Methods for predicting the risk of poisonous mushroom poisoning
By constructing an exposure-response relationship model between meteorological factors and the number of cases of poisonous mushroom poisoning, and combining it with meteorological forecast data, the problem of rapidly and accurately predicting the risk of poisonous mushroom poisoning was solved, improving prediction accuracy and early warning effectiveness, and reducing the risk of poisoning.
Patent Information
- Application Number
- CN202611417380.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-14
- Publication Date
- 2026-11-10
AI Technical Summary
Current technology lacks a method for rapidly and accurately predicting the risk of poisonous mushroom poisoning, which prevents health authorities and the general public from obtaining timely and effective health warnings, resulting in a high incidence and mortality rate of poisonous mushroom poisoning.
By acquiring the daily number of cases and meteorological monitoring data of the target area, an exposure-response model was constructed between meteorological factors and the number of cases of poisonous mushroom poisoning. The risk of poisoning in the future period was predicted using meteorological forecast data and assessed in conjunction with the number of cases in historical periods.
It enables accurate and rapid prediction of the risk of poisonous mushroom poisoning, improves the accuracy of prediction and the efficiency of monitoring and early warning, and reduces the incidence and mortality of poisoning.
Smart Images

Figure CN123069326A_ABST
Abstract
Description
Technical Field
[0001] This disclosure relates to the field of disease prediction technology, and in particular to a method for predicting the risk of poisonous mushroom poisoning. Background Technology
[0002] Poisonous mushrooms, also known as toxic fungi or poisonous mushrooms, belong to the macrofungi class. Some varieties have high mortality rates. Due to the lack of effective methods for quickly distinguishing between poisonous and edible mushrooms, ordinary residents often suffer acute poisoning from mistakenly picking and eating poisonous mushrooms. This type of poisoning exhibits regional and seasonal patterns, often showing characteristics of family clusters and group outbreaks, posing a significant social threat. Mushroom poisoning has become one of the most fatal foodborne illnesses in my country. Its clinical manifestations are complex and diverse. Most patients initially present with gastrointestinal symptoms such as nausea, vomiting, abdominal pain, and diarrhea. Subsequently, depending on the toxins contained in the ingested mushroom, different target organ damage may occur, even leading to organ failure and death.
[0003] Because there is currently a lack of rapid and accurate methods for predicting poisonous mushroom poisoning, national and provincial health departments mostly rely on past cases of poisonous mushroom poisoning to issue health alerts during peak seasons in their respective areas. These alerts often focus on the biological morphology of common poisonous mushrooms, symptoms of poisoning, and emergency treatment measures, lacking specific warnings based on local natural conditions regarding potentially poisonous mushrooms. Furthermore, the timing and frequency of these alerts are largely based on experience, failing to provide timely and accurate health warnings to health departments and the general public.
[0004] This disclosure primarily aims to provide risk prediction and early warning information for poisonous mushroom poisoning to managers, technical personnel, and the general public. For managers, it can provide early warning information to national, provincial, and municipal health departments, enabling timely prevention and control of poisonous mushroom poisoning and reducing the incidence and mortality rates. For technical personnel, it can provide empirical evidence for further research on the risk of poisonous mushroom poisoning based on meteorological factors. For the general public, it can provide timely early warnings of poisonous mushroom poisoning, effectively reducing the risk of poisonous mushroom poisoning.
[0005] Therefore, accurately and quickly predicting the risk of poisoning from poisonous mushrooms is of great significance to the health and lives of residents. Summary of the Invention
[0006] In view of this, the purpose of this disclosure is to provide a method, apparatus and electronic device that can accurately and quickly predict the risk of poisoning from poisonous mushrooms.
[0007] To achieve the above objectives, this disclosure adopts the following technical solution: acquiring the daily number of cases and meteorological monitoring data of the target area within a preset historical period, as well as the meteorological forecast data of the target area within a preset future period, wherein both the meteorological monitoring data and the meteorological forecast data include daily data of multiple types of meteorological factors; using the meteorological forecast data as an input variable, based on a pre-constructed exposure-response relationship model between meteorological factors and the number of cases of poisonous mushroom poisoning, and the daily number of cases within the historical period, predicting the daily number of cases of poisonous mushroom poisoning in the target area attributable to meteorological factors within the preset future period; and based on the predicted daily number of cases of poisonous mushroom poisoning in the target area attributable to meteorological factors within the preset future period, assessing the risk of poisonous mushroom poisoning in the target area within the preset future period, and obtaining assessment data.
[0008] The beneficial effects of the method for predicting the risk of poisonous mushroom poisoning provided in this disclosure are as follows: By using an exposure-response relationship model between target meteorological factors and the number of cases of poisonous mushroom poisoning, the correspondence between meteorological factors and the number of cases of poisonous mushroom poisoning can be characterized. By combining the daily number of cases in the target area during historical periods with meteorological forecast data for a preset future period, the daily number of cases of poisonous mushroom poisoning attributable to target meteorological factors in the target area during a preset future period can be accurately and quickly predicted. This can improve the accuracy of poisonous mushroom poisoning prediction and help improve the efficiency of monitoring and early warning work. Attached Figure Description
[0009] The above and other objects, features and advantages of this disclosure will become clearer from the following description of embodiments of this disclosure with reference to the accompanying drawings.
[0010] Figure 1 This is a flowchart illustrating an embodiment of the method for predicting the risk of poisonous mushroom poisoning according to the present disclosure;
[0011] Figure 2 This is a flowchart illustrating the process of determining an exposure-response relationship model according to one embodiment of the method for predicting the risk of poisonous mushroom poisoning disclosed herein;
[0012] Figure 3 This is a flowchart illustrating the process of determining the parameters of the cross-basis function and confounding variables in a model according to one embodiment of the method for predicting the risk of poisonous mushroom poisoning disclosed herein.
[0013] Figure 4 This is a flowchart illustrating the process of determining the parameters of the cross-basis function and confounding variables in a model according to yet another embodiment of the method for predicting the risk of poisonous mushroom poisoning disclosed herein.
[0014] Figure 5 This is a flowchart of confidence interval estimation in one embodiment of the method for predicting the risk of poisonous mushroom poisoning according to the present disclosure;
[0015] Figure 6 This is a graphical visualization of the results of the method for predicting the risk of poisonous mushroom poisoning according to the present disclosure;
[0016] Figure 7 This is a flowchart for estimating the number of excess cases in one embodiment of the method for predicting the risk of poisonous mushroom poisoning according to this disclosure;
[0017] Figure 8 This is a flowchart illustrating the process of building a model according to one embodiment of the method for predicting the risk of poisonous mushroom poisoning disclosed herein. Detailed Implementation
[0018] The present disclosure is described below based on embodiments. Those skilled in the art should understand that the accompanying drawings provided herein are for illustrative purposes and are not necessarily drawn to scale.
[0019] Unless the context explicitly requires it, the words "comprising," "including," and similar terms throughout the specification and claims should be interpreted as encompassing rather than being exclusive or exhaustive; that is, meaning "including but not limited to."
[0020] like Figure 1 As shown, this disclosure provides a method for predicting the risk of poisonous mushroom poisoning, which may include the following steps.
[0021] Step S100: Obtain the daily number of cases in the target area within a preset historical period and the weather forecast data for the target area within a preset future period.
[0022] The meteorological forecast data includes daily data on multiple types of meteorological factors, such as daily data on rainfall, temperature, and humidity.
[0023] In this embodiment, the target area refers to the area to be predicted, which may be an administrative region or a natural region.
[0024] As an example, monitoring records of poisonous mushroom poisoning cases in the target area during a preset historical period can be collected in advance from a foodborne disease surveillance system, and the daily number of cases during that preset historical period can be determined. Simultaneously, weather forecast data for the target area during a preset future period can be obtained from the China Meteorological Data Network. This weather forecast data can include daily data on multiple types of meteorological factors; for example, if the preset future period is the next 60 days, the weather forecast data could include daily temperature, humidity, precipitation, etc., for the next 60 days.
[0025] The preset historical time period must be of a certain length to meet the model fitting requirements. The preset future time period can be set according to actual needs (such as the computing performance of the equipment and the required accuracy of the prediction), for example, it can be two weeks, one month, etc.
[0026] Step S200: Using meteorological forecast data as input variables, based on the pre-constructed exposure-response relationship model between the target meteorological factor and the number of cases of poisonous mushroom poisoning, and the daily number of cases in a preset historical period, predict the daily number of cases of poisonous mushroom poisoning in the target area attributable to the target meteorological factor in a preset future period.
[0027] In this embodiment, based on the close correlation between the growth patterns of poisonous mushrooms and meteorological factors, an exposure-response model can be constructed between target meteorological factors and the number of cases of poisonous mushroom poisoning. This model can then be used to characterize the correspondence between meteorological factors and the number of cases of poisonous mushroom poisoning. As an example, the exposure-response model between meteorological factors and the number of cases of poisonous mushroom poisoning can employ a DLNM model (Distribution Lag Nonlinear Model). The target meteorological factor can be a pre-defined meteorological factor or can be determined from multiple meteorological factors based on actual conditions.
[0028] In this way, meteorological forecast data can be used as input variables, and combined with the daily number of cases in the target area during a preset historical period, the daily number of cases of poisonous mushroom poisoning in the target area attributable to the target meteorological factors in a preset future period can be predicted.
[0029] In an optional implementation of this embodiment, the types of poisonous mushrooms that have been previously reported in the target area and the number of cases for each type can be determined based on the monitoring records of poisonous mushroom poisoning cases. Then, one or more types of poisonous mushrooms with the highest number of cases are added to the evaluation data as the types of poisonous mushrooms with the highest risk of poisoning. This enriches the information contained in the evaluation data, thereby providing higher reference value for the early warning and monitoring of poisonous mushrooms.
[0030] In this embodiment, a pre-constructed exposure-response relationship model between target meteorological factors and the number of cases of poisonous mushroom poisoning is used to characterize the close relationship between the growth pattern of poisonous mushrooms and meteorological factors. By combining the daily number of cases in the target area during historical periods with meteorological forecast data for a preset future period, the daily number of cases of poisonous mushroom poisoning attributable to target meteorological factors in the target area during the preset future period can be accurately and quickly predicted. This can improve the accuracy of poisonous mushroom poisoning prediction and help improve the efficiency of monitoring and early warning work.
[0031] As mentioned above, the target meteorological factor can be a preset meteorological factor, such as precipitation, or it can be determined from multiple meteorological factors based on historical meteorological monitoring data of the target area as the target meteorological factor. In some embodiments, such as Figure 2 As shown, the methods for determining the exposure-response relationship model between target meteorological factors and the number of cases of poisonous mushroom poisoning include:
[0032] S210. Based on the daily number of cases and meteorological monitoring data within a historical period, exposure-response relationship models were constructed for multiple candidate meteorological factors.
[0033] In this step, an exposure-response model is constructed for each candidate meteorological factor. Candidate meteorological factors can be selected from precipitation, relative humidity, average wind speed, sunshine duration, daily average temperature, daily minimum temperature, and daily maximum temperature. When constructing the model for each candidate meteorological factor, it can be used as the core variable, with at least some other candidate meteorological factors, day of the week variables, and holiday variables as covariates. A generalized Poisson regression combined with a distributed lag nonlinear model is used for fitting to construct the exposure-response model for that candidate meteorological factor. Specifically, a separate cross-basis function is constructed for each candidate meteorological factor, and at least some of the other meteorological factors are included in the model as confounding factors using natural cubic spline functions.
[0034] For example, the multiple candidate meteorological factors include temperature-related meteorological factors and non-temperature-related meteorological factors. Non-temperature-related meteorological factors may include, for example, precipitation, relative humidity, average wind speed, and sunshine duration. Temperature-related meteorological factors may include, for example, daily average temperature, daily minimum temperature, and daily maximum temperature. If the candidate meteorological factor serving as the core variable is a temperature-related meteorological factor, the covariates are all non-temperature-related meteorological factors. If the candidate meteorological factor serving as the core variable is a non-temperature-related meteorological factor, the covariates are other non-temperature-related meteorological factors besides the core variable, as well as the daily average minimum temperature and the daily average minimum temperature.
[0035] S220. Based on a candidate exposure-response relationship model of multiple candidate meteorological factors, the target meteorological factor is determined from the multiple candidate meteorological factors, and an exposure-response relationship model between the target meteorological factor and the number of cases of poisonous mushroom poisoning is determined.
[0036] In this step, exposure-response relationship models and professional knowledge can be combined. For example, exposure-response relationship models can be used to find that increased rainfall can increase the risk of mushroom poisoning. At the same time, from a professional point of view, rain is conducive to mushroom growth. After comprehensively comparing different meteorological factors, the optimal factor that is meaningful to the model and can be professionally explained can be selected.
[0037] In another embodiment, the target meteorological factors can be determined using a quantitative screening algorithm based on the multi-factor distributed lag nonlinear model (DLNM) constructed in S210. Specifically, the significance hypothesis of the overall effect parameters of the candidate exposure-response relationship models of all candidate meteorological factors is tested, the model P-value corresponding to each core meteorological factor is calculated, and a significance threshold of P < 0.05 is set. Meteorological factors that pass the significance test and have statistically effective associations are screened, while pseudo-correlated meteorological variables that have no significant lag nonlinear associations and are invalid as a whole are eliminated, forming a preliminary set of effective meteorological factor candidates. This ensures the statistically true correlation between the screened variables and the incidence of poisonous mushroom poisoning. Then, the model is optimized by quantifying the fitting effect and lag effect. Specifically, the core effect indicators of the candidate exposure-response relationship models of each candidate meteorological factor are extracted, including the main effect amplitude of the meteorological factor on the incidence of poisonous mushroom poisoning, the duration of the lag effect, and the nonlinear fit degree. The overall fitting effect of different meteorological factor models is compared. The optimal model is selected first, based on its effect characteristics closely matching the growth and disease development patterns of poisonous mushrooms, its strong fitting curve stability, and the absence of invalid fluctuations. The corresponding meteorological factors are then selected as the optimal candidate meteorological factors. This effectively avoids inferior models with poor fitting and weak effects, ensuring the effectiveness and adaptability of the prediction model.
[0038] Based on the quantitative screening results obtained from the above algorithm, the final verification was completed by combining the professional mechanisms of poisonous mushroom growth and reproduction and wild poisoning exposure, eliminating statistically significant but biologically contradictory outliers. For example, if the algorithm screens precipitation as the optimal factor, the rationality of the positive correlation between meteorological factors and the number of poisoning cases is verified by combining the professional mechanisms of increased soil moisture after rain, concentrated germination of poisonous mushroom fruiting bodies, and increased risk of accidental ingestion by people in the wild. Finally, meteorological factors that simultaneously satisfy statistical significance, optimal model effect fit, and explainable professional mechanisms are determined as target meteorological factors.
[0039] This step utilizes an algorithm system that combines quasi-Poisson distribution modeling, statistical significance screening, model effect optimization, and professional mechanism verification to accurately adapt to the lagged and nonlinear effects of meteorological factors on the onset of poisonous mushroom poisoning. It abandons the subjectivity of traditional single-experience judgment, and achieves standardized and normalized screening of target meteorological factors and related models. This provides core model support for the accurate prediction of the daily number of poisonous mushroom poisoning cases in the future, effectively improving the accuracy of risk prediction and the reliability of early warning.
[0040] In this embodiment, the association between meteorological factors and the number of cases of poisonous mushroom poisoning can be characterized by cross-basis functions. By using at least some other candidate meteorological factors, day of the week variables, and holiday variables as covariates, the exposure-response relationship model between meteorological factors and the number of cases of poisonous mushroom poisoning can more accurately represent the correspondence between meteorological factors and the number of cases of poisonous mushroom poisoning, thereby improving the accuracy of subsequent predictions of the risk of poisonous mushroom poisoning based on meteorological factors.
[0041] In some embodiments, the exposure-response relationship model between the target meteorological factor and the number of cases of poisonous mushroom poisoning is as follows:
[0042] Log[E(Y t )]=α+ns(Conf t ,3)+ns(time, df×y)+γDOW t +δHOLIDAY t +βVar t,l
[0043] In the formula, This represents the expected number of cases of mushroom poisoning on day t. Represents the intercept term; ns represents the natural spline smoothing function; Conf t This indicates confounding factors, which include multiple meteorological factors other than the target meteorological factor. For example, as mentioned earlier, meteorological factors include temperature-related meteorological factors and non-temperature-related meteorological factors. When the target meteorological factor is a temperature-related meteorological factor, the confounding factors include all non-temperature-related meteorological factors. When the target meteorological factor is a non-temperature-related meteorological factor, the confounding factors include other non-temperature-related meteorological factors besides the target meteorological factor, as well as the daily average minimum temperature. `ns(time, df×y)` is used to control the long-term time trend, where `df=3` and `y` is the number of years covered by the historical time period. `γ` represents... The estimated values of the regression coefficients; The day of the week effect is represented by a dummy variable; δ represents... The estimated values of the regression coefficients; Dumb variables representing the holiday effect; Var t,l The cross-basis function representing the target meteorological factors of the day is uniformly set to a maximum lag of 7 days. Both the exposure dimension and the lag dimension adopt natural cubic spline functions, and the degrees of freedom are set to 3. The effect parameter represents the target meteorological factor.
[0044] In this embodiment, The day-of-the-week effect can be controlled numerically; for example, when day t is Monday, Tuesday, ... The values can be 1, 2, 3, ... 7. The holiday effect can be controlled numerically; for example, when day t is a holiday. The value of can be l, when day t is a non-holiday. The value of can be 0.
[0045] Cross-basis functions are used to characterize the association between target meteorological factors and the number of cases of poisonous mushroom poisoning. For example, when the predictor variable is precipitation, Var t,l This indicates the correlation between precipitation and the number of cases of poisonous mushroom poisoning, with a lag of, for example, 7 days.
[0046] Understandably, exposure-response relationship models for other candidate meteorological factors can also be constructed with reference to the target meteorological factor.
[0047] After determining the exposure-response relationship model, the sensitivity of the exposure-response relationship model between the target meteorological factor and the number of cases of poisonous mushroom poisoning can be analyzed. Based on the analysis results, the robustness of the exposure-response relationship model between the meteorological factor and the number of cases of poisonous mushroom poisoning can be evaluated, and the parameters of the model can be optimized based on the robustness evaluation results to improve the prediction accuracy of the model.
[0048] As an example, sensitivity analysis can be performed by changing the maximum lag time and degrees of freedom of confounding factors in the exposure-response relationship model between the target meteorological factor and the number of cases of poisonous mushroom poisoning.
[0049] In this embodiment, the robustness of the parameters of the exposure-response relationship model between the target meteorological factor and the number of cases of poisonous mushroom poisoning is assessed by sensitivity analysis, thereby improving the reliability of the model's prediction results.
[0050] After determining the exposure-response model of the target meteorological factor, the following formula can be used to predict the daily number of cases of poisonous mushroom poisoning in the target area attributable to the target meteorological factor in a preset future period:
[0051]
[0052] In the formula, The daily number of cases of poisonous mushroom poisoning attributed to the target meteorological factor is represented; E represents the average daily number of cases of poisonous mushroom poisoning during the historical period, determined based on the daily number of cases during the historical period; e is a natural constant. To predict the cumulative effect function value under meteorological conditions, The total cumulative association coefficient extracted from the exposure-response model. These are the daily target meteorological factor observations in the meteorological forecast data. The optimal meteorological factor level corresponding to the lowest risk of mushroom poisoning is obtained by solving the exposure-disease risk curve fitted by the exposure-response relationship model. For optimal exposure level The corresponding cumulative effect function value.
[0053] To quantify the uncertainty of the prediction results, the Monte Carlo simulation method (≥1000 simulations) can be used to estimate the 95% confidence interval of the risk of disease and the number of cases attributable to the disease. By repeatedly sampling the key parameters of the model (such as cross-basis function coefficients, spline node positions, etc.), the predicted value is recalculated after each sampling. Finally, the upper and lower limits of the confidence interval are determined by the ranking method to ensure the reliability and reference value of the prediction results (see the description of the following embodiments).
[0054] In one embodiment, after constructing the exposure-response relationship model, the parameters of the cross basis functions and the degrees of freedom of confounding factors in the exposure-response relationship model can be adjusted. For example... Figure 3 As shown, for example, it may include the following steps:
[0055] Step 310: Arrange and combine the parameters of the cross basis function and the degrees of freedom of the confounding factors in the exposure-response relationship model between the meteorological factors corresponding to the target area and the number of cases of poisonous mushroom poisoning within their respective ranges, and determine the model fit evaluation value corresponding to each combination of values.
[0056] Step 320: Based on the numerical combination corresponding to the model fitting effect evaluation value with the smallest value, adjust the parameters of the cross basis function and the degrees of freedom of the confounding factors in the exposure-response relationship model between meteorological factors and the number of cases of poisonous mushroom poisoning in the target area.
[0057] In this embodiment, the model fit evaluation value is used to characterize the fit of the initial exposure-response relationship model. The smaller the model fit evaluation value, the higher the predictive accuracy of the exposure-response relationship model between meteorological factors and the number of cases of poisonous mushroom poisoning. As an example, the model fit evaluation value corresponding to each combination of values can be determined based on Akaike's Information Criterion for quasi-Poisson (QAIC).
[0058] In a specific example, the exposure dimension of the cross-basis function can be a natural cubic spline function with three equally spaced nodes and a degree of freedom of 3. The lag dimension can be a natural cubic spline function with one equally spaced node and a degree of freedom of 3. The degree of freedom of the time trend control variable can be set to 3, and the degree of freedom of other meteorological factors besides the predictor variable can also be set to 3.
[0059] In this embodiment, the overall model fitting effect of the target area is evaluated, and the parameters of the cross basis function and the degrees of freedom of the confounding factors in the initial exposure-response relationship model corresponding to the target area are adjusted according to the evaluation results. This ensures that the overall model fitting effect of the target area is optimal, thereby ensuring that the exposure-response relationship model between meteorological factors and the number of cases of poisonous mushroom poisoning can more accurately represent the correspondence between meteorological factors and the risk of poisonous mushroom poisoning.
[0060] In some alternative implementations of this embodiment, the target region includes multiple sub-regions, such as... Figure 4 As shown, the adjustment method may also include the following steps.
[0061] Step S410: Obtain monitoring data for each sub-region, and determine the exposure-response relationship model between meteorological factors and the number of cases of poisonous mushroom poisoning in each sub-region based on the monitoring data.
[0062] As an example, the target area can be a provincial-level administrative region, and the sub-regions can be municipal-level administrative regions within the target area. An exposure-response relationship model is constructed for each sub-region. By applying generalized Poisson regression combined with a distributed lag nonlinear model at the city level, the exposure-response relationship model between meteorological factors and the number of cases of poisonous mushroom poisoning can be obtained for each sub-region.
[0063] Step S420: Arrange and combine the parameters of the cross basis function and the degrees of freedom of the confounding factors in the exposure-response relationship model between meteorological factors and the number of cases of poisonous mushroom poisoning in each sub-region within their respective ranges, and determine the model fitting effect evaluation value corresponding to each combination of values, thereby obtaining multiple model fitting effect evaluation values for each sub-region.
[0064] Step S430: Based on the multiple model fitting effect evaluation values corresponding to each sub-region, arrange and combine the model fitting effect evaluation values corresponding to all sub-regions, and determine the sum of all model effect evaluation values corresponding to each combination.
[0065] Step S440: Based on the combination of model effect evaluation values corresponding to the minimum sum of evaluation values, determine the parameters of the cross-basis function and the degrees of freedom of the confounding factors corresponding to the combination method, and use them as the parameters of the cross-basis function and the degrees of freedom of the confounding factors in the exposure-response relationship model between meteorological factors and the number of cases of poisonous mushroom poisoning in the target area and each sub-area respectively.
[0066] In this embodiment, an exposure-response model is constructed to connect meteorological factors and the number of cases of poisonous mushroom poisoning in each sub-region. The parameters of the cross-basis function and the degrees of freedom of the confounding factors are combined to determine multiple model performance evaluation values for each sub-region. These evaluation values are then combined and summed to obtain a total sum, which characterizes the prediction accuracy of the exposure-response model at the target region level. Finally, the combination of model performance evaluation values for each sub-region that minimizes the sum of the evaluation values is determined, and the corresponding cross-basis function is then used to determine the appropriate combination. The parameters and degrees of freedom of the confounding factors of the function are used as the parameters and degrees of freedom of the cross basis function in the exposure-response relationship model between meteorological factors and the number of cases of poisonous mushroom poisoning in the target region and each sub-region. This allows us to obtain the exposure-response relationship model between meteorological factors and the number of cases of poisonous mushroom poisoning, as well as the regional exposure-response relationship model between meteorological factors and the number of cases of poisonous mushroom poisoning in each sub-region. This ensures that the exposure-response relationship model between meteorological factors and the number of cases of poisonous mushroom poisoning, as well as the regional exposure-response relationship model between meteorological factors and the number of cases of poisonous mushroom poisoning in each region, can more accurately characterize the correspondence between meteorological factors and the risk of poisonous mushroom poisoning.
[0067] based on Figure 4 The implementation shown can assess the risk of mushroom poisoning in a target area within a preset future time period in the following way: assess the risk of mushroom poisoning in each sub-area within the preset future time period to obtain assessment data for each sub-area; and determine the assessment data for the target area based on the assessment data of all sub-areas.
[0068] As an example, after using an exposure-response relationship model between regional meteorological factors and the number of cases of poisonous mushroom poisoning to predict each sub-region and obtain assessment data for each sub-region, a multivariate meta-analysis algorithm can be used to merge the assessment data of each sub-region to obtain the overall assessment data of the target region, which helps to further improve the accuracy of risk prediction.
[0069] In some embodiments, the method of the present invention for predicting the risk of poisonous mushroom poisoning can also use an interval estimation approach to predict the number of cases in a target area using an exposure-response relationship model between meteorological factors and the number of cases of poisonous mushroom poisoning. Figure 5 As shown, the steps may include the following.
[0070] Step S510: Based on the Monte Carlo algorithm, simulate the parameters of the cross basis function of the target meteorological factors to obtain a preset number of estimated parameters.
[0071] The parameters of the cross-basis function for the predictor variables include the regression coefficients of the cross-basis function parameters and their covariance matrix.
[0072] Step S520: Substitute each estimated parameter into the exposure-response relationship model between the target meteorological factor and the number of cases of poisonous mushroom poisoning, and predict the number of cases of poisonous mushroom poisoning in the target area attributable to the target meteorological factor in a preset future period under multiple climate models, and obtain the prediction results corresponding to each estimated parameter under multiple climate models.
[0073] As an example, climate models may include those developed by the China Meteorological Administration (BCC-CSM1-1, BCC-CSM1-1-M), Beijing Normal University (BNU-ESM), and the First Institute of Oceanography of the State Oceanic Administration of China (FIO-ESM).
[0074] Step S530: The number of cases of poisonous mushroom poisoning on the same date in the prediction results corresponding to each estimated parameter under multiple climate models is summed and averaged to obtain the number of cases of poisonous mushroom poisoning attributable to the target meteorological factor corresponding to that estimated parameter.
[0075] Here, the same date can be the same day, the same week, the same month, or other time units.
[0076] Step S540: Based on the predetermined confidence interval, perform statistical analysis on the number of poisonous mushroom poisoning cases attributable to the target meteorological factors corresponding to all estimated parameters, and obtain the daily number of poisonous mushroom poisoning cases attributable to the target meteorological factors in the target area during the preset future time period.
[0077] As an example, the Monte Carlo algorithm can be used to simulate the two-dimensional cross-basis function parameters (corresponding regression coefficients and their covariance matrix) corresponding to precipitation (i.e., the predictor variable) in the exposure-response model between meteorological factors and the number of cases of mushroom poisoning. Assuming that the above parameters follow a multivariate normal distribution, 1000 pairs of estimated parameters for the best linear unbiased prediction (BLUP) fitting the relationship between precipitation and the number of cases can be generated. Then, combined with daily precipitation data under different meteorological models, the attributable number of cases related to daily precipitation in 1000 preset future periods under each climate model is predicted. Next, the attributable number of cases related to daily precipitation is integrated by date and the mean is calculated to obtain 1000 attributable number of cases of cases due to precipitation on different dates, and the interval from 2.5% to 97.5% is determined as the 95% confidence interval (eCI).
[0078] In practice, the uncertainty in predicting the number of poisonous mushroom poisoning cases using an exposure-response model based on meteorological factors mainly stems from the correlation between the independent variable and the number of cases. To reduce this uncertainty, this embodiment can estimate the confidence interval of the prediction results from the exposure-response model based on meteorological factors, thereby improving the reliability of the poisonous mushroom poisoning risk assessment.
[0079] In some optional implementations of this embodiment, Figure 5 The illustrated embodiments may also include the following steps.
[0080] Step S550: Based on the confidence interval and the exposure-response relationship model between the target meteorological factors in each region and the number of cases of poisonous mushroom poisoning, statistical analysis is performed on the number of cases of poisonous mushroom poisoning in each sub-region in the preset future time period to obtain the statistical analysis results.
[0081] Step S560: Based on the statistical analysis results and the geographical location information of each sub-region, a spatiotemporal map is generated through graphical visualization processing. The spatiotemporal map represents the risk of mushroom poisoning in each sub-region within a preset future time period.
[0082] Among them, the spatiotemporal map can represent the risk of poisonous mushroom poisoning in each sub-region in the form of images within a preset future time period.
[0083] In a specific example, the target area can be a provincial-level administrative region, and the sub-regions can be municipal-level administrative units. First, map data corresponding to the target area can be obtained, which may include the ranges corresponding to each sub-region. After estimating the number of cases of poisonous mushroom poisoning in each sub-region, the predicted number of cases can be marked within the range corresponding to each sub-region on the map, thus obtaining a spatiotemporal map of the region.
[0084] For example, based on a pre-determined risk level and corresponding color label, the risk level of each sub-region can be determined according to the predicted number of cases of poisonous mushroom poisoning in each sub-region. Then, the image range corresponding to each sub-region in the map data is marked with the color corresponding to the risk level, thus obtaining a spatiotemporal map of the region. Figure 6 An exemplary graphical visualization result is shown.
[0085] In some embodiments, the method disclosed herein for predicting the risk of poisonous mushroom poisoning can also utilize an exposure-response model between meteorological factors and the number of cases of poisonous mushroom poisoning to estimate the excess number of cases attributed to meteorological factors, such as... Figure 7 As shown, the process may include the following steps.
[0086] Step S610: Based on the exposure-response relationship model between meteorological factors and the number of cases of poisonous mushroom poisoning, determine the relative risk of the target area under different values of the meteorological factors represented by the predictor variables.
[0087] Step S620: Determine the value of the meteorological factor represented by the predictor variable when the relative risk is the lowest as the reference value of the predictor variable.
[0088] Step S630: Compare the reference values with the daily data of the predicted variables in the weather forecast data, and determine the days that are higher than the reference values and the days that are lower than the reference values as unsuitable days.
[0089] Step S640: Based on the exposure-response relationship model between meteorological factors and the number of cases of poisonous mushroom poisoning, predict the number of cases of poisonous mushroom poisoning corresponding to unsuitable days, so as to estimate the number of excess cases attributed to meteorological factors represented by the predictor variables.
[0090] In this embodiment, a reference value for the predictor variable is determined based on the relative risk level. By comparing the reference value with the daily data of the predictor variable in the meteorological forecast data, an unsuitable day is determined. Then, the number of excess cases corresponding to the unsuitable day is predicted. This represents the number of excess cases attributed to the meteorological factors represented by the predictor variable, which can provide data support for studying the correlation between the meteorological factors represented by the predictor variable and the number of cases of poisonous mushroom poisoning.
[0091] In one embodiment, the method for predicting the risk of poisonous mushroom poisoning further includes:
[0092] Based on the exposure-response relationship model of multiple candidate meteorological factors, the risk of mushroom poisoning in the target area in a preset future period is assessed, and the poisoning risk of each candidate meteorological factor is obtained.
[0093] As an example, a range of case numbers corresponding to each risk level can be pre-defined, with a higher number of cases corresponding to a higher risk level. Then, based on the predicted daily number of poisonous mushroom poisoning cases attributed to meteorological factors in the target area within a preset future period, the risk level corresponding to each time interval of the target area in the preset future period can be determined as assessment data for the target area. Here, the time interval can be a day, a week, or a month.
[0094] In one embodiment, based on a constructed exposure-response relationship model of multiple candidate meteorological factors, the risk of mushroom poisoning in a target area within a preset future time period is assessed to obtain the poisoning risk of each candidate meteorological factor. This includes: obtaining the effect parameters in the exposure-response relationship model of each candidate meteorological factor. These effect parameters are the parameters in the constructed exposure-response relationship model. Each candidate meteorological factor corresponds to an effect parameter. Thus, the relative risk (RR) of each candidate meteorological factor can be determined based on the effect parameter corresponding to each candidate meteorological factor. The relative risk (RR) is used to represent the magnitude of the risk. RR = exp(β), where exp(x) is an exponential function representing e to the power of x.
[0095] In one embodiment, the target area includes multiple sub-regions. The target area may be a provincial-level administrative region, and the sub-regions may be municipal-level administrative units. For each sub-region, an exposure-response relationship model for each candidate meteorological factor is constructed. For example, the exposure-response relationship model can be constructed by referring to the method used to construct the target meteorological factor. Then, based on the effect parameters in the exposure-response relationship models for each candidate meteorological factor corresponding to each sub-region, the effect parameters for each candidate meteorological factor in the target area are determined. For example, for each candidate meteorological factor, the mean of the effect parameters corresponding to that candidate meteorological factor in each sub-region is used as the effect parameters for each candidate meteorological factor in the target area.
[0096] In one embodiment, a multivariate meta-regression analysis method with restricted maximum likelihood estimation is used to merge the effect parameters in the exposure-response relationship models of each candidate meteorological factor corresponding to each sub-region, thereby obtaining the effect parameters of each candidate meteorological factor in the target region. By merging the effect parameters corresponding to the same meteorological factor in each sub-region through meta-merging, it is ensured that the final effect parameters can more accurately reflect the degree of danger in the target region.
[0097] In some embodiments, the method disclosed herein for predicting the risk of poisonous mushroom poisoning may further include the following steps: predicting the number of excess cases in each sub-region attributable to the meteorological factors represented by the predictor variable based on an exposure-response relationship model between meteorological factors and the number of cases of poisonous mushroom poisoning in each region; and generating a spatiotemporal map by graphical visualization processing based on the geographical location information of each sub-region and the number of excess cases attributable to the meteorological factors represented by the predictor variable.
[0098] In this embodiment, the exposure-response relationship model between regional meteorological factors and the number of cases of poisonous mushroom poisoning in each sub-region can be used to predict the excess number of cases in the sub-region, which can be used as the risk of poisonous mushroom poisoning in the sub-region in a preset future period. Based on this, a spatiotemporal map can be generated, which can intuitively present the risk of poisonous mushroom poisoning in each sub-region.
[0099] It will be readily understood by those skilled in the art that, without conflict, the above-mentioned preferred solutions can be freely combined and superimposed.
[0100] The above are merely preferred embodiments of this disclosure and are not intended to limit this disclosure. Various modifications and variations can be made to this disclosure by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this disclosure should be included within the protection scope of this disclosure.
Claims
1. A method for predicting the risk of poisonous mushroom poisoning, characterized in that, The method includes: The system acquires the daily number of cases in the target area within a preset historical period and the weather forecast data for the target area within a preset future period. The weather forecast data includes daily data for multiple types of meteorological factors. Using the meteorological forecast data as input variables, and based on a pre-constructed exposure-response relationship model between the target meteorological factor and the number of cases of poisonous mushroom poisoning, as well as the daily number of cases within the preset historical period, the daily number of cases of poisonous mushroom poisoning in the target area attributable to the target meteorological factor is predicted in the preset future period.
2. The method according to claim 1, characterized in that, The method for determining the exposure-response relationship model between the target meteorological factor and the number of cases of poisonous mushroom poisoning includes: Based on the daily number of cases and meteorological monitoring data during the historical period, exposure-response relationship models were constructed for multiple candidate meteorological factors. Based on the candidate exposure-response relationship model of multiple candidate meteorological factors, the target meteorological factor is determined from the multiple candidate meteorological factors, and the exposure-response relationship model between the target meteorological factor and the number of cases of poisonous mushroom poisoning is determined.
3. The method according to claim 2, characterized in that, The exposure-response relationship models were constructed for multiple candidate meteorological factors, including: Using the candidate meteorological factors as core variables, and at least some other candidate meteorological factors, day of the week variables, and holiday variables as covariates, a generalized Poisson regression combined with a distributed lag nonlinear model is used for fitting to construct an exposure-response relationship model for the candidate meteorological factors.
4. The method according to claim 3, characterized in that, Multiple candidate meteorological factors include temperature-related meteorological factors and non-temperature-related meteorological factors. If the candidate meteorological factor used as the core variable is a temperature-related meteorological factor, then all non-temperature-related meteorological factors are used as covariates. If the candidate meteorological factor used as the core variable is a non-temperature-related meteorological factor, then other non-temperature-related meteorological factors besides the core variable and the daily average minimum temperature are used as covariates.
5. The method according to claim 2, characterized in that, The exposure-response model between the target meteorological factor and the number of cases of poisonous mushroom poisoning is as follows: Log[E(Y t )]=α+ns(Conf t ,3)+ns(time,df×y)+γDOW t +δHOLIDAY t +βVar t,l In the formula, This represents the expected number of cases of mushroom poisoning on day t. Represents the intercept term; ns represents the natural spline smoothing function; Conf t The term represents confounding factors, which include multiple meteorological factors other than the target meteorological factor. ns(time, df×y) is used to control the long-term time trend, where df=3 and y is the number of years covered by the historical time period; γ represents... The estimated values of the regression coefficients; The day of the week effect is represented by a dummy variable; δ represents... The estimated values of the regression coefficients; Dumb variables representing the holiday effect; Var t,l The cross-basis function representing the target meteorological factors of the day is uniformly set to a maximum lag of 7 days. Both the exposure dimension and the lag dimension adopt natural cubic spline functions, and the degrees of freedom are set to 3. The effect parameter represents the target meteorological factor.
6. The method according to claim 5, characterized in that, Using the meteorological forecast data as input variables, and based on a pre-constructed exposure-response relationship model between the target meteorological factor and the number of cases of poisonous mushroom poisoning, as well as the daily number of cases during the historical period, the daily number of cases of poisonous mushroom poisoning in the target area attributable to the target meteorological factor in the preset future period is predicted. This includes: predicting the daily number of cases of poisonous mushroom poisoning in the target area attributable to the target meteorological factor in the preset future period using the following formula: ; In the formula, The daily number of cases of poisonous mushroom poisoning attributed to the target meteorological factor is represented; E represents the average daily number of cases of poisonous mushroom poisoning during the historical period, determined based on the daily number of cases during the historical period; e is a natural constant. To predict the cumulative effect function value under meteorological conditions, The total cumulative association coefficient extracted from the exposure-response model is given. The daily target meteorological factor observation values in the aforementioned meteorological forecast data. The optimal meteorological factor level corresponding to the lowest risk of mushroom poisoning is obtained by solving the exposure-disease risk curve fitted by the exposure-response relationship model. For optimal exposure level The corresponding cumulative effect function value.
7. The method according to claim 5, characterized in that, The method further includes: Based on the exposure-response relationship model of multiple candidate meteorological factors, the risk of mushroom poisoning in the target area during the preset future time period is assessed to obtain the poisoning risk of each candidate meteorological factor.
8. The method according to claim 7, characterized in that, The exposure-response model based on multiple candidate meteorological factors is used to assess the risk of mushroom poisoning in the target area during the preset future time period, obtaining the poisoning risk of each candidate meteorological factor, including: Obtain the effect parameters in the exposure-response relationship model for each of the candidate meteorological factors; The relative hazard of each candidate meteorological factor is determined based on the effect parameters corresponding to each candidate meteorological factor.
9. The method according to claim 8, characterized in that, The target area includes multiple sub-regions, and for each sub-region, an exposure-response relationship model for each candidate meteorological factor is constructed. Based on the effect parameters in the exposure-response relationship model of each candidate meteorological factor corresponding to each sub-region, the effect parameters of each candidate meteorological factor in the target region are determined.
10. The method according to claim 9, characterized in that, The effect parameters in the exposure-response relationship models of the candidate meteorological factors corresponding to each of the sub-regions were merged using the multivariate meta-regression analysis method with restricted maximum likelihood estimation to obtain the effect parameters of each candidate meteorological factor in the target region.
