Method and device for constructing health portrait of user, equipment and medium

Abstract

The invention discloses a method and device for constructing a health portrait of a user, equipment and a medium. The method comprises the following steps: firstly, collecting user data, and preprocessing the user data to obtain user index data; obtaining a user disease risk level label according to the user index data and a disease risk level rule; inputting the user index data into a data trend prediction model, and predicting the index data trend of the user; inputting the user index data into the disease risk prediction model to predict the disease risk of the user; and finally, combining the disease risk level label of the user, the health index trend of the user and the disease risk, and constructing a health portrait of the user. In the process, multi-aspect user data is collected to comprehensively analyze user health data, a large model is accurately utilized to evaluate user health conditions and predict disease risks, and a multi-information user health portrait is constructed.

Description

Technical Field

[0001] The present application relates to the field of medical health management technology, and in particular to a method, device, equipment and medium for constructing a user health profile. Background Art

[0002] With the development of medical information technology, personal medical and health data has shown exponential growth. The accumulation of data such as electronic medical records, medical records and physical examination reports has provided a rich source of information for personal health management.

[0003] Therefore, a user health profile can be built for the user based on various data information in order to more intuitively understand the health status of different individual users and meet the needs of personalized health management.

[0004] The current method of constructing a user health profile focuses on the analysis of a single health indicator, which results in an incomplete health profile, insufficient personalization, and a lack of a comprehensive evaluation model, which results in the inability to dynamically predict the user's future health trends. Based on the above problems, it is difficult to meet the needs of personalized health management. Therefore, this application needs to propose a method to solve the above problems. Summary of the invention

[0005] The present application provides a method, device, equipment and medium for constructing a user health profile, which collects various user data, analyzes and evaluates the user data, and predicts the user's future health trends, thereby constructing a user health profile with multiple information.

[0006] In a first aspect, the present application provides a method for constructing a user health profile, the method comprising:

[0007] Collecting user data and preprocessing the user data to obtain user indicator data, wherein the user data includes basic information and behavior information of the user; wherein the basic information is data on the user's identity and health status, and the behavior information is activity or behavior data related to the user's health status;

[0008] Obtain the user's disease risk level label based on the user's indicator data and disease risk level rules;

[0009] Inputting user indicator data into a data trend prediction model to predict the indicator data trend of the user, wherein the indicator data prediction model is a trained neural network model, and the training data of the indicator data prediction model is historical indicator data and the trend of the historical indicator data changing over time;

[0010] Inputting the user indicator data into a disease risk prediction model to predict the user's disease risk, wherein the disease risk prediction model is a trained neural network model, and the training samples of the disease risk prediction model are historical indicator data and corresponding historical disease risk scores;

[0011] Combine the user's disease risk level label, the user's health indicator trend and disease risk to build a user health profile.

[0012] Optionally, preprocess the user data to obtain user indicator data, including:

[0013] Perform multimodal data processing on user data to extract key information;

[0014] Extract features of key information according to defined indicators, and extract user basic indicators from user data, wherein the defined indicators include user basic indicators, health basic indicators, and health extended indicators;

[0015] Perform data cleaning and standardization on the data in the user basic indicators to obtain user indicator data.

[0016] Optionally, according to the user indicator data and the disease risk level rule, a user disease risk level label is obtained, including:

[0017] Determine the user's disease theme based on the user's indicator data and health theme rules, wherein the health theme rules are rules for evaluating various diseases using indicator data;

[0018] Substitute the user indicator data into the disease risk level rule of the disease theme to obtain the user disease risk level label, wherein the disease risk level rule is a rule for evaluating the level of various diseases based on the indicator data.

[0019] Optionally, the training process of the data trend prediction model includes:

[0020] Obtain the user's historical indicator data and the trend of historical indicator data over time;

[0021] The historical indicator data and the trend of historical indicator data changing over time are input as training samples into the initial trend model for training until the initial trend model reaches the target value, thereby obtaining a data trend prediction model.

[0022] Optionally, the training process of the disease risk prediction model includes:

[0023] Obtain the user's historical indicator data and historical disease risk scores;

[0024] The historical indicator data and historical disease risk scores are input as training samples into the initial risk model for training until the initial risk model reaches the target value to obtain the disease risk prediction model.

[0025] Optionally, the method further comprises:

[0026] According to the user's disease risk level label, the big model is used to obtain healthy living guidance suggestions corresponding to the disease risk level label.

[0027] Optionally, the method further comprises:

[0028] Set the user's health control goals through parameters;

[0029] Determining, according to the health control target, an improvement suggestion for the user from the improvement suggestion set, wherein the improvement suggestion set has a corresponding relationship with the health control target;

[0030] According to the improvement suggestions of the user, a multi-faceted guidance and suggestion plan is generated, wherein the multi-faceted guidance and suggestion plan includes a lifestyle guidance and suggestion plan, a diet structure guidance and suggestion plan, and a sports guidance and suggestion plan.

[0031] In a second aspect, an embodiment of the present application provides a device for constructing a user health profile, the device comprising:

[0032] A preprocessing unit, used to collect user data and preprocess the user data to obtain user indicator data, wherein the user data includes basic information and behavior information of the user; wherein the basic information is data of the user's identity and health status, and the behavior information is activity or behavior data related to the user's health status;

[0033] An obtaining unit, used to obtain a user disease risk level label according to user indicator data and disease risk level rules;

[0034] A prediction unit, used to input user indicator data into a data trend prediction model to predict the indicator data trend of the user, wherein the indicator data prediction model is a trained neural network model, and the training data of the indicator data prediction model is historical indicator data and the trend of historical indicator data changing over time;

[0035] The prediction unit is further used to input the user indicator data into a disease risk prediction model to predict the user's disease risk, wherein the disease risk prediction model is a trained neural network model, and the training samples of the disease risk prediction model are historical indicator data and corresponding historical disease risk scores;

[0036] The construction unit is used to combine the user's disease risk level label, the user's health indicator trend and the disease risk to construct a user health profile.

[0037] Optionally, the preprocessing unit is specifically configured to:

[0038] Perform multimodal data processing on user data to extract key information;

[0039] Extract features of key information according to defined indicators, and extract user basic indicators from user data, wherein the defined indicators include user basic indicators, health basic indicators, and health extended indicators;

[0040] Perform data cleaning and standardization on the data in the user basic indicators to obtain user indicator data.

[0041] Optionally, obtaining a unit, specifically for:

[0042] Determine the user's disease theme based on the user's indicator data and health theme rules, wherein the health theme rules are rules for evaluating various diseases using indicator data;

[0043] Substitute the user indicator data into the disease risk level rule of the disease theme to obtain the user disease risk level label, wherein the disease risk level rule is a rule for evaluating the level of various diseases based on the indicator data.

[0044] Optionally, the device further comprises a training unit, configured to:

[0045] Obtain the user's historical indicator data and the trend of historical indicator data over time;

[0046] The historical indicator data and the trend of historical indicator data changing over time are input as training samples into the initial trend model for training until the initial trend model reaches the target value, thereby obtaining a data trend prediction model.

[0047] Optionally, the device further comprises a training unit, configured to:

[0048] Obtain the user's historical indicator data and historical disease risk scores;

[0049] The historical indicator data and historical disease risk scores are input as training samples into the initial risk model for training until the initial risk model reaches the target value to obtain the disease risk prediction model.

[0050] Optionally, the obtaining unit is further configured to:

[0051] According to the user's disease risk level label, the big model is used to obtain healthy living guidance suggestions corresponding to the disease risk level label.

[0052] Optionally, the device further comprises a generating unit, configured to:

[0053] Set the user's health control goals through parameters;

[0054] Determining, according to the health control target, an improvement suggestion for the user from the improvement suggestion set, wherein the improvement suggestion set has a corresponding relationship with the health control target;

[0055] According to the improvement suggestions of the user, a multi-faceted guidance and suggestion plan is generated, wherein the multi-faceted guidance and suggestion plan includes a lifestyle guidance and suggestion plan, a diet structure guidance and suggestion plan, and a sports guidance and suggestion plan.

[0056] In a third aspect, the present application provides an electronic device, the electronic device comprising a memory and a processor:

[0057] The memory is used to store computer programs;

[0058] The processor is used to execute the method provided by the first aspect according to the computer program.

[0059] In a fourth aspect, the present application further provides a computer-readable storage medium, which is used to store a computer program, and the computer program is used to execute the method provided in the first aspect above.

[0060] It can be seen that this application has the following beneficial effects:

[0061] The present application provides a method for constructing a user health portrait, firstly collecting user data, and preprocessing the user data to obtain user index data; obtaining the user disease risk level label according to the user index data and the disease risk level rule; inputting the user index data into the data trend prediction model to predict the index data trend of the user; and inputting the user index data into the disease risk prediction model to predict the user's disease risk; finally, combining the user's disease risk level label, the user's health index trend and the disease risk to construct a user health portrait. In this process, by collecting multi-faceted user data and comprehensively analyzing the user's health data, accurately using the large model to evaluate the user's health status and predict disease risks, a user health portrait with multiple information is constructed. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present application. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0063] Figure 1 A flowchart of an embodiment of a method for constructing a user health portrait in an embodiment of the present application;

[0064] Figure 2A flow chart of definition indicators for a method of constructing a user health profile in an embodiment of the present application;

[0065] Figure 3 This is a flow chart of another embodiment of a method for constructing a user health portrait in an embodiment of the present application;

[0066] Figure 4 A schematic diagram of the structure of a device for constructing a user health portrait in an embodiment of the present application;

[0067] Figure 5 This is a schematic diagram of the structure of an electronic device in an embodiment of the present application. DETAILED DESCRIPTION

[0068] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0069] It should be noted that the user 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 relevant laws, regulations and standards of relevant countries and regions.

[0070] Currently, there are several shortcomings in building user health portraits:

[0071] (1) Data source limitations: Many existing methods may rely on a single data source, such as electronic medical records or physical examination reports, without comprehensively considering multi-source data, such as social media, wearable device data, etc., which may lead to an incomplete portrait.

[0072] (2) Insufficient personalization: Existing methods may not be able to fully consider individual differences, such as living habits and environmental factors, resulting in insufficient personalization of portraits.

[0073] (3) Lack of dynamic prediction capabilities: Some methods may only provide static health assessments without the ability to dynamically predict users’ future health trends.

[0074] (4) Poor algorithm transparency and explainability: Some methods using complex machine learning models may lack transparency and explainability, making it difficult for users and medical professionals to understand the decision logic of the model.

[0075] (5) Lack of interdisciplinary integration: The construction of a health profile often requires knowledge from multiple disciplines such as medicine, psychology, and nutrition. Existing methods may not integrate this interdisciplinary knowledge well.

[0076] (6) Difficulty in large-scale application: Some methods may perform well in small-scale applications, but face problems such as performance degradation and cost increase when applied on a large scale.

[0077] In the embodiments of the present application, by combining big data and big model technology, the above technology is used to extract features and predict trends of original data such as electronic medical records, medical records, physical examination reports, and manual entries to build a personalized user health portrait.

[0078] In specific implementation, the method may, for example, include: first collecting user data, and preprocessing the user data to obtain user indicator data, wherein the user data includes basic information and behavior information of the user; wherein the basic information is data on the user's identity and health status, and the behavior information is activity or behavior data related to the user's health status; obtaining a user's disease risk level label based on the user indicator data and disease risk level rules; inputting the user indicator data into a data trend prediction model to predict the indicator data trend of the user, wherein the indicator data prediction model is a trained neural network model, and the training data of the indicator data prediction model is historical indicator data and the trend of historical indicator data over time; inputting the user indicator data into a disease risk prediction model to predict the user's disease risk, wherein the disease risk prediction model is a trained neural network model, and the training samples of the disease risk prediction model are historical indicator data and corresponding historical disease risk scores; finally, combining the user's disease risk level label, the user's health indicator trend and the disease risk to construct a user health portrait.

[0079] It can be seen that the method provided by the implementation of this application comprehensively analyzes user health data by collecting various user data, accurately assesses user health status and predicts disease risks, and constructs a user health portrait with multiple information.

[0080] To facilitate understanding of the specific implementation of the method for constructing a user health profile provided in the embodiment of the present application, an explanation will be given below with reference to the accompanying drawings.

[0081] It should be noted that the subject implementing the method of constructing a user health portrait may be a device for constructing a user health portrait provided in an embodiment of the present application, and the device for constructing a user health portrait may be carried in an electronic device or a functional module of an electronic device. The electronic device in the embodiment of the present application may be any device capable of implementing the method for constructing a user health portrait in the embodiment of the present application, such as an Internet of Things (IoT) device.

[0082] Figure 1 A flowchart of a method for constructing a user health profile provided in an embodiment of the present application. The method can be applied to a device for constructing a user health profile, and the device for constructing a user health profile can be, for example, Figure 4 The device 400 for constructing a user health profile is shown, or the device for constructing a user health profile may also be integrated in Figure 5 Functional modules in the electronic device 500 are shown.

[0083] like Figure 1 As shown, the method includes the following S101 to S105:

[0084] S101: Collect user data and pre-process the user data to obtain user indicator data, wherein the user data includes basic information and behavior information of the user; wherein the basic information is data on the user's identity and health status, and the behavior information is activity or behavior data related to the user's health status.

[0085] In order to construct a more comprehensive personalized user health portrait, the embodiment of the present application first collects user data and pre-processes the user data. After obtaining the user indicator data, the user disease risk level label is obtained according to the user indicator data and the disease risk level rules; the user indicator data is then input into the data trend prediction model to predict the user's indicator data trend and the user indicator data is input into the disease risk prediction model to predict the user's disease risk. Finally, the user disease risk level label, the user's health indicator trend and the disease risk are combined to construct a user health portrait. In the embodiment of the present application, the user indicator data is obtained by S101, which makes preliminary preparations for the subsequent acquisition of the user's disease risk level label, the user's health indicator trend and the disease risk.

[0086] As an example, S101 may include: S1011, collecting user data; S1012, preprocessing the user data to obtain user indicator data.

[0087] The user data in S1011 above includes the user's basic information and behavior information; the basic information is the data of the user's identity and health status, and the behavior information is the activity or behavior data related to the user's health status. The above user data can be obtained through electronic medical records, medical records, physical examination reports, manual entry, etc., or through App to collect user data in different dimensions, including but not limited to the user's basic information (such as gender, age, region, etc.), as well as electronic medical records, physical examination reports, medical data, behavioral habits (such as product usage behavior), health habits (such as exercise, diet, etc.) and health equipment data generated in different time periods.

[0088] In the above S1012, the user basic indicators contained in the user data can be extracted by means of large model feature extraction, optical character recognition technology OCR, multimodal processing technology, system fault tolerance processing, etc., and cleaned and standardized (such as: unit conversion) to finally obtain user indicator data. The specific process may include: first, extract text information from pictures and scanned files uploaded by users using OCR technology, and extract key information by combining multimodal data such as text and images; extract features of key information according to defined indicators, and extract user basic indicators in user data, wherein the defined indicators include user basic indicators and health basic indicators; clean and standardize the data in the user basic indicators to obtain user indicator data. Among them, data cleaning is to identify and process missing values, outliers and duplicate records, and standardization is to unify data formats and convert data of different units of measurement into standard units.

[0089] It should be noted that the above-defined indicators include basic user indicators, basic health indicators and extended health indicators, including: (1) Basic user indicators: basic information such as the user's name, gender, and education level. (2) Health status indicators: user health status data, describing a certain time (the time point of obtaining data), user health events and event results. For example: On a certain day of a certain month of a certain year, Zhang San's blood pressure at rest was 190 mmHg systolic and 120 mmHg diastolic. Specifically, it can include: Physiological health data: the user's physical symptoms and organ functions, daily life functions, physical activity functions, etc., which can be obtained through physical examination data, such as blood pressure, blood sugar, blood lipids, liver and kidney function indicators, etc.; Mental health data: including the user's mental state, including emotional state, stress level, sleep quality, etc., which can be achieved through mental health scales or regular mental health assessments; Social health data: including the user's health status in social relationships, social support and social activities, which may involve social frequency, social participation and the acquisition of social resources. (3) Define extended health indicators: Lifestyle data: including lifestyle habits such as diet, exercise, and sleep, which have a direct impact on health; medical history data: understanding the user's past medical conditions, which helps assess the risk of certain diseases; family medical history data: understanding the genetic diseases in the user's family, which helps assess the risk of certain diseases; occupational history data: recording the user's occupational exposure, living environment, and possible health risk factors in lifestyle habits.

[0090] S102: Obtain a user disease risk level label according to the user indicator data and disease risk level rules.

[0091] As an example, S102 may include: determining the user's disease theme based on user indicator data and health theme rules, wherein the health theme rules are rules for evaluating various types of diseases using indicator data; substituting the user indicator data into the disease risk level rules of the disease theme to obtain the user's disease risk level label, wherein the disease risk level rules are rules for evaluating various types of disease levels based on indicator data.

[0092] In the embodiment of the present application, the health theme rules include reference indicators of predefined health themes, that is, each health theme needs to have which indicators of the above-mentioned user basic indicators, health basic indicators and health extended indicators as reference indicators. Specific references can be made to some national, local or industry diagnosis and treatment guidelines, such as: "Guidelines for the Prevention and Treatment of Dyslipidemia in Adults in China (Revised Edition 2016)", "Chinese Medical Association Clinical Diagnosis and Treatment Guidelines for Lung Cancer (2023 Edition)", "Guidelines for Primary Diagnosis and Treatment of Dyslipidemia (Practical Edition 2019)" and "Guidelines for Early Screening, Diagnosis and Prevention of Chronic Kidney Disease (2022)". Among them, specific disease themes may include: Hypertension: Long-term high blood pressure may not have obvious symptoms, but it will increase the risk of damage to organs such as the heart, blood vessels, and kidneys. Diabetes: Due to insufficient insulin secretion or insufficient response of the body to insulin, blood sugar levels rise, and long-term existence may lead to a variety of complications. Coronary heart disease: Insufficient blood supply to the coronary arteries of the heart may lead to heart problems such as angina pectoris and myocardial infarction. Chronic obstructive pulmonary disease (COPD): including chronic bronchitis and emphysema, leading to persistent respiratory symptoms and decreased activity tolerance. Cerebrovascular disease: such as stroke, brain tissue is damaged due to rupture or blockage of blood vessels in the brain. Chronic kidney disease: Gradual loss of kidney function, which may eventually require dialysis or a kidney transplant. Osteoarthritis: Wear and tear of joint cartilage, leading to joint pain and limited movement. Osteoporosis: Loss of bone density, bones become fragile and prone to fractures. Asthma: Chronic inflammation of the respiratory tract, leading to airway constriction, increased sputum, and difficulty breathing. Autoimmune diseases: such as rheumatoid arthritis and lupus erythematosus, in which the immune system mistakenly attacks its own tissues. Obesity: Excessive accumulation of fat in the body increases the risk of a variety of chronic diseases. Mental illnesses: such as depression, anxiety, bipolar disorder, etc., affect mood and behavior. Parkinson's disease: A neurodegenerative disease that causes impaired movement and cognitive function. Liver disease: including non-alcoholic fatty liver disease, hepatitis B and C, etc., which may develop into cirrhosis or liver cancer. Cancer: Malignant tumors that can affect any part of the body.

[0093] Among them, the above health theme rules take the hypertension theme as an example. When the patient is in a resting state, the systolic blood pressure is 190 mmHg and the diastolic blood pressure is 130 mmHg. We can judge that the patient suffers from hypertension. Among them, the judgment formula for the hypertension theme is: normal blood pressure: systolic blood pressure (SBP) <120 mmHg and diastolic blood pressure (DBP) <80 mmHg; Stage 1 hypertension: SBP 120-139 mmHg or DBP 80-89 mmHg; Stage 2 hypertension: SBP ≥140 mmHg or DBP ≥90 mmHg.

[0094] The disease risk level rules for the disease themes in the embodiment of the present application are divided into six levels based on the trend of chronic disease risk changes: high risk, high risk, moderate risk, low risk, low risk concept, and past history; that is, based on the user indicator data, it can be judged whether one of these risk stages exists, and combined with the specific disease theme, the user's disease risk level label can be obtained.

[0095] In addition, after obtaining the user's disease risk level label, the embodiment of the present application can also use the big model to obtain healthy living guidance suggestions corresponding to the disease risk level label based on the user's disease risk level label.

[0096] S103: Inputting the user indicator data into a data trend prediction model to predict the indicator data trend of the user, wherein the indicator data prediction model is a trained neural network model, and the training data of the indicator data prediction model is historical indicator data and the trend of historical indicator data changing over time.

[0097] As an example, the training process of the data trend prediction model in S103 may include: obtaining the user's historical indicator data and the trend of the historical indicator data changing over time; inputting the historical indicator data and the trend of the historical indicator data changing over time as training samples into the trend initial model for training until the trend initial model reaches the target value, thereby obtaining the data trend prediction model.

[0098] In the embodiments of the present application, time series analysis, statistical models or machine learning algorithms, such as support vector machine (SVM), random forest and other models, can be used as initial models to analyze and predict the trend of user indicator data.

[0099] S104: Inputting the user indicator data into a disease risk prediction model to predict the user's disease risk, wherein the disease risk prediction model is a trained neural network model, and the training samples of the disease risk prediction model are historical indicator data and corresponding historical disease risk scores.

[0100] As an example, the training process of the disease risk prediction model in S104 may include: obtaining the user's historical indicator data and historical disease risk scores; inputting the historical indicator data and historical disease risk scores as training samples into the risk initial model for training until the risk initial model reaches the target value, thereby obtaining a disease risk prediction model.

[0101] In the embodiment of the present application, the risk prediction model can use support vector machine (SVM), BP neural network, random forest, naive Bayes and other algorithms, wherein SVM is a supervised learning algorithm, suitable for small samples, high-dimensional data, can solve the classification problems in disease monitoring and bioinformatics, especially in the prediction of chronic diseases. BP neural network is a multi-layer feedforward network, suitable for data sets with large sample sizes, with parallel processing capabilities and self-learning capabilities, and is widely used in cardiovascular disease prediction and other aspects. The random forest algorithm is suitable for processing high-dimensional data, feature missing data and unbalanced data, can provide unbiased estimates, and is commonly used in the prediction of chronic diseases such as diabetes and cardiovascular diseases. And the naive Bayes algorithm is based on Bayesian decision theory, suitable for the situation where attributes are independent of each other, is not very sensitive to missing data, is suitable for text classification and natural language processing, and can also be used for liver disease and TCM syndrome risk prediction. When the disease risk prediction is specifically carried out, the appropriate machine learning algorithm can be selected according to the characteristics and needs of the data to conduct in-depth analysis and prediction of the user's health status.

[0102] In the embodiment of the present application, the historical indicator data for training the disease risk prediction model may also include biochemical indicators: measurement values ​​of biochemical components in blood and urine; imaging data: medical imaging data such as X-rays, CT, MRI, and gene expression data: expression levels of related genes.

[0103] S105: Combine the user's disease risk level label, the user's health indicator trend and disease risk to construct a user health profile.

[0104] After obtaining the user's disease risk level label, the user's health indicator trend and disease risk, they can be combined to build a user health profile. The user health profile information in the embodiment of the present application comes from a comprehensive source, and will analyze and evaluate the user data and predict the user's future health trend, and has the ability to dynamically predict the user's future health trend.

[0105] After obtaining the user's health image, it is also possible to scientifically propose user health guidance suggestions including lifestyle, diet structure and exercise suggestions from multiple relevant angles based on the information on the user's health image. The specific process may include: setting the user's health control goal through parameters; determining the user's improvement suggestion from a set of improvement suggestions based on the health control goal, and the improvement suggestion set has a corresponding relationship with the health control goal; finally, generating a multi-faceted guidance suggestion plan based on the user's improvement suggestion, and the multi-faceted guidance suggestion plan includes a lifestyle guidance suggestion plan, a diet structure guidance suggestion plan and an exercise guidance suggestion plan.

[0106] The guidance and suggestion plan in the embodiment of the present application needs to be customized according to the specific situation of the user to ensure the practicality and effectiveness of the suggestion. At the same time, the guidance and suggestion plan also needs to be easy to understand and execute so that the user can stick to it for a long time.

[0107] It can be seen that after obtaining user data from multiple aspects, the embodiment of the present application performs preprocessing to obtain user indicator data, obtains the user's disease risk level label based on the user indicator data, and obtains the user's health indicator trend and disease risk based on the model, and finally combines the user's disease risk level label, the user's health indicator trend and disease risk to construct a multi-information user health portrait. And provide comprehensive health guidance based on the user's health picture, and finally realize personalized healthy life guidance and disease prevention suggestions.

[0108] In order to make the method provided in the embodiment of the present application clearer and easier to understand, Figure 2 A specific example is used to illustrate.

[0109] like Figure 2 As shown, the implementation may include:

[0110] S201: Define user basic indicators, health basic indicators and health extended indicators as reference indicators for defining indicators and defining health topics.

[0111] like Figure 3 As shown in the definition indicator flow chart, the embodiment of the present application is to build a user health portrait to evaluate the user's health indicator trends and disease risks. It is necessary to collect data from multiple dimensions, and it is necessary to obtain user indicator data based on the defined user basic information, basic health indicators and extended health indicators. The specific definition content is as known from the above embodiment and will not be repeated here. It is also necessary to pre-define which indicators of the above user basic indicators, basic health indicators and extended health indicators required for each health theme as reference indicators. For example, the hypertension theme needs to refer to the family medical history data in the extended health indicators and the blood pressure indicator data in the basic health indicators.

[0112] S202: Collect user data.

[0113] In the embodiments of the present application, there are many ways to collect user data. By increasing the data sources of the user's health portrait, the data accuracy of the user's health portrait can be improved. The way to collect user data can be obtained through electronic medical records, medical records, physical examination reports, manual entry, etc., or through App to collect user data in different dimensions, including but not limited to basic information of users (such as gender, age, region, etc.), and electronic medical records, physical examination reports, medical data, behavioral habits (such as product use behavior), health habits (such as exercise, diet, etc.) and health equipment data generated in different time periods.

[0114] S203: Preprocess the user data to obtain user index data.

[0115] The embodiment of the present application requires user indicator data as input data of the model, so the user data needs to be preprocessed, mainly including first performing multimodal data processing on the user data to extract key information; performing feature extraction on the key information according to the definition indicators obtained in the above S201, and extracting the user basic indicators in the user data; finally, performing data cleaning and standardization on the data in the user basic indicators to obtain user indicator data.

[0116] S204: Determine the disease theme of the user based on the user indicator data and health theme rules, where the health theme rules are rules for evaluating various types of diseases using indicator data.

[0117] After obtaining the user indicator data, the user's disease risk level label can be determined based on the user indicator data. First, the user's disease theme needs to be determined based on the user indicator data and health theme rules. The specific health theme rules are shown in the above embodiment and will not be repeated here.

[0118] S205: Substituting the user indicator data into the disease risk level rule of the disease theme to obtain the user disease risk level label, wherein the disease risk level rule is a rule for evaluating the level of each disease according to the indicator data.

[0119] After determining the user's disease theme, the current user's disease risk level can also be determined, and the user can be labeled with a disease risk level. The specific disease risk level rules are as shown in the above embodiment and will not be described in detail here.

[0120] S206: Based on the disease risk level label of the user, use the big model to obtain healthy life guidance suggestions corresponding to the disease risk level label.

[0121] After obtaining the user's disease risk level label, the corresponding large model can be used to directly obtain healthy living guidance suggestions based on the disease risk level. For example, if the user's disease risk level is intermediate hypertension, the corresponding healthy living guidance suggestions may be regular measurement of blood pressure, blood sugar, blood lipids, and adjustment of lifestyle habits.

[0122] S207: Inputting the user indicator data into a data trend prediction model to predict the indicator data trend of the user, wherein the indicator data prediction model is a trained neural network model, and the training data of the indicator data prediction model is historical indicator data and the trend of historical indicator data changing over time.

[0123] In the embodiment of the present application, not only can the disease risk level of the user be obtained, but the trend of the user's indicator data can also be predicted based on the user's indicator data, so as to better understand the user's health condition.

[0124] S208: Inputting the user indicator data into a disease risk prediction model to predict the user's disease risk, wherein the disease risk prediction model is a trained neural network model, and the training samples of the disease risk prediction model are historical indicator data and corresponding historical disease risk scores.

[0125] In the embodiment of the present application, not only can the user's disease risk level be obtained, but the user's disease risk can also be predicted based on the user's indicator data, thereby better understanding the user's health condition.

[0126] S209: Combine the user's disease risk level label, the user's health indicator trend and disease risk to construct a user health profile.

[0127] Finally, in order to enrich the amount of information in the user's health portrait, the embodiment of the present application combines the user's disease risk level label, the user's health indicator trend and disease risk to construct a user health portrait.

[0128] S210: Setting a health control target of the user through parameters, and determining an improvement suggestion for the user from an improvement suggestion set according to the health control target.

[0129] In the embodiment of the present application, a set of improvement suggestions is pre-stored, and the set of improvement suggestions has a certain correspondence with the health control target, that is, after the user sets the user's health control target through parameters, the user's improvement suggestions can be determined from the set of improvement suggestions according to the health control target and the correspondence. For example, the user's health control target is determined to be blood pressure control: the goal is to maintain it at 90-139 / 60-89mmHg. Blood sugar control: the goal is to maintain it at 3.9-6.1mmol / L. Body mass index (BMI): the goal is to drop to 18.5-23.9kg / m2. Total cholesterol: the goal is to be lower than 5.2mmol / L. Triglycerides: the goal is to be lower than 1.70mmol / L. It is recommended to control fat intake. Smoking: the goal is not to smoke. It is recommended to take smoking cessation measures. Drinking: the goal is to limit drinking. It is recommended to reduce alcohol intake. Salt amount: the goal is to reduce it to 1 salt spoon. It is recommended to reduce salt intake. For different goals, the user's improvement suggestions can be determined separately, for example, suggestions for blood pressure control: regularly measure blood pressure, blood sugar, blood lipids, and adjust living habits. Recommendations for blood sugar control: It is recommended to monitor blood sugar levels regularly. Recommendations for body mass index: It is recommended to control body weight through diet and exercise. Recommendations for total cholesterol: It is recommended to control dietary cholesterol intake.

[0130] S211: Generate a multi-faceted guidance and suggestion plan based on the user's improvement suggestions, wherein the multi-faceted guidance and suggestion plan includes a lifestyle guidance and suggestion plan, a diet structure guidance and suggestion plan, and a sports guidance and suggestion plan.

[0131] In addition, in the embodiments of the present application, various guidance and suggestion plans are generated based on the user's improvement suggestions, such as (1) Lifestyle guidance and suggestion plan: Use a health radar chart to visualize the user's health habits and provide personalized suggestions based on the radar chart results. Analyze the user's health status and point out the advantages and areas that need improvement. (2) Diet structure suggestions: Point out indicators of deficiencies in the diet, such as high salt, high sugar, high fat, etc. Provide specific dietary improvement suggestions, such as increasing the intake of vegetables and fruits and reducing processed foods. (3) Healthy exercise suggestions: Choose mild to moderate aerobic exercise, such as brisk walking, jogging, swimming, Tai Chi, etc. The purpose of exercise is to enhance metabolism and physical fitness, and it should not be excessive. Strictly implement the exercise plan formulated by the doctor. The embodiments of the present application can provide comprehensive and personalized healthy life guidance and disease prevention suggestions, provide users with scientific health guidance, and are efficient, accurate and practical.

[0132] This embodiment provides a method for constructing a user health portrait, comprehensively analyzing user health data based on big data and big models, and accurately assessing the user's health status and predicting disease risks, thereby providing personalized healthy living guidance and disease prevention recommendations.

[0133] See also Figure 4 , the embodiment of the present application provides a device 400 for constructing a user health profile, the device comprising:

[0134] The preprocessing unit 401 is used to collect user data and preprocess the user data to obtain user indicator data, wherein the user data includes basic information and behavior information of the user; wherein the basic information is data of the user's identity and health status, and the behavior information is activity or behavior data related to the user's health status;

[0135] The obtaining unit 402 is used to obtain a user disease risk level label according to the user indicator data and the disease risk level rule;

[0136] A prediction unit 403 is used to input the user indicator data into a data trend prediction model to predict the indicator data trend of the user, wherein the indicator data prediction model is a trained neural network model, and the training data of the indicator data prediction model is historical indicator data and the trend of historical indicator data changing over time;

[0137] The prediction unit 403 is further used to input the user indicator data into a disease risk prediction model to predict the user's disease risk, wherein the disease risk prediction model is a trained neural network model, and the training samples of the disease risk prediction model are historical indicator data and corresponding historical disease risk scores;

[0138] The construction unit 404 is used to combine the user's disease risk level label, the user's health indicator trend and the disease risk to construct a user health profile.

[0139] Optionally, the preprocessing unit 401 is specifically configured to:

[0140] Perform multimodal data processing on user data to extract key information;

[0141] Extract features of key information according to defined indicators, and extract user basic indicators from user data, wherein the defined indicators include user basic indicators, health basic indicators, and health extended indicators;

[0142] Perform data cleaning and standardization on the data in the user basic indicators to obtain user indicator data.

[0143] Optionally, the obtaining unit 402 is specifically configured to:

[0144] Determine the user's disease theme based on the user's indicator data and health theme rules, wherein the health theme rules are rules for evaluating various diseases using indicator data;

[0145] Substitute the user indicator data into the disease risk level rule of the disease theme to obtain the user disease risk level label, wherein the disease risk level rule is a rule for evaluating the level of various diseases based on the indicator data.

[0146] Optionally, the apparatus 400 further includes a training unit, configured to:

[0147] Obtain the user's historical indicator data and the trend of historical indicator data over time;

[0148] The historical indicator data and the trend of historical indicator data changing over time are input as training samples into the initial trend model for training until the initial trend model reaches the target value, thereby obtaining a data trend prediction model.

[0149] Optionally, the apparatus 400 further includes a training unit, configured to:

[0150] Obtain the user's historical indicator data and historical disease risk scores;

[0151] The historical indicator data and historical disease risk scores are input as training samples into the initial risk model for training until the initial risk model reaches the target value to obtain the disease risk prediction model.

[0152] Optionally, the obtaining unit 402 is further configured to:

[0153] According to the user's disease risk level label, the big model is used to obtain healthy living guidance suggestions corresponding to the disease risk level label.

[0154] Optionally, the apparatus 400 further includes a generating unit, configured to:

[0155] Set the user's health control goals through parameters;

[0156] Determining, according to the health control target, an improvement suggestion for the user from the improvement suggestion set, wherein the improvement suggestion set has a corresponding relationship with the health control target;

[0157] According to the improvement suggestions of the user, a multi-faceted guidance and suggestion plan is generated, wherein the multi-faceted guidance and suggestion plan includes a lifestyle guidance and suggestion plan, a diet structure guidance and suggestion plan, and a sports guidance and suggestion plan.

[0158] It should be noted that the specific implementation method and effect of the device 400 for constructing a user health profile can be found in the above Figure 1 or Figure 2 The relevant descriptions in the provided methods will not be repeated here.

[0159] The present application embodiment also provides an electronic device 500, such as Figure 5As shown, the device 500 includes a memory 501 and a processor 502:

[0160] The memory 501 is used to store computer programs;

[0161] The processor 502 is used to execute the above Figure 1 or Figure 2 The method provided.

[0162] In addition, the present application also provides a computer-readable storage medium, the computer-readable storage medium is used to store a computer program, and the computer program is used to execute Figure 1 or Figure 2 The method provided.

[0163] Through the description of the above implementation methods, it can be known that those skilled in the art can clearly understand that all or part of the steps in the above-mentioned embodiment method can be implemented by means of software plus a general hardware platform. Based on such an understanding, the technical solution of the present application can be embodied in the form of a software product, which can be stored in a storage medium, such as a read-only memory (ROM) / RAM, a magnetic disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network communication device such as a router) to execute the methods described in each embodiment of the present application or some parts of the embodiments.

[0164] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment. The device embodiment described above is only schematic, in which the modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules can be selected according to actual needs to achieve the goals of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative work.

[0165] The above description is merely an exemplary embodiment of the present application and is not intended to limit the protection scope of the present application.

Claims

1. A method for constructing a user health profile, characterized in that: The method comprises: Collecting user data and preprocessing the user data to obtain user indicator data, wherein the user data includes basic information and behavior information of the user; wherein the basic information is data of the user's identity and health status, and the behavior information is activity or behavior data related to the user's health status; Obtaining a user disease risk level label according to the user indicator data and disease risk level rules; Inputting the user indicator data into a data trend prediction model to predict the indicator data trend of the user, wherein the indicator data prediction model is a trained neural network model, and the training data of the indicator data prediction model is historical indicator data and the trend of the historical indicator data changing over time; Inputting the user indicator data into a disease risk prediction model to predict the disease risk of the user, wherein the disease risk prediction model is a trained neural network model, and the training samples of the disease risk prediction model are historical indicator data and corresponding historical disease risk scores; The user's disease risk level label, the user's health indicator trend and disease risk are combined to construct a user health profile.

2. The method according to claim 1, characterized in that The preprocessing of the user data to obtain user indicator data includes: Performing multimodal data processing on the user data to extract key information; Perform feature extraction on the key information according to the defined indicators, and extract user basic indicators from the user data, wherein the defined indicators include user basic indicators, health basic indicators, and health extended indicators; The data in the user basic indicators are cleaned and standardized to obtain the user indicator data.

3. The method according to claim 1, characterized in that The step of obtaining a user disease risk level label according to the user indicator data and the disease risk level rule includes: Determining the disease theme of the user based on the user indicator data and health theme rules, wherein the health theme rules are rules for evaluating various types of diseases using indicator data; Substitute the user indicator data into the disease risk level rule of the disease subject to obtain the user disease risk level label, wherein the disease risk level rule is a rule for evaluating various disease levels based on the indicator data.

4. The method according to claim 1, characterized in that The training process of the data trend prediction model includes: Obtaining historical indicator data of the user and the trend of changes in the historical indicator data over time; The historical indicator data and the trend of the historical indicator data changing over time are input as training samples into the trend initial model for training until the trend initial model reaches the target value, thereby obtaining the data trend prediction model.

5. The method according to claim 1, characterized in that The training process of the disease risk prediction model includes: Obtaining historical indicator data and historical disease risk scores of the user; The historical indicator data and historical disease risk scores are input as training samples into the initial risk model for training until the initial risk model reaches the target value, thereby obtaining the disease risk prediction model.

6. The method according to claim 1, characterized in that The method further comprises: According to the user's disease risk level label, a large model is used to obtain healthy living guidance suggestions corresponding to the disease risk level label.

7. The method according to claim 1, characterized in that The method further comprises: Setting the user's health control target through parameters; Determining, according to the health control target, an improvement suggestion for the user from an improvement suggestion set, wherein the improvement suggestion set has a corresponding relationship with the health control target; According to the improvement suggestions of the user, a multi-faceted guidance and suggestion plan is generated, wherein the multi-faceted guidance and suggestion plan includes a lifestyle guidance and suggestion plan, a diet structure guidance and suggestion plan, and a sports guidance and suggestion plan.

8. A method for constructing a user health profile, characterized in that: The method comprises: A preprocessing unit, used to collect user data and preprocess the user data to obtain user indicator data, wherein the user data includes basic information and behavior information of the user; wherein the basic information refers to data indicating the identity and health status of the user, and the behavior information refers to activity or behavior data related to the health status of the user; An obtaining unit, used to obtain a user disease risk level label according to the user indicator data and the disease risk level rule; A prediction unit, used for inputting the user indicator data into a data trend prediction model to predict the indicator data trend of the user, wherein the indicator data prediction model is a trained neural network model, and the training data of the indicator data prediction model is historical indicator data and the trend of the historical indicator data changing over time; The prediction unit is further used to input the user indicator data into a disease risk prediction model to predict the disease risk of the user, wherein the disease risk prediction model is a trained neural network model, and the training samples of the disease risk prediction model are historical indicator data and corresponding historical disease risk scores; The construction unit is used to combine the user's disease risk level label, the user's health indicator trend and disease risk to construct a user health portrait.

9. An electronic device, characterized in that: The device comprises a memory and a processor, and the electronic device is used to execute a program stored in the memory to run the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium is used to store a computer program, and the computer program is used to execute the method according to any one of claims 1 to 7.

Images