Health management scheme generation method, related device and computer storage medium

Abstract

The invention provides a health management scheme generation method, a related device and a computer storage medium. The method comprises the following steps: collecting physiological data, clinical data and input data of a user; determining a blood glucose fluctuation prediction trend of the user according to the physiological data and the input data; then, according to the clinical data, determining a current risk assessment result of the user; and finally, according to the blood glucose fluctuation prediction trend of the user, the current risk assessment result of the user, the target body index and the living habit, determining a health management scheme. Therefore, the intelligent level and accuracy of diabetes management are effectively improved, a more scientific, efficient and personalized health management scheme is provided for diabetic patients, the risk of diabetic complications is reduced, and the life quality of the patients is improved.

Description

Technical Field

[0001] The present application relates to the field of large model technology, and in particular to a method for generating a health management plan, related devices, and computer storage media. Background Art

[0002] Traditional diabetes management mainly relies on the experience of medical staff and simple data analysis tools, and intervenes through regular blood sugar testing, standardized diet and exercise plans, and fixed medication guidance.

[0003] However, this management model has significant flaws: on the one hand, individual patients have large differences in genes, lifestyle habits, disease progression, etc., and standardized plans are difficult to meet personalized needs; on the other hand, multi-source heterogeneous health data (such as real-time blood sugar fluctuations, dietary details, exercise intensity, etc.) cannot be fully integrated and deeply analyzed, resulting in a lack of timeliness and accuracy in the prediction and intervention of the disease. Summary of the Invention

[0004] In view of this, the present application provides a method for generating a health management plan, related devices and computer storage media, which effectively improve the intelligence level and accuracy of diabetes management, provide diabetic patients with a more scientific, efficient and personalized health management plan, help reduce the risk of diabetic complications, and improve the quality of life of patients.

[0005] The first aspect of the present application provides a method for generating a health management plan, comprising:

[0006] Collecting user data; wherein the user data includes: physiological data, clinical data and input data;

[0007] Determining a predicted trend of blood sugar fluctuation of the user based on the physiological data and the recorded data;

[0008] determining a current risk assessment result of the user based on the clinical data;

[0009] A health management plan is determined based on the user's predicted blood sugar fluctuation trend, the user's current risk assessment results, target physical indicators and living habits.

[0010] Optionally, determining the predicted trend of the user's blood sugar fluctuation based on the physiological data and the input data includes:

[0011] Preprocessing the physiological data and input data to obtain a first data set;

[0012] Performing fusion feature extraction on the data in the first data set to obtain a first fusion feature of the user;

[0013] The second fusion feature of the user is input into a prediction model, and the predicted trend of the user's blood sugar fluctuation is output; wherein the prediction model includes a trained large language model and a fusion model of a time recurrent network.

[0014] Optionally, determining a current risk assessment result of the user based on the clinical data includes:

[0015] Preprocessing the clinical data to obtain a second data set;

[0016] Performing fusion feature extraction on the data in the second data set to obtain a second fusion feature of the user;

[0017] The second fusion feature of the user is input into a risk assessment model, and a risk assessment result is output; wherein the risk assessment model includes a multi-label classification algorithm.

[0018] Optionally, determining a health management plan based on the user's predicted blood sugar fluctuation trend, the user's current risk assessment results, target physical indicators, and lifestyle habits includes:

[0019] For each preset plan, determine the plan's fitness value based on the user's predicted blood sugar fluctuation trend, the user's current risk assessment results, target physical indicators, and lifestyle habits; wherein the preset plan includes encoding of various parameters for dietary recommendations, exercise plans, and medication guidance;

[0020] The solution with the highest fitness value is selected as the target solution;

[0021] The target solution is decoded to obtain a health management solution.

[0022] Optionally, after determining a health management plan based on the user's predicted blood sugar fluctuation trend, the user's current risk assessment results, target physical indicators, and lifestyle habits, the method further includes:

[0023] Real-time monitoring of the user's health status; wherein the user's health status includes at least: blood sugar index, the user's current risk assessment results, physical indicators and the implementation of the user's health management plan;

[0024] If the user's health status triggers the preset threshold, an early warning message is generated.

[0025] Optionally, after determining a health management plan based on the user's predicted blood sugar fluctuation trend, the user's current risk assessment results, target physical indicators, and lifestyle habits, the method further includes:

[0026] Determine an optimal adjustment action pair for the current user's health status based on the current user's health status; wherein the adjustment action includes an action adjustment method and a parameter change range; and the adjustment action pair includes at least two adjustment actions;

[0027] An optimized health management plan is generated based on the optimal adjustment action pair.

[0028] Optionally, determining the optimal adjustment action pair for the current user's health status according to the current user's health status includes:

[0029] Initialize a two-dimensional quality value table; wherein the number of rows in the two-dimensional quality value table corresponds to the health dimension, and the number of columns corresponds to the adjustment action;

[0030] At each time step, a greedy strategy is used to select the target adjustment action pair based on the current health dimension data;

[0031] Executing the target adjustment action pair, collecting the user's subsequent health data in real time, and determining new health dimension data and immediate rewards;

[0032] Iteratively update the two-dimensional quality value table according to the preset quality learning update formula;

[0033] When the changes of multiple consecutive iterations of the two-dimensional quality value table are less than a specific threshold, the target adjustment action pair with the largest quality value is taken as the optimal adjustment action pair for the current user's health status.

[0034] A second aspect of the present application provides a device for generating a health management plan, comprising:

[0035] A data collection unit, configured to collect user data, wherein the user data includes physiological data, clinical data, and input data;

[0036] a trend prediction unit, configured to determine a predicted trend of blood sugar fluctuation of a user based on the physiological data and the input data;

[0037] a risk assessment unit, configured to determine a current risk assessment result of the user based on the clinical data;

[0038] A health management plan determination unit is used to determine a health management plan based on the user's blood sugar fluctuation prediction trend, the user's current risk assessment results, target physical indicators and living habits.

[0039] Optionally, the trend prediction unit includes:

[0040] a first preprocessing unit, configured to preprocess the physiological data and the input data to obtain a first data set;

[0041] a first fusion feature extraction unit, configured to perform fusion feature extraction on the data in the first data set to obtain a first fusion feature of the user;

[0042] The prediction subunit is used to input the second fusion feature of the user into the prediction model and output the predicted trend of the user's blood sugar fluctuation; wherein the prediction model includes a trained large language model and a fusion model of a time recurrent network.

[0043] Optionally, the risk assessment unit includes:

[0044] a second preprocessing unit, configured to preprocess the clinical data to obtain a second data set;

[0045] a second fusion feature extraction unit, configured to perform fusion feature extraction on the data in the second data set to obtain a second fusion feature of the user;

[0046] An evaluation subunit is configured to input the second fused feature of the user into a risk assessment model and output a risk assessment result; wherein the risk assessment model includes a multi-label classification algorithm.

[0047] Optionally, the health management plan determination unit includes:

[0048] a fitness value determination unit, configured to determine, for each preset plan, a fitness value based on the user's predicted blood sugar fluctuation trend, the user's current risk assessment results, target physical indicators, and lifestyle habits; wherein the preset plan includes encoding of various parameters for dietary recommendations, exercise plans, and medication guidance;

[0049] A target solution determination unit is used to select the solution with the highest solution fitness value as the target solution;

[0050] A decoding unit is used to decode the target solution to obtain a health management solution.

[0051] Optionally, the device for generating a health management plan further includes:

[0052] A monitoring unit for monitoring the user's health status in real time; wherein the user's health status includes at least: blood sugar index, the user's current risk assessment result, physical index and the implementation of the user's health management plan;

[0053] The early warning unit is used to generate early warning information if the user's health status triggers a preset threshold.

[0054] Optionally, the device for generating a health management plan further includes:

[0055] An adjustment determination unit, configured to determine an optimal adjustment action pair for the current user's health status based on the current user's health status; wherein the adjustment action includes an action adjustment method and a parameter variation range; and the adjustment action pair includes at least two adjustment actions;

[0056] A generating unit is used to generate an optimized health management plan based on the optimal adjustment action pair.

[0057] Optionally, the adjustment determination unit includes:

[0058] an initialization unit, configured to initialize a two-dimensional quality value table, wherein the number of rows in the two-dimensional quality value table corresponds to the health dimension, and the number of columns corresponds to the adjustment action;

[0059] The selection unit is used to select the target adjustment action pair based on the current health dimension data at each time step using a greedy strategy;

[0060] an execution unit, configured to execute the target adjustment action pair, collect subsequent health data of the user in real time, and determine new health dimension data and immediate rewards;

[0061] An updating unit, configured to iteratively update the two-dimensional quality value table according to a preset quality learning update formula;

[0062] The adjustment determination subunit is used to take the target adjustment action pair with the largest quality value as the optimal adjustment action pair for the current user's health status when the continuous multiple iteration changes of the two-dimensional quality value table are less than a specific threshold.

[0063] A third aspect of the present application provides an electronic device, including:

[0064] one or more processors;

[0065] a storage device having one or more programs stored thereon;

[0066] When the one or more programs are executed by the one or more processors, the one or more processors implement the method for generating a health management plan as described in any one of the first aspects.

[0067] A fourth aspect of the present application provides a computer storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the method for generating a health management plan as described in any one of the first aspects is implemented.

[0068] As can be seen from the above scheme, this application provides a method for generating a health management plan, related devices and computer storage media, which collects the user's physiological data, clinical data and input data; determines the user's blood sugar fluctuation prediction trend based on the physiological data and input data; then, determines the user's current risk assessment result based on the clinical data; and finally, determines the health management plan based on the user's blood sugar fluctuation prediction trend, the user's current risk assessment result, target body indicators and living habits. This effectively improves the intelligence level and accuracy of diabetes management, provides diabetic patients with a more scientific, efficient and personalized health management plan, helps reduce the risk of diabetic complications, and improves the quality of life of patients. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0070] Figure 1 A specific flow chart of a method for generating a health management plan provided in an embodiment of the present application;

[0071] Figure 2 This is an overall architecture diagram of a data collection method provided in another embodiment of the present application;

[0072] Figure 3 This is a flowchart of a method for determining a user's blood sugar fluctuation prediction trend provided by another embodiment of the present application;

[0073] Figure 4 A flowchart of a method for determining a user's current risk assessment result provided in another embodiment of the present application;

[0074] Figure 5 A flowchart of a method for determining a health management plan provided in another embodiment of the present application;

[0075] Figure 6 A flowchart of a method for dynamically adjusting a health management plan provided in another embodiment of the present application;

[0076] Figure 7 A flowchart of a method for determining an optimal adjustment action pair for a current user's health status provided by another embodiment of the present application;

[0077] Figure 8 A schematic diagram of a device for generating a health management plan provided in another embodiment of the present application;

[0078] Figure 9 A schematic diagram of an electronic device for implementing a method for determining an answer to a question provided in another embodiment of the present application. DETAILED DESCRIPTION

[0079] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0080] As used herein, the term "including" and its variations are open-ended, i.e., "including but not limited to." The term "based on" means "based, at least in part, on." The term "one embodiment" means "at least one embodiment," the term "another embodiment" means "at least one additional embodiment," and the term "some embodiments" means "at least some embodiments." Other terms are defined in the following description.

[0081] It should be noted that the information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of relevant countries and regions.

[0082] It should be noted that the concepts of "first" and "second" mentioned in this application are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0083] It should be noted that the modifications of "one" and "multiple" mentioned in this application are illustrative rather than restrictive. Those skilled in the art should understand that unless otherwise clearly indicated in the context, they should be understood as "one or more".

[0084] The present application embodiment provides a method for generating a health management plan, such as Figure 1 As shown, the specific steps include:

[0085] S101. Collect user data.

[0086] Among them, user data includes but is not limited to physiological data, clinical data, input data, basic information, etc., which are not limited here.

[0087] Physiological data includes but is not limited to blood sugar data, personal vital signs data, etc.; clinical data includes but is not limited to historical cases, test reports, etc.; input data includes but is not limited to diet, exercise, current symptoms, etc. independently entered by users on the mobile terminal; basic information includes but is not limited to user age, gender, family medical history, etc., which are not limited here.

[0088] During the actual application of this application, a "end-edge-cloud" collaborative data collection architecture was constructed to collect physiological data in real time through wearable devices (such as continuous blood glucose monitors and smart watches); access to the Hospital Information System (HIS) system was used to obtain historical medical records and test reports; and a mobile application was developed for patients to independently enter diet, exercise and symptom information.

[0089] In addition, data cleaning algorithms are used to remove noise data, natural language processing technology is used to parse medical record texts, and data standardization technology is combined to convert multi-source heterogeneous data into a unified format, providing a high-quality data foundation for subsequent analysis.

[0090] Specifically, outlier detection algorithms (such as the 3σ principle) can be used, but are not limited to, to eliminate erroneous data. Natural language processing (NLP) techniques can be used to parse medical records and convert unstructured data into structured information. Data mapping and normalization can be used to eliminate format differences between different data sources and create a unified diabetes patient dataset.

[0091] like Figure 2 The figure shows the overall architecture of a data collection method provided by an embodiment of the present application, which provides a reliable, secure, low-cost, and elastically scalable data synchronization platform capable of spanning heterogeneous data storage systems. Data collection combines multiple data sources, provides full / incremental data synchronization channels under different network environments, and has a visual wizard task configuration method, which can quickly implement enterprise-level data collection functions and reduce the development cost of data collection; data collection provides wizard-style configuration guidance, eliminating the need for tedious script step configuration, and can quickly complete the configuration of data synchronization tasks through visual filling and next step guidance; data collection supports mutual data synchronization between multiple homogeneous and heterogeneous data sources, such as relational databases, file systems, Hadoop, NoSQL and other data sources, to meet different data migration scenarios; data collection supports multiple data synchronization methods such as full / incremental / full+incremental, single table and batch table, to meet different data synchronization needs.

[0092] S102: Determine the user's blood sugar fluctuation prediction trend based on the physiological data and the input data.

[0093] Optionally, in another embodiment of the present application, an implementation of step S102 is as follows: Figure 3 Shown, including:

[0094] S301 : Preprocess physiological data and input data to obtain a first data set.

[0095] Among them, the preprocessing methods include but are not limited to standardization, word segmentation, etc., which are not limited here.

[0096] For example, blood sugar data and personal vital signs data are standardized, converted into a uniform numerical range, and normalization formulas are used to make data of different indicators comparable. Medical records require operations such as word segmentation and stop word removal to convert them into computer-processable text sequences.

[0097] Blood glucose data includes blood glucose measurements at different time points, along with relevant data that may influence blood glucose levels, such as dietary records, exercise data, medication information, and sleep duration. The collected data is cleaned to remove outliers and missing values, and missing data is filled in using methods such as interpolation and mean filling. The data is then normalized, with values ​​mapped to the [0, 1] range.

[0098] S302: Perform fusion feature extraction on the data in the first data set to obtain a first fusion feature of the user.

[0099] In the specific implementation of this application, feature extraction methods include, but are not limited to, extracting statistical features such as mean, variance, and trend for blood glucose data and personal vital sign data. For medical records, word embedding technologies (such as Word2Vec and GloVe) or pre-trained language models (such as BERT) can be used to extract semantic features and map each word to a low-dimensional vector, which is not limited here.

[0100] The method of performing feature fusion may be, but is not limited to, early fusion, late fusion, hybrid fusion, etc., and is not limited here.

[0101] Early fusion: If the correlation between blood glucose data, medical records, and personal vital sign data is considered strong, an early fusion architecture can be used. The extracted features from different modalities are directly concatenated or combined through a specific fusion module to generate a unified multimodal feature representation, which is then input into the large prediction model for subsequent processing.

[0102] Late fusion: If each modality's data needs to be fully explored within its own network's strengths, a late fusion architecture can be chosen. Blood glucose data features, medical record text features, and individual vital sign data features are fed into their respective neural networks for processing. Later in the model, before the output layer, these features are fused through methods such as feature concatenation and weighted summation.

[0103] Hybrid fusion: Combining the advantages of early fusion and late fusion, early fusion is used at the bottom layer of the model for some closely related modalities (such as blood sugar data and personal vital signs data) to quickly establish basic cross-modal connections; at the middle and high layers of the model, late fusion is used for modal data such as medical records that require in-depth independent analysis to give full play to their independent processing advantages.

[0104] Of course, in the specific implementation process of this application, a multimodal attention mechanism can also be introduced to enhance the model's ability to understand complex health data.

[0105] For example, for each time step or feature dimension, attention weights are calculated between blood glucose data, medical record text data, and personal vital sign data. The importance of each modal data at the current moment or dimension can be represented by calculating the similarity (such as dot product or cosine similarity) between feature vectors of different modalities and then converting it into weights through the softmax function. Based on the calculated attention weights, the features of different modalities are weighted and summed to obtain a multimodal feature representation processed by the attention mechanism. In this way, the model can automatically focus on more important modal data in different tasks and scenarios. For example, when diagnosing diabetes-related diseases, symptom descriptions and blood glucose data in medical record text may be given higher weights.

[0106] Specifically, effective time series features are extracted from the preprocessed data, such as the time stamp's hour, day of the week, and holiday status, to capture the periodicity of blood sugar changes. The first- and second-order differences of blood sugar are calculated to obtain dynamic features such as the rate of change and acceleration of blood sugar. These features are then fused with the original blood sugar data to construct a richer feature vector.

[0107] S303: Input the second fusion feature of the user into the prediction model, and output the predicted trend of the blood sugar fluctuation of the user.

[0108] The prediction model includes a trained large language model and a fusion model of a temporal recurrent network. Large language models include, but are not limited to, DeepSeek, and temporal recurrent networks include, but are not limited to, long short-term memory (LSTM) networks. These are not limited here.

[0109] In the specific implementation of this application, the training and optimization process of the prediction model includes defining an appropriate loss function based on the specific application task (such as disease diagnosis, condition prediction, etc.). For classification tasks, the cross entropy loss function can be used; for regression tasks, the mean square error loss function can be used.

[0110] The preprocessed multi-source data is fed into the DeepSeek model, which incorporates a multimodal attention mechanism. Training is performed using optimization algorithms such as stochastic gradient descent (SGD), Adagrad, and Adadelta. The model parameters are continuously adjusted to minimize the loss function, allowing the model to learn the connections and patterns between the multi-source data. Finally, through methods such as cross-validation, model hyperparameters such as the number of heads in the attention mechanism, the learning rate, and the number of hidden layer neurons are adjusted to achieve optimal model performance.

[0111] During the specific implementation of this application, the trained model can also be evaluated using a test dataset, using metrics such as accuracy, recall, F1 value, and root mean square error to measure the model's performance on the multi-source data joint modeling task. Based on the evaluation results, the model's shortcomings can be analyzed, such as whether there is overfitting or insufficient attention to certain modal data. The model can be improved by increasing the amount of data, adjusting the model structure, and optimizing the attention mechanism to continuously enhance its ability to jointly model multi-source data such as blood glucose data, medical records, and personal vital signs.

[0112] Specifically, the user's second fusion feature is input into the prediction model. First, the trained large language model is called to extract the fusion feature to obtain a feature vector. Then, the feature vector is input into the time recurrent network, and the output is the blood sugar fluctuation prediction trend.

[0113] S103. Determine the user's current risk assessment result based on the clinical data.

[0114] It should be noted that, in the actual application process of this application, it is not limited to executing step S102 first and then executing step S103. Step S102 and step S103 can be performed simultaneously, or step S103 can be performed first and then step S102. Figure 1 This is just an exemplary description, taking the example of first executing step S102 and then executing step S103.

[0115] Optionally, in another embodiment of the present application, an implementation of step S103 is as follows: Figure 4 Shown, including:

[0116] S401. Preprocess the clinical data to obtain a second data set.

[0117] S402: Perform fusion feature extraction on the data in the second data set to obtain a second fusion feature of the user.

[0118] The specific implementation of step S401 and step S402 can refer to step S301 and step S302, which will not be repeated here.

[0119] S403: Input the user's second fusion feature into the risk assessment model, and output a risk assessment result.

[0120] Among them, the risk assessment model includes a multi-label classification algorithm.

[0121] In the specific implementation of this application, the process of constructing a risk assessment model includes: first, collecting clinical data of diabetic patients, including but not limited to blood sugar, blood pressure, blood lipids, renal function indicators (such as creatinine and urea nitrogen), fundus examination results, medical history records, etc. Professional medical experts are invited to annotate this data and, according to the diagnostic criteria of six common complications such as diabetic nephropathy and retinopathy, grade and label whether the patient has a certain complication and the severity of the disease, to construct a multi-label classification dataset. Then, feature selection algorithms such as chi-square test and mutual information are used to screen out features with a high correlation with the six types of complications, reduce data dimensions, and reduce the complexity of model training. At the same time, numerical features are standardized, and categorical features are one-hot encoded or label encoded. For image data (such as fundus images), a convolutional neural network (CNN) is used for feature extraction, and the extracted features are fused with other clinical data features. Ultimately, algorithms suitable for multi-label classification were selected, such as Binary Relevance (which transforms a multi-label problem into multiple binary classification problems), Label Powerset (which treats each label combination as a single category for classification), and Classifier Chains (which classifies each label sequentially, taking into account the dependencies between labels), to construct a complication risk assessment model. Using the cross-entropy loss function as the optimization objective, the model was trained using a gradient descent algorithm, and model parameters were adjusted to enable the model to accurately provide graded warnings for the six types of complications.

[0122] In the actual application of this application, the model can be evaluated on the test set using multi-label classification evaluation metrics such as, but not limited to, accuracy, recall, and F1 value. The model's predictive performance for different complication categories is analyzed. For categories with low prediction accuracy, the model structure is adjusted or relevant features are added. By continuously optimizing the model, the early warning accuracy rate reaches above 90%. This is not limited here.

[0123] S104. Determine a health management plan based on the user's predicted blood sugar fluctuation trend, the user's current risk assessment results, target body indicators, and lifestyle habits.

[0124] Among them, target physical indicators include but are not limited to blood sugar control targets, body mass index, blood pressure, etc., and lifestyle habits include but are not limited to eating habits, exercise preferences, work and rest routines, etc., which are not limited here.

[0125] Through multi-dimensional data perception and preprocessing, integrating comprehensive patient health information, and combining the powerful multimodal data processing capabilities of the DeepSeek large-scale model, we can accurately predict blood sugar trends (with an error rate within ±5%) and efficiently assess the risk of diabetes complications (with an early warning accuracy rate exceeding 90%). Compared to traditional management methods, this can proactively identify changes in the condition, provide patients with more targeted prevention and intervention measures, effectively reduce the risk of complications, and safeguard their health.

[0126] During the specific implementation of this application, diabetes knowledge and suggestions can be pushed to patients in the form of pictures, texts, and videos through channels such as but not limited to mobile applications and text messages. At the same time, reminder services such as timely measurement and medication can be set up to promote the implementation of the plan.

[0127] Optionally, in another embodiment of the present application, an implementation of step S104 is as follows: Figure 5 Shown, including:

[0128] S501. For each preset plan, determine the plan fitness value based on the user's blood sugar fluctuation prediction trend, the user's current risk assessment results, target body indicators, and lifestyle habits.

[0129] Among them, the preset plan includes the encoding of various parameters of dietary recommendations, exercise plans and medication instructions.

[0130] In the specific implementation process of this application, it is first necessary to pre-encode various parameters of dietary recommendations, exercise plans, and medication guidance, such as calorie distribution (ratio of carbohydrates, protein, and fat), nutrient ratio (vitamin and mineral intake) in dietary recommendations, exercise intensity (expressed in heart rate ranges) and frequency (number of exercise days per week and duration of each exercise session) in exercise plans, and dosage adjustment rules in medication guidance (coefficients for adjusting drug dosage based on changes in blood sugar levels), and convert them into genetic coding. A certain number of individuals are randomly generated to form an initial population, each of which represents a possible three-dimensional management plan, namely the preset plan.

[0131] The fitness function is then designed by comprehensively considering the patient's physical indicators (such as blood sugar control targets, body mass index, blood pressure, etc.), lifestyle habits (eating habits, exercise preferences, sleep and rest patterns, etc.), as well as the blood sugar trend prediction results and complication risk assessment results. For example, for blood sugar prediction results, plans that bring predicted blood sugar values ​​closer to the target blood sugar range are given a higher fitness score. For complication risk, plans that reduce the probability of high-risk complications receive higher scores. Furthermore, plans that are more feasible based on the patient's lifestyle habits also receive higher fitness scores. The fitness function is used to evaluate each individual in the population and calculate its fitness value.

[0132] Next, for each individual, methods such as roulette wheel selection and tournament selection are used to select superior individuals from the current population based on their fitness. This increases the probability that individuals with high fitness will be selected and enter the next generation. A crossover operation is performed on these selected individuals, randomly exchanging portions of their genes with a certain crossover probability to generate new individuals and increase population diversity. Random mutations are performed on individual genes with a low mutation probability, altering their values ​​and preventing the algorithm from falling into local optima, allowing further exploration of the solution space. This genetic operation is repeated to continuously update the population and calculate the fitness values ​​of individuals in the next generation.

[0133] S502: The solution with the highest fitness value is selected as the target solution.

[0134] Continuing with the above example, in actual application, an iteration termination condition will also be set, such as reaching the maximum number of iterations or the population fitness value tending to be stable.

[0135] When the termination condition is met, the individual with the highest fitness is selected from the final population as the target solution.

[0136] S503: Decode the target solution to obtain a health management solution.

[0137] Specifically, the target plan is decoded into specific dietary recommendations, exercise plans and medication instructions to generate a personalized health management plan.

[0138] This application uses genetic algorithms and deep learning technology to fully consider the individual differences of patients and generate personalized three-dimensional management plans including diet, exercise, and medication. Whether it is a patient with special eating habits or an individual with different exercise abilities, they can obtain a management strategy that suits their needs, significantly improving the patient's compliance with the management plan, thereby enhancing the management effect. By actively identifying the patient's potential health risks, such as predicting the risk of increased blood sugar due to diet and pushing adjustment suggestions. Reminders and notifications can also be sent based on the medication cycle and follow-up time, and personalized health tips can be pushed based on daily health data to provide continuous and caring services and improve patient management experience and satisfaction.

[0139] Optionally, in another embodiment of the present application, an implementation of the method for generating a health management plan further includes:

[0140] Monitor the user's health status in real time; the user's health status includes at least: blood sugar index, the user's current risk assessment results, physical indicators and the implementation of the user's health management plan; if the user's health status triggers the preset threshold, an early warning message is generated.

[0141] During the specific implementation of this application, real-time collection of patient health data can be achieved through a variety of devices. For example: using a continuous glucose monitor (CGM) to collect blood glucose data at high frequency, obtaining a blood glucose value every 5-15 minutes; a smart bracelet or sports watch to record the patient's exercise data such as exercise duration, intensity, heart rate, etc.; smart scales, blood pressure monitors, blood lipid testers and other equipment to collect BMI, blood pressure, blood lipids and other physical indicators data. At the same time, a supporting APP is developed to allow patients to manually enter information such as diet content, medication time and dosage. All collected data are uniformly transmitted to the central database for format conversion and standardization, eliminating data silos and realizing the integration of multi-source data.

[0142] Based on data visualization technology, a patient health data dashboard is built. Web front-end frameworks (such as Vue.js and React) and visualization libraries (such as ECharts and Highcharts) are used to display the integrated data in the form of intuitive charts. For example, a line chart can be used to present the changing trend of blood sugar levels over the past 24 hours in real time, with the target blood sugar range marked as a reference line; a dashboard can be used to display the comparison between the current values ​​of physical indicators such as BMI, blood pressure, and blood lipids and the normal range; and a progress bar or status icon can be used to display the completion of diet, exercise, and medication plans. The dashboard allows patients and medical staff to customize the display content and time range according to their needs, making it easy to quickly obtain key information.

[0143] In the actual application process of this application, this application can also establish a patient health data dashboard to display key indicators such as blood sugar changes and plan execution status in real time. When the data triggers the preset threshold (such as blood sugar exceeding the standard for three consecutive times), an early warning will be automatically issued.

[0144] The preset thresholds include, but are not limited to, setting a warning for blood sugar data when the blood sugar value exceeds 7.0mmol / L (fasting) or 11.1mmol / L (2 hours after a meal) for three consecutive times; when executing an exercise plan, if the actual exercise time is less than 50% of the plan for three consecutive days, a warning will be triggered; and for medication, if the medication is not recorded for more than 1 hour after the prescribed time, a reminder will be issued. When real-time data triggers the preset threshold, the system automatically issues an alert through various methods such as app push, SMS, and email, and highlights the abnormal data in a striking color (such as red) on the data dashboard, reminding patients to pay attention and take timely measures. Medical staff can also obtain warning information simultaneously for remote guidance.

[0145] In the specific implementation process of this application, the user's health status can also be adjusted in real time according to the user's health status to obtain an optimized health management plan, which can more effectively improve the patient's health management level. An implementation method of the health management plan generation method is as follows: Figure 6 As shown, it also includes:

[0146] S601: Determine an optimal adjustment action pair for the current user's health status based on the current user's health status.

[0147] The adjustment action includes the adjustment method and parameter change range of the action; and the adjustment action pair includes at least two adjustment actions.

[0148] First, we need to construct a health state space and adjustment action controls. Specifically, we need to build a multidimensional state space encompassing blood sugar indicators, physical indicators, plan execution status, and complication risk status, and quantify and encode the data in each dimension. We also need to define an action space encompassing three types of plan adjustment actions: diet, exercise, and medication. We also need to clarify the specific adjustment methods and parameter variation ranges for each action, providing a data foundation for the quality learning algorithm.

[0149] The quality learning algorithm includes but is not limited to Q-learning, etc., which is not limited here.

[0150] Taking Q-learning as an example, it is necessary to set reward and punishment rules based on multiple dimensions such as blood sugar control, improvement of physical indicators, implementation of the plan, and reduction of complication risks. Through reasonable score settings, the algorithm is guided to learn in the direction of optimizing the management plan, ensuring that the algorithm can evaluate the pros and cons of adjustment actions based on patient health feedback.

[0151] Optionally, in another embodiment of the present application, an implementation of step S601 is as follows: Figure 7 Shown, including:

[0152] S701: Initialize a two-dimensional quality value table.

[0153] The number of rows in the two-dimensional quality value table corresponds to the health dimension, and the number of columns corresponds to the adjustment action.

[0154] Specifically, all elements are initialized to 0 and are used to store the expected cumulative reward of the health state-adjustment action pair.

[0155] S702. At each time step, a greedy strategy is used to select a target adjustment action pair based on the current health dimension data.

[0156] S703. Execute the target adjustment action pair, collect the user's subsequent health data in real time, and determine new health dimension data and immediate rewards.

[0157] Among them, the instant reward is calculated according to the reward mechanism.

[0158] Specifically, the initial exploration rate ε and the decay formula can be set to balance the exploration of new strategies and the utilization of existing experience.

[0159] S704: Iteratively update the two-dimensional quality value table according to a preset quality learning update formula.

[0160] According to the Q-learning update formula, combined with the set learning rate and discount factor, the two-dimensional quality value table is continuously updated iteratively.

[0161] S705 : When the changes of the two-dimensional quality value table in multiple consecutive iterations are less than a specific threshold, the target adjustment action pair with the largest quality value is used as the optimal adjustment action pair for the current user's health status.

[0162] S602: Generate an optimized health management plan based on the optimal adjustment action pair.

[0163] During the implementation of this application, patient health data will be continuously collected and fully stored in a database based on real-time data monitoring. In addition to routine data such as blood sugar and physical indicators, the content and time of each management plan adjustment will be recorded, as well as the patient's subjective feedback on the adjustment plan (such as physical sensations and difficulty of implementing the plan). A data backup and archiving mechanism will be established to ensure data security and integrity, providing rich data resources for subsequent model iterations.

[0164] During the specific implementation of this application, the large amount of collected data can also be analyzed regularly (such as weekly or monthly), and the model evaluation indicators in machine learning (such as mean square error, accuracy, F1 value, etc.) can be used to evaluate the effectiveness of the current blood sugar trend prediction model, complication risk assessment model, and parameters in the Q-learning algorithm. If it is found that the model prediction error increases or the scheme adjustment effect is not good, the model can be retrained using new data and the model structure and parameters can be adjusted. For example, the number of layers or neurons in the deep learning model can be increased, and the exploration rate decay rate and reward weight distribution in the Q-learning algorithm can be optimized to make the model better adapt to changes in the patient's condition and individual differences.

[0165] It's understandable that this application closely connects four key processes: data collection, intelligent analysis (including blood sugar trend prediction, complication risk assessment, and Q-learning algorithm optimization), intervention (adjusting diet, exercise, and medication regimens), and feedback (collecting patient health data changes and subjective feedback), forming a continuously circulating closed loop. Each completed closed loop optimizes the entire management process based on new data and feedback, continuously improving the accuracy and effectiveness of the diabetes management plan and achieving dynamic, personalized diabetes management.

[0166] Through the above steps of real-time data monitoring, adaptive adjustment of plans and closed-loop management optimization, a complete diabetes closed-loop management system is built, which can effectively improve the health management level of diabetic patients.

[0167] During the actual application of this application, the management effect can be regularly evaluated from the dimensions of blood sugar control, complications, quality of life, etc., and the results can be fed back to both doctors and patients. The plan can be optimized based on the feedback, chronic disease follow-up can be carried out regularly, and experience can be accumulated to improve subsequent management.

[0168] Specifically, algorithms such as principal component analysis and linear discriminant analysis can be used for data dimensionality reduction and feature extraction to uncover key information from follow-up data. DeepSeek combined with the LSTM algorithm can be used to predict blood sugar trends, a multi-label classification algorithm can be used to assess complication risks, a genetic algorithm can be used to generate personalized management plans, and clustering algorithms can be used to segment patient groups and analyze the characteristics of different groups. A key focus is on the correlation between the effectiveness of management plan implementation and changes in patient health indicators. For example, analysis can be performed on blood sugar fluctuations after dietary adjustments and the impact of exercise plans on weight and blood sugar, providing a basis for plan iteration.

[0169] As can be seen from the above scheme, this application provides a method for generating a health management plan. After collecting the user's physiological data, clinical data, and input data; determining the user's blood sugar fluctuation prediction trend based on the physiological data and input data; then, determining the user's current risk assessment result based on the clinical data; and finally, determining the health management plan based on the user's blood sugar fluctuation prediction trend, the user's current risk assessment result, target body indicators, and lifestyle habits. This effectively improves the intelligence level and accuracy of diabetes management, provides diabetic patients with a more scientific, efficient, and personalized health management plan, helps reduce the risk of diabetes complications, and improves the quality of life of patients.

[0170] Another embodiment of the present application provides a device for generating a health management plan, such as Figure 8 As shown, specifically including:

[0171] The data collection unit 801 is used to collect user data.

[0172] Among them, user data includes: physiological data, clinical data and input data.

[0173] The trend prediction unit 802 is used to determine the user's blood sugar fluctuation prediction trend based on the physiological data and the input data.

[0174] Optionally, in another embodiment of the present application, an implementation of the trend prediction unit 802 includes:

[0175] The first preprocessing unit is used to preprocess the physiological data and the input data to obtain a first data set.

[0176] The first fusion feature extraction unit is used to perform fusion feature extraction on the data in the first data set to obtain the first fusion feature of the user.

[0177] The prediction subunit is used to input the user's second fusion feature into the prediction model and output the user's blood sugar fluctuation prediction trend.

[0178] Among them, the prediction model includes a trained large language model and a fusion model of a time recurrent network.

[0179] For the specific working process of the units disclosed in the above embodiments of the present application, please refer to the corresponding method embodiments, such as Figure 3 As shown, no further details are given here.

[0180] The risk assessment unit 803 is used to determine the user's current risk assessment result based on clinical data.

[0181] Optionally, in another embodiment of the present application, an implementation of the risk assessment unit 803 includes:

[0182] The second preprocessing unit is used to preprocess the clinical data to obtain a second data set.

[0183] The second fusion feature extraction unit is used to perform fusion feature extraction on the data in the second data set to obtain the second fusion feature of the user.

[0184] The evaluation subunit is used to input the user's second fusion feature into the risk assessment model and output a risk assessment result.

[0185] Among them, the risk assessment model includes a multi-label classification algorithm.

[0186] For the specific working process of the units disclosed in the above embodiments of the present application, please refer to the corresponding method embodiments, such as Figure 4 As shown, no further details are given here.

[0187] The health management plan determination unit 804 is used to determine a health management plan based on the user's blood sugar fluctuation prediction trend, the user's current risk assessment results, target physical indicators and living habits.

[0188] For the specific working process of the units disclosed in the above embodiments of the present application, please refer to the corresponding method embodiments, such as Figure 1 As shown, no further details are given here.

[0189] Optionally, in another embodiment of the present application, an implementation of the health management plan determination unit 804 includes:

[0190] The fitness value determination unit is used to determine the fitness value of each preset plan based on the user's blood sugar fluctuation prediction trend, the user's current risk assessment results, target physical indicators and living habits.

[0191] Among them, the preset plan includes the encoding of various parameters of dietary recommendations, exercise plans and medication instructions.

[0192] The target solution determination unit is used to select the solution with the highest solution fitness value as the target solution.

[0193] The decoding unit is used to decode the target solution to obtain a health management solution.

[0194] For the specific working process of the units disclosed in the above embodiments of the present application, please refer to the corresponding method embodiments, such as Figure 5 As shown, no further details are given here.

[0195] Optionally, in another embodiment of the present application, an implementation of the device for generating a health management plan further includes:

[0196] The monitoring unit is used to monitor the user's health status in real time.

[0197] The user's health status includes at least: blood sugar index, the user's current risk assessment results, physical indicators and the implementation of the user's health management plan.

[0198] The early warning unit is used to generate early warning information if the user's health status triggers a preset threshold.

[0199] The specific working process of the units disclosed in the above embodiments of this application can be found in the corresponding method embodiments and will not be repeated here.

[0200] Optionally, in another embodiment of the application, an embodiment of the device for generating a health management plan further includes:

[0201] The adjustment determination unit is used to determine the optimal adjustment action pair for the current user's health status according to the current user's health status.

[0202] The adjustment action includes the adjustment method and parameter change range of the action; and the adjustment action pair includes at least two adjustment actions.

[0203] The generation unit is used to generate an optimized health management plan based on the optimal adjustment action pair.

[0204] For the specific working process of the units disclosed in the above embodiments of the present application, please refer to the corresponding method embodiments, such as Figure 6 As shown, no further details are given here.

[0205] Optionally, in another implementation of the application, an implementation of the adjustment determination unit further includes:

[0206] The initialization unit is used to initialize the two-dimensional quality value table.

[0207] The number of rows in the two-dimensional quality value table corresponds to the health dimension, and the number of columns corresponds to the adjustment action.

[0208] The selection unit is used to select the target adjustment action pair using a greedy strategy based on the current health dimension data at each time step.

[0209] The execution unit is used to execute the target adjustment action pair, collect the user's subsequent health data in real time, and determine new health dimension data and immediate rewards.

[0210] The updating unit is used to iteratively update the two-dimensional quality value table according to a preset quality learning update formula.

[0211] The adjustment determination subunit is used to take the target adjustment action pair with the largest quality value as the optimal adjustment action pair for the current user's health status when the continuous multiple iteration changes of the two-dimensional quality value table are less than a specific threshold.

[0212] For the specific working process of the units disclosed in the above embodiments of the present application, please refer to the corresponding method embodiments, such as Figure 7 As shown, no further details are given here.

[0213] As can be seen from the above scheme, this application provides a device for generating a health management plan. After collecting the user's physiological data, clinical data, and input data, the device determines the user's blood sugar fluctuation prediction trend based on the physiological data and input data; then, based on the clinical data, the device determines the user's current risk assessment result; and finally, the device determines the health management plan based on the user's blood sugar fluctuation prediction trend, the user's current risk assessment result, target body indicators, and lifestyle habits. This effectively improves the intelligence level and accuracy of diabetes management, provides diabetic patients with a more scientific, efficient, and personalized health management plan, helps reduce the risk of diabetes complications, and improves the quality of life of patients.

[0214] The functions described above herein may be performed, at least in part, by one or more hardware logic components. For example, and without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0215] Another embodiment of the present application provides an electronic device, such as Figure 9 Shown, including:

[0216] One or more processors 901.

[0217] The storage device 902 stores one or more programs.

[0218] When the one or more programs are executed by the one or more processors 901 , the one or more processors 901 implement the method for generating a health management plan as described in the above embodiments.

[0219] Another embodiment of the present application provides a computer storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the method for generating a health management plan as described in the above embodiment is implemented.

[0220] In the context of this application, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or apparatus. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of machine-readable storage media may include an electrical connection based on one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), optical fibers, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0221] It should be noted that the computer-readable medium referred to in this application may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. Computer-readable storage media may include, but are not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or components, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to, an electrical connection having one or more conductors, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In this application, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device, or component. In this application, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. This propagated data signal may take a variety of forms, including, but not limited to, electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer-readable medium may be transmitted using any suitable medium, including but not limited to wire, optical cable, RF (radio frequency), or any suitable combination thereof.

[0222] The computer-readable medium may be included in the electronic device, or may exist independently without being incorporated into the electronic device.

[0223] Another embodiment of the present application provides a computer program product, which, when executed, is used to execute the above-mentioned method for generating a health management plan.

[0224] In particular, according to an embodiment of the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present application includes a computer program product, which includes a computer program carried on a non-transitory computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network via a communication device, or installed from a storage device, or installed from a ROM. When the computer program is executed by a processing device, the above-mentioned functions defined in the method of the embodiment of the present application are performed.

[0225] Although the subject matter has been described in terms of structural features and / or method logic actions, it should be understood that the subject matter defined in this application is not necessarily limited to the specific features or actions described above. On the contrary, the specific features and actions described above are merely example forms of implementing this application.

[0226] Although several specific implementation details are included in the above discussion, these should not be construed as limiting the scope of this application. Certain features described in the context of separate embodiments may also be implemented in combination in a single embodiment. Conversely, various features described in the context of a single embodiment may also be implemented in multiple embodiments individually or in any suitable sub-combination.

[0227] The above description is merely a preferred embodiment of the present application and an illustration of the technical principles employed. Those skilled in the art should understand that the scope of application of this application is not limited to technical solutions formed by a specific combination of the aforementioned technical features. It also encompasses other technical solutions formed by any combination of the aforementioned technical features or their equivalents, without departing from the aforementioned application concept. For example, a technical solution formed by replacing the aforementioned features with (but not limited to) technical features with similar functions claimed in this application.

Claims

1. A method for generating a health management plan, characterized in that: include: Collecting user data; wherein the user data includes: physiological data, clinical data and input data; Determining a predicted trend of blood sugar fluctuation of the user based on the physiological data and the recorded data; determining a current risk assessment result of the user based on the clinical data; A health management plan is determined based on the user's predicted blood sugar fluctuation trend, the user's current risk assessment results, target physical indicators and living habits.

2. The method for generating a health management plan according to claim 1, characterized in that: Determining the predicted trend of blood sugar fluctuation of the user based on the physiological data and the recorded data includes: Preprocessing the physiological data and input data to obtain a first data set; Performing fusion feature extraction on the data in the first data set to obtain a first fusion feature of the user; The second fusion feature of the user is input into a prediction model, and the predicted trend of the user's blood sugar fluctuation is output; wherein the prediction model includes a trained large language model and a fusion model of a time recurrent network.

3. The method for generating a health management plan according to claim 1, wherein: Determining the user's current risk assessment result based on the clinical data includes: Preprocessing the clinical data to obtain a second data set; Performing fusion feature extraction on the data in the second data set to obtain a second fusion feature of the user; The second fusion feature of the user is input into a risk assessment model, and a risk assessment result is output; wherein the risk assessment model includes a multi-label classification algorithm.

4. The method for generating a health management plan according to claim 1, wherein: Determining a health management plan based on the user's predicted blood sugar fluctuation trend, the user's current risk assessment results, target physical indicators, and lifestyle habits includes: For each preset plan, determine the plan's fitness value based on the user's predicted blood sugar fluctuation trend, the user's current risk assessment results, target physical indicators, and lifestyle habits; wherein the preset plan includes encoding of various parameters for dietary recommendations, exercise plans, and medication guidance; The solution with the highest fitness value is selected as the target solution; The target solution is decoded to obtain a health management solution.

5. The method for generating a health management plan according to claim 1, wherein: After determining a health management plan based on the user's predicted blood sugar fluctuation trend, the user's current risk assessment results, target physical indicators, and lifestyle habits, the method further includes: Real-time monitoring of the user's health status; wherein the user's health status includes at least: blood sugar index, the user's current risk assessment results, physical indicators and the implementation of the user's health management plan; If the user's health status triggers the preset threshold, an early warning message is generated.

6. The method for generating a health management plan according to claim 5, characterized in that: After determining a health management plan based on the user's predicted blood sugar fluctuation trend, the user's current risk assessment results, target physical indicators, and lifestyle habits, the method further includes: Determine an optimal adjustment action pair for the current user's health status based on the current user's health status; wherein the adjustment action includes an action adjustment method and a parameter change range; and the adjustment action pair includes at least two adjustment actions; An optimized health management plan is generated based on the optimal adjustment action pair.

7. The method for generating a health management plan according to claim 6, characterized in that: The determining of the optimal adjustment action pair for the current user's health status according to the current user's health status includes: Initialize a two-dimensional quality value table; wherein the number of rows in the two-dimensional quality value table corresponds to the health dimension, and the number of columns corresponds to the adjustment action; At each time step, a greedy strategy is used to select the target adjustment action pair based on the current health dimension data; Executing the target adjustment action pair, collecting the user's subsequent health data in real time, and determining new health dimension data and immediate rewards; Iteratively update the two-dimensional quality value table according to the preset quality learning update formula; When the changes of multiple consecutive iterations of the two-dimensional quality value table are less than a specific threshold, the target adjustment action pair with the largest quality value is taken as the optimal adjustment action pair for the current user's health status.

8. A device for generating a health management plan, characterized in that: include: A data collection unit, configured to collect user data, wherein the user data includes physiological data, clinical data, and input data; a trend prediction unit, configured to determine a predicted trend of blood sugar fluctuation of a user based on the physiological data and the input data; a risk assessment unit, configured to determine a current risk assessment result of the user based on the clinical data; A health management plan determination unit is used to determine a health management plan based on the user's blood sugar fluctuation prediction trend, the user's current risk assessment results, target physical indicators and living habits.

9. An electronic device, characterized in that: include: one or more processors; a storage device having one or more programs stored thereon; When the one or more programs are executed by the one or more processors, the one or more processors are enabled to implement the method for generating a health management plan according to any one of claims 1 to 7.

10. A computer storage medium, characterized in that A computer program is stored thereon, wherein when the computer program is executed by a processor, the method for generating a health management plan according to any one of claims 1 to 7 is implemented.

Images