Health management method and device based on machine learning, equipment and medium

Abstract

The invention provides a health management method and device based on machine learning, equipment and a medium, and relates to the technical field of intelligent health management, and the method comprises the steps: obtaining health data of a target user in response to a prediction request for a target chronic disease triggered by the target user; according to the health data, target health data corresponding to the target chronic disease is matched from a configured mapping table of different types of chronic diseases and different health data; taking target health data in the health data as target data; inputting the target data into a pre-constructed target chronic disease auxiliary prediction model to obtain an auxiliary prediction result; determining a comprehensive auxiliary prediction result of the target chronic disease based on the auxiliary prediction result; and displaying the comprehensive auxiliary prediction result of the target chronic disease. According to the method, auxiliary prediction of various chronic diseases can be comprehensively and accurately carried out.

Description

Technical Field

[0001] This application relates to the technical field of intelligent health management. Specifically, it relates to a health management method, device, equipment, and medium based on machine learning. Background Art

[0002] With the continuous development of the medical and health field, the management and prevention of chronic diseases have become a research hotspot. In the prior art, the monitoring and management of chronic diseases mainly rely on regular physical examinations, personal health records, and doctors' clinical diagnoses. For example, diabetic patients need to regularly measure their blood glucose levels and adjust their diet and medication according to the test results; hypertensive patients need to regularly measure their blood pressure and adjust their lifestyle and drug treatment plans according to the blood pressure values.

[0003] Common chronic disease management methods include using portable health monitoring devices, such as blood glucose meters, blood pressure monitors, etc. These devices can help patients understand their own health status at any time. In addition, the applications of smartphones and wearable devices are also becoming increasingly widespread. They can collect users' physiological data and analyze it through applications to help users better manage their health.

[0004] However, there is currently a lack of technology for intelligently assisting in predicting chronic diseases based on users' health data and promptly conducting chronic disease health management. Summary of the Invention

[0005] The purpose of the embodiments of this application is to provide a health management method, device, equipment, and medium based on machine learning to solve the above problems existing in the prior art and to comprehensively and accurately assist in predicting various chronic diseases.

[0006] In a first aspect, the present invention provides a health management method based on machine learning. The method may include:

[0007] In response to a prediction request for a target chronic disease triggered by a target user, obtain the health data of the target user;

[0008] From the configured mapping table of different types of chronic diseases and different health data, match the target health data corresponding to the target chronic disease;

[0009] Use the target health data in the health data as target data;

[0010] The target data are respectively input into a pre-constructed first target chronic disease auxiliary prediction model, a second target chronic disease auxiliary prediction model and a third target chronic disease auxiliary prediction model to obtain a first auxiliary prediction result, a second auxiliary prediction result and a third auxiliary prediction result; wherein the first target chronic disease auxiliary prediction model is used to predict whether the target user is a high-risk population for the target chronic disease; the second target chronic disease auxiliary prediction model is used to predict whether the target user is a suspected population for the target chronic disease; and the third target chronic disease auxiliary prediction model is used to predict whether the target user is a confirmed population for the target chronic disease;

[0011] Determining a comprehensive auxiliary prediction result of the target chronic disease based on the first auxiliary prediction result, the second auxiliary prediction result, and the third auxiliary prediction result;

[0012] The comprehensive auxiliary prediction results of the target chronic disease are displayed.

[0013] In an optional embodiment, before inputting the target data into the pre-built first target chronic disease auxiliary prediction model, the second target chronic disease auxiliary prediction model and the third target chronic disease auxiliary prediction model, the method further includes:

[0014] From the configured mapping table of different target chronic disease auxiliary prediction models and different health data, respectively match the first health data, the second health data and the third health data corresponding to the first target chronic disease auxiliary prediction model, the second target chronic disease auxiliary prediction model and the third target chronic disease auxiliary prediction model;

[0015] The first health data, the second health data and the third health data in the target data are used as the first target data, the second target data and the third target data.

[0016] In an optional embodiment, before inputting the target data into the pre-built first target chronic disease auxiliary prediction model, the second target chronic disease auxiliary prediction model and the third target chronic disease auxiliary prediction model, the method further includes:

[0017] respectively acquiring data types of the first target data, the second target data, and the third target data;

[0018] From the configured comparison table of different data types and corresponding data processing methods, respectively match the first target processing method, the second target processing method and the third target processing method corresponding to the data types of the first target data, the second target data and the third target data;

[0019] The first target data, the second target data and the third target data are processed using the first target processing method, the second target processing method and the third target processing method to obtain processed first target data, second target data and third target data.

[0020] In an optional implementation, the target data is respectively input into a pre-constructed first target chronic disease auxiliary prediction model, a second target chronic disease auxiliary prediction model, and a third target chronic disease auxiliary prediction model, including:

[0021] Inputting the preprocessed first target data into the first target chronic disease auxiliary prediction model to obtain a first auxiliary prediction result;

[0022] Inputting the preprocessed second target data into the second target chronic disease auxiliary prediction model to obtain a second auxiliary prediction result;

[0023] The preprocessed third target data is input into the third target chronic disease auxiliary prediction model to obtain a third auxiliary prediction result.

[0024] In an optional embodiment, the method further comprises:

[0025] If the target data does not include the first health data, the second health data and / or the third health data, the auxiliary prediction result of the target chronic disease auxiliary prediction model using the corresponding health data as the target data is empty.

[0026] In an optional embodiment, determining a comprehensive auxiliary prediction result of the target chronic disease based on the first auxiliary prediction result, the second auxiliary prediction result, and the third auxiliary prediction result includes:

[0027] Obtaining the priorities of the pre-configured first auxiliary prediction result, the second auxiliary prediction result, and the third auxiliary prediction result;

[0028] The auxiliary prediction result having the highest priority and not being empty among the first auxiliary prediction result, the second auxiliary prediction result and the third auxiliary prediction result is determined as the comprehensive auxiliary prediction result.

[0029] In an optional embodiment, after determining the comprehensive auxiliary prediction result of the target chronic disease, the method further includes:

[0030] Matching the target chronic disease and the target health management measure corresponding to the comprehensive auxiliary prediction result from the configured mapping table of different target chronic diseases, different comprehensive auxiliary prediction results and different health management measures;

[0031] The target health management measure is displayed.

[0032] In a second aspect, the present invention provides a health management device based on machine learning, the device comprising:

[0033] A response unit, configured to obtain health data of a target user in response to a prediction request for a target chronic disease triggered by the target user;

[0034] A matching unit, configured to match the target health data corresponding to the target chronic disease from a configured mapping table of different types of chronic diseases and different health data; and use the target health data in the health data as the target data;

[0035] The auxiliary prediction unit is used to input the target data into a pre-constructed first target chronic disease auxiliary prediction model, a second target chronic disease auxiliary prediction model and a third target chronic disease auxiliary prediction model, respectively, to obtain a first auxiliary prediction result, a second auxiliary prediction result and a third auxiliary prediction result; wherein the first target chronic disease auxiliary prediction model is used to predict whether the target user is a high-risk population for the target chronic disease; the second target chronic disease auxiliary prediction model is used to predict whether the target user is a suspected population for the target chronic disease; the third target chronic disease auxiliary prediction model is used to predict whether the target user is a confirmed population for the target chronic disease;

[0036] a determining unit, configured to determine a comprehensive auxiliary prediction result of the target chronic disease based on the first auxiliary prediction result, the second auxiliary prediction result, and the third auxiliary prediction result;

[0037] The display unit is used to display the comprehensive auxiliary prediction results of the target chronic disease.

[0038] In a third aspect, the present invention provides an electronic device, the electronic device comprising a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus;

[0039] Memory, used to store computer programs;

[0040] The processor is used to implement the method described in any of the above-mentioned embodiments when executing the program stored in the memory.

[0041] In a fourth aspect, the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method described in any one of the aforementioned embodiments is implemented.

[0042] This application improves the accuracy of auxiliary prediction results by using different processing methods for different data and combining multiple target chronic disease auxiliary prediction models; the automated and intelligent auxiliary prediction method reduces the workload of doctors and reduces medical costs; the rapid auxiliary prediction shortens the waiting time of users and improves the efficiency of medical services; the comprehensive auxiliary prediction results for users are combined with the specific circumstances of users to provide comprehensive and personalized health management measures, thereby improving the pertinence and effectiveness of health management measures;

[0043] The machine learning-based health management method of this application can be applied to a wide range of medical and health industries through cooperation with technology manufacturers, such as integrating it with hospital information systems, disease prevention and control systems, basic public health information systems, commercial physical examinations and other scenarios, and applying it to hospital intelligent all-in-one machines to achieve automated auxiliary prediction of chronic diseases, assist in the implementation of medical and health behaviors in the above business scenarios, reduce prediction error rates, and improve the effectiveness of chronic disease health management.

[0044] This application obtains a large amount of health data of target users and enriches the data basis of the prediction model. Through data cleaning, standardization, data mapping and other methods, it realizes the preprocessing of the acquired data and improves the accuracy and consistency of the data. At the same time, it uses bag-of-words model, convolutional neural network and other technologies to effectively extract features of unstructured text and image data, providing strong support for subsequent auxiliary prediction model training.

[0045] This application achieves comprehensive and accurate auxiliary prediction of chronic diseases, providing medical workers with a reliable basis for decision-making; based on the auxiliary prediction results and risk levels, targeted health management measures are formulated, and specific lifestyle adjustment suggestions are provided to users. This not only helps to improve the accuracy of auxiliary prediction, but also promotes users' self-management awareness and enhances the prevention effect.

[0046] This application presents the auxiliary prediction results to users and medical workers in a visual manner, simplifying the process for doctors to obtain user-assisted prediction results. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments of the present application will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0048] Figure 1 A flowchart of a health management method based on machine learning provided in an embodiment of the present application;

[0049] Figure 2 A schematic diagram of the structure of a health management device based on machine learning provided in an embodiment of the present application;

[0050] Figure 3 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0051] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0052] The health management method based on machine learning provided in the embodiment of the present application can be applied in a server or in a terminal with strong computing power. The server can be a physical server, a server cluster or a distributed system composed of multiple physical servers, or a cloud server providing cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content distribution networks (Content Delivery Network, CDN), and basic cloud computing services such as big data and artificial intelligence platforms. The terminal can be a user equipment (User Equipment, UE) such as a mobile phone, a smart phone, a laptop, a digital broadcast receiver, a personal digital assistant (PDA), a tablet computer (PAD), a handheld device, a vehicle-mounted device, a wearable device, a computing device or other processing equipment connected to a wireless modem, a mobile station (Mobile Station, MS), a mobile terminal (Mobile Terminal), etc. The terminal and the server can be directly or indirectly connected by a wired or wireless communication method, and this application is not limited here.

[0053] The preferred embodiments of the present application are described below in conjunction with the drawings in the specification. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present application, and are not used to limit the present application. In addition, the embodiments and features in the embodiments of the present application may be combined with each other if there is no conflict.

[0054] Figure 1 A flowchart of a health management method based on machine learning provided in an embodiment of the present application. Figure 1 As shown, the method may include:

[0055] S110. In response to a prediction request for a target chronic disease triggered by a target user, obtain health data of the target user; match target health data corresponding to the target chronic disease from a configured mapping table of different types of chronic diseases and different health data; and use the target health data in the health data as target data.

[0056] In an embodiment of the present application, the target user may trigger prediction requests for multiple target chronic diseases at the same time, or may only trigger a prediction request for one target chronic disease; specifically, the target chronic diseases include hypertension, diabetes, chronic kidney disease, and chronic obstructive pulmonary disease.

[0057] In an embodiment of the present application, health data includes: personal data and health test data; wherein, personal data includes: age, gender, past medical history, height, weight, BMI, blood pressure, pulse, family history, living habits, living environment and occupational exposure history, etc.; health test data includes: clinical test data, imaging test data and genome test data; wherein, clinical test data includes medical records, test data and physical examination data, etc.; imaging test data includes X-rays, CT scans, MRI and ultrasound images, etc.; genome test data includes gene sequences and gene expression spectra, etc.

[0058] In an embodiment of the present application, health data can be obtained from the electronic medical record system of a medical institution, the public health monitoring network, the family doctor contract service record, the disease control reporting system of a disease prevention and control agency, and relevant business information of maternal and child health service agencies and commercial physical examination agencies after authorization by the target user.

[0059] In an embodiment of the present application, after the health data is acquired, the acquired health data needs to be cleaned and standardized to remove noise and missing values.

[0060] In the embodiments of the present application, different types of chronic diseases require different health data, so it is necessary to match the required health data according to the type of target chronic disease; for example, hypertension mainly requires blood pressure data, medical history and family history, etc.; while diabetes mainly requires blood sugar and glycated hemoglobin, etc.

[0061] In an embodiment of the present application, the target health data corresponding to the target chronic disease includes all health data required by the first target chronic disease auxiliary prediction model, the second target chronic disease auxiliary prediction model and the third target chronic disease auxiliary prediction model corresponding to the target chronic disease.

[0062] In the embodiment of the present application, after determining the target data, it is also necessary to:

[0063] From the mapping table of different configured target chronic disease auxiliary prediction models and different health data, match the first health data, second health data and third health data corresponding to the first target chronic disease auxiliary prediction model, the second target chronic disease auxiliary prediction model and the third target chronic disease auxiliary prediction model respectively; use the first health data, the second health data and the third health data in the target data as the first target data, the second target data and the third target data.

[0064] In an embodiment of the present application, different auxiliary prediction models for the same target chronic disease may require the same or different target data; for example, when the target chronic disease is hypertension, the first target chronic disease auxiliary prediction model may only require eating habits, family history and BMI index; but the third target chronic disease auxiliary prediction model requires systolic blood pressure and diastolic blood pressure; therefore, it is necessary to match the required health data (i.e., the input data of the auxiliary prediction model) in advance according to different target chronic disease auxiliary prediction models.

[0065] In an embodiment of the present application, the health data required by any target chronic disease auxiliary prediction model may not be acquired, that is, the target data does not contain the health data required by a certain target chronic disease auxiliary prediction model; at this time, the auxiliary prediction result of the target chronic disease auxiliary prediction model with the corresponding health data as the target data is empty; for example, the input data of the first target chronic disease auxiliary prediction model is the blood pressure value, and the health data of the target user does not contain the blood pressure value. At this time, the first auxiliary prediction result of the first target chronic disease auxiliary prediction model with the blood pressure value as input is empty.

[0066] In the embodiment of the present application, after determining the first target data, the second target data and the third target data or determining the target data, or after obtaining the health data of the target user, the target data needs to be processed, and the specific steps include:

[0067] Respectively obtain the data types of the first target data, the second target data, and the third target data; respectively match the first target processing method, the second target processing method, and the third target processing method corresponding to the data types of the first target data, the second target data, and the third target data from a configured comparison table of different data types and corresponding data processing methods; use the first target processing method, the second target processing method, and the third target processing method to process the first target data, the second target data, and the third target data to obtain processed first target data, second target data, and third target data.

[0068] In the embodiment of the present application, since the health data includes test data and personal data, and the processing methods of different test data and different personal data are not the same, different processing methods are mapped in advance for data of different data types or target data.

[0069] In the embodiment of the present application, data processing includes:

[0070] Standardize the nouns and formats of target data with multiple nouns, establish corresponding connections between data from different sources through mapping relationships, so that the nouns and formats of data referring to the same object are unified, so that the differential data can be uniformly converted into a data format that meets industry or institutional standards, thereby ensuring the accuracy and consistency of the data; for example, different testing devices have different nouns for blood sugar, including fasting blood sugar or blood sugar, and the nouns and formats of fasting blood sugar or blood sugar are unified;

[0071] For laboratory test data in clinical test data, thresholds were set according to medical standards to convert continuous variables into binary or categorical variables;

[0072] For personal data such as past medical history, family history, living habits, living environment, occupational exposure history, medical records, and test reports, a language model pre-trained using the bag-of-words (BoW) model is used to encode unstructured text and extract meaningful screening model features;

[0073] For image detection data, convolutional neural networks (CNNs) are used to automatically learn hierarchical feature representations from images of image detection data.

[0074] S120. Input the target data into the pre-built first target chronic disease auxiliary prediction model, second target chronic disease auxiliary prediction model and third target chronic disease auxiliary prediction model, respectively, to obtain a first auxiliary prediction result, a second auxiliary prediction result and a third auxiliary prediction result.

[0075] In the embodiment of the present application, the target data is respectively input into the pre-built first target chronic disease auxiliary prediction model, the second target chronic disease auxiliary prediction model and the third target chronic disease auxiliary prediction model, including:

[0076] The preprocessed first target data is input into the first target chronic disease auxiliary prediction model to obtain the first auxiliary prediction result; the preprocessed second target data is input into the second target chronic disease auxiliary prediction model to obtain the second auxiliary prediction result; the preprocessed third target data is input into the third target chronic disease auxiliary prediction model to obtain the third auxiliary prediction result.

[0077] In an embodiment of the present application, the first target chronic disease auxiliary prediction model is used to predict whether the target user is a high-risk group for the target chronic disease; the second target chronic disease auxiliary prediction model is used to predict whether the target user is a suspected group for the target chronic disease; and the third target chronic disease auxiliary prediction model is used to predict whether the target user is a confirmed group for the target chronic disease.

[0078] In another embodiment of the present application, the method may further include:

[0079] The preprocessed first target data is input into the first target chronic disease auxiliary prediction model to obtain the first auxiliary prediction result; if the first auxiliary prediction result is that the target user is a high-risk group for the target chronic disease, the preprocessed second target data is input into the second target chronic disease auxiliary prediction model to obtain the second auxiliary prediction result; if the second auxiliary prediction result is that the target user is a suspected group for the target chronic disease, the preprocessed third target data is input into the third target chronic disease auxiliary prediction model to obtain the third auxiliary prediction result; the auxiliary prediction results are obtained by hierarchical means to save computing resources and avoid waste of resources.

[0080] In an embodiment of the present application, a method for constructing a first target chronic disease auxiliary prediction model includes:

[0081] Obtain data on target patients with various chronic diseases of different ages, genders, nationalities, regions and lifestyles from public databases or health records, questionnaires, health monitoring equipment, and medical records, including data of various types and attributes such as age, gender, disease history, family history, BMI, waist circumference, smoking habits, eating habits, exercise and lifestyle habits, job type, living environment, biomarker data, and physiological indicator data;

[0082] Utilize statistical methods and correlation analysis models, regression analysis models, cluster analysis models, and factor analysis models to conduct in-depth analysis of the above data to determine the connections and patterns between different types of data and the target chronic disease, thereby determining the high-risk factors for the target chronic disease and the potential connections and patterns between the target chronic disease and the high-risk factors;

[0083] On the basis of the above analysis, the coupling coefficients between high-risk factors and various chronic diseases are determined to quantify the degree of association between high-risk factors and target chronic diseases; based on the coupling coefficients of all high-risk factors and various chronic diseases, a coupling relationship map between target chronic diseases and high-risk factors is constructed; based on the coupling relationship map between target chronic diseases and high-risk factors combined with the expert database, a high-risk population identification rule base is determined; based on the high-risk population identification rule base, an auxiliary prediction model for the first target chronic disease is constructed.

[0084] In the embodiment of the present application, the determination of each rule in the high-risk population identification rule base must comply with the following principles:

[0085] Scientificity: Rulemaking must be based on scientific research methods and reliable data sources to ensure the accuracy and effectiveness of the rules; Comprehensiveness: All possible high-risk factors and types of chronic diseases should be considered to ensure the comprehensiveness and applicability of the rules; Operability: The rules should be simple and clear, easy to understand and operate, so as to be convenient for use in practical applications; Dynamism: The rules should have a certain degree of flexibility and adjustability to adapt to new scientific discoveries and practical needs.

[0086] In an embodiment of the present application, each rule in the high-risk population identification rule base includes conditions such as the name, quantity, type, and size of the coupling coefficient of the high-risk factor, as well as the criteria and identification results for individuals to be identified as high-risk groups when these conditions are met; each rule in the high-risk population identification rule base is verified using actual cases; the verification results show that each rule has high accuracy and reliability, and can provide strong support for chronic disease prevention work.

[0087] Specifically, physiological indicator data refers to parameters that reflect the physiological function or state of the human body; for example, blood pressure, blood sugar, heart rate, respiratory rate, etc. Physiological indicator data are usually measured and monitored by medical equipment and can reflect the patient's overall health status and disease status.

[0088] Medical history refers to the diseases that the patient has suffered in the past, including the type of disease, diagnosis time, treatment, etc. Family history refers to the diseases that the patient's immediate family members have suffered. These data can reflect the patient's disease background and the impact of genetic factors on the occurrence and development of the disease.

[0089] Lifestyle and environmental factors are important factors that affect human health and the occurrence of diseases. Common lifestyle data include smoking, drinking, diet, exercise, sleep, psychological emotions, etc.; environmental factor data include living environment, occupational exposure, etc. These data can be obtained through questionnaires, health monitoring equipment, etc., which can assist in predicting patients' health risks and analyzing the impact of lifestyle and environmental factors on the occurrence, development and intervention control of diseases.

[0090] Biomarkers refer to substances in organisms or biological samples that can reflect a certain biological state or pathological process; for example, proteins, genes, metabolites, etc. The content or activity changes of biomarkers in biological samples are used to characterize the patient's physiological state, disease progression and treatment effect.

[0091] In the embodiments of the present application, correlation analysis is used to evaluate whether there is a statistical correlation between high-risk factors and chronic diseases. By calculating correlation coefficients and performing chi-square tests, etc., it can be determined which high-risk factors are closely related to the occurrence and development of chronic diseases; regression analysis is used to quantify the impact of high-risk factors on the probability or severity of chronic diseases. By establishing a regression model, the effect size of high-risk factors can be estimated and its statistical significance can be evaluated; cluster analysis is used to group high-risk factors or chronic diseases with similar characteristics. Through cluster analysis, potential associations between high-risk factors and their common mechanisms of action in different chronic diseases can be discovered; factor analysis is used to simplify the data structure and extract the main high-risk factors that affect the occurrence of chronic diseases; multiple related high-risk factors are summarized into a few independent factors, so as to more clearly understand the impact on chronic diseases.

[0092] In the embodiments of the present application, the coupling coefficient is a quantitative indicator used to describe the degree of association between high-risk factors and chronic diseases. By calculating the coupling coefficient, the contribution of different high-risk factors to the risk of chronic diseases can be evaluated, and which factors are key targets for the prevention and treatment of chronic diseases can be determined.

[0093] In the embodiment of the present application, the preprocessed first target data is input into the first target chronic disease auxiliary prediction model to obtain a first auxiliary prediction result, including:

[0094] The preprocessed first target data is matched with each rule in the first target chronic disease auxiliary prediction model, including the number, type, coupling coefficient, etc. of high-risk factors, one by one; if the matching result meets the conditions configured in the rule, the first auxiliary prediction result is that the target user is a high-risk group for the target chronic disease.

[0095] In an embodiment of the present application, a method for constructing a second target chronic disease auxiliary prediction model includes:

[0096] Obtain data on target patients with various chronic diseases of different ages, genders, nationalities, regions and lifestyles from public databases or health records, questionnaires, health monitoring equipment, and medical records, including data of various types and attributes such as age, gender, disease history, family history, BMI, waist circumference, smoking habits, eating habits, exercise and lifestyle habits, job type, living environment, biomarker data, and physiological indicator data;

[0097] Based on the data obtained above, determine the qualitative screening factors of each target chronic disease and analyze the coupling relationship between the qualitative screening factors and chronic diseases;

[0098] According to the analysis results, key qualitative screening factors with strong correlation and high significance with chronic diseases are screened out; based on the key qualitative screening factors, a logical framework is constructed to clarify which qualitative factors or combinations of factors can be used to indicate or predict a certain chronic disease; reasonable thresholds are set for each screening qualitative factor or factor combination, and qualitative screening rules are determined at the same time, that is, under what conditions the relevant qualitative screening factors or combination conditions meet the qualitative screening requirements; combined with clinical experience and expert opinions, the preliminary rules are adjusted and optimized to obtain the second target chronic disease auxiliary prediction model.

[0099] In the present application example, the coupling relationship between qualitative screening factors and chronic diseases is analyzed, including:

[0100] Correlation analysis: The correlation between each qualitative screening factor and chronic diseases was evaluated by statistical methods (such as chi-square test, Spearman rank correlation analysis, etc.); these analysis results revealed which factors had significant statistical associations with chronic diseases;

[0101] Importance ranking: Use machine learning algorithms (such as random forests, logistic regression, etc.) to rank the factors by importance to determine which factors contribute most to the prediction of chronic diseases. These ranking results help identify key screening factors.

[0102] Calculation of coupling coefficient: Based on the above analysis results, the coupling coefficient is determined by comprehensively considering the strength of correlation, importance ranking, and clinical practice experience; specifically, the contribution of each factor can be quantified into a specific value, namely the coupling coefficient, by using weighted summation, product method, or probability calculation method based on Bayesian network. The coupling coefficient reflects the closeness of the relationship between the screening factor and the chronic disease.

[0103] In the embodiment of the present application, the coupling relationship between the qualitative screening factors and chronic diseases includes:

[0104] Direct association: Some screening factors (such as specific signs and symptoms) directly point to a certain chronic disease, such as dizziness, headache and other symptoms in patients with hypertension; Indirect impact: Some screening factors may indirectly cause chronic diseases by affecting other physiological or psychological processes, such as bad living habits (such as smoking and drinking) may increase the risk of cardiovascular disease; Combination effect: The combination of multiple screening factors may jointly determine the probability of occurrence of chronic diseases. This combination effect is particularly important in the screening of complex chronic diseases (such as diabetes and cancer).

[0105] In the embodiment of the present application, the construction of the third target chronic disease auxiliary prediction model includes:

[0106] Obtain test and examination data of target patients with each target chronic disease of different ages, genders, nationalities, regions and living habits from public databases or health archives, questionnaires, health monitoring equipment, and medical records, especially data with quantitative characteristics, such as blood cholesterol levels, blood sugar concentrations, blood pressure readings, nodule sizes in images, etc.; analyze the coupling relationship between quantitative screening factors and chronic diseases, and extract the coupling coefficients between various chronic diseases and quantitative screening factors; determine qualitative screening rules based on the coupling relationship between quantitative screening factors and chronic diseases and the coupling coefficients; and construct a third target chronic disease auxiliary prediction model based on qualitative screening rules.

[0107] In the embodiment of the present application, the determination of the qualitative screening rules includes:

[0108] The key screening factors are screened according to the size of the coupling coefficient; reasonable thresholds or ranges are set for the key screening factors based on clinical experience and expert opinions.

[0109] In the examples of the present application, the coupling relationship between the quantitative screening factors and chronic diseases is analyzed, including:

[0110] Use statistical methods (such as mean, standard deviation, coefficient of variation, etc.) to describe the distribution characteristics of the data, and use correlation analysis (such as Pearson correlation coefficient, Spearman rank correlation coefficient, etc.) to evaluate the linear or nonlinear relationship between screening factors and chronic diseases; use regression analysis (such as linear regression, logistic regression, etc.) to explore the predictive role of screening factors on the risk of chronic diseases, and use machine learning algorithms (such as decision trees, random forests, etc.) to explore complex nonlinear relationships.

[0111] In the embodiments of the present application, there is a direct correlation between the quantitative screening factor and the results of the quantitative screening of chronic diseases. Specifically, the numerical changes in the screening factor can reflect the changing trend of the patient's chronic disease risk. For example, as the cholesterol level in the blood increases, the patient's risk of cardiovascular disease will also increase accordingly.

[0112] In actual applications, when the target data of the target user exceeds or falls below these thresholds, the screening rules are triggered, prompting the target user to have reached the threshold for diagnosing a certain chronic disease. For example, the lung function index FEV1 / FVC is used as a quantitative screening factor for COPD, the coupling coefficient is 1, and the threshold of the screening factor is set to 70%. The rule for quantitative screening of COPD with FEV1 / FVC<70% is set. When the rule for quantitative screening of COPD is met, the system outputs the third auxiliary prediction result as "diagnosed population", otherwise it is "undiagnosed population". In clinical practice, FEV1 / FVC<70% is the gold standard for the diagnosis of COPD and is widely used in the early identification, diagnosis and disease management of COPD.

[0113] S130. Determine a comprehensive auxiliary prediction result of the target chronic disease based on the first auxiliary prediction result, the second auxiliary prediction result, and the third auxiliary prediction result; and display the comprehensive auxiliary prediction result of the target chronic disease.

[0114] In the embodiment of the present application, based on the first auxiliary prediction result, the second auxiliary prediction result and the third auxiliary prediction result, a comprehensive auxiliary prediction result of the target chronic disease is determined, including:

[0115] Obtain the priorities of the pre-configured first auxiliary prediction result, the second auxiliary prediction result, and the third auxiliary prediction result; and determine the auxiliary prediction result with the highest priority and not empty among the first auxiliary prediction result, the second auxiliary prediction result, and the third auxiliary prediction result as the comprehensive auxiliary prediction result.

[0116] In the embodiment of the present application, the priority order of the first auxiliary prediction result, the second auxiliary prediction result and the third auxiliary prediction result is: the third auxiliary prediction result is greater than the second auxiliary prediction result; the second auxiliary prediction result is greater than the first auxiliary prediction result.

[0117] In an embodiment of the present application, when the third auxiliary prediction result has been obtained and is not empty, the third auxiliary prediction result is directly used as the comprehensive auxiliary prediction result; if the third auxiliary prediction result is empty, the second auxiliary prediction result is used as the comprehensive auxiliary prediction result; if the second auxiliary prediction result is empty, the first auxiliary prediction result is used as the comprehensive auxiliary prediction result.

[0118] For example, when the third auxiliary prediction result is not empty, and the third auxiliary prediction result is that the target user is a confirmed population or an unconfirmed population of the target chronic disease, then the comprehensive auxiliary prediction result is that the target user is a confirmed population or an unconfirmed population of the target chronic disease; when the third auxiliary prediction result is empty, the second auxiliary prediction result is that the target user is a suspected population or an unsuspected population of the target chronic disease.

[0119] In the embodiment of the present application, after determining the comprehensive auxiliary prediction result of the target chronic disease, the method further includes:

[0120] From the configured mapping table of different target chronic diseases, different comprehensive auxiliary prediction results and different health management measures, match the target chronic diseases and the target health management measures corresponding to the comprehensive auxiliary prediction results; and display the target health management measures.

[0121] In the embodiment of the present application, different health management measures are matched for different comprehensive auxiliary prediction results and different target chronic diseases, including:

[0122] Obtain target data corresponding to an auxiliary prediction model with a higher priority than the comprehensive auxiliary prediction result; display the corresponding target data so that the target user can obtain the corresponding target data and perform an auxiliary prediction with a higher priority.

[0123] For example, when the comprehensive auxiliary prediction result is the first auxiliary prediction result, and the first auxiliary prediction result is that the target user is a high-risk group for the target chronic disease, the health management measure is to recommend the target user to obtain the second target data corresponding to the second auxiliary prediction model and the third target data corresponding to the third auxiliary prediction model, so as to perform the second auxiliary prediction and the third auxiliary prediction of the target chronic disease, so as to obtain more accurate results.

[0124] In an embodiment of the present application, health management measures include: prompting the target user to complete screening items, test items and examination items related to each target chronic disease; and providing the target user with follow-up and other services related to the corresponding target chronic disease, as well as the service frequency, number of services and service start time of each service.

[0125] In an embodiment of the present application, when the target user has multiple target chronic diseases, the health management measures corresponding to each target chronic disease are integrated, arranged and optimized, including: automatically eliminating inappropriate examination items based on the contraindications of the examination or test items. For example, if it is clear that the patient is hypertensive, he needs to undergo regular oral glucose tolerance tests to screen for diabetes. However, if the patient has hypertension and diabetes, the oral glucose tolerance test item needs to be eliminated to ensure the stability and safety of the diabetes condition; based on the service items recommended for multiple diseases, the same service items are deduplicated and merged to ensure the rights and interests of the target users and the integrity of the plan.

[0126] In the embodiment of the present application, after obtaining the health management measures of the target user, the method further includes:

[0127] By personalizing the health management measures for target users and combining the collected health data of target users, we provide them with targeted health advice and lifestyle adjustment suggestions, including dietary advice, exercise advice, physical sign monitoring, mental health, weight control, Chinese medicine conditioning and other dimensions for comprehensive health education information.

[0128] In the embodiment of the present application, the health management measures for the target user are personalized, including:

[0129] Use statistical models or machine learning algorithms to evaluate the risk level and risk factors of the target chronic disease to the target user's health risks; obtain the target user's health management goals; determine personalized health management measures based on the health management goals and risk level and risk factors; adjust the target user's health management measures based on the personalized health management measures.

[0130] In an embodiment of the present application, the generated target health management measures and comprehensive auxiliary prediction results of the target chronic disease are sent to the terminal corresponding to the target user or the terminal of the medical worker corresponding to the target user for visual display, so that the medical worker can intuitively obtain the comprehensive auxiliary prediction results and further target health management measures of the target user. At the same time, each medical worker provides suggestions for further examinations and obtains feedback from each medical worker on the target health management measures and comprehensive auxiliary prediction results, and corrects and optimizes the responsive auxiliary prediction model based on the feedback to improve the accuracy of the model.

[0131] Corresponding to the above method, the embodiment of the present application also provides a health management device based on machine learning, such as Figure 2 As shown, the health management device based on machine learning includes:

[0132] A response unit 210 is used to obtain health data of a target user in response to a prediction request for a target chronic disease triggered by the target user;

[0133] A matching unit 220 is used to match target health data corresponding to a target chronic disease from a configured mapping table of different types of chronic diseases and different health data; and use the target health data in the health data as target data;

[0134] The auxiliary prediction unit 230 is used to input the target data into the pre-built first target chronic disease auxiliary prediction model, the second target chronic disease auxiliary prediction model and the third target chronic disease auxiliary prediction model, respectively, to obtain the first auxiliary prediction result, the second auxiliary prediction result and the third auxiliary prediction result; wherein the first target chronic disease auxiliary prediction model is used to predict whether the target user is a high-risk group for the target chronic disease; the second target chronic disease auxiliary prediction model is used to predict whether the target user is a suspected group for the target chronic disease; the third target chronic disease auxiliary prediction model is used to predict whether the target user is a confirmed group for the target chronic disease;

[0135] A determination unit 240, configured to determine a comprehensive auxiliary prediction result of the target chronic disease based on the first auxiliary prediction result, the second auxiliary prediction result, and the third auxiliary prediction result;

[0136] The display unit 250 is used to display the comprehensive auxiliary prediction results of the target chronic disease.

[0137] The functions of each functional unit of the health management device based on machine learning provided in the above embodiments of the present application can be achieved through the above-mentioned method steps. Therefore, the specific working process and beneficial effects of each unit in the health management device based on machine learning provided in the embodiments of the present application will not be repeated here.

[0138] The present application also provides an electronic device, such as Figure 3 As shown, it includes a processor 310 , a communication interface 320 , a memory 330 and a communication bus 340 , wherein the processor 310 , the communication interface 320 , and the memory 330 communicate with each other via the communication bus 340 .

[0139] Memory 330, for storing computer programs;

[0140] The processor 310 is used to execute the program stored in the memory 330 to implement the following steps:

[0141] In response to a prediction request for a target chronic disease triggered by a target user, obtaining health data of the target user;

[0142] Match the target health data corresponding to the target chronic disease from the configured mapping table of different types of chronic diseases and different health data;

[0143] Using target health data in the health data as target data;

[0144] The target data are respectively input into the pre-constructed first target chronic disease auxiliary prediction model, the second target chronic disease auxiliary prediction model and the third target chronic disease auxiliary prediction model to obtain the first auxiliary prediction result, the second auxiliary prediction result and the third auxiliary prediction result; wherein the first target chronic disease auxiliary prediction model is used to predict whether the target user is a high-risk population for the target chronic disease; the second target chronic disease auxiliary prediction model is used to predict whether the target user is a suspected population for the target chronic disease; the third target chronic disease auxiliary prediction model is used to predict whether the target user is a confirmed population for the target chronic disease;

[0145] Determining a comprehensive auxiliary prediction result of the target chronic disease based on the first auxiliary prediction result, the second auxiliary prediction result, and the third auxiliary prediction result;

[0146] The comprehensive auxiliary prediction results of the target chronic diseases are displayed.

[0147] The communication bus mentioned above can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The communication bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus.

[0148] The communication interface is used for communication between the above electronic device and other devices.

[0149] The memory may include a random access memory (RAM) or a non-volatile memory (NVM), such as at least one disk memory. Optionally, the memory may also be at least one storage device located away from the aforementioned processor.

[0150] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0151] The implementation methods and beneficial effects of the components of the electronic device in the above embodiments to solve the problems can be seen in Figure 1 The various steps in the illustrated embodiment are implemented, therefore, the specific working process and beneficial effects of the electronic device provided by the embodiment of the present application are not repeated here.

[0152] In another embodiment provided in the present application, a computer-readable storage medium is also provided, in which instructions are stored. When the computer-readable storage medium is executed on a computer, the computer executes the health management method based on machine learning described in any of the above embodiments.

[0153] In another embodiment provided in the present application, a computer program product comprising instructions is also provided, which, when executed on a computer, enables the computer to execute the health management method based on machine learning described in any of the above embodiments.

[0154] Those skilled in the art will appreciate that the embodiments in the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt a complete hardware embodiment, a complete software embodiment, or a form of an embodiment combining software and hardware. Moreover, the present application may adopt a form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0155] Embodiments of the present application are described with reference to the flowcharts and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, and the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing device generate a means for implementing the functions specified in one flow Figure 1 one flow or multiple flows and / or blocks Figure 1 or multiple blocks.

[0156] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including an instruction means that implements the functions specified in one flow Figure 1 one flow or multiple flows and / or blocks Figure 1 or multiple blocks.

[0157] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, so that the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one flow Figure 1 one flow or multiple flows and / or blocks Figure 1 or multiple blocks.

[0158] Although the preferred embodiments in the embodiments of the present application have been described, those skilled in the art can make additional changes and modifications once they know the basic creative concepts. Therefore, the appended claims are intended to be construed as including the preferred embodiments and all changes and modifications falling within the scope of the embodiments of the present application.

[0159] Obviously, those skilled in the art can make various changes and variations to the embodiments in the embodiments of the present application without departing from the spirit and scope of the embodiments in the embodiments of the present application. Thus, if these modifications and variations of the embodiments in the embodiments of the present application fall within the scope of the claims of the embodiments of the present application and their equivalent technologies, the embodiments of the present application are also intended to include these changes and variations.

Claims

1. A health management method based on machine learning, characterized in that: The method comprises: In response to a prediction request for a target chronic disease triggered by a target user, obtaining health data of the target user; Matching the target health data corresponding to the target chronic disease from the configured mapping table of different types of chronic diseases and different health data; using target health data in the health data as target data; The target data are respectively input into a pre-constructed first target chronic disease auxiliary prediction model, a second target chronic disease auxiliary prediction model and a third target chronic disease auxiliary prediction model to obtain a first auxiliary prediction result, a second auxiliary prediction result and a third auxiliary prediction result; wherein the first target chronic disease auxiliary prediction model is used to predict whether the target user is a high-risk population for the target chronic disease; the second target chronic disease auxiliary prediction model is used to predict whether the target user is a suspected population for the target chronic disease; and the third target chronic disease auxiliary prediction model is used to predict whether the target user is a confirmed population for the target chronic disease; Determining a comprehensive auxiliary prediction result of the target chronic disease based on the first auxiliary prediction result, the second auxiliary prediction result, and the third auxiliary prediction result; The comprehensive auxiliary prediction results of the target chronic disease are displayed.

2. The method according to claim 1, characterized in that Before inputting the target data into the pre-built first target chronic disease auxiliary prediction model, the second target chronic disease auxiliary prediction model and the third target chronic disease auxiliary prediction model, the method further includes: From the configured mapping table of different target chronic disease auxiliary prediction models and different health data, respectively match the first health data, the second health data and the third health data corresponding to the first target chronic disease auxiliary prediction model, the second target chronic disease auxiliary prediction model and the third target chronic disease auxiliary prediction model; The first health data, the second health data and the third health data in the target data are used as the first target data, the second target data and the third target data.

3. The method according to claim 2, characterized in that Before inputting the target data into the pre-built first target chronic disease auxiliary prediction model, the second target chronic disease auxiliary prediction model and the third target chronic disease auxiliary prediction model, the method further includes: respectively acquiring data types of the first target data, the second target data, and the third target data; From the configured comparison table of different data types and corresponding data processing methods, respectively match the first target processing method, the second target processing method and the third target processing method corresponding to the data types of the first target data, the second target data and the third target data; The first target data, the second target data and the third target data are processed using the first target processing method, the second target processing method and the third target processing method to obtain processed first target data, second target data and third target data.

4. The method according to claim 3, characterized in that The target data are respectively input into a pre-built first target chronic disease auxiliary prediction model, a second target chronic disease auxiliary prediction model and a third target chronic disease auxiliary prediction model, including: Inputting the preprocessed first target data into the first target chronic disease auxiliary prediction model to obtain a first auxiliary prediction result; Inputting the preprocessed second target data into the second target chronic disease auxiliary prediction model to obtain a second auxiliary prediction result; The preprocessed third target data is input into the third target chronic disease auxiliary prediction model to obtain a third auxiliary prediction result.

5. The method according to claim 4, characterized in that The method further comprises: If the target data does not include the first health data, the second health data and / or the third health data, the auxiliary prediction result of the target chronic disease auxiliary prediction model using the corresponding health data as the target data is empty.

6. The method according to claim 5, characterized in that Determining a comprehensive auxiliary prediction result of the target chronic disease based on the first auxiliary prediction result, the second auxiliary prediction result, and the third auxiliary prediction result includes: Obtaining the priorities of the pre-configured first auxiliary prediction result, the second auxiliary prediction result, and the third auxiliary prediction result; The auxiliary prediction result having the highest priority and not being empty among the first auxiliary prediction result, the second auxiliary prediction result and the third auxiliary prediction result is determined as the comprehensive auxiliary prediction result.

7. The method according to claim 1, characterized in that After determining the comprehensive auxiliary prediction result of the target chronic disease, the method further includes: From the configured mapping table of different target chronic diseases, different comprehensive auxiliary prediction results and different health management measures, match the target chronic disease and the target health management measure corresponding to the comprehensive auxiliary prediction result; The target health management measure is displayed.

8. A health management device based on machine learning, characterized in that: The device comprises: A response unit, configured to obtain health data of a target user in response to a prediction request for a target chronic disease triggered by the target user; A matching unit, configured to match the target health data corresponding to the target chronic disease from a configured mapping table of different types of chronic diseases and different health data; and use the target health data in the health data as the target data; The auxiliary prediction unit is used to input the target data into a pre-constructed first target chronic disease auxiliary prediction model, a second target chronic disease auxiliary prediction model and a third target chronic disease auxiliary prediction model, respectively, to obtain a first auxiliary prediction result, a second auxiliary prediction result and a third auxiliary prediction result; wherein the first target chronic disease auxiliary prediction model is used to predict whether the target user is a high-risk population for the target chronic disease; the second target chronic disease auxiliary prediction model is used to predict whether the target user is a suspected population for the target chronic disease; the third target chronic disease auxiliary prediction model is used to predict whether the target user is a confirmed population for the target chronic disease; a determining unit, configured to determine a comprehensive auxiliary prediction result of the target chronic disease based on the first auxiliary prediction result, the second auxiliary prediction result, and the third auxiliary prediction result; The display unit is used to display the comprehensive auxiliary prediction results of the target chronic disease.

9. An electronic device, characterized in that: The electronic device comprises a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus; Memory, used to store computer programs; A processor, for implementing any of the methods described in claims 1-7 when executing a program stored in a memory.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images