Customer health information management system and method

Abstract

The invention discloses a customer health information management system and method, and relates to the technical field of health information management, and the management system comprises a data source summarization module, a customer type classification module, a health state evaluation module, an adjustment feedback correction module and a tracking prompt module. The technical key points are as follows: the system not only ensures that high-risk customers can obtain timely and effective monitoring and management, but also avoids resource waste caused by excessive inspection by dynamically adjusting the inspection frequency; meanwhile, a machine learning model and a cloud medical clinical guide database are fully utilized, the accuracy of correction coefficients and the selectivity of characteristic indexes are improved, and the setting of the examination frequency is more scientific and reasonable; in addition, the introduction of a comprehensive correction mechanism further improves the health monitoring effect of high-age customers and customers with disease history at the same time, and provides powerful support for preventing diseases and improving the health level of the customers.

Description

Technical Field

[0001] The present invention relates to the technical field of health information management, and specifically to a customer health information management system and method. Background Art

[0002] Health information management refers to the process of collecting, organizing, analyzing, and utilizing relevant information such as the health status, health risk factors, and medical service utilization of individuals or groups; it involves multiple fields such as clinical, epidemiology, and demography. Through effective management, it is possible to better understand the health status and needs of patients, provide a basis for formulating personalized care plans and providing high-quality care services, and is of great significance for improving the quality of medical services, promoting individual health, and controlling medical costs; a health information management system supports the realization of this process, records and manages personal health information, and provides technical support for health management.

[0003] In the existing document with the publication number CN103123663A and the name "User Health Management Method and System", it is pointed out that its technical solution includes the following steps: the client provides a user registration port for the user to fill in personal information and medical information, and generates a virtual person bound to the user's medical information based on the personal information and medical information; the health assistant platform regularly obtains the user's physiological index information from the hospital information management system, generates health index information based on the physiological index information, combines the health index information and personal information to generate a health index and a health reminder queue and saves them; the client regularly reads the health index and reminder information, drives the virtual person to display the user's health status according to the health index, and pushes health reminders on time; it can improve the convenience of users' health management and enhance the user experience; Combined with the above document and the prior art, in the process of managing customer health information, conventional solutions can all complete the health degree assessment of target customers based on the physiological index information of the customers. In the case of no obvious abnormality or discomfort, the order or frequency of the reminder queue can be changed according to the size of the health degree assessment value. For example: the higher the health degree of customer A, the lower the reminder frequency and the more backward the queue, which is convenient for giving priority reminders to customers with lower health degrees. However, the specific reminder frequency based solely on the health degree is not accurate. For some customers who are slightly older or have a medical history, the reminder frequency is relatively more frequent, but the inspection frequency of the traditional solution is fixed or lacks flexibility and personalization, so it cannot provide more powerful support for preventing diseases or improving the health level of customers. Summary of the Invention

[0004] (I) Technical Problems to be Solved In view of the deficiencies of the prior art, the present invention provides a customer health information management system and method. By dynamically adjusting the examination frequency, the system not only ensures that high-risk customers can be monitored and managed in a timely and effective manner, but also avoids waste of resources caused by excessive examinations. At the same time, the solution also makes full use of machine learning models and cloud medical clinical guideline databases to improve the accuracy of correction coefficients and the selectivity of characteristic indicators, making the setting of examination frequency more scientific and reasonable. In addition, the introduction of a comprehensive correction mechanism further enhances the health monitoring effect for customers who meet both the high age and the presence of a disease history, providing strong support for disease prevention and improving the health level of customers, and solving the problems raised in the background technology.

[0005] (2) Technical solution To achieve the above objectives, the present invention is realized through the following technical solutions: A customer health information management system, comprising: A data source aggregation module, which collects relevant data sets of target customers and performs data processing operations; A customer type classification module, which constructs a classification standard setting engine; Inputs the relevant data sets of target customers and outputs the classification results of customer types; A health status evaluation module, which builds a health evaluation model and calculates the status valuation of target customers based on relevant data sets, and generates a preliminary examination frequency based on the status valuation; An adjustment feedback correction module, which obtains and analyzes the classification results of customer types; When the target customer belongs to the customer type of high age group, the age correction mechanism is triggered; When the target customer belongs to the customer type with a corresponding disease history, a representative analysis strategy is first executed, and then the determination adjustment mechanism is triggered; When the target customer meets both the high age group and the customer type with a corresponding disease history, the comprehensive correction mechanism is triggered; When the target customer belongs to other normal customer types, the established pointing mechanism is triggered; Dynamically adjusts the preliminary examination frequency according to the operation results of the corresponding mechanism to obtain the current examination frequency; A tracking reminder module, according to the current examination frequency of the target customer, after completing the current examination action, sends the time of the next examination day to the mobile terminal of the target customer through wireless communication technology. If no information on the target customer's examination action is detected on the given next examination day, a reminder warning strategy is executed.

[0006] Furthermore, the relevant data sets at least include physiological data, basic information, and psychological data; Among them, the physiological data includes blood pressure and heart rate; The basic information includes name, age, weight, height, and medical history data; The psychological data is the mental health score; The content of performing data processing operations on the relevant data set is: cleaning each data in the relevant data set, removing outliers, and performing standardization processing on each data.

[0007] Furthermore, the process of constructing the classification criterion setting engine is as follows: Condition 1: Set an age threshold, and classify the target customers whose age exceeds the age threshold into the customer type of the high-age group; Condition 2: Determine a disease list from the disease type database. When any one of the diseases in the disease list is included in the medical history data of the target customer, classify it into the customer type with the corresponding disease history; among them, when the number of disease types in the medical history data of the target customer exceeds 1, according to the most recent examination result, select the disease type with the maximum degree of exceeding the standard as the representative of the disease type, and classify it into the customer type with the corresponding disease history, and simultaneously determine the value of the number of disease types; Condition 3: Classify the target customers who do not meet Conditions 1 and 2 into other normal customer types; The classification result of outputting the customer type includes: The target customer belongs to any one or several of the customer type of the high-age group, the customer type with the corresponding disease history, and other normal customer types.

[0008] Furthermore, the index types based on the relevant data set include: physiological data, weight and height in the basic information, and psychological data; When building a health assessment model, the formula is as follows: ; (1) ; (2) In the formula, H represents the state estimate. There are n indicators related to the health status, which are respectively denoted as x1, x2,..., xn. Each indicator has its corresponding standard value, denoted as s1, s2,.., sn. The deviation between the corresponding indicator and its standard value is calculated through the function f(xi, si), and wi represents the weight coefficient of the i-th indicator, and the value range is all [0, 1]; The method of generating the preliminary examination frequency based on the state estimate is: Establish a model that associates the state estimate with the examination frequency, and obtain a formula based on a linear relationship: ; In the formula, F0 represents the preliminary examination frequency, F_base represents the basic examination frequency, and k1 represents the adjustment coefficient, and k1 > 0.

[0009] Furthermore, the process of triggering the age correction mechanism is as follows: Introduce an age adjustment factor to correct the preliminary examination frequency in the following manner: ; (3) ; (4) In the formula, F_mf represents the corrected value of the examination frequency, and α(Age) represents a function of age Age.

[0010] Furthermore, the representative analysis strategy executed first is as follows: Candidate indication: Query the index data related to the corresponding disease history based on the cloud medical clinical guideline database to obtain a list of candidate indicators closely related to the corresponding disease history; Statistical analysis: Use the existing target customer data and adopt correlation analysis technology to conduct statistical analysis on the candidate indicators in the candidate indicator list to evaluate the correlation between the candidate indicators and the corresponding disease state or progression; Feature algorithm: Based on the results of the statistical analysis, apply the feature importance ranking algorithm based on the tree model to output the corresponding indicator data with the highest feature importance and use it as the feature indicator corresponding to the corresponding disease history; The process of the subsequent triggered judgment adjustment mechanism is as follows: Adopt a machine learning model to calculate the correction coefficient, extract the number value of the disease types in the classification result, construct a corrected examination model, input the correction coefficient and the number value of the disease types corresponding to the target customer, and output the corrected value of the examination frequency; Among them, the steps of using the machine learning model to calculate the correction coefficient are as follows: Data preparation: Based on the results of the representative analysis strategy, collect the historical data of the characteristic indicators corresponding to the target customer and set a target variable to represent the future health status of the target customer; Feature engineering: Extract the statistical parameter of the feature indicator from the historical data as a feature, and use a rolling window to calculate the moving average and moving standard deviation; among them, the statistical parameters include: mean, standard deviation, maximum value, minimum value, change rate; Model training: Use the trained random forest regression model to predict the target variable; Correction coefficient calculation: Input the extracted features into the trained random forest regression model to obtain the predicted probability value of the target variable; set a correction coefficient calculation engine, and the engine settings are as follows: When the target variable probability value < 0.3, set the correction coefficient to 1.0; When 0.3 ≤ target variable probability value < 0.6, set the correction coefficient to 1.5; When the probability value of the target variable ≥ 0.6, the correction coefficient is set to 2.0.

[0011] Furthermore, when running the correction check model, the formula used is as follows: ; In the formula, k2 represents the correction coefficient, w1 and w2 represent the weight coefficients, and their value ranges are both [0, 1]. Nr represents the number value of disease types.

[0012] Furthermore, the process of triggering the comprehensive correction mechanism is as follows: For customer types that simultaneously meet the high - age customer type and the corresponding disease history, a correction formula that combines the two is constructed, and the formula is as follows: ; In the formula, γ represents a regulation coefficient greater than 1; The process of triggering the established pointing mechanism is as follows: Maintain the preliminary inspection frequency as the current inspection frequency.

[0013] Furthermore, the content of the implemented prompt and warning strategy is as follows: The time of the next inspection day is sent to the mobile device of the target customer a second time through wireless communication technology. If the information indicating that the target customer has performed an inspection action has not been detected on the second day after the next inspection day, the prompt information is synchronously sent to the target customer and their bound terminal; among them, the bound terminal is the relative number reserved by the target customer in the medical institution.

[0014] A customer health information management method includes the following steps: S1. Collect the relevant data set of the target customer and perform data processing operations; S2. Build a classification standard setting engine; Input the relevant data set of the target customer and output the classification result of the customer type; S3. Build a health assessment model and calculate the status estimate of the target customer based on the relevant data set, and generate a preliminary inspection frequency based on the status estimate; S4. Obtain and analyze the classification result of the customer type; When the target customer belongs to the high - age customer type, trigger the age correction mechanism; When the target customer belongs to the customer type with a corresponding disease history, first execute the representative analysis strategy, and then trigger the determination adjustment mechanism; When the target customer simultaneously meets the high - age and the customer type with a corresponding disease history, trigger the comprehensive correction mechanism; When the target customer belongs to other normal customer types, trigger the established pointing mechanism; Dynamically adjust the preliminary inspection frequency according to the operation result of the corresponding mechanism to obtain the current inspection frequency; S5. According to the current inspection frequency of the target customer, after completing the current inspection action, send the time of the next inspection day to the mobile terminal of the target customer through wireless communication technology. If no information on the target customer's inspection action is detected on the given next inspection day, then execute the prompt warning strategy.

[0015] (III) Beneficial effects The present invention provides a customer health information management system and method, having the following beneficial effects: (1) This solution constructs a classification standard setting engine. According to the matching situation of age thresholds, medical history data and disease lists, and the principle of maximizing the degree of exceeding the standard, the target customers are accurately classified, and types such as high-age customers, customers with corresponding disease histories, and other normal customers are output. This not only solves the problems of unclear customer type division and inconsistent classification standards, but also improves the accuracy and pertinence of customer classification; (2) This solution realizes the quantitative evaluation of the health status of target customers, and dynamically generates a preliminary inspection frequency based on the evaluation results. By building a health assessment model and comprehensively considering multi-dimensional indicators such as physiology and psychology to calculate the status valuation, it not only comprehensively reflects the health status of customers, but also can realize the automatic adjustment of the inspection frequency according to the size of the customer's health risk by establishing a linear relationship model between the status valuation and the inspection frequency, initially ensuring the pertinence and timeliness of health inspections; (3) This solution realizes the dynamic and personalized adjustment of the inspection frequency of target customers, comprehensively considering multiple factors such as customer age, disease history, and disease complexity. Through the age correction mechanism, representative analysis strategy and decision adjustment mechanism, comprehensive correction mechanism, and established pointing mechanism, the inspection frequency correction value of each customer is accurately calculated, effectively solving the problems of fixed inspection frequency, lack of flexibility and personalization in traditional health management, and improving the pertinence and efficiency of health management; (4) By dynamically adjusting the inspection frequency, this solution not only ensures that high-risk customers can be monitored and managed in a timely and effective manner, but also avoids resource waste caused by over-inspection; at the same time, this solution also makes full use of machine learning models and cloud medical clinical guideline databases to improve the accuracy of correction coefficients and the selectivity of characteristic indicators, making the setting of inspection frequencies more scientific and reasonable; in addition, the introduction of the comprehensive correction mechanism further improves the health monitoring effect for customers who simultaneously meet the conditions of high age and having a disease history, provides strong support for preventing diseases and improving the health level of customers, and realizes the personalization and precision of health management. Description of the drawings

[0016] Figure 1Schematic diagram of the component modules within the system of the present invention; Figure 2 Schematic diagram of the system architecture of the present invention; Figure 3 Schematic diagram of the method flow of the present invention. Detailed implementation manners

[0017] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0018] Embodiment 1: Please refer to Figure 1 and Figure 2 , this embodiment provides a customer health information management system; The customer health information management system aims to provide a comprehensive, efficient, and secure health management platform for individual users and medical institutions by integrating modern information technology and medical health knowledge; the system will cover multiple aspects such as data collection, health assessment, and health status prediction, and use advanced technologies such as analysis or intelligent algorithms to realize the customization and implementation of personalized health management solutions, which can handle different types of customers and provide targeted inspection frequency suggestions, and can guarantee the health and safety of customers to a certain extent; This management system includes the design of each functional module, namely, a data source aggregation module, a customer type classification module, a health status assessment module, an adjustment feedback correction module, and a tracking reminder module; the core of the entire system lies in: being able to provide targeted inspection suggestions according to the customer type, further improving the accuracy and guaranteeing the management effect; The module descriptions are as follows: The data source aggregation module collects relevant data sets of target customers and performs data processing actions; Among them, the target customers include at least one customer, usually several customers, and the customer types are different. Each customer has its corresponding relevant data set; the ways to collect relevant data sets include at least manual input, docking with medical institutions, and detection by wearable devices; usually in this embodiment, the only way to collect relevant data sets is docking with medical institutions, using API interfaces to dock with third-party medical institution systems to obtain professional medical records and obtain the required relevant data sets from them; The relevant data sets include at least physiological data, basic information, and psychological data; Specifically, the physiological data includes blood pressure and heart rate, the basic information includes name, age, weight, height, and medical history data, and the psychological data is the mental health score (directly using the scale score). The content of the data processing actions on the relevant data sets is as follows: Clean each data in the relevant data sets to remove outliers or missing values; and perform standardization or normalization processing on each data to ensure that indicators with different dimensions can be compared and integrated.

[0019] The customer type classification module constructs a classification standard setting engine, inputs the relevant data sets of the target customers, and outputs the classification results of the customer types; Among them, the process of constructing the classification standard setting engine is as follows: Condition 1: Set an age threshold, and classify the target customers whose age exceeds the age threshold as the customer type of the high-age group; Condition 2: Determine the disease list from the disease type database. When any one of the diseases in the disease list is included in the medical history data of the target customer, classify it as the customer type with the corresponding disease history; among them, when the number of disease types in the medical history data of the target customer exceeds 1, according to the result of the most recent examination, select the disease type with the maximum degree of exceeding the standard as the representative of the disease type, and classify it as the customer type with the corresponding disease history, and simultaneously determine the value of the number of disease types; Condition 3: Classify the target customers who do not meet Conditions 1 and 2 as other normal customer types; It should be noted that the age threshold is usually set at 60 years old; the disease type database records the names of many types of diseases currently suffered by humans, and the obtained disease list is dozens or hundreds of types; when Condition 2 is met, when the disease type of the target customer is 1, directly use it as the customer type with the corresponding disease history; when the disease type of the target customer exceeds 1, use the representative of the disease type as the customer type with the corresponding disease history; the degree of exceeding the standard indicates the amount by which the most representative indicator parameter under the corresponding type of disease exceeds its corresponding normal range value, that is, the degree of exceeding the standard. Then, each type of disease has a degree of exceeding the standard, and the maximum value is extracted, which is the maximum degree of exceeding the standard, and the corresponding disease type is also used as the representative of the disease type; The classification results of the output customer types include: the target customer belongs to any one or several of the customer type of the high-age group, the customer type with the corresponding disease history, and other normal customer types; among them, some target customers may simultaneously meet the customer type of the high-age group and the customer type with the corresponding disease history.

[0020] By adopting the above technical solutions, the comprehensive collection and efficient processing of the data sets related to target customers are realized, including the cleaning, standardization or normalization of physiological data, basic information and psychological data, ensuring the quality and comparability of the data; Meanwhile, a classification standard setting engine is constructed. According to the age threshold, the matching situation between the medical history data and the disease list, and the principle of maximizing the degree of exceeding the standard, the target customers are accurately classified, and types such as customers in the high-age group, customers with corresponding disease histories, and other normal customers are output; This solution not only solves the problems of unclear customer type division and inconsistent classification standards, but also improves the accuracy and pertinence of customer classification, making the demand analysis and service strategy formulation for different customer types more scientific and reasonable; especially for target customers who meet multiple conditions, it can comprehensively consider factors such as their age and disease history, providing strong support for providing personalized and differentiated medical service and management solutions, and effectively improving the accuracy and satisfaction of customer service.

[0021] The health status evaluation module builds a health evaluation model and calculates the status valuation of the target customer based on the relevant data sets, and generates a preliminary examination frequency based on the status valuation for the target customer to perform timely health examination operations accordingly; Among them, based on the physiological data, weight and height in the basic information, and psychological data in the relevant data sets; When building the health evaluation model, the formula used is as follows: ; (1) ; (2) In the formula, H represents the status valuation. There are n indicators related to the health status, which are respectively denoted as x1, x2,..., xn. Here, n = 4, x1 represents blood pressure, x2 represents heart rate, x3 represents body mass index, and x4 represents mental health score. Each indicator has a corresponding standard value or normal range, denoted as s1, s2,.., sn. The deviation between the corresponding indicator and its standard value is calculated through the function f(xi, si); wi represents the weight coefficient of the i-th indicator, which reflects the importance of this indicator in evaluating the customer's health status, and the value range is [0, 1]; the function f(xi, si) is a function for calculating the deviation between the actual measured value or score of the i-th indicator and the standard value; Logical overview: Select indicators that are closely related to the customer's health status. These indicators should be able to comprehensively reflect the customer's physical, mental, and social health conditions; for each selected indicator, we need to set a standard value or normal range. This standard value can be a statistical quantity such as the population mean, median, percentile, etc., or an ideal value or health range determined based on medical expertise; the setting of the weight coefficient wi is crucial, and it should be based on professional knowledge and experience to reflect the relative importance of each indicator in evaluating the customer's health status; for example, for some key indicators (such as blood pressure, blood sugar), we can assign them larger weights; while for some secondary indicators (such as height, body mass index), we can assign them smaller weights; finally, by the method of weighted summation, the deviations of each indicator are combined to obtain the customer's status estimate H; The way to generate the preliminary examination frequency based on the status estimate is as follows: Establish a model that correlates the status estimate with the examination frequency to obtain a formula based on a linear relationship: ; In the formula, F0 represents the preliminary examination frequency, F_base represents the examination frequency when the status estimate is a certain reference value (such as 0 or the ideal value of the healthy state), that is, the basic examination frequency, and k1 represents the adjustment coefficient, and k1 > 0; Brief explanation: k1 is a positive number that determines how the examination frequency should change when the status estimate H increases or decreases by one unit. The value of k1 can be adjusted according to the actual situation to reflect the sensitivity between the status estimate and the examination frequency; in this model, there is a linear relationship between the status estimate H and the examination frequency F, which means that when the status estimate increases, the examination frequency will also increase accordingly to monitor the customer's health condition more frequently; Example: Suppose the customer's status estimate H is calculated by the previously mentioned health assessment model, ranging from 0 to 10, and the higher the value, the greater the health risk; we set the basic examination frequency to once a year (F_base = 1 time / year), and the adjustment coefficient k1 = 0.2 times / year / unit of status estimate; If the customer's status estimate H = 2, then the examination frequency calculated according to the formula is: 1.4 times / year; This means that the customer needs to undergo an examination approximately every 9 months; If the customer's status estimate H = 5, then the examination frequency calculated according to the formula is: 2 times / year; This means that the customer needs to undergo an examination every 6 months; If the customer's status estimate H = 8, then the examination frequency calculated according to the formula is: 2.6 times / year; This means that customers need to be examined approximately once every 4.5 months; The above frequencies are set when no obvious abnormalities are found during the examination. If abnormalities or problems are found during the current examination, treatment or surgery will be carried out immediately, which is not part of the specific description or protection of this application.

[0022] By adopting the above technical solution, a quantitative assessment of the health status of target customers is achieved, and the initial examination frequency is dynamically generated based on the assessment results. This solution builds a health assessment model, comprehensively considering multi-dimensional indicators such as physiology and psychology to calculate the status valuation, which not only comprehensively reflects the health status of customers but also realizes the automatic adjustment of the examination frequency according to the size of the customer's health risk by establishing a linear relationship model between the status valuation and the examination frequency, ensuring the pertinence and timeliness of health examinations. This technical solution effectively solves the problems of fixed examination frequency and lack of personalization in traditional health management, improves the accuracy and efficiency of health management. By dynamically adjusting the examination frequency, it not only avoids the waste of resources caused by over-examination but also ensures that high-risk customers can be monitored and managed in a timely and effective manner, providing strong support for disease prevention and improving the health level of customers.

[0023] The adjustment feedback correction module obtains and analyzes the classification results of customer types; When the target customer belongs to the customer type of high age group, the age correction mechanism is triggered; When the target customer belongs to the customer type with a corresponding disease history, the representative analysis strategy is first executed, and then the determination adjustment mechanism is triggered; When the target customer simultaneously meets the customer types of high age group and having a corresponding disease history, the comprehensive correction mechanism is triggered; When the target customer belongs to other normal customer types, the established pointing mechanism is triggered; Dynamically adjust the initial examination frequency according to the operation results of the corresponding mechanism, and remind the target customer to perform health examination operations according to the current examination frequency (under the condition that the established pointing mechanism is not triggered, the examination frequency correction value is used as the current examination frequency; under the condition that the established pointing mechanism is triggered, the initial examination frequency is not adjusted, and the initial examination frequency is used as the current examination frequency). Among them, the process of triggering the age correction mechanism is as follows: Introduce an age adjustment factor to correct the initial examination frequency in the following way: ; (3) ; (4) In the formula, \(F_{mf}\) represents the inspection frequency correction value, \(\alpha(Age)\) represents a function of age \(Age\), which adjusts the inspection frequency according to the age. \(\alpha\) represents the age adjustment factor, and \(\alpha(Age)\) is the linear age adjustment factor. For customers aged 60 and above, the inspection frequency increases by 5% for each additional year. When \(Age \lt 60\), \(\alpha(Age)=1\), and of course, this age correction mechanism will not be triggered. Brief description: \(\alpha(Age)\) is the core, which reflects the impact of age on the inspection frequency. Generally speaking, as age increases, the value of \(\alpha(Age)\) should gradually increase to reflect the need for more frequent health monitoring of elderly customers. The specific form of \(\alpha(Age)\) can be adjusted personalized according to the actual situation. For example, it can be set that before a certain age (such as 60 years old), \(\alpha(Age)\) remains 1, that is, the inspection frequency is not adjusted. After this age, \(\alpha(Age)\) increases linearly or non-linearly with age.

[0024] The content of the representative analysis strategy executed first is as follows: Candidate indication: According to the cloud medical clinical guideline database, query the index data related to the corresponding disease history to obtain a list of index candidates closely related to the corresponding disease history. Among them, the cloud medical clinical guideline database stores medical literature, including the main pathophysiological mechanisms, common complications, and known related biomarkers or physiological parameters of various diseases. Statistical analysis: Using the existing target customer data, perform statistical analysis on the candidate indicators in the list of candidate indicators by using correlation analysis techniques to evaluate the correlation between the candidate indicators and the corresponding disease status or progression. Among them, the existing target customer data includes relevant data sets and other various types of detected index data, which will not be elaborated here. The correlation analysis techniques include any one of the Pearson correlation coefficient and the Spearman rank correlation coefficient, and it is also possible to choose to combine the correlation analysis techniques with regression analysis techniques (such as linear regression, logistic regression). Feature algorithm: According to the statistical analysis results, apply the feature importance ranking algorithm based on the tree model to output the corresponding index data with the highest feature importance and use it as the feature index corresponding to the corresponding disease history. Among them, the feature importance ranking algorithm based on the tree model is the random forest. Using this feature selection algorithm to further screen the most representative indicators is the required feature index. An overall example of the representative analysis strategy: Suppose we want to select a most representative index parameter to monitor the disease progression for patients with type 2 diabetes. Candidate Indication: According to the cloud-based medical clinical guideline database, query the index data related to the corresponding disease history. Type 2 diabetes is related to blood glucose level, glycated hemoglobin (HbA1c), and insulin resistance index (HOMA-IR). Statistical Analysis: Collect data from a group of type 2 diabetes patients, perform correlation analysis and regression analysis on these indicators, and find that HbA1c has a significant correlation with the disease progression of patients (such as the incidence of complications). Feature Algorithm: Apply the random forest algorithm to evaluate the importance of each indicator, and find that HbA1c has the highest feature importance in the model. Therefore, it is used as the feature indicator corresponding to the corresponding disease history (type 2 diabetes).

[0025] The process of triggering the judgment adjustment mechanism is as follows: Use a machine learning model to calculate the correction coefficient, extract the number value of the disease type in the classification result, construct a correction inspection model, input the correction coefficient and the number value of the disease type corresponding to the target customer, and output the correction value of the inspection frequency. Among them, the steps of using a machine learning model to calculate the correction coefficient are as follows: Data Preparation: According to the results of the representative analysis strategy, collect the historical data of the feature indicators corresponding to the target customer, and set a target variable to represent the future health status of the target customer (such as disease onset, deterioration of the condition, etc.). Feature Engineering: Extract the statistical parameter of the feature indicator from the historical data as a feature, and use a rolling window to calculate the moving average and moving standard deviation; among them, the statistical parameters include: mean, standard deviation, maximum value, minimum value, and change rate. Model Training: Use a random forest regression model to predict the target variable; use historical data to train the random forest regression model, evaluate the performance of the random forest regression model through cross-validation technology, and adjust the parameters of the random forest regression model to optimize the prediction accuracy. Correction Coefficient Calculation: Input the extracted features into the trained random forest regression model to obtain the predicted probability value of the target variable; set a correction coefficient calculation engine, and the engine settings are as follows: When the target variable probability value < 0.3, set the correction coefficient to 1.0; When 0.3 ≤ target variable probability value < 0.6, set the correction coefficient to 1.5; When the target variable probability value ≥ 0.6, set the correction coefficient to 2.0; An example of the overall steps is as follows: Suppose we have a dataset on the blood glucose levels of diabetes patients, and our goal is to predict the risk of deterioration of the condition within the next three months based on the change in blood glucose levels. Data Preparation: Collect the blood glucose level data of diabetic patients in the past year, as well as the target variable indicating whether the condition will deteriorate in the next three months; Feature Engineering: Extract the mean, standard deviation, maximum value, minimum value, and rate of change of blood glucose levels as features; Calculate the moving average and moving standard deviation of blood glucose levels in the past three months; Model Training: Use a random forest regression model to predict the risk of condition deterioration in the next three months (expressed as a probability between 0 and 1); Evaluate the model performance through cross-validation and adjust the model parameters; Correction Coefficient Calculation: For the data of new diabetic patients, input the extracted features into the model to obtain the predicted probability of condition deterioration; Set up a correction coefficient calculation engine, and the engine is set as: If the predicted probability < 0.3, the correction coefficient = 1.0 (do not change the examination frequency); If 0.3 <= predicted probability < 0.6, the correction coefficient = 1.5 (increase the examination frequency by 50%); If the predicted probability >= 0.6, the correction coefficient = 2.0 (double the examination frequency).

[0026] When running the correction check model, the formula is as follows: ; In the formula, k2 represents the correction coefficient, w1 and w2 represent the weight coefficients, and their value ranges are both [0, 1], and Nr represents the number value of disease types; The weight coefficients are determined by the coefficient of variation method, which is a method of weighting each index according to the degree of variation between the current value and the target value of each evaluation index; If the numerical difference of a certain index is large and can clearly distinguish each evaluated object, it means that the resolution information of this index is rich, so a larger weight should be given to this index; On the contrary, if the numerical differences of each evaluated object on a certain index are small, then the ability of this index to distinguish each evaluation object is weak, so a smaller weight should be given to this index; This method directly uses the information contained in each index and calculates the weight of the index, so it has objectivity; Logical Explanation: k2 reflects the adjustment of the original examination frequency based on a certain standard (such as age, health status, etc.); Nr reflects the complexity of the patient's disease, and the more disease types, the corresponding increase in the examination frequency; w1*Nr: directly increase the examination frequency according to the number of disease types; w2*k2*Nr*F0 is a product term to ensure that when k2, F0, and Nr act simultaneously, there is an additional increase in the examination frequency; Advantage Explanation: Compared with traditional weighted calculations, its advantages mainly lie in comprehensiveness and dynamics; Traditional weighted calculations usually simply perform linear weighting on various factors, while this formula realizes the interaction and dynamic adjustment between factors by introducing a product term; this means that when the correction coefficient, the original examination frequency, and the number of existing disease types of the patient change simultaneously, the corrected examination frequency will have an additional increase or adjustment, more accurately reflecting the actual examination needs under the combined action of these factors; in addition, this formula has high flexibility and can adapt to different clinical scenarios and patient needs by adjusting the weight coefficients, making the setting of the examination frequency more personalized and scientific; therefore, the design of this formula has significant advantages in comprehensively considering multiple factors and their interactions; Example verification: Assume: The correction coefficient k2 = 1.2; The original examination frequency F0 = 2 times / year; The number value N of the existing disease types of the target customer = 3; The weight coefficients w1 = 0.5, w2 = 0.1; Calculate the corrected examination frequency: F_mf = 4.62 times / year; Comparison: 1. Only considering the correction coefficient and F0: F_mf = 2.4 times / year;

[0027] 2. Only considering F0 and Nr: F_mf = 3.5 times / year; It can be seen that the corrected examination frequency F_mf = 4.62 times / year is greater than the result of 2.4 times / year when only considering the correction coefficient and F0, and the result of 3.5 times / year when only considering F0 and Nr, which verifies the effectiveness of the formula.

[0028] The process of triggering the comprehensive correction mechanism is as follows: For customer types that simultaneously meet the high-age customer type and the corresponding disease history, construct a correction formula that combines the two, and the formula is as follows: ; In the formula, γ represents a regulation coefficient greater than 1; Logical explanation: F0*α(Age) represents the corrected frequency considering only the age factor, and k2 + w1*Nr + w2*k2*Nr*γ represents that on the basis of the disease history correction, the influence of the number of disease types is added, and γ ensures that the corrected frequency is higher; γ>1: ensures that when considering both age and disease history simultaneously, the corrected frequency will be higher than when considering only one of them; no more redundant example proofs are given here.

[0029] The process of triggering the established pointing mechanism is as follows: Maintain the preliminary examination frequency to perform real-time health examination operations.

[0030] By adopting the above technical solutions, the dynamic and personalized adjustment of the examination frequency of target customers is achieved; this solution comprehensively considers multiple factors such as customer age, medical history, and disease complexity, and accurately calculates the examination frequency correction value for each customer through an age correction mechanism, a representative analysis strategy and determination adjustment mechanism, a comprehensive correction mechanism, and a predefined pointing mechanism; This technical solution effectively solves the problems of fixed examination frequency, lack of flexibility and personalization in traditional health management, and improves the pertinence and efficiency of health management; by dynamically adjusting the examination frequency, it not only ensures that high-risk customers can be monitored and managed in a timely and effective manner, but also avoids the waste of resources caused by over-examination; at the same time, this solution also makes full use of machine learning models and cloud medical clinical guideline databases to improve the accuracy of correction coefficients and the selectivity of characteristic indicators, making the setting of examination frequency more scientific and reasonable; in addition, the introduction of the comprehensive correction mechanism further improves the health monitoring effect for customers who meet both high age and have a medical history, provides strong support for preventing diseases and improving the health level of customers, and realizes the personalization and precision of health management.

[0031] The tracking and reminder module, based on the current examination frequency of the target customer, after completing the current examination action, sends the time of the next examination day to the mobile terminal of the target customer through wireless communication technology. If no information on the target customer's examination action is detected on the given next examination day, a reminder and warning strategy is executed; Among them, the mobile terminal of the target customer is usually a mobile phone, and it is sent in the form of text messages or voice calls; For example: if the current examination frequency is 12 times a year, a text message with the content "A second physical examination is required on the 1st of next month" will be sent to the mobile phone of the target user on the fixed date of each month, such as the 1st; The content of the executed reminder and warning strategy is as follows: The time of the next examination day is sent to the mobile terminal of the target customer again through wireless communication technology. If no information on the target customer's examination action is detected on the second day after the next examination day, the reminder information will be synchronously sent to the target customer and their bound terminal. The bound terminal is the relative number reserved by the target customer in the medical institution, which is convenient to send the reminder information to the relative number to realize the supervision of the target customer by the relative.

[0032] By adopting the above technical solutions, the effective tracking and timely reminder of the examination schedule of target customers are achieved; Based on the current examination frequency of the target customers, this solution automatically calculates and sends the time of the next examination day to the customers' mobile terminals to ensure that the customers can conduct health examinations on time. If the customers fail to have the examinations on time, a reminder and warning strategy will be executed. By sending the examination day reminder twice and synchronously sending it to the relatives' numbers, the reminder effect is enhanced. The supervision of the relatives is utilized to encourage the customers to complete the examinations on time, effectively solving the problem that customers forget or neglect the examination schedule, improving the compliance and timeliness of health examinations, and providing a strong guarantee for the health management of the customers.

[0033] Embodiment 2: Please refer to Figure 3 , based on Embodiment 1, this embodiment also provides a method for managing customers' health information, including the following steps: S1. Collect relevant data sets of the target customers and perform data processing operations; S2. Build a classification standard setting engine; Input the relevant data sets of the target customers and output the classification results of the customer types; S3. Build a health assessment model and calculate the status valuation of the target customers based on the relevant data sets, and generate a preliminary examination frequency based on the status valuation; S4. Obtain and analyze the classification results of the customer types; When the target customer belongs to the customer type of the high-age group, trigger the age correction mechanism; When the target customer belongs to the customer type with a corresponding disease history, first execute the representative analysis strategy, and then trigger the judgment adjustment mechanism; When the target customer simultaneously meets the customer types of the high-age group and having a corresponding disease history, trigger the comprehensive correction mechanism; When the target customer belongs to other normal customer types, trigger the established pointing mechanism; Dynamically adjust the preliminary examination frequency according to the operation results of the corresponding mechanisms to obtain the current examination frequency; S5. Based on the current examination frequency of the target customers, after completing the current examination operation, send the time of the next examination day to the mobile terminal of the target customers through wireless communication technology. If no information indicating that the target customer has performed the examination operation is detected on the given next examination day, execute the reminder and warning strategy.

[0034] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. Those of ordinary skill in the art will realize that the units and algorithm steps of the examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or in a combination of computer software and electronic hardware. Whether these functions are executed in hardware or software depends on the specific application and design constraints of the technical solution.

[0035] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units. They may be located in one place or distributed over multiple network units. Some or all of the units can be selected according to actual needs to achieve the objectives of the solution of this embodiment.

[0036] As described above, the above is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should all be covered within the protection scope of the present application.

Claims

1. A customer health information management system, characterized in that, It includes: A data source aggregation module that collects relevant data sets of target customers and performs data processing operations; A customer type classification module that constructs a classification standard setting engine; Inputs the relevant data sets of the target customers and outputs the classification results of the customer types; A health status assessment module that builds a health assessment model, calculates the status valuation of the target customers based on the relevant data sets, and generates a preliminary examination frequency based on the status valuation; An adjustment feedback correction module that obtains and analyzes the classification results of the customer types; When the target customer belongs to the customer type of the high-age group, an age correction mechanism is triggered; When the target customer belongs to the customer type with a corresponding disease history, a representative analysis strategy is first executed, and then a determination adjustment mechanism is triggered; When the target customer simultaneously meets the customer types of the high-age group and having a corresponding disease history, a comprehensive correction mechanism is triggered; When the target customer belongs to other normal customer types, a predefined pointing mechanism is triggered; Dynamically adjusts the preliminary examination frequency according to the operation results of the corresponding mechanisms to obtain the current examination frequency; A tracking reminder module that, based on the current examination frequency of the target customer, after completing the current examination operation, sends the time of the next examination day to the mobile terminal of the target customer through wireless communication technology. If no information about the target customer's examination operation is detected on the given next examination day, a reminder warning strategy is executed.

2. The customer health information management system according to claim 1, wherein: The relevant data sets at least include physiological data, basic information, and psychological data; Among them, the physiological data includes blood pressure and heart rate; The basic information includes name, age, weight, height, and medical history data; The psychological data is the mental health score; The content of performing data processing operations on the relevant data sets is: cleaning each data in the relevant data sets, removing outliers, and performing standardization processing on each data.

3. The customer health information management system according to claim 2, wherein: The process of constructing the classification standard setting engine is as follows: Condition 1: Set an age threshold, and classify target customers whose age exceeds the age threshold as the customer type of the high-age group; Condition 2: Determine a disease list from the disease type database. When any one of the disease list is included in the medical history data of the target customer, classify it as the customer type with a corresponding disease history; among them, when the number of disease types included in the medical history data of the target customer exceeds 1, select the disease type with the maximum degree of exceeding the standard as the representative of the disease type according to the nearest previous examination result, classify it as the customer type with a corresponding disease history, and simultaneously determine the number value of the disease types; Condition 3: Classify target customers who do not meet Conditions 1 and 2 as other normal customer types; The output classification results of the customer types include: The target customer belongs to any one or several of the customer types of the high-age group, the customer type with a corresponding disease history, and other normal customer types.

4. The customer health information management system according to claim 2, characterized in that: The indicators based on the relevant data sets include: physiological data, weight and height in the basic information, and psychological data; When building the health assessment model, the formula used is as follows: ； （1） ； （2） Wherein, H represents the state estimation. Suppose there are n indicators related to the health state, denoted as x1, x2, ..., xn respectively. Each indicator has its corresponding standard value, denoted as s1, s2, .., sn. The deviation between the corresponding indicator and its standard value is calculated by the function f(xi, si), and wi represents the weight coefficient of the i-th indicator, and the value range is [0, 1]; The method for generating the preliminary inspection frequency based on the state estimation is as follows: Establish a model that associates the state estimation with the inspection frequency, and output the preliminary inspection frequency F0.

5. The customer health information management system according to claim 4, wherein: The process of triggering the age correction mechanism is as follows: Introduce an age adjustment factor to correct the preliminary inspection frequency in the following way: ； （3） ； （4） Wherein, F_mf represents the inspection frequency correction value, and α(Age) represents a function of age Age.

6. The customer health information management system according to claim 3, characterized in that: The representative analysis strategy executed first is as follows: Candidate indication: According to the cloud medical clinical guideline database, query the indicator data related to the corresponding disease history to obtain a list of candidate indicators closely related to the corresponding disease history; Statistical analysis: Using the existing target customer data, adopt correlation analysis technology to conduct statistical analysis on the candidate indicators in the list of candidate indicators, and evaluate the correlation between the candidate indicators and the corresponding disease state or progress; Feature algorithm: According to the statistical analysis results, apply the feature importance ranking algorithm based on the tree model to output the corresponding indicator data with the highest feature importance, and use it as the feature indicator corresponding to the corresponding disease history.

7. The customer health information management system according to claim 6, characterized in that: The process of the subsequent triggered determination adjustment mechanism is as follows: Adopt a machine learning model to calculate the correction coefficient, extract the number value of the disease type in the classification result, construct a corrected inspection model, input the correction coefficient and the number value of the disease type corresponding to the target customer, and output the inspection frequency correction value; Among them, the steps of using a machine learning model to calculate the correction coefficient are as follows: Data preparation: According to the results of the representative analysis strategy, collect the historical data of the corresponding feature indicators of the target customer, and set a target variable to represent the future health status of the target customer; Feature engineering: Extract the statistical parameter of the feature indicator from the historical data as a feature, and calculate the moving average and moving standard deviation using a rolling window; among them, the statistical parameters include: mean, standard deviation, maximum value, minimum value, change rate; Model training: Use the trained random forest regression model to predict the target variable; Correction coefficient calculation: Input the extracted features into the trained random forest regression model to obtain the predicted probability value of the target variable; set a correction coefficient calculation engine to determine the set correction coefficient according to the range of the target variable probability value.

8. The customer health information management system according to claim 7, characterized in that: The process of triggering the comprehensive correction mechanism is: For customers who simultaneously meet the high-age customer type and the corresponding disease history type, construct a correction formula that combines the two to obtain the inspection frequency correction value F_mf; The process of triggering the established pointing mechanism is as follows: Maintain the preliminary inspection frequency as the current inspection frequency.

9. The customer health information management system according to claim 1, wherein: The content of the executed prompt and warning strategy is as follows: The time of the next inspection date is sent to the mobile device of the target customer for the second time through wireless communication technology. If the information indicating that the target customer has performed the inspection action has not been detected on the second day after the next inspection date, the prompt message is synchronously sent to the target customer and their bound terminal; where the bound terminal is the relative number reserved by the target customer in the medical institution.

10. A method for managing customer health information, characterized in that: It includes the following steps: S1. Collect the relevant data set of the target customer and perform data processing actions; S2. Build a classification standard setting engine; Input the relevant data set of the target customer and output the classification result of the customer type; S3. Build a health assessment model and calculate the status estimate of the target customer based on the relevant data set, and generate a preliminary inspection frequency based on the status estimate; S4. Obtain and analyze the classification result of the customer type; When the target customer belongs to the customer type of the high-age group, trigger the age correction mechanism; When the target customer belongs to the customer type with a corresponding disease history, first execute the representative analysis strategy, and then trigger the judgment adjustment mechanism; When the target customer simultaneously meets the customer types of the high-age group and having a corresponding disease history, trigger the comprehensive correction mechanism; When the target customer belongs to other normal customer types, trigger the established pointing mechanism; Dynamically adjust the preliminary inspection frequency according to the operation result of the corresponding mechanism to obtain the current inspection frequency; S5. According to the current inspection frequency of the target customer, after completing the current inspection action, send the time of the next inspection date to the mobile device of the target customer through wireless communication technology. If the information indicating that the target customer has performed the inspection action has not been detected on the given next inspection date, execute the prompt warning strategy.

Images