GIS-based resident health management method

Abstract

The invention discloses a GIS-based resident health management method, which comprises the following steps: step 1, data acquisition, step 2, data processing, step 3, cleaned data are mined and associated, an association relationship among house, resident and medical data is established through a main index, and after spatial data is associated with filing information of resident health archives, the information of the resident health archives is acquired; precise positioning of residents on a space information system can be realized; step 4, carrying out visual display; 5, real-time monitoring and early warning are carried out; step 6, artificial intelligence and machine learning technology; and step 7, a health intervention module. By collecting and analyzing resident health data in real time, real-time monitoring and accurate management of resident health states are realized, and by solving core problems of data association, space integration, value mining and the like, the practicability and decision support capability of the GIS resident health management system are remarkably improved, and the real-time monitoring and accurate management of resident health states are realized. And a key technical guarantee is provided for precise and intelligent resident health management.

Description

Technical Field

[0001] This invention is a GIS-based method for managing residents' health, belonging to the field of flow metering in the natural gas industry. Background Technology

[0002] With the rapid development of technology, Geographic Information Systems (GIS) have been widely used in many fields, including urban planning, environmental monitoring, and agricultural management. GIS technology is increasingly being applied in resident health management.

[0003] GIS technology can integrate hospital electronic medical records, community health records, wearable device monitoring data, and third-party health APP data to build a dynamic health database with "one person, one file" and establish a standard system to normalize heterogeneous data.

[0004] GIS technology can overlay spatial data such as satellite remote sensing imagery, administrative divisions, and transportation networks to accurately locate the service boundaries of each medical grid. Combined with GPS technology, it can pinpoint residents' locations in real time, enabling rapid location determination for rescue in emergencies such as falls.

[0005] By mapping the spatial distribution of diseases, areas with high disease incidence rates can be visually displayed, and the spatial clustering of diseases can be identified. A GIS-based disease risk assessment model, integrating factors such as geographical environment, socioeconomic status, and residents' health status, can quantitatively assess residents' risk of contracting diseases.

[0006] Visualization: Based on the four-level administrative structure of city, county, town and village, a hierarchical and progressive display system is constructed. Combined with visualization methods such as heat maps, icon labels and data dashboards, medical resources and services are accurately presented, making it convenient for managers at all levels and medical and health personnel to obtain the information they need.

[0007] GIS technology can help upgrade primary healthcare services: Primary healthcare institutions have leveraged GIS technology to achieve precise allocation of service resources. By dividing healthcare into refined grids and linking key information such as chronic disease management and maternal and child health care to specific buildings, health services can proactively reach every risk point, improving both the primary care first-visit rate and the rate of standardized chronic disease management. This drives a shift in health management models from traditional passive management to proactive management. Residents can check their health scores, schedule appointments with family doctors, and receive AI health reminders anytime through relevant mini-programs. Simultaneously, the system dynamically tracks the physiological indicators of patients with chronic diseases, automatically triggering alerts when abnormal values ​​exceed thresholds, achieving a closed-loop chronic disease management system and improving the rate of standardized patient management.

[0008] However, existing GIS-based resident health management systems may differ across regions and institutions, leading to inconsistent data formats and interfaces, poor system compatibility and collaboration, and difficulty in achieving large-scale inter-regional health data sharing and collaborative management. When integrating multi-source data, quality issues such as inaccurate and incomplete data may arise. In existing systems, resident health data (such as medical records and health files) and cross-departmental data (public security, land resources, civil affairs, etc.) are often stored separately with inconsistent formats, making effective correlation difficult. For example, village clinic names differ between basic medical and basic public health systems (e.g., "Lijia Health Clinic" versus "Luocheng Street Lijia Health Room"), causing data integration difficulties. Spatial data such as resident addresses and medical institution locations lack correlation with health data (such as disease diagnosis and treatment records), making it impossible to analyze health problems through spatial dimensions. For instance, knowing only the number of hypertension patients without being able to pinpoint their distribution area makes it difficult to develop targeted prevention and control strategies. Existing technologies primarily focus on data storage and simple statistics, lacking in-depth data mining and failing to discover key information such as disease incidence patterns and health influencing factors from massive amounts of data. For example, simply counting the total number of diabetes patients in a certain area cannot identify high-incidence clusters and possible triggers. Summary of the Invention

[0009] The technical problem to be solved by this invention is to provide a GIS-based method for resident health management that addresses the above-mentioned shortcomings. By collecting and analyzing resident health data in real time, it enables real-time monitoring and precise management of residents' health status. By solving core issues such as data association, spatial integration, and value mining, it significantly improves the practicality and decision support capabilities of the GIS-based resident health management system, providing key technical support for precise and intelligent resident health management.

[0010] To solve the above technical problems, the present invention adopts the following technical solution: A GIS-based method for resident health management includes the following steps: Step 1, Data Collection: To acquire residents’ full life-cycle data, and relying on the city’s health information platform, all medical institutions, laboratories, examinations, and residents’ health records in the city will be synchronized and updated every 15 minutes via ETL and CDC. Step 2, Data Processing: The collected data is cleaned, integrated, and standardized to ensure its accuracy and consistency. Step 3: Mining and linking the cleaned data. Establishing the relationship between housing, residents, and medical data through the master index. After linking spatial data with the filing information of residents' health records, accurate positioning of residents in the spatial information system can be achieved. Step 4, Visualization; The analysis results are visualized in the form of maps, charts, reports, etc., so that users can intuitively understand the health status of residents; Step 5: Real-time monitoring and early warning; Step 6: Artificial intelligence and machine learning technologies; Step 7, Health Intervention Module; Based on the analysis results, the system automatically generates health intervention recommendations and pushes these recommendations to relevant residents and medical institutions via mobile applications and SMS notifications to achieve precise health management.

[0011] Furthermore, in step 2, data cleaning includes converting non-standard village clinic names for basic medical care and basic public health, standardizing data based on the national health information platform, establishing a standardized data dictionary, and linking spatial data with non-spatial data to form a complete resident health database.

[0012] Furthermore, step 3 includes the following steps: Step 3.1, Construction of the main index system; Establish unique master index identifiers for housing, resident, and medical data respectively; assign a unique resident health master index to each resident; assign a unique geocode master index to each housing unit; and assign a unique institution master index to each medical institution. Step 3.2, Extracting data association fields; Extract fields that can be used for association from the cleaned resident health records; extract address code and administrative division field from housing data; extract institution code, address, and service coverage field from medical data; Step 3.3, Spatial data standardization processing; Spatial data such as residential addresses and medical institution locations are uniformly converted into a standard geographic coordinate format to ensure that all spatial data are in the same coordinate system. Step 3.4, Data Mining; Step 3.5: Establish relationships between the data.

[0013] Furthermore, step 3.4, data mining includes the following steps: Step 3.4.1, Association rule mining; Based on the primary index, we can mine the association rules between houses and residents; and mine the association rules between residents and medical data. Step 3.4.2, Entity Recognition and Extraction; From the cleaned health data, identify health-related entities such as patients with hypertension and diabetes, and extract their personal information, diagnosis time, and disease status data; Identify agricultural entities such as vegetable greenhouses, schools, and enterprises from spatial data and data from relevant departments, and extract their location, scale, and business scope information; Step 3.4.3, Cluster Analysis; Cluster analysis was performed on the spatial distribution of disease patients to identify high-incidence areas; agricultural entities such as vegetable greenhouses, schools, and enterprises were clustered by geographical location to analyze their potential correlation with the health status of surrounding residents. Step 3.4.4: Determine the clustering objects and prepare the data; Define the target entities for clustering: Conduct cluster analysis on health-related entities such as patients with hypertension and diabetes, as well as spatial entities such as vegetable greenhouses, schools, and enterprises; Extract key data fields: For healthy entities, extract their latitude and longitude coordinates and diagnosis time information; for spatial entities, extract their latitude and longitude coordinates and scale information. Data standardization: Ensure that the spatial coordinates of all entities are in the same coordinate system, eliminate coordinate differences between different data sources, and ensure the accuracy of clustering; Step 3.4.5: Set clustering parameters; Step 3.4.6: Perform cluster analysis; The standardized entity data is input into the clustering algorithm, and entities that are close to each other are grouped into the same cluster by calculating the spatial distance between them. Step 3.4.7, Clustering results verification and adjustment; Verify the rationality of the clustering results in conjunction with actual business scenarios, and determine whether there are any logical deviations; If the results do not meet expectations, adjust the clustering parameters and re-execute the clustering until a valid result is obtained. Each cluster represents a patient cluster area, which is marked on a map through a spatial information system to intuitively display the spatial distribution characteristics of hypertension and diabetes. By correlating and analyzing the health data of the cluster area with that of surrounding residents, we can provide a basis for research on health influencing factors. Step 3.4.8, Application of clustering results; By linking clustering results with spatial information systems and visually presenting the spatial clustering characteristics of entities, the results can be intuitively displayed.

[0014] Furthermore, step 3.5, establishing relationships between data, includes the following steps: Step 3.5.1, association based on the primary index; Linking houses with residents: By matching the geocoding master index corresponding to the address information in the resident's health record with the geocoding master index of the house data, a one-to-one or one-to-many relationship between "residents and houses" is established. Resident-Medical Linkage: By utilizing the resident master index and the institution master index of medical institutions, a relationship of "resident-medical institution-medical data" is established through the medical records and health information data in the resident's medical records. Step 3.5.2: Linking spatial data with health records; The information in residents' health records is linked with the standardized spatial coordinates of residents' addresses, so that each resident corresponds to a unique spatial location in the spatial information system. For patients with diseases, their personal master index is associated with the corresponding spatial coordinates to enable the patient's location in the spatial information system; for agricultural entities such as vegetable greenhouses, schools, and enterprises, their standardized spatial coordinates are directly associated with the spatial locations of surrounding residents to analyze the spatial correlation between the distribution of entities and residents' health. Step 3.5.3, Multi-entity association integration; Using spatial location as a link, it integrates data from disease patients, agricultural entities such as vegetable greenhouses, schools, and businesses.

[0015] Furthermore, step 3.4.1, the association rule mining, specifically includes the following steps: Step 3.4.1.1: Determine the target data and prepare the data. Identify related entities: Select the entities or attributes whose relationships need to be mined; Step 3.4.1.2, Data preprocessing; A dataset for correlation analysis is formed by linking multiple source data based on the main index. Discretizing the data makes rule extraction easier; Step 3.4.1.3: Select the association rule algorithm; The Apriori algorithm, in its core approach, generates association rules by iteratively filtering high-frequency itemsets layer by layer, as detailed below: Itemset definition: An entity attribute is considered as an "item", and a combination of multiple items is called an "itemset"; Frequent itemset mining: By setting a minimum support, itemsets that appear frequently are selected. Rule generation: Extract rules for "previous item → subsequent item" from the frequent itemset, and filter valid rules by minimum confidence. Step 3.4.1.3, parameter setting and rule filtering; Minimum support: set based on data size and entity relationships; Minimum confidence level: set according to business needs; Pruning optimization: Eliminate redundant rules.

[0016] Furthermore, step 3.4.3, cluster analysis specifically includes the following steps: The core formula for cluster analysis is based on spatial distance calculation and density determination: Step 3.4.3.1: Determine the neighborhood radius, which is used to determine the spatial distance threshold for whether entities belong to the same cluster; For any data point p, its neighborhood is the set of all data points whose spatial distance from p is less than or equal to the neighborhood radius ε, expressed by the formula: ; in, The spatial distance between data points p and q; Step 3.4.3.2: Determine the core points and the minimum number of entities required to form a cluster; If the neighborhood radius of data point p contains at least the minimum number of samples If there are 10 data points, then p is the core point, and the formula is expressed as: ; Step 3.4.3.3, density reachability relationship; if a data point chain exists. ,in , And each It is the core point, and If q is so high that it is reachable from the density of p, then q is said to be attainable from the density of p.

[0017] Furthermore, step 3.4.5, setting the clustering parameters, specifically includes the following steps: Step 3.4.5.1: Setting parameters for spatial distribution clustering of disease patients; Neighborhood radius ε: set according to the population density of the area; Minimum number of samples : Set based on disease incidence rate; Step 3.4.5.2: Setting parameters for clusters of agricultural, business, and school entities such as vegetable greenhouses; Neighborhood radius ε: Adjusted according to entity type; Minimum number of samples : Set according to the density of the entity.

[0018] Furthermore, step 3.4.6, performing cluster analysis, specifically includes the following steps: Step 3.4.6.1, clustering of disease patients is performed; Traverse all patient data points, calculate the number of samples in the neighborhood of each point, and identify core points; Based on density reachability, the core point and all points within its ε-neighborhood are divided into the same cluster, forming a patient clustering region; Isolated cases are considered sporadic cases and are not included in clusters. Step 3.4.6.2: Clustering of agricultural entities such as vegetable greenhouses, enterprises, and schools; Calculate the number of samples in the ε-neighborhood of each entity to identify core entities; Based on density reachability, the core entity and its neighboring entities are grouped into the same cluster; Isolated entities are not included in clusters.

[0019] The present invention adopts the above technical solution and has the following technical effects compared with the prior art: The GIS-based resident health management method achieves real-time monitoring and precise management of residents' health status by collecting and analyzing residents' health data in real time. By solving core issues such as data correlation, spatial integration, and value mining, it significantly improves the practicality and decision support capabilities of the GIS resident health management system, providing key technical support for precise and intelligent resident health management.

[0020] 1. Improve data quality and integrity; By cleaning, standardizing, and linking multi-source data, we ensure the accuracy (e.g., standardized gender and institution names), timeliness (synchronous updates every 15 minutes), and completeness (integration of data from multiple departments such as medical care, public security, and land resources) of residents' full life cycle data, providing a reliable data foundation for health management.

[0021] 2. Achieve precise spatial positioning and visualization; By linking spatial data with health records, residents, patients, medical institutions, agricultural entities such as vegetable greenhouses can be accurately located in the GIS system. Combined with visualization methods such as maps and heat maps, the spatial distribution of health status can be presented intuitively (such as areas with a high concentration of hypertension patients and townships with a high incidence of diseases), helping managers to quickly grasp the regional health situation.

[0022] 3. Supporting scientific decision-making and precise intervention; By conducting entity correlation analysis (such as the correlation between enterprise distribution and respiratory diseases of surrounding residents), we can provide a basis for research on health influencing factors and help formulate targeted policies (such as adjusting enterprise environmental protection standards).

[0023] After identifying high-incidence areas of diseases through cluster analysis, medical resources can be precisely allocated (such as setting up chronic disease management stations in high-incidence areas) to improve the efficiency of health intervention.

[0024] 4. Promote health management throughout the entire life cycle; Based on the integrity and relevance of data throughout the entire life cycle, health data from birth to old age can be traced through spatial dimensions. For example, by analyzing the relationship between vaccination rates and geographical location of children around a school, refined management of the entire process from prevention to treatment can be achieved. Attached Figure Description

[0025] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the accompanying drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn to scale.

[0026] Figure 1 This is the layer distribution of hypertension patients in this invention; Figure 2 This is the layer distribution of the vegetable greenhouse in this invention; Figure 3 This is a visualization of the disease spectrum in this invention. Detailed Implementation

[0027] An example of a GIS-based resident health management method includes the following steps: Step 1, Data Collection: By acquiring residents' full life-cycle data and relying on the city's public health information platform, data from all medical institutions, laboratories, examination centers, and residents' health records are synchronized and updated every 15 minutes via ETL and CDC. Through a unified interface and standardized data formats (HL7, FHIR), seamless integration of data from public security, land resources, agriculture, civil affairs, human resources and social security, and education departments is achieved, ensuring the completeness, timeliness, and accuracy of residents' full life-cycle data.

[0028] Step 2, Data Processing: The collected data underwent cleaning, integration, and standardization to ensure accuracy and consistency. Data cleaning included converting non-standard village clinic names in basic medical and basic public health categories. For example, a basic medical clinic might be named "Lijia Health Station," and a basic public health clinic might be named "Luocheng Street Lijia Health Clinic." Data standardization relied on the national health information platform to establish a standardized data dictionary. For instance, while medical services use "male" and basic public health services use "male," our platform standardized the data to "1 male, 2 females." Through this data cleansing, we ensured the standardization and consistency of the data.

[0029] At the same time, spatial data (such as residents' addresses and the location of medical institutions) are linked with non-spatial data (such as health data) to form a complete resident health database.

[0030] Step 3 involves mining and correlating the cleaned data. By establishing relationships between housing, resident, and medical data through the master index, and linking spatial data with resident health record information, precise location of residents within the spatial information system can be achieved. Figure 1 and Figure 3Currently, precise location tracking has been achieved for multiple entities, including diseases, vegetable greenhouses, agriculture, schools, and enterprises.

[0031] Step 3.1, Construction of the main index system; Establish unique master index identifiers for housing, residents, medical data, etc. Assign a unique resident health master index to each resident (such as an encrypted unique code associated with the ID card number), assign a unique geocode master index to each house (such as a spatial code combined with the real estate number of the land and resources department), and assign a unique institution master index to each medical institution (such as an institution code registered with the health department).

[0032] Ensure that the primary index remains stable throughout the entire data lifecycle and does not change with data updates, serving as the core identifier for data association.

[0033] Step 3.2, Extracting data association fields; Extract relevant fields from the cleaned resident health records that can be used for association, such as the resident's ID number, address information, and the name of the medical institution where they received treatment.

[0034] Extract fields such as address code and administrative division from housing data; extract fields such as institution code, address, and service coverage from medical data.

[0035] Step 3.3, Spatial data standardization processing; Spatial data such as residential addresses and medical institution locations are uniformly converted into a standard geographic coordinate format (such as latitude and longitude coordinates) to ensure that all spatial data are in the same coordinate system. For example, "No. 10, Lijia Village, Luocheng Street" is converted into specific coordinates of XX degrees North latitude and XX degrees East longitude using address resolution technology.

[0036] Step 3.4, Data Mining; Step 3.4.1, Association rule mining; Based on the primary index, we can mine the association rules between houses and residents. For example, by matching the primary index of residents' addresses with the primary index of house geocodes, we can determine the correspondence between residents and the houses they live in.

[0037] Discover the association rules between residents and medical data, such as matching the resident master index with the institution master index of the medical institution to obtain related data such as residents' medical records and treatment information.

[0038] Step 3.4.1, Association rule mining; Based on the primary index, we can mine the association rules between houses and residents. For example, by matching the primary index of residents' addresses with the primary index of house geocodes, we can determine the correspondence between residents and the houses they live in.

[0039] Discover the association rules between residents and medical data, such as matching the resident master index with the institution master index of the medical institution to obtain related data such as residents' medical records and treatment information.

[0040] Step 3.4.1.1: Determine the target data and prepare the data. Identify related entities: Select the entities or attributes whose relationships need to be mined, for example: Residents' health attributes (diagnosis of diseases, etc.) and spatial attributes (address coordinates, whether there are enterprises / vegetable greenhouses, etc. within 500 meters). Residents' medical treatment behavior (such as frequency of visits and choice of medical institutions) and personal attributes (age, gender, occupation); Housing attributes (such as building age and distance to surrounding medical resources) and residents' health indicators (such as physical examination pass rate).

[0041] Step 3.4.1.2, Data preprocessing; A dataset for correlation analysis is formed by linking multi-source data based on the master index. For example, the "resident master index" can be linked to "health records (including disease diagnoses)," "house address coordinates," and "surrounding entity distribution (location of businesses / schools)." Discretize the data, such as dividing "age" into "0-18 years old", "19-65 years old" and "66 years old and above", and dividing "distance from the enterprise" into "<500 meters", "500-1000 meters" and ">1000 meters", to facilitate rule extraction.

[0042] Step 3.4.1.3: Select the association rule algorithm; The Apriori algorithm is employed, and its core principle is to generate association rules by iteratively filtering high-frequency itemsets (frequently occurring combinations of entities) layer by layer. Specifically: Itemset definition: Entity attributes are treated as "items". For example, "hypertension = yes", "there are vegetable greenhouses and other agricultural facilities nearby = yes", and "age = 66 years and above" are all items. A combination of multiple items is an "itemset" (e.g., {hypertension = yes, there are vegetable greenhouses and other agricultural facilities nearby = yes}). Frequent itemset mining: By setting a minimum support (e.g., 5%, meaning a combination appears in the dataset at a rate ≥ 5%), frequently occurring itemsets are selected. For example, if the combination "hypertension = yes and there is a business within 500 meters = yes" accounts for 8% (≥ 5%) of the dataset, it is considered a frequent itemset. Rule generation: Extract rules for "preceding item → subsequent item" from the frequent itemset, and filter valid rules by minimum confidence (e.g., 60%, meaning the probability of the rule being true is ≥60%). For example, the rule "There is a business within 500 meters = yes → high blood pressure = yes" has a confidence of 70% (≥60%), so it is a valid association rule.

[0043] Step 3.4.1.3, parameter setting and rule filtering; Minimum support: Set according to the data scale and entity association. For example, when analyzing "hypertension and enterprise distribution", considering that the incidence of hypertension is about 20%, the minimum support can be set to 3% (to ensure that the rule covers enough samples). Minimum confidence level: Set according to business needs. For example, for disease prevention and control, a higher confidence level (such as 70%) is required to ensure the reliability of the rules; for exploring potential influencing factors, it can be appropriately reduced (such as 50%). Pruning optimization: Eliminate redundant rules (e.g., if both "rule A→B" and "rule A+C→B" exist, retain the simpler rule A→B if their confidence levels are similar).

[0044] The above methods can be used to uncover hidden correlations between residents' health and spatial entities and personal attributes from multi-source data, providing data support for health interventions (such as adjusting corporate environmental standards and optimizing traffic around schools), and realizing a closed loop of "data association → pattern discovery → precise management".

[0045] Step 3.4.2, Entity Recognition and Extraction; From the cleaned health data, health-related entities such as patients with hypertension and diabetes are identified, and their personal information, diagnosis time, and condition data are extracted.

[0046] Identify agricultural entities such as vegetable greenhouses, schools, and enterprises from spatial data and data from relevant departments, and extract information such as their location, scale, and business scope (e.g., the number of students in schools and the industry type of enterprises).

[0047] Step 3.4.3, Cluster Analysis Cluster analysis was performed on the spatial distribution of disease patients to identify areas with high disease incidence; entities such as schools and businesses were clustered by geographical location to analyze their potential correlation with the health status of surrounding residents.

[0048] The core formula for cluster analysis is based on spatial distance calculation and density determination: Step 3.4.3.1: Determine the neighborhood radius, which is used to determine the spatial distance threshold for whether entities belong to the same cluster; For any data point p, its neighborhood is the set of all data points whose spatial distance from p is less than or equal to the neighborhood radius ε, expressed by the formula: ; in, This represents the spatial distance between data points p and q (for latitude and longitude coordinates, Euclidean distance or spherical distance is usually used for calculation).

[0049] Step 3.4.3.2: Determine the core points and the minimum number of entities required to form a cluster; If the neighborhood radius of data point p contains at least the minimum number of samples If there are 10 data points, then p is the core point, and the formula is expressed as: ; Step 3.4.3.3, density reachability relationship; if a data point chain exists. ,in , And each It is the core point, and If q is so high that it is reachable from the density of p, then q is said to be attainable from the density of p.

[0050] Step 3.4.4: Determine the clustering objects and prepare the data; Define the target entities for clustering: Conduct cluster analysis on health-related entities such as patients with diseases such as hypertension and diabetes, as well as spatial entities such as agricultural entities such as vegetable greenhouses, schools, and enterprises.

[0051] Extract key data fields: For healthy entities, extract their latitude and longitude coordinates, diagnosis time, and other information; for spatial entities, extract their latitude and longitude coordinates, scale, and other information.

[0052] Data standardization: Ensure that the spatial coordinates of all entities are in the same coordinate system, eliminate coordinate differences from different data sources, and guarantee the accuracy of clustering.

[0053] Step 3.4.5: Set clustering parameters; Step 3.4.5.1: Setting parameters for spatial distribution clustering of disease patients; Neighborhood radius ε: set according to the population density of the area, for example, 500 meters in urban areas (dense area) and 1000 meters in rural areas (sparse area).

[0054] Minimum number of samples Combined with the disease incidence rate, for example, it is set to 8 (that is, when there are at least 8 patients in a certain area and the distance between them is ≤ ε, it is determined to be a cluster area).

[0055] Step 3.4.5.2: Setting parameters for clusters of agricultural entities such as vegetable greenhouses and businesses. Neighborhood radius ε: Adjusted according to entity type. For example, agricultural clusters such as vegetable greenhouses are set to 1000 meters (because agricultural land is widely distributed), while enterprise clusters (such as industrial zones) are set to 800 meters.

[0056] Minimum number of samples The value is set according to the density of entities. For example, agricultural facilities such as vegetable greenhouses are set to 5 (at least 5 greenhouses are clustered together), and enterprises are set to 3 (at least 3 enterprises are clustered together).

[0057] Step 3.4.6, perform cluster analysis Standardized entity data is input into a clustering algorithm, which calculates the spatial distance between entities (such as Euclidean distance) and groups entities that are close together into the same cluster. For example, clustering data of diabetic patients can result in multiple patient clusters.

[0058] Step 3.4.6.1, clustering of disease patients is performed; Iterate through all patient data points, calculate the number of samples in the neighborhood of each point, and identify core points (e.g., if there are 12 patients within 500 meters of a given patient, and...). =8, then the patient is the core point).

[0059] Based on density reachability, the core point and all points within its ε-neighborhood are divided into the same cluster, forming a patient cluster area (a high-incidence area of ​​the disease).

[0060] Isolated cases (non-core points and not within the ε-neighborhood of any core point) are considered sporadic cases and are not included in the cluster.

[0061] Step 3.4.6.2, clustering of agricultural entities such as vegetable greenhouses and enterprises; Calculate the number of samples within the ε-neighborhood of each entity to identify core entities (e.g., if there are 6 other greenhouses within 1000 meters of a certain vegetable greenhouse, etc.). =5, then the greenhouse is the core point).

[0062] By using density reachability relationships, core entities and their neighboring entities are grouped into the same cluster, forming agricultural clusters such as "vegetable greenhouses" and "enterprise clusters".

[0063] Isolated entities (such as a single business or a single greenhouse) are not included in the cluster area.

[0064] Step 3.4.7, Clustering Results Validation and Adjustment Verify the rationality of clustering results by combining them with actual business scenarios. For example, compare the clustering results of hypertension patients with information such as the distribution of local medical institutions and population density to determine whether there are any logical biases.

[0065] If the results do not meet expectations, adjust the clustering parameters (such as increasing / decreasing the neighborhood radius) and re-execute the clustering until a valid result is obtained.

[0066] Each cluster represents a patient cluster area, which is marked on a map through a spatial information system to intuitively display the spatial distribution characteristics of the disease.

[0067] Correlation analysis between clustered areas and surrounding residents' health data (such as the incidence of hypertension) can be conducted to identify potential associations such as "a high density of hypertension patients around an agricultural cluster such as a vegetable greenhouse," providing a basis for research on health influencing factors.

[0068] Step 3.4.8, Application of Clustering Results By linking clustering results with spatial information systems and visually displaying them (such as marking clustered areas with different colors on a map), the spatial clustering characteristics of entities can be presented intuitively. For example, the locational relationship between areas with a high concentration of hypertension patients and surrounding businesses, vegetable greenhouses, and other agricultural facilities can be shown, providing a basis for analyzing health influencing factors.

[0069] The above clustering methods can accurately identify the spatial clustering characteristics of different entities and effectively handle irregularly shaped clusters, thus meeting the needs for analyzing spatial distribution patterns in health management scenarios.

[0070] Step 3.5: Establish relationships between data; Step 3.5.1, association based on the primary index; House-Resident Association: By matching the geocoding master index corresponding to the address information in the resident's health record with the geocoding master index of the house data, a one-to-one or one-to-many relationship between "resident-house" is established (such as multiple residents living in the same house).

[0071] Resident-Medical Linkage: By utilizing the resident master index and the medical institution master index, and through data such as residents' medical records and medical information in health records, a relationship of "resident-medical institution-medical data" is established.

[0072] Step 3.5.2: Linking spatial data with health records; The information in residents' health records (including the resident master index, basic information, etc.) is linked with the standardized spatial coordinates of residents' addresses, so that each resident corresponds to a unique spatial location in the spatial information system.

[0073] For patients with diseases, their personal master index is associated with the corresponding spatial coordinates to enable the patient's location in the spatial information system; for agricultural entities such as vegetable greenhouses, schools, and enterprises, their standardized spatial coordinates are directly associated with the spatial locations of surrounding residents to analyze the spatial correlation between entity distribution and residents' health.

[0074] Step 3.5.3, Multi-entity association integration; Using spatial location as a link, data from entities such as disease patients, agricultural facilities like vegetable greenhouses, schools, and businesses can be integrated. For example, through spatial overlay analysis, the distribution of hypertension patients around a particular agricultural facility, or the incidence trend of diabetes patients around a school, can be examined, establishing spatial relationships between multiple entities.

[0075] Step 4, Visualization; The analysis results are visualized in the form of maps, charts, and reports, allowing users to intuitively understand residents' health status. Interactive functions are also provided, allowing users to customize queries and analyses according to their needs. Taking disease spectrum as an example, backend big data analysis can realize the disease spectrum of the entire city and its townships and villages, and visualize it in various new formats such as icons.

[0076] Step 4.1, Map Display: Health data is displayed in the form of a map, which intuitively reflects the spatial distribution of health status.

[0077] Maps can include various types such as administrative division maps, topographic maps, and satellite images to meet the needs of different users.

[0078] Step 4.2, Chart Display: Use charts such as bar charts, line charts, and pie charts to display the statistical results and trends of health data, such as... Figure 3 As shown.

[0079] Charts can show changes in a single indicator or the relationships between multiple indicators.

[0080] Step 4.3, Heatmap Display: By using color depth to represent the level of health status or disease risk, the spatial concentration of health problems can be visually displayed.

[0081] Step 5: Real-time monitoring and early warning technology; Step 5.1, Real-time monitoring: It can acquire the latest health data in real time and monitor and analyze it in real time.

[0082] Set an early warning threshold; when the data exceeds the threshold, the system will automatically trigger an early warning mechanism.

[0083] Step 5.2, Early Warning Mechanism: When the system detects abnormal health data or potential health risks, it automatically issues an alert. The alert can be sent to relevant personnel via SMS, email, mobile application push notifications, etc.

[0084] Step 5.3, Intervention measures: Based on the early warning information, the system automatically generates corresponding intervention recommendations. These interventions may include aspects such as health education, allocation of medical resources, and disease prevention and control.

[0085] Step 6: Artificial intelligence and machine learning technologies; Step 6.1, Health Trend Prediction: By using historical health data and training models with machine learning algorithms, we can predict future health trends for residents.

[0086] The prediction results can provide a scientific basis for health management and intervention.

[0087] Step 6.2, Risk Assessment: By combining residents' personal information and health data, artificial intelligence technology is used to assess residents' disease risks. The risk assessment results can help medical institutions and government departments develop more precise health management strategies.

[0088] Step 6.3, Intelligent Decision Support: Leveraging artificial intelligence and machine learning technologies, the system provides intelligent decision support for health management and interventions. It can automatically adjust intervention strategies and resource allocation plans based on real-time data and predictive results.

[0089] Step 7, Health Intervention Module; Based on the analysis results, the system automatically generates health intervention recommendations, such as disease prevention, health education, and medical resource allocation. Simultaneously, these recommendations are pushed to relevant residents and medical institutions via mobile applications and SMS notifications, enabling precise health management.

[0090] The description of this invention is given for illustrative and descriptive purposes only and is not intended to be exhaustive or to limit the invention to the forms disclosed. Many modifications and variations will be apparent to those skilled in the art. The embodiments were chosen and described in order to better illustrate the principles and practical application of the invention and to enable those skilled in the art to understand the invention and design various embodiments with various modifications suitable for a particular purpose.

Claims

1. A GIS-based method for resident health management, characterized in that: Includes the following steps: Step 1, Data Collection: To acquire residents’ full life-cycle data, and relying on the city’s health information platform, all medical institutions, laboratories, examinations, and residents’ health records in the city will be synchronized and updated every 15 minutes via ETL and CDC. Step 2, Data Processing: The collected data is cleaned, integrated, and standardized to ensure its accuracy and consistency. Step 3: Mining and linking the cleaned data. Establishing the relationship between housing, residents, and medical data through the master index. After linking spatial data with the filing information of residents' health records, accurate positioning of residents in the spatial information system can be achieved. Step 4, Visualization; The analysis results are visualized in the form of maps, charts, reports, etc., so that users can intuitively understand the health status of residents; Step 5: Real-time monitoring and early warning; Step 6: Artificial intelligence and machine learning technologies; Step 7, Health Intervention Module; Based on the analysis results, the system automatically generates health intervention recommendations and pushes these recommendations to relevant residents and medical institutions via mobile applications and SMS notifications to achieve precise health management.

2. The GIS-based resident health management method as described in claim 1, characterized in that: In step 2, data cleaning includes converting non-standard village clinic names for basic medical care and basic public health. Data standardization relies on the national health information platform to establish a standardized data dictionary. At the same time, spatial data is linked with non-spatial data to form a complete resident health database.

3. The GIS-based resident health management method as described in claim 1, characterized in that: Step 3 includes the following steps: Step 3.1, Construction of the main index system; Establish unique master index identifiers for housing, resident, and medical data respectively; assign a unique resident health master index to each resident; assign a unique geocode master index to each housing unit; and assign a unique institution master index to each medical institution. Step 3.2, Extracting data association fields; Extract fields that can be used for association from the cleaned resident health records; extract address code and administrative division field from housing data; extract institution code, address, and service coverage field from medical data; Step 3.3, Spatial data standardization processing; Spatial data such as residential addresses and medical institution locations are uniformly converted into a standard geographic coordinate format to ensure that all spatial data are in the same coordinate system. Step 3.4, Data Mining; Step 3.5: Establish relationships between the data.

4. The GIS-based resident health management method as described in claim 3, characterized in that: Step 3.4, data mining includes the following steps: Step 3.4.1, Association rule mining; Based on the primary index, we can mine the association rules between houses and residents; and mine the association rules between residents and medical data. Step 3.4.2, Entity Recognition and Extraction; From the cleaned health data, identify health-related entities such as patients with hypertension and diabetes, and extract their personal information, diagnosis time, and disease status data; Identify agricultural, school, and business entities from spatial data and data from relevant departments, and extract their location, scale, and business scope information; Step 3.4.3, Cluster Analysis; Cluster analysis was performed on the spatial distribution of disease patients to identify high-incidence areas; agricultural, school, and business entities were clustered by geographical location to analyze their potential correlation with the health status of surrounding residents. Step 3.4.4: Determine the clustering objects and prepare the data; Define the target entities for clustering: Conduct cluster analysis on health-related entities such as patients with hypertension and diabetes, as well as spatial entities such as agriculture, schools, and enterprises; Extract key data fields: For healthy entities, extract their latitude and longitude coordinates and diagnosis time information; for spatial entities, extract their latitude and longitude coordinates and scale information. Data standardization: Ensure that the spatial coordinates of all entities are in the same coordinate system, eliminate coordinate differences between different data sources, and ensure the accuracy of clustering; Step 3.4.5: Set clustering parameters; Step 3.4.6: Perform cluster analysis; The standardized entity data is input into the clustering algorithm, and entities that are close to each other are grouped into the same cluster by calculating the spatial distance between them. Step 3.4.7, Clustering results verification and adjustment; Verify the rationality of the clustering results in conjunction with actual business scenarios, and determine whether there are any logical deviations; If the results do not meet expectations, adjust the clustering parameters and re-execute the clustering until a valid result is obtained. Each cluster represents a patient cluster area, which is marked on a map through a spatial information system to intuitively display the spatial distribution characteristics of hypertension and diabetes. By correlating and analyzing the health data of the cluster area with that of surrounding residents, we can provide a basis for research on health influencing factors. Step 3.4.8, Application of clustering results; By linking clustering results with spatial information systems and visually presenting the spatial clustering characteristics of entities, the results can be intuitively displayed.

5. A GIS-based resident health management method as described in claim 3, characterized in that: Step 3.5, establishing relationships between data, includes the following steps: Step 3.5.1, association based on the primary index; House-Resident Association: By matching the geocoding master index corresponding to the address information in the resident's health record with the geocoding master index of the house data, a one-to-one or one-to-many relationship between "resident-house" is established. Resident-Medical Linkage: By utilizing the resident master index and the institution master index of medical institutions, a relationship of "resident-medical institution-medical data" is established through the resident's medical records and health information data. Step 3.5.2: Linking spatial data with health records; The information in residents' health records is linked with the standardized spatial coordinates of residents' addresses, so that each resident corresponds to a unique spatial location in the spatial information system. For patients with diseases, their personal master index is associated with the corresponding spatial coordinates to realize the patient's location in the spatial information system; for agricultural, school, and enterprise entities, their standardized spatial coordinates are directly associated with the spatial location of surrounding residents to analyze the spatial correlation between entity distribution and residents' health. Step 3.5.3, Multi-entity association integration; Using spatial location as a link, it integrates data from disease patients, agriculture, schools, and business entities.

6. A GIS-based resident health management method as described in claim 4, characterized in that: Step 3.4.1, association rule mining, specifically includes the following steps: Step 3.4.1.1: Determine the target data and prepare the data. Identify related entities: Select the entities or attributes whose relationships need to be mined; Step 3.4.1.2, Data preprocessing; A dataset for correlation analysis is formed by linking multiple source data based on the main index. Discretizing the data makes rule extraction easier; Step 3.4.1.3: Select the association rule algorithm; The Apriori algorithm, in its core approach, generates association rules by iteratively filtering high-frequency itemsets layer by layer, as detailed below: Itemset definition: An entity attribute is considered as an "item", and a combination of multiple items is called an "itemset"; Frequent itemset mining: By setting a minimum support, itemsets that appear frequently are selected. Rule generation: Extract rules for "previous item → subsequent item" from the frequent item set, and filter valid rules by minimum confidence; Step 3.4.1.3, parameter setting and rule filtering; Minimum support: set based on data size and entity relationships; Minimum confidence level: set according to business needs; Pruning optimization: Eliminate redundant rules.

7. A GIS-based resident health management method as described in claim 4, characterized in that: Step 3.4.3, cluster analysis specifically includes the following steps: The core formula for cluster analysis is based on spatial distance calculation and density determination: Step 3.4.3.1: Determine the neighborhood radius, which is used to determine the spatial distance threshold for whether entities belong to the same cluster; For any data point p, its neighborhood is the set of all data points whose spatial distance from p is less than or equal to the neighborhood radius ε, expressed by the formula: ; Where is the spatial distance between data points p and q; Step 3.4.3.2: Determine the core points and the minimum number of entities required to form a cluster; If the neighborhood radius of data point p contains at least the minimum number of data points, then p is a core point, expressed by the formula: ; Step 3.4.3.3, Density reachability relation; If there exists a chain of data points, where , , and each is a core point, and , then q is said to be density reachable from p.

8. A GIS-based resident health management method as described in claim 4, characterized in that: Step 3.4.5, setting the clustering parameters, specifically includes the following steps: Step 3.4.5.1: Setting parameters for spatial distribution clustering of disease patients; Neighborhood radius ε: set according to the population density of the area; Minimum sample size: set based on disease incidence rate; Step 3.4.5.2: Setting parameters for clusters of agricultural, business, and school entities; Neighborhood radius ε: Adjusted according to entity type; Minimum sample size: set according to entity density.

9. A GIS-based resident health management method as described in claim 4, characterized in that: Step 3.4.6, performing cluster analysis, specifically includes the following steps: Step 3.4.6.1, clustering of disease patients is performed; Traverse all patient data points, calculate the number of samples in the neighborhood of each point, and identify core points; Based on density reachability, the core point and all points within its ε-neighborhood are divided into the same cluster, forming a patient clustering region; Isolated cases are considered sporadic cases and are not included in clusters. Step 3.4.6.2: Clustering of agricultural, business, and school entities is performed; Calculate the number of samples in the ε-neighborhood of each entity to identify core entities; Based on density reachability, the core entity and its neighboring entities are grouped into the same cluster; Isolated entities are not included in clusters.

Images