A method, apparatus, and storage medium for time-series causal analysis of environmental change and the risk of infectious disease incidence.

Abstract

This invention discloses a time-series causal analysis method, computer device, and storage medium for environmental change and infectious disease incidence risk. The invention can process environmental data time series and infectious disease incidence time series using convergent cross-mapping algorithms, Peter-Clark instantaneous conditional independence tests, and causal forest algorithms, respectively. The obtained first, second, and third causal relationship information represent the causal relationship between the environmental data time series and the infectious disease incidence time series from different perspectives. This allows multiple causal relationship algorithms to complement each other and strengthen the causal relationships identified by each algorithm. It can more effectively explain the causal patterns of environmental change represented by environmental data time series and infectious disease incidence time series represented by infectious disease incidence time series, aiming to provide a more reliable answer to the driving effect of environmental factors on infectious diseases. This invention has wide applications in the field of environmental health technology.

Description

Technical Field

[0001] This invention relates to the field of environmental health technology, and in particular to a time-series causal analysis method, computer device, and storage medium for environmental change and the risk of infectious disease incidence. Background Technology

[0002] A causal relationship exists between the incidence of infectious diseases and environmental changes. For example, with the intensification of climate change, the transmission patterns of infectious diseases may change, and host susceptibility, social factors, and environmental conditions may affect the activity of infectious diseases. A deep understanding of potential environmental drivers has become an important means of preventing infectious diseases. Through time-series causal inference, the causal relationship between environmental changes (such as climate change, land-use change, and changes in the social and demographic environment) and the incidence of infectious diseases can be revealed. This understanding helps identify which environmental factors are key drivers of increased infectious disease risk, and allows for the prediction of future infectious disease trends using historical data, thus providing a scientific basis for public health interventions. Summary of the Invention

[0003] To address the current technical problems of lacking stable standards, low efficiency, and susceptibility to errors in causal analysis of environmental change and infectious disease incidence risk, the present invention aims to provide a time-series causal analysis method, computer device, and storage medium for environmental change and infectious disease incidence risk.

[0004] On one hand, embodiments of the present invention include a time-series causal analysis method for environmental change and the risk of infectious disease incidence, the method comprising the following steps:

[0005] Acquire and process time series environmental data and time series of infectious disease incidence data;

[0006] The environmental data time series and the infectious disease incidence time series are subjected to convergent cross-mapping processing to evaluate the causal impact of candidate driving factors on infectious disease activities and obtain first causal relationship information.

[0007] Peter-Clark instantaneous conditional independence tests were performed on the environmental data time series and the infectious disease incidence time series to obtain the causal dependence between the visual lag and contemporaneous data, thus obtaining second causal relationship information.

[0008] The environmental data time series and the infectious disease incidence time series are processed by the causal forest algorithm to quantitatively analyze the linear and nonlinear relationships between environmental factors and infectious disease activities, and obtain third causal relationship information.

[0009] Based on the first causal relationship information, the second causal relationship information, and the third causal relationship information, the causal relationship between the environmental data time series and the infectious disease incidence time series is determined by comprehensive consideration.

[0010] Further, the convergent cross-mapping processing of the environmental data time series and the infectious disease incidence time series to evaluate the causal impact of candidate driving factors on infectious disease activity and obtain first causal relationship information includes:

[0011] Set the embedding dimension and lag time;

[0012] Embed the environmental data time series into the X phase space;

[0013] Embed the time series of the number of infectious disease cases into the Y-phase space;

[0014] The state in the Y phase space is reconstructed from the neighborhood information in the embedded X phase space, and the Pearson correlation coefficient between the predicted Y and the actually observed Y is calculated.

[0015] The direction and strength of the causal relationship are confirmed based on the Pearson correlation coefficient.

[0016] Once the causality test is passed, a linear model is fitted at each point in the X phase space to obtain the fitting coefficients. Commonly used methods such as heat maps can be used to show the spatiotemporal distribution of the impact of environmental factors on infectious disease activities analyzed by S-map.

[0017] The Pearson correlation coefficient is used as the first causal relationship information.

[0018] Furthermore, the Peter-Clark instantaneous conditional independence test is performed on the environmental data time series and the infectious disease incidence time series to obtain the causal dependency between the visual lag and contemporaneity, thus obtaining second causal relationship information, including:

[0019] Set the maximum time lag value according to information criteria;

[0020] In the PC phase, within the range of the maximum time lag, partial correlation analysis is used as an independence test method. This involves performing conditional independence tests on each pair of variables X and Y, controlling for the influence of other variables, and then calculating the partial correlation coefficients of X and Y to identify potential driving variables (parent nodes) and causal relationships. The MCI phase, building upon the preliminary causality identified in the PC phase, further calculates partial correlation coefficients—that is, calculating the partial correlation coefficient for each lag time of the variable X→Y—to analyze the strength of the lagged causal relationship and determine the causal strength and significance between the environmental data time series and the infectious disease incidence time series corresponding to multiple different lag times.

[0021] Generate a directed acyclic graph based on each of the described causal paths;

[0022] The autocorrelation strength, causal path, strength (lag number), and significance of each variable are determined based on the directed acyclic graph.

[0023] The partial correlation coefficient is used as the second causal relationship information.

[0024] Furthermore, determining the autocorrelation strength, causal path, strength (lag number), and significance of each variable based on the directed acyclic graph includes:

[0025] The directed acyclic graph is visualized to obtain multiple nodes and the edges connecting the nodes;

[0026] The autocorrelation strength of each variable is determined based on the color of the node;

[0027] Determine whether the causal relationship is delayed based on the curvature of the edge;

[0028] The causal strength is determined based on the color of the edge.

[0029] Furthermore, the step of processing the environmental data time series and the infectious disease incidence time series using a causal forest algorithm to obtain third causal relationship information includes:

[0030] Construct a causal forest model;

[0031] The causal forest model was trained using the environmental data time series and the infectious disease incidence time series.

[0032] The trained causal forest model was used to calculate the average treatment effect and the conditional average causal effect.

[0033] The conditional average causal effect is used as the third causal relationship information.

[0034] Furthermore, the construction of the causal forest model includes:

[0035] According to the formula

[0036] Y t =f(X) t )+∈

[0037] Establish the causal forest model; where Y t X represents the data in the time series of the number of cases of the aforementioned infectious disease; t ...

[0038] Further, determining the causal relationship between the environmental data time series and the infectious disease incidence time series based on the first causal relationship information, the second causal relationship information, and the third causal relationship information includes:

[0039] The first causal relationship information, the second causal relationship information, and the third causal relationship information are weighted and fused.

[0040] Based on the weighted fusion results, the causal relationship between the environmental data time series and the infectious disease incidence time series is determined.

[0041] Furthermore, the weighted fusion of the first causal relationship information, the second causal relationship information, and the third causal relationship information includes:

[0042] The causal information obtained from the above three methods (Pearson correlation coefficient, partial correlation coefficient, and conditional average causal effect) is standardized / normalized, and the results are S0. CCM (X→Y), S PCMCI+ (X→Y), S CF (X→Y);

[0043] Based on the applicability or reliability of each method in a specific scenario, assign weights W. CCM W PCMCI+ W CF ;

[0044] Weighting criteria: a. Method characteristics (e.g., CCM is suitable for dynamic and non-linear data, PCMCI+ assumes independence and can control confounding, causal forest assumes random data distribution and provides quantitative effects, etc.); b. Significance of causal association (P-value, methods with higher significance are given greater weight); c. Consistency of results: relationships with higher consistency between methods are given greater weight; d. Data characteristics (e.g., when the sample size is small, causal forest may be weaker, while CCM and PCMCI+ are more robust).

[0045] Overall causal strength = S CCM (X→Y)·W CCM +S PCMCI+ (X→Y)·W PCMCI+ +S CF (X→Y)·W CF

[0046] The combined causal strength is used as the result of weighted fusion.

[0047] On the other hand, embodiments of the present invention also include a computer device, including a memory and a processor, the memory for storing at least one program, and the processor for loading at least one program to execute the time-series causal analysis method of environmental changes and infectious disease incidence risk in the embodiments.

[0048] On the other hand, embodiments of the present invention also include a computer-readable storage medium storing a processor-executable program, which, when executed by a processor, is used to perform the time-series causal analysis method for environmental changes and the risk of infectious disease outbreaks in the embodiments.

[0049] The beneficial effects of this invention are as follows: The time-series causal analysis method for environmental changes and infectious disease incidence risk in the embodiments can use convergent cross-mapping algorithm, Peter-Clark instantaneous conditional independence test algorithm, and causal forest algorithm to process environmental data time series and infectious disease incidence time series, respectively. The obtained first causal relationship information, second causal relationship information, and third causal relationship information represent the causal relationship between environmental data time series and infectious disease incidence time series from different aspects. Finally, the causal relationship between environmental data time series and infectious disease incidence time series is determined based on the first causal relationship information, second causal relationship information, and third causal relationship information. This allows multiple causal relationship algorithms to complement each other and strengthen the causal relationships identified by each algorithm. It can more effectively explain the causal relationship of environmental changes represented by environmental data time series and infectious disease incidence patterns represented by infectious disease incidence time series, in order to provide a more reliable answer to the driving effect of environmental factors on infectious diseases. Attached Figure Description

[0050] Figure 1 This is a schematic diagram illustrating the steps of the time-series causal analysis method for environmental changes and the risk of infectious disease incidence in the embodiments.

[0051] Figure 2 This is a flowchart illustrating the time-series causal analysis method for environmental changes and the risk of infectious disease outbreaks in this embodiment. Detailed Implementation

[0052] When assessing the causal relationship between environmental change and infectious disease incidence, focusing solely on linear or complex nonlinear models may not adequately control for potential temporal confounding factors. For example, using CCM and PCMCI+ without assuming a linear relationship or specific functional form typically results in standardized effect measures (such as correlation coefficients or partial correlation coefficients), emphasizing the existence and directionality of causality, but failing to derive regression coefficients or hazard ratios to quantitatively assess how changes in environmental factors lead to changes in infectious disease incidence.

[0053] Causal forests are a nonparametric method based on random forests, designed to estimate treatment effects and capture potential nonlinear relationships. Causal forests can flexibly handle high-dimensional data, and are particularly suitable for situations with complex interactions and nonlinear effects. Their core idea is to estimate the conditionally averaged causal effect (CATE) by constructing multiple decision trees.

[0054] Therefore, a dynamic data comprehensive analysis method (“CCM”—“PCMCI+”—“CF”) can be constructed to infer the time series causal relationship between environmental change and the risk of infectious disease incidence. This method can solve the problems of causal structure inference and quantitative estimation of influence coefficients. The combination of the three methods can cross-validate the results and provide a more robust causal explanation, in order to provide a more reliable answer to the driving effect of environmental factors on infectious diseases.

[0055] Based on the above principles, this embodiment provides a time-series causal analysis method for environmental changes and the risk of infectious disease outbreaks.

[0056] Reference Figure 1 The time-series causal analysis method for environmental change and the risk of infectious disease incidence includes the following steps:

[0057] S1. Obtain time series of environmental data and time series of infectious disease incidence rates;

[0058] S2. Perform convergent cross-mapping processing on environmental data time series and infectious disease incidence time series to evaluate the causal impact of candidate driving factors on infectious disease activities and obtain first causal relationship information;

[0059] S3. Perform Peter-Clark instantaneous conditional independence tests on environmental data time series and infectious disease incidence time series to obtain visual lag and contemporaneous causal dependencies, and obtain secondary causal relationship information;

[0060] S4. Process environmental data time series and infectious disease incidence time series using causal forest algorithm to quantitatively analyze the linear and nonlinear relationships between environmental factors and infectious disease activities, and obtain third causal relationship information;

[0061] S5. Based on the first causal relationship information, the second causal relationship information, and the third causal relationship information, determine the causal relationship between the environmental data time series and the infectious disease incidence time series.

[0062] In this embodiment, the process of steps S1-S5 is as follows: Figure 2 As shown.

[0063] In step S1, the environmental data time series to be acquired is data related to the natural or social environment, arranged according to its collection time or occurrence time. The time unit for environmental data time series and infectious disease incidence time series can be daily, weekly, etc.

[0064] For example, we can statistically analyze meteorological data (which is related to the natural environment) within a unit of time (e.g., daily or weekly), pollutant data (which is related to the natural environment) within a unit of time, and population movement (which is related to the social environment) within a unit of time, to obtain environmental data time series.

[0065] Taking meteorological data as an example, when executing step S1, meteorological data such as regional unit time (e.g., daily or weekly) average temperature, relative humidity, precipitation, and wind speed can be obtained from data sources such as the National Meteorological Data Center or publicly available atmospheric reanalysis datasets (e.g., the European Centre for Medium-Range Weather Forecasts reanalysis dataset, the U.S. National Center for Environmental Prediction reanalysis dataset). This data can then be used to form environmental data time series arranged by multiple average temperatures, environmental data time series arranged by multiple relative humidities, environmental data time series arranged by multiple precipitation amounts, and environmental data time series arranged by multiple wind speeds.

[0066] Similarly, in step S1, the infectious disease incidence time series is also a time series composed of the incidence of a certain infectious disease (which can be data in the form of the number of cases or the incidence rate, etc.) within multiple unit time periods. Specifically, data on infectious disease cases reported per unit time can be obtained from national or regional disease control and prevention centers, or data on infectious disease etiological surveillance conducted by sentinel hospitals can be collected from local infectious disease sentinel surveillance systems, and data on the number of outpatients, emergency rooms, and inpatients per unit time (or medical record front page data) can be obtained from local medical institutions, and then uniformly converted into an infectious disease incidence time series.

[0067] In step S2, the Convergent Cross-mapping (CCM) method based on State Space Reconstruction (SSR) is used to process the environmental data time series and infectious disease incidence time series obtained in step S1, obtaining the first causal relationship information obtained from the CCM processing. The first causal relationship information can be used to filter preliminary causal relationships, and the resulting correlation coefficient represents the causal influence of candidate driving factors (environmental data time series) on infectious disease activity (infectious disease incidence time series).

[0068] In step S3, the Peter and Clark Momentary Conditional Independence Plus (PCMCI+) test is used to process the environmental data time series and the infectious disease incidence time series obtained in step S1, yielding the second causal relationship information obtained from the Peter and Clark Momentary Conditional Independence Plus test. This second causal relationship information represents the causal dependency between the environmental data time series and the infectious disease incidence time series, indicating both lag and contemporaneous relationships.

[0069] In step S4, the causal forest algorithm (CF) is used to process the environmental data time series and infectious disease incidence time series obtained in step S1 to obtain the third causal relationship information. The third causal relationship information can quantitatively represent the linear and nonlinear relationship between environmental factors (environmental data time series) and infectious disease activity (infectious disease incidence time series).

[0070] In step S5, the causal relationship between the environmental data time series and the infectious disease incidence time series can be determined based on any one, two, or all of the first causal relationship information, the second causal relationship information, and the third causal relationship information.

[0071] In this embodiment, by executing steps S1-S5, the convergent cross-mapping algorithm, the Peter-Clark instantaneous conditional independence test, and the causal forest algorithm can be used to process the environmental data time series and the infectious disease incidence time series, respectively. The obtained first causal relationship information, second causal relationship information, and third causal relationship information represent the causal relationship between the environmental data time series and the infectious disease incidence time series from different aspects. Finally, the causal relationship between the environmental data time series and the infectious disease incidence time series is determined based on the first causal relationship information, the second causal relationship information, and the third causal relationship information. This allows multiple causal relationship algorithms to complement each other and strengthen the causal relationships identified by each algorithm. It can more effectively explain the causal relationship between the environmental changes represented by the environmental data time series and the infectious disease incidence patterns represented by the infectious disease incidence time series, in order to provide a more reliable answer to the driving effect of environmental factors on infectious diseases.

[0072] In this embodiment, when performing step S2, which involves performing convergent cross-mapping processing on the environmental data time series and the infectious disease incidence time series to obtain the first causal relationship information, the following steps can be specifically executed:

[0073] S201. Set the embedding dimension and lag time;

[0074] S202. Embed environmental data time series into X phase space;

[0075] S203. Embed the time series of infectious disease incidence rates into the Y-phase space;

[0076] S204. Reconstruct the state in the Y phase space using the neighborhood information in the embedded X phase space, and calculate the Pearson correlation coefficient between the predicted Y and the actually observed Y;

[0077] S205. Confirm the direction and strength of the causal relationship based on the Pearson correlation coefficient;

[0078] S206. When the causality test is passed, a linear model is fitted at each point in the X phase space to obtain the fitting coefficients. Commonly used heat maps and other methods can be used to show the spatiotemporal distribution of the impact of environmental factors on infectious disease activities analyzed by S-map.

[0079] S207. Use the Pearson correlation coefficient as the primary causal information.

[0080] Before executing steps S201-S207, data loading and preprocessing can be performed. During data loading and preprocessing, environmental data time series and infectious disease incidence time series are loaded, and missing data in the environmental data time series and infectious disease incidence time series are checked and processed. Each data point in the environmental data time series and infectious disease incidence time series is standardized or normalized to eliminate the influence of units.

[0081] Step S201 is the parameter selection step: embedding dimension (E) and lag time (τ), where E ranges from... The lower bound E=2 ensures that the embedding dimension contains at least one external variable. The upper bound is determined based on the time series length n of the environmental data time series and the infectious disease incidence time series, using Leave-One-Out Cross-Validation (LOO-CV). Within the range of testing each E, using univariate predictability (such as correlation coefficient) as an indicator, the optimal embedding dimension E value is found as the value of the embedding dimension to be set in step S201. The lag time τ is determined by the autocorrelation function or cross-correlation function as the value of the lag time to be set in step S201, so as to maximize the predictability of univariate prediction.

[0082] Steps S202-S203 are the steps for constructing the phase space: based on Takens' embedding theorem, the delayed coordinate method is used to embed the driving variable environmental factors, i.e., the environmental data time series, into the X phase space, and the response variable, i.e., the infectious disease incidence time series, into the Y phase space.

[0083] Step S204 is the cross-mapping step: This step uses neighborhood information in one phase space to reconstruct the state in another phase space. The specific steps are as follows:

[0084] a. Select a target point in the X phase space (usually the embedding point corresponding to time t);

[0085] b. Find the k nearest neighbors of the target point selected in step a (usually k = E + 1);

[0086] c. Use the weighted average of the Euclidean distances of these neighboring points found in step b to predict the current time point t, and map the predicted current time point t into the Y phase space; calculate the correlation statistical measure (specifically, the Pearson correlation coefficient) between the mapped (predicted) points in the phase space Y and the points actually observed in the phase space Y.

[0087] Step S205 is the causality test step: analyze the results of the cross-mapping in step S204 to confirm the direction and strength of the causal relationship; specifically, if the correlation statistic approaches 1 (e.g., the correlation statistic is greater than a threshold such as 0.7), it indicates that phase space X can predict phase space Y well, which means that phase space X may have a causal influence on phase space Y. Therefore, the causality test can be determined to be passed, and steps S206 and so on can be executed. Because CCM has convergence, the longer the time series used, the smaller the estimation error of the cross-mapping will be. Therefore, longer environmental data time series and infectious disease incidence time series with larger n can be used.

[0088] Step S206 involves constructing the S-map model: After passing the causality test in step S205, a multivariate S-map is used to fit a linear model at each point in the phase space X. The fitting coefficients of the fitted linear model are obtained, and common methods such as heatmaps are used to visualize the spatiotemporal changes in the impact of environmental factors on infectious disease activity analyzed by the S-map. In step S207, the Pearson correlation coefficient confirms the direction and strength of the causal relationship and can be used as primary causal information.

[0089] In this embodiment, when performing step S3, which involves performing the Peter-Clark instantaneous conditional independence test on the environmental data time series and the infectious disease incidence time series to obtain the second causal relationship information, the following steps can be specifically performed:

[0090] S301. Set the maximum time lag value L according to information criteria. max ;

[0091] In the S302.PC phase, within the range of the maximum time lag value, partial correlation analysis is used as the independence test method. That is, conditional independence is tested for each pair of variables X and Y. After controlling for the influence of other variables, the partial correlation coefficients of X and Y are calculated to identify potential driving variables (parent nodes) and causal relationships. The MCI phase, based on the preliminary causality identified in the PC phase, further calculates the partial correlation coefficients. That is, it calculates the partial correlation coefficients for each lag time of the variable X→Y, analyzes the strength of the lagged causal relationship, and obtains the causal strength and significance between the environmental data time series and the infectious disease incidence time series corresponding to multiple different lag times.

[0092] S303. Generate a directed acyclic graph based on each causal path;

[0093] S304. Determine the autocorrelation strength, causal path, strength (lag number), and significance of each variable based on the directed acyclic graph.

[0094] S305. Use the partial correlation coefficient as the second causal relationship information.

[0095] Before executing steps S301-S305, data loading and preprocessing can be performed. During data loading and preprocessing, environmental data time series and infectious disease incidence time series are loaded, and missing data in the environmental data time series and infectious disease incidence time series are checked and processed. Each data point in the environmental data time series and infectious disease incidence time series is standardized or normalized to eliminate the influence of dimensions.

[0096] Step S301 is the parameter selection step: There may be a lag between the environmental data time series and the infectious disease incidence time series. The maximum time lag value L can be set through information criteria (AIC, BIC). max。 Maximum time lag value L max This represents the maximum lag time between environmental data time series and infectious disease incidence time series, i.e., the lag time ranges from [0, L]. max ].

[0097] Step S302 is the step of identifying PCMCI+ causal relationships: In the PC stage, within the range of the maximum time lag value, partial correlation analysis is used as an independence test method, that is, conditional independence is tested for each pair of variables X and Y. After controlling for the influence of other variables, the partial correlation coefficients of X and Y are calculated to identify potential driving variables (parent nodes) and causal relationships. The MCI stage, based on the preliminary causality identified in the PC stage, further calculates the partial correlation coefficients, that is, calculates the partial correlation coefficients for each lag time of the variable X→Y, analyzes the strength of the lagged causal relationship, and obtains the causal strength and significance between the environmental data time series and the infectious disease incidence time series corresponding to multiple different lag times.

[0098] Steps S303-S304 involve analysis using a directed acyclic graph (DAG): DAGs can visualize lagged and contemporaneous causal relationships between environmental data time series and infectious disease incidence time series. Specifically, curved and straight edges between nodes represent lagged and contemporaneous causal relationships, respectively. Lagged values ​​are labeled on the curved edges, node colors represent autocorrelation strength (i.e., auto-MCI, autocorrelation instantaneous conditional independence), and edge colors represent causal strength estimated through partial correlation (i.e., cross-MCI, cross-cross instantaneous conditional independence). Therefore, in step S305, the partial correlation coefficient can be used as secondary causal information.

[0099] In this embodiment, when performing step S4, which involves processing the environmental data time series and the infectious disease incidence time series using the causal forest algorithm to obtain the third causal relationship information, the following steps can be specifically performed:

[0100] S401. Construct a causal forest model;

[0101] S402. Train the causal forest model using time series of environmental data and time series of infectious disease incidence data;

[0102] S403. Calculate the average treatment effect and conditional average causal effect using the trained causal forest model;

[0103] S404. Use the conditional average causal effect as the third causal relationship information.

[0104] Before executing steps S401-S404, data loading and preprocessing can be performed. During this process, environmental data time series and infectious disease incidence time series are loaded, and missing data in these time series is checked and addressed. Data from both time series are standardized or normalized to eliminate the influence of dimensions. Furthermore, during data loading and preprocessing, holiday effects and day-of-week effects in the environmental data time series and infectious disease incidence time series can be treated as dummy variables, and infectious disease data from the previous time unit can be introduced as lagged variables to control for potential autocorrelation effects.

[0105] Steps S401-S402 are the steps for building and training the causal forest model: In step S401, the constructed causal forest model has...

[0106] Y t =f(X) t )+∈

[0107] The form shown. Where Y... t Data from a time series of infectious disease incidence rates; X t The environmental data time series includes air pollutants, meteorological factors, holiday effects, day-of-week effects, lagged variables, etc., f is a nonlinear relationship function learned through causal forest, and ∈ represents the random error term.

[0108] In step S402, the causal forest model is trained using time series data of environmental data and time series data of infectious disease incidence. Specifically, the nonlinear relationship function f contains multiple parameters, which take values ​​of t = 1, 2, 3...n, thereby generating the data Y... t and X t The input is fed into the causal forest model, and the values ​​of the parameters in the nonlinear relationship function f are adjusted until the calculated random error term ∈ converges, thus completing the training of the causal forest model. The trained causal forest model can calculate the average treatment effect (ATE), that is, use the average treatment effect to measure overall environmental factors (such as atmospheric particulate matter PM2.5). 2.5 It includes functions such as increasing the impact of unit quantity on the risk of infectious disease incidence.

[0109] Step S403 is the statistical analysis step: using the trained causal forest model, the conditional average treatment effect (CATE) under different conditions (e.g., different concentration ranges or meteorological conditions) is further calculated. The conditional average treatment effect can quantitatively assess the causal impact of a specific environmental indicator on the risk of infectious disease incidence. The results analysis may include 95% confidence intervals for both ATE and CATE to clarify the statistical significance of the causal effect.

[0110] In step S404, the conditional average causal effect calculated in step S403 is used as the third causal relationship information.

[0111] In this embodiment, if the causal relationships between the infectious disease incidence time series represented by the first, second, and third causal relationship information and the environmental data time series are consistent, a more comprehensive and reliable causal conclusion can be drawn, providing a more robust causal explanation. If the causal relationships represented by the three causal relationship information are inconsistent, it does not necessarily mean that the causal analysis has failed. A more comprehensive and reliable causal conclusion can be obtained through analytical method assumptions, data processing, external verification, fusion of multiple results, and biological rationality. Specifically, further analysis can be performed through the following steps:

[0112] a. Analytical Method Assumptions: Consider the advantages and disadvantages and applicability of different methods, analyze the degree of matching between the data and the assumptions, and determine which methods have more reliable results. For example, CCM is suitable for nonlinear and large complex dynamic systems, and can clearly define the causal direction between two variables. PCMCI+ is suitable for high-dimensional time series data, assumes that the independence test is valid and can control confounding, and is easy to identify instantaneous and lagged causal relationships. Causal forest assumes that the data is randomly distributed and does not provide a clear causal direction, but can provide quantitative causal estimation coefficients.

[0113] b. Examine the data: Perform sensitivity analysis on data preprocessing and parameter settings to observe whether inconsistent results are caused by these factors;

[0114] c. External validation: Evaluate which method’s causal inference is more reliable by using validation sets and other validation methods to assess their predictive power, robustness, and consistency with known domain knowledge. If the results of a method do not match the majority of methods or domain knowledge, it may be necessary to exclude the results of that method.

[0115] d. Results fusion: The causal relationships inferred from the three methods are combined into a union, which serves as a candidate set for further verification. This may include common causal relationships or causal relationships specific to certain methods, which can be discussed together. Alternatively, different weights can be assigned to the results based on the theoretical basis and applicability of each method's assumptions, and the results can be comprehensively evaluated.

[0116] e. Comprehensive explanation: Clearly identify the source of the discrepancy, such as certain causal relationships being significant in dynamic systems but not captured by the causal forest model; or certain complex interaction effects being discovered by the causal forest but not significant in the Bayesian network; interpret the results in a hierarchical manner, with multiple methods consistently supporting a highly credible causal relationship, and contradictory results requiring further verification of the causal relationship; finally, combine relevant domain knowledge to explain the contradictions.

[0117] In this embodiment, the specific forms of the first causal relationship information obtained by executing steps S201-S207 are Pearson correlation coefficients and visual heatmaps, the specific forms of the second causal relationship information obtained by executing steps S301-S305 are partial correlation coefficients and directed acyclic graphs, and the specific forms of the third causal relationship information obtained by executing steps S401-S404 are conditional average causal effects. They all quantify the causal relationship between environmental data time series and infectious disease incidence time series from their respective perspectives. Therefore, the first causal relationship information, the second causal relationship information, and the third causal relationship information can be dimensionless, so that summation and other operations can be performed between the first causal relationship information, the second causal relationship information, and the third causal relationship information.

[0118] In this embodiment, when performing step S5, which is to determine the causal relationship between the environmental data time series and the infectious disease incidence time series based on the first causal relationship information, the second causal relationship information, and the third causal relationship information, the following steps can be specifically performed:

[0119] S501. Data preprocessing and standardization or normalization of causal information output by all methods;

[0120] S502. Determine the weights;

[0121] S503. Weighted fusion of the first causal relationship information, the second causal relationship information, and the third causal relationship information to determine the causal relationship between the environmental data time series and the infectious disease incidence time series;

[0122] S504. Visualization and Consistency Verification.

[0123] In step S501, the causal direction, causal strength indices (Pearson correlation coefficient, partial correlation coefficient, and conditional average causal effect), and their corresponding significance p-values ​​are collected for the three methods: CCM, PCMCI+, and causal forest. The causal strength indices are standardized or normalized; all p-values ​​are converted to -log(P), which can be used as weights later.

[0124] In step S502, the principles for weight allocation can be based on: a. Method characteristics (e.g., CCM is suitable for dynamically nonlinear data, PCMCI+ assumes the independence test is valid and can control confounding, causal forest assumes random data distribution and provides quantitative effects, etc.); b. Significance of causal association (P-value, methods with higher significance are given greater weight); c. Consistency of results: relationships with higher consistency between methods are given greater weight; d. Data characteristics (e.g., when the sample size is small, causal forest may be weaker, while CCM and PCMCI+ are more robust); Preliminary allocation suggestion: if the data is highly nonlinear, CCM has a higher weight (W). CCM W is 0.5. PCMCI+ W is 0.3. CF (0.2), if the confounding variables are large, the PCMCI+ weight is higher, etc.

[0125] In step S5003, the causal strength of the three methods is calculated as follows: Combined causal strength = S CCM (X→Y)·W CCM +S PCMCI+ (X→Y)·W PCMCI+ +S CF (X→Y)·W CF Confidence intervals are calculated for the weighted results to assess the consistency of results among different methods.

[0126] In step S504, a heatmap or network diagram is generated to visualize the weighted causal relationship strength. Causal relationships with low confidence are further verified and comprehensively interpreted in conjunction with actual data.

[0127] A computer program can be written to execute the time-series causal analysis method for environmental changes and infectious disease incidence risk in this embodiment. The computer program can be written into a computer device or storage medium. When the computer program is read out and run, the time-series causal analysis method for environmental changes and infectious disease incidence risk in this embodiment can be executed, thereby achieving the same technical effect as the time-series causal analysis method for environmental changes and infectious disease incidence risk in the embodiment.

[0128] It should be noted that, unless otherwise specified, when a feature is referred to as "fixed" or "connected" to another feature, it can be directly fixed or connected to the other feature, or indirectly fixed or connected to the other feature. Furthermore, the descriptions of "upper," "lower," "left," and "right" used in this disclosure are only relative to the relative positional relationships of the components of this disclosure in the accompanying drawings. The singular forms "a" and "the" used in this disclosure are also intended to include the plural forms, unless the context clearly indicates otherwise. Moreover, unless otherwise defined, all technical and scientific terms used in this embodiment have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in this embodiment specification is only for describing particular embodiments and is not intended to limit the invention. The term "and / or" as used in this embodiment includes any combination of one or more of the associated listed items.

[0129] It should be understood that although the terms first, second, third, etc., may be used to describe various elements in this disclosure, these elements should not be limited to these terms. These terms are only used to distinguish elements of the same type from each other. For example, a first element may also be referred to as a second element without departing from the scope of this disclosure, and similarly, a second element may also be referred to as a first element. The use of any and all instances or exemplary language (“e.g.,” “such as,” etc.) provided in this embodiment is intended only to better illustrate embodiments of the invention and, unless otherwise required, does not impose a limitation on the scope of the invention.

[0130] It should be recognized that embodiments of the present invention can be implemented or carried out by computer hardware, a combination of hardware and software, or by computer instructions stored in a non-transitory computer-readable storage medium. The method can be implemented using standard programming techniques—including a non-transitory computer-readable storage medium configured with a computer program, wherein such a storage medium causes the computer to operate in a specific and predefined manner—according to the methods and drawings described in the specific embodiments. Each program can be implemented in a high-level procedural or object-oriented programming language to communicate with the computer system. However, if desired, the program can be implemented in assembly or machine language. In any case, the language can be a compiled or interpreted language. Furthermore, for this purpose, the program can run on a programmed application-specific integrated circuit (ASIC).

[0131] Furthermore, the procedures described in this embodiment can be performed in any suitable order unless otherwise indicated by this embodiment or clearly contradicted by the context. The procedures (or variations and / or combinations thereof) described in this embodiment can be executed under the control of one or more computer systems configured with executable instructions, and can be implemented by hardware or a combination thereof as code (e.g., executable instructions, one or more computer programs, or one or more applications) that commonly executes on one or more processors. A computer program includes multiple instructions executable by one or more processors.

[0132] Furthermore, the method can be implemented in any suitable type of computing platform, including but not limited to personal computers, minicomputers, mainframes, workstations, networked or distributed computing environments, standalone or integrated computer platforms, or in communication with charged particle tools or other imaging devices, etc. Aspects of the invention can be implemented as machine-readable code stored on a non-transitory storage medium or device, whether removable or integrated into a computing platform, such as a hard disk, optical read and / or write storage medium, RAM, ROM, etc., such that it is readable by a programmable computer, and when the storage medium or device is read by the computer, it can be used to configure and operate the computer to perform the processes described herein. Furthermore, the machine-readable code, or portions thereof, can be transmitted via wired or wireless networks. The invention of this embodiment includes these and other different types of non-transitory computer-readable storage media when such media comprises instructions or programs that implement the steps above in conjunction with a microprocessor or other data processor. When programmed according to the methods and techniques of the invention, the invention also includes the computer itself.

[0133] A computer program can be applied to input data to perform the functions of this embodiment, thereby transforming the input data to generate output data stored in non-volatile memory. The output information can also be applied to one or more output devices, such as a display. In a preferred embodiment of the invention, the transformed data represents physical and tangible objects, including specific visual depictions of physical and tangible objects generated on the display.

[0134] The above are merely preferred embodiments of the present invention. The present invention is not limited to the above-described embodiments. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention, as long as they achieve the technical effects of the present invention by the same means, should be included within the scope of protection of the present invention. Within the scope of protection of the present invention, the technical solutions and / or implementation methods can have various modifications and variations.

Claims

1. A time-series causal analysis method for environmental change and the risk of infectious disease incidence, characterized in that, The time-series causal analysis method for the relationship between environmental change and the risk of infectious disease incidence includes: Acquire and process time series environmental data and time series of infectious disease incidence data; The environmental data time series and the infectious disease incidence time series are subjected to convergent cross-mapping processing to evaluate the causal impact of candidate driving factors on infectious disease activities and obtain first causal relationship information. Peter-Clark instantaneous conditional independence tests were performed on the environmental data time series and the infectious disease incidence time series to obtain the causal dependence between the visual lag and contemporaneous data, thus obtaining second causal relationship information. The environmental data time series and the infectious disease incidence time series are processed by the causal forest algorithm to quantitatively analyze the linear and nonlinear relationships between environmental factors and infectious disease activities, and obtain third causal relationship information. Based on the first causal relationship information, the second causal relationship information, and the third causal relationship information, the causal relationship between the environmental data time series and the infectious disease incidence time series is determined by comprehensive consideration. The process of performing convergent cross-mapping on the environmental data time series and the infectious disease incidence time series to evaluate the causal impact of candidate driving factors on infectious disease activity and obtain first causal relationship information includes: Set the embedding dimension and lag time; Embed the environmental data time series into the X phase space; Embed the time series of the number of infectious disease cases into the Y-phase space; The state in the Y phase space is reconstructed from the neighborhood information in the embedded X phase space, and the Pearson correlation coefficient between the predicted Y and the actually observed Y is calculated. The direction and strength of the causal relationship are confirmed based on the Pearson correlation coefficient. Once the causality test is passed, a linear model is fitted at each point in the X phase space to obtain the fitting coefficients, demonstrating the spatiotemporal distribution of the impact of environmental factors on infectious disease activities analyzed by S-map. The Pearson correlation coefficient is used as the first causal relationship information; The Peter-Clark instantaneous conditional independence test is performed on the environmental data time series and the infectious disease incidence time series to obtain the causal dependence between the visual lag and contemporaneous data, thus obtaining second causal relationship information, including: Set the maximum time lag value according to information criteria; During the PC phase, within the range of the maximum time lag value, partial correlation analysis is used as an independence test method. That is, conditional independence is tested for each pair of variables X and Y. After controlling for the influence of other variables, the partial correlation coefficients of X and Y are calculated to identify potential driving variables and causal relationships. Generate a directed acyclic graph based on the causal relationships described above; The autocorrelation strength, causal path, strength, and significance of each variable are determined based on the directed acyclic graph. The partial correlation coefficient is used as the second causal relationship information; The step of processing the environmental data time series and the infectious disease incidence time series using a causal forest algorithm to obtain third causal relationship information includes: Construct a causal forest model; The causal forest model was trained using the environmental data time series and the infectious disease incidence time series. The trained causal forest model was used to calculate the average treatment effect and the conditional average causal effect. The conditional average causal effect is used as the third causal relationship information.

2. The time-series causal analysis method for environmental change and infectious disease incidence risk according to claim 1, characterized in that, The determination of the autocorrelation strength, causal path, strength, and significance of each variable based on the directed acyclic graph includes: The directed acyclic graph is visualized to obtain multiple nodes and the edges connecting the nodes; The autocorrelation strength of each variable is determined based on the color of the node; Determine whether the causal relationship is delayed based on the curvature of the edge; The causal strength is determined based on the color of the edge.

3. The time-series causal analysis method for environmental change and infectious disease incidence risk according to claim 1, characterized in that, The construction of the causal forest model includes: According to the formula Establish the causal forest model; wherein, The data refers to the time series of the number of cases of the aforementioned infectious disease; The data refers to the time series data of the environmental data; It is a non-linear relation function learned through causal forest; This is the random error term.

4. The time-series causal analysis method for environmental change and infectious disease incidence risk according to claim 1, characterized in that, Determining the causal relationship between the environmental data time series and the infectious disease incidence time series based on the first causal relationship information, the second causal relationship information, and the third causal relationship information includes: The first causal relationship information, the second causal relationship information, and the third causal relationship information are weighted and fused. Based on the weighted fusion results, the causal relationship between the environmental data time series and the infectious disease incidence time series is determined.

5. The time-series causal analysis method for environmental change and infectious disease incidence risk according to claim 4, characterized in that, The weighted fusion of the first causal relationship information, the second causal relationship information, and the third causal relationship information includes: The Pearson correlation coefficient, partial correlation coefficient, and conditional average causal effect were standardized; the standardized Pearson correlation coefficient was S. CCM (X→Y), the standardized partial correlation coefficient is S PCMCI+ (X→Y), the standardized conditional average causal effect is S. CF (X→Y); Weights are assigned based on the methodological characteristics of Pearson correlation coefficient, partial correlation coefficient, and conditional average causal effect, as well as the significance of causal association and consistency of results; the weight corresponding to the Pearson correlation coefficient is W. CCM The weight corresponding to the partial correlation coefficient is W. PCMCI+ The weight corresponding to the conditional average causal effect is W. CF ; According to the formula Overall causal strength = S CCM (X→Y)·W CCM + S PCMCI+ (X→Y)·W PCMCI+ + S CF (X→Y)·W CF Calculate the overall causal strength; The combined causal strength is used as the result of weighted fusion.

6. A computer device, characterized in that, It includes a memory and a processor, the memory being used to store at least one program, and the processor being used to load at least one program to execute the time-series causal analysis method for environmental change and infectious disease incidence risk as described in any one of claims 1-5.

7. A computer-readable storage medium storing a processor-executable program, characterized in that, The processor-executable program, when executed by the processor, is used to perform the time-series causal analysis method for environmental change and the risk of infectious disease incidence as described in any one of claims 1-5.

Images