A regional intelligent chronic disease collaborative management method and system

Abstract

This invention relates to the field of health information management technology, specifically a regional digitalized collaborative management method and system for chronic diseases. It unifies and aggregates physical examination and diagnostic data from medical institutions at all levels, resident behavioral characteristics, and the physical signs and behaviors of authorized monitoring subjects to construct a standardized, interoperable data foundation, eliminating information barriers and supporting closed-loop management throughout the entire process. Based on physical examination diagnoses and resident behavioral characteristics, a behavior-chronic disease correlation map is constructed to accurately identify promoting and hindering behaviors. Through temporal and spatial trend analysis and root cause localization of various chronic diseases, precise attribution at the group level is achieved, enabling the development of differentiated intervention strategies for different regions and changing the methods of health education. Utilizing the physical signs and behaviors of authorized monitoring subjects, individual trend characteristics are dynamically analyzed and root causes are located, achieving automatic identification of disease stages, upgrading from regular follow-up visits to continuous, intelligent individual monitoring, and providing timely warnings and preventing the development of complications.

Description

Technical Field

[0001] This invention relates to the field of health information management technology, specifically a regional digitalized collaborative management method and system for chronic diseases. Background Technology

[0002] To improve the prevention and control of chronic diseases, various regions are actively promoting the coordinated development of regional medical and health services and building integrated service systems such as urban medical groups and medical consortia. However, in practice, regional chronic disease management still faces the following technical bottlenecks:

[0003] (1) Data silos and collaboration barriers are prominent.

[0004] There is a lack of unified data interoperability standards among medical institutions at all levels within the region (general hospitals, primary healthcare institutions, and public health institutions). Patient data, including physical examination records, diagnostic records, laboratory test results, and home monitoring information, generated from visits to different institutions, are scattered across heterogeneous business systems, making effective aggregation and sharing impossible.

[0005] (2) Chronic disease intervention lacks precision and regional adaptability.

[0006] Most existing chronic disease management systems focus on individual-level health record management and follow-up reminders, failing to fully utilize the behavioral characteristics of regional residents to guide prevention and intervention.

[0007] (3) The individual disease monitoring and early warning mechanism is not perfect.

[0008] Existing systems primarily manage the condition of patients with chronic diseases through regular follow-up visits and manual visits.

[0009] The identification of disease progression stages is crude: There is a lack of dynamic stage division models based on continuous time-series data (indicators from previous physical examinations, vital signs at home, and individual behavioral characteristics), making it impossible to accurately determine whether a patient is in a stable period, a fluctuating period, a risk period, or a complication period. Summary of the Invention

[0010] In view of this, the purpose of this invention is to provide a regional digitalized collaborative management method and system for chronic diseases, so as to solve the above-mentioned technical problems.

[0011] To achieve the above objectives, the present invention adopts the following technical solution:

[0012] The present invention provides a regional digitalized collaborative management method for chronic diseases, comprising the following steps:

[0013] The system acquires declassified physical examination and diagnostic data from medical institutions at all levels in the target area, behavioral characteristic data of residents in the target area, and vital sign and behavioral characteristic data of authorized monitoring subjects. The physical examination and diagnostic data includes various functional physical fitness indicators and diagnostic conclusions.

[0014] Based on the physical examination and diagnosis data and the residents' behavioral characteristics data, a behavior-chronic disease association map of the target area is constructed. The behavior-chronic disease association map represents the associated behaviors of various chronic diseases, including promoting behaviors and hindering behaviors.

[0015] Based on multiple cycles of physical examination and diagnostic data of the target area, trend analysis was performed on various chronic diseases in the target area to obtain the temporal and spatial trend characteristics of various chronic diseases; and trend analysis was performed on the vital sign data of authorized monitoring subjects to obtain individual trend characteristics.

[0016] The root causes of the time-trend features are determined by analyzing the behavior-chronic disease association map and resident behavioral characteristic data; and the root causes of the individual trend features are determined by analyzing the behavioral characteristic data of the authorized monitoring subjects and the behavior-chronic disease association map.

[0017] Information management is based on the temporal trend characteristics, spatial trend characteristics, temporal trend root causes, individual trend characteristics, and individual trend root causes of various chronic diseases. The information management includes diffusion blocking management, group publicity management, and individual monitoring management.

[0018] In one embodiment of this application, constructing a behavior-chronic disease association map of a target area based on the physical examination diagnostic data and the resident behavioral characteristic data includes:

[0019] The target area and its physical examination and diagnosis data are divided into multiple geographical units and their physical examination and diagnosis data.

[0020] Calculate the incidence rate of chronic diseases in multiple geographic units And extract feature vectors from resident behavior data of multiple geographical units. , where the feature vector This includes per capita salt intake, fat intake score, vegetable and fruit consumption compliance rate, smoking rate, per capita alcohol consumption, per capita daily physical activity, and the proportion of the population engaged in physical activity. For chronic disease indexing, For geographic unit indexing;

[0021] Construct a spatial autoregressive model, wherein the mathematical expression of the spatial autoregressive model is:

[0022]

[0023] In the formula, These are the spatial autoregressive coefficients. For geographic unit index, Indicates the first The first geographical unit The incidence of chronic diseases in China A collection of geographic unit indexes. This is the spatial weight matrix. For the regression coefficient vector, This is the random error term;

[0024] Chronic disease incidence rates based on multiple geographic units and eigenvectors The spatial autoregressive model is fitted to obtain a regression coefficient vector and a P-value vector, wherein the regression coefficient vector includes regression coefficients of multiple behavioral features, and the P-value vector includes P-values ​​of multiple behavioral features.

[0025] A behavior-chronic disease association map is constructed based on the regression coefficient vector and the p-value vector.

[0026] In one embodiment of this application, the incidence rate of chronic diseases in multiple geographical units is calculated. And extract feature vectors from resident behavior data of multiple geographical units. ,include:

[0027] Extracting cases of various chronic diseases from physical examination and diagnosis data of geographical units;

[0028] The incidence rate of chronic diseases in multiple geographic units is obtained by calculating the ratio of the number of cases of each chronic disease to the total population of the geographic unit. ;

[0029] The data includes per capita salt intake, fat intake score, vegetable and fruit intake compliance rate, smoking rate, per capita alcohol consumption, per capita daily exercise volume, and the proportion of people exercising, extracted from statistical data for multiple geographic units. These data are then standardized to obtain feature vectors representing the behavioral characteristics of residents in multiple geographic units. .

[0030] In one embodiment of this application, constructing a behavior-chronic disease association map based on the regression coefficient vector and the p-value vector includes:

[0031] For each chronic disease, target behavioral features with P values ​​less than a preset association threshold and regression coefficients greater than a preset regression threshold are selected based on the regression coefficient vector and the P-value vector.

[0032] The association direction is determined based on the regression coefficient of the target behavioral feature, wherein when the regression coefficient is positive, the target behavioral feature is a promoting behavior; when the regression coefficient is negative, the target behavioral feature is a hindering behavior.

[0033] The regression coefficients of the target behavioral characteristics for each chronic disease are standardized to obtain the association weights between the target behavioral characteristics and the chronic disease.

[0034] A node set is constructed based on multiple chronic diseases and multiple behavioral characteristics, and an edge set is constructed based on the association relationship, association weight, and association direction between each chronic disease and the target behavioral characteristic;

[0035] Construct a behavior-chronic disease association graph based on the node set and the edge set.

[0036] In one embodiment of this application, trend analysis is performed on various chronic diseases in the target area based on multiple periods of physical examination and diagnostic data, to obtain the temporal and spatial trend characteristics of various chronic diseases, including:

[0037] Extract the incidence rates of chronic diseases in the target region over multiple periods; perform linear fitting based on the incidence rates of chronic diseases over multiple periods to obtain the overall slope of change; and construct time trend features based on the overall slope of change.

[0038] Chronic disease incidence rates of multiple geographic units in the target area for each period are extracted, and an incidence rate heatmap for each period is constructed based on the chronic disease incidence rates of multiple geographic units. Diffusion features are extracted from the incidence rate heatmaps of multiple periods to obtain spatial trend features, wherein the diffusion features include a marker indicating whether diffusion has occurred, diffusion speed, and diffusion direction.

[0039] In one embodiment of this application, trend analysis is performed on the vital sign data of authorized monitoring subjects to obtain individual trend characteristics, including:

[0040] Obtain chronic disease tags of authorized monitoring subjects, and obtain associated vital signs and relationships related to the chronic disease tags, wherein the relationships include positive and negative associations;

[0041] The relevant physical condition values ​​at multiple monitoring time points are extracted from the vital sign data, and linear fitting is performed on the relevant physical condition values ​​at multiple monitoring time points to obtain the slope of the change in the relevant physical condition; and personal trend features are constructed based on the slope of the change in the relevant physical condition.

[0042] In one embodiment of this application, the root cause localization of the time trend features is performed based on the behavior-chronic disease association map and resident behavioral characteristic data to obtain the time trend root cause, including:

[0043] When the absolute value of the time trend feature of any target chronic disease is greater than or equal to the preset slope threshold, extract the resident behavior feature data sequence of multiple time periods before the current time period;

[0044] Based on the behavior-chronic disease association map, target behavioral features of the target chronic disease are extracted, and the values ​​of the target behavioral features in the resident behavioral feature data sequence are linearly fitted to obtain the slope of the target behavioral features.

[0045] Target behavioral features whose absolute slope is greater than or equal to a preset slope threshold and whose correlation matches are used as root causes of time trends.

[0046] In one embodiment of this application, the root cause of an individual's trend characteristics is determined by analyzing the behavioral characteristic data of the authorized monitoring object and the behavior-chronic disease association map, thereby obtaining the root cause of the individual trend, including:

[0047] When the absolute value of the individual trend feature is greater than or equal to a preset slope threshold, the target behavioral feature associated with the chronic disease label of the authorized monitoring object is determined based on the behavior-chronic disease association map.

[0048] Linear fitting is performed on the values ​​of the target behavioral features at multiple time points in the behavioral feature data to obtain the slope of the target behavioral features;

[0049] Target behavioral features whose absolute slope is greater than or equal to a preset slope threshold and whose correlation matches are taken as individual trend root causes.

[0050] In one embodiment of this application, information management is performed based on the temporal trend characteristics, spatial trend characteristics, temporal trend root causes, individual trend characteristics, and individual trend root causes of multiple chronic diseases, including:

[0051] When time-trend characteristics indicate an increase in the incidence of chronic diseases, group outreach should be conducted based on the root causes of the time trend.

[0052] When the individual trend characteristics indicate that the condition is improving or worsening, the monitoring results will be sent to the authorized monitoring subjects;

[0053] Based on the spatial trend characteristics of various chronic diseases, chronic disease diffusion analysis is performed. When there is a diffusion trend of chronic diseases and the diffusion speed is greater than the set speed threshold, high-risk areas are delineated based on the diffusion direction and diffusion speed. The high-risk areas are then sent to the diffusion blocking planning module as the data basis for diffusion blocking schemes.

[0054] This application also provides a regional digitalized chronic disease collaborative management system, including:

[0055] The acquisition module is used to acquire declassified physical examination and diagnosis data from medical institutions at all levels in the target area, behavioral characteristic data of residents in the target area, and vital sign data and behavioral characteristic data of authorized monitoring subjects. The physical examination and diagnosis data includes various functional physical fitness indicators and diagnostic conclusions.

[0056] The association map construction module is used to construct a behavior-chronic disease association map of the target area based on the physical examination and diagnosis data and the residents' behavioral characteristic data. The behavior-chronic disease association map represents the association behaviors of various chronic diseases, including promoting behaviors and hindering behaviors.

[0057] The trend analysis module is used to perform trend analysis on various chronic diseases in the target area based on multiple periods of physical examination and diagnosis data, and to obtain the temporal and spatial trend characteristics of various chronic diseases; as well as to perform trend analysis on the vital sign data of authorized monitoring subjects to obtain individual trend characteristics;

[0058] The root cause localization module is used to perform root cause localization on the time trend features based on the behavior-chronic disease association map and resident behavioral characteristic data to obtain the time trend root cause; and to perform root cause localization on the individual trend features based on the behavioral characteristic data of the authorized monitoring subjects and the behavior-chronic disease association map to obtain the individual trend root cause.

[0059] The information management module is used for information management based on the temporal trend characteristics, spatial trend characteristics, temporal trend root causes, individual trend characteristics, and individual trend root causes of various chronic diseases. The information management includes diffusion blocking management, group publicity management, and individual monitoring management.

[0060] The beneficial effects of this invention are as follows: This invention provides a regional digitalized chronic disease collaborative management method and system. By uniformly aggregating physical examination and diagnostic data from medical institutions at all levels, resident behavioral characteristics, and the physical signs and behaviors of authorized monitoring subjects, it constructs a standardized and interoperable data foundation, eliminating information barriers and supporting closed-loop management throughout the entire process. Based on physical examination diagnoses and resident behavioral characteristics, it constructs a behavior-chronic disease correlation map, accurately identifying promoting and hindering behaviors. Through temporal and spatial trend analysis and root cause localization of various chronic diseases, it achieves precise attribution at the group level, enabling the development of differentiated intervention strategies for different regions and changing the methods of public education.

[0061] By utilizing the vital signs and behavioral data of authorized monitoring subjects, we can dynamically analyze individual trend characteristics and pinpoint root causes, enabling automatic identification of disease stages. This upgrades regular follow-up visits to continuous and intelligent individual monitoring, providing timely warnings and preventing the development of complications, thereby enhancing patient safety and self-management capabilities. Attached Figure Description

[0062] The present invention will be further described below with reference to the accompanying drawings and embodiments:

[0063] Figure 1 This is a system architecture diagram of a regional digitalized chronic disease collaborative management system shown in one embodiment of this application;

[0064] Figure 2 This is a flowchart illustrating a regional digitalized chronic disease collaborative management method in one embodiment of this application;

[0065] Figure 3 This is an example of a color gradient heatmap in one embodiment of this application;

[0066] Figure 4 This is a structural diagram of a regional digitalized chronic disease collaborative management system shown in one embodiment of this application. Detailed Implementation

[0067] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that, unless otherwise specified, the following embodiments and features described therein can be combined with each other.

[0068] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Therefore, the drawings only show the layers related to the present invention and are not drawn according to the actual number, shape and size ratio of the layers in the actual implementation. In the actual implementation, the form and number of each layer can be arbitrarily changed, and the layer layout may also be more complex.

[0069] Numerous details are explored in the following description to provide a more thorough explanation of embodiments of the invention; however, it will be apparent to those skilled in the art that embodiments of the invention may be practiced without these specific details.

[0070] Figure 1 This is a system architecture diagram of a regional digitalized chronic disease collaborative management system shown in one embodiment of this application. Figure 1 As shown, the system adopts a five-layer architecture, comprising a data access layer 110, an intelligent data processing layer 120, a unified FHIR (Fast Healthcare Interoperability Resources) resource layer 130, an intelligent collaborative service layer 140, and an application layer 150. The FHIR standard serves as the core data model connecting all layers. Specifically:

[0071] Data access layer 110;

[0072] The system's data sources are comprised of raw data collected from multiple sources, including business systems of medical institutions at all levels, public health platforms, wearable devices, community surveys, and mobile applications. This layer enables data access through interface calls, database extraction, message buses, and IoT gateways. It supports integration with core business systems such as HIS (Hospital Information System), LIS (Laboratory Information System), EMR (Electronic Medical Record), and PACS (Picture Archiving and Communication System). It also supports real-time vital sign data streams from wearable devices, resident behavioral characteristic data (exercise, diet, smoking and alcohol consumption, travel terrain, etc.), and individual behavioral data of authorized monitoring subjects. During data access, the system uses a unified data acquisition component for parsing and format conversion, and performs preliminary quality checks on the accessed data to ensure the reliability of subsequent standardized processing.

[0073] Intelligent data processing layer 120;

[0074] The system cleans, transforms, and standardizes the incoming raw medical data, behavioral feature data, and vital sign data. This layer establishes a unified data governance rule system for field standardization, coding standardization, and anomaly handling. During standardization, the system uses the FHIR resource model as the core data structure, converting the original business table structure into standard FHIR resources through data mapping rules. The system has a built-in FHIR mapping agent that converts the raw data into data objects conforming to the FHIR JSON / XML format according to the mapping rules. For example: Patient information → Patient Resource, Medical records → Encounter Resource, Test results → Observation Resource, Diagnostic information → Condition Resource. In addition, this layer adds a behavioral feature extraction and aggregation module: it performs spatiotemporal aggregation on resident behavioral feature data (including regional dietary habits, smoking and drinking habits, exercise levels, terrain and travel patterns, etc.) to form feature vectors organized by geographical units and behavioral types, and supports multi-period dynamic updates. Simultaneously, a behavior-chronic disease association graph is constructed: using spatial autoregressive models or geographically weighted regression, the association between behavioral characteristics and chronic disease prevalence is quantified, and the weights of promoting and hindering behaviors are output and stored in a graph structure. This process enables the transformation of heterogeneous data into a unified standard resource model and completes the graph-based accumulation of behavioral knowledge.

[0075] Unified FHIR resource layer 130;

[0076] The core of the system architecture is the unified management of standardized medical data resources and behavior-chronic disease association graphs. This layer achieves unified storage and management of various resources (including Patient, Encounter, Observation, Condition, Medication Request, etc.) through the construction of FHIR services, and supports resource version control, resource relationship management, and resource retrieval. Simultaneously, this layer extends the storage of behavior-chronic disease association graph nodes and edge data, which can be standardized according to FHIR extension mechanisms (such as extending resources or using evidence resources). The system establishes a resource index structure according to FHIR specifications, enabling the formation of a complete data association network between resources, for example: Patient → Encounter → Observation → Condition, and supports rapid retrieval of aggregated trend features and root cause results by geographical unit, time range, and disease type.

[0077] Intelligent Collaborative Service Layer 140;

[0078] Based on the FHIR standard, the system provides a unified data access interface and intelligent analysis services. It delivers external data services through a RESTful API (a web application interface designed based on the REST architectural style), strictly adhering to the FHIR API specification (resource reading, retrieval, creation, and updating). The interface layer implements unified authentication and authorization through an API gateway, supporting security mechanisms such as OAuth2 and Token. A new intelligent analysis service module has been added to this layer, including:

[0079] Trend analysis service: Based on multi-period physical examination and diagnostic data, calculate the time trend characteristics (linear slope, coefficient of variation, inflection point) and spatial trend characteristics (global Moran's I, local spatial autocorrelation, hotspot migration trajectory) of various chronic diseases, as well as the personal trend characteristics (physiological index coefficient of variation, changes in medication adherence, risk score, etc.) of authorized monitoring subjects.

[0080] Root Cause Analysis Service: Combining behavior-chronic disease association maps with resident behavioral characteristic data, attribution analysis is performed on time trend characteristics to output time trend root causes (dominant behaviors and their contribution); combining the behavioral characteristic data and association maps of authorized monitoring subjects, attribution analysis is performed on individual trend characteristics to output individual trend root causes (such as uncontrolled eating, missed medication, insufficient exercise, etc.).

[0081] Information management services: Based on trend characteristics and root cause results, these services provide diffusion blocking management (identifying the spatiotemporal clustering risk of chronic diseases, providing early warnings and blocking potential transmission / spread), group outreach management (generating differentiated educational content and distributing it by region), and individual monitoring management (dynamic tiered early warning and personalized intervention task scheduling). These services can be flexibly invoked by the application layer through standardized interfaces.

[0082] Application layer 150;

[0083] Based on standardized data services and intelligent collaborative services, we build medical application systems for different users, including:

[0084] Clinical data sharing: Supports doctors to access complete patient diagnosis and monitoring data across institutions.

[0085] Scientific data analysis: Provides disease-specific cohort construction and behavior-disease association exploration.

[0086] Healthcare quality supervision: Displaying regional chronic disease trend indicators and root cause analysis reports.

[0087] Regional medical collaboration: enabling two-way referrals, remote consultations, and prescription transfers.

[0088] Group promotion management application: Automatically generates promotional content for hot spots and pushes it through multiple channels.

[0089] Individual monitoring and management application: providing a closed loop of early warning and intervention tasks for family doctors and patients.

[0090] Applications in diffusion blocking management: spatiotemporal clustering detection, automatic early warning, and initiation of joint prevention and control.

[0091] Through the FHIR standard interface, various application systems can flexibly call platform data and intelligent services to achieve cross-system data sharing, business collaboration and precise chronic disease management, and comprehensively improve the level of convenience and benefits for the people.

[0092] Figure 2 This is a flowchart illustrating a regional digitalized chronic disease collaborative management method in one embodiment of this application, such as... Figure 2 This embodiment of a regional digitalized collaborative management method for chronic diseases may include steps S210 to S250:

[0093] S210, acquire declassified physical examination and diagnosis data from medical institutions at all levels in the target area, behavioral characteristic data of residents in the target area, and vital sign data and behavioral characteristic data of authorized monitoring subjects, wherein the physical examination and diagnosis data includes multiple functional physical fitness indicators and diagnostic conclusions;

[0094] This application requires the acquisition of three types of data: physical examination and diagnosis data (from medical institutions at all levels), resident behavioral characteristic data (from multi-source aggregation), and vital sign data and behavioral characteristic data of authorized monitoring subjects (from individual authorized collection). The acquisition scenarios are described below and specific examples are provided.

[0095] (1) Physical examination and diagnostic data;

[0096] Data is collected periodically through the HIS / EMR / LIS systems of various medical institutions within the city's medical group (general hospitals, community health service centers, township health centers, and physical examination centers) via interface calls, database extraction, or message bus methods. The data undergoes anonymization (removing direct identifiers such as names, ID numbers, and mobile phone numbers) and is uniformly linked to the resident master index (EMPI).

[0097] (2) Resident behavioral characteristics data;

[0098] Instead of directly collecting traceable individual behaviors, it aggregates multi-source indirect data to geographic units (community / street scale) to form group behavioral characteristics. Main sources:

[0099] Aggregated data from wearable devices (anonymized after user authorization): statistical values ​​such as steps, heart rate, and sleep.

[0100] Community Questionnaire / Health Records: Sampling survey of typical residents' dietary habits, exercise frequency, and smoking and drinking history.

[0101] Consumer data (in cooperation with supermarkets / e-commerce): regional sales of condiments (salt, soy sauce), cooking oil, sugar, and alcoholic beverages, to estimate per capita intake.

[0102] Geographic information: Digital elevation model (DEM), road network, distribution of parks and green spaces.

[0103] Mobile signaling / GPS trajectory (anonymized aggregation): Residents' travel distance, activity radius, and type of residence.

[0104] The data example is shown in the table below:

[0105] Table 1. Examples of Resident Behavioral Characteristics Data

[0106]

[0107] (3) Vital signs and behavioral characteristics data of the authorized monitoring subjects;

[0108] After obtaining informed consent from the patient or resident, data will be continuously collected through the following methods:

[0109] Wearable devices (smart bracelets / watches, continuous glucose monitors, smart blood pressure monitors): automatically upload heart rate, steps, sleep, blood pressure, blood oxygen, blood sugar, etc.

[0110] Mobile health app: Patients actively record their diet logs, medication check-ins, and symptom self-assessments.

[0111] Home medical devices: Bluetooth blood pressure monitor and blood glucose meter, which automatically sync after measurement.

[0112] Base station positioning / GPS (after authorization): activity trajectory, mode of travel, and duration of stay.

[0113] Data is uploaded in real-time or near real-time and linked to the patient's master index for individual trend analysis and root cause localization.

[0114] S220, construct a behavior-chronic disease association map of the target area based on the physical examination and diagnosis data and the residents' behavioral characteristics data, wherein the behavior-chronic disease association map represents the associated behaviors of various chronic diseases, and the associated behaviors include promoting behaviors and hindering behaviors;

[0115] In regional chronic disease management, the association between behavioral factors (diet, exercise, smoking, alcohol consumption, etc.) and chronic disease prevalence often exhibits nonlinearity, lag, and spatial heterogeneity. Traditional statistical methods (such as ordinary linear regression) ignore spatial dependencies between geographical units, leading to biased parameter estimates; moreover, simply calculating correlation coefficients cannot distinguish between "promoting behaviors" and "hindering behaviors," nor can it quantify the relative contributions of different behaviors. Therefore, a spatially explicit, interpretable, and causally oriented modeling method is needed to construct a behavior-chronic disease association map, providing a knowledge base for subsequent trend attribution and precise intervention. This method extracts significantly associated behaviors through a spatial autoregressive model and solidifies them in a graph structure, realizing the transformation from "data" to "knowledge."

[0116] The method for constructing the behavior-chronic disease association map in this application includes:

[0117] S221, the target area and the physical examination and diagnosis data of the target area are divided to obtain multiple geographical units and physical examination and diagnosis data of multiple geographical units;

[0118] Discretize the continuous space into basic analytical units. (Such as community neighborhood committees, grids, or 1km×1km grids). When dividing the area, it is necessary to ensure that each unit has a sufficient number of chronic disease cases (usually ≥ 50 cases) to stably estimate the prevalence rate, while preserving spatial heterogeneity.

[0119] S222, Calculate the incidence rate of chronic diseases in multiple geographic units. And extract feature vectors from resident behavior data of multiple geographical units. , where the feature vector This includes per capita salt intake, fat intake score, vegetable and fruit consumption compliance rate, smoking rate, per capita alcohol consumption, per capita daily physical activity, and the proportion of the population engaged in physical activity. For chronic disease indexing, For geographic unit indexing;

[0120] Specifically, it includes:

[0121] S222-1, extracting cases of various chronic diseases from physical examination and diagnosis data of geographical units;

[0122] S222-2, calculate the ratio of the number of cases of each chronic disease to the total population of the geographic unit, and obtain the incidence rate of chronic diseases for multiple geographic units. ;

[0123]

[0124] In the formula, Geographical unit Corresponding chronic diseases The number of cases, Geographical unit The total resident population.

[0125] S222-3 extracts per capita salt intake, fat intake score, vegetable and fruit intake compliance rate, smoking rate, per capita alcohol consumption, per capita daily exercise volume, and the proportion of people exercising from statistical data for multiple geographic units. It then standardizes these data for multiple geographic units to obtain feature vectors of resident behavioral characteristics data for each geographic unit. .

[0126] Average salt intake Fat intake score Vegetable and fruit compliance rate Smoking rate per capita alcohol consumption Average daily exercise volume and the proportion of the sports population The mathematical expressions are as follows:

[0127]

[0128] In the formula, Representing geographical units Total sales volume (kg) of salt equivalent of condiments (salt, soy sauce, sauces) in supermarkets / e-commerce in the current quarter.

[0129] This indicates the salt equivalent conversion factor (1 kg soy sauce ≈ 0.18 kg salt, direct salt coefficient = 1).

[0130] Representing geographical units The total population.

[0131] This indicates the number of days in the statistics.

[0132]

[0133]

[0134]

[0135]

[0136]

[0137] In the formula, This indicates the sales volume of various types of alcohol. This refers to the alcohol content (e.g., beer 4% → 0.04%). This represents the density of ethanol. To enable authorized wearable device users to combine, Indicates user In the The number of steps taken each day.

[0138] The above calculation process transforms abstract behaviors into calculable regional indicators, standardizes and eliminates differences in feature scales, and ensures comparability of regression coefficients; thus providing high-quality input for spatial autoregressive models.

[0139] S223, Construct a spatial autoregressive model, wherein the mathematical expression of the spatial autoregressive model is:

[0140]

[0141] In the formula, These are the spatial autoregressive coefficients. For geographic unit index, Indicates the first The first geographical unit The incidence of chronic diseases in China A collection of geographic unit indexes. This is the spatial weight matrix. For the regression coefficient vector, This is the random error term;

[0142] Ordinary linear regression requires independent samples, but geographical data exhibits spatial autocorrelation (considering the mutual influence and spread of disease rates among adjacent units). The spatial autoregression model in this application introduces a spatial lag term ρ into the equation. We can explicitly model the spillover effect of neighbor prevalence to estimate the net effect of behavioral characteristics without bias.

[0143] Specifically, the spatial weight matrix Represented as:

[0144]

[0145] The above autoregressive model eliminates spurious correlations caused by spatial dependence, accurately separates the independent contributions of behavioral factors from spatial spillover effects, and provides reliable coefficient estimates for the graph.

[0146] S224, Chronic disease incidence rates based on multiple geographic units and eigenvectors The spatial autoregressive model is fitted to obtain a regression coefficient vector and a P-value vector, wherein the regression coefficient vector includes regression coefficients of multiple behavioral features, and the P-value vector includes P-values ​​of multiple behavioral features.

[0147] This application employs maximum likelihood estimation (MLE) to fit the S-model. The fitting process is existing technology and will not be elaborated here. The fitting process obtains statistically rigorous coefficient significance, avoiding over-interpretation of random fluctuations and providing a mathematical basis for subsequent screening.

[0148] S225, Construct a behavior-chronic disease association map based on the regression coefficient vector and the p-value vector. The general idea is to map significant behavioral features with sufficiently large effect sizes to graph nodes, connect them to chronic disease nodes with directed edges, use edge weights to reflect the association strength, and use edge colors (or symbols) to represent facilitation / inhibition. Specifically, this includes:

[0149] S225-1, For each chronic disease, target behavioral features with P values ​​less than a preset association threshold and regression coefficients greater than a preset regression threshold are selected based on the regression coefficient vector and the P-value vector.

[0150] First, this application filters target behavioral features by setting association thresholds and regression thresholds, for example:

[0151] Preset association threshold: (This can be adjusted to a stricter value, such as 0.01, to control the false positive rate).

[0152] Preset regression threshold: (The minimum meaningful effect size of the absolute value of the standardized coefficient can be set according to the importance of the business.)

[0153] S225-2, Determine the association direction based on the regression coefficient of the target behavioral feature, wherein when the regression coefficient is positive, the target behavioral feature is a promoting behavior; when the regression coefficient is negative, the target behavioral feature is a hindering behavior.

[0154] S225-3, standardize the regression coefficients of the target behavioral characteristics for each chronic disease to obtain the association weights between the target behavioral characteristics and the chronic disease;

[0155] For the regression coefficients of each significant behavioral feature, Min-Max normalization to the [0,1] interval is used as the edge weights.

[0156] S225-4: Construct a node set based on multiple chronic diseases and multiple behavioral characteristics, and construct an edge set based on the association relationship, association weight and association direction between each chronic disease and the target behavioral characteristic;

[0157] S225-5, Construct a behavior-chronic disease association graph based on the node set and the edge set.

[0158] Node set ,in:

[0159] For all satisfied and behavioral characteristics;

[0160] For chronic disease labels (such as hypertension, diabetes, etc.) in the study.

[0161] Edge set For each chronic disease and each behavioral feature that meets the conditions Create directed edges and store the following attributes: (1) weight (2) Direction type (Positive - promotes, negative - hinders).

[0162] This application uses a graph database (Neo4j) to store node labels and edge attributes, thereby transforming complex regression results into an intuitive graph structure that supports quick queries for "what are the most relevant promoting behaviors for hypertension"; edge weights can be used to quantify intervention priorities; and the graph can be iteratively updated over time (refitted after adding new period data).

[0163] The aforementioned map construction method overcomes estimation bias caused by spatial dependence between geographical units through a spatial autoregressive model, achieves cross-indicator comparability by utilizing standardized behavioral characteristics, and ensures the statistical reliability and practical application value of the map through dual threshold screening (significance + effect size). The resulting map provides interpretable knowledge for subsequent spatiotemporal trend attribution, personalized intervention, group outreach, and diffusion blocking, and is the core knowledge layer of the regional digitalized chronic disease collaborative management system.

[0164] S230, based on multiple periods of physical examination and diagnosis data of the target area, performs trend analysis on various chronic diseases in the target area to obtain the temporal and spatial trend characteristics of various chronic diseases; and performs trend analysis on the vital sign data of authorized monitoring subjects to obtain individual trend characteristics;

[0165] Trend analysis serves as a bridge connecting historical data and future interventions. Regional spatiotemporal trends reveal the evolutionary patterns of chronic diseases across time and space, identifying high-incidence clusters and spread patterns; individual trends track the trajectory of disease changes in authorized patients. Together, they provide quantitative evidence for preventing the spread of diseases, community outreach, and individual monitoring.

[0166] The trend analysis in this application does not rely on real-time behavioral data, but directly utilizes multi-period physical examination diagnostic data and the vital sign data of authorized subjects, ensuring operability and robustness. The trend analysis process includes:

[0167] (1) Regional spatiotemporal trend characteristics

[0168] S231, extract the incidence rate of chronic diseases in the target region over multiple periods; and perform linear fitting based on the incidence rate of chronic diseases over multiple periods to obtain the overall slope of change; and construct time trend features based on the overall slope of change;

[0169] Extract the incidence rate of chronic diseases in a target region over multiple consecutive time windows t=1,2,…,T (usually quarterly or semi-annually). (Summary of the entire region). By fitting the time series using linear regression, the slope is obtained, reflecting whether the chronic disease is increasing (>0) or decreasing (<0) at the overall level, as well as the average rate of change.

[0170] The mathematical expression for the linear fitting model is:

[0171]

[0172] In the formula, The intercept is... The slope of the overall change. The time window number, These are independent and identically distributed errors.

[0173] After constructing the above model, the least squares method is used to ensemble the data, and then the significance is verified. If the slope is significant enough, the result is output, thereby quickly judging the overall trend of chronic disease deterioration or improvement in the region and providing a macro-based basis for health decision-making. The positive or negative value and magnitude of the slope can be directly used to compare the rate of change of different chronic diseases.

[0174] S232, extract the incidence rates of chronic diseases for multiple geographical units in the target area for each period, construct an incidence rate heatmap for each period based on the incidence rates of chronic diseases for multiple geographical units, and extract diffusion features from the incidence rate heatmaps for multiple periods to obtain spatial trend features, wherein the diffusion features include a flag indicating whether diffusion has occurred, diffusion speed, and diffusion direction.

[0175] For each time window t, based on the incidence rate of each unit Heatmaps are generated. By comparing heatmaps over consecutive time windows, the spatial diffusion characteristics of high-incidence areas of the disease are extracted: whether they are expanding, in which direction they are moving, and at what speed. This draws on the concept of "gravimeter shift" in the spatial epidemiology of infectious diseases.

[0176] In the heat map, Classify by quantiles (e.g., 5 levels: very high, high, medium, low, very low), or use a color gradient directly. Figure 3 Here is an example of a color gradient heatmap in one embodiment of this application, such as... Figure 3 As shown, geographic units are divided using map boundaries, forming... Figure 3The heatmap example shown. A threshold is also set. (For example, take the top 20 percentile of the incidence rate of all units), and mark the units with the top 20 percentile of the incidence rate as "high-incidence units".

[0177] Then, the centroid position of the high-incidence unit is extracted, and the rate of change of the centroid position in different time windows is calculated. If the rate is greater than the set value, it indicates that there is directional movement. At the same time, the movement speed and direction are calculated to obtain the spatial trend characteristics.

[0178] The heatmap in this application visually illustrates the shifts in "hotspots" of chronic diseases, such as whether high-incidence areas of hypertension are spreading from old urban areas to newer areas. These diffusion characteristics can be directly used for diffusion prevention management: if rapid spread is detected, intervention resources can be deployed in advance at the edges of high-incidence areas.

[0179] (2) Individual trend characteristics;

[0180] For authorized patients with chronic diseases, the specific chronic disease (k) is first identified. Then, based on the clinical knowledge base or behavior-chronic disease association map, the 1-3 most relevant physical indicators are determined (e.g., hypertension is associated with systolic blood pressure (SBP) and diastolic blood pressure (DBP); diabetes is associated with fasting blood glucose (FBG) and glycated hemoglobin (HbA1c); hyperlipidemia is associated with total cholesterol (TC) and triglycerides (TG)). The associations are also defined: positive association (elevated indicators reflect worsening condition, such as blood pressure) and negative association (elevated indicators reflect improvement, such as HDL, but these are usually not included in routine monitoring; more commonly, "the lower the target value, the better," meaning higher values ​​indicate greater risk, all of which are positive associations). In this protocol, "positive association" is defined as higher indicator values ​​representing worsening condition (e.g., blood pressure, blood glucose, blood lipids), while "negative association" is rarely used in routine monitoring and can be ignored. The specific process includes:

[0181] S233, obtain the chronic disease label of the authorized monitoring object, and obtain the associated vital signs and associations related to the chronic disease label, wherein the associations include positive associations and negative associations;

[0182] S234, extract the values ​​of related physical constitution at multiple monitoring time points from the vital signs data, and perform linear fitting on the values ​​of related physical constitution at multiple monitoring time points to obtain the slope of change of related physical constitution; and construct personal trend features based on the slope of change of related physical constitution.

[0183] The linear fitting model and its fitting process are consistent with those described above and will not be repeated here. This process compresses the chaotic time-series vital sign data into an intuitive "trend slope," facilitating doctors' rapid understanding of the patient's condition dynamics. The standardized relative rate of change allows for comparisons across vital signs, identifying key deteriorating factors. Combined with subsequent root cause localization, personalized intervention recommendations can be generated.

[0184] S240, based on the behavior-chronic disease association map and resident behavioral characteristic data, root cause localization is performed on the time trend characteristics to obtain the time trend root cause; and based on the behavioral characteristic data of the authorized monitoring subjects and the behavior-chronic disease association map, root cause localization is performed on the individual trend characteristics to obtain the individual trend root cause;

[0185] Root cause analysis is a crucial link between "trend analysis" and "precision intervention." Regional time-based root cause analysis reveals the main behavioral factors leading to an overall increase (or decrease) in the prevalence of a chronic disease; individual trend root cause analysis identifies the specific behavioral reasons for the worsening of an individual's condition. Both utilize pre-constructed behavior-chronic disease association maps as a knowledge base, combined with the slope of recent behavioral characteristic data changes, to automate and make attribution more interpretable. Specifically, the process of time-based root cause analysis includes:

[0186] S241, when the absolute value of the time trend feature of any target chronic disease is greater than or equal to the preset slope threshold, extract the resident behavior feature data sequence of multiple time periods before the current time period;

[0187] S242, Based on the behavior-chronic disease association map, extract the target behavior features of the target chronic disease, and perform linear fitting on the values ​​of the target behavior features in the resident behavior feature data sequence to obtain the slope of the target behavior features;

[0188] S243, target behavioral features whose absolute slope value is greater than or equal to a preset slope threshold and whose correlation matches are used as root causes of time trends.

[0189] When the overall incidence rate of chronic diseases in a region changes significantly (the slope exceeds a threshold), the system retrospectively examines resident behavioral characteristic data from multiple past time periods, focusing only on behavioral characteristics (i.e., target behavioral characteristics) that are significantly associated with the chronic disease in the association graph. A linear fit is performed on the change of each target behavioral characteristic over time. If it also shows significant changes (the slope exceeds a threshold) and the direction of change is consistent with the promoting / inhibiting effects defined in the association graph, then this behavior is determined to be one of the main causes of the current time trend.

[0190] When the incidence of chronic diseases just begins to rise and reach an inflection point, it is possible to pinpoint which behavioral factors are likely contributing to the increase, facilitating early intervention. Root causes directly point to specific behaviors (such as "rapidly increasing high-salt diets"), providing clear targets for community outreach and resource allocation. No manual review is required; the system generates root cause reports periodically without needing to be manually reviewed.

[0191] For individuals authorized for monitoring, when the slope of change in key vital signs exceeds a threshold (e.g., persistently rising blood pressure), the system utilizes the individual's behavioral characteristic data time series (including dietary records, exercise volume, medication adherence, etc.) and combines it with target behavioral characteristics related to the individual's chronic disease in the association map to fit the slope of change for each behavioral characteristic. If a behavioral characteristic itself shows a significant deterioration trend, and its direction of change is consistent with the promoting / inhibiting effects defined in the map, then this behavior is determined to be the individual root cause of the individual's disease exacerbation. The individual trend root cause localization process includes:

[0192] S244, when the absolute value of the personal trend feature is greater than or equal to a preset slope threshold, the target behavioral feature associated with the chronic disease label of the authorized monitoring object is determined based on the behavior-chronic disease association map;

[0193] For example: the slope of changes in key physical signs (such as blood pressure) for patient i suffering from chronic disease k. .when Root cause localization is triggered at that time. It can be set to a monthly relative change of 5% or an absolute change (such as a monthly increase of 3 mmHg in systolic blood pressure).

[0194] Extract behavioral characteristics of patient i over the past L time points (e.g., weekly summary data over the past 12 weeks). ,in To monitor the week number Behavioral characteristics may include:

[0195] Average daily steps (exercise volume).

[0196] Frequency of high-salt / high-fat diets (from a food diary);

[0197] Medication adherence (attendance rate);

[0198] Frequency of drinking, etc.

[0199] S245, perform linear fitting on the values ​​of the target behavioral features at multiple time points in the behavioral feature data to obtain the slope of the target behavioral features;

[0200] The principle of linear fitting has been described above and will not be repeated here.

[0201] S246, target behavioral features whose absolute slope is greater than or equal to a preset slope threshold and whose correlation matches are taken as personal trend root causes.

[0202] Unlike group-based root causes, individual root causes directly reflect the patient's specific behavioral deterioration points (e.g., a 20% increase in medication miss rate over four consecutive weeks), allowing for highly customized interventions. With the influx of new data, root causes are automatically updated weekly or monthly, enabling real-time tracking of disease progression. The system directly identifies key behavioral problems, improving management efficiency.

[0203] The root cause localization module utilizes a behavior-chronic disease association map as a static knowledge base, combining it with time series and trend slopes of regional or individual behavioral characteristics. By matching directional consistency, it automatically identifies the dominant behavioral factors leading to the exacerbation (or improvement) of chronic diseases. This method has the following advantages:

[0204] Data-driven approach: Quantitative attribution to avoid subjective bias.

[0205] Highly interpretable: Outputs specific behavior names and directions of change, making it easy to directly translate into intervention measures.

[0206] Layered adaptation: It supports both regional population (public health decision-making) and individual (clinical management) scales.

[0207] Together with the aforementioned trend analysis module, it forms a complete closed loop of "trend monitoring → root cause identification → information management", significantly improving the accuracy and proactivity of chronic disease management.

[0208] S250 manages information based on the temporal trend characteristics, spatial trend characteristics, temporal trend root causes, individual trend characteristics, and individual trend root causes of various chronic diseases. The information management includes diffusion blocking management, group publicity management, and individual monitoring management.

[0209] Information management is the decision-making and execution layer that translates trend analysis and root cause identification results into practical actions. Based on different triggering conditions, it automatically executes three types of management actions: group outreach (targeting rising regional incidence rates), individual feedback (targeting changes in individual disease conditions), and spread prevention (targeting the spatial spread of chronic diseases). The principles, key parameters, and beneficial effects of each scenario are explained below:

[0210] S251, When the time trend characteristics indicate an increase in the incidence of chronic diseases, conduct group advocacy based on the root causes of the time trend.

[0211] When the system detects that the overall incidence rate of a certain chronic disease k in a region shows a significant upward trend (i.e., the time trend slope), and When the time trend root cause localization results (S241-S243) are invoked, a list of dominant behavioral characteristics leading to the increase (e.g., "high-salt diet slope +0.32") is obtained. Based on these root cause behaviors, the system matches corresponding themes (e.g., "reduce salt and control blood pressure") from the promotional material library, and combines them with hotspot areas in the spatial trend characteristics to generate differentiated promotional content, which is then pushed to the target audience through multiple channels.

[0212] S252, when the individual trend characteristics indicate that the condition is improving or worsening, the monitoring results are sent to the authorized monitoring subjects;

[0213] For each authorized monitoring subject i, the system periodically (e.g., weekly) calculates their individual trend characteristics, including the slope of changes in key vital signs. Based on the sign and magnitude of this slope, the condition is determined as "improving," "worsening," or "stable" (in between). The determination results, the trend graph of key vital signs, and (if worsening) a list of individual trend root causes are sent to the patient and their family via mobile application or SMS, and simultaneously pushed to the responsible family physician.

[0214] S253, based on the spatial trend characteristics of various chronic diseases, performs chronic disease diffusion analysis, and when there is a diffusion trend of chronic diseases and the diffusion speed is greater than the set speed threshold, high-risk areas are delineated based on the diffusion direction and diffusion speed, and the high-risk areas are sent to the diffusion blocking planning module as the data basis for the diffusion blocking scheme.

[0215] By utilizing diffusion characteristics (diffusion markers, diffusion speed, and diffusion direction) in spatial trend features, when a high-incidence area of ​​chronic diseases is detected to be significantly expanding and the diffusion speed exceeds a set threshold, the system automatically delineates high-risk areas—i.e., geographical units that may be affected in the future. This area information is sent to the diffusion blocking planning module for developing blocking plans such as pre-deployment of resources, enhanced screening, and targeted publicity.

[0216] This invention discloses a regional digitalized collaborative management method for chronic diseases. It constructs a standardized, interoperable data foundation by uniformly aggregating physical examination and diagnostic data from medical institutions at all levels, resident behavioral characteristics, and the physical signs and behaviors of authorized monitoring subjects. This eliminates information barriers and supports closed-loop management throughout the entire process. Based on physical examination diagnoses and resident behavioral characteristics, a behavior-chronic disease correlation map is constructed to accurately identify promoting and hindering behaviors. Through temporal and spatial trend analysis and root cause localization of various chronic diseases, precise attribution at the group level is achieved, enabling the development of differentiated intervention strategies for different regions and changing the methods of health education.

[0217] By utilizing the vital signs and behavioral data of authorized monitoring subjects, we can dynamically analyze individual trend characteristics and pinpoint root causes, enabling automatic identification of disease stages. This upgrades regular follow-up visits to continuous and intelligent individual monitoring, providing timely warnings and preventing the development of complications, thereby enhancing patient safety and self-management capabilities.

[0218] like Figure 4 As shown, this application also provides a regional digitalized chronic disease collaborative management system, including:

[0219] The acquisition module is used to acquire declassified physical examination and diagnosis data from medical institutions at all levels in the target area, behavioral characteristic data of residents in the target area, and vital sign data and behavioral characteristic data of authorized monitoring subjects. The physical examination and diagnosis data includes various functional physical fitness indicators and diagnostic conclusions.

[0220] The association map construction module is used to construct a behavior-chronic disease association map of the target area based on the physical examination and diagnosis data and the residents' behavioral characteristic data. The behavior-chronic disease association map represents the association behaviors of various chronic diseases, including promoting behaviors and hindering behaviors.

[0221] The trend analysis module is used to perform trend analysis on various chronic diseases in the target area based on multiple periods of physical examination and diagnosis data, and to obtain the temporal and spatial trend characteristics of various chronic diseases; as well as to perform trend analysis on the vital sign data of authorized monitoring subjects to obtain individual trend characteristics;

[0222] The root cause localization module is used to perform root cause localization on the time trend features based on the behavior-chronic disease association map and resident behavioral characteristic data to obtain the time trend root cause; and to perform root cause localization on the individual trend features based on the behavioral characteristic data of the authorized monitoring subjects and the behavior-chronic disease association map to obtain the individual trend root cause.

[0223] The information management module is used for information management based on the temporal trend characteristics, spatial trend characteristics, temporal trend root causes, individual trend characteristics, and individual trend root causes of various chronic diseases. The information management includes diffusion blocking management, group publicity management, and individual monitoring management.

[0224] This invention discloses a regional digitalized chronic disease collaborative management system. By unifying and aggregating physical examination and diagnostic data from medical institutions at all levels, resident behavioral characteristics, and the physical signs and behaviors of authorized monitoring subjects, it constructs a standardized and interoperable data foundation, eliminating information barriers and supporting closed-loop management throughout the entire process. Based on physical examination diagnoses and resident behavioral characteristics, it constructs a behavior-chronic disease correlation map to accurately identify promoting and hindering behaviors. Through temporal and spatial trend analysis and root cause localization of various chronic diseases, it achieves precise attribution at the group level, enabling the development of differentiated intervention strategies for different regions and changing the way health education is conducted. Utilizing the physical signs and behaviors of authorized monitoring subjects, it dynamically analyzes individual trend characteristics and locates root causes, achieving automatic identification of disease stages. This upgrades from regular follow-up visits to continuous, intelligent individual monitoring, providing timely warnings and preventing the development of complications, thus improving patient safety and self-management capabilities.

[0225] This embodiment also provides an electronic terminal, including: a processor and a memory;

[0226] The memory is used to store computer programs, and the processor is used to execute the computer programs stored in the memory so that the terminal performs any of the methods in this embodiment.

[0227] As will be understood by those skilled in the art, the computer-readable storage medium described in this embodiment allows for the implementation of all or part of the steps in the above method embodiments by computer program-related hardware. The aforementioned computer program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0228] The electronic terminal provided in this embodiment includes a processor, a memory, a transceiver, and a communication interface. The memory and the communication interface are connected to the processor and the transceiver and complete communication between them. The memory is used to store computer programs, the communication interface is used to perform communication, and the processor and the transceiver are used to run the computer programs, so that the electronic terminal performs the steps of the above method.

[0229] In this embodiment, the memory may include random access memory (RAM) and may also include non-volatile memory, such as at least one disk storage device.

[0230] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0231] In the above embodiments, although the present application has been described in conjunction with specific embodiments thereof, many substitutions, modifications, and variations of these embodiments will be apparent to those skilled in the art based on the foregoing description. The embodiments of the present application are intended to cover all such substitutions, modifications, and variations falling within the broad scope of the appended claims.

[0232] The above embodiments are merely illustrative of the principles and effects of this application and are not intended to limit this application. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of this application. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in this application should still be covered by the claims of this application.

Claims

1. A regional digital and intelligent collaborative management method for chronic diseases, characterized in that, Including the following steps: The system acquires declassified physical examination and diagnostic data from medical institutions at all levels in the target area, behavioral characteristic data of residents in the target area, and vital sign and behavioral characteristic data of authorized monitoring subjects. The physical examination and diagnostic data includes various functional physical fitness indicators and diagnostic conclusions. Based on the physical examination and diagnosis data and the residents' behavioral characteristics data, a behavior-chronic disease association map of the target area is constructed. The behavior-chronic disease association map represents the associated behaviors of various chronic diseases, including promoting behaviors and hindering behaviors. Based on multiple cycles of physical examination and diagnostic data of the target area, trend analysis was performed on various chronic diseases in the target area to obtain the temporal and spatial trend characteristics of various chronic diseases; and trend analysis was performed on the vital sign data of authorized monitoring subjects to obtain individual trend characteristics. The root causes of the time-trend features are determined by analyzing the behavior-chronic disease association map and resident behavioral characteristic data; and the root causes of the individual trend features are determined by analyzing the behavioral characteristic data of the authorized monitoring subjects and the behavior-chronic disease association map. Information management is based on the temporal trend characteristics, spatial trend characteristics, temporal trend root causes, individual trend characteristics, and individual trend root causes of various chronic diseases. The information management includes diffusion blocking management, group publicity management, and individual monitoring management.

2. The regional digitalized collaborative management method for chronic diseases according to claim 1, characterized in that, Based on the physical examination and diagnostic data and the residents' behavioral characteristic data, a behavior-chronic disease association map of the target area is constructed, including: The target area and its physical examination and diagnosis data are divided into multiple geographical units and their physical examination and diagnosis data. Calculate the incidence rate of chronic diseases in multiple geographic units And extract feature vectors from resident behavior data of multiple geographical units. , where the feature vector This includes per capita salt intake, fat intake score, vegetable and fruit consumption compliance rate, smoking rate, per capita alcohol consumption, per capita daily physical activity, and the proportion of the population engaged in physical activity. For chronic disease indexing, For geographic unit indexing; Construct a spatial autoregressive model, wherein the mathematical expression of the spatial autoregressive model is: In the formula, These are the spatial autoregressive coefficients. For geographic unit index, Indicates the first The first geographical unit The incidence of chronic diseases in China A collection of geographic unit indexes. This is the spatial weight matrix. For the regression coefficient vector, This is the random error term; Chronic disease incidence rates based on multiple geographic units and eigenvectors The spatial autoregressive model is fitted to obtain a regression coefficient vector and a P-value vector, wherein the regression coefficient vector includes regression coefficients of multiple behavioral features, and the P-value vector includes P-values ​​of multiple behavioral features. A behavior-chronic disease association map is constructed based on the regression coefficient vector and the p-value vector.

3. The regional digitalized collaborative management method for chronic diseases according to claim 2, characterized in that, Calculate the incidence rate of chronic diseases in multiple geographic units And extract feature vectors from resident behavior data of multiple geographical units. ,include: Extracting cases of various chronic diseases from physical examination and diagnosis data of geographical units; The incidence rate of chronic diseases in multiple geographic units is obtained by calculating the ratio of the number of cases of each chronic disease to the total population of the geographic unit. ; The data includes per capita salt intake, fat intake score, vegetable and fruit intake compliance rate, smoking rate, per capita alcohol consumption, per capita daily exercise volume, and the proportion of people exercising, extracted from statistical data for multiple geographic units. These data are then standardized to obtain feature vectors representing the behavioral characteristics of residents in multiple geographic units. .

4. The regional digitalized collaborative management method for chronic diseases according to claim 2, characterized in that, Constructing a behavior-chronic disease association map based on the regression coefficient vector and the p-value vector, including: For each chronic disease, target behavioral features with P values ​​less than a preset association threshold and regression coefficients greater than a preset regression threshold are selected based on the regression coefficient vector and the P-value vector. The association direction is determined based on the regression coefficient of the target behavioral feature, wherein when the regression coefficient is positive, the target behavioral feature is a promoting behavior; when the regression coefficient is negative, the target behavioral feature is a hindering behavior. The regression coefficients of the target behavioral characteristics for each chronic disease are standardized to obtain the association weights between the target behavioral characteristics and the chronic disease. A node set is constructed based on multiple chronic diseases and multiple behavioral characteristics, and an edge set is constructed based on the association relationship, association weight, and association direction between each chronic disease and the target behavioral characteristic; Construct a behavior-chronic disease association graph based on the node set and the edge set.

5. The regional digitalized collaborative management method for chronic diseases according to claim 1, characterized in that, Based on multiple periods of physical examination and diagnostic data from the target region, trend analysis was performed on various chronic diseases in the target region to obtain the temporal and spatial trend characteristics of various chronic diseases, including: Extract the incidence rates of chronic diseases in the target region over multiple periods; perform linear fitting based on the incidence rates of chronic diseases over multiple periods to obtain the overall slope of change; and construct time trend features based on the overall slope of change. Chronic disease incidence rates of multiple geographic units in the target area for each period are extracted, and an incidence rate heatmap for each period is constructed based on the chronic disease incidence rates of multiple geographic units. Diffusion features are extracted from the incidence rate heatmaps of multiple periods to obtain spatial trend features, wherein the diffusion features include a marker indicating whether diffusion has occurred, diffusion speed, and diffusion direction.

6. The regional digital intelligent collaborative management method for chronic diseases according to claim 1, characterized in that, Trend analysis was performed on the vital sign data of authorized monitoring subjects to obtain individual trend characteristics, including: Obtain chronic disease tags of authorized monitoring subjects, and obtain associated vital signs and relationships related to the chronic disease tags, wherein the relationships include positive and negative associations; The relevant physical condition values ​​at multiple monitoring time points are extracted from the vital sign data, and linear fitting is performed on the relevant physical condition values ​​at multiple monitoring time points to obtain the slope of the change in the relevant physical condition; and personal trend features are constructed based on the slope of the change in the relevant physical condition.

7. The regional digitalized collaborative management method for chronic diseases according to claim 1, characterized in that, Based on the aforementioned behavior-chronic disease association map and resident behavioral characteristic data, the root causes of the time trend features are located to obtain the time trend root causes, including: When the absolute value of the time trend feature of any target chronic disease is greater than or equal to the preset slope threshold, extract the resident behavior feature data sequence of multiple time periods before the current time period; Based on the behavior-chronic disease association map, target behavioral features of the target chronic disease are extracted, and the values ​​of the target behavioral features in the resident behavioral feature data sequence are linearly fitted to obtain the slope of the target behavioral features. Target behavioral features whose absolute slope is greater than or equal to a preset slope threshold and whose correlation matches are used as root causes of time trends.

8. The regional digitalized collaborative management method for chronic diseases according to claim 1, characterized in that, Based on the behavioral characteristic data of the authorized monitoring subjects and the behavior-chronic disease association map, the root causes of the individual trend characteristics are located to obtain the individual trend root causes, including: When the absolute value of the individual trend feature is greater than or equal to a preset slope threshold, the target behavioral feature associated with the chronic disease label of the authorized monitoring object is determined based on the behavior-chronic disease association map. Linear fitting is performed on the values ​​of the target behavioral features at multiple time points in the behavioral feature data to obtain the slope of the target behavioral features; Target behavioral features whose absolute slope is greater than or equal to a preset slope threshold and whose correlation matches are taken as individual trend root causes.

9. A regional digitalized collaborative management method for chronic diseases according to claim 1, characterized in that, Information management is based on the temporal trend characteristics, spatial trend characteristics, temporal trend root causes, individual trend characteristics, and individual trend root causes of various chronic diseases, including: When time-trend characteristics indicate an increase in the incidence of chronic diseases, group outreach should be conducted based on the root causes of the time trend. When the individual trend characteristics indicate that the condition is improving or worsening, the monitoring results will be sent to the authorized monitoring subjects; Based on the spatial trend characteristics of various chronic diseases, chronic disease diffusion analysis is performed. When there is a diffusion trend of chronic diseases and the diffusion speed is greater than the set speed threshold, high-risk areas are delineated based on the diffusion direction and diffusion speed. The high-risk areas are then sent to the diffusion blocking planning module as the data basis for diffusion blocking schemes.

10. A regional digitalized chronic disease collaborative management system, characterized in that, include: The acquisition module is used to acquire declassified physical examination and diagnosis data from medical institutions at all levels in the target area, behavioral characteristic data of residents in the target area, and vital sign data and behavioral characteristic data of authorized monitoring subjects. The physical examination and diagnosis data includes various functional physical fitness indicators and diagnostic conclusions. The association map construction module is used to construct a behavior-chronic disease association map of the target area based on the physical examination and diagnosis data and the residents' behavioral characteristic data. The behavior-chronic disease association map represents the association behaviors of various chronic diseases, including promoting behaviors and hindering behaviors. The trend analysis module is used to perform trend analysis on various chronic diseases in the target area based on multiple periods of physical examination and diagnosis data, and to obtain the temporal and spatial trend characteristics of various chronic diseases; as well as to perform trend analysis on the vital sign data of authorized monitoring subjects to obtain individual trend characteristics; The root cause localization module is used to perform root cause localization on the time trend features based on the behavior-chronic disease association map and resident behavioral characteristic data to obtain the time trend root cause; and to perform root cause localization on the individual trend features based on the behavioral characteristic data of the authorized monitoring subjects and the behavior-chronic disease association map to obtain the individual trend root cause. The information management module is used for information management based on the temporal trend characteristics, spatial trend characteristics, temporal trend root causes, individual trend characteristics, and individual trend root causes of various chronic diseases. The information management includes diffusion blocking management, group publicity management, and individual monitoring management.

Images