A medical-physical integrated intelligent obesity stratified management system covering multiple population groups

Abstract

This invention discloses a multi-population integrated intelligent obesity stratification management system, belonging to the field of intelligent health management technology. It includes: a multi-population stratification and classification module, configured to automatically determine stratification based on the user's life cycle stage, body mass index, and self-reported health status, and generate a unique digital health record for each user containing stratification level identifiers and personalized physiological constraints; and a multi-source data acquisition module, configured to integrate user-reported data via a standardized API interface, wearable device data synchronized in real-time via Bluetooth BLE protocol, and institutional diagnostic data asynchronously acquired through a medical data exchange platform. This system achieves precise stratification of pregnant women with gestational diabetes, multi-source data fusion, personalized intervention, cross-sectoral resource integration, and a closed-loop medical-physical collaboration, effectively ensuring maternal and infant safety and improving intervention outcomes.

Description

Technical Field

[0001] This invention belongs to the field of intelligent health management technology, specifically involving a medical-physical collaborative intelligent obesity hierarchical management system covering multiple population groups. Background Technology

[0002] Currently, overweight and obesity have become a global public health challenge. According to relevant studies, the overweight and obesity rate in my country continues to rise, and the burden of obesity-related diseases has increased significantly. The incidence of diabetes, hypertension, hyperlipidemia, polycystic ovary syndrome, non-alcoholic fatty liver disease, asthma, and cardiovascular and cerebrovascular diseases (such as coronary heart disease and cerebral infarction) is showing an upward trend year by year, and the affected population is gradually becoming younger. Due to differences in physical constitution and the different stages of life, each person needs a different weight loss management plan.

[0003] Obesity management is not a single-dimensional weight control, but a systematic project involving multiple disciplines such as nutrition, sports medicine, endocrinology, and clinical medicine. It requires a collaborative process of "assessment-intervention-monitoring-adjustment" and the integration of various means such as nutritional guidance, exercise prescription, medical intervention, and behavioral management to achieve scientific and sustainable weight management results.

[0004] However, existing obesity management systems are poorly adapted and do not fully consider the physiological differences among different populations; there is insufficient collaboration between medical and physical health services, and there is a lack of effective connection between general population management and specialist treatment for patients with complications; service resources are scattered, and intervention programs are difficult to implement in specific scenarios.

[0005] Therefore, how to build an intelligent obesity management system that can cover the entire life cycle of the population, realize a closed loop of medical and physical collaboration, and integrate cross-border service resources has become an urgent technical problem to be solved in this field.

[0006] The information disclosed in this background section is intended only to enhance the understanding of the overall background of the invention and should not be construed as an admission or in any way implying that the information constitutes prior art known to those skilled in the art. Summary of the Invention

[0007] The purpose of this invention is to provide a medical-physical collaborative intelligent obesity stratification management system covering multiple population groups. It can achieve precise stratification of pregnant women with gestational diabetes, multi-source data fusion, personalized intervention, cross-border resource integration, and medical-physical collaborative closed loop, effectively ensuring maternal and infant safety and improving intervention effects.

[0008] To achieve the above objectives, a specific embodiment of the present invention provides the following technical solution: A medical-physical collaborative intelligent obesity stratified management system covering multiple population groups includes: The whole life cycle population stratification module is configured to automatically stratify based on the user's life cycle stage, body mass index and self-reported health status, and call the built-in multi-condition decision tree algorithm to generate a unique digital health record for each user, which includes stratification level identifiers and personalized physiological constraints. The standardized multi-source data fusion acquisition module is configured to integrate user-reported data, wearable device data synchronized in real time via Bluetooth BLE protocol, and institutional diagnosis and treatment data acquired asynchronously through a medical data exchange platform via a standardized API interface, and to perform preprocessing on all imported data to unify the format and align the timestamps. The tiered intervention module has a built-in scalable rule knowledge graph adapted to different user groups. The nodes of the rule knowledge graph represent user tiers, and the edges represent the mapping relationship with various intervention resources. Based on the user's digital profile, the module dynamically assembles and pushes a graph-matched personalized intervention service package through graph query and matching algorithms. The cross-border service resource scheduling module is configured to maintain a dynamic service resource graph, with nodes representing various cooperative service organizations and edges representing their service capabilities and suitable user tags. Based on the user's intervention plan, this module intelligently matches and assembles scenario-based combined intervention service packages from different service providers, and pushes them to the user and the corresponding organizations through a unified interface. The human-computer interaction collaboration module, as a middleware layer, adopts an event-driven architecture and asynchronous message queues. It defines a set of RESTful API service interaction protocols based on HTTPS to enable information exchange between users and various service organizations. This module is responsible for encapsulating the user's intervention plan execution data into standard event messages and pushing them to relevant organizations, as well as receiving service progress data from organizations to update user profiles.

[0009] In one or more embodiments of the present invention, in the full life cycle population stratification determination module, the life cycle stages include children, adolescents, pregnant women, postpartum women, and the elderly; the personalized physiological constraints are stored in their digital archives in the form of scalable key-value pairs.

[0010] In one or more embodiments of the present invention, the standardized multi-source data fusion acquisition module is further configured to: clean the continuous physiological data acquired via Bluetooth BLE protocol using an outlier removal algorithm within a sliding time window; and perform automatic desensitization processing based on sensitive fields on the diagnostic and treatment data obtained from medical institutions before storage.

[0011] In one or more embodiments of the present invention, the intervention resources mapped by the rule knowledge graph in the hierarchical intervention module include general health promotion programs, special program templates adapted to the physiological characteristics of specific life cycle stages, and medical linkage triggering rules for connecting with external specialist medical services.

[0012] In one or more embodiments of the present invention, the tiered intervention module embeds the medical linkage triggering rules in the special plan template for people with complications. When user data meets the preset clinical abnormal conditions, the system prioritizes pushing medical advice and generates a referral service request according to the user's command.

[0013] In one or more embodiments of the present invention, the service institution types maintained in the cross-border service resource scheduling module include at least one or more of the following: medical institution specialist outpatient clinics, maternal and child rehabilitation institutions, professional fitness institutions, nutrition meal preparation institutions, and community elderly care service centers.

[0014] In one or more embodiments of the present invention, an AI evaluation submodule is further included, which is configured to perform machine learning algorithm analysis on the fused data processed by the multi-source data acquisition module, and output weight trend prediction, personalized diet and exercise suggestions and disease risk warnings.

[0015] In one or more embodiments of the present invention, the AI ​​evaluation submodule is connected to a VR immersive compliance enhancement unit, which is used to match VR motion scenarios according to user stratification and to dynamically optimize the motion plan by collecting user action completion data in the VR scenario and feeding it back to the AI ​​evaluation submodule.

[0016] In one or more embodiments of the present invention, the multi-source data acquisition module uses image recognition technology to assist in recording the user's food intake data and uses it as one of the data sources for active reporting.

[0017] In one or more embodiments of the present invention, the system further includes a data security and privacy protection unit configured to de-identify data obtained from medical institutions upon access, implement hierarchical encryption on stored identity information and behavioral data, and use API signature verification and HTTPS encrypted channels for all cross-institutional data transmission.

[0018] Compared with the prior art, the medical-physical collaborative intelligent obesity stratified management system of the present invention covers multiple population groups. By introducing the life cycle stage as the stratification dimension and using a multi-condition decision tree algorithm for automated stratification determination, the present invention can provide intervention programs adapted to the physiological characteristics of different life cycle groups such as children, adolescents, pregnant women, postpartum women, and the elderly. By embedding medical linkage trigger rules into the hierarchical intervention module, when the physiological data of patients with complications meet preset clinical abnormality conditions, the system can prioritize pushing medical advice and automatically generate referral service requests based on user commands. Simultaneously, the human-machine collaboration module pushes user data to cooperating medical institutions, realizing medical-physical collaboration in "diagnosis and treatment-management". By constructing a dynamic service resource map and using a map matching algorithm, this invention can transform intervention strategies into executable external service combinations. This invention integrates three types of data sources—user-initiated reporting, wearable device real-time monitoring, and medical institution medical records—through a multi-source data acquisition module, and achieves bidirectional synchronization of intervention plan execution data and institutional service feedback data through the human-machine collaboration module, constructing a complete data closed loop from user profiling, plan generation, service execution to effect evaluation. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a diagram of the system core module architecture in one embodiment of the present invention; Figure 2 This is a business process diagram of a medical-physical collaborative intelligent obesity hierarchical management system covering multiple population groups, as described in one embodiment of the present invention. Detailed Implementation

[0021] To enable those skilled in the art to better understand the technical solutions in this disclosure, the technical solutions in the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this disclosure, and not all embodiments. Based on the embodiments in this disclosure, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this disclosure.

[0022] Example 1: As Figure 1 and Figure 2 As shown, the medical-physical collaborative intelligent obesity stratified management system provided by this invention, covering multiple population groups, adopts a modular and layered architecture design. It includes a full-lifecycle population stratification determination module, a standardized multi-source data fusion and acquisition module, a stratified intervention module, a cross-border service resource scheduling module, and a human-computer interaction and collaboration module. Data and instructions are transmitted between these modules through standardized interfaces, forming a complete intelligent management closed loop from user profiling to service delivery.

[0023] This system is deployed based on a microservice architecture, with each module capable of independent deployment and horizontal scaling. Inter-module communication utilizes a RESTful API over HTTPS. For cross-module calls requiring high reliability, a message queue is introduced to achieve asynchronous decoupling, ensuring high availability and scalability of the system.

[0024] Example 2: This example is a specific implementation of the full life-cycle population stratification module. The core function of this module is to achieve accurate user stratification, laying the foundation for subsequent personalized intervention. The stratification rules are stored in a structured knowledge base. The core stratification logic is based on calculating the Body Mass Index (BMI) based on the user's height and weight. Referring to the latest obesity-related industry standards, treatment guidelines, and technical specifications issued by the National Health Commission, the module categorizes different population groups by weight status and classifies them according to age and special physiological conditions, as detailed below: For average adults (18-59 years old), weight is calculated using BMI = weight (kg / height). 2 (m) 2 Automatically detects BMI ≥ 28 kg / m² 2 Obese, 24kg / m 2 ≤BMI<28kg / m 2 Overweight, 18.5 kg / m 2 ≤BMI<24kg / m 2 For those within the normal weight range or with a BMI < 18.5 kg / m² 2 Being underweight; The BMI criteria for children (3-11 years old) adopt the "Growth Standards for Children Under 7 Years Old" (WS / T 423-2022) and "Screening for Overweight and Obesity in School-Age Children and Adolescents" (WS / T 586-2018) issued by the National Health Commission, without using fixed adult thresholds: For children aged 3-5 years, the age-specific BMI standard deviation method is used, with the median BMI for the same age and sex as the benchmark. A BMI below 2 standard deviations is considered underweight, above 1 standard deviation but not more than 2 standard deviations is considered overweight, and ≥2 standard deviations is considered obese; For children aged 6-11 years, the age-specific and sex-specific BMI percentile method is used, with a BMI ≥85th percentile and <95th percentile for the same age and sex being overweight, and a BMI ≥95th percentile being obese. The BMI determination criteria for adolescents (12-18 years old) follow the WS / T 586-2018 standard, using age- and gender-specific BMI percentile methods to determine overweight and obesity, and the adult BMI determination criteria are adopted when they reach the age of 18. The BMI determination and weight management standards for pregnant women (during pregnancy) are based on pre-pregnancy BMI, classified according to the Chinese adult BMI standard, and the range of weight gain during pregnancy is modified and assessed according to the "Recommended Values ​​for Weight Gain in Pregnant Women" (WS / T 801-2022). Obesity is not directly determined by the real-time BMI during pregnancy. For postpartum women (within 6 months postpartum), the pre-pregnancy BMI is used as a reference, and the weight status is gradually returned to the adult BMI standard, combined with the postpartum physiological recovery. For older adults (≥60 years old), the Chinese adult BMI standard should be used in principle. In clinical assessment, muscle mass, nutritional status and activity level can be combined for comprehensive judgment, and the lower limit of low body weight should be appropriately relaxed.

[0025] Based on user age and specific physiological conditions, users are categorized into children (3-11 years old), adolescents (12-18 years old), pregnant women (during pregnancy), postpartum women (within 6 months postpartum), elderly people (≥60 years old), or ordinary adults (18-59 years old, with normal BMI, seeking to improve their physical and health). Ordinary adults are the default category, covering those not included in the specific stages.

[0026] Based on user self-reports or personal medical information accessed by the system through a medical data exchange platform, the system identifies whether the user has complications such as diabetes, hypertension, hyperlipidemia, polycystic ovary syndrome, non-alcoholic fatty liver disease, asthma, and cardiovascular and cerebrovascular diseases (such as coronary heart disease and cerebral infarction), and generates complication labels.

[0027] The stratification process is executed automatically using a multi-condition decision tree algorithm. When a user registers or updates their health information, the system performs stratification determination in real time and synchronously associates it with preset physiological constraints. These physiological constraints are stored in the user's digital health record in an extensible key-value pair format. For example: Automatically bind rules such as "daily calorie intake must meet the combined needs of basal metabolism and growth and development, and protein intake must not be less than 1.2g / kg body weight" for adolescents; For pregnant women, restrictions include "avoiding raw and cold foods, limiting caffeine intake, and keeping exercise heart rate below 140 beats per minute"; Rules such as "exercise intensity should be based on a Borg subjective fatigue score of 11-13, avoid explosive movements, and monitor blood pressure for early warning" are set for the elderly population; For individuals with complications, medical-related constraints such as "frequency of blood glucose monitoring, medication reminders, and abnormal value warning thresholds" are imposed.

[0028] The digital health profile generated by this module includes: basic user information, hierarchical level identifiers, a list of personalized physiological constraints, and historical hierarchical change records, which serve as the core input for subsequent modules.

[0029] Example 3: This example is a specific implementation of a standardized multi-source data fusion and acquisition module. This module is responsible for aggregating health data from different sources and in different formats, providing a comprehensive and real-time data foundation for the system. Data is aggregated through three types of standardized interfaces: (1) User-initiated reporting channel supports users to enter data such as diet records, subjective feelings, and exercise check-ins through WeChat mini program, mobile APP and PC web page. Among them, diet records adopt image recognition assisted technology. After the user takes a photo of the meal, the system automatically identifies the types of ingredients and estimates the intake, reducing the burden of manual entry for the user.

[0030] (2) The wearable device access channel is based on the Bluetooth BLE 5.0 protocol, and connects in real time to physiological indicators monitored by devices such as smart bracelets, smartwatches, body fat scales, and continuous glucose monitors, including heart rate, steps, exercise duration, calories burned, body fat percentage, blood pressure, blood glucose, and sleep. For the collected continuous physiological data, an outlier removal algorithm within a sliding time window is used for data cleaning, such as removing abnormal heart rate values ​​caused by loose device wearing based on the 3σ principle.

[0031] (3) The institutional data integration channel follows the HL7 FHIR medical information exchange standard and asynchronously connects to the hospital information system and electronic health record system through the medical data exchange platform to obtain structured medical data such as the user's personal medical records, test results, medication details, and diagnostic certificates.

[0032] This module performs a unified preprocessing workflow for all imported data: First, it standardizes the format, converting data from different sources into a unified data format defined internally by the system; then, it aligns timestamps, calibrating data from different devices and systems based on a unified timeline; finally, it cleans and completes the data, using interpolation or collaborative filtering based on similar user groups to fill in missing values. The preprocessed data is stored in a distributed time-series database, supporting efficient querying and analysis.

[0033] In addition, this module also extends to VR sports simulation devices to solve the problem of exercise execution in a fun way and lower the "behavioral threshold" of obesity management. By collecting data such as the user's limb range of motion, movement completion, exercise duration, and heart rate changes in the virtual scene in real time, it can be used for subsequent compliance assessment and exercise program optimization.

[0034] Example 4: This example demonstrates a specific implementation of the hierarchical intervention module. This module incorporates a population-adaptive, scalable rule knowledge graph. Its core design principle is to semantically map user hierarchical tags to intervention resources. The knowledge graph nodes include two types: User stratification nodes include: children - obese - without complications, pregnant women - overweight - gestational hypertension, elderly - overweight - diabetes, etc. Intervention resource nodes include general health promotion programs, specialized program templates adapted to the physiological characteristics of specific life cycle stages, and triggering rules for connecting with external specialist medical services.

[0035] The edges of the knowledge graph represent the mapping relationship between hierarchical nodes and intervention resource nodes, as well as the combination relationship between different intervention resources.

[0036] Once a user has completed their stratification, this module dynamically assembles a personalized intervention service package based on the stratification tags in their digital profile, using a graph query and matching algorithm. Specifically, it employs a breadth-first search algorithm based on graph traversal, starting from the user's stratification node and searching along the mapped edges for reachable intervention resource nodes. Multiple candidate paths are then sorted and filtered according to preset weight rules. Referring to relevant research, using graph neural networks for task graph matching can effectively improve the accuracy of resource matching.

[0037] Examples of specific intervention strategies for different population groups are as follows: For children, the intervention plan prioritizes ensuring their growth and development are not negatively impacted, with a focus on establishing a reminder mechanism for guardians. Regarding diet, the "Rainbow Diet" guides dietary structure to ensure nutritional balance; for exercise, fun and height-promoting activities such as rope skipping, basketball, and swimming are recommended; simultaneously, parent-child interactive tasks are pushed through a guardian's app to increase family involvement.

[0038] For adolescents, the strategy ensures a daily protein intake of at least 1.2g / kg of body weight to meet their growth and development needs. Exercise programs should incorporate high-intensity interval training and resistance training, focusing on activities that promote bone development and body shaping. For those who struggle to maintain exercise routines, VR basketball games and VR music rhythm exercises are recommended. These activities, through visual and auditory immersion, reduce fatigue, enhance enjoyment, and gradually improve exercise capacity and adherence.

[0039] For pregnant and postpartum women, the dietary strategy automatically identifies and excludes foods that may affect fetal development or breastfeeding. Exercise recommendations focus on low-intensity walking, prenatal yoga, and pelvic floor muscle training, with a set upper limit threshold for exercise intensity. An automatic warning is issued when the heart rate exceeds 140 beats per minute. For the high-risk group of "pregnant women - overweight - gestational hypertension," the intervention plan must be strictly under the guidance of a doctor, adhering to three principles: "prioritizing maternal and infant safety, controlling weight and blood pressure simultaneously, and minimizing stimulation and burden." This involves avoiding risky foods in the diet, limiting the safe intensity of exercise, and linking real-time monitoring to ensure that the intervention meets weight management needs while protecting fetal development and the health of the pregnant woman.

[0040] For the elderly population, a special blood pressure monitoring and early warning mechanism is implemented to monitor blood pressure changes in real time during exercise. Exercise programs are limited to joint-friendly activities such as Tai Chi, Baduanjin, and water walking to avoid the risk of falls and joint injuries. The program also incorporates fall prevention education and balance training. Utilizing wearable devices (such as smart bracelets with blood pressure monitoring), key indicators such as heart rate and blood pressure are continuously collected daily. Through algorithmic analysis by a tiered intervention module, potential risks such as sudden increases or decreases in blood pressure and abnormal heart rate fluctuations (e.g., falls, sudden cardiovascular or cerebrovascular events) are predicted, and early warning information is pushed to family members and pre-set doctors.

[0041] For individuals with complications, the rule base incorporates clinical early warning logic. For example, when a user's BMI is ≥30 and they are already categorized as having type 2 diabetes, the strategy will prioritize recommending a referral to an endocrinologist, along with a suitable low glycemic index diet plan and low-intensity activity guidance for diabetic patients. If a user's blood glucose data exceeds a preset threshold three times consecutively, such as fasting blood glucose ≥7.0 mmol / L or 2-hour postprandial blood glucose ≥11.1 mmol / L, the system will automatically trigger a medical linkage rule and generate a medical reminder.

[0042] This module also integrates an AI assessment submodule, employing a random forest algorithm to model and analyze the fused multi-source data. Input features include: demographic characteristics, BMI trends, dietary intake data, exercise execution data, physiological indicators collected by wearable devices, and past medical history. Output results include: a 4-week weight change trend prediction, personalized dietary adjustment suggestions, exercise program optimization suggestions, and complication risk warnings. The AI ​​assessment results, as a supplement to the rule-based knowledge graph output, together constitute a complete personalized intervention strategy.

[0043] Example 5: This example is a specific implementation of the cross-boundary service resource scheduling module. This module is responsible for transforming the intervention strategies generated by the layered intervention module into executable external service combinations. Its core is a dynamic service resource registration and matching mechanism.

[0044] The system maintains a dynamically updated service resource graph. Each node in the graph represents a cooperating service provider, and node attributes include: provider ID, provider type, and service capability tag. Edges in the graph represent the collaborative relationships between providers and the compatibility of service combinations. Provider types include medical institution specialist clinics, maternal and child health rehabilitation centers, professional fitness institutions, nutrition catering companies, and community elderly care service centers, etc.

[0045] Upon receiving the intervention strategy generated by the hierarchical intervention module, this module decomposes the strategy into multiple service requirement vectors. For example, for postpartum obese female users, the intervention strategy includes two measures: core muscle training and lactation nutrition guidance. The system converts these two measures into service requirement vectors and performs matching queries in the service resource graph.

[0046] The matching process employs a graph similarity-based retrieval algorithm. First, the demand vector is transformed into a query graph pattern. Then, service nodes with matching tags and currently available are searched in the service resource graph. Finally, candidate nodes are ranked based on factors such as geographical location, user preferences, and service evaluation. For scenarios requiring combined services, such as combining offline courses with online consultations, the system further searches the graph for existing combinations of nodes with established collaborative relationships, or generates the optimal combination scheme through reinforcement learning algorithms.

[0047] After matching is complete, the system packages the selected service resources into scenario-based, combined intervention service packages. Taking "postpartum obese women" as an example, the service package includes: "postpartum rehabilitation yoga courses" provided by partner gym A and "online breastfeeding nutrition consultation services" provided by partner nutrition institution B. The service package also generates a schedule guide for users and a to-do list for service providers.

[0048] Drawing on relevant practices, integrating multidisciplinary and multi-institutional service resources through an information platform can create a closed-loop health service system covering the entire process of "screening-assessment-intervention-feedback." Clinical data shows that adopting a multidisciplinary collaborative management model can improve patient treatment adherence to 91%.

[0049] Example 6: This example is a specific implementation of the human-computer interaction collaboration module. As the collaboration hub between the system and the external service ecosystem, this module adopts an event-driven architecture and asynchronous message queues to achieve closed-loop execution of the service package throughout the entire process.

[0050] This module is built on a message queue middleware and uses a publish-subscribe pattern to achieve asynchronous communication between components within the module. The main components include: Event producers include a layered intervention module that generates new solution events, a cross-border service resource scheduling module that generates service package distribution events, and a user terminal that generates data reporting events, etc. The message queue uses a message middleware that supports high concurrency and persistent storage, and sets up independent topics for different types of events, such as "intervention plan execution event", "service progress feedback event", and "early warning event". Event consumers include institutional adapters, which convert system events into formats that can be accepted by various institutions; user notification services, which push notifications to users; and profile update services, which update user profiles based on feedback data. It also supports the generation of dynamic and visual charts for the entire weight loss cycle of users, including line trend charts and bar charts of core indicators such as weight and body fat percentage, which intuitively present the trajectory of indicator changes and weight loss results.

[0051] Taking the management of patients with complications as an example, the collaborative process of this module is explained: When a user's wearable device continuously collects abnormal physiological data, such as fasting blood glucose ≥7.0mmol / L for 3 consecutive days, the multi-source data acquisition module will import the data into the system, and the hierarchical intervention module will identify the conditions that meet the preset clinical abnormality and trigger the medical linkage rule. This rule generates a "referral suggestion event" and publishes it to the "medical collaboration" topic.

[0052] After the event consumer of the human-computer interaction collaboration module listens for this event, it performs the following operations: Through the institution adapter, referral suggestions and users' recent health data are encapsulated in the format required by the target medical institution and pushed to the institution's receiving interface via the HTTPS protocol; At the same time, through the user notification service, medical reminders are pushed to the user's mobile app, showing recommended doctors, available appointment times, and a list of documents to be brought. The event is recorded in the persistent storage of the message queue to ensure that the event is not lost even if the organization interface is temporarily unavailable, and can be automatically retried after recovery.

[0053] After a doctor completes their diagnosis and treatment of a user, they provide a treatment plan, including but not limited to exercise, nutrition, and medication. The user, following the doctor's instructions, provides feedback on the latest data through a dedicated interface. Upon receiving this feedback data, this module triggers a file update service to update the user's digital health record and notifies the tiered intervention module to reassess the intervention strategy, forming a complete closed loop of "monitoring-early warning-referral-feedback-optimization".

[0054] This module provides a unified dual-platform interactive interface; users can view service execution progress in real time, receive institutional notifications, and complete service evaluations through WeChat mini-programs, apps, and other terminals. Service providers such as doctors, health managers, and fitness coaches can access user intervention progress through a dedicated server interface (web or institutional app) and dynamically adjust service content or treatment recommendations based on the actual situation.

[0055] This system strictly adheres to the "Personal Information Protection Law of the People's Republic of China" and relevant industry security management standards for health and medical data. It establishes a comprehensive, end-to-end data security governance system that aligns with regulations concerning personal information protection and medical data management. It employs a tiered encryption storage strategy, independently encrypting user identity information and classifying and encrypting behavioral and physiological health data to strengthen data storage security. A complete user authorization management interface is provided, allowing users to independently control the scope of access and usage of their personal data, achieving autonomous data access control. All data access logs are recorded synchronously throughout the process, preserving complete access traces to meet the compliance requirements of subsequent security audits and source tracing. For cross-institutional data transmission scenarios, a dedicated collaborative service bus is used for data interaction, with API signature verification and HTTPS encrypted transmission channels enabled throughout the process to strictly ensure the confidentiality and integrity of data transmission. For sensitive diagnostic and treatment data obtained from external medical institutions, sensitive fields are automatically anonymized before storage, shielding core privacy information to further mitigate the risks of data leakage and misuse, comprehensively meeting data security compliance management requirements.

[0056] It will be apparent to those skilled in the art that this disclosure is not limited to the details of the exemplary embodiments described above, and that this disclosure can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of this disclosure is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within this disclosure. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0057] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A medical-physical collaborative intelligent obesity stratified management system covering multiple population groups, characterized in that: include: The whole life cycle population stratification module is configured to automatically stratify based on the user's life cycle stage, body mass index and self-reported health status, and call the built-in multi-condition decision tree algorithm to generate a unique digital health record for each user, which includes stratification level identifiers and personalized physiological constraints. The standardized multi-source data fusion acquisition module is configured to integrate user-reported data, wearable device data synchronized in real time via Bluetooth BLE protocol, and institutional diagnosis and treatment data acquired asynchronously through a medical data exchange platform via a standardized API interface, and to perform preprocessing on all imported data to unify the format and align the timestamps. The tiered intervention module has a built-in scalable rule knowledge graph adapted to different user groups. The nodes of the rule knowledge graph represent user tiers, and the edges represent the mapping relationship with various intervention resources. Based on the user's digital profile, the module dynamically assembles and pushes a graph-matched personalized intervention service package through graph query and matching algorithms. The cross-border service resource scheduling module is configured to maintain a dynamic service resource graph, with nodes representing various cooperative service organizations and edges representing their service capabilities and suitable user tags. Based on the user's intervention plan, this module intelligently matches and assembles scenario-based combined intervention service packages from different service providers, and pushes them to the user and the corresponding organizations through a unified interface. The human-computer interaction collaboration module, as a middleware layer, adopts an event-driven architecture and asynchronous message queues. It defines a set of RESTful API service interaction protocols based on HTTPS to enable information exchange between users and various service organizations. This module is responsible for encapsulating the user's intervention plan execution data into standard event messages and pushing them to relevant organizations, as well as receiving service progress data from organizations to update user profiles.

2. The system according to claim 1, characterized in that, In the full life cycle population stratification determination module, the life cycle stages include children, adolescents, pregnant women, postpartum women, and the elderly; the personalized physiological constraints are stored in their digital archives in the form of scalable key-value pairs.

3. The system according to claim 1, characterized in that, The standardized multi-source data fusion acquisition module is further configured to: clean the continuous physiological data acquired via Bluetooth BLE protocol using an outlier removal algorithm within a sliding time window; and perform automatic desensitization processing based on sensitive fields on the diagnostic and treatment data obtained from medical institutions before storage.

4. The system according to claim 1, characterized in that, The intervention resources mapped by the rule knowledge graph in the hierarchical intervention module include general health promotion programs, special program templates adapted to the physiological characteristics of specific life cycle stages, and medical linkage trigger rules for connecting with external specialist medical services.

5. The system according to claim 4, characterized in that, The tiered intervention module includes a medical linkage triggering rule embedded in the special treatment plan template for people with complications. When user data meets the preset clinical abnormality conditions, the system prioritizes pushing medical advice and generates a referral service request based on the user's command.

6. The system according to claim 1, characterized in that, The service institution types maintained in the cross-border service resource scheduling module include at least one or more of the following: medical institution specialist outpatient clinics, maternal and child rehabilitation institutions, professional fitness institutions, nutrition meal preparation institutions, and community elderly care service centers.

7. The system according to claim 1, characterized in that, It also includes an AI assessment submodule, which is configured to perform machine learning algorithm analysis on the fused data processed by the multi-source data acquisition module, and output weight trend prediction, personalized diet and exercise suggestions and disease risk warnings.

8. The system according to claim 7, characterized in that, The AI ​​evaluation submodule is connected to a VR immersive compliance enhancement unit. The VR immersive compliance enhancement unit is used to match VR exercise scenarios according to user stratification and to collect user action completion data in VR scenarios and feed it back to the AI ​​evaluation submodule to dynamically optimize exercise plans.

9. The system according to claim 1, characterized in that, The multi-source data acquisition module uses image recognition technology to help record the user's food intake data and uses it as one of the data sources for proactive reporting.

10. The system according to claim 1, characterized in that, The system also includes a data security and privacy protection unit, which is configured to de-identify data obtained from medical institutions upon access, implement hierarchical encryption for stored identity information and behavioral data, and use API signature verification and HTTPS encrypted channels for all cross-institutional data transmission.

Images