User health data management method and system based on cloud platform

By dividing outpatient types and correlation analysis of user health data, building a search library, identifying health risk factors and conducting trend assessment, the problem of inaccurate correlation and trend assessment of health status search in traditional methods is solved, and personalized health management and Real-time data push improves the efficiency and accuracy of health management.

CN118969287AInactive Publication Date: 2024-11-15GUANGDONG MAIKE MEDICAL TECH CO LTD

Patent Information

Application Number
CN202411435066.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-15
Publication Date
2024-11-15
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

Traditional cloud-based user health data management methods cannot quickly retrieve health status and inaccurately evaluate health trends, resulting in the inability to provide personalized health management solutions.

Method used

By obtaining the user's historical physical examination report data, conducting outpatient type classification and correlation overlap analysis, building a content-related intermediate layer shard search library, identifying health risk factors, and using these factors to evaluate health trends and behavioral corrections to achieve real-time data push .

Benefits of technology

It improves the utilization rate and management efficiency of health data, can quickly correlate health status, improves the accuracy of health trend assessment, provides personalized health management suggestions, and enhances the convenience and effectiveness of users' health management.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118969287A_ABST
    Figure CN118969287A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of health data management, in particular to a user health data management method and system based on a cloud platform. The method comprises the following steps: performing outpatient service category division on historical physical examination report data to which contents belong to obtain outpatient service category data to which physical examination contents belong; according to the outpatient service type data to which the physical examination content belongs, performing content association retrieval library construction on the historical physical examination report data to obtain a content association intermediate layer fragment retrieval library; health trend evaluation is carried out according to the content association intermediate layer fragment retrieval library, and health trend evaluation data is obtained; performing user health behavior correction based on the health trend evaluation data to obtain health behavior correction data; and pushing the health behavior correction data based on the cloud platform to execute user health data management. The health data management technology is improved, so that the health data management technology is more perfect.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of health data management, and in particular to a user health data management method and system based on a cloud platform. Background Art

[0002] With the popularity of smart wearable devices, mobile health applications, and home medical devices, users are able to generate a large amount of health data, including heart rate, blood pressure, exercise steps, sleep quality, etc. However, these data are often scattered in different devices and applications, lacking a unified management method. The previous health data management model faced problems such as data silos, insufficient privacy protection, and difficulty in data sharing, making it difficult to provide users with comprehensive and personalized health guidance. The cloud platform-based health data management method integrates multi-source data through the cloud, using distributed databases, efficient data transmission protocols, and data encryption technology to ensure the security, integrity, and real-time nature of the data. Through the powerful computing power of the cloud platform, these data can be analyzed in real time to generate personalized health assessment reports and provide users with dynamic health risk predictions and intervention recommendations. In addition, the method supports multi-party data sharing and cross-platform collaboration, allowing users to authorize medical institutions, fitness coaches, or family members to access their health data to achieve precise health management. The scalability and elastic architecture of the cloud platform can adapt to the growing amount of data and computing needs, while continuously optimizing the accuracy and effectiveness of health recommendations through artificial intelligence algorithms and machine learning models. This management method not only improves the utilization rate of health data, but also promotes the development of personalized medicine and preventive medicine, which is of great significance to improving public health and promoting digital health transformation. However, the traditional user health data management method based on cloud platform has the problems of inability to quickly retrieve and associate health status and inaccurate assessment of health trends. Summary of the invention

[0003] Based on this, it is necessary to provide a user health data management method and system based on a cloud platform to solve at least one of the above technical problems.

[0004] To achieve the above purpose, a user health data management method based on a cloud platform is provided, the method comprising the following steps: Step S1: Obtain historical physical examination report data uploaded by the user; classify the historical physical examination report data into the type of clinic to which the content belongs, and obtain the type of clinic to which the physical examination content belongs; Step S2: performing overlapping analysis of the clinics to which the contents of the physical examination belong according to the clinic category data of the physical examination contents on the historical physical examination report data to obtain overlapping data of the clinics to which the contents belong; constructing a content-related retrieval library for the overlapping data of the clinics to which the contents belong to obtain a content-related intermediate layer fragmented retrieval library; Step S3: Identify health risk factors based on the content-related intermediate layer shard retrieval library to obtain health risk factor data; perform health trend assessment based on the health risk factor data to obtain health trend assessment data; Step S4: Correct the user's health behavior based on the health trend assessment data to obtain health behavior correction data; push the health behavior correction data based on the cloud platform to perform user health data management.

[0005] The present invention can systematically divide the physical examination content into different outpatient categories by classifying the outpatient types of the historical physical examination report data uploaded by the user. The complex physical examination data can be structured so that the data of each physical examination report is easier to manage and analyze. Secondly, this classification not only improves the efficiency of data processing, but also lays a solid foundation for subsequent analysis, such as performing more accurate health risk assessment according to the outpatient type. By performing association overlap analysis on the outpatient type data to which the physical examination content belongs, the connection between different outpatient categories is further excavated. The key to this step is that it can reveal the content overlap or association between different outpatient clinics, thereby discovering potential health problems. For example, the examinations of the same patient in different outpatient clinics show some repeated or related health problems, and this analysis can provide a more comprehensive understanding of the patient's health status. Subsequently, these overlapping analysis results will be used to construct a content association retrieval library, which not only integrates the associated data between outpatient clinics, but also subdivides the retrieval library into middle-layer fragments, effectively improving the speed and accuracy of data retrieval. This fragmented retrieval library can quickly locate relevant information in large-scale data, provide support for medical decision-making, and also make further personalized health analysis and prediction possible. In general, this technical solution significantly improves the utilization value of physical examination report data and optimizes the health management process through systematic data processing and association analysis. By using the content-related intermediate layer fragmentation retrieval library to identify health risk factors, potential health risk factor data can be accurately extracted. The core advantage of this process is that it can identify key factors related to health risks from complex physical examination data, such as abnormal changes in certain indicators or specific health signs. This accurate identification can not only help doctors or health managers better understand the health status of patients, but also provide basic data for further health trend assessment. Subsequently, health trend assessment based on these health risk factor data can provide insights into the dynamic changes in the patient's health status, such as the increasing trend of risk factors or the improvement of health indicators. The beneficial effect of this trend assessment is that it can help detect potential health problems early and provide a scientific basis for the formulation of personalized health management plans. Personalized health intervention measures are formulated based on health trend assessment data. For example, based on the assessment results, the system can recommend users to adjust their eating habits, increase exercise, or conduct regular health checks. Health behavior correction data can effectively reflect the implementation of these interventions and their actual impact on user health. Furthermore, by pushing health behavior correction data based on the cloud platform, real-time health management and feedback can be achieved. The application of this cloud platform not only makes the health management process more efficient and convenient, but also ensures that users can get timely advice and reminders on health management through automated data push functions, thereby promoting continuous health behavior improvement.In general, this technical solution provides users with a comprehensive and personalized health management solution through accurate health risk identification, trend assessment and behavior correction, combined with advanced cloud platform technology, greatly improving the effectiveness and convenience of health management. Therefore, the present invention is an improvement on a traditional user health data management method based on a cloud platform, solving the problem that a traditional user health data management method based on a cloud platform cannot quickly retrieve and associate health status, and the problem of inaccurate health trend assessment. It can quickly retrieve and associate health status, and improve the accuracy of health trend assessment.

[0006] Preferably, step S1 comprises the following steps: Step S11: Obtain historical physical examination report data uploaded by the user; Step S12: Clean the historical physical examination report data to obtain cleaned data of the historical physical examination report; Step S13: labeling the cleaned data of the historical physical examination report to obtain labeled data of the historical physical examination report; Step S14: classify the labeled data of historical physical examination reports according to the type of clinic to which the contents belong, and obtain the data on the type of clinic to which the physical examination contents belong.

[0007] In step S11 of the present invention, the historical physical examination report data uploaded by the user is first obtained, which lays the foundation for the entire data processing process. In step S12, the data is cleaned. The key to this step is to remove noise and inconsistency in the data to ensure the quality and accuracy of the data. The cleaned data is neater and helps with subsequent analysis and processing. Then, in step S13, the historical physical examination report cleaning data is labeled, the data is classified into different categories, and each piece of data is given a specific label. The labeling process not only makes the data more structured, but also enables more efficient information retrieval and management in subsequent analysis. Finally, step S14 divides the outpatient category to which the content belongs according to the labeled data, and classifies the physical examination data by outpatient category. This division helps to systematically manage physical examination data, so that data of different outpatient categories can be independently analyzed and integrated, and clear outpatient category data is provided. This step-by-step processing method not only improves the processing accuracy and effectiveness of the data, but also provides a more reliable basis for subsequent analysis and decision-making, thereby promoting the comprehensive optimization of health data management and analysis.

[0008] Preferably, step S2 comprises the following steps: Step S21: performing overlapping analysis of clinic-related contents on the historical physical examination report data according to the clinic-related types of the physical examination contents, and obtaining overlapping clinic-related contents; Step S22: evaluating the correlation complexity of the overlapping data of the clinics to which the content belongs, and obtaining the correlation complexity data of the clinics to which the content belongs; Step S23: constructing a content-related search library for the overlapping data of the clinics to which the content belongs according to the complexity data of the clinics to which the content belongs, and obtaining a content-related search library; Step S24: vertically slice the content-related middle layer of the content-related search library to obtain a content-related middle layer sliced ​​search library.

[0009] In step S21 of the present invention, the correlation overlap analysis is first performed on the outpatient categories to which the physical examination content belongs. The core of this step is to reveal the content overlap between different outpatient categories. In this way, similar or repeated health information can be found, thereby effectively integrating related data. Subsequently, step S22 analyzes the complexity of the relationship between different outpatient categories by evaluating the correlation complexity of these overlapping data. This complexity evaluation can help identify the complex correlation structure in the data, thereby providing a basis for building a more accurate retrieval system. Step S23 constructs a content-related retrieval library based on the complexity data. This process organizes complex overlapping data into a structured retrieval library, making information search and analysis more efficient. Finally, step S24 performs middle-layer vertical sharding on the content-related retrieval library to form a content-related middle-layer sharding retrieval library. This sharding technology helps to split large data sets into small fragments that are easier to manage and query, thereby improving the speed and accuracy of data retrieval. Overall, these steps improve the management capabilities of health data through systematic data analysis and optimization, so that key information can be obtained faster and more accurately when processing complex outpatient data, thereby supporting more effective health management decisions.

[0010] Preferably, step S23 includes the following steps: Step S231: performing overlapping area aggregation analysis on the overlapping data associated with the clinics to which the content belongs, and obtaining content overlapping area aggregation data; Step S232: constructing a content structure hierarchical network diagram according to the content overlap area aggregation data and the clinic association complexity data to which the content belongs, to obtain a content structure hierarchical network diagram; Step S233: performing factor correlation derivation analysis on the content structure hierarchical network diagram to obtain content factor correlation derivation data; Step S234: designing a hierarchical bubble priority index for the content structure hierarchical network diagram according to the content factor association derivation data to obtain a content hierarchical bubble priority index; Step S235: performing index balance optimization on the content-level bubble priority index through the AVL tree to obtain a content-level bubble priority balanced index; Step S236: constructing a content-related search library for the overlapping data related to the outpatient clinic to which the content belongs according to the content-level bubble priority balance index and the content factor-related derivation data, and obtaining a content-related search library.

[0011] The present invention greatly improves the construction efficiency and accuracy of the content-related retrieval library through a refined data processing flow. First, step S231 can identify and integrate health information areas that appear repeatedly in different outpatient categories by performing overlapping area aggregation analysis on the overlapping data associated with the outpatient clinics to which the content belongs. This aggregation analysis can significantly reduce data redundancy and provide more concentrated and useful information for subsequent analysis. Then, step S232 uses the content overlapping area aggregation data and complexity data to construct a content structure hierarchical network diagram. This network diagram shows the hierarchical relationship and structure of the data more clearly, which helps to identify the association and its importance between different data points. Step S233 performs factor association derivation analysis on this basis to reveal the key factors and their relationships in the network diagram, which helps to deeply understand the complex interactions between data. Step S234 designs a hierarchical bubble priority index for the network diagram based on the derived factor relationship. This index optimization technology improves the access speed and retrieval efficiency of the data. Subsequently, step S235 performs balanced optimization of the bubble priority index through the AVL tree to ensure rapid response and efficient processing during data retrieval. Finally, step S236 combines the optimized index and factor association derivation data to construct a content association search library. Overall, this series of steps greatly improves the organizational structure and search efficiency of the search library through refined data processing and optimization technology, making it more efficient to obtain and use information when processing complex health data, and promoting accurate health data analysis and management.

[0012] Preferably, step S24 comprises the following steps: Step S241: performing search semantic domain distribution analysis on the content-related search library to obtain search semantic domain distribution data; Step S242: performing context dependency sparsity analysis according to the retrieval semantic domain distribution data to obtain retrieval semantic context dependency sparsity data; Step S243: performing multi-directional reference retrieval semantic extraction on the retrieval semantic domain distribution data according to the retrieval semantic context dependency sparsity data to obtain multi-directional reference retrieval semantic data; Step S244: dividing the search semantic domain distribution data into search content intermediate levels based on the multi-directional reference search semantic data to obtain search content intermediate level data; Step S245: perform vertical slicing of the content middle layer according to the multi-directional reference retrieval semantic data and the retrieval content middle-level data to obtain a content-related middle-layer slicing retrieval library.

[0013] The present invention greatly improves the performance and effect of the content-related intermediate layer fragmented retrieval library through a series of refined analysis and processing steps. Step S241 first performs a retrieval semantic domain distribution analysis on the content-related retrieval library. The main purpose of this step is to identify and map the distribution of different retrieval semantic domains, so as to understand the overall layout of semantic information in the data. This lays the foundation for further semantic analysis. Then, step S242 uses the retrieval semantic domain distribution data to perform context-dependent sparsity analysis to evaluate the sparsity of semantics in different contexts. This analysis helps to identify which semantic areas are relatively sparse in a specific context, thereby optimizing semantic matching and context understanding during the retrieval process. Step S243 performs multi-directional reference retrieval semantic extraction on this basis, and obtains more comprehensive and accurate retrieval semantic data through multi-dimensional semantic reference extraction. This extraction can provide richer semantic information and enhance the understanding ability of the retrieval system. Subsequently, step S244 divides the retrieval semantic domain distribution data into intermediate levels according to the extracted multi-directional reference retrieval semantic data, and organizes the retrieval content into a hierarchical structure. This division can effectively manage and optimize the retrieval content, making the retrieval process more efficient. Finally, step S245 performs vertical sharding of the content middle layer according to the multi-directional reference retrieval semantic data and the middle-level data to form a content-related middle-layer sharding retrieval library. This sharding technology not only improves the data retrieval speed, but also optimizes the system's processing capabilities for complex queries. Overall, these steps significantly improve the accuracy and efficiency of the retrieval library through in-depth semantic analysis and optimization, ensuring that information retrieval and analysis can be performed quickly and accurately when processing large and complex health data.

[0014] Preferably, step S3 comprises the following steps: Step S31: searching for group basic physical data of users based on historical physical examination report data to obtain group basic physical data of users; Step S32: identifying health risk factors of the user's basic group data based on the content-related intermediate layer fragment retrieval library to obtain health risk factor data; Step S33: Perform health trend assessment on the user's basic group data based on the health risk factor data to obtain health trend assessment data.

[0015] In step S3, the present invention effectively improves the management and analysis capabilities of user health data through a series of precise analysis steps. Step S31 first retrieves group user basic body data based on historical physical examination report data. The goal of this step is to extract the user's basic body information from large-scale data to form a comprehensive basic body data set. This provides a wide range of basic data for subsequent health analysis. Next, step S32 uses the content-related intermediate layer fragmentation retrieval library to identify health risk factors for these user basic body group data. This process helps to reveal potential health problems and risks by analyzing the user's health data and identifying health risk factors. This identification can provide a key basis for formulating personalized health management strategies. Finally, step S33 evaluates the health trend of the user's basic body data based on the identified health risk factors. This step evaluates the user's health trend by comprehensively analyzing the health risk factors, revealing potential health changes and development trends. This evaluation can help predict future health conditions and provide users with more forward-looking health advice. Overall, these steps greatly improve the depth and accuracy of health data analysis through systematic health data retrieval and risk assessment, making health management more personalized and forward-looking, thereby supporting more effective health intervention and management decisions.

[0016] Preferably, step S32 includes the following steps: Step S321: perform a physical examination evaluation content similarity assessment on the user's physical basic group data according to the content-related intermediate layer segmented retrieval library to obtain physical examination evaluation content similarity data; Step S322: performing cluster analysis on the physical examination evaluation content similarity data to obtain physical examination evaluation content cluster data; Step S323: reconstructing the individual health trajectory according to the physical examination evaluation content similarity data and the physical examination evaluation content clustering data to obtain individual health trajectory reconstruction data; Step S324: using a preset individual risk identification model to identify individual risk characteristic factors of the individual health trajectory reconstruction data and the physical examination evaluation content clustering data, and obtaining individual risk characteristic factor data; Step S325: performing multivariate sensitivity analysis on individual risk characteristic factor data to obtain risk multivariate sensitivity data; Step S326: Identify health risk factors based on individual risk characteristic factor data and risk multivariate sensitivity data to obtain health risk factor data.

[0017] The present invention comprehensively optimizes the identification and evaluation of health risk factors through a series of sophisticated analysis steps. Step S321 evaluates the similarity of the physical examination evaluation content of the user's basic group data according to the content-related intermediate layer fragmentation retrieval library. The core of this step is to analyze the similarities in the user's physical examination report and reveal the similarity relationship between different physical examination evaluation contents. This similarity evaluation can help discover the commonalities and patterns in the data and provide basic data for subsequent analysis. Next, step S322 performs cluster analysis on the physical examination evaluation content similarity data, groups the physical examination evaluation contents with similar characteristics, and obtains the physical examination evaluation content cluster data. This cluster analysis helps to organize a large amount of physical examination data into meaningful categories, so as to more clearly understand the health patterns and trends in the data. Subsequently, step S323 combines the physical examination evaluation content similarity data and cluster data to reconstruct the individual health trajectory. This process reconstructs the health change trajectory of the user by comprehensively analyzing the user's historical physical examination data and provides an individualized health trend view. Step S324 uses the preset individual risk identification model to identify individual risk characteristic factors for the reconstructed health trajectory data and physical examination evaluation content clustering data. The goal of this step is to identify the health risk factors faced by each individual and provide targeted information for risk management. Step S325 performs multivariate sensitivity analysis on the identified individual risk characteristic factor data to evaluate the impact of different risk factors on health, thereby revealing which factors are more sensitive to health risks. This multivariate analysis can help identify key health risk factors and formulate more accurate intervention measures. Finally, step S326 identifies health risk factors based on individual risk characteristic factor data and risk multivariate sensitivity data, and identifies the main health risk factors based on various types of information. This process ensures the comprehensive identification and accurate assessment of the user's health risks, and provides a solid foundation for the formulation of personalized health management plans. Overall, these steps have greatly improved the accuracy and practicality of health risk factor identification through multi-level and multi-dimensional data analysis and optimization, which helps to achieve more accurate and effective health management.

[0018] Preferably, step S33 includes the following steps: Step S331: Perform dynamic cross-feature association analysis on the health risk factor data to obtain dynamic health feature association data; Step S332: performing trend path analysis on the user's basic group data based on the dynamic health feature association data to obtain health trend path data; Step S333: Perform health trend assessment on the user's basic group data based on the dynamic health feature association data and the health trend path data to obtain health trend assessment data.

[0019] The present invention significantly improves the accuracy and comprehensiveness of health trend assessment through a series of complex analysis steps. Step S331 performs dynamic cross-feature correlation analysis on health risk factor data. The core of this step is to analyze the relationship between different health risk factors and their changing dynamics to obtain dynamic health feature correlation data. This dynamic analysis can reveal the mutual influence of risk factors at different time points and conditions, and provide a deeper understanding of health risks. Next, step S332 uses dynamic health feature correlation data to perform trend path analysis on the user's basic group data. This process reveals the trend path of health status by applying dynamic feature data to group basic data, and provides a time series view of individual health changes. This path analysis helps to identify patterns and changes in health trends and provides a reliable basis for health prediction. Finally, step S333 evaluates the health trend of the user's basic group data based on dynamic health feature correlation data and health trend path data. This comprehensive evaluation process comprehensively analyzes the user's health trend by combining dynamic features and trend paths to obtain health trend evaluation data. This evaluation can provide an in-depth prediction of future health status and help formulate more accurate health management strategies. Overall, these steps improve insights into health trends through comprehensive analysis and dynamic data processing, making health management more proactive and personalized.

[0020] Preferably, step S4 comprises the following steps: Step S41: normalizing the health trend assessment data to obtain health trend assessment normalized data; Step S42: Correcting the user's health behavior based on the health trend assessment normalized data to obtain health behavior correction data; Step S43: Push the health behavior correction data based on the cloud platform to perform user health data management.

[0021] The present invention significantly improves the effect and accuracy of user health behavior management through a series of orderly processing steps. Step S41 normalizes the health trend assessment data, converts data of different ranges and scales into a unified standard, and thus obtains the normalized data of health trend assessment. This process ensures the consistency and comparability of the data, making the subsequent analysis more accurate and reliable, and facilitating comparison between different data sources and user groups. Then, step S42 corrects the user's health behavior based on the normalized data. By conducting an in-depth analysis of the health trend assessment data, identifying and correcting bad health behaviors, and thus obtaining health behavior correction data. The core of this process is to formulate specific health improvement suggestions based on the data analysis results, and implement these suggestions into the user's daily behavior to help the user improve health habits. Finally, step S43 uses the cloud platform to push the health behavior correction data. This process transmits the corrective measures and personalized health management solutions to the user in real time through the cloud platform, realizing the dynamic management and real-time update of health data. With the support of the cloud platform, users can get the latest health advice and guidance at any time, enhancing the convenience and real-time nature of health management. Overall, these steps ensure the efficiency, accuracy and personalization of the health management process through data normalization, behavior correction and cloud platform push, helping users achieve better health results and quality of life.

[0022] Preferably, the present invention also provides a user health data management system based on a cloud platform, which is used to execute the user health data management method based on a cloud platform as described above, and the user health data management system based on a cloud platform includes: Outpatient clinic classification data, used to obtain historical physical examination report data uploaded by users; historical physical examination report data is classified into the types of outpatient clinics to which the contents belong, and the outpatient clinic type data to which the physical examination contents belong is obtained; The outpatient clinic association analysis module is used to perform an outpatient clinic association overlap analysis on the historical physical examination report data according to the outpatient clinic type data to which the physical examination content belongs, and obtain the outpatient clinic association overlap data to which the content belongs; a content association retrieval library is constructed for the outpatient clinic association overlap data to obtain a content association intermediate layer fragmentation retrieval library; The health trend analysis module is used to identify health risk factors based on the content-related intermediate layer fragment retrieval library to obtain health risk factor data; perform health trend assessment based on the health risk factor data to obtain health trend assessment data; The user health data management module is used to correct the user's health behavior based on the health trend assessment data to obtain the health behavior correction data; and push the health behavior correction data based on the cloud platform to perform user health data management.

[0023] The beneficial effect of the present invention is that by classifying the types of outpatient clinics for the historical physical examination report data uploaded by the user, the physical examination content can be systematically divided into different outpatient clinic categories. The complex physical examination data can be structured so that the data of each physical examination report is easier to manage and analyze. Secondly, this classification not only improves the efficiency of data processing, but also lays a solid foundation for subsequent analysis, such as more accurate health risk assessment based on the types of outpatient clinics. By performing association overlap analysis on the outpatient clinic type data to which the physical examination content belongs, the connection between different outpatient clinic categories is further explored. The key to this step is that it can reveal the content overlap or association between different outpatient clinics, thereby discovering potential health problems. For example, the examinations of the same patient in different outpatient clinics show some repeated or related health problems, and this analysis can provide a more comprehensive understanding of the patient's health status. Subsequently, these overlapping analysis results will be used to construct a content association retrieval library, which not only integrates the associated data between outpatient clinics, but also subdivides the retrieval library into middle-layer fragments, effectively improving the speed and accuracy of data retrieval. This fragmented retrieval library can quickly locate relevant information in large-scale data, provide support for medical decision-making, and also make further personalized health analysis and prediction possible. In general, this technical solution significantly improves the utilization value of physical examination report data and optimizes the health management process through systematic data processing and association analysis. By using the content-related intermediate layer fragmentation retrieval library to identify health risk factors, potential health risk factor data can be accurately extracted. The core advantage of this process is that it can identify key factors related to health risks from complex physical examination data, such as abnormal changes in certain indicators or specific health signs. This accurate identification can not only help doctors or health managers better understand the health status of patients, but also provide basic data for further health trend assessment. Subsequently, health trend assessment based on these health risk factor data can provide insights into the dynamic changes in the patient's health status, such as the increasing trend of risk factors or the improvement of health indicators. The beneficial effect of this trend assessment is that it can help detect potential health problems early and provide a scientific basis for the formulation of personalized health management plans. Personalized health intervention measures are formulated based on health trend assessment data. For example, based on the assessment results, the system can recommend users to adjust their eating habits, increase exercise, or conduct regular health checks. Health behavior correction data can effectively reflect the implementation of these interventions and their actual impact on user health. Furthermore, by pushing health behavior correction data based on the cloud platform, real-time health management and feedback can be achieved. The application of this cloud platform not only makes the health management process more efficient and convenient, but also ensures that users can get timely advice and reminders on health management through automated data push functions, thereby promoting continuous health behavior improvement.In general, this technical solution provides users with a comprehensive and personalized health management solution through accurate health risk identification, trend assessment and behavior correction, combined with advanced cloud platform technology, greatly improving the effectiveness and convenience of health management. Therefore, the present invention is an improvement on a traditional user health data management method based on a cloud platform, solving the problem that a traditional user health data management method based on a cloud platform cannot quickly retrieve and associate health status, and the problem of inaccurate health trend assessment. It can quickly retrieve and associate health status, and improve the accuracy of health trend assessment. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 A schematic diagram of the steps of a user health data management method based on a cloud platform; Figure 2 for Figure 1 Detailed implementation steps of step S2 in FIG.

[0025] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0026] The technical method of the present invention is described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.

[0027] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0028] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0029] To achieve this, please refer to Figure 1 to Figure 2 , a user health data management method based on a cloud platform, the method comprising the following steps: Step S1: Obtain historical physical examination report data uploaded by the user; classify the historical physical examination report data into the type of clinic to which the content belongs, and obtain the type of clinic to which the physical examination content belongs; Step S2: performing overlapping analysis of the clinics to which the contents of the physical examination belong according to the clinic category data of the physical examination contents on the historical physical examination report data to obtain overlapping data of the clinics to which the contents belong; constructing a content-related retrieval library for the overlapping data of the clinics to which the contents belong to obtain a content-related intermediate layer fragmented retrieval library; Step S3: Identify health risk factors based on the content-related intermediate layer shard retrieval library to obtain health risk factor data; perform health trend assessment based on the health risk factor data to obtain health trend assessment data; Step S4: Correct the user's health behavior based on the health trend assessment data to obtain health behavior correction data; push the health behavior correction data based on the cloud platform to perform user health data management.

[0030] In the embodiment of the present invention, reference Figure 1 The above is a schematic diagram of the steps of a user health data management method based on a cloud platform of the present invention. In this example, the user health data management method based on a cloud platform includes the following steps: Step S1: Obtain historical physical examination report data uploaded by the user; classify the historical physical examination report data into the type of clinic to which the content belongs, and obtain the type of clinic to which the physical examination content belongs; In an embodiment of the present invention, when obtaining the historical physical examination report data uploaded by the user, the physical examination report uploaded by the user is received and stored through the file transfer protocol (such as FTP or HTTP / HTTPS) built into the cloud platform. First, the system standardizes the report file format to ensure that all physical examination report data formats are consistent, such as converting PDF, pictures, and text files of different formats into structured text data. Then, a regular expression-based text parsing method is used to scan the report content word by word, extract key medical terms, examination items, numerical values, and diagnostic conclusions, and segment and label the data according to preset field standards. In the labeling process, the report content is pre-classified according to a specific medical terminology library and rules formulated by experts to determine the attribution of each examination item or diagnostic data in the report. Finally, a multi-level keyword matching and rule mapping method is used to divide the contents in the physical examination report into the outpatient category to which it belongs, such as internal medicine, surgery, cardiovascular department, etc., and the outpatient category data to which the physical examination content belongs is output to ensure data accuracy and consistency.

[0031] Step S2: performing overlapping analysis of the clinics to which the contents of the physical examination belong according to the clinic category data of the physical examination contents on the historical physical examination report data to obtain overlapping data of the clinics to which the contents belong; constructing a content-related retrieval library for the overlapping data of the clinics to which the contents belong to obtain a content-related intermediate layer fragmented retrieval library; In the embodiment of the present invention, when the historical physical examination report data is subjected to the overlapping analysis of the outpatient clinics to which the contents belong according to the outpatient clinic category data obtained in step S1, it is first necessary to establish the association rules between the outpatient clinic categories. After the physical examination contents of each report are mapped to the outpatient clinic categories, the construction method of the outpatient clinic association matrix is ​​adopted to summarize the association situations of all outpatient clinic categories in the historical data, and the association strength between each outpatient clinic is calculated by statistically analyzing the repetition rate and the frequency of simultaneous occurrence between the physical examination items. In the specific implementation process, cluster analysis and association rule mining techniques are used, based on the logical association between the physical examination items, the relationship between the outpatient clinic categories is gradually subdivided through hierarchical clustering to form the overlapping data of the outpatient clinics to which the contents belong. Subsequently, the overlapping data is further analyzed, and the content-related retrieval library is constructed in combination with the depth and breadth indicators of the content overlap. When the library is constructed, the fragmented storage and indexing technology is adopted to store the overlapping data in the retrieval library in fragments, and a fast retrieval index is constructed for each fragment, so as to ensure that the content-related intermediate layer fragmented retrieval library can realize fast and accurate content positioning and extraction in subsequent retrieval operations.

[0032] Step S3: Identify health risk factors based on the content-related intermediate layer shard retrieval library to obtain health risk factor data; perform health trend assessment based on the health risk factor data to obtain health trend assessment data; In the embodiment of the present invention, when health risk factors are identified according to the content-associated intermediate layer shard retrieval library, it is first necessary to classify, retrieve and extract the data in the shard retrieval library according to the predefined health risk indicators. Through methods such as keyword matching, content relevance calculation and frequency analysis, the content related to common health risk factors, such as hypertension, diabetes, hyperlipidemia and other risk factors, is screened out from the retrieval library. Then, these contents are compared and labeled item by item using the rule-defined method, and the factors that lead to health risks are marked and quantified, such as the physical examination indicators for marking hypertension-related factors are blood pressure values, abnormal electrocardiograms, signs of arteriosclerosis, etc. Further, all marked health risk factors are summarized and health risk factor data is output. On the basis of identifying health risk factors, the time series analysis method is used to evaluate the change trend of these risk factors, comprehensively considering factors such as the change rate, fluctuation frequency, and abnormal data points in the historical data, and calculating the long-term trend and short-term fluctuation of individual health, and outputting health trend evaluation data. The health trend evaluation data is specifically reflected in the change trajectory of risk factors over time, such as the gradual increase in blood sugar levels or the increase in the frequency of abnormal heart rate, which is convenient for subsequent health intervention and management.

[0033] Step S4: Correct the user's health behavior based on the health trend assessment data to obtain health behavior correction data; push the health behavior correction data based on the cloud platform to perform user health data management.

[0034] In the embodiment of the present invention, when correcting the user's health behavior based on the health trend assessment data, the bad behavior patterns in the health trend are first identified and analyzed, and the specific data points that deviate from the healthy behavior are identified by comparing the health trend assessment data with the preset health behavior reference standard, such as blood sugar fluctuations caused by irregular diet, cardiovascular problems caused by insufficient exercise, etc. For these deviant behaviors, a rule-driven behavior correction suggestion generation method is adopted to formulate personalized health behavior correction strategies according to the assessed health trends, such as suggestions to increase aerobic exercise, adjust dietary structure, and have a regular work and rest schedule. Subsequently, the generated health behavior correction data is formatted into content that is easy for users to understand, including specific action suggestions and expected health improvement effects. Finally, the health behavior correction data is accurately pushed to the user terminal through the data push service of the cloud platform. During the push process, the real-time and reliability of data push are ensured through the message queue management mechanism of the cloud platform, and the user's response to the correction suggestion is tracked to form a closed-loop management system. This push process not only ensures that users can obtain health behavior suggestions in a timely manner, but also provides data support for subsequent health management, and further improves the overall framework of user health data management.

[0035] Preferably, step S1 comprises the following steps: Step S11: Obtain historical physical examination report data uploaded by the user; Step S12: Clean the historical physical examination report data to obtain cleaned data of the historical physical examination report; Step S13: labeling the cleaned data of the historical physical examination report to obtain labeled data of the historical physical examination report; Step S14: classify the labeled data of historical physical examination reports according to the type of clinic to which the contents belong, and obtain the data on the type of clinic to which the physical examination contents belong.

[0036] In an embodiment of the present invention, when obtaining historical physical examination report data uploaded by a user, a file receiving module based on a cloud platform receives and classifies the report files uploaded by the user in real time through a standardized data transmission protocol (such as FTP, HTTP / HTTPS). During the receiving process, the file transmission module performs format verification and inspection on the uploaded report data to ensure the integrity and losslessness of the file. When receiving the file, the physical examination report is stored in a designated database of the cloud platform through a distributed file system, and a unique identifier is generated for subsequent call and management during operation. To ensure data security, the receiving process uses an encryption protocol (such as TLS / SSL) to encrypt data transmission, and data backup is performed when the database is stored to ensure that the data will not be lost or tampered with. When cleaning the historical physical examination report data, the original physical examination report data received is first pre-processed by a data cleaning engine to remove redundant information, such as irrelevant headers and footers, pictures, blank characters, and data items with incorrect formats. Using string parsing and matching rules, the system scans the content in the report line by line to filter out data that does not meet the medical terminology standards, such as garbled characters, duplicate records, and missing values. The values ​​of key data items, such as blood pressure and blood sugar, are unified in units and standardized in format to ensure data consistency. For data items with incomplete or obviously abnormal content, the missing value filling method or logical verification mechanism defined by the rules is used to complete or correct the data to maintain data integrity. After the cleaning is completed, the historical physical examination report cleaning data is generated, and the error log in the cleaning process is recorded in the system log for subsequent verification and adjustment. When labeling the physical examination data of the historical physical examination report cleaning data, the rule-based text annotation technology is used to label the contents of the cleaned physical examination report item by item. During the labeling process, the system uses the built-in medical terminology library to classify and label each examination item, diagnostic conclusion and indicator value in the physical examination data, such as labeling "blood pressure" as "cardiovascular indicator" and labeling "fasting blood sugar" as "metabolic indicator". The labeling operation also includes unit conversion and standardization of numerical data to ensure the uniformity of data annotation. For polysemous words or ambiguous content in the text description, multiple rounds of matching and context analysis are used to ensure the accuracy of the label. During the labeling process, all labels are stored in a structured data table, and the labeled data of historical physical examination reports are generated to facilitate subsequent classification and analysis operations. When the labeled data of historical physical examination reports are divided into the types of clinics to which the content belongs, the predefined clinic classification rules are used to match the labeled data. By comparing the medical clinic fields to which each label belongs, the labeled data is assigned to specific clinic types. The clinic classification process is strictly based on national or industry medical classification standards. For example, data containing "electrocardiogram" and "blood pressure" are classified as cardiovascular medicine, and data containing "liver function" and "blood lipids" are classified as gastroenterology.In the division operation, by building a multi-level classification structure, the multiple category labels in the report are screened and associated step by step to ensure that each data is accurately classified. The division results form the outpatient category data to which the physical examination content belongs, and after output, they are provided to the subsequent association overlap analysis steps to ensure that the system can carry out targeted content management and health assessment.

[0037] Preferably, step S2 comprises the following steps: Step S21: performing overlapping analysis of clinic-related contents on the historical physical examination report data according to the clinic-related types of the physical examination contents, and obtaining overlapping clinic-related contents; Step S22: evaluating the correlation complexity of the overlapping data of the clinics to which the content belongs, and obtaining the correlation complexity data of the clinics to which the content belongs; Step S23: constructing a content-related search library for the overlapping data of the clinics to which the content belongs according to the complexity data of the clinics to which the content belongs, and obtaining a content-related search library; Step S24: vertically slice the content-related middle layer of the content-related search library to obtain a content-related middle layer sliced ​​search library.

[0038] As an example of the present invention, refer to Figure 2 As shown, in this example, step S2 includes: Step S21: performing overlapping analysis of clinic-related contents on the historical physical examination report data according to the clinic-related types of the physical examination contents, and obtaining overlapping clinic-related contents; In an embodiment of the present invention, when performing an analysis of the association overlap of the outpatient clinics to which the historical physical examination report data belongs according to the outpatient clinic category data to which the physical examination content belongs, it is first necessary to construct an association analysis matrix between outpatient clinic categories. By scanning the data in the physical examination report, the overlapping contents of different outpatient clinic categories are counted, such as physical examination items involved in multiple outpatient clinics at the same time or repeated diagnostic conclusions. The association overlap analysis adopts an item-by-item comparison method, uses matching rules and outpatient clinic classification standards, calculates the association degree of items belonging to different outpatient clinic categories in the same patient report, and counts the frequency of occurrence of cross-data between categories. In the analysis, based on the overlap and association strength of different physical examination items, the association overlap data of the outpatient clinics to which the content belongs is generated to ensure that the data covers all potential cross-outpatient clinic contents. The output data is used to evaluate the association complexity between outpatient clinic categories and provide basic data for the subsequent construction of a retrieval library.

[0039] Step S22: evaluating the correlation complexity of the overlapping data of the clinics to which the content belongs, and obtaining the correlation complexity data of the clinics to which the content belongs; In an embodiment of the present invention, when evaluating the complexity of the association of the overlapping data associated with the outpatient clinics to which the content belongs, the overlapping data is first grouped and clustered, and the complexity of the data is quantified based on the association frequency between the outpatient clinics, the number of repeated items, and the degree of data intersection. During the evaluation process, a complexity scoring model is constructed to analyze the diversity and association level of the outpatient clinic-related data, such as the depth of intersection of multiple outpatient items, the number of association paths, and the degree of mutual influence. When calculating the association strength and complexity index, a weighted scoring method is used to comprehensively consider the distribution of associated data between each outpatient clinic category to obtain the complexity value of each outpatient clinic association. The complexity data of the outpatient clinics to which the content belongs is output, and the evaluation results provide a basis for the construction of a content-related retrieval library and lay a data foundation for the subsequent optimization of the retrieval structure.

[0040] Step S23: constructing a content-related search library for the overlapping data of the clinics to which the content belongs according to the complexity data of the clinics to which the content belongs, and obtaining a content-related search library; In an embodiment of the present invention, when constructing a content-related retrieval library for overlapping data associated with the outpatient clinic to which the content belongs based on the complexity data associated with the outpatient clinic to which the content belongs, firstly, the overlapping data is hierarchically divided and indexed according to the complexity evaluation result, and a partition storage scheme is established according to the data complexity distribution. The library construction process adopts the inverted index technology to index the overlapping content and associated paths to ensure that the associated content can be quickly located during retrieval. Through content segmentation and index optimization, a multi-level content-related retrieval structure is established to ensure efficient retrieval in a complex data environment. In the retrieval library, data with higher complexity is stored separately, and special retrieval tags are attached to ensure accurate search of highly complex associated content. The output content-related retrieval library has the functions of fast retrieval and hierarchical calling, providing reliable support for subsequent data management and risk identification.

[0041] Step S24: vertically slice the content-related middle layer of the content-related search library to obtain a content-related middle layer sliced ​​search library.

[0042] In an embodiment of the present invention, when the content-related retrieval library is vertically sharded at the content middle layer, a sharding technology based on vertical segmentation is adopted to vertically cut the associated data in the retrieval library according to the outpatient category and the complexity of the association. During the sharding process, a multi-level vertical sharding structure is constructed according to the depth of content association and the complexity of the retrieval path, and high-frequency associated content and low-frequency associated content are stored separately to form an independent index of the content shards. The vertical sharding operation ensures that each shard contains only associated data of a specific category or a specific complexity, thereby reducing the interference of cross-shard retrieval. The shard library is independently optimized and compressed to improve the retrieval efficiency, and finally a content-related middle-layer sharded retrieval library is obtained, which provides fast and accurate data access capabilities for subsequent health risk factor identification and health assessment processes.

[0043] Preferably, step S23 includes the following steps: Step S231: performing overlapping area aggregation analysis on the overlapping data associated with the clinics to which the content belongs, and obtaining content overlapping area aggregation data; Step S232: constructing a content structure hierarchical network diagram according to the content overlap area aggregation data and the clinic association complexity data to which the content belongs, to obtain a content structure hierarchical network diagram; Step S233: performing factor correlation derivation analysis on the content structure hierarchical network diagram to obtain content factor correlation derivation data; Step S234: designing a hierarchical bubble priority index for the content structure hierarchical network diagram according to the content factor association derivation data to obtain a content hierarchical bubble priority index; Step S235: performing index balance optimization on the content-level bubble priority index through the AVL tree to obtain a content-level bubble priority balanced index; Step S236: constructing a content-related search library for the overlapping data related to the outpatient clinic to which the content belongs according to the content-level bubble priority balance index and the content factor-related derivation data, and obtaining a content-related search library.

[0044] In the embodiment of the present invention, when the overlapping area aggregation analysis is performed on the overlapping data associated with the outpatient clinic to which the content belongs, the associated overlapping data is firstly aggregated by region, and the data is grouped according to the outpatient clinic category and the overlapping frequency. The overlapping areas between each outpatient clinic category are scanned item by item to identify the high-frequency overlapping content and its distribution characteristics. The aggregation analysis of the overlapping area adopts regional density calculation and association matrix operation, and the core area with the highest overlap in each outpatient clinic category is calibrated through multi-level overlapping distinction. The aggregation result is output as the content overlapping area aggregation data, which includes the association density, cross frequency and corresponding specific items of each overlapping area, and provides accurate data support for the subsequent structural hierarchy construction. When the content structure hierarchy network diagram is constructed according to the content overlapping area aggregation data and the complexity data associated with the outpatient clinic to which the content belongs, the associated nodes in the overlapping area aggregation data are first used as the basic nodes of the network diagram for diagram construction. According to the association complexity data, according to the interaction strength and hierarchy between the nodes, the relationship between the nodes is defined as a directed edge to form a hierarchical structure network. In the process of constructing the network diagram, the weight setting of the node link strength is used to ensure that the high-frequency associated nodes are in the core level of the network. The hierarchical division of the network graph strictly follows the aggregation results of the overlapping areas, and the content structure hierarchical network graph is output to form a complete hierarchical association structure, which provides a basis for the subsequent factor association derivation. When the factor association derivation analysis of the content structure hierarchical network graph is carried out, the factor analysis method is used to derive the association strength of each node in the network graph layer by layer. During the derivation and analysis process, the key association factors are identified and marked by calculating the edge weights between nodes, and the path evaluation is performed on their propagation paths in the network graph. The derivation process requires full path analysis of the factor association path, analyzing the interaction mode and intensity between different factors, so as to obtain complete factor association derivation data. The output content factor association derivation data clearly shows the hierarchical association relationship between factors and the control influence of key nodes, providing the necessary association basis for subsequent index design. When the hierarchical bubble priority index design of the content structure hierarchical network graph is carried out according to the content factor association derivation data, the bubble priority sorting algorithm is used to prioritize the hierarchical factors in the network graph. In the index design, nodes with high hierarchical levels and strong association strength are placed at the top of the index, and the priority position of each node is dynamically adjusted through the bubble process to optimize the retrieval efficiency. During the design of hierarchical bubble priority index, the index nodes need to be dynamically tuned according to the factor strength in the associated derivation data to ensure that the most commonly used associated factors are in the optimal retrieval position. The output content hierarchical bubble priority index has efficient hierarchical access capabilities, providing a basic framework for subsequent index balance optimization. When optimizing the index balance of the content hierarchical bubble priority index through the AVL tree, the self-balancing mechanism of the balanced tree is used to adjust the node height and rotate the bubble priority index.During the balance optimization process, the node balance adjustment is automatically performed according to the node association strength and index access frequency to avoid long paths during index access. The balance adjustment uses single rotation and double rotation operations to reconstruct the unbalanced nodes to ensure that the tree structure always remains balanced. The optimized index tree has an efficient balanced retrieval path, and outputs the content level bubble priority balanced index, which provides a stable balanced structure for the rapid retrieval function of the content association retrieval library. When constructing the content association retrieval library for the overlapping data of the outpatient clinic to which the content belongs according to the content level bubble priority balanced index and the content factor association derivation data, the balanced index is first used as the core structure of the retrieval library to ensure the efficiency and accuracy of the retrieval process. In the construction of the retrieval library, the hierarchical relationship in the factor association derivation data is used to logically group and independently store the library content, and the hierarchical location of the retrieval data is realized quickly. Through the organic combination of index and derivation data, efficient data search and rapid acquisition of related content are achieved. The final output content association retrieval library has a powerful hierarchical association retrieval capability, providing all-round support for user health data management on the cloud platform.

[0045] Preferably, step S24 comprises the following steps: Step S241: performing search semantic domain distribution analysis on the content-related search library to obtain search semantic domain distribution data; Step S242: performing context dependency sparsity analysis according to the retrieval semantic domain distribution data to obtain retrieval semantic context dependency sparsity data; Step S243: performing multi-directional reference retrieval semantic extraction on the retrieval semantic domain distribution data according to the retrieval semantic context dependency sparsity data to obtain multi-directional reference retrieval semantic data; Step S244: dividing the search semantic domain distribution data into search content intermediate levels based on the multi-directional reference search semantic data to obtain search content intermediate level data; Step S245: perform vertical slicing of the content middle layer according to the multi-directional reference retrieval semantic data and the retrieval content middle-level data to obtain a content-related middle-layer slicing retrieval library.

[0046] In an embodiment of the present invention, when performing a search semantic domain distribution analysis on a content-related search library, firstly, a comprehensive vocabulary decomposition is performed on the semantic content in the search library, and the search data is classified and labeled according to the frequency of use, context relevance and semantic category of each semantic unit. The word frequency statistics and semantic clustering technology are used to analyze the distribution characteristics of each semantic unit in the search library in detail, and similar semantic contents are classified to form a specific semantic domain. The distribution of each semantic domain is graphically represented, and the relevance and difference between the semantic domains are clarified by calculating the similarity and overlap between domains. Finally, the search semantic domain distribution data is obtained, in which the relative position and proportion of each semantic domain in the search library are clarified, laying the foundation for subsequent sparsity analysis. When performing context-dependent sparsity analysis based on the search semantic domain distribution data, relying on the hierarchical relationship of the semantic domain distribution, a context-dependent scan is performed for each semantic domain, and by analyzing the context connection strength between the semantic domains, the area with relatively sparse context connection is identified. During the sparsity analysis, point-to-point dependency statistics and distribution map evaluation are used to quantify the sparsity of the semantic domain through a multidimensional matrix, highlighting the frequency and location of sparsely connected and discontinuous semantic fragments. The contextual sparsity data specifically marks the high sparsity areas, providing a basis for semantic extraction. The output retrieval semantic context dependency sparsity data shows the dependency weaknesses between semantic domains, helping to optimize the direction of semantic extraction. When performing multi-directional reference retrieval semantic extraction on the retrieval semantic domain distribution data based on the retrieval semantic context dependency sparsity data, the key semantic nodes with cross-semantic domain connection attributes are extracted by multi-directional reference scanning of the sparsity marked areas. During the extraction process, the multi-directional reference strategy of weak context association is used to perform semantic analysis from multiple directions at the same time to ensure that the extracted semantic information has high complementarity and integrity. The recursive reference retrieval technology is used to perform multi-level recursive scanning on each node with high sparsity, and similar semantic units are uniformly summarized to form complete multi-directional reference retrieval semantic data. This data contains the semantic nodes and their connection paths referenced across domains, providing refined semantic information for subsequent content division. When dividing the retrieval content middle level of the retrieval semantic domain distribution data based on multi-directional reference retrieval semantic data, the structural advantages of multi-directional reference semantics are used to hierarchically reconstruct the content in the retrieval library according to the reference relationship. During the division process, the core nodes are classified into the upper level according to the node priority and reference level in the semantic data, and the secondary nodes are classified into the lower level in order of reference frequency. Through the hierarchical operation, the semantic content of each level is reordered to construct a multi-level semantic structure, making the access path between semantics more intuitive. The output retrieval content middle level data clarifies the position of each semantic node in the hierarchical structure, providing a structural basis for subsequent vertical sharding.When performing vertical sharding of the content middle layer based on multi-directional reference retrieval semantic data and retrieval content middle-level data, first perform a vertical layered scan of the middle-level data, and independently shard the content according to the hierarchy and semantic relationship. During the sharding process, highly related semantic content is placed in the same shard strictly according to the hierarchical guidance in the multi-directional reference semantic data to reduce the complexity of cross-shard retrieval. Vertical sharding forms logically clear retrieval library fragments through layer-by-layer division and intra-layer aggregation operations to ensure the access efficiency of data in the vertical direction. The final output content-related middle-layer shard retrieval library has fast hierarchical positioning capabilities and precise semantic access characteristics, providing strong retrieval support for user health data management on the cloud platform.

[0047] Preferably, step S3 comprises the following steps: Step S31: searching for group basic physical data of users based on historical physical examination report data to obtain group basic physical data of users; Step S32: identifying health risk factors of the user's basic group data based on the content-related intermediate layer fragment retrieval library to obtain health risk factor data; Step S33: Perform health trend assessment on the user's basic group data based on the health risk factor data to obtain health trend assessment data.

[0048] In an embodiment of the present invention, when searching for group user basic body data based on historical physical examination report data, the data in the historical physical examination report is first preliminarily screened according to dimensions such as gender, age, and occupation to form a representative group sample set. On this basis, a structured parsing method is applied to extract key health indicator data such as blood pressure, blood sugar, and heart rate in each physical examination report to construct a comprehensive user health data set. By clustering and statistically analyzing the data set, the general health characteristics and abnormal indicator distribution in the group are identified to ensure the diversity and representativeness of the data. The user's basic body group data finally output contains the basic health characteristics of different groups, providing large-scale data support for the identification of subsequent risk factors. When the user's basic body group data is identified for health risk factors according to the content-related intermediate layer sharding retrieval library, the efficient retrieval capability of the sharding retrieval library is used to perform deep matching on each health indicator. First, the user's basic body group data is input into the intermediate layer sharding retrieval library, and the risk factor standard library in the retrieval library is used as the basis to compare each indicator one by one to identify indicators that meet specific health risk characteristics. The specific operation includes automatic screening of risk factors such as abnormal blood pressure and prediabetes symptoms. The group data is quickly located and labeled using the risk factor thresholds and rules defined in the search library. The output health risk factor data clearly points out the high-risk individuals in the group and their corresponding risk categories, laying the foundation for health trend assessment. When evaluating the health trend of the user's basic group data based on the health risk factor data, the correlation analysis method between risk factors and group health indicators is used to comprehensively evaluate the health change trend of the group. In the specific implementation process, the identified health risk factors are analyzed in time series to identify the change patterns and development trends of risk factors, such as the continued upward trend of hypertension in a certain age group. By comparing and analyzing the frequency and severity of various health risks in different time periods, a trend curve is generated, and trend prediction and risk warning are carried out in combination with historical data. The output health trend assessment data describes in detail the changing trajectory of the group's health status, providing a scientific basis for further intervention and management of user health behaviors.

[0049] Preferably, step S32 includes the following steps: Step S321: perform a physical examination evaluation content similarity assessment on the user's physical basic group data according to the content-related intermediate layer segmented retrieval library to obtain physical examination evaluation content similarity data; Step S322: performing cluster analysis on the physical examination evaluation content similarity data to obtain physical examination evaluation content cluster data; Step S323: reconstructing the individual health trajectory according to the physical examination evaluation content similarity data and the physical examination evaluation content clustering data to obtain individual health trajectory reconstruction data; Step S324: using a preset individual risk identification model to identify individual risk characteristic factors of the individual health trajectory reconstruction data and the physical examination evaluation content clustering data, and obtaining individual risk characteristic factor data; Step S325: performing multivariate sensitivity analysis on individual risk characteristic factor data to obtain risk multivariate sensitivity data; Step S326: Identify health risk factors based on individual risk characteristic factor data and risk multivariate sensitivity data to obtain health risk factor data.

[0050] In an embodiment of the present invention, when the user's basic body group data is evaluated for the similarity of the physical examination evaluation content according to the content-associated intermediate layer fragmented retrieval library, the user's basic body group data is first matched with the reference standard in the retrieval library. By quantifying the values ​​of various health indicators in the physical examination report, similarity calculations are performed using algorithms such as Euclidean distance and cosine similarity, and the indicators in the physical examination report are compared with the known health evaluation content in the retrieval library. In specific operations, the system calculates the similarity of indicators of different individuals item by item to screen out highly similar evaluation content. The output physical examination evaluation content similarity data not only includes the similarity score, but also clarifies the similarity level of each physical examination evaluation content in terms of health indicators, providing data support for subsequent cluster analysis. When clustering analysis is performed on the physical examination evaluation content similarity data, the similarity data is input into a hierarchical clustering algorithm or a K-means clustering algorithm, and the individuals are classified using the similarity features in the data. During the operation, by selecting appropriate cluster centers, the physical examination evaluation contents with high similarity are grouped, and the boundaries of each cluster are adjusted in combination with the distance matrix, so that the sample differences within the same cluster are minimized and the differences between clusters are maximized. The output physical examination evaluation content clustering data includes the center points of each cluster, the distribution characteristics of internal samples and the corresponding health evaluation mode description, which lays a classification foundation for the reconstruction of individual health trajectories. When reconstructing individual health trajectories based on the physical examination evaluation content similarity data and physical examination evaluation content clustering data, the individual data in each cluster are first arranged in time series to determine the changes in the health status of the individual at different time nodes. By using the historical data of multiple physical examinations, the individual's health change process is reconstructed in the form of trajectories. During the operation, a continuous health change path is constructed by superimposing data from different periods. In the process of trajectory reconstruction, it is necessary to focus on analyzing the transition between each cluster and the dynamic evolution of health status. The output individual health trajectory reconstruction data not only shows the individual's health change trend, but also provides a continuous health dynamic reference for the deep identification of individual risks. When using the preset individual risk identification model to identify individual risk characteristic factors from the individual health trajectory reconstruction data and the physical examination evaluation content clustering data, the trajectory reconstruction data and the clustering results are cross-analyzed. In the operation, the key nodes in the trajectory are first determined, such as the time point of health deterioration and the drastic fluctuation of characteristic indicators, and then the factors that cause health risks are matched one by one in combination with the factor feature library preset by the risk identification model. This process includes the extraction of risk indicators, the evaluation of the amplitude of changes, and the marking of abnormal points. The output individual risk characteristic factor data identifies in detail the potential risk points and related characteristics of individuals in the health trajectory, laying the foundation for further sensitivity analysis. When performing multivariate sensitivity analysis on individual risk characteristic factor data, multiple regression analysis and factor sensitivity test methods are used to quantitatively evaluate the degree of influence of each risk factor under different variable combinations.During the analysis process, the weights or values ​​of multiple influencing factors are gradually changed to observe their impact on the changes in health risk characteristics, and highly sensitive variable combinations are identified through sensitivity curves. The output risk multivariate sensitivity data not only shows the sensitivity of each risk factor to health changes, but also clarifies the key factors that lead to a sharp deterioration in health, providing a multi-angle analysis basis for the final identification of risk factors. When identifying health risk factors based on individual risk characteristic factor data and risk multivariate sensitivity data, the factors with higher sensitivity are first screened out from the risk characteristic factors and compared with the dynamic data of individual health trajectories. Combined with the performance of risk factors in different time periods and variable states, the key indicator differences between high-risk and low-risk groups are compared, and the main health risk factors are finally locked by gradually eliminating low-correlation factors. The output health risk factor data includes the sources of health risks that individuals are most susceptible to and their performance characteristics, providing targeted risk warning information for health management.

[0051] Preferably, step S33 includes the following steps: Step S331: Perform dynamic cross-feature association analysis on the health risk factor data to obtain dynamic health feature association data; Step S332: performing trend path analysis on the user's basic group data based on the dynamic health feature association data to obtain health trend path data; Step S333: Perform health trend assessment on the user's basic group data based on the dynamic health feature association data and the health trend path data to obtain health trend assessment data.

[0052] In the embodiment of the present invention, when the dynamic cross-feature correlation analysis is performed on the health risk factor data, the dynamic interaction relationship between different health risk factors is evaluated one by one by constructing a multivariate cross matrix. The specific operation includes selecting multiple time nodes, cross-comparing and analyzing the change trend of the health risk factor with the corresponding characteristic variables, and identifying the factor combination with higher correlation. The correlation coefficient and cross entropy analysis method are used to quantify the mutual influence intensity between the factors to reveal the dynamic correlation characteristics. During the operation, the more complex factor relationships are hierarchically summarized through a step-by-step analysis method, and the strong and weak associations between the dynamic health characteristics are clarified. The output dynamic health feature association data describes in detail the interactive influence mode of each health risk factor and its dynamic characteristics under different conditions, providing necessary data support for subsequent trend path analysis. When the trend path analysis of the user's basic group data based on the dynamic health feature association data is performed, the individual indicators in the group data are first matched with the dynamic health feature association data, and the path of health changes is analyzed in multiple stages using the time series analysis method. During the analysis process, the dynamic path planning algorithm is used to hierarchically deconstruct the impact path of the health risk factor, and gradually present the evolution trajectory of the health state. The changes in health characteristics at each time node will be mapped to a specific path node to form a complete trend path diagram. By superimposing the trend paths of multiple individuals, the overall trend representing the evolution of the health status of the group is obtained. The output health trend path data accurately shows the dynamic changes of each key health characteristic in different periods, provides a clear trend direction and potential change inflection points, and lays an intuitive data foundation for subsequent health trend evaluation. When evaluating the health trend of the user's basic group data based on the dynamic health characteristic association data and the health trend path data, the node characteristics in the trend path are first matched with the association strength of the dynamic health characteristics to evaluate the impact of each health status change node on the overall trend. During the evaluation process, the path integral method is used to accumulate the changes in each section of the trend path, and the trend of rising, falling or stable state is quantitatively evaluated in combination with the association data of health characteristics. A detailed trend evaluation is performed on the key nodes in each evaluation cycle to obtain the overall evaluation data of the group health changes. The output health trend evaluation data not only includes the trend of group health, but also clearly marks the risk increase interval and health improvement interval, providing an accurate trend prediction basis for subsequent health management decisions.

[0053] Preferably, step S4 comprises the following steps: Step S41: normalizing the health trend assessment data to obtain health trend assessment normalized data; Step S42: Correcting the user's health behavior based on the health trend assessment normalized data to obtain health behavior correction data; Step S43: Push the health behavior correction data based on the cloud platform to perform user health data management.

[0054] In an embodiment of the present invention, when the health trend assessment data is normalized, the dimension differences between different health indicators are eliminated through data standardization operations, so that the indicators can be compared and analyzed at the same scale. The specific steps include normalizing the numerical range of each health feature in the health trend assessment data, and using the minimum-maximum scaling method to convert each feature value to the interval of [0, 1]. In the operation, the maximum and minimum values ​​of each health feature are first calculated, and each data point is scaled using a formula to maintain the integrity of the relative change trend of the data. During the normalization process, the continuity and consistency of the data are strictly checked to ensure that the processed data can be directly used for the analysis of subsequent health behavior correction. The output health trend assessment normalized data is the balanced processing result of different health features, which provides standardized basic data for the next step of health behavior correction, and is convenient for identifying health anomalies and improvement priorities. When correcting the user's health behavior based on the health trend assessment normalized data, the part of the user's behavior that deviates from the health standard is identified through comparative analysis with the established health behavior norms, and a corresponding correction plan is formulated. In the specific implementation, the normalized data is first mapped to the preset health behavior criteria, and the deviation detection algorithm is used to accurately evaluate the gap between the user's current behavior status and the ideal health status. For the identified health risk behaviors, the corrective measures in the health intervention strategy library are matched to gradually adjust the user's health behavior. During the correction process, the implementation of each corrective measure and its actual impact on health indicators are recorded. The core of the operation lies in accurate correction. By gradually optimizing the user's health behavior habits, the output health behavior correction data records the process and results of behavior adjustment in detail, laying a practical foundation for further management and data push of the cloud platform. When the health behavior correction data is pushed based on the cloud platform to perform user health data management, the health behavior correction data is first uploaded to the health management module of the cloud platform, and the data is matched and analyzed according to the user's personalized health profile. During the push process, the correction suggestions, health behavior adjustment prompts and subsequent monitoring requirements are pushed to the user end in real time in the form of a combination of pictures and texts through the preset data interface. The health management module tracks and collects feedback on the pushed correction information in real time to evaluate the effectiveness of the push measures. The operation steps include setting up the data push protocol, dynamically generating push content, and real-time feedback of feedback data, ensuring that the correction data is managed in a closed loop during the execution process. The interactive push system of the cloud platform can continue to guide and adjust the user's health behavior after correction, ultimately achieving long-term optimization of health behavior and dynamic management of health data.

[0055] Preferably, the present invention also provides a user health data management system based on a cloud platform, which is used to execute the user health data management method based on a cloud platform as described above, and the user health data management system based on a cloud platform includes: Outpatient clinic classification data, used to obtain historical physical examination report data uploaded by users; historical physical examination report data is classified into the types of outpatient clinics to which the contents belong, and the outpatient clinic type data to which the physical examination contents belong is obtained; The outpatient clinic association analysis module is used to perform an outpatient clinic association overlap analysis on the historical physical examination report data according to the outpatient clinic type data to which the physical examination content belongs, and obtain the outpatient clinic association overlap data to which the content belongs; a content association retrieval library is constructed for the outpatient clinic association overlap data to obtain a content association intermediate layer fragmentation retrieval library; The health trend analysis module is used to identify health risk factors based on the content-related intermediate layer fragment retrieval library to obtain health risk factor data; perform health trend assessment based on the health risk factor data to obtain health trend assessment data; The user health data management module is used to correct the user's health behavior based on the health trend assessment data to obtain the health behavior correction data; and push the health behavior correction data based on the cloud platform to perform user health data management.

[0056] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0057] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A user health data management method based on cloud platform, It is characterized in that The following steps are involved: Step S1: Obtain historical physical examination report data uploaded by the user; classify the historical physical examination report data into the type of clinic to which the content belongs, and obtain the type of clinic to which the physical examination content belongs; Step S2: performing overlapping analysis of clinic-related contents on the historical physical examination report data according to the clinic-related types of the physical examination contents, and obtaining overlapping clinic-related contents; Construct a content-related retrieval library for the overlapping data of the clinics to which the content belongs, and obtain a content-related intermediate layer fragmented retrieval library; Step S3: Identify health risk factors according to the content-related intermediate layer shard retrieval library to obtain health risk factor data; Perform health trend assessment based on health risk factor data to obtain health trend assessment data; Step S4: Correcting the user's health behavior based on the health trend assessment data to obtain health behavior correction data; The health behavior correction data is pushed based on the cloud platform to perform user health data management.

2. The user health data management method based on the cloud platform according to claim 1, It is characterized in that Step S1 includes the following steps: Step S11: Obtain historical physical examination report data uploaded by the user; Step S12: Clean the historical physical examination report data to obtain cleaned data of the historical physical examination report; Step S13: labeling the cleaned data of the historical physical examination report to obtain labeled data of the historical physical examination report; Step S14: Classify the labeled data of historical physical examination reports according to the type of clinic to which the contents belong, and obtain data on the type of clinic to which the physical examination contents belong.

3. The user health data management method based on the cloud platform according to claim 1, It is characterized in that Step S2 includes the following steps: Step S21: performing overlapping analysis of clinic-related contents on the historical physical examination report data according to the clinic-related types of the physical examination contents, and obtaining overlapping clinic-related contents; Step S22: evaluating the correlation complexity of the overlapping data of the clinics to which the content belongs, and obtaining the correlation complexity data of the clinics to which the content belongs; Step S23: constructing a content-related search library for the overlapping data of the clinics to which the content belongs according to the complexity data of the clinics to which the content belongs, and obtaining a content-related search library; Step S24: vertically slice the content-related middle layer of the content-related search library to obtain a content-related middle layer sliced ​​search library.

4. The user health data management method based on the cloud platform according to claim 3, It is characterized in that Step S23 includes the following steps: Step S231: performing overlapping area aggregation analysis on the overlapping data associated with the clinics to which the content belongs, and obtaining content overlapping area aggregation data; Step S232: constructing a content structure hierarchical network diagram according to the content overlap area aggregation data and the clinic association complexity data to which the content belongs, to obtain a content structure hierarchical network diagram; Step S233: performing factor correlation derivation analysis on the content structure hierarchical network diagram to obtain content factor correlation derivation data; Step S234: designing a hierarchical bubble priority index for the content structure hierarchical network diagram according to the content factor association derivation data to obtain a content hierarchical bubble priority index; Step S235: performing index balance optimization on the content-level bubble priority index through the AVL tree to obtain a content-level bubble priority balanced index; Step S236: constructing a content-related search library for the overlapping data related to the outpatient clinic to which the content belongs according to the content-level bubble priority balance index and the content factor-related derivation data, and obtaining a content-related search library.

5. The user health data management method based on the cloud platform according to claim 3, It is characterized in that Step S24 includes the following steps: Step S241: performing search semantic domain distribution analysis on the content-related search library to obtain search semantic domain distribution data; Step S242: performing context dependency sparsity analysis according to the retrieval semantic domain distribution data to obtain retrieval semantic context dependency sparsity data; Step S243: performing multi-directional reference retrieval semantic extraction on the retrieval semantic domain distribution data according to the retrieval semantic context dependency sparsity data to obtain multi-directional reference retrieval semantic data; Step S244: dividing the search semantic domain distribution data into search content intermediate levels based on the multi-directional reference search semantic data to obtain search content intermediate level data; Step S245: perform vertical slicing of the content middle layer according to the multi-directional reference retrieval semantic data and the retrieval content middle-level data to obtain a content-related middle-layer slicing retrieval library.

6. The user health data management method based on the cloud platform according to claim 1, It is characterized in that Step S3 includes the following steps: Step S31: searching for group basic physical data of users based on historical physical examination report data to obtain group basic physical data of users; Step S32: identifying health risk factors of the user's basic group data based on the content-related intermediate layer fragment retrieval library to obtain health risk factor data; Step S33: Perform health trend assessment on the user's basic group data based on the health risk factor data to obtain health trend assessment data.

7. The user health data management method based on the cloud platform according to claim 6, It is characterized in that Step S32 includes the following steps: Step S321: perform a physical examination evaluation content similarity assessment on the user's physical basic group data according to the content-related intermediate layer segmented retrieval library to obtain physical examination evaluation content similarity data; Step S322: performing cluster analysis on the physical examination evaluation content similarity data to obtain physical examination evaluation content cluster data; Step S323: reconstructing the individual health trajectory according to the physical examination evaluation content similarity data and the physical examination evaluation content clustering data to obtain individual health trajectory reconstruction data; Step S324: using a preset individual risk identification model to identify individual risk characteristic factors of the individual health trajectory reconstruction data and the physical examination evaluation content clustering data, and obtaining individual risk characteristic factor data; Step S325: performing multivariate sensitivity analysis on individual risk characteristic factor data to obtain risk multivariate sensitivity data; Step S326: Identify health risk factors based on individual risk characteristic factor data and risk multivariate sensitivity data to obtain health risk factor data.

8. The user health data management method based on the cloud platform according to claim 6, It is characterized in that Step S33 includes the following steps: Step S331: Perform dynamic cross-feature correlation analysis on the health risk factor data to obtain dynamic health feature correlation data; Step S332: performing trend path analysis on the user's basic group data based on the dynamic health feature association data to obtain health trend path data; Step S333: Perform health trend assessment on the user's basic group data based on the dynamic health feature association data and the health trend path data to obtain health trend assessment data.

9. The user health data management method based on the cloud platform according to claim 1, It is characterized in that Step S4 includes the following steps: Step S41: normalizing the health trend assessment data to obtain health trend assessment normalized data; Step S42: Correcting the user's health behavior based on the health trend assessment normalized data to obtain health behavior correction data; Step S43: Push the health behavior correction data based on the cloud platform to perform user health data management.

10. A user health data management system based on cloud platform, It is characterized in that Used to execute the user health data management method based on the cloud platform as claimed in claim 1, the user health data management system based on the cloud platform comprises: Outpatient clinic classification data, used to obtain historical physical examination report data uploaded by users; historical physical examination report data is classified into the types of outpatient clinics to which the contents belong, and the outpatient clinic type data to which the physical examination contents belong is obtained; The outpatient clinic association analysis module is used to perform an outpatient clinic association overlap analysis on the historical physical examination report data according to the outpatient clinic type data to which the physical examination content belongs, and obtain the outpatient clinic association overlap data to which the content belongs; a content association retrieval library is constructed for the outpatient clinic association overlap data to obtain a content association intermediate layer fragmentation retrieval library; The health trend analysis module is used to identify health risk factors based on the content-related intermediate layer fragment retrieval library to obtain health risk factor data; perform health trend assessment based on the health risk factor data to obtain health trend assessment data; The user health data management module is used to correct the user's health behavior based on the health trend assessment data to obtain the health behavior correction data; and push the health behavior correction data based on the cloud platform to perform user health data management.

Citation Information

Patent Citations

  • Medical information intelligent management system based on cloud platform

    CN118213057A

  • Internet-based physical examination report data analysis and management system

    CN118471420A

  • Health managing method and storage medium

    US20220037002A1

Cited By

  • User health intervention feature management system and method based on multi-source data

    CN120977473A