A method and system for integrated community monitoring and early warning of key infectious diseases

Abstract

The application provides an integrated community monitoring and early warning method and system for key infectious diseases, collects multi-source data such as suspected case geographic position, contact degree, clinical symptoms, environment and abnormal death; then calculates spatial aggregation, case contact, symptoms, environment and abnormal death components through spatial centroid offset and contact network; linear superposition is carried out to obtain a comprehensive risk composite value and determine a risk level; finally, the overall spatial aggregation is calculated based on the risk quantization value of each community to generate a regional level early warning signal. The application constructs an integrated community monitoring and early warning system of 'human-environment-animal', improves the early identification accuracy and efficiency of infectious diseases, and realizes early judgment, accurate early warning and rapid response of key infectious disease risks.

Description

Technical Field

[0001] This invention belongs to the interdisciplinary field of public health and information technology, specifically relating to an integrated community monitoring and early warning method for key infectious diseases. Background Technology

[0002] The infectious disease surveillance and early warning system is the core of public health emergency management, and its performance directly determines whether the prevention and control goals of "early detection, early reporting, early verification, early confirmation, and early warning" of an epidemic can be achieved. With the advancement of intelligent multi-point triggering surveillance systems, strengthening grassroots surveillance capabilities at the community level, especially for early warning of vector-borne infectious diseases and zoonotic diseases, has become a strategic priority for enhancing the resilience of the public health system. The key to current technological development lies in how to effectively integrate early reporting data from village doctors, community grid workers, veterinarians, and other grassroots sources to construct a regionalized early warning model that can dynamically integrate multi-source heterogeneous information such as spatial distribution, clinical symptoms, ecological environment, and abnormal events.

[0003] Currently, community-level infectious disease surveillance technologies primarily rely on a "passive reporting + symptom screening" model. Typical implementations include: establishing a hierarchical reporting process for case information, conducting post-incident epidemiological investigations, and setting static thresholds for early warning based on historical data. These methods are effective in identifying cluster outbreaks in traditional epidemics where data is standardized and transmission chains are clear. However, when the focus of surveillance shifts to diseases with complex ecological backgrounds and atypical symptoms, traditional surveillance methods reveal the following shortcomings: 1. Limited spatial analysis dimensions: Currently, "spatial clustering" is often simply equated with the total number of cases exceeding the standard within an administrative region, which fails to accurately depict the potential dynamics of epidemic transmission.

[0004] 2. Insufficient use of clinical symptoms: Symptom judgment relies heavily on fixed threshold alarms for a single indicator (such as body temperature), resulting in low signal recognition rate.

[0005] 3. Insufficient ecological and environmental factors: Ecological risk signals such as environmental factors and abnormal animal deaths are generally not included in the early warning and assessment system.

[0006] Therefore, there is an urgent need in this field to break through existing technological bottlenecks and develop a new invention for community monitoring and early warning. Summary of the Invention

[0007] The purpose of this invention is to address the above-mentioned problems by proposing an integrated community monitoring and early warning method and system for key infectious diseases.

[0008] The technical invention of this invention is: In a first aspect, the present invention provides an integrated community monitoring and early warning method for key infectious diseases, comprising the following steps: Step S1: For suspected cases in the target monitoring community, obtain the family geographical coordinates, family contact degree, estimated occupational contact degree, thrombocytopenia indicator, bleeding indicator, rash indicator, as well as the green coverage rate of the target monitoring community, the current month number, the number of abnormal human deaths, and the number of abnormal animal deaths. Step S2: Calculate the spatial clustering component based on the family geographical coordinates of suspected cases; calculate the case contact network component based on the number of family members and occupational exposure information of each suspected case; calculate the corresponding symptom components based on the thrombocytopenia indicator, bleeding indicator, and rash indicator of each suspected case; calculate the green coverage rate component and monthly risk component of the target monitoring community as environmental components; calculate the abnormal human mortality component and abnormal animal mortality component as abnormal mortality components. Step S3: Linearly superimpose the above components to obtain the comprehensive risk composite value, and compare the comprehensive risk composite value with the preset threshold range to determine whether the risk level of the target monitoring area is low risk, medium risk or high risk, and quantify the risk level into a risk quantification value. Step S4: Determine the overall spatial clustering component based on the risk quantification values ​​of multiple target monitoring communities and the overall spatial centroid location. Simultaneously, calculate the algebraic sum of the overall risk quantification values ​​and the overall risk value, and generate an overall early warning signal covering the entire monitoring range based on the overall risk value.

[0009] Furthermore, the computation of spatial clustering components in S2 includes: Obtain the home geographic coordinates of all suspected cases within the target monitoring community. Calculate its centroid The spatial clustering components are determined according to the following formula. ;

[0010] in, n Total number of suspected cases i The suspected case number, This is the preset adjustment coefficient.

[0011] Furthermore, the calculation of the case contact network components in S2 includes: For the first i One suspected case, based on the number of its family members Determine household contact levels And combined with their occupational exposure information, the estimated occupational exposure value is obtained. Then, the contact network components of the case are determined according to the following formula. ;

[0012] in, This is the preset adjustment coefficient.

[0013] Furthermore, the symptom component described in S2 includes a thrombocytopenic component. Bleeding tendency and the amount of rash The results are obtained by calculating using the following formulas respectively.

[0014] in, The variable is a binary judgment variable corresponding to the presence of symptoms. The value is 1 when the symptoms occur and 0 when the symptoms do not occur. These are the corresponding preset adjustment coefficients.

[0015] Furthermore, the environmental component in S2 is obtained using the following formula for the green coverage component. Monthly risk component ;

[0016] in, This represents the average annual percentage of green coverage in the area. This is the preset adjustment coefficient; t This is the current month's sequence number. This represents the theoretical maximum monthly risk value. This is a regional adjustment coefficient; The abnormal mortality component in S2 is obtained using the following formula. and abnormal animal mortality ;

[0017] in, The number of abnormal human deaths, The number of abnormal animal deaths, and These are the corresponding preset adjustment coefficients.

[0018] Furthermore, the composite risk value described in S3 Determined by the following formula: .

[0019] Furthermore, the overall spatial clustering component described in S4 The calculations include: Based on the coordinates of the administrative center of the community monitored by multiple targets And the centroid coordinates of the entire space comprised of all the target monitoring communities. The overall spatial clustering components are determined using the following formula. ;

[0020] in, m The target is to monitor the total number of communities. j Number the target monitoring community; For the number j The community's risk quantification value, This is the preset adjustment coefficient.

[0021] Furthermore, the algebraic sum of the overall risk quantification value is calculated in S4. and overall risk value The following formulas are used to determine the values ​​respectively: .

[0022] Furthermore, the adjustment coefficients for each component in S2 The model is optimized by training a machine learning model based on historical epidemic data. The machine learning model is an XGBoost regression model. The input is the original value of each component in the same historical period, and the output is the number of confirmed cases that actually occurred within a preset time after the corresponding time point. The coefficients are adjusted in reverse by minimizing the prediction error.

[0023] Secondly, the present invention provides an integrated community monitoring and early warning system for key infectious diseases, comprising: The data acquisition module is configured to collect monitoring data from the target monitoring community. The monitoring data includes: individual case data: the geographical coordinates of the suspected case's home, the degree of contact within the family, the estimated value of occupational contact, and clinical manifestations such as thrombocytopenia, bleeding, and rash; community environmental data: the green coverage rate of the target monitoring community and the current month number; and community abnormal event data: the number of abnormal human deaths and the number of abnormal animal deaths in the target monitoring community within a preset statistical period. The risk component calculation module is connected to the data acquisition module and is configured to calculate risk components in multiple dimensions based on the monitoring data, including: spatial clustering component, case contact network component, symptom component, environmental component, green coverage component, monthly risk component, and abnormal mortality component. The community risk assessment module is connected to the risk component calculation module and is configured to linearly superimpose the components to generate a comprehensive risk composite value for the target monitoring community; compare the comprehensive risk composite value with a preset threshold range to determine the risk level of the target monitoring community, and quantify the risk level into a community risk quantification value; The regional early warning generation module is configured to acquire the community risk quantification values ​​and their corresponding spatial centroid locations of multiple target monitoring communities; calculate the overall spatial clustering component based on the spatial centroid location of each community; perform algebraic summation on the community risk quantification values ​​of each community to calculate the overall risk quantification value; calculate the overall risk value covering the entire monitoring range based on the overall spatial clustering component and the overall risk quantification value; and generate a regional-level overall early warning signal based on the overall risk value.

[0024] The integrated community monitoring and early warning method and system for key infectious diseases provided by this invention has the following significant beneficial effects: 1. This invention achieves deep fusion and full-chain risk assessment of multi-source heterogeneous data. By systematically integrating multi-source heterogeneous data from three dimensions—individual cases, community environment, and abnormal events—this invention constructs a monitoring data foundation covering the entire chain of "humans, environment, and animals." The method not only incorporates individual contact risks such as clinical symptoms of suspected cases, family members, and occupational exposure, but also innovatively introduces environmental and ecological risk factors such as community green coverage, monthly and seasonal effects, and abnormal human and animal mortality events. Through quantitative calculation and linear synthesis of spatial clustering components, case contact network components, symptom components, environmental components, and abnormal mortality components, a comprehensive and three-dimensional assessment of community infectious disease risks is achieved, overcoming the limitations of traditional monitoring methods that suffer from single data dimensions and fragmented chains.

[0025] 2. A multi-level, quantifiable hierarchical early warning system has been constructed. This invention establishes a multi-level early warning mechanism with a multi-level network system from village-level communities to townships, counties, cities, and provinces. At the community level, by comparing the comprehensive risk composite value with a preset threshold range, the risk level can be objectively quantified as low, medium, or high risk, and the community risk quantification value is output. At the township level, the quantified risk values ​​and spatial centroid information of multiple communities are further integrated to calculate the overall spatial clustering component and the overall risk value, thereby generating an overall early warning signal covering a wider monitoring range. This hierarchical quantification system connects the early warning path from grassroots communities to higher-level regions, providing a direct decision-making basis for implementing precise and differentiated hierarchical prevention and control responses.

[0026] 3. A machine learning-based parameter adaptive optimization mechanism is introduced to improve model accuracy and adaptability. In this invention, the key adjustment coefficients used to calculate each risk component are not fixed empirical values, but are optimized and determined through an AI machine learning model trained on historical epidemic data. This AI model takes the original values ​​of each component from the same historical period as input and the actual comprehensive risk as the learning objective. By minimizing the prediction error, it adjusts each coefficient in reverse, enabling the AI ​​model to automatically learn the actual contribution weight of each risk factor in different regions and periods from historical data. This achieves adaptive optimization and continuous iteration of the early warning model parameters, significantly improving the accuracy, reliability, and adaptability to different spatiotemporal scenarios in risk assessment.

[0027] 4. This invention provides a quantitative and structured foundation for the early intelligent identification of clustered outbreaks based on spatiotemporal big data. The standardized risk assessment framework constructed in this invention provides a crucial quantitative foundation and structural support for the subsequent early and automatic identification of clustered outbreaks. The community risk quantification value output by the system, the spatial clustering component calculated based on family geographical coordinates, and the structured case data including occupation, symptoms, and other dimensions together constitute an analyzable dataset of temporal, spatial, and population characteristics. When the risk quantification value or spatial clustering of a certain area exceeds a threshold within a specific time period, or when a cluster of cases with similar high-risk characteristics appears, the system can automatically trigger a clustering warning. Simultaneously, the explicit geographical location information (such as family coordinates and community centroid) in this invention can be directly used for the visualization of case distribution and risk hotspots, thereby quickly locating the spatiotemporal clustering range and providing clear guidance for subsequent precise tracing and investigation.

[0028] Other features and advantages of the present invention will be described in detail in the following detailed description section. Attached Figure Description

[0029] The above and other objects, features and advantages of the present invention will become more apparent from the more detailed description of exemplary embodiments of the invention in conjunction with the accompanying drawings, wherein the same reference numerals generally represent the same components in the exemplary embodiments of the invention.

[0030] Figure 1 A flowchart of an integrated community monitoring and early warning method for key infectious diseases according to the present invention is shown. Detailed Implementation

[0031] Preferred embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While preferred embodiments of the invention are shown in the drawings, it should be understood that the invention can be implemented in various forms and should not be limited to the embodiments set forth herein.

[0032] Example 1 Figure 1 A flowchart of an integrated community monitoring and early warning method for key infectious diseases according to the present invention is shown.

[0033] like Figure 1 As shown, this invention provides an integrated community monitoring and early warning method for key infectious diseases, comprising the following steps: Step S1: For suspected cases in the target monitoring community, obtain the family geographical coordinates, family contact degree, estimated occupational contact degree, thrombocytopenia indicator, bleeding indicator, rash indicator, as well as the green coverage rate of the target monitoring community, the current month number, the number of abnormal human deaths, and the number of abnormal animal deaths. In one possible implementation, household geographic coordinates are collected by village-level grid workers using handheld terminals and recorded in the WGS84 coordinate system with an accuracy controlled within 10 meters. Intra-household contact levels are calculated based on the number of family members. Occupational contact level estimates are categorized into high, medium, and low levels based on occupation type, assigned values ​​of 0.8, 0.5, and 0.2 respectively, adjusted based on weekly fieldwork hours. Thrombocytopenia, bleeding, and rash indicators are represented as binary variables, with 1 indicating the presence of corresponding clinical manifestations and 0 indicating their absence; these are entered after initial screening by primary healthcare institutions.

[0034] Step S2: Calculate the spatial clustering component based on the family geographical coordinates of suspected cases; calculate the case contact network component based on the number of family members and occupational exposure information of each suspected case; calculate the corresponding symptom components based on the thrombocytopenia indicator, bleeding indicator, and rash indicator of each suspected case; calculate the green coverage rate component and monthly risk component of the target monitoring community as environmental components; calculate the abnormal human mortality component and abnormal animal mortality component as abnormal mortality components. Specifically, calculating the spatial clustering component includes: obtaining the home geographic location coordinates of all suspected cases within the target monitoring community. Calculate its centroid The spatial clustering components are determined according to the following formula. ;

[0035] in, n Total number of suspected cases i The suspected case number, This is the preset adjustment coefficient.

[0036] In one embodiment, 10 suspected cases were identified within the target monitoring community, and their home geographical coordinates were known through prior data collection. By calculating the average of these coordinates, the centroid was determined to be located near a public green space in the community center. Further analysis revealed that the homes of 6 cases were less than 200 meters from the centroid, showing strong spatial clustering, while the remaining 4 cases were more dispersed. A comprehensive calculation of the distance deviation yielded a spatial clustering component, reflecting the concentration of case distribution. This helps identify potentially high-risk areas within the community and provides a basis for subsequent prevention and control measures.

[0037] The calculation of the case contact network components includes: for the first i One suspected case, based on the number of its family members Determine household contact levels And combined with their occupational exposure information, the estimated occupational exposure value is obtained. Then, the contact network components of the case are determined according to the following formula. ;

[0038] in, This is the preset adjustment coefficient.

[0039] In one embodiment, a suspected case in the community had five family members, thus the household contact level was 4. This case was a tea picker, whose work environment involved contact with ticks in outdoor hilly areas; therefore, the occupational contact level was set to a high level. Another case was an elderly person living alone, with zero household contact; their occupation was an indoor urban work environment, so the occupational contact level was set to a low level. By comprehensively calculating the household and occupational contact levels for all cases, a case contact network component was obtained, reflecting the overall transmission risk of cases within the community.

[0040] In S2, the symptom component includes a thrombocytopenic component. Bleeding tendency and the amount of rash The results are obtained by calculating using the following formulas respectively.

[0041] in, The variable is a binary judgment variable corresponding to the presence of symptoms. The value is 1 when the symptoms occur and 0 when the symptoms do not occur. These are the corresponding preset adjustment coefficients.

[0042] In one embodiment, comprehensive physical examinations and blood tests were performed on 10 suspected cases. Of these, 3 cases showed platelet counts below the normal range, therefore a thrombocytopenia indicator value of 1; 2 cases were found to have unexplained petechiae during the physical examination, therefore a bleeding indicator value of 1; and 1 case had a noticeable rash during the skin examination, therefore a rash indicator value of 1. The relevant symptom indicator values ​​for the remaining cases were all 0. These data provided a direct basis for subsequent calculation of symptom components.

[0043] The environmental component in S2 is obtained using the following formula for the green vegetation coverage component. Monthly risk component ;

[0044] in, This represents the average annual percentage of green coverage in the area. This is a preset adjustment coefficient; t This is the current month's sequence number. This represents the theoretical maximum monthly risk value. This is a regional adjustment coefficient; The abnormal mortality component in S2 is obtained using the following formula. and abnormal animal mortality ;

[0045] in, The number of abnormal human deaths, The number of abnormal animal deaths, and These are the corresponding preset adjustment coefficients.

[0046] In one embodiment, for a community monitoring scenario, remote sensing data provided by the local environmental protection department confirms that the green coverage rate of the target monitoring community is 28%. The current time is August, so the month number is 8. The management department reported three abnormal human deaths in the past month, mainly caused by sudden illnesses; and seven abnormal animal deaths, involving both domestic pets and wild animals. These data allow for a preliminary assessment of the impact of the community environment and these abnormal events on the spread of infectious diseases.

[0047] Step S3: Linearly superimpose the above components to obtain the comprehensive risk composite value. The comprehensive risk composite value is compared with the preset threshold range to determine the risk level of the target monitoring area as low risk, medium risk or high risk, and the risk level is quantified into a risk quantification value.

[0048] In one embodiment, the composite risk value is calculated by linearly superimposing spatial clustering, case contact network, symptom, environmental, and abnormal mortality components to obtain a composite value that characterizes the overall risk level of the community. This value is then compared to a preset threshold range, which can be set based on historical epidemic data, for example, divided into low, medium, and high risk levels. Finally, a corresponding risk quantification value is determined based on the risk level for subsequent regional analysis.

[0049] In a specific implementation scenario, the target monitoring community obtains various component values, such as spatial clustering and case contact network, through prior calculations. These component values ​​are then linearly superimposed to obtain a comprehensive risk composite value. For example, if the preset threshold ranges are 0 to 30 for low risk, 31 to 60 for medium risk, and 61 to 100 for high risk, and the comprehensive risk composite value of this community is 45, then its risk level is determined to be medium risk, and the medium risk is quantified into a specific risk quantification value according to preset rules.

[0050] Step S4: Determine the overall spatial clustering component based on the risk quantification values ​​of multiple target monitoring communities and the overall spatial centroid location. Simultaneously, calculate the algebraic sum of the overall risk quantification values ​​and the overall risk value, and generate an overall early warning signal covering the entire monitoring range based on the overall risk value.

[0051] In one embodiment, after quantifying the risks at each community level, it is necessary to broaden the scope to the entire monitoring area, such as all streets and communities in a prefecture-level city, all townships in a county, or key epidemic-affected counties and districts in a province. These communities are treated as risk units, and their spatial relationships and the strength of their respective risk quantification values ​​are used to characterize the epidemic clustering situation and overall pressure level on a larger scale.

[0052] Specifically, the overall spatial clustering component The calculations include: Step S41: Based on the coordinates of the administrative center of the community being monitored from multiple targets. And the centroid coordinates of the entire space comprised of all the target monitoring communities. The overall spatial clustering components are determined using the following formula. ;

[0053] in, m The target is to monitor the total number of communities. j Number the target monitoring community; For the number j The community's risk quantification value, This is the preset adjustment coefficient.

[0054] In one embodiment, the overall centroid is determined by weighting the risk quantification values; that is, communities with higher risk quantification values ​​have a greater weight in determining the location of the overall centroid. This weighted centroid reflects the actual spatial shift of the current risk center of gravity better than an equally weighted centroid. After obtaining the overall centroid, the spatial deviation of each community representative point from the overall centroid is calculated, and these deviations are summarized and statistically analyzed to obtain the overall spatial clustering component. The larger the value of this component, the more concentrated the high-risk communities are spatially, indicating a potential for local outbreaks or contiguous spread within a large area.

[0055] When the monitoring range is large and there are many communities, preliminary spatial clustering of the communities can be performed first. Communities with similar geographical locations and similar risk quantification values ​​can be grouped into larger risk blocks. Then, the clustering can be calculated on a block-by-block basis, which can reduce noise and improve the significance of large-scale clustering features.

[0056] Step S42: Calculate the algebraic sum of the overall risk quantification values. and overall risk value The following formulas are used to determine the values ​​respectively:

[0057] In one embodiment, if the risk quantification value increases significantly during the detection period, it indicates that the number and intensity of risk communities are increasing, and the overall pressure is intensifying. Adding the algebraic sum of the overall risk quantification value and the overall spatial clustering component reflects, firstly, whether the risk has formed a spatial clustering and concentration pattern, and secondly, yields the total risk load.

[0058] Specifically, the adjustment coefficients for each component in S2 The model is optimized by training a machine learning model based on historical epidemic data. The machine learning model is an XGBoost regression model. The input is the original value of each component in the same historical period, and the output is the number of confirmed cases that actually occurred within a preset time after the corresponding time point. The coefficients are adjusted in reverse by minimizing the prediction error.

[0059] In one embodiment, the model employs the XGBoost algorithm, whose input feature vector includes the historical raw values ​​of the aforementioned nine components, labeled with the actual number of confirmed cases occurring within one week after that point in time. During model training, by minimizing the squared loss function, the model can automatically learn the nonlinear correlation weights between each parameter and the outbreak under different geographical environments (such as southern forest areas and northern arid areas). The machine learning-based dynamic coefficient adjustment mechanism adopted in this invention ensures that the system can serve as a general platform, quickly adapting to the monitoring of key infectious diseases in different regions across the country, without requiring complex manual parameter tuning.

[0060] Example 2 This invention provides an integrated community monitoring and early warning system for key infectious diseases, comprising: The data acquisition module is configured to collect monitoring data from the target monitoring community. The monitoring data includes: individual case data: the geographical coordinates of the suspected case's home, the degree of contact within the family, the estimated value of occupational contact, and clinical manifestations such as thrombocytopenia, bleeding, and rash; community environmental data: the green coverage rate of the target monitoring community and the current month number; and community abnormal event data: the number of abnormal human deaths and the number of abnormal animal deaths in the target monitoring community within a preset statistical period. The risk component calculation module is connected to the data acquisition module and is configured to calculate risk components in multiple dimensions based on the monitoring data, including: spatial clustering component, case contact network component, symptom component, environmental component, green coverage component, monthly risk component, and abnormal mortality component. The community risk assessment module is connected to the risk component calculation module and is configured to linearly superimpose the components to generate a comprehensive risk composite value for the target monitoring community; compare the comprehensive risk composite value with a preset threshold range to determine the risk level of the target monitoring community, and quantify the risk level into a community risk quantification value; The regional early warning generation module is configured to acquire the community risk quantification values ​​and their corresponding spatial centroid locations of multiple target monitoring communities; calculate the overall spatial clustering component based on the spatial centroid location of each community; perform algebraic summation on the community risk quantification values ​​of each community to calculate the overall risk quantification value; calculate the overall risk value covering the entire monitoring range based on the overall spatial clustering component and the overall risk quantification value; and generate a regional-level overall early warning signal based on the overall risk value.

[0061] The various embodiments of the present invention have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments.

Claims

1. An integrated community monitoring and early warning method for key infectious diseases, characterized in that, Includes the following steps: Step S1: For suspected cases in the target monitoring community, obtain the family geographical coordinates, family contact degree, estimated occupational contact degree, thrombocytopenia indicator, bleeding indicator, rash indicator, as well as the green coverage rate of the target monitoring community, the current month number, the number of abnormal human deaths, and the number of abnormal animal deaths. Step S2: Calculate the spatial clustering component based on the family geographic location coordinates of suspected cases; calculate the case contact network component based on the number of family members and occupational exposure information of each suspected case; Based on the thrombocytopenia, bleeding, and rash indicators of each suspected case, the corresponding symptom components were calculated; the green coverage rate component and the monthly risk component of the target monitoring community were calculated as environmental components. Calculate the abnormal mortality components of humans and animals as abnormal mortality components; Step S3: Linearly superimpose the above components to obtain the comprehensive risk composite value, and compare the comprehensive risk composite value with the preset threshold range to determine whether the risk level of the target monitoring area is low risk, medium risk or high risk, and quantify the risk level into a risk quantification value. Step S4: Determine the overall spatial clustering component based on the risk quantification values ​​of multiple target monitoring communities and the overall spatial centroid location. Simultaneously, calculate the algebraic sum of the overall risk quantification values ​​and the overall risk value, and generate an overall early warning signal covering the entire monitoring range based on the overall risk value.

2. The integrated community monitoring and early warning method for key infectious diseases according to claim 1, characterized in that... In S2, the computation of spatial clustering components includes: Obtain the home geographic coordinates of all suspected cases within the target monitoring community. Calculate its centroid The spatial clustering components are determined according to the following formula. ; ； in, n Total number of suspected cases i The suspected case number, This is the preset adjustment coefficient.

3. The integrated community monitoring and early warning method for key infectious diseases according to claim 1, characterized in that... In S2, the calculation of the case contact network components includes: For the i One suspected case, based on the number of its family members Determine household contact levels And combined with their occupational exposure information, the estimated occupational exposure value is obtained. Then, the contact network components of the case are determined according to the following formula. ; ； in, This is the preset adjustment coefficient.

4. The integrated community monitoring and early warning method for key infectious diseases according to claim 1, characterized in that... In S2, the symptom component includes a thrombocytopenic component. Bleeding tendency and the amount of rash The results are obtained by calculating using the following formulas respectively. ； in, The variable is a binary judgment variable corresponding to the presence of symptoms. The value is 1 when the symptoms occur and 0 when the symptoms do not occur. These are the corresponding preset adjustment coefficients.

5. The integrated community monitoring and early warning method for key infectious diseases according to claim 1, characterized in that, The environmental component in S2 is obtained using the following formula for the green vegetation coverage component. Monthly risk component ; ； in, This represents the average annual percentage of green coverage in the area. This is a preset adjustment coefficient; t This is the current month's sequence number. This represents the theoretical maximum monthly risk value. This is a regional adjustment coefficient; The abnormal mortality component in S2 is obtained using the following formula. and abnormal animal mortality ; ； in, The number of abnormal human deaths, The number of abnormal animal deaths, and These are the corresponding preset adjustment coefficients.

6. The integrated community monitoring and early warning method for key infectious diseases according to claim 1, characterized in that... In S3, the comprehensive risk composite value Determined by the following formula: 。 7. The integrated community monitoring and early warning method for key infectious diseases according to claim 1, characterized in that... In S4, the overall spatial clustering component The calculations include: Based on the coordinates of the administrative center of the community monitored by multiple targets And the centroid coordinates of the entire space comprised of all the target monitoring communities. The overall spatial clustering components are determined using the following formula. ; ; in, m The target is to monitor the total number of communities. j Number the target monitoring community; For the number j The community's risk quantification value, This is the preset adjustment coefficient.

8. The integrated community monitoring and early warning method for key infectious diseases according to claim 7, characterized in that... In S4, the algebraic sum of the overall risk quantification value is calculated. and overall risk value The following formulas are used to determine the values ​​respectively: 。 9. The integrated community monitoring and early warning method for key infectious diseases according to claim 1, characterized in that... In S2, the adjustment coefficients for each component The model is optimized by training a machine learning model based on historical epidemic data. The machine learning model is an XGBoost regression model. The input is the original value of each component in the same historical period, and the output is the number of confirmed cases that actually occurred within a preset time after the corresponding time point. The coefficients are adjusted in reverse by minimizing the prediction error.

10. An integrated community monitoring and early warning system for key infectious diseases, characterized in that, include: The data acquisition module is configured to collect monitoring data from the target monitoring community. The monitoring data includes: individual case data: the geographical coordinates of the suspected case's home, the degree of contact within the family, the estimated value of occupational contact, and clinical manifestation information such as thrombocytopenia, bleeding, and rash; community environmental data: the green coverage rate of the target monitoring community and the current month number; and community abnormal event data: the number of abnormal human deaths and the number of abnormal animal deaths in the target monitoring community within a preset statistical period. The risk component calculation module is connected to the data acquisition module and is configured to calculate risk components in multiple dimensions based on the monitoring data, including: spatial clustering component, case contact network component, symptom component, environmental component, green coverage component, monthly risk component, and abnormal mortality component. The community risk assessment module is connected to the risk component calculation module and is configured to linearly superimpose the components to generate a comprehensive risk composite value for the target monitoring community; compare the comprehensive risk composite value with a preset threshold range to determine the risk level of the target monitoring community, and quantify the risk level into a community risk quantification value; The regional early warning generation module is configured to acquire the community risk quantification values ​​and their corresponding spatial centroid locations of multiple target monitoring communities; calculate the overall spatial clustering component based on the spatial centroid location of each community; perform algebraic summation on the community risk quantification values ​​of each community to calculate the overall risk quantification value; calculate the overall risk value covering the entire monitoring range based on the overall spatial clustering component and the overall risk quantification value; and generate a regional-level overall early warning signal based on the overall risk value.

Images