and ignore regional heterogeneity and biological lag characteristics. This enables refined spatiotemporal risk early warning of mosquito-borne infectious diseases, providing quantitative decision support for differentiated public health management and dynamic allocation of health resources.

[0027] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for early warning of infectious disease risks based on the nonlinear coupling of endogenous carrying capacity and exogenous input, characterized in that, Includes the following steps: Step 1: Based on the Ross-Macdonald human-mosquito coupled differential equations, the basic reproduction number including the mosquito's incubation period is derived using the next-generation matrix method. Then, the baseline reproduction number under no environmental disturbance is obtained based on the baseline environmental parameters. Step 2: Define all seven propagation parameters of the model as functions of the environmental vector. Taking the bite rate as an example, construct the exponential relationship between the parameters and multiple environmental indicators. After simplification by logarithmic transformation, obtain the expression for the regeneration number that changes dynamically with the environment. Step 3: Distinguish between immediate and delayed environmental effects, use variable difference to characterize the magnitude of environmental changes, construct a dynamic factor lag effect formula with the help of a distributed lag model, and combine it with the immediate environmental term correction to obtain a dynamic propagation capacity calculation formula that takes into account the lag. Step 4: Divide the influencing factors into static and dynamic indicators, construct a static two-dimensional feature matrix and a dynamic three-dimensional time series tensor respectively, obtain the standardized comprehensive factor through factor dimensionality reduction and standardization, and calculate the time series difference of each comprehensive factor. Step 5: Based on the divided static and dynamic factor sets, substitute them into the previous formula to construct the regional time-sharing effective regeneration number. Mathematical expression; combining the quantitative correlation between new cases and effective reproduction number, a panel negative binomial regression model is built, and the coefficients of each factor and lag coefficient are solved by regression using historical case data to determine the effective reproduction number of each spatiotemporal unit; Step Six: Improve the traditional spatial gravity model by introducing the effective regeneration number of the target location to represent the local environment's acceptance capacity. Combine this with the scale of outflow of infected individuals from the source and spatial traffic impedance to construct an intercity directional transmission weight formula. Then, accumulate the weights of all outflow directions to calculate the total risk of external input in a single region. Step 7: Couple the effective regeneration number, which characterizes the local endogenous carrying capacity, with the value of external input risk to complete the spatiotemporal risk quantification and early warning of regional infectious diseases.

2. The infectious disease risk early warning method based on the nonlinear coupling of endogenous carrying capacity and exogenous input as described in claim 1, characterized in that, The Ross-Macdonald model described in step one includes formulas for changes in population infection rates and mosquito infection rates, from which the baseline reproduction number during the out-of-band incubation period T is derived. The baseline regeneration number is obtained after fixing the baseline environment values. .

3. The infectious disease risk early warning method based on the nonlinear coupling of endogenous carrying capacity and exogenous input as described in claim 1, characterized in that, In step two, an exponential function of the bite rate is constructed using vegetation coverage, population density, and the proportion of healthcare workers. After exponentialization and logarithmic linear integration of all parameters, the environment-dependent regeneration number is derived. .

4. The infectious disease risk early warning method based on the nonlinear coupling of endogenous bearing capacity and exogenous input as described in claim 1, characterized in that, Step three first performs a difference operation on the original environmental indicators to obtain the environmental change. Then, the cumulative lag effect of the dynamic environment is calculated using a multi-order distributed lag formula. Combined with the immediate environmental index term, a formula for calculating the propagation capacity that takes into account both immediate and lag environmental impacts is obtained. .

5. The infectious disease risk early warning method based on the nonlinear coupling of endogenous bearing capacity and exogenous input as described in claim 1, characterized in that, Step four's static indicators include land cover, socio-economic, and healthcare indicators, constituting... Static matrix; dynamic indicators include meteorological and air quality indicators, constituting Temporal tensor; After factor transformation and Z-score standardization, the difference between adjacent standardized factors is obtained.

6. The infectious disease risk early warning method based on the nonlinear coupling of endogenous bearing capacity and exogenous input as described in claim 1, characterized in that, In step five, the set of the first m static factors and the remaining dynamic factors D are distinguished, and the effective reproduction number is constructed using the standardized factor differences: Based on the case generation pattern, a panel negative binomial regression was established: Coefficients are obtained through regression fitting. , .

7. The infectious disease risk early warning method based on the nonlinear coupling of endogenous bearing capacity and exogenous input as described in claim 1, characterized in that, Step six involves improving the gravity model to obtain the intercity directional propagation weights. Space impedance adopts Calculate the risk of exogenous input in a single region. .

8. An infectious disease risk early warning system based on the method of any one of claims 1 to 7, characterized in that, include: Data acquisition module: Collects static land, economic, and medical data, as well as dynamic meteorological and air quality monitoring data for each region; Data preprocessing module: realizes the construction of indicator matrix / tensor, factor transformation, standardization and factor difference calculation; Parameter fitting module: Run panel negative binomial regression to solve for static factor coefficients and dynamic lag coefficients; Bearing capacity calculation module: Substitute fitting parameters for spatiotemporal unit calculation ; External risk calculation module: Substituting the improved gravity formula into the calculation of intercity propagation weights and regional exogenous input risks ; Risk warning module: Combining local endogenous carrying capacity and external input risks, it outputs the spatiotemporal outbreak warning results of regional infectious diseases.

Images