Dynamic response type cloud platform for chronic disease self-management

Abstract

The invention provides a dynamic response type cloud platform for chronic disease self-management, and relates to the technical field of artificial intelligence. Comprising the steps that physiological data of a user are collected in real time through a data collection module, health management suggestions are generated through a data analysis module, medical literatures and user feedback are integrated in combination with a knowledge graph module, and a personalized health management suggestion report is provided. The user interaction module realizes real-time data display and feedback, the feedback analysis module generates a user portrait, the data packaging module generates packaged data, and the large model reasoning module quickly updates a health management suggestion report. And the cloud infrastructure module ensures data storage security and access control, so that the health management experience of the user is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of artificial intelligence, in particular to a dynamic response cloud platform for chronic disease self-management. BACKGROUND

[0002] In recent years, with the rapid development of information technology, especially the widespread popularity of wearable devices and cloud computing technology, the field of chronic disease management has undergone significant changes. Wearable devices such as smart watches, health bands, etc. can monitor users' physiological indicators in real time, including heart rate, blood pressure, blood sugar, and other key health data. The widespread use of such devices enables chronic disease patients to easily obtain real-time feedback on their health status in daily life. At the same time, these devices are usually equipped with corresponding mobile applications to transmit data to users' smartphones or computers, not only improving the convenience of data acquisition, but also encouraging patients' initiative in self-health management.

[0003] With the development of technology, cloud computing technology provides strong data support for chronic disease management. Cloud platforms can store multi-source health data and support efficient analysis and processing of data. The convenience of accessing data through the cloud greatly enhances cross-platform health management capabilities, allowing users to view historical records and health reports anytime, anywhere. In addition, cloud computing also supports secure data transmission and backup, ensuring user privacy and data security, thereby enhancing user trust and reliance on technology.

[0004] Although existing technologies have made progress in data collection and cloud storage, there are still many shortcomings in practical application. First, many wearable devices and health applications lack the ability to provide personalized health management. Current systems often provide static health recommendations, but cannot dynamically adjust to changes in user health status. This neglect of individual differences can limit the effectiveness of health management programs. For example, for diabetic patients with fluctuating blood sugar levels, fixed medication reminders and lifestyle recommendations often cannot meet their actual needs, so dynamic response capabilities are particularly important.

[0005] Secondly, the data compatibility between existing devices and applications is low, often requiring users to use multiple applications to manage different health data such as blood pressure, weight, and exercise volume. The complexity of this operation makes it difficult for users to take advantage of these technologies, leading to data homogenization and information silos, making it difficult for users to obtain a comprehensive view of health management.

[0006] In addition, the privacy concerns of users remain an important obstacle to the promotion of chronic disease management technology. Although existing systems generally take various security measures in data transmission and storage, in many cases, users still lack sufficient trust in how their health data is stored and utilized. This can lead some users to be reluctant to actively monitor their health data due to concerns about privacy breaches, thereby affecting self-management effectiveness. SUMMARY

[0007] In order to overcome the shortcomings of the prior art, the purpose of the present application is to provide a dynamic response type cloud platform for chronic disease self-management. The present application solves the problems of lack of dynamic response capability, personalized service and data compatibility in the prior art chronic disease management technology, which leads to complex operation and insufficient privacy protection in health management for users.

[0008] To achieve the above purpose, the present application provides the following scheme:

[0009] A dynamic response type cloud platform for chronic disease self-management, comprising:

[0010] A data collection module for collecting real-time physiological data of a user by connecting an intelligent wearable device;

[0011] A data analysis module for performing time series analysis and multi-index correlation analysis on the physiological data using a constructed machine learning model to generate health management recommendations;

[0012] A knowledge graph module for integrating medical literature, clinical guidelines and user feedback information to construct a health knowledge graph, and using the constructed health knowledge graph to perform semantic reasoning on the physiological data and the health management recommendations to generate an individualized health management recommendation report, the individualized health management recommendation report including a first health status assessment result and at least one intervention measure;

[0013] A user interaction module for designing a user interface to allow the user to view their own physiological data, historical records and individualized health management recommendation reports in real time, and to receive user feedback information on the health management recommendation reports, the user interface dynamically adjusting the display content according to the user type;

[0014] A feedback analysis module for extracting multi-dimensional features of the feedback information to generate a specific character profile;

[0015] A data packaging module for packaging the generated specific character profile and real-time physiological data to generate packaged data;

[0016] The large model inference module is used to receive packaged data and the personalized health management suggestion report at the current moment, and to perform rapid inference in the large model to obtain an updated health management suggestion report. The updated health management suggestion report includes personalized intervention measures and second health status assessment results generated based on real-time physiological data and user profiles.

[0017] The cloud infrastructure module is used to store user physiological data and health management advice reports using transport layer encryption protocols and data at rest encryption algorithms, and to manage data access permissions based on role-based access control policies.

[0018] Preferably, it further includes:

[0019] The health community module, connected to the data analysis module and the user interaction module, is used to build a health user community to enable users to exchange experiences and generate exchange experience data to be transmitted to the data analysis module to optimize the personalized health management suggestion report.

[0020] Preferably, the physiological data includes:

[0021] Heart rate, blood pressure, blood sugar, body temperature, and activity level.

[0022] Preferably, the data analysis module includes:

[0023] The data preprocessing submodule is used to perform noise removal, missing value imputation and normalization on the physiological data to obtain preprocessed physiological data.

[0024] The time series analysis submodule is used to perform time series analysis on the processed physiological data using the constructed machine learning model, and to extract the user's health change trend sequence data.

[0025] The correlation analysis submodule is used to analyze the processed physiological data based on the relationships between multiple indicators to obtain correlation analysis data.

[0026] It is recommended to generate a submodule to combine correlation analysis data and the user's health change trend sequence data to generate health management suggestions;

[0027] The expression for the machine learning model is:

[0028] ;

[0029] in, For example, predictive health indicators; The intercept of the model; For each of the characteristics, the regression coefficients are... , corresponding to features ; For input physiological data; This is the error term.

[0030] Preferably, the user interaction module includes:

[0031] The user interface design submodule is used to design a user-friendly user interface, which includes a user login interface, a main operation interface, and a settings interface.

[0032] The data visualization submodule is used to visually display real-time collected physiological data in the form of charts and tables;

[0033] The Health Management Recommendations Display submodule is used to display personalized health management recommendations reports;

[0034] The user feedback collection submodule is used by users to evaluate personalized health management advice reports and data displays, and allows users to input free text feedback.

[0035] Preferably, the feedback analysis module includes:

[0036] The user feedback collection submodule is used to collect user feedback on the health management advice report, including user satisfaction ratings, suggestions, and changes in health behaviors;

[0037] The feature extraction submodule is used to extract multi-dimensional features from feedback information using natural language processing technology and sentiment analysis algorithms. The multi-dimensional features include: behavioral features, preference features, physiological response features, and disease features.

[0038] The user profile building submodule is used to generate specific user profiles through a time-series fusion algorithm.

[0039] The expression for the specific portrait is:

[0040] ;

[0041] in, For the final generated portrait of a specific character; The number of features included; The user's historical feedback feature vector; This is for extracting multidimensional feature vectors, such as behavioral features and preference features; Characteristics of changes in users' health behaviors; These are the weight coefficients for historical feedback features, extracted features, and health behavior features, respectively, used to adjust the influence of each feature in the generated profile; This is a vector of contextual information, such as the user's environment and emotional attitude; The weighting coefficients for contextual information are used to control their influence in the final portrait of a specific person.

[0042] Preferably, the data packaging module includes:

[0043] The data preparation submodule is used to extract target information from real-time physiological data and user profiles. The target information includes user ID, timestamp, health indicators and user profile features.

[0044] The data formatting submodule is used to format the target information, convert it into a unified data structure, and obtain formatted data.

[0045] The data encryption submodule is used to encrypt the formatted data to obtain encrypted data;

[0046] The data packaging submodule is used to package the encrypted data and corresponding metadata into a data packet to obtain the packaged data.

[0047] Preferably, the large model inference module includes:

[0048] The model loading submodule is used to load large pre-trained models from the local server.

[0049] The inference request construction submodule is used to construct an inference request based on the packaged data, including specifying the inference task and setting inference parameters;

[0050] The large model inference submodule is used to send inference requests to the inference engine, execute large model inference, and obtain prediction results;

[0051] The results post-processing submodule is used to post-process the prediction results of the large model to obtain an updated health management recommendation report.

[0052] The output feedback submodule is used to send the updated health management advice report back to the user interaction module.

[0053] Preferably, the cloud infrastructure module includes:

[0054] The data collection submodule is used to configure the transport layer encryption protocol to collect user physiological data and health management recommendation reports;

[0055] The data storage submodule is used to encrypt the stored user physiological data and health management suggestion reports using a static data encryption algorithm;

[0056] The role definition submodule is used to create clearly defined user roles and grant specific access permissions to various roles according to preset requirements;

[0057] The access control policy configuration submodule is used to implement role-based access control policies according to role definitions;

[0058] The data access request processing submodule is used to verify the user's identity when the user requests access to data stored in the cloud, and to decide whether to allow the access request based on the user's role and access permissions.

[0059] A dynamic, responsive approach to chronic disease self-management includes:

[0060] By connecting to smart wearable devices, users' physiological data can be collected in real time.

[0061] By using a pre-built machine learning model, time series analysis and multi-indicator correlation analysis are performed on physiological data to generate health management recommendations.

[0062] By integrating medical literature, clinical guidelines and user feedback, a health knowledge graph is constructed. The constructed health knowledge graph is then used to perform semantic reasoning on the physiological data and health management recommendations to generate a personalized health management recommendation report. The personalized health management recommendation report includes: the first health status assessment result and at least one intervention measure.

[0063] The user interface is designed to allow users to view their physiological data, historical records, and personalized health management advice reports in real time, and to receive user feedback on the health management advice reports. The user interface dynamically adjusts the displayed content according to the user type.

[0064] Extract the multidimensional features of the feedback information to generate a specific person profile;

[0065] The generated specific person portrait and real-time physiological data are packaged together to generate packaged data;

[0066] The system receives packaged data and a personalized health management recommendation report for the current moment, and performs rapid reasoning in a large model to obtain an updated health management recommendation report. The updated health management recommendation report includes personalized intervention measures and second health status assessment results generated based on real-time physiological data and user profiles.

[0067] The system utilizes transport layer encryption protocols and data at rest encryption algorithms to store user physiological data and health management advice reports, and manages data access permissions based on role-based access control policies.

[0068] The present invention discloses the following technical effects:

[0069] This invention provides a dynamic, responsive cloud platform for chronic disease self-management, comprising: a data collection module for collecting users' physiological data in real time via a connected smart wearable device; a data analysis module for performing time-series analysis and multi-indicator correlation analysis on the physiological data using a pre-constructed machine learning model to generate health management recommendations; a knowledge graph module for integrating medical literature, clinical guidelines, and user feedback to construct a health knowledge graph, and using the constructed health knowledge graph to perform semantic reasoning on the physiological data and health management recommendations to generate a personalized health management recommendation report, the personalized health management recommendation report including: a first health status assessment result and at least one intervention measure; and a user interaction module for designing a user interface that allows users to view their physiological data, historical records, and personalized health management recommendation reports in real time. The system integrates data collection, analysis, knowledge graph construction, user interaction, feedback analysis, packaging, and large-model inference modules to provide a dynamic and responsive health management solution. The user interface dynamically adjusts its display content based on user type. A feedback analysis module extracts multi-dimensional features from the feedback to generate a specific user profile. A data packaging module packages the generated user profile and real-time physiological data to create packaged data. A large-model inference module receives the packaged data and the current personalized health management suggestion report, performs rapid inference within a large model, and obtains an updated health management suggestion report. This updated report includes personalized intervention measures and a second health status assessment result generated based on real-time physiological data and the user profile. A cloud infrastructure module stores user physiological data and health management suggestion reports using transport layer encryption protocols and static data encryption algorithms, and manages data access permissions based on role-based access control policies. This invention provides a dynamic and responsive health management solution by integrating modules for data collection, analysis, knowledge graph construction, user interaction, feedback analysis, packaging, and large-model inference. The platform can collect and analyze user physiological data in real time, combine personalized health management suggestions and user feedback to generate accurate health status assessments and intervention measures, thereby improving the self-management ability and health level of patients with chronic diseases. In addition, data security is ensured by using transport layer encryption and data encryption algorithms, and data access permissions are managed through role-based access control policies, which effectively protects user privacy and enhances user trust. Attached Figure Description

[0070] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0071] Figure 1This is a schematic diagram of a dynamic and responsive cloud platform structure for chronic disease self-management provided in an embodiment of the present invention.

[0072] Explanation of reference numerals in the attached figures:

[0073] 1-Data collection module, 2-Data analysis module, 3-Knowledge graph module, 4-User interaction module, 5-Feedback analysis module, 6-Data packaging module, 7-Large model inference module, 8-Cloud infrastructure module. Detailed Implementation

[0074] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0075] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0076] like Figure 1 As shown, this invention provides a dynamic, responsive cloud platform for chronic disease self-management, comprising:

[0077] Data collection module 1 is used to collect users' physiological data in real time by connecting to smart wearable devices;

[0078] Data analysis module 2 is used to perform time series analysis and multi-indicator correlation analysis on physiological data using the constructed machine learning model to generate health management recommendations;

[0079] Knowledge graph module 3 is used to integrate medical literature, clinical guidelines and user feedback information to construct a health knowledge graph. The constructed health knowledge graph is used to perform semantic reasoning on the physiological data and the health management suggestions to generate a personalized health management suggestion report. The personalized health management suggestion report includes: the first health status assessment result and at least one intervention measure.

[0080] User interaction module 4 is used to design the user interface, allowing users to view their physiological data, historical records and personalized health management advice reports in real time, and to receive user feedback on the health management advice reports. The user interface dynamically adjusts the displayed content according to the user type.

[0081] Feedback analysis module 5 is used to extract multidimensional features of the feedback information to generate a specific person profile;

[0082] Data packaging module 6 is used to package the generated specific person portrait and real-time physiological data to generate packaged data;

[0083] The large model inference module 7 is used to receive packaged data and the personalized health management suggestion report at the current moment, and to perform rapid inference in the large model to obtain an updated health management suggestion report. The updated health management suggestion report includes personalized intervention measures and second health status assessment results generated based on real-time physiological data and user profiles.

[0084] Cloud Infrastructure Module 8 is used to store user physiological data and health management advice reports using transport layer encryption protocols and data at rest encryption algorithms, and to manage data access permissions based on role-based access control policies.

[0085] Furthermore, it also includes:

[0086] The health community module, connected to the data analysis module 2 and the user interaction module 4, is used to build a health user community to enable users to exchange experiences and generate exchange experience data to be transmitted to the data analysis module 2 to optimize the personalized health management suggestion report.

[0087] Specifically, the health community module in this embodiment, through its connection with the data analysis module 2 and the user interaction module 4, can construct an integrated health user community platform. On this platform, users can freely share their personal health management experiences and success stories, which not only provides users with valuable reference materials but also creates a mutually supportive and encouraging social environment. The exchange of experiences generated during user interaction can be transformed into valuable information, further enriching the content of the backend data analysis.

[0088] Based on user feedback and experience sharing, the health community module in this embodiment regularly transmits the collected shared experience data to the data analysis module 2. This process ensures the continuous optimization and updating of personalized health management advice reports. By integrating actual feedback and experience data from community users, health management advice can better align with users' actual needs and behavioral patterns, thereby improving the effectiveness and applicability of the advice.

[0089] To enhance user interaction and engagement, the health community module in this embodiment features diverse communication functions, including topic discussions, experience sharing, and health activity organization. Through these functions, users can not only receive personalized health management advice but also gain emotional support and practical assistance through these interactions. This approach, combining community-based experience sharing with data analysis, not only enhances user engagement but also provides a more comprehensive and accurate basis for intelligent health management, ultimately promoting better self-management for patients with chronic diseases.

[0090] Furthermore, the physiological data includes:

[0091] Heart rate, blood pressure, blood sugar, body temperature, and activity level.

[0092] Furthermore, the data analysis module 2 includes:

[0093] The data preprocessing submodule is used to perform noise removal, missing value imputation and normalization on the physiological data to obtain preprocessed physiological data.

[0094] The time series analysis submodule is used to perform time series analysis on the processed physiological data using the constructed machine learning model, and to extract the user's health change trend sequence data.

[0095] The correlation analysis submodule is used to analyze the processed physiological data based on the relationships between multiple indicators to obtain correlation analysis data.

[0096] It is recommended to generate a submodule to combine correlation analysis data and the user's health change trend sequence data to generate health management suggestions;

[0097] The expression for the machine learning model is:

[0098] ;

[0099] in, For example, predictive health indicators; The intercept of the model; For each of the characteristics, the regression coefficients are... , corresponding to features ; For input physiological data; This is the error term.

[0100] Specifically, to realize the function of data analysis module 2, this embodiment first performs noise reduction, missing value imputation, and normalization on the collected physiological data through the data preprocessing submodule to obtain high-quality preprocessed physiological data. This process ensures the accuracy and consistency of the data, making the next step of analysis more reliable. Noise reduction refers to removing noise introduced by sensor errors, while missing value imputation involves reasonably estimating the missing items in the data through interpolation or other algorithms. Normalization involves scaling the data proportionally to make it comparable on the same scale.

[0101] The subsequent time series analysis submodule utilizes the constructed machine learning model to perform time series analysis on the processed physiological data, extracting the user's health change trend sequence data. Time series analysis is a statistical method designed to identify patterns in data changes over time. Through the user's historical physiological data, the model can identify long-term trends, seasonal variations, and sudden events. The user's health change trend is based on the performance of multidimensional physiological indicators (such as heart rate, blood pressure, blood sugar, body temperature, and activity level) over time, providing a data foundation for subsequent health management recommendations.

[0102] The correlation analysis submodule extracts interrelated health indicators from the physiological data processed by multi-indicator relationship analysis and generates health management suggestions by combining them with the aforementioned user health change trend sequence data. This suggestion generation submodule uses the analyzed correlation data to provide personalized health management suggestions for the user's status, such as recommending specific foods based on changes in heart rate and blood sugar. Furthermore, the machine learning model's expression creates predictions for health indicators by weighted summing of multiple features from the input physiological data. The regression coefficients of these features reflect the importance of each feature to changes in health indicators, while the model's intercept is the predicted value of health indicators in the absence of feature input. Through these analyses, users can obtain practical health management suggestions to help them improve their lifestyle and health status.

[0103] Furthermore, the user interaction module 4 includes:

[0104] The user interface design submodule is used to design a user-friendly user interface, which includes a user login interface, a main operation interface, and a settings interface.

[0105] The data visualization submodule is used to visually display real-time collected physiological data in the form of charts and tables;

[0106] The Health Management Recommendations Display submodule is used to display personalized health management recommendations reports;

[0107] The user feedback collection submodule is used by users to evaluate personalized health management advice reports and data displays, and allows users to input free text feedback.

[0108] Specifically, to achieve the functions of data analysis module 2, this embodiment employs specific steps to process the collected physiological data. The main task of the data preprocessing submodule is to ensure the quality of the raw data. First, noise reduction is performed, using filtering algorithms to remove invalid data introduced by sensor errors, environmental interference, etc., thus cleaning the user's physiological data. Next, missing value imputation is performed, using linear interpolation or the nearest neighbor method to reasonably estimate missing data, ensuring the integrity of the data used. Finally, normalization is applied to transform physiological data of different magnitudes to the same scale, facilitating subsequent data analysis and comparison operations. These processing steps ensure data quality, laying a solid foundation for accurate data analysis.

[0109] This embodiment utilizes a time series analysis submodule for in-depth data analysis. This submodule employs the pre-built machine learning model to analyze pre-processed physiological data, focusing on extracting the user's health trend sequence data. By performing time series analysis on multiple dimensions such as heart rate, blood pressure, blood sugar, body temperature, and activity level, the model can automatically identify long-term trends, seasonal variations, and abnormal states in health status. These health trends will help understand the user's health status at different time periods, providing necessary data support for generating personalized health management recommendations.

[0110] The correlation analysis submodule will perform multi-indicator relationship analysis on physiological data based on the aforementioned user health change trends. In this step, the module will extract the interrelationships between various health indicators and generate health management suggestions. These suggestions are not only based on the identified trends but also consider the user's individual circumstances. For example, when the model detects a correlation between a user's heart rate and blood sugar levels, it may suggest that the user adjust their dietary habits. By combining correlation analysis with user health change trends, this embodiment can provide users with scientific and reasonable personalized health management suggestions, thereby supporting users' health decisions and lifestyle improvements.

[0111] Furthermore, the feedback analysis module 5 includes:

[0112] The user feedback collection submodule is used to collect user feedback on the health management advice report, including user satisfaction ratings, suggestions, and changes in health behaviors;

[0113] The feature extraction submodule is used to extract multi-dimensional features from feedback information using natural language processing technology and sentiment analysis algorithms. The multi-dimensional features include: behavioral features, preference features, physiological response features, and disease features.

[0114] The user profile building submodule is used to generate specific user profiles through a time-series fusion algorithm.

[0115] The expression for the specific portrait is:

[0116] ;

[0117] in, For the final generated portrait of a specific character; The number of features included; The user's historical feedback feature vector; This is for extracting multidimensional feature vectors, such as behavioral features and preference features; Characteristics of changes in users' health behaviors; These are the weight coefficients for historical feedback features, extracted features, and health behavior features, respectively, used to adjust the influence of each feature in the generated profile; This is a vector of contextual information, such as the user's environment and emotional attitude; The weighting coefficients for contextual information are used to control their influence in the final portrait of a specific person.

[0118] Specifically, to implement the functions of feedback analysis module 5, this embodiment designs a user feedback collection submodule, which is specifically used to collect user feedback information on health management suggestion reports. This feedback information includes user satisfaction ratings, personal opinions and suggestions, and changes in health behaviors. User satisfaction ratings reflect the user's acceptance and satisfaction level with the suggestions; opinions and suggestions provide the user's intuitive understanding of the suggestion content; and changes in health behaviors record the actual changes the user experiences after implementing the suggestions, reflecting the effectiveness of the suggestions. Through the collection and organization of the above feedback information, the feature extraction submodule uses natural language processing technology and sentiment analysis algorithms to conduct in-depth analysis of the collected feedback information to extract multi-dimensional features. These multi-dimensional features include behavioral features, preference features, physiological response features, and disease features. Behavioral features focus on the user's health behavior, such as exercise frequency and dietary habits; preference features record the user's personalized preferences, such as the degree of liking for certain healthy foods; physiological response features focus on the user's physiological changes after accepting the suggestions; and disease features are characteristics specific to the user's known health conditions. By analyzing these features, the user profile construction submodule processes the extracted feature data using a time-series fusion algorithm to generate a specific user profile. This specific user profile integrates historical feedback feature vectors, extracted multidimensional feature vectors, and user health behavior change features to form a comprehensive representation of the user's health status. The historical feedback feature vectors contain the user's feedback records on health management suggestions over a period of time. The extracted multidimensional feature vectors are obtained through natural language processing and sentiment analysis as mentioned above, while the health behavior change features reflect the user's actual actions in implementing health suggestions. Each feature will have a corresponding weight coefficient to adjust its influence on the final user profile. Furthermore, the module will consider contextual information vectors, such as the user's environment and emotional state, and set corresponding weight coefficients for contextual information to ensure that the generated user profile comprehensively reflects the user's life background and psychological state, thereby improving the personalization and targeting of health management suggestions. Through the above steps, this embodiment can generate accurate and practical specific user profiles, helping to improve the effectiveness of user health management.

[0119] Furthermore, the data packaging module 6 includes:

[0120] The data preparation submodule is used to extract target information from real-time physiological data and user profiles. The target information includes user ID, timestamp, health indicators and user profile features.

[0121] The data formatting submodule is used to format the target information, convert it into a unified data structure, and obtain formatted data.

[0122] The data encryption submodule is used to encrypt the formatted data to obtain encrypted data;

[0123] The data packaging submodule is used to package the encrypted data and corresponding metadata into a data packet to obtain the packaged data.

[0124] Specifically, to implement the functions of data packaging module 6, this embodiment first designs a data preparation submodule, which is responsible for extracting target information from real-time physiological data and user profiles. Target information includes user ID, timestamp, health indicators, and user profile features. The user ID uniquely identifies each user, ensuring data can be associated with a specific user. The timestamp records the specific time the data was generated, allowing for tracking of the time series of health changes. Health indicators cover physiological parameters such as heart rate, blood pressure, and blood sugar, while user profile features provide important information about the user's personalized health status, laying the foundation for subsequent data processing.

[0125] The data display formatting submodule in this embodiment formats the extracted target information to ensure that the data conforms to uniform structural requirements. This step standardizes all target information by establishing certain data format specifications, making it consistent and identifiable. The formatted data can be effectively read and used by subsequent modules. In addition, this process also includes necessary data type checks and validity verification to ensure that the entered data meets expectations in terms of accuracy and consistency.

[0126] Immediately after formatting, the data encryption submodule encrypts the data. The purpose of data encryption is to protect user privacy and sensitive information, preventing unauthorized access and data leakage. This embodiment employs advanced encryption algorithms to ensure data security during storage and transmission. After encryption, the data packaging submodule packages the encrypted data along with corresponding metadata (such as data source, processing time, etc.) to form a complete data packet. This data packet facilitates subsequent storage, transmission, and analysis, ensuring the security and traceability of user health data throughout the process. Through the design and implementation of the above steps, the packaging module in this embodiment ensures data reliability, integrity, and security, further enhancing user trust and user experience within the entire health management platform.

[0127] Furthermore, the large model inference module 7 includes:

[0128] The model loading submodule is used to load large pre-trained models from the local server.

[0129] The inference request construction submodule is used to construct an inference request based on the packaged data, including specifying the inference task and setting inference parameters;

[0130] The large model inference submodule is used to send inference requests to the inference engine, execute large model inference, and obtain prediction results;

[0131] The results post-processing submodule is used to post-process the prediction results of the large model to obtain an updated health management recommendation report.

[0132] The output feedback submodule is used to feed back the updated health management advice report to the user interaction module 4.

[0133] Specifically, to implement the functionality of the large model inference module 7, this embodiment first designs a model loading submodule, which is used to load pre-trained large models from the local server. During this process, the module ensures that the loaded model version matches the needs of the data analysis, including the predictive capabilities of various health indicators. The purpose of this step is to prepare sufficient computational power and accuracy for the subsequent inference process and to lay the foundation for personalized health management recommendations. Furthermore, a model integrity check must be performed simultaneously during model loading to ensure that the loaded model is not damaged and to maintain the stability of its inference performance.

[0134] The inference request construction submodule in this embodiment constructs inference requests based on the packaged data. This process is transformed into a structured request that includes specifying the inference task and setting inference parameters. Through explicit task instructions, it ensures that the inference engine can execute the target prediction. For example, the inference task might involve predicting a user's health trends or evaluating the latest health management recommendations; the inference parameters can define various details, such as the accuracy requirements of the inference and the maximum number of iterations. This flexible request construction mechanism ensures that the inference process of a large model can be highly tailored to specific user needs.

[0135] After the inference request is sent to the inference engine, the large model inference submodule executes the specific inference process and obtains the prediction results. This submodule efficiently processes user inference requests and promptly returns the corresponding health management prediction results. Next, the result post-processing submodule processes the large model's prediction results to generate an updated health management recommendation report. This post-processing step may include multi-dimensional data visualization, data smoothing, and intelligent recommendation generation, ensuring that the final report is not only accurate but also easy for users to understand. Finally, the output feedback submodule feeds the updated health management recommendation report back to the user interaction module 4, enabling users to obtain the latest personalized health advice in real time, thereby further enhancing user engagement and experience. This entire process design ensures the efficiency of the inference phase and the practicality of the results.

[0136] Furthermore, the cloud infrastructure module 8 includes:

[0137] The data collection submodule is used to configure the transport layer encryption protocol to collect user physiological data and health management recommendation reports;

[0138] The data storage submodule is used to encrypt the stored user physiological data and health management suggestion reports using a static data encryption algorithm;

[0139] The role definition submodule is used to create clearly defined user roles and grant specific access permissions to various roles according to preset requirements;

[0140] The access control policy configuration submodule is used to implement role-based access control policies according to role definitions;

[0141] The data access request processing submodule is used to verify the user's identity when the user requests access to data stored in the cloud, and to decide whether to allow the access request based on the user's role and access permissions.

[0142] Specifically, to implement the functions of cloud infrastructure module 8, this embodiment first designs a data collection submodule. This module is specifically designed to configure transport layer encryption protocols to ensure the security of user physiological data and health management recommendation reports during the collection process. In this process, the module will encrypt the data through Secure Sockets Layer (SSL) or Transport Layer Security (TLS) protocols to ensure the confidentiality and integrity of the data during transmission. By employing these encryption protocols, data can be prevented from being stolen or tampered with by unauthorized third parties, thereby protecting user privacy and data security.

[0143] The data display and storage submodule in this embodiment is used to encrypt stored user physiological data and health management recommendation reports using a static data encryption algorithm. This process ensures comprehensive protection of user data during the storage phase. In specific implementations, the module can choose a symmetric encryption algorithm, such as Advanced Encryption Standard (AES), to encrypt the stored data to prevent data leakage inside or outside the data center. By encrypting the stored data, even if the data is illegally accessed on the storage medium, attackers cannot obtain valid user information, thereby enhancing the security of user data.

[0144] The role definition submodule in this embodiment plays a crucial role in the system's security framework. It establishes clearly defined user roles and assigns specific access permissions to various roles according to preset requirements. Roles can be further subdivided based on user identity and needs, such as administrators, healthcare providers, and ordinary users. Based on this, the access control policy configuration submodule implements role-based access control policies according to the role definitions, ensuring that users can only access the data they are authorized to access when accessing data stored in the cloud. The data access request processing submodule is responsible for verifying the user's identity and determining whether to allow the request based on their role and access permissions when a user requests access to data in the cloud. This entire mechanism ensures a high level of security and reliability for the system during data storage and access, effectively preventing the leakage and misuse of sensitive data.

[0145] A dynamic, responsive approach to chronic disease self-management includes:

[0146] By connecting to smart wearable devices, users' physiological data can be collected in real time.

[0147] By using a pre-built machine learning model, time series analysis and multi-indicator correlation analysis are performed on physiological data to generate health management recommendations.

[0148] By integrating medical literature, clinical guidelines and user feedback, a health knowledge graph is constructed. The constructed health knowledge graph is then used to perform semantic reasoning on the physiological data and health management recommendations to generate a personalized health management recommendation report. The personalized health management recommendation report includes: the first health status assessment result and at least one intervention measure.

[0149] The user interface is designed to allow users to view their physiological data, historical records, and personalized health management advice reports in real time, and to receive user feedback on the health management advice reports. The user interface dynamically adjusts the displayed content according to the user type.

[0150] Extract the multidimensional features of the feedback information to generate a specific person profile;

[0151] The generated specific person portrait and real-time physiological data are packaged together to generate packaged data;

[0152] The system receives packaged data and a personalized health management recommendation report for the current moment, and performs rapid reasoning in a large model to obtain an updated health management recommendation report. The updated health management recommendation report includes personalized intervention measures and second health status assessment results generated based on real-time physiological data and user profiles.

[0153] The system utilizes transport layer encryption protocols and data at rest encryption algorithms to store user physiological data and health management advice reports, and manages data access permissions based on role-based access control policies.

[0154] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0155] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A dynamic, responsive cloud platform for chronic disease self-management, characterized in that: include: The data collection module is used to collect users' physiological data in real time by connecting to smart wearable devices; The data analysis module is used to perform time series analysis and multi-indicator correlation analysis on physiological data using the built machine learning model to generate health management recommendations. The knowledge graph module is used to integrate medical literature, clinical guidelines and user feedback information to construct a health knowledge graph. The constructed health knowledge graph is used to perform semantic reasoning on the physiological data and the health management suggestions to generate a personalized health management suggestion report. The personalized health management suggestion report includes: the first health status assessment result and at least one intervention measure. The user interaction module is used to design the user interface, allowing users to view their physiological data, historical records, and personalized health management advice reports in real time, and to receive user feedback on the health management advice reports. The user interface dynamically adjusts the displayed content according to the user type. The feedback analysis module is used to extract multidimensional features from the feedback information to generate a specific person profile; The data packaging module is used to package the generated specific person portrait and real-time physiological data to generate packaged data. The large model inference module is used to receive packaged data and the personalized health management suggestion report at the current moment, and to perform rapid inference in the large model to obtain an updated health management suggestion report. The updated health management suggestion report includes personalized intervention measures and second health status assessment results generated based on real-time physiological data and user profiles. The cloud infrastructure module is used to store user physiological data and health management advice reports using transport layer encryption protocols and data at rest encryption algorithms, and to manage data access permissions based on role-based access control policies.

2. The dynamic responsive cloud platform for chronic disease self-management according to claim 1, characterized in that, Also includes: The health community module, connected to the data analysis module and the user interaction module, is used to build a health user community to enable users to exchange experiences and generate exchange experience data to be transmitted to the data analysis module to optimize the personalized health management suggestion report.

3. The dynamic responsive cloud platform for chronic disease self-management according to claim 1, characterized in that, The physiological data includes: Heart rate, blood pressure, blood sugar, body temperature, and activity level.

4. The dynamic responsive cloud platform for chronic disease self-management according to claim 1, characterized in that, The data analysis module includes: The data preprocessing submodule is used to perform noise removal, missing value imputation and normalization on the physiological data to obtain preprocessed physiological data. The time series analysis submodule is used to perform time series analysis on the processed physiological data using the constructed machine learning model, and to extract the user's health change trend sequence data. The correlation analysis submodule is used to analyze the processed physiological data based on the relationships between multiple indicators to obtain correlation analysis data. It is recommended to generate a submodule to combine correlation analysis data and the user's health change trend sequence data to generate health management suggestions; The expression for the machine learning model is: ； in, For example, predictive health indicators; The intercept of the model; For each of the characteristics, the regression coefficients are... , corresponding to features ; For input physiological data; This is the error term.

5. A dynamic, responsive cloud platform for chronic disease self-management according to claim 1, characterized in that, The user interaction module includes: The user interface design submodule is used to design a user-friendly user interface, which includes a user login interface, a main operation interface, and a settings interface. The data visualization submodule is used to visually display real-time collected physiological data in the form of charts and tables; The Health Management Recommendations Display submodule is used to display personalized health management recommendations reports; The user feedback collection submodule is used by users to evaluate personalized health management advice reports and data displays, and allows users to input free text feedback.

6. The dynamic responsive cloud platform for chronic disease self-management according to claim 1, characterized in that, The feedback analysis module includes: The user feedback collection submodule is used to collect user feedback on the health management advice report, including user satisfaction ratings, suggestions, and changes in health behaviors; The feature extraction submodule is used to extract multi-dimensional features from feedback information using natural language processing technology and sentiment analysis algorithms. The multi-dimensional features include: behavioral features, preference features, physiological response features, and disease features. The user profile building submodule is used to generate specific user profiles through a time-series fusion algorithm. The expression for the specific portrait is: ； in, For the final generated portrait of a specific character; The number of features included; The user's historical feedback feature vector; This is for extracting multidimensional feature vectors, such as behavioral features and preference features; Characteristics of changes in users' health behaviors; These are the weight coefficients for historical feedback features, extracted features, and health behavior features, respectively, used to adjust the influence of each feature in the generated profile; This is a vector of contextual information, such as the user's environment and emotional attitude; The weighting coefficients for contextual information are used to control their influence in the final portrait of a specific person.

7. A dynamic, responsive cloud platform for chronic disease self-management according to claim 1, characterized in that, The data packaging module includes: The data preparation submodule is used to extract target information from real-time physiological data and user profiles. The target information includes user ID, timestamp, health indicators and user profile features. The data formatting submodule is used to format the target information, convert it into a unified data structure, and obtain formatted data. The data encryption submodule is used to encrypt the formatted data to obtain encrypted data; The data packaging submodule is used to package the encrypted data and corresponding metadata into a data packet to obtain the packaged data.

8. A dynamic, responsive cloud platform for chronic disease self-management according to claim 1, characterized in that, The large model inference module includes: The model loading submodule is used to load large pre-trained models from the local server. The inference request construction submodule is used to construct an inference request based on the packaged data, including specifying the inference task and setting inference parameters; The large model inference submodule is used to send inference requests to the inference engine, execute large model inference, and obtain prediction results; The results post-processing submodule is used to post-process the prediction results of the large model to obtain an updated health management recommendation report. The output feedback submodule is used to send the updated health management advice report back to the user interaction module.

9. A dynamic, responsive cloud platform for chronic disease self-management according to claim 1, characterized in that, The cloud infrastructure module includes: The data collection submodule is used to configure the transport layer encryption protocol to collect user physiological data and health management recommendation reports; The data storage submodule is used to encrypt the stored user physiological data and health management suggestion reports using a static data encryption algorithm; The role definition submodule is used to create clearly defined user roles and grant specific access permissions to various roles according to preset requirements; The access control policy configuration submodule is used to implement role-based access control policies according to role definitions; The data access request processing submodule is used to verify the user's identity when the user requests access to data stored in the cloud, and to decide whether to allow the access request based on the user's role and access permissions.

10. A dynamic, responsive method for chronic disease self-management, characterized in that, include: By connecting to smart wearable devices, users' physiological data can be collected in real time. By using a pre-built machine learning model, time series analysis and multi-indicator correlation analysis are performed on physiological data to generate health management recommendations. By integrating medical literature, clinical guidelines and user feedback, a health knowledge graph is constructed. The constructed health knowledge graph is then used to perform semantic reasoning on the physiological data and health management recommendations to generate a personalized health management recommendation report. The personalized health management recommendation report includes: the first health status assessment result and at least one intervention measure. The user interface is designed to allow users to view their physiological data, historical records, and personalized health management advice reports in real time, and to receive user feedback on the health management advice reports. The user interface dynamically adjusts the displayed content according to the user type. Extract the multidimensional features of the feedback information to generate a specific person profile; The generated specific person portrait and real-time physiological data are packaged together to generate packaged data; The system receives packaged data and a personalized health management recommendation report for the current moment, and performs rapid reasoning in a large model to obtain an updated health management recommendation report. The updated health management recommendation report includes personalized intervention measures and second health status assessment results generated based on real-time physiological data and user profiles. The system utilizes transport layer encryption protocols and data at rest encryption algorithms to store user physiological data and health management advice reports, and manages data access permissions based on role-based access control policies.

Images