A comprehensive evaluation method, device, medium and product for equalization of basic medical services

Abstract

This application discloses a method, equipment, medium, and product for comprehensive evaluation of equalization of basic medical services, relating to the field of medical geography. The method includes: spatial processing of research data to construct a unified geospatial database; calculating multiple evaluation index values ​​representing the spatial allocation level of medical resources and the spatial convenience of residents' access to medical care through spatial analysis to form an original decision matrix; using the entropy weight method, calculating the objective weights of each evaluation index based on the degree of variation of the multiple evaluation index values ​​across different evaluation areas; processing the original decision matrix using the objective weights based on the TOPSIS model to calculate the comprehensive evaluation index of equalization of basic medical services for each evaluation area; and performing spatial hierarchical mapping based on the comprehensive evaluation index of equalization of basic medical services to generate a multidimensional diagnostic report, thereby avoiding evaluation distortion, accurately reflecting residents' comprehensive employment opportunities in accessing medical services, and enhancing public credibility.

Description

Technical Field

[0001] This application relates to the field of medical geography, and in particular to a method, equipment, medium, and product for comprehensive evaluation of the equalization of basic medical services. Background Technology

[0002] Equal access to basic medical services, as the cornerstone of the modern public service system, is a core issue in safeguarding citizens' right to health, promoting social equity, and achieving the strategic goal of "Healthy China." Its core essence lies in ensuring that all citizens, regardless of their place of residence, socioeconomic status, or identity, can access basic medical services of similar quality fairly, conveniently, and affordably when needed. Scientifically, accurately, and dynamically "diagnosing" the current state of equal access to basic medical services, identifying its spatial pattern, areas of weakness, and key constraints, becomes a fundamental prerequisite for governments at all levels to optimize the allocation of medical resources, plan and layout facilities, and formulate public health policies.

[0003] However, the mainstream evaluation methods widely used in academic and practical fields both domestically and internationally suffer from a series of interconnected and unresolved technical shortcomings in terms of theoretical framework, technical approach, and practicality. This often leads to evaluation results being out of touch with the complex geographical and socioeconomic contexts of the real world, specifically manifested in the following three aspects: First, evaluation methods based on macroeconomic statistical indicators suffer from "geospatial blindness" and "accessibility decoupling." These methods rely entirely on macroeconomic and social statistics aggregated at the administrative division level (such as the number of hospital beds per thousand people, the number of tertiary hospitals per ten thousand people, etc.) to construct relative intensity indicators and calculate comprehensive scores using linear weighting, principal component analysis, and other methods. This method assumes that resources are evenly distributed within administrative units, completely ignoring the actual geographical distribution and matching relationship between resources and population. For example, a region may meet the "number of hospital beds per capita" standard, but high-quality medical resources are highly concentrated in the core area, requiring residents in peripheral areas to pay extremely high costs to access services. This evaluation result masks serious inequalities within administrative regions and only reflects a static "ownership" situation, failing to address the time and distance costs required for residents to access services, thus completely decoupling from true "accessibility."

[0004] Second, medical geography evaluation methods based on spatial accessibility suffer from a single evaluation dimension and neglect of demand heterogeneity. These methods, which incorporate GIS technology, focus on the single dimension of "spatial accessibility," often employing the two-step search algorithm (2SFCA) and its variants. They overemphasize spatial distance or temporal thresholds, severing the intrinsic connection with the "allocation level" of medical resources (e.g., high accessibility may simply be due to a high density of low-level clinics, failing to reflect the coverage of high-quality resources). Furthermore, traditional models treat all populations as homogeneous demand points, failing to differentiate the intensity of differentiated medical needs among higher-risk groups such as the elderly and children, leading to evaluation results that deviate from actual social needs.

[0005] Third, the subjective weighting issue in comprehensive evaluation leads to a lack of objective benchmarks and dynamic adaptability in the results. Traditional methods widely employ subjective weighting methods such as the Delphi method and the Analytic Hierarchy Process (AHP), where weights are highly dependent on expert experience. Weights assigned by experts with different backgrounds vary significantly, resulting in poor repeatability and comparability. Furthermore, static subjective weights cannot reflect the actual distribution differences of indicator data across different evaluation objects. For example, when an indicator's value is close to its upper limit in all regions, its discriminatory power is extremely low, and its contribution (weight) to the final evaluation should be minimal. However, subjective weighting may still assign it a high weight, leading to a disconnect between weight allocation and the actual discriminatory power of the data, thus distorting the final evaluation results.

[0006] In summary, existing technological systems have profound limitations in the "spatial dimension," "system dimension," and "objective dimension," often leading to difficulties in medical resource allocation decisions due to insufficient evidence, imprecise targeting, and challenges in evaluating effects. Therefore, there is an urgent need for an intelligent method that can deeply integrate multi-source geospatial big data, automatically construct a multi-dimensional indicator system that considers both "resource allocation" and "spatial accessibility," and objectively and dynamically determine weights based on the inherent characteristics of the data itself, ultimately achieving comprehensive evaluation and spatial visualization output. Summary of the Invention

[0007] The purpose of this application is to provide a comprehensive evaluation method, equipment, medium, and product for equalizing basic medical services, in order to solve the problems of existing methods, such as evaluation distortion due to lack of spatial information, failure to reflect comprehensive medical access opportunities due to single evaluation dimensions, and lack of credibility of results due to strong subjectivity in weight determination.

[0008] To achieve the above objectives, this application provides the following solution: Firstly, this application provides a comprehensive evaluation method for equalizing basic medical services, including: Acquire research data for the study area and perform spatial processing on the research data to construct a unified geospatial database; the research data includes basic geographic framework data, medical resource thematic data, population and social statistics data, and transportation network attribute data; Based on the geospatial database, multiple evaluation index values ​​representing the spatial allocation level of medical resources and the convenience of residents' access to medical care are calculated through spatial analysis to form an original decision matrix; The entropy weight method is used to calculate the objective weight of each evaluation index based on the degree of variation of the values ​​of the multiple evaluation indicators in each evaluation region. Based on the TOPSIS model, the original decision matrix is ​​processed using the objective weights to calculate the comprehensive evaluation index of basic medical service equalization in each evaluation region. Spatial hierarchical mapping is performed based on the comprehensive evaluation index for equalization of basic medical services, and a multidimensional diagnostic report is generated.

[0009] Secondly, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-mentioned comprehensive evaluation method for equalization of basic medical services.

[0010] Thirdly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the aforementioned comprehensive evaluation method for equalization of basic medical services.

[0011] Fourthly, this application provides a computer program product, including a computer program that, when executed by a processor, implements the aforementioned comprehensive evaluation method for equalization of basic medical services.

[0012] According to the specific embodiments provided in this application, this application has the following technical effects: This application acquires research data from the study area and spatializes it to construct a unified geospatial database. The research data includes basic geographic framework data, medical resource thematic data, population and socio-statistical data, and transportation network attribute data. By using multiple types of research data, the application avoids evaluation distortion. Based on the geospatial database, spatial analysis is used to calculate multiple evaluation indicators characterizing the spatial allocation level of medical resources and the spatial convenience of residents' access to medical care, forming an original decision matrix. This matrix encompasses both the spatial coverage of medical institutions to various residential units and accessibility based on supply and demand matching and actual travel distance, truly reflecting the comprehensive employment opportunities for residents to access medical services and overcoming the one-sidedness of traditional methods. This application introduces the entropy weight method, automatically calculating weights based entirely on the degree of data variation of each indicator across regions. The greater the indicator variation, the smaller the entropy value, and the higher the weight, thus objectively identifying the key factors leading to equalization differences. This process is data-driven and repeatable, significantly improving the objectivity and credibility of the evaluation. This application constructs a fully automated pipeline from multi-source data fusion and governance, spatial modeling, indicator calculation, entropy weight assignment, TOPSIS comprehensive evaluation to visualization report generation. Users only need to provide basic data, and the computer can automatically complete all calculations and outputs, supporting batch processing of any number of evaluation areas to achieve routine monitoring. Attached Figure Description

[0013] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0014] Figure 1 A schematic diagram of a comprehensive evaluation method for equalizing basic medical services provided in an embodiment of this application; Figure 2 A schematic diagram of the core process of a comprehensive evaluation method for equalization of basic medical services provided in an embodiment of this application; Figure 3 This is a schematic diagram of a data spatialization processing flow provided in an embodiment of this application; Figure 4 A schematic diagram illustrating the enhanced two-step move search algorithm (E2SFCA) for reachingability calculation according to an embodiment of this application; Figure 5 A flowchart illustrating the logic of calculating index weights using the entropy weight method, provided in an embodiment of this application. Figure 6 This is a schematic diagram illustrating the principle of calculating the comprehensive evaluation index using the TOPSIS model provided in an embodiment of this application. Detailed Implementation

[0015] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0016] To make the objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0017] Example 1 like Figure 1 As shown in the embodiments of this application, a comprehensive evaluation method for equalizing basic medical services is provided, including: S1: Acquire research data for the study area and perform spatial processing on the research data to construct a unified geospatial database; the research data includes basic geographic framework data, medical resource thematic data, population and social statistics data, and transportation network attribute data.

[0018] S2: Based on the geospatial database, multiple evaluation index values ​​representing the spatial allocation level of medical resources and the convenience of residents' access to medical care are calculated through spatial analysis to form an original decision matrix.

[0019] S3: Using the entropy weight method, the objective weight of each evaluation index is calculated based on the degree of variation of the values ​​of the multiple evaluation indicators in each evaluation region.

[0020] S4: Based on the TOPSIS model, the original decision matrix is ​​processed using the objective weights to calculate the comprehensive evaluation index of equalization of basic medical services in each evaluation region.

[0021] S5: Based on the comprehensive evaluation index for equalization of basic medical services, spatial hierarchical mapping is performed, and a multidimensional diagnostic report is generated.

[0022] like Figure 2 As shown, the core process of this application is as follows: First, spatial fusion and governance of multi-source heterogeneous data are carried out to construct a unified geospatial database; second, based on GIS spatial analysis technology, basic indicators covering multiple dimensions such as coverage, accessibility, and configuration intensity are calculated; then, the objective weights of each indicator are automatically calculated using information entropy theory; furthermore, the TOPSIS model is used to calculate the relative proximity of each region to the ideal service level, resulting in an equalization comprehensive index; finally, the index is hierarchically visualized to form evaluation results.

[0023] In an exemplary embodiment, S1 specifically includes: S11: The Monte Carlo-kernel density fusion algorithm is used to perform refined spatial distribution modeling of the population and social statistics data to generate population density raster data.

[0024] S12: Based on the basic geographic framework data and medical resource thematic data, and using the preset service capacity adjustment coefficient, calculate the standardized service capacity value of each medical institution according to the level of the medical institution and the number of beds.

[0025] S13: Construct a topology network model based on the traffic network attribute data, assign travel time costs to each road segment in the road network, and construct a traffic network cost model.

[0026] S14: Write the population density raster data, the standardized service capacity value, and the transportation network cost model into the geospatial database.

[0027] In practical applications, such as Figure 3 As shown, S1 specifically includes: Step 1.1, Multimodal data acquisition, the acquired data includes: (1) Basic geographic framework data: sourced from national or provincial basic geographic information centers, such as provincial geographic national conditions monitoring data, including vector datasets of administrative division boundaries (areas), settlements (points / areas), and transportation networks (lines). The format is Shapefile or GeoJSON, and the spatial reference is uniformly CGCS2000 coordinate system.

[0028] (2) Medical resource data. This data was obtained from the health management department's information system and stored in a structured relational table. Key fields include: medical institution name, geographical coordinates (latitude and longitude), institution level (tertiary, secondary, primary, etc.), approved number of beds, number of on-duty doctors, number of on-duty nurses, and address.

[0029] (3) Population and social statistics. Obtained from the statistics department, including: total population by township / street, age structure (especially the number of people aged 65 and above), and estimated number of people at potential health risks. At the same time, obtain higher-precision gridded population distribution data (such as 100m×100m grid).

[0030] (4) Traffic network attribute data. Associated with road network vector data, define the traffic attributes of each road segment, including: road grade (expressway, national highway, provincial highway, county road, township road), design speed (km / h), and one-way / two-way traffic restrictions.

[0031] Spatial processing is required for population data, health-risk populations, and hospital data, so that they can be used in subsequent spatial analysis and calculation of basic medical service equalization indicators.

[0032] Step 1.2, Spatialization of Population Data Step 1.2.1 Refined Spatial Distribution Modeling of Population Data. For the aggregated statistical population data, the method employs an improved Monte Carlo-kernel density fusion algorithm for downscaling and spatialization.

[0033] Population data undergoes cleaning and formatting to ensure basic data quality, eliminate errors and missing data, and provide reliable input for subsequent Monte Carlo simulations, avoiding a "garbage in, garbage out" approach. Simultaneously, standardized data specifications are implemented to enable the integration of multi-source, heterogeneous population data into a single spatial database, supporting cross-regional and cross-year comparative analysis.

[0034] The specific processing procedure includes: Missing value detection and handling. Specifically: Scan the table to identify null values, NULL values, or invalid placeholders (such as "-", " / ") in the population field. If the proportion of missing values ​​is below a threshold (e.g., 5%), use mean imputation to fill in the missing values ​​with the average population density of adjacent administrative regions at the same level or the province as a whole. If the proportion of missing values ​​is high, obtain estimated values ​​through spatial interpolation or by weighting from higher-level statistical units. All filling operations are logged for subsequent quality control traceability.

[0035] Outlier Identification and Correction. Two outlier detection methods are employed: First, based on statistical rules, the mean and standard deviation of the population in each unit are calculated, and data exceeding the mean ± 3 times the standard deviation are marked as suspected outliers. Second, based on spatial consistency, the population of the target unit is compared with that of its neighboring units; if the difference is too large (e.g., the population of a township suddenly becomes more than 10 times that of surrounding townships), it is marked. There are three correction methods: First, automatic comparison with historical data; if the data changes abruptly in a given year without significant administrative division adjustments, trend extrapolation is used for correction. Second, for obvious data entry errors (e.g., decimal point misalignment), correction is performed based on data from neighboring or higher-level units. Third, if automatic correction is not possible, a warning message is generated for manual verification.

[0036] Data format standardization. This includes unifying field names, mapping field names from different sources to internal standard fields (such as POP_TOTAL, POP_65PLUS, ADCODE); data type conversion, ensuring that population values ​​are integers or floating-point numbers, and administrative division codes are character types; encoding consistency, with administrative division codes uniformly using the national standard GB / T2260 to ensure matching with GIS base map codes; and timestamp processing, recording the year of data in a unified format of YYYY-MM-DD for easy comparison across multiple periods.

[0037] Data matching and association. Population tables and administrative division vector layers are connected via ADCODE to ensure each statistical unit has a corresponding spatial geometric object. If enclaves or special areas exist, spatial location checks are added during the connection process to avoid incorrect attribution.

[0038] Construct a population attribute table. After cleaning, a structured population attribute table is generated, containing the following core fields: ADCODE: Administrative division code (primary key); NAME: Administrative division name; POP_TOTAL: Total population; POP_ELDER: Population aged 65 and above (or other health risk groups); POP_DENSITY: Population density (optional, used for spatialization later); YEAR: Data year; SOURCE: Data source identifier. The cleaned and formatted population attribute table is linked with the administrative division vector layer to form a population statistics unit surface layer with attributes, which serves as the input for the Monte Carlo simulation.

[0039] (2) Using building outline data (from provincial geographic national conditions monitoring data) as the distribution base, it is assumed that the population is evenly distributed within the buildings.

[0040] (3) Perform a Monte Carlo simulation within each administrative unit (e.g., street). Let the total population of the unit be... ,conduct random assignment (n times) In each allocation, a person is randomly assigned a building within that unit, and the probability of being selected is proportional to the building's area.

[0041] (4) Perform kernel density estimation on the simulated population count results of all buildings to generate a continuous population density distribution surface grid. This grid serves as the basis for all subsequent calculations involving the "population" index.

[0042] Step 1.3 Quantifying the Service Capacity of Medical Institutions. The raw data for medical institutions typically comes from administrative records or statistical reports of health management departments, commonly in Excel, CSV, or database tables. Each record contains the following core fields: Medical Institution ID: A unique code for a medical institution.

[0043] Name of medical institution: such as "The First Affiliated Hospital of a Medical University in a certain province".

[0044] Institutional levels: Hospitals are classified into tertiary hospitals, secondary hospitals, primary hospitals, and unclassified hospitals according to the standards of the National Health Commission.

[0045] Number of beds: Approved number of beds (integer).

[0046] Number of medical staff: number of licensed physicians and registered nurses (optional).

[0047] Geographic coordinates: The geographical coordinates of the medical institution, including longitude and latitude.

[0048] Address: Detailed house number or landmark description, such as "No. XX, XX Road, XX District, City A".

[0049] Calculate a standardized service capacity value for each medical institution As the supply scale in the accessibility model The input is [the input]. The calculation formula is: .in, The number of beds is a known quantity. This is an adjustment coefficient based on hospital level (e.g., α=1.5 for tertiary hospitals, α=1.2 for secondary hospitals, and α=1.0 for primary hospitals). This step equivalently converts the resources of hospitals at different levels.

[0050] Specifically, (1) Data cleaning and formatting. Preprocessing of the original medical institution data includes 1) Deduplication and matching: checking for multiple records of the same medical institution and deduplicating based on name and address. 2) Field standardization: unifying the level description (e.g., unifying "Grade A" and "Grade B" to "Grade III") and converting the number of beds field to integers. 3) Address normalization: segmenting and completing unstructured addresses, removing redundant information, and improving the success rate of geocoding.

[0051] (2) Geocoding Processing. For some medical institutions that lack geographic coordinate information, address information can be used to convert text addresses into spatial coordinates. Using multi-source geocoding service adapters such as Gaode Map API, Baidu Map API, and National Geographic Information Public Service Platform, a normalized address string is passed in, and a request is made to return latitude and longitude coordinates (GCJ-02, CGCS2000, or WGS-84 coordinate system). If the returned coordinates are in the State Bureau of Surveying and Mapping coordinate system (GCJ-02) or WGS-84 coordinate system, the coordinate system needs to be further converted to the CGCS2000 geographic coordinate system to ensure consistency with the basic geographic data.

[0052] (3) Service capacity assignment. To reflect the differences in service intensity among hospitals of different levels, this application introduces a service capacity adjustment coefficient α to weight the bed size and form a standardized service capacity value Cap_j: Coefficient setting rules: Tertiary hospitals: α=1.5 (assume the functions of regional medical centers, with a large service radius and strong technical capabilities) Level II hospitals: α=1.2 (regional medical centers with comprehensive service capabilities) Level 1 hospitals and unclassified hospitals: α=1.0 (primary healthcare institutions with limited service coverage) Service capacity calculation: in: The approved number of beds in medical institution j (a known quantity, derived from raw data). Adjustment coefficient (known quantity) determined based on the institution's level. Standardized service capacity value (unknown quantity, supply scale used for subsequent accessibility calculations) ) Extended considerations If the data includes the number of medical staff, a more complex service capacity index can be constructed, such as: in The preset weights can be automatically optimized using the entropy weight method or set by experts.

[0053] (4) Generate a spatial point layer for medical institutions. After all medical institution data has been processed, export the data to a standard spatial data format (such as Shapefile, GeoJSON) and store it in the PostGIS spatial database. The generated layer contains the following core attribute fields: hos_id: unique identifier; name: name of medical institution; grade: grade (1 / 2 / 3 / other); beds: original number of beds; alpha: adjustment coefficient; capacity: standardized service capacity value. Cap j geom: Spatial coordinates. This layer serves as the base input for subsequent buffer analysis and reachability calculations.

[0054] Step 1.4 Road Network Dataset Construction. Road network data comes from national or provincial basic geographic information databases (such as 1:50,000 or 1:10,000 topographic maps), open street maps (OSM), or commercial navigation data. Raw road network data is typically stored as vector line features, and each road contains the following attributes: Road classification: expressway, national highway, provincial highway, county road, township road, urban arterial road, secondary arterial road, branch road, etc.

[0055] Road length: calculated geometrically (unit: meters).

[0056] Design speed: The standard speed (unit: km / h) set according to the road grade, or the average traffic speed based on field surveys.

[0057] Traffic direction: two-way, one-way, no passage, etc. (optional).

[0058] A topological network model is constructed based on road network vector data and its attributes. Each edge (road segment) in the network is assigned a travel time cost. .in, The geometric length of the road segment is a known quantity. The design speed for this road segment is known. This network model is used to calculate the shortest time path between any two points based on road travel.

[0059] Specifically: (1) Data cleaning and topology preprocessing. Data cleaning and topology preprocessing are performed on the road network data, including firstly, geometric inspection and repair, checking for topology errors such as hanging lines, pseudo nodes, and duplicate lines, and using GIS tools to automatically repair them; ensuring that roads are correctly connected at intersections (breaking intersecting lines) to prepare for network construction; secondly, attribute integrity verification, assigning default values ​​to records with missing design speeds according to road level (see Table 1); for fields with missing road names, they can be left blank or assigned "unknown road".

[0060] Table 1. Comparison of Road Grades and Default Design Speeds

[0061] (2) Construction of the traffic network dataset. The network dataset is the foundation for path analysis. This application constructs it using the following steps: Create a network feature class. Import the cleaned road line features into a network dataset building tool (such as ArcGIS Network Analyst or pgRouting); define the network connectivity: typically select "Endpoint Connectivity" (connecting any two roads at their intersection) to ensure traffic continuity.

[0062] Set the toll cost attribute. Calculate the toll time cost for each road segment. As a fundamental impedance in network analysis: in: Road segment length (unit: kilometers, calculated geometrically, known quantity); The design speed of this road section (unit: km / h, known quantity); Travel time (unit: minutes, unknown).

[0063] If there are one-way streets or turning restrictions, the direction of travel (such as forward, reverse, or two-way) must be defined in the attributes.

[0064] Add elevation / terrain effects (optional). In mountainous or complex terrain areas, a slope factor can be introduced to correct travel time. in This is the slope reduction factor (e.g., speed decreases when the slope is >5%).

[0065] (3) Generate the route dataset. After the traffic network dataset is constructed, it is stored in a special format (such as ArcGIS Network Dataset or PostGIS pgrouting network table), which includes: Node table: Coordinates of all road intersections.

[0066] Side table: Each road segment and its attributes (length, travel time, direction, etc.).

[0067] Topological relationships: the connections between nodes and edges.

[0068] This path dataset supports the following analysis functions: Shortest path query: Given a starting point and a destination, return the total travel time and the path geometry.

[0069] Nearest facility analysis: Find the nearest medical facilities and routes to residential areas.

[0070] Service area generation: Calculates the area that can be reached from a point within a given time / distance.

[0071] (4) Data Output and Updates. The generated network dataset is stored in the PostGIS spatial database and used in conjunction with medical institution point layers, population density rasters, etc. To maintain timeliness, the network dataset needs to be reconstructed periodically (e.g., annually) from the latest road network data, or the travel time cost needs to be dynamically updated via a real-time traffic information API.

[0072] Step 1.5 Constructing the geospatial database. After completing the spatialization processing of multi-source data, this application stores the results in a spatial database through the following steps: First, the population density raster, medical institution point layer, and road network line layer were uniformly converted to the CGCS2000 coordinate system. Using the raster2pgsql and shp2pgsql command-line tools in the GDAL / OGR library, the raster and vector data were imported into a predefined table structure in the PostGIS spatial database, and GIST spatial indexes were created for key attribute fields and spatial geometry columns to accelerate subsequent spatial queries. Second, during the import process, record count verification, geometric validity checks, and spatial extent verification were performed to ensure data integrity and accuracy. Finally, the source, year, number of records, and import time of each data type were recorded in the metadata table, forming a traceable data version and providing a unified, efficient, and reproducible spatial data foundation for subsequent indicator calculations.

[0073] Through the above data entry process, this application achieves the following technical effects: (1) Unified data management. Heterogeneous spatial data are stored in a unified PostGIS database, eliminating differences in data format and coordinate system, and providing standardized data services for subsequent spatial analysis.

[0074] (2) Query performance optimization. Spatial query efficiency is greatly improved by using spatial indexes (GIST) and attribute indexes, and fast index calculation is supported for large-scale regions.

[0075] (3) Traceability. The metadata table and version management mechanism record the source, year, number of records and import time of the data, ensuring that the evaluation results are reproducible and auditable.

[0076] (4) Automation and standardization: The entire warehousing process can be executed automatically by computer programs without human intervention, ensuring the standardization and consistency of data processing and laying a solid foundation for subsequent multi-stage automated evaluation.

[0077] At this point, the first phase is complete, and the unified geospatial database is ready to be used at any time in the second phase (multi-level evaluation index calculation).

[0078] In one exemplary embodiment, the multiple evaluation indicators include: coverage rates of residential communities, administrative villages, populations, and people at health risks within the service radius of hospitals of different levels; the ratio of medical staff to the covered population; accessibility indicators based on spatial supply and demand matching; and the average distance indicator for nearby medical treatment networks.

[0079] In an exemplary embodiment, the accessibility index based on spatial supply and demand matching is calculated using an enhanced two-step move search method, specifically including: Step 1: For each medical institution, within a search range with a network access time threshold as the radius, calculate the Gaussian decay weight of each settlement, calculate the weighted total demand based on the Gaussian decay weight, and then obtain the supply-demand ratio of each medical institution.

[0080] Step 2: For each settlement, within a search range with the network access time threshold as the radius, obtain the supply-demand ratio of accessible medical institutions, and then apply Gaussian decay weighting to perform weighted summation to obtain the accessibility score of each settlement. Finally, aggregate the accessibility scores of all settlements in each evaluation area to obtain the accessibility index of the corresponding evaluation area.

[0081] In practical applications, the aim is to quantify the supply level of basic medical services and the ease with which residents can access them from a spatial perspective. First, a scientific, comprehensive, and operable evaluation index system needs to be constructed. Then, GIS spatial analysis technology is used to calculate the values ​​of each index one by one.

[0082] Step 2.1, the process of constructing the evaluation index system The construction of the indicator system follows these principles: Scientific Principle. The indicators should objectively reflect the level of medical resource allocation and the convenience of residents accessing medical care, and have a solid theoretical basis.

[0083] The principle of systematicity. It takes into account both the "resource supply" and "resident access" dimensions to form an organic whole.

[0084] The principle of operability: The data required for the indicators can be obtained through existing channels, the calculation methods are clear, and they can be repeated.

[0085] Spatial principle. Indicators must reflect geographical spatial distribution characteristics, avoiding the use of purely statistical data.

[0086] Based on the above principles, this application starts from two criterion levels (resource spatial allocation level and residents' access to medical care space) and refines them into 13 tertiary indicators. The specific construction steps are as follows: 2.1.1 Target Decomposition The overall goal of the evaluation index framework in this application is to obtain a basic medical service equalization index and conduct multi-level evaluations. The first is "Resource Spatial Allocation Level," which mainly measures the spatial distribution of medical services from a supply perspective, including the breadth of coverage of medical institutions to residential areas, administrative villages, populations, and people at health risks, as well as the matching strength between the number of medical personnel and the service population. The second is "Residents' Spatial Convenience in Accessing Medical Services," which focuses on reflecting the ease with which residents access medical services from a demand perspective. It is measured through two indicators: spatial accessibility (a comprehensive indicator considering supply, demand, and distance attenuation) and minimum impedance (the actual network distance for residents to reach the nearest medical institution). Together, these two indicators constitute a comprehensive evaluation system for the spatial equity of medical services.

[0087] 2.1.2 Basis for Indicator Selection D1~D4: Residential communities and administrative villages were selected as representative settlements. The differences in service radius of hospitals of different levels (hospital as a whole, secondary and tertiary) were considered to reveal the spatial coverage of medical services at different levels.

[0088] D5~D 10 By taking the population and health-risk groups as the main demand subjects, the proportion of the population covered by hospitals at all levels is calculated, which directly reflects the coverage of the service population. Among them, health-risk groups (such as the elderly over 65 years old) are the key targets of medical services, and listing them separately helps to assess the protection level of vulnerable groups.

[0089] D 11 The ratio of medical staff to the population covered reflects the intensity of service and avoids the one-sidedness of using only the number of beds.

[0090] D 12 The enhanced two-step move search (E2SFCA) method is used to calculate accessibility. This method is widely recognized internationally as an effective tool for measuring spatial supply and demand matching. The introduction of Gaussian decay is more in line with actual medical treatment behavior.

[0091] D 13 The average distance to the nearest medical network directly reflects the spatial resistance to residents' access to medical care and serves as a supplementary indicator of accessibility.

[0092] 2.1.3 Constructing a 13-item evaluation index system After the above analysis and screening, 13 tertiary indicators were finally determined, as shown in Table 2.

[0093] Table 2 Evaluation Index System for Equalization of Basic Medical Services Considering Spatial Characteristics

[0094] 2.2 Calculation process of evaluation indicators Before the index calculation, the first stage, namely data spatialization processing, was completed, generating the following basic layers: residential community point layer (including population attributes), administrative village point layer, population density raster (pixel values ​​represent population numbers), health risk population density raster (optional, or extracted proportionally from the population raster), medical institution point layer (including level, number of beds, and standardized service capacity values), and traffic network dataset (including travel time costs). All layers use a unified projection coordinate system (such as Albers equal area projection) to ensure the accuracy of spatial analysis.

[0095] The base layer is the data stored in the geospatial database. In other words, the geospatial database actually stores the base layer. Think of the geospatial database as a container, and the base layer is the general term for the tables stored within it. There are many specific layers, such as a residential area point layer and a medical institution point layer. Alternatively, it can be stated as, "The first stage, step 1, has been completed, resulting in the geospatial database, which contains the following spatial data layers."

[0096] 2.2.1 Coverage indicators (D1~D) 10 )calculate General steps: Determine the service radius. Based on the "Urban Residential Area Planning and Design Standard" (GB50180-2018) and a survey of residents' travel intentions, determine the service radius (straight-line distance) for hospitals of different levels.

[0097] General hospitals (including Level 1 and unrated hospitals): 1km Level 2 hospital: 5km Tertiary hospital: 20km Considering the road curvature coefficient, an equivalent transformation can be performed based on the network distance in practical applications. However, to simplify the calculation, this application adopts a straight buffer zone.

[0098] Create buffer zones. Generate circular buffer surface layers with corresponding radii centered on the medical institution points. Create separate buffer zones for different hospital levels, such as "Hospital Buffer Zone," "Level II Hospital Buffer Zone," and "Level III Hospital Buffer Zone."

[0099] Spatial Connectivity and Statistics Spatially overlay the buffer layer with the target layer (such as residential area points, administrative village points, population raster).

[0100] For point targets, use the "Spatial Connection" tool to determine whether a point falls into the buffer and count the number of points that fall into the buffer.

[0101] For raster targets, use the "Partition Statistics" tool to calculate the sum of raster cell values ​​(i.e., total population) within each buffer.

[0102] When multiple buffers overlap, a point or grid may be covered by multiple hospitals. To avoid double-counting the total covered population, this application employs a "population allocation method," whereby if a geographic unit is covered by k hospital buffers, the population of that unit is evenly distributed among these k hospitals, with each hospital counting 1 / k of the population. The sum of the population covered by all hospitals ultimately represents the actual total covered population (without duplication).

[0103] Calculate coverage rate. The total population covered by a certain type of hospital. Divide by the total population of the region Multiply this by 100% to get the population coverage rate of that type of hospital. For the coverage rate of residential communities / administrative villages, it is the number of coverage points divided by the total number of points.

[0104] The following uses D7 (population coverage within the service radius of tertiary hospitals) as an example to explain in detail: Input: Tertiary hospital point layer, population density raster layer, total population of the region (Known).

[0105] step: Create a tertiary hospital buffer zone (radius 20km).

[0106] Overlay the buffers with the population raster and use partition statistics to obtain the original population summation within each buffer. .

[0107] Identifying overlapping regions: Each cell is labeled with which hospital buffers it is covered by (e.g., using binary encoding) through raster calculations. If a cell is covered by multiple hospitals, its population value is evenly distributed among those hospitals.

[0108] Recalculate the adjusted population coverage for each hospital And sum them up to get the total population covered. .

[0109] Calculate coverage .

[0110] Output: D7 value (percentage).

[0111] The same applies to other coverage metrics; simply replace the target layer (residential community points, administrative village points, and health risk grids).

[0112] 2.2.2 Medical staff to population coverage ratio (D 11 )calculate This indicator reflects the intensity of healthcare service supply, and its calculation formula is as follows: in: , The number of practicing physicians and registered nurses in medical institution j (known quantities, from medical resource topic data).

[0113] : The total population covered within the service radius of all hospitals (i.e., the denominator in D5), in people.

[0114] Multiplying by 10,000 is to get "the number of healthcare workers per 10,000 people".

[0115] Note: If a region has a very small population, the ratio may be abnormally high, but this is rare in reality and can be smoothed out in subsequent normalization.

[0116] 2.2.3 Accessibility based on spatial supply and demand matching (D 12 )calculate This application employs the Enhanced Two-Step Shift Search (E2SFCA) method, which introduces a Gaussian decay function to make the reachability more realistic.

[0117] Figure 4 This diagram illustrates the principle of the two-step move-forward search method (E2SFCA) incorporating Gaussian decay weights, showcasing the two-step reachability calculation process centered on the supply and demand points under distance decay weight constraints. The overall process consists of three... Figure 4 It consists of (a), (b) and (c), where, Figure 4 In the diagram, (a) represents the distance decay weighting function. Figure 4 (b) and (c) in the text correspond to the first and second steps of the two-step move search method, respectively. Figure 4 (a) in the figure: Gaussian decay function curve (distance decay weight), showing the effect of network travel time. Changing distance decay weight When the passage time When the search threshold is 30 minutes, the weight decreases with distance according to a Gaussian function; when At that time, the weight is 0. The graph uses a blue curve to represent the decay function and a gray dashed line to mark the search threshold. The key points (0,1) and were marked. The right side shows the segmented calculation formula for the attenuation weight.

[0118] Figure 4 (b) The first step is to calculate the supply-demand ratio. A five-pointed star symbol ★ is drawn in the center to represent a medical institution. The label next to it reads " (Supply capacity, such as the number of beds). The search radius is represented by a dashed circle centered on a pentagram.) Multiple small blue dots ● are drawn inside the dashed circle to represent settlements. Each dot is marked with " (Population) and corresponding decay weight A solid blue line with arrows points from residential areas to medical institutions, indicating the weighted demand flow. The formula is labeled below: Figure 4 Step (b) 2: Calculate the accessibility of settlements A i A small blue dot ● is drawn in the center to represent a settlement. The search radius is represented by a dashed circle centered on the dot. Multiple red five-pointed stars (★) are drawn inside the dashed circle, representing medical institutions. Each five-pointed star is marked with " (The supply-demand ratio calculated in the first step) and the corresponding decay weight A solid red line with arrows points from medical institutions to residential areas, indicating the weighted flow of services. The formula is labeled below: Accessibility index based on spatial supply and demand matching (D 12 The specific calculation process is as follows: Step 1: Calculate the supply-demand ratio for each medical institution. For each medical institution : (1) Determine the search threshold (e.g., a 30-minute drive), search for all entries in the network dataset. Residential areas accessible within a short period of time .

[0119] (2) Calculate the number of settlements. arrive Network access time (Known quantities, obtained through network analysis).

[0120] (3) Calculate the weights using the Gaussian decay function. : This function is in The time value is 1, in The time value is 0, and it decays smoothly in the middle.

[0121] (4) Calculate the weighted total demand for the services provided by the medical institution: in residential area The population (a known quantity).

[0122] (5) Calculate the supply-demand ratio: in Standardized service capacity value for medical institutions (Known quantities, from the first stage, i.e., step 1).

[0123] Step 2: Calculate the accessibility of each settlement. For each settlement : (1) Find all in Medical institutions that can be reached within a short period of time .

[0124] (2) For each accessible medical institution, extract its supply-demand ratio. And apply Gaussian decay weights again. .

[0125] (3) Summing up yields the settlements. Accessibility score: Step 3: Calculation of Regional Accessibility Indicators All residential areas in the region Calculate the average (or population-weighted average) to obtain the overall accessibility index for the region. : in This represents the number of settlements within the area. A higher value indicates better spatial accessibility for residents to access medical services.

[0126] Parameter description: , , All of them are known quantities or calculable quantities.

[0127] The preset threshold (e.g., 30 minutes) can be adjusted according to the actual situation in the area.

[0128] The constant in the Gaussian decay function It is a normalization factor that ensures the function range is [0,1].

[0129] 2.2.4 Average distance to nearest medical care network (D) 13 )calculate This indicator directly reflects the travel distance (time) of residents to the nearest medical institution.

[0130] Calculation steps: Residential areas (or population centers) are designated as "event points," and medical institutions are designated as "facility points."

[0131] Use the "Nearest Facility Point" solver in network analysis to calculate the shortest network path distance (in km) from each settlement to the nearest medical facility.

[0132] Get the shortest distance to each settlement .

[0133] Calculate the average value for all settlements: If population weighting is considered, it can be changed to population-weighted average.

[0134] Note: When a residential area is within the service area of ​​a medical institution, the distance may be 0 or very small; if it is far from the medical institution, the distance will be large. This indicator is a cost-based indicator, and the smaller the value, the better.

[0135] 2.2.5 Overlap and Edge Processing Strategies Buffer overlap. As mentioned earlier, the "population allocation method" is used to avoid double counting.

[0136] Regional boundary effect. Residential areas near regional boundaries may be covered by hospitals outside the region, but this application evaluates the level of service within the region, therefore cross-regional medical care is not considered (unless specifically defined). For greater precision, the search scope can be expanded to adjacent regions, but this will increase data requirements.

[0137] Boundaries in network analysis: The network dataset should cover the entire study area and a certain range around it to ensure that settlements near the boundary can be correctly connected to the road network.

[0138] 2.3 Output of index calculation results After all indicators are calculated, an m×n original decision matrix is ​​generated. ,in: Rows: m evaluation regions (e.g., prefecture-level administrative regions) Column: n indicators (D1~D) 13 ) This matrix will serve as input for the next stage (entropy weight method). Each cell in the matrix... That is, the first The region in the Calculated values ​​for each indicator.

[0139] To ensure data quality, the following verifications are required: Check for missing values ​​(if a region cannot calculate a certain indicator due to data issues, mark it and use interpolation or removal).

[0140] Check for outliers (such as coverage exceeding 100%, which is usually caused by data errors and requires backtracking).

[0141] In an exemplary embodiment, S3 specifically includes: S31: The original decision matrix is ​​normalized using a linear scaling transformation method based on the range to obtain a normalized decision matrix. Benefit-type indicators are normalized using a forward normalization formula, while cost-type indicators are normalized using a backward normalization formula, ensuring all indicator values ​​are mapped to the [0,1] interval with a consistent direction (larger values ​​are better).

[0142] S32: Based on the normalized decision matrix, calculate the feature weight of each evaluation index, and calculate the information entropy of each evaluation index based on the feature weight.

[0143] S33: Calculate the difference coefficient of each evaluation index based on the information entropy, normalize and sum the difference coefficients to obtain the objective weight of each evaluation index.

[0144] In practical applications, the core task is to objectively assign weights to the various evaluation indicators calculated in the second stage, i.e., step 2. The scientific nature of the weight determination is crucial to the credibility of the comprehensive evaluation results. To avoid the subjectivity and arbitrariness of traditional methods (such as expert scoring and equal weighting), this application introduces the entropy weighting method from information theory. The basic principle of the entropy weighting method is that the smaller the information entropy of an indicator, the greater the degree of variation of the indicator among different evaluation objects, and the greater the amount of information it provides, thus it should be given a higher weight; conversely, the larger the information entropy of an indicator, the smaller the difference in the value of each object on that indicator, the smaller its contribution to distinguishing evaluation objects, and the lower its weight should be. This method is entirely data-driven and can objectively reflect the importance of each indicator in the evaluation system. The entire weighting process does not rely on any expert experience and is entirely based on the degree of variation of the original data, avoiding subjective bias and controversy.

[0145] Figure 5 This demonstrates the complete computational process of the entropy weight method, with an m×n original decision matrix as input. After normalization, feature weighting, entropy calculation, and difference coefficient calculation, the final weight vector is obtained. A top-down sequential structure is adopted, with the calculation formula noted in each step box.

[0146] 3.1 Data normalization and decision matrix construction.

[0147] Input an m×n original decision matrix : Number of evaluation areas (For example, the 14 prefecture-level administrative regions of a certain province or autonomous region).

[0148] Number of evaluation indicators (Indicator numbers D1 to D) 13 ).

[0149] Original decision matrix ,in Indicates the first The region in the The original values ​​for each indicator. This matrix is ​​calculated in the second stage (S2), and all data are known quantities.

[0150] Processing procedure: Indicator attribute discrimination Efficiency indicators: The higher the value, the higher the level of equalization, including D1~D 12 .

[0151] Cost-related indicators: The smaller the value, the higher the level of equalization, including D. 13 (Average distance to nearest medical care network).

[0152] Normalization (dimensionless): To eliminate the influence of different dimensions and orders of magnitude of indicators, the original data needs to be normalized, mapping each indicator value to the [0,1] interval. This application uses the range normalization method, and the specific formula is as follows: For efficiency-type indicators: in, and They represent the first The minimum and maximum values ​​of this indicator across all regions. When hour, ;when hour, For cost-related indicators: At this time, when (i.e., when the distance is minimum) ;when hour, This satisfies the "smaller is better" conversion.

[0153] Constructing the normalized decision matrix: After normalization, the normalized decision matrix is ​​obtained. All elements satisfy .

[0154] If a certain indicator has the same value in all regions (i.e.) If the denominator of the normalization formula is zero, then the indicator has no discriminative power and should be assigned a weight of zero directly, or removed in subsequent processing. It is necessary to automatically detect such cases and mark the indicator's weight as 0.

[0155] 3.2 Calculate the information entropy and difference coefficient of the indicators Processing procedure: Calculate the proportion of features : in, Indicates the first The first item under the indicator The contribution of each region. To ensure that subsequent logarithmic calculations are meaningful, when When, define .

[0156] Calculate the first Information entropy of the indicator : in, This is an adjustment coefficient, its function is to make... The value range is controlled between [0,1]. When the contribution of all regions... When they are completely equal, When the contribution of a certain region approaches 1 (i.e., other regions are almost zero), Approaching 0. The smaller the information entropy, the greater the degree of variation of the indicator and the greater the amount of information it carries.

[0157] Calculate the coefficient of difference : Coefficient of difference Directly reflects the first The degree of variation of the indicator across regions The larger the value, the greater the contribution of the indicator to differentiating the level of equalization among regions, and the higher its weight should be.

[0158] Parameter description: : No. The information entropy of the item index (an unknown quantity, calculated by the formula).

[0159] : No. The coefficient of variation (unknown quantity) of the indicators.

[0160] Total number of evaluation areas (known quantity).

[0161] : constant, from Determine (known quantities).

[0162] : Natural logarithm.

[0163] Example Explanation: Suppose that a certain indicator (such as the population coverage rate of tertiary hospitals, D7) has a very high value in some regions and a very low value or even zero value in other regions. In this case, the data distribution of this indicator is highly discrete, and its entropy value... Smaller, coefficient of difference A larger value results in a higher final weight; conversely, if a certain indicator (such as the coverage rate of general hospitals and residential communities, D1) has similar values ​​in different regions, its entropy value is close to 1, the difference coefficient is close to 0, and the weight is very low.

[0164] 3.3 Calculate the weight of the indicators Processing procedure: Normalize the difference coefficients to obtain the weights of each indicator. : in, Indicates the first The objective weight of each indicator satisfies .

[0165] The weight vector output will be the calculated weight vector. The data is stored and passed to the next stage (TOPSIS comprehensive evaluation). Simultaneously, a weight distribution table is created, listing the entropy value, difference coefficient, and weight value of each indicator for user reference.

[0166] Special case handling: If the coefficient of difference of a certain indicator (i.e., all regions have the same indicator value), then its weight This indicates that the indicator does not contribute to the evaluation and can be removed. At the same time, the weight calculation results need to be normalized and verified to ensure that the sum of the weights is 1 (allowing for minor floating-point errors).

[0167] 3.4 Technical Effects of This Stage By employing the entropy weight method, this application achieves the following beneficial effects: By minimizing subjective factors in multi-indicator evaluations, the scientific rigor and credibility of the evaluations are improved.

[0168] The weights are dynamically adjusted according to the data, making them suitable for horizontal or vertical comparisons across different periods and regions.

[0169] It helps identify key influencing factors and provides data support for subsequent decision-making recommendations.

[0170] The weight vector output in this stage It is the core input for the next stage of TOPSIS comprehensive evaluation, and together with the original indicator matrix obtained in the second stage, it forms the basis for weighted decision-making, ensuring that the final comprehensive evaluation index is objective and reliable.

[0171] In an exemplary embodiment, S4 specifically includes: S41: Based on the TOPSIS model, the original decision matrix is ​​vector normalized, and a weighted normalized decision matrix is ​​constructed by combining the objective weights.

[0172] S42: Based on the attribute type of each evaluation index, select the optimal and worst values ​​of each index from the weighted normalized decision matrix to form the positive ideal solution vector and the negative ideal solution vector, respectively.

[0173] S43: Calculate the Euclidean distance from the weighted normalized decision vector corresponding to each evaluation region to the positive ideal solution vector and the negative ideal solution vector, respectively.

[0174] S44: Calculate the relative proximity of each evaluation region based on the Euclidean distance, and use the relative proximity as the comprehensive evaluation index.

[0175] In practical applications, the core task of S4 is to comprehensively rank each evaluation region based on the objective weights determined in the third stage using the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS), and calculate the comprehensive evaluation index for the equalization of basic medical services in each region. The basic idea of ​​the TOPSIS method is to define a positive ideal solution (a virtual solution where all indicators reach their optimal values) and a negative ideal solution (a virtual solution where all indicators reach their worst values). Then, the distance between each evaluated object and the positive and negative ideal solutions is calculated, and finally, the relative proximity is used as the basis for comprehensive evaluation. This method has advantages such as intuitive geometric meaning, simple calculation, and ranking of results, and is widely used in multi-indicator decision-making.

[0176] To visually demonstrate the geometric significance of the TOPSIS method, this application provides Figure 6 This diagram illustrates the principle of the TOPSIS model for calculating the comprehensive evaluation index. It uses a two-dimensional plane to represent two evaluation indicators (in practice, this is a 13-dimensional plane, simplified here), with points... , , This represents three regions that are yet to be evaluated. Figure 6 The positive ideal solution is marked in the middle. and negative ideal solution Connect each point to the ideal solution with a dashed line and mark the distance. and .

[0177] The specific implementation steps for this stage are as follows: 4.1 Construct a weighted normalized decision matrix.

[0178] enter: Original decision matrix ,in To evaluate the number of regions, The number of evaluation indicators is n=13. This matrix is ​​calculated in the second stage (S2).

[0179] Weight vector It is calculated by the third stage (S3) entropy weight method.

[0180] Processing procedure: Vector normalization: To eliminate the influence of different indicator dimensions while preserving the relative magnitudes of the indicator values, this application employs vector normalization to standardize the original decision matrix. The calculation formula is as follows: Where the denominator is the first The sum of the squares of the indicators across all regions is the square root of the sum of their squares. After normalization, the sum of the squares of all indicator values ​​is 1, thus eliminating the dimensions.

[0181] Weighted processing: Multiplying the normalized matrix by the weight vector yields the weighted normalized decision matrix. : The weighted matrix contains information from the original indicators and also reflects the differences in importance among the indicators.

[0182] Output: Weighted normalized decision matrix .

[0183] Parameter description: : No. The region in the The original value (known quantity) of the indicator.

[0184] : No. The weight of each indicator (a known quantity, calculated from stage S3).

[0185] : The normalized value of the vector (intermediate variable).

[0186] : Weighted normalized value (intermediate variable, used in subsequent steps).

[0187] 4.2 Determining the Positive and Negative Ideal Solutions Processing procedure: Define the type of ideal solution: For benefit-type indicators (the larger the value, the better), the positive ideal solution takes the maximum value, and the negative ideal solution takes the minimum value.

[0188] For cost-related indicators (the smaller the value, the better), the positive ideal solution takes the minimum value, and the negative ideal solution takes the maximum value.

[0189] Calculate the ideal solution and negative ideal solution : Output: Positive ideal solution vector and negative ideal solution vector .

[0190] Parameter description: : No. The optimal value of this indicator across all regions (unknown quantity, from) (Calculation in Chinese).

[0191] : No. The worst value of this indicator across all regions (unknown quantity, from) (Calculation in Chinese).

[0192] 4.3 Calculating Distance and Relative Proximity Processing procedure: Calculate Euclidean distance: Each evaluation object Distance to the ideal solution and distance to the negative ideal solution Calculate using the following formula: The smaller the distance, the closer the evaluated object is to the ideal solution.

[0193] Calculate the relative proximity (comprehensive evaluation index): The range of values ​​is .when When this occurs, it indicates that the region coincides with the ideal solution (i.e., all indicators have reached their optimal state). When the value is 0, it indicates that the region coincides with the negative ideal solution (i.e., all indicators reach the worst state). The higher the value, the higher the level of equalization of basic medical services in the region.

[0194] Sorting and ranking: according to The values ​​are sorted in descending order across all regions to determine the ranking of their level of equalization. Meanwhile, The values ​​can serve as the basis for subsequent classification and visualization.

[0195] Output: Comprehensive evaluation index vector of each region .

[0196] according to A sorted list in descending order.

[0197] Key role: pass This value can not only compare the level of equalization in different regions, but also quantify the gap between each region and the ideal state, providing precise targets for policy making.

[0198] Example 2 Example (taking 14 cities in a certain province as an example) For ease of understanding, let's assume that the initial three stages of calculation yielded the original matrix of 13 indicators for the 14 cities. and weight vector This section only considers three regions (City A, City B, and City C) and two indicators (D7 and D). 12 The calculation process is demonstrated using an example (13 indicators in actual application), as shown in Table 3.

[0199] Original data (hypothetical values): Table 3. Original values ​​of medical service equalization evaluation indicators for each city

[0200] Weights (derived from the entropy weight method): , .

[0201] Vector normalization: First, calculate the square and square root of each column: D7: D 12 : As shown in Table 4, the normalized matrix : Table 4. Normalized Matrix of Medical Service Equalization Evaluation Indicators in Each City

[0202] As shown in Table 5, the weighted normalized decision matrix : Table 5. Weighted Normalized Decision Matrix of Medical Service Equalization Evaluation Indicators for Each City

[0203] Determine the positive and negative ideal solutions: Both indicators are benefit-oriented, therefore: Positive Ideal Solution Negative ideal solution .

[0204] Calculate distance and relative proximity: City A: City B: City C: Calculation results: In City A In City B City C This directly reflects that City A has the highest level of equalization, City C has the lowest, and City B is in the middle.

[0205] 4.5 Technical Effects of This Stage Objectively quantify the gap: through The value can accurately measure the gap between each region and the ideal equal state, providing a quantitative basis for prioritizing resource allocation.

[0206] The results can be sorted: The values ​​are naturally sortable, which facilitates horizontal comparison and vertical tracking.

[0207] Seamless integration with the entropy weight method: The weights are objectively determined by the entropy weight method, and there is no subjective intervention in the entire evaluation process, from indicator calculation to final index generation, which ensures the scientific nature and repeatability of the evaluation.

[0208] In an exemplary embodiment, S5 specifically includes: using the natural discontinuity grading method to divide the comprehensive evaluation index into multiple levels, and configuring corresponding spatial filling styles for each level to generate a spatial distribution thematic map.

[0209] In one exemplary embodiment, the multidimensional diagnostic report includes a comprehensive evaluation overview, analysis of key weakness indicators, spatial fairness diagnosis, and resource allocation optimization suggestions.

[0210] In practical applications, the core task of S5 is to calculate the comprehensive evaluation index for each region. This process transforms data into intuitive and understandable visualizations, generating structured diagnostic reports that provide decision-makers with precise recommendations for healthcare resource allocation. Visualizations can be created using professional GIS software such as ArcGIS, supporting multiple output formats to meet the needs of users at different levels.

[0211] 5.1 Equalization of Horizontal Spatial Differentiation Mapping The aim is to visually display the numerical distribution of the comprehensive evaluation index in the form of thematic maps, revealing the spatial pattern and regional differences in the level of equalization of basic medical services.

[0212] Technical process: Hierarchical processing Input: Comprehensive evaluation index of each region ( ).

[0213] Method: This application employs the Jenks Natural Breaks (NBR) grading method to automatically determine the grading threshold. This method uses iterative calculations to minimize the internal variance of each level and maximize the variance between each level, thereby optimally reflecting the natural clustering characteristics of the data.

[0214] Suppose it needs to be divided into Each level (preset in this application) These correspond to Level I (high equalization), Level II (relatively high), Level III (medium), and Level IV (low equalization).

[0215] The computer executes Jenks' algorithm to calculate... A breakpoint, The range is divided into A range.

[0216] Alternatively, the standard deviation grading method (divided by multiples of the standard deviation with the mean as the center) or the quantile grading method (making each level contain the same number of regions) can be used, which users can specify in the configuration file as needed.

[0217] Thematic map generation Base map preparation. This involves accessing basic GIS geographic data, including administrative boundary layers, major water system layers, and main transportation road layers.

[0218] Grading and color scheme. Based on the grading results, each administrative division element is assigned a corresponding grayscale or pattern fill (patent drawings are usually in black and white line art, so different densities of dots, lines, or grid fills are used to represent grade differences, for example: Level I: Blank fill (white) Level II: Sparse horizontal line fill Level III: Dense Vertical Line Fill Level IV: Cross-grid fill) Overlapping elements Distribution of medical institutions: The locations of tertiary and secondary hospitals are represented by specific symbols (such as hollow circles and solid triangles).

[0219] Transportation network: Major roads are represented by thin solid lines.

[0220] Population density: Optional, represented by grayscale rendering or point density map.

[0221] Map finishing touches: Automatically add legends, scale bars, north arrows, titles, map approval numbers, and other information. Legends must clearly indicate the fill styles and meanings corresponding to each level.

[0222] Output format: Can output as high-resolution raster images (such as TIFF, PNG, with a resolution of at least 300 DPI) to meet the requirements of patent illustrations and printing. Can also output as vector formats (such as SVG, PDF) for easy subsequent editing and scaling.

[0223] The key function of this step is that the spatial differentiation map enables decision-makers to clearly identify the regional distribution of equalization levels at a glance, quickly pinpoint weak areas that require special attention, and provide intuitive spatial guidance for resource allocation.

[0224] 5.2 Automatic generation of multidimensional diagnostic reports The goal is to integrate the intermediate results and final conclusions of the entire evaluation process into a structured, readable report that is easy to archive, report, and analyze in depth.

[0225] Report structure: Overall Evaluation Overview Overall statistics. Number of statistical regions, equalization index. The mean, maximum, minimum, and standard deviation of the value.

[0226] Ranking list. (By...) The ranking is reduced to show all regions and their scores, with the top three and bottom three regions marked.

[0227] Distribution by tier. Statistics on the number and percentage of regions in each tier are presented in a pie chart or bar chart.

[0228] Analysis of key weakness indicators: For regions with low equalization levels (Level IV), identify and analyze their three lowest-ranked indicators (i.e., those with the largest gap from the optimal value). Generate attribution analysis text for each weakest indicator, for example: "L City's D7 (tertiary hospital population coverage rate) is 0%, far below the regional average of 31.42%, indicating that the lack of tertiary hospital coverage in this region is the primary factor leading to the low equalization level." Simultaneously, radar charts of each indicator for the region can be displayed, comparing them with the optimal and average values.

[0229] Spatial equity diagnosis: Lorenz curve and Gini coefficient modules are introduced to analyze the equity of medical resources (such as beds and medical staff) distributed by population.

[0230] Calculation steps: The regions are arranged in ascending order of per capita resources (such as the number of beds per thousand people).

[0231] Plot the Lorenz curve: the horizontal axis represents the cumulative percentage of population, and the vertical axis represents the cumulative percentage of resources.

[0232] Calculate the Gini coefficient ,in This represents the cumulative population percentage. This represents the cumulative resource ratio.

[0233] Outputs the Gini coefficient value and fairness level (e.g., <0.2 indicates absolute average, 0.2-0.3 indicates relatively average, 0.3-0.4 indicates reasonable, 0.4-0.5 indicates large disparity, >0.5 indicates extreme disparity). It can generate a Lorenz curve to visually demonstrate the degree to which resource allocation deviates from the absolute average.

[0234] Resource allocation optimization suggestions: Based on the analysis of shortcomings and the diagnosis of spatial fairness, targeted optimization suggestions are formulated.

[0235] Suggested generation logic: If a region has an extremely low D7 (population coverage rate of tertiary hospitals) and the accessibility of tertiary hospitals in the surrounding areas is good, it is recommended to strengthen transportation connections or establish a hierarchical diagnosis and treatment referral mechanism.

[0236] If D7 is extremely low and there are no tertiary hospitals nearby, it is recommended to plan and build a new branch of a tertiary hospital in the area or upgrade the existing hospital.

[0237] If D 11 The ratio of medical staff to healthcare workers is too low. It is recommended to increase the number of medical staff or recruit more talent.

[0238] If D13 The average distance to nearby medical care is too large. It is recommended to increase the number of primary healthcare institutions or optimize the road network.

[0239] Each suggestion is accompanied by an estimated expected effect, such as: "It is recommended to plan and build a new secondary hospital in the eastern new district of City A, which can improve the D3 and D6 indicators of approximately 100,000 people and is expected to increase the equalization index of the region by 0.15." Example 3 The method described in this application will be explained in detail below using an application example from a certain province or autonomous region (hereinafter referred to as "the province"). This embodiment is intended to demonstrate the feasibility and practicality of this application and is not intended to limit the scope of protection of this application. Those skilled in the art can adjust the parameter settings and data processing methods based on the teachings of this embodiment and the characteristics of the specific evaluation area.

[0240] Study area and data preparation: 1.1 Study Area This embodiment uses 14 prefecture-level administrative regions (including cities A, B, C, D, E, F, G, H, J, K, L, M, N, and O) in a province or autonomous region as evaluation units. The total area of ​​the study region is 1.66 million square kilometers, with a complex geographical environment, uneven population distribution, and significant differences in the allocation of medical resources, making it suitable as a typical example for evaluating the equalization of basic medical services.

[0241] 1.2 Data Acquisition In accordance with the requirements of the first phase (S1) of this application, the following four types of data shall be collected: Basic geographic framework data: sourced from the Surveying and Mapping Science Research Institute of a certain province / autonomous region, including: Boundaries of 14 prefecture-level administrative divisions (surface layer), scale 1:1,000,000, coordinate system CGCS2000.

[0242] Residential data: including urban residential communities (3,856 in total) and administrative villages (8,790 in total), each with a unique identifier.

[0243] Road network data: Includes road elements such as expressways, national highways, provincial highways, county roads, and township roads, totaling 125,000 road segments. Each road segment records its length, grade, and design speed.

[0244] Medical resource data: sourced from a provincial health commission's 2020 statistical report, including: There are a total of 1,862 medical institutions at all levels, including 42 tertiary hospitals, 168 secondary hospitals, and 1,652 primary and unclassified hospitals.

[0245] Each record includes the institution's name, level, number of beds, number of licensed physicians, number of registered nurses, and detailed address.

[0246] Population and social statistics: sourced from a provincial statistical yearbook (2020) and the Seventh National Population Census Bulletin, including: Total population by district / county (25.85 million permanent residents in the district).

[0247] Age structure: The elderly population aged 65 and above accounts for 8.5% of the total population, serving as a proxy for people at health risk.

[0248] Gridded population distribution data: The WorldPop dataset (100m resolution) was used to assist in spatialization.

[0249] Traffic network attribute data: Design speeds are set for different grades of roads according to the "Technical Standards for Highway Engineering" and urban road design specifications, as shown in Table 6.

[0250] Table 6 Design Speeds for Roads of Different Grades

[0251] 1.3 Data Spatialization Processing Refined spatial distribution of population data: An improved Monte Carlo-kernel density fusion algorithm is employed. First, building outlines extracted from high-resolution remote sensing imagery are used as the distribution base. Then, the population of each district / county is randomly assigned to buildings based on building area weights (number of simulations). Then, kernel density estimation (500m bandwidth) is performed on the allocation results to generate a population density raster with a resolution of 100m. This raster covers the entire area, and the cell value represents the population within that grid.

[0252] Quantifying the service capacity of medical institutions: setting adjustment coefficients based on hospital level (Tertiary Hospital) Secondary hospitals Level 1 and unrated ), calculate the standardized service capacity value for each medical institution. For example, a tertiary hospital in a certain province has 3,000 beds. .

[0253] Traffic network cost model construction: Construct a topological network based on road network data, and assign a travel time cost to each edge. (Unit: minutes). The network dataset was built using ArcGIS Network Analyst and supports shortest path and nearest facility point analysis.

[0254] 2. Calculation of evaluation indicators 2.1 Construction of the indicator system Thirteen tertiary indicators were constructed according to Table 2. In this embodiment, the hospital service radius was set based on residents' travel intentions and actual road conditions: Service radius of general hospitals (including level one): 1km (15-minute walk away) Service radius of the secondary hospital: 5km (9 minutes by car) Service radius of tertiary hospitals: 20km (36 minutes by car) Service radius of medical institutions in administrative villages: 3km 2.2 Calculation of Coverage-Related Indicators (Taking D7 as an Example) Explanation of the calculation process based on the population coverage rate of tertiary hospitals: 42 tertiary hospital locations were extracted from the medical institution point layer, and a buffer zone with a radius of 20km was created. A straight-line buffer zone was used as an approximation to account for actual road traffic.

[0255] Overlay the tertiary hospital buffer layer with the population density raster, and use the partition statistics tool to calculate the original total population within each buffer.

[0256] Handling buffer overlap: A population allocation method is used, that is, for a grid cell that is covered by k buffers, its population is evenly distributed to k hospitals, and then the population covered by each hospital is re-accumulated.

[0257] Calculate the total population covered by tertiary hospitals in the entire region Divide by the total population of the district 10,000, to get D7= Calculations show that the population coverage rate of tertiary hospitals in a certain province is 31.42%. The D7 values ​​for each prefecture-level city are shown in Table 7.

[0258] 2.3 Calculation of Accessibility Index (D) 12 ) An enhanced two-step move search (E2SFCA) method is employed, with a threshold time set. Minutes (driving distance), Gaussian decay function parameters are the same as before.

[0259] Step 1: Calculate the supply-demand ratio R j Taking a tertiary hospital in City A as an example, its standardized service capabilities... Through network analysis, search for settlements reachable within 30 minutes (using the population grid center as the demand point) and calculate the weighted total demand. , to obtain R j =4500 / 268500=0.0168.

[0260] Step 2: Calculate the accessibility A of the settlement. i For each settlement, add up the R... of all hospitals reachable within 30 minutes. jThen, Gaussian attenuation is applied again. Finally, the accessibility values ​​of each settlement are obtained, ranging from 0 to 1.5.

[0261] Regional aggregation: The accessibility of each settlement is weighted by population to obtain the accessibility index D for each prefecture-level city. 12 For example, City A, D 12 =1.42, J City D 12 =0.05.

[0262] 2.4 Calculation of Medical Treatment Distance Index (D) 13 ) Starting from each residential point (population grid center) and ending at a medical institution, the network analysis "nearest facility point" solver is invoked to calculate the network distance from each residential point to the nearest medical institution. The arithmetic mean of the nearest distances to all residential points within each prefecture-level city is then calculated to obtain D. 13 For example, the average distance to the nearest medical facility is 1.87 km in City A and 4.53 km in City O.

[0263] 2.5 Summary of Indicators After calculating all 13 indicators, the original decision matrix is ​​obtained. (14 rows × 13 columns). Some data is shown in Table 7. To save space, only examples of key indicators are listed.

[0264] Table 7. Original values ​​of some medical service equalization evaluation indicators in 14 cities of a certain province (example)

[0265] 3. Objective weight assignment based on entropy weight method 3.1 Data Normalization Preprocessing For the original decision matrix Perform range normalization. Taking D7 as an example, the maximum value is 92.62 and the minimum value is 0. Therefore, the normalized value for City A is... City C Cost-related indicator D 13 Inverse normalization is used to obtain the normalized matrix. .

[0266] 3.2 Calculating Information Entropy and Weights Calculated based on 14 regions and 13 indicators: Adjustment coefficient .

[0267] Calculate the characteristic weight of each indicator Then, the information entropy can be obtained. Coefficient of difference Weight The results are shown in Table 8.

[0268] Table 8 Calculation results using the entropy weight method

[0269] It can be seen that D4 (coverage rate of residential communities with tertiary hospitals), D7 (population coverage rate of tertiary hospitals), and D 10 (Coverage rate of people at health risk in tertiary hospitals), D 12 Accessibility had the highest weighting, totaling approximately 78%, indicating that the availability and spatial accessibility of tertiary hospitals are the core factors affecting the equalization of basic medical services in a certain province.

[0270] 4. Calculation of the comprehensive equalization index based on TOPSIS 4.1 Constructing a weighted normalized decision matrix relative to the original matrix Perform vector normalization to obtain Then multiply by the weight The weighted normalized decision matrix is ​​obtained. Some of the results are shown in Table 9.

[0271] Table 9 Weighted Normalized Decision Matrix (Partial)

[0272] 4.2 Determining the Positive and Negative Ideal Solutions For benefit-related indicators, take the maximum value; for cost-related indicators, take the minimum value. Therefore: Positive Ideal Solution Negative ideal solution 4.3 Calculating Distance and Relative Proximity Taking City A as an example: Similar to calculating all regions, a comprehensive evaluation index is obtained. The rankings are shown in Table 10.

[0273] Table 10. Comprehensive Evaluation Index of Equalization of Medical Services in 14 Cities of a Certain Province

[0274] 5. Results visualization and decision support.

[0275] 5.1 Equalization of horizontal spatial differentiation mapping.

[0276] according to The values ​​were used to classify the 14 cities into 4 levels using the natural discontinuity grading method: Level I (High Equalization): (City A only) Level II (Higher): (City D) Level III (Intermediate): (City B, City M, City E, City L) Level IV (Low Equalization): (The remaining 8 cities) 5.2 Compilation of Multidimensional Diagnostic Reports The following is an excerpt from the preparation of the diagnostic report: Overall Evaluation Overview: The Whole Region The average value was 0.256, the maximum value was 0.968 (City A), and the minimum value was 0.047 (City O). Level IV regions accounted for 57.1%, indicating that the overall level of equalization in the province was relatively low.

[0277] Analysis of key weakness indicators: Taking City O in a Level IV region as an example, its three lowest-ranked indicators are: D4 (Tertiary hospital residential area coverage rate = 0%), D7 (Tertiary hospital population coverage rate = 0%), and D... 12 (Accessibility = 0.07). Attribution analysis: "The lack of tertiary hospitals in City O, coupled with its remote location and low road network density, results in extremely poor accessibility for medical care for residents, which is the main reason for the low level of equalization." Spatial equity diagnosis: The Gini coefficient for the distribution of medical beds by population is calculated to be 0.38, which is in the "significant gap" range. The Lorenz curve shows that 20% of the population owns 35% of the beds, indicating that resources are concentrated in a few cities.

[0278] Resource allocation optimization suggestions: For areas such as O City and H City that lack tertiary hospitals, it is recommended to plan and build a new branch of a tertiary hospital in the regional capital or upgrade the existing secondary hospitals. It is estimated that this could increase the local D7 by more than 30% and C_i by 0.1 to 0.2.

[0279] For areas with poor accessibility, such as Tacheng and F City, it is recommended to optimize road network connections, increase the number of township health centers, and shorten the distance for residents to seek medical treatment.

[0280] For the entire Xinjiang region, it is recommended to strengthen telemedicine collaboration so that residents in remote areas can share high-quality medical resources.

[0281] This embodiment successfully applied the method of this application to quantitatively evaluate the equalization of basic medical services in 14 cities of a province. It identified City A as having the highest level of equalization, while Cities O and H were the lowest. The study also revealed that the configuration and spatial accessibility of tertiary hospitals are key influencing factors. The evaluation results are consistent with the actual situation, demonstrating the feasibility and effectiveness of the method. This application can be extended to evaluations at the provincial or municipal levels, and the indicator system and parameters can be adjusted according to actual needs, demonstrating broad applicability.

[0282] Example 4 In an exemplary embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments. The computer device can be a server or a terminal. The computer device includes a processor, a memory, an input / output interface (I / O), and a communication interface. The processor, memory, and I / O interface are connected via a system bus, and the communication interface is connected to the system bus via the I / O interface. The processor of the computer device provides computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The database of the computer device stores data to be processed. The I / O interface of the computer device is used for exchanging information between the processor and external devices. The communication interface of the computer device is used for communicating with an external terminal via a network connection. When the computer program is executed by the processor, it implements the above-described methods.

[0283] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0284] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0285] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0286] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by hardware related to computer program instructions. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0287] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0288] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0289] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A comprehensive evaluation method for equalizing basic medical services, characterized in that, include: Acquire research data for the study area, and perform spatialization processing on the research data to construct a unified geospatial database; The research data includes basic geographic framework data, medical resource thematic data, population and social statistics data, and transportation network attribute data; Based on the geospatial database, multiple evaluation index values ​​representing the spatial allocation level of medical resources and the convenience of residents' access to medical care are calculated through spatial analysis to form an original decision matrix; The entropy weight method is used to calculate the objective weight of each evaluation index based on the degree of variation of the values ​​of the multiple evaluation indicators in each evaluation region. Based on the TOPSIS model, the original decision matrix is ​​processed using the objective weights to calculate the comprehensive evaluation index of basic medical service equalization in each evaluation region. Spatial hierarchical mapping is performed based on the comprehensive evaluation index for equalization of basic medical services, and a multidimensional diagnostic report is generated.

2. The comprehensive evaluation method for equalization of basic medical services according to claim 1, characterized in that, The evaluation indicators include: coverage rates of residential communities, administrative villages, populations, and people at health risks within the service radius of hospitals of different levels; the ratio of medical staff to the covered population; accessibility indicators based on spatial supply and demand matching; and the average distance to the nearest medical network.

3. The comprehensive evaluation method for equalization of basic medical services according to claim 2, characterized in that, The accessibility index based on spatial supply and demand matching is calculated using an enhanced two-step move search method, specifically including: Step 1: For each medical institution, within a search range with a network access time threshold as the radius, calculate the Gaussian decay weight of each settlement, calculate the weighted total demand based on the Gaussian decay weight, and then obtain the supply-demand ratio of each medical institution. Step 2: For each settlement, within a search range with the network access time threshold as the radius, obtain the supply-demand ratio of accessible medical institutions, and then apply Gaussian decay weighting to perform weighted summation to obtain the accessibility score of each settlement. Finally, aggregate the accessibility scores of all settlements in each evaluation area to obtain the accessibility index of the corresponding evaluation area.

4. The comprehensive evaluation method for equalization of basic medical services according to claim 1, characterized in that, The entropy weight method is used to calculate the objective weight of each evaluation indicator based on the degree of variation of the values ​​of the multiple evaluation indicators across different evaluation regions. Specifically, this includes: The original decision matrix is ​​subjected to range normalization to obtain a normalized decision matrix; wherein, benefit-type indicators and cost-type indicators adopt different normalization mapping rules; Based on the normalized decision matrix, the feature weight of each evaluation index is calculated, and the information entropy of each evaluation index is calculated based on the feature weight. The difference coefficients of each evaluation indicator are calculated based on the information entropy, and the difference coefficients are normalized and summed to obtain the objective weights of each evaluation indicator.

5. The comprehensive evaluation method for equalization of basic medical services according to claim 1, characterized in that, Based on the TOPSIS model, the original decision matrix is ​​processed using the objective weights to calculate the comprehensive evaluation index for the equalization of basic medical services in each evaluation region, specifically including: Based on the TOPSIS model, the original decision matrix is ​​vector normalized, and a weighted normalized decision matrix is ​​constructed by combining the objective weights. Based on the attribute type of each evaluation index, the optimal and worst values ​​of each index are selected from the weighted normalized decision matrix to form the positive ideal solution vector and the negative ideal solution vector, respectively. Calculate the Euclidean distance from the weighted normalized vector corresponding to each evaluation region to the positive ideal solution vector and the negative ideal solution vector, respectively; The relative proximity of each evaluation region is calculated based on the Euclidean distance, and the relative proximity is used as the comprehensive evaluation index.

6. The comprehensive evaluation method for equalization of basic medical services according to claim 1, characterized in that, Spatial hierarchical mapping is performed based on the comprehensive evaluation index for equalization of basic medical services, specifically including: The comprehensive evaluation index is divided into multiple levels using the natural discontinuity grading method, and a corresponding spatial filling style is configured for each level to generate a spatial distribution thematic map.

7. The comprehensive evaluation method for equalization of basic medical services according to claim 1, characterized in that, The multidimensional diagnostic report includes a comprehensive evaluation overview, analysis of key weakness indicators, spatial fairness diagnosis, and recommendations for resource allocation optimization.

8. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the comprehensive evaluation method for equalization of basic medical services as described in any one of claims 1-7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the comprehensive evaluation method for equalization of basic medical services as described in any one of claims 1-7.

10. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the comprehensive evaluation method for equalization of basic medical services as described in any one of claims 1-7.

Images