Intelligent management and control method and system for health data

Abstract

The invention discloses a health data intelligent management and control method and system, and belongs to the field of data management. The method comprises the following steps: inputting a to-be-verified detection attribute of a target user into a predefined association attribute mapping table, and outputting an association attribute set; according to the association attribute set, downloading multi-attribute historical time series data of the target user passing the blockchain evidence storage, executing frequent pattern mining, and constructing confidence data; and when the to-be-verified detection data is inconsistent with the confidence data, generating a quality abnormality identifier and sending the quality abnormality identifier to a management terminal. According to the method, the problems of weak applicability and insufficient accuracy of static verification of health detection data in the prior art are solved, and the technical defect that subsequent health management and health management decisions are affected due to the fact that the health condition of the user is difficult to reflect reliably due to incomplete and inaccurate data verification is overcome.

Description

Technical Field

[0001] This invention relates to the field of data management, and more particularly to a method and system for intelligent management and control of health data. Background Technology

[0002] In today's society, with the continuous advancement of medical technology and people's increasing emphasis on health management, the accuracy and reliability of health monitoring data have become crucial. However, traditional methods for verifying the quality of health monitoring data mainly rely on static verification based on fixed thresholds. While this method is simple and easy to implement, it has significant limitations. Specifically, fixed thresholds cannot adapt to data fluctuations under different users and different testing scenarios, resulting in weak applicability of the verification results and difficulty in comprehensively and accurately reflecting the user's health status. Especially in the field of maternal and child health services, due to the significant differences between the physiological characteristics of pregnant women and children and adults, their health monitoring data is more complex and variable. Traditional verification methods often cannot effectively cope with these special situations, leading to low data quality, which in turn affects subsequent health management and decision-making.

[0003] With the rapid development of big data technology, effectively managing and utilizing massive amounts of health monitoring data, and improving the completeness, accuracy, timeliness, and consistency of the data, has become an urgent problem to be solved. At the same time, management researchers also need a powerful data analysis and intelligent decision-making support system to formulate more scientific and reasonable health management strategies based on deep data mining and precise analysis. Summary of the Invention

[0004] This invention addresses the technical problem in existing technologies where static verification of health monitoring data results in weak applicability and insufficient accuracy, making it difficult to comprehensively and accurately reflect the user's health status and thus affecting subsequent health management and hygiene management decisions. It provides a method and system for intelligent management and control of health data to solve this problem.

[0005] The technical solution of the present invention to solve the above technical problems is as follows: In a first aspect, the present invention provides a method for intelligent management and control of health data, comprising: inputting the detection attributes of the target user's unverified detection data into a predefined associated attribute mapping table, and outputting an associated attribute set; downloading the target user's multi-attribute historical time-series data stored on the blockchain according to the associated attribute set, performing frequency mining, and constructing detection attribute confidence data; when the unverified detection data is inconsistent with the detection attribute confidence data, generating a quality anomaly identifier for the unverified detection data, and sending it to a health data management terminal.

[0006] Secondly, the present invention provides a health data intelligent management and control system, the system comprising: an attribute association module, used to input the detection attributes of the target user's unverified detection data into a predefined association attribute mapping table and output an association attribute set; a data mining module, used to download the target user's multi-attribute historical time-series data stored on the blockchain according to the association attribute set, perform frequency mining, and construct detection attribute confidence data; and an anomaly identification module, used to generate a quality anomaly identification for the unverified detection data when the unverified detection data is inconsistent with the detection attribute confidence data, and send it to the health data management terminal.

[0007] The beneficial effects of this invention are as follows: by inputting the target user's unverified detection attributes into a predefined associated attribute mapping table to output an associated attribute set, and then downloading historical time-series data stored in the blockchain for frequent mining to construct confidence data, and generating a quality anomaly identifier when the unverified data and the confidence data are inconsistent and sending it to the management terminal, the applicability and accuracy of health detection data verification are improved, the effect of data quality management is enhanced, and more reliable data support is provided for health management. Attached Figure Description

[0008] Figure 1 This is a flowchart illustrating the intelligent health data management method provided by the present invention.

[0009] Figure 2 This is a schematic diagram of the structure of the intelligent health data management and control system provided by the present invention.

[0010] Figure labeling: Attribute association module 11, data mining module 12, anomaly identification module 13. Detailed Implementation

[0011] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0012] In the description of this invention, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0013] In the description of this invention, the term "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this invention is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use the invention. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that the invention can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid obscuring the description of the invention with unnecessary detail. Therefore, the invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed herein. Example

[0014] like Figure 1 As shown, this embodiment of the invention provides a method for intelligent management and control of health data, including: S10: Input the detection attributes of the target user's detection data to be verified into a predefined associated attribute mapping table, and output the associated attribute set.

[0015] For example, in the intelligent health data management method, the detection attributes of the target user's test data to be verified are first used as input, and a predefined association attribute mapping table is used to accurately output the corresponding association attribute set. The target user refers to the specific user targeted by the intelligent health data management method, that is, the data owner whose health test data is to be verified. Test data to be verified refers to test-related data belonging to the user that still needs verification. This data may come from various testing activities, such as medical testing and quality testing. Before verification, its accuracy and reliability are uncertain, requiring further confirmation processes to ensure the data is authentic, valid, and usable. For example, if a user has undergone genetic testing, the preliminary data obtained by the testing institution belongs to the user's test data to be verified before passing a series of verification steps.

[0016] In this context, the detection attribute refers to the specific attribute in the test data that characterizes a particular health indicator or characteristic, such as blood pressure or blood sugar levels. The associated attribute mapping table is a pre-built data structure that stores the mapping relationships between the detection attribute and other attributes that may influence it (i.e., associated attributes). By inputting the detection attribute, the system can quickly locate and output a set of related associated attributes. These associated attributes may include the user's age, gender, lifestyle habits, and past medical history, which together constitute the multidimensional factors influencing the detection attribute, forming the associated attribute set. For example, when analyzing the blood pressure test data of a hypertensive patient, the system might include age, weight, and family medical history as associated attributes in subsequent data analysis.

[0017] By accurately identifying and correlating the multidimensional factors that influence the detection attributes, a more comprehensive and in-depth data foundation is provided for subsequent data analysis and decision-making, which helps to improve the accuracy and effectiveness of health data management.

[0018] S20: Based on the aforementioned associated attribute set, download the multi-attribute historical time-series data of the target user stored on the blockchain, perform frequency mining, and construct confidence data for the detection attributes.

[0019] Furthermore, based on the output set of related attributes, the download process is automatically triggered to obtain the target user's multi-attribute historical time-series data, which is stored on the blockchain. Here, the set of related attributes serves as a filtering criterion, ensuring the relevance and specificity of the downloaded data. Blockchain storage guarantees the authenticity and immutability of the data, providing a reliable data source for subsequent analysis. The multi-attribute historical time-series data covers multiple health indicator records of the user at different points in time, such as blood pressure, heart rate, and blood sugar levels. This data is organized in a time series format, making it easy to track changes in the user's health status.

[0020] Subsequently, the system performs frequency mining on these historical time-series data to identify frequently occurring and representative patterns or regularities in the dataset. These patterns or regularities form the basis of the confidence data for the detection attributes. For example, when analyzing the historical blood glucose data of a diabetic patient, the system may find that the patient's blood glucose level fluctuates regularly within specific time periods each day; this finding is then incorporated into the confidence data for the detection attributes.

[0021] By constructing confidence data for detection attributes, users can be provided with more personalized and accurate health assessments and suggestions. At the same time, it also provides a scientific basis for subsequent health intervention measures, significantly improving the intelligence level and practicality of health data management.

[0022] S30: When the test data to be verified is inconsistent with the test attribute confidence data, a quality anomaly identifier is generated for the test data to be verified and sent to the health data management terminal.

[0023] Specifically, in the health data management process, the system continuously monitors and compares the data to be verified with the established confidence data for the test attributes. When a significant deviation is detected between the data to be verified, such as a pregnant woman's blood pressure value at a specific stage of pregnancy, and the range of confidence data for the test attributes constructed based on her historical health data and related attributes (such as age, gestational age, and past medical history), the data to be verified is determined to be inconsistent with the confidence data. At this time, the system will automatically generate a quality anomaly label for the data to be verified. This label clearly indicates the specific type and degree of the data anomaly, such as "blood pressure value exceeds the expected range." Subsequently, the quality anomaly label will be immediately sent to the health data management terminal to ensure that medical staff or health management personnel can quickly be aware of this anomaly.

[0024] Taking maternal and child health as an example, if the system detects that a pregnant woman's blood sugar level suddenly rises and exceeds the normal range predicted based on her historical data and current gestational age, the system will immediately generate a quality abnormality indicator and notify relevant medical staff so that timely intervention measures can be taken to ensure the health of the mother and baby.

[0025] The above steps not only improve the real-time nature and accuracy of health data management, but also provide strong decision support for medical staff, helping to identify and address potential health risks in a timely manner.

[0026] In a preferred embodiment, the detection attributes of the target user's detection data to be verified are input into a predefined association attribute mapping table, and the association attribute set is output, including: Obtain preset detection attributes, as well as a set of fuzzy interference type attributes and a set of fuzzy interference quantification attributes set through the health data management terminal. Among them, the fuzzy interference attributes represent the attributes that do not have a clear degree of correlation with the preset detection attributes.

[0027] Based on the preset detection attributes and the fuzzy interference attribute set, collect one-to-one corresponding preset detection attribute record data and fuzzy interference attribute record dataset, perform correlation analysis, and obtain the correlation attribute set.

[0028] Based on the preset detection attributes and the associated attribute set, the associated attribute mapping table is constructed.

[0029] Specifically, the process begins by acquiring preset detection attributes, such as the pregnant woman's weight gain during pregnancy. This attribute is the core of subsequent analysis. Simultaneously, a fuzzy interference type attribute set and a fuzzy interference quantification attribute set are set through the health data management terminal. The fuzzy interference attributes are used to characterize attributes that may affect the preset detection attributes, but whose degree of correlation is not yet clearly defined. In the context of maternal and child health, the fuzzy interference quantification attribute set can specifically include parameters such as the user's gender (although in this example, primarily pregnant women, gender classification remains a basic attribute), age (e.g., the pregnant woman's age may affect the normal range of weight gain during pregnancy), and body temperature (changes in body temperature during pregnancy may indirectly reflect physical condition, thus affecting weight management).

[0030] Next, the system collects corresponding pre-defined detection attribute records (such as weekly weight gain records for pregnant women) and fuzzy interference attribute records (such as age, body temperature, etc. for corresponding weeks) based on the pre-defined detection attributes and fuzzy interference attribute sets, and performs correlation analysis. The correlation analysis process aims to reveal the potential relationship between fuzzy interference attributes and pre-defined detection attributes, thereby filtering out the truly significant correlated attributes that have an impact on the pre-defined detection attributes, forming a correlated attribute set.

[0031] Finally, based on the preset detection attributes and the selected set of related attributes, a correlation attribute mapping table is constructed. This mapping table provides crucial information for subsequent intelligent management of health data, enabling the system to more accurately understand the interactions between various attributes, thereby improving the accuracy of data analysis and decision support. For example, when analyzing pregnant women's weight gain data, the correlation attribute mapping table can identify the specific impact of fuzzy interference attributes such as age and body temperature on weight gain, providing medical staff with more scientific advice on pregnancy weight management, effectively preventing pregnancy complications, and ensuring maternal and infant health.

[0032] In a preferred embodiment, based on the preset detection attribute and the fuzzy interference attribute set, one-to-one corresponding preset detection attribute record data and fuzzy interference attribute record dataset are collected, and correlation analysis is performed to obtain the correlation attribute set, including: Obtain the pre-configured fuzzy interference attribute deviation threshold set of the health data management terminal.

[0033] When the preset detection attribute is a type attribute, the first fuzzy interference attribute is extracted from the fuzzy interference attribute set and set as a variable attribute. The remaining fuzzy interference attributes are set as constant attributes. Variable attributes refer to deviations greater than deviation thresholds, while constant attributes refer to deviations less than or equal to deviation thresholds.

[0034] Based on the fuzzy interferometric attribute deviation threshold set, and constrained by the variable attributes and the constant attributes, the preset detection attribute record data and the fuzzy interferometric attribute record dataset are integrated to obtain the first multi-cluster variable attribute record data and the first multi-cluster preset detection attribute record data. The proportion of multiple record data with the preset detection attributes changing as the variable attributes change is statistically analyzed.

[0035] When the proportion of multiple recorded data is greater than or equal to the ratio threshold, the first fuzzy interference attribute is added to the associated attribute set.

[0036] Optionally, in order to accurately identify the associated attributes that are closely related to the preset detection attributes, a set of fuzzy interference attribute deviation thresholds pre-configured by the health data management terminal is first obtained. This set of thresholds is used to define the degree of deviation of the fuzzy interference attributes relative to the preset detection attributes.

[0037] When a preset detection attribute is determined to be a type attribute, such as the pregnancy stage (early, middle, or late), the system selects one attribute from the fuzzy interference attribute set as a variable attribute, such as the pregnant woman's age. The remaining fuzzy interference attributes are treated as constant attributes, such as body temperature (assuming that body temperature fluctuations are small during this stage and do not affect the pregnancy stage classification). Variable attributes refer to those whose deviation values ​​exceed a preset deviation threshold, while constant attributes are those whose deviation values ​​are within the threshold.

[0038] Subsequently, based on the fuzzy interferometric attribute deviation threshold set, variable attributes and constant attributes are used as constraints. Preset detection attribute record data (such as records of pregnant women at different stages of pregnancy) and fuzzy interferometric attribute record dataset (such as data on the age and body temperature of the corresponding pregnant women) are integrated to divide the data into multiple clusters that meet the constraints of variable attributes and constant attributes, namely the first multiple cluster variable attribute record data and the first multiple cluster preset detection attribute record data.

[0039] Further statistical analysis was conducted on the percentage of recorded data where the preset detection attribute (pregnancy stage) changed with the variable attribute (age). If this percentage exceeded or was equal to the preset ratio threshold, it indicated that the fuzzy interference attribute of age had a significant impact on the pregnancy stage. Therefore, the age attribute was added to the associated attribute set.

[0040] This process, through refined correlation analysis, can accurately identify fuzzy interference attributes that have a substantial impact on preset detection attributes, providing a more precise basis for subsequent health data analysis and decision-making. For example, in maternal and child health services, it helps medical staff understand the physiological changes of pregnant women of different ages at different stages of pregnancy, thereby developing more personalized health management plans.

[0041] In a preferred embodiment, based on the preset detection attribute and the fuzzy interference attribute set, one-to-one corresponding preset detection attribute record data and fuzzy interference attribute record dataset are collected, and correlation analysis is performed to obtain the correlation attribute set, including: When the preset detection attribute is a quantization attribute, the fuzzy interference attribute set is segmented to obtain a fuzzy interference type attribute set and a fuzzy interference quantization attribute set.

[0042] Without changing the set of fuzzy interference type attributes, based on the preset detection attribute record data and the fuzzy interference attribute record dataset, perform grey relational analysis on the preset detection attributes and the fuzzy interference quantization attribute set, extract fuzzy interference quantization attributes with a correlation degree greater than or equal to the correlation degree threshold, and add them to the correlation attribute set.

[0043] The first fuzzy interference type attribute of the fuzzy interference type attribute set is set as a variable attribute, and the remaining fuzzy interference type attributes and the fuzzy interference quantization attribute set are set as constant attributes. The preset detection attribute record data and the fuzzy interference attribute record dataset are integrated to obtain the second multi-cluster variable attribute record data and the second multi-cluster preset detection attribute record data. The average value of the fluctuation of multiple preset detection attributes as the variable attribute type changes is statistically analyzed.

[0044] The weighted average is calculated by taking the proportion of records within multiple clusters as the weight and calculating the average of the fluctuation values ​​of the multiple preset detection attributes.

[0045] When the weighted mean is greater than or equal to the weighted mean threshold, the first fuzzy interference type attribute is added to the associated attribute set.

[0046] Furthermore, when the preset detection attribute is a quantitative attribute, such as the pregnant woman's weight gain during pregnancy, the fuzzy interference attribute set is segmented to clarify the fuzzy interference type attribute set (such as the pregnant woman's pregnancy history type, including primiparous women, multiparous women, etc.) and the fuzzy interference quantitative attribute set (such as the pregnant woman's pre-pregnancy BMI index).

[0047] Next, keeping the fuzzy interference type attribute set unchanged, grey relational analysis is performed on the preset detection attributes and the fuzzy interference quantification attribute set based on the collected preset detection attribute record data (such as the weight gain records of pregnant women at various stages) and the fuzzy interference attribute record dataset (such as the pre-pregnancy BMI index records of the corresponding pregnant women). The grey relational analysis process aims to quantify the degree of correlation between the preset detection attributes and the fuzzy interference quantification attributes, and extract the fuzzy interference quantification attributes with a correlation degree greater than or equal to the preset correlation degree threshold. These attributes are then added to the correlation attribute set, indicating that they have a significant impact on the preset detection attributes.

[0048] Subsequently, the system uses the first fuzzy interference type attribute (e.g., primiparous women in the pregnancy history category) from the fuzzy interference type attribute set as the variable attribute, and treats the remaining fuzzy interference type attributes and the fuzzy interference quantization attribute set as constant attributes. By integrating the preset detection attribute record data and the fuzzy interference attribute record dataset, the system obtains the second multi-cluster variable attribute record data and the second multi-cluster preset detection attribute record data. Furthermore, the system statistically analyzes the average fluctuation of multiple preset detection attributes (pregnancy weight gain) as the variable attribute type (primiparous vs. multiparous) changes, in order to quantify the impact of different pregnancy history types on maternal weight gain.

[0049] To more comprehensively assess this impact, a weighted average is calculated by weighting the average fluctuations of multiple preset detection attributes using the proportion of records within multiple clusters as weights. When the weighted average is greater than or equal to a preset weighted average threshold, it indicates that the variable attribute (pregnancy history type) has a sufficiently significant impact on the preset detection attribute (pregnancy weight gain). Therefore, this first fuzzy interference type attribute (primiparous women in pregnancy history type) is added to the associated attribute set.

[0050] Through refined correlation analysis and weighted mean calculation, fuzzy interference attributes that have a substantial impact on preset detection attributes can be accurately identified. Whether it is a quantitative attribute or a typological attribute, it helps medical staff to deeply understand the differences in weight gain during pregnancy among pregnant women with different pregnancy histories, thereby providing pregnant women with more personalized weight management advice and effectively preventing the health risks caused by excessive weight gain during pregnancy.

[0051] In a preferred embodiment, based on the associated attribute set, multi-attribute historical time-series data of the target user stored on the blockchain is downloaded, frequency mining is performed, and confidence data for detection attributes is constructed, including: Using the aforementioned multi-attribute historical time-series data as constraints, a consistent modality user sample set is retrieved.

[0052] Obtain the pre-configured detection attribute deviation threshold of the health data management terminal.

[0053] Based on the detection attribute deviation threshold, cluster analysis is performed on the set of detection attribute record values ​​of the consistent modality user sample set to obtain multi-cluster detection attribute record values.

[0054] Clusters with fewer than or equal to a preset number of frequent detection attribute records are deleted to obtain multiple clusters of retained detection attribute records, which are then added to the detection attribute confidence data.

[0055] In detail, based on the constructed set of related attributes, the target user's multi-attribute historical time-series data, which is stored on the blockchain, is downloaded. This data covers multiple health indicator records of the user at different points in time, such as blood pressure, blood sugar, and heart rate. The authenticity and immutability of the data are guaranteed by blockchain technology.

[0056] Subsequently, the system uses multi-attribute historical time-series data as constraints to retrieve a user sample set with the same modality (i.e., similar health characteristics, lifestyle habits, medical history, etc.) as the target user from the vast user database of the health data management terminal. This step ensures that the data analyzed later has comparability and reference value.

[0057] Furthermore, the pre-configured detection attribute deviation threshold of the health data management terminal is obtained, which is used to define the normal fluctuation range of the detection attribute record values. Based on this threshold, cluster analysis is performed on the set of detection attribute record values ​​of the consistent modality user sample set. The record values ​​are divided into multiple clusters using a specific clustering algorithm (such as K-means clustering), and the record values ​​within each cluster have high similarity.

[0058] After completing the cluster analysis, the system further filters the detection attribute record values ​​within each cluster. Specifically, clusters with fewer than or equal to a preset number of frequent detection attribute record values ​​are deleted, as these clusters may not provide valuable reference information due to data sparsity or outlier interference. After filtering, multiple clusters of retained detection attribute record values ​​are obtained, representing the common value range or pattern of the detection attribute under the target user's modality.

[0059] Finally, the retained test attribute records are added to the test attribute confidence data, serving as an important basis for subsequent health assessments and decisions. For example, if the target user is a pregnant woman, the test attribute confidence data constructed through the above process may include the normal range of weight gain during pregnancy, the reasonable range of blood pressure fluctuations, etc. This data helps medical staff to more accurately assess the health status of pregnant women and promptly identify and address potential health risks.

[0060] In a preferred embodiment, clusters with a number of detection attribute record values ​​less than or equal to a preset frequency number are deleted to obtain multiple clusters of retained detection attribute record values, which are then added to the detection attribute confidence data, including: From the multi-cluster retention detection attribute record values, extract the maximum record value and the minimum record value of the first cluster retention detection attribute from the first cluster retention detection attribute record values.

[0061] When the maximum record value of the first cluster retention detection attribute is the same as the minimum record value of the first cluster retention detection attribute, any record value of the first cluster retention detection attribute is set as the confidence discrete point of the detection attribute.

[0062] When the maximum record value of the first cluster retained detection attribute and the minimum record value of the first cluster retained detection attribute are different, a continuous confidence interval for the detection attribute is constructed.

[0063] The discrete points of the detection attribute confidence and the continuous intervals of the detection attribute confidence are added to the detection attribute confidence data.

[0064] Preferably, after filtering the detection attribute record values ​​within a cluster, the retained multi-cluster detection attribute record values ​​need further processing to construct more refined detection attribute confidence data. Specifically, the first cluster of retained detection attribute record values ​​is extracted from the retained clusters, and the maximum and minimum record values ​​within this cluster are determined. These two extreme values ​​reflect the fluctuation range of the detection attribute record values ​​within the cluster. If the maximum and minimum record values ​​of the first cluster are the same, it indicates that all record values ​​within the cluster are consistent. In this case, any record value is randomly selected and set as the detection attribute confidence discrete point. This discrete point represents the stable value of the specific detection attribute in the consistent modality user sample set, providing a clear reference for health assessment.

[0065] Conversely, if the maximum and minimum recorded values ​​of the detected attribute in the first cluster are not the same, it indicates fluctuations in the recorded values ​​of the detected attribute within the cluster. In this case, a continuous confidence interval for the detected attribute is constructed, defined by the minimum and maximum recorded values, covering all possible values ​​of the detected attribute within the cluster. This continuous interval can more comprehensively reflect the distribution of the detected attribute in the consistent modality user sample set, providing a more flexible and comprehensive reference range for health assessment.

[0066] Finally, the processed confidence discrete points and confidence continuous intervals of the detection attributes are uniformly added to the detection attribute confidence data. For example, if processing is performed on a pregnant woman's health indicator (such as hemoglobin level), the resulting confidence discrete points might represent an ideal hemoglobin level value, while the confidence continuous intervals might provide a normal range of hemoglobin level fluctuations. Such detection attribute confidence data not only helps healthcare professionals more accurately assess the health status of pregnant women but also provides strong support for developing personalized health management plans, significantly improving the scientific rigor and practicality of health data management.

[0067] In a preferred embodiment, retrieving a consistent modality user sample set based on the multi-attribute historical time-series data includes: Using the aforementioned multi-attribute historical time-series data as constraints, a sample set of users with the same modality is retrieved.

[0068] When the amount of data in the same modal user sample set meets the convergence data requirement, the same modal user sample set is set as the consistent modal user sample set.

[0069] When the amount of data in the same modal user sample set does not meet the convergence data amount, based on the multi-attribute historical time-series data, fault tolerance conditions are configured through the health data management terminal, and similar modal user sample sets that meet the fault tolerance conditions are retrieved.

[0070] When the sum of the data amounts of the same modality user sample set and the similar modality user sample set meets the convergence data amount, the same modality user sample set and the similar modality user sample set are added to the consistent modality user sample set.

[0071] Otherwise, perform a loop search.

[0072] For example, using the target user's multi-attribute historical time-series data as a key constraint, the system retrieves a consistent modality user sample set. Specifically, the system first retrieves a user sample set that has the exact same modality as the target user (i.e., completely identical health characteristics, lifestyle habits, medical history, etc.) based on the multi-attribute historical time-series data; this is the consistent modality user sample set. Then, it determines whether the data volume of the consistent modality user sample set meets the preset convergence data volume requirement. The convergence data volume is a threshold used to ensure the reliability and stability of the analysis results. When the data volume of the consistent modality user sample set reaches or exceeds this threshold, the system directly sets it as the consistent modality user sample set for subsequent analysis.

[0073] However, if the amount of data in the same modal user sample set does not meet the convergence data requirement, the system will use multi-attribute historical time-series data as a benchmark and further search for similar modal user sample sets that meet the fault tolerance conditions configured in the health data management terminal. The fault tolerance conditions allow the system to accept a certain degree of modal difference during the search process, thereby expanding the scope of the sample set and improving the comprehensiveness and accuracy of the analysis.

[0074] After obtaining the sample set of users with similar modalities, the data volume of the sample set of users with the same modalities and the sample set of users with similar modalities are summed, and it is determined whether the summed data volume meets the convergence data volume requirement. If it does, the two sample sets are merged and added to the consistent modality sample set; if it does not, the system continues to search in a loop until a consistent modality sample set that meets the conditions is found.

[0075] For example, if the target user is a pregnant woman with a specific pregnancy complication, the system first retrieves samples of other pregnant women with the exact same medical history and health characteristics. If the sample size is insufficient, it retrieves samples of other pregnant women with similar pregnancy complications or similar health risks using a fault-tolerant approach. Finally, these two sets of samples are merged to form a consistent modality user sample set, providing strong support for subsequent health assessments and decision-making.

[0076] The above steps significantly improve the scientific rigor and practicality of health data management, helping medical professionals to more accurately assess the health status of target users and develop personalized health management plans.

[0077] In a preferred embodiment, when the detection data to be verified is inconsistent with the detection attribute confidence data, including: From the confidence data of the detection attributes, extract several discrete confidence points and several continuous confidence intervals of the detection attributes.

[0078] For each of the several confidence discrete points of the detection attributes, copy the value by a factor of 2 to construct several discrete point binary tree nodes.

[0079] Based on the confidence intervals of the aforementioned detection attributes, construct several confidence interval binary tree nodes.

[0080] By combining the discrete point binary tree nodes and the confidence interval binary tree nodes, a multi-level binary tree is constructed.

[0081] The detection data to be verified is input into the multi-level binary tree. If no new binary tree node is generated during traversal, it is considered consistent; otherwise, it is considered inconsistent.

[0082] Specifically, in the intelligent verification process of health data, when the data to be verified is inconsistent with the pre-built confidence data of the detection attributes, the system will perform a series of refined comparison operations. Specifically, it extracts several discrete confidence points and several continuous confidence intervals of the detection attributes from the confidence data. These discrete points and continuous intervals represent the normal range or stable values ​​of health indicators and are important bases for health assessment.

[0083] Next, each confidence discrete point of the detection attribute is copied twice to construct several discrete point binary tree nodes. The purpose of this step is to provide base nodes for subsequent binary tree traversal. Each node contains the value of a discrete point and its copy, so that they can be compared during the traversal.

[0084] Simultaneously, based on several continuous confidence intervals of detection attributes, several confidence interval binary tree nodes are constructed. Each confidence interval binary tree node represents the normal value range of a health indicator, and these interval information can be efficiently stored and retrieved through the binary tree structure.

[0085] Subsequently, the system combines these discrete point binary tree nodes and confidence interval binary tree nodes to construct a multi-level binary tree. This multi-level binary tree is a complex data structure that integrates all the key information of the detection attribute confidence data, providing comprehensive support for subsequent comparison of the detection data to be verified.

[0086] Finally, the detection data to be verified is input into a multi-level binary tree for traversal. During the traversal, the system checks whether the detection data to be verified matches a node (discrete point or confidence interval) in the binary tree. If no new binary tree node is generated during the traversal (i.e., the detection data to be verified completely matches an existing node), it is considered that the detection data to be verified is consistent with the detection attribute confidence data; otherwise, if a new binary tree node is generated during the traversal (i.e., the detection data to be verified does not match any existing node), it is considered inconsistent.

[0087] For example, if the test data to be verified is a biochemical indicator value of a pregnant woman, the system will compare it with pre-built confidence data for the test attributes. If the indicator value falls within a certain confidence interval or matches a certain discrete point, the system determines that they are consistent; otherwise, they are determined to be inconsistent, and may trigger further health assessments or interventions. The technical effect of this step is that it significantly improves the accuracy and efficiency of health data verification, providing strong support for personalized health management.

[0088] In a preferred embodiment, a multi-level binary tree is constructed by combining the plurality of discrete point binary tree nodes and the plurality of confidence interval binary tree nodes, including: Construct binary tree traversal rules.

[0089] When a binary tree node cannot fall into the confidence interval, or its value is different from that of the discrete point binary tree node, a new binary tree node is generated.

[0090] When a node falls into the confidence interval binary tree node, or has the same value as a discrete point binary tree node, proceed to the next level of traversal.

[0091] If no new binary tree node is generated after traversing all levels, output 1, where 1 represents consistency.

[0092] Based on the binary tree traversal rules, and combining the discrete point binary tree nodes and the confidence interval binary tree nodes, the multi-level binary tree is constructed.

[0093] Specifically, in the process of constructing intelligent verification of health data, binary tree traversal rules are first established to ensure that the data to be verified can be efficiently and accurately compared with the confidence data of the detected attributes. Specifically, the system combines several pre-constructed discrete-point binary tree nodes and several confidence interval binary tree nodes to build a multi-level binary tree structure.

[0094] When traversing multi-level binary trees, the system follows these rules: If the data to be verified cannot fall within the interval represented by any confidence interval binary tree node, or if its value is different from the value of any discrete point binary tree node, then the data is determined to be inconsistent with the confidence data of the detection attribute, and a new binary tree node is generated to record this inconsistency. Conversely, if the data to be verified can fall within the interval represented by a confidence interval binary tree node, or if its value is the same as the value of a discrete point binary tree node, then the data is determined to be consistent with the confidence data of the detection attribute at the current level, and the system proceeds to the next level for a more in-depth traversal and comparison.

[0095] The traversal process continues until all levels of the multi-level binary tree have been traversed. If no new binary tree nodes are generated during the entire traversal process, the system determines that the data to be verified is completely consistent with the confidence data of the detected attributes, and outputs an identifier (such as 1) to indicate this result.

[0096] For example, if the data to be verified is a key physiological indicator value of a pregnant woman, the system will input it into a multi-level binary tree for traversal and comparison. If the indicator value matches a certain confidence interval or discrete point at a certain level, the system will continue to traverse the next level to further confirm its consistency; if no inconsistencies are found throughout the entire traversal process, the system determines that the indicator value is normal and no further action is required.

[0097] The above steps significantly improve the automation and accuracy of health data verification, reduce the need for manual intervention, and enhance the efficiency and reliability of health assessments. By constructing a multi-level binary tree and adhering to strict traversal rules, the system can efficiently process large amounts of health data, providing strong support for personalized health management.

[0098] The intelligent health data management method provided in this invention has at least the following technical effects: 1. By inputting the detection attributes of the target user's unverified detection data into a predefined associated attribute mapping table and outputting an associated attribute set, efficient identification of fuzzy interference attributes that affect the preset detection attributes is achieved. Combined with the associated attribute set, the target user's multi-attribute historical time-series data stored on the blockchain is downloaded and frequency mining is performed to construct detection attribute confidence data, which significantly improves the efficiency and accuracy of health data mining and provides a reliable data foundation for subsequent health assessment.

[0099] 2. In the process of retrieving a consistent modality user sample set, using multi-attribute historical time-series data as constraints and combining fault-tolerant conditions to retrieve a similar modality user sample set, and dynamically adjusting the sample set to meet the convergence data volume requirements, not only improves the diversity and representativeness of the sample set, but also constructs more accurate and reliable detection attribute confidence data through steps such as cluster analysis and deletion of low-frequency clusters, providing strong support for the intelligent management and control of health data.

[0100] 3. By constructing a multi-level binary tree and combining discrete point binary tree nodes and confidence interval binary tree nodes, a rapid comparison between the detection data to be verified and the confidence data of the detection attributes is realized. This not only improves the efficiency and accuracy of the comparison, but also enables real-time monitoring and early warning of health data quality by generating quality anomaly indicators and sending them to the health data management terminal in a timely manner, providing timely and effective data support for personalized health management. Example

[0101] like Figure 2 As shown, based on the same inventive concept as the intelligent health data management method provided in Embodiment 1, this embodiment of the invention also provides an intelligent health data management system, the system comprising: The attribute association module 11 is used to input the detection attributes of the target user's detection data to be verified into a predefined association attribute mapping table and output an association attribute set.

[0102] Data mining module 12 is used to download multi-attribute historical time-series data of target users that are stored on the blockchain based on the associated attribute set, perform frequency mining, and construct confidence data for detection attributes.

[0103] The anomaly identification module 13 is used to generate a quality anomaly identification for the test data to be verified when the test data to be verified is inconsistent with the test attribute confidence data, and send it to the health data management terminal.

[0104] Furthermore, the attribute association module 11 is also used to perform the following steps: Obtain preset detection attributes, as well as a set of fuzzy interference type attributes and a set of fuzzy interference quantification attributes set through a health data management terminal, wherein the fuzzy interference attributes represent attributes that do not have a clearly defined degree of correlation and affect the preset detection attributes; based on the preset detection attributes and the fuzzy interference attribute set, collect one-to-one corresponding preset detection attribute record data and fuzzy interference attribute record dataset, perform correlation analysis, and obtain the correlation attribute set; based on the preset detection attributes and the correlation attribute set, construct the correlation attribute mapping table.

[0105] Furthermore, the attribute association module 11 is also used to perform the following steps: Obtain a pre-configured set of fuzzy interference attribute deviation thresholds for the health data management terminal; when the preset detection attribute is a type attribute, extract the first fuzzy interference attribute from the fuzzy interference attribute set and set it as a variable attribute, and set the remaining fuzzy interference attributes as constant attributes. Variable attributes refer to deviations greater than the deviation threshold, and constant attributes refer to deviations less than or equal to the deviation threshold; based on the fuzzy interference attribute deviation threshold set, and constrained by the variable attribute and the constant attribute, integrate the preset detection attribute record data and the fuzzy interference attribute record dataset to obtain the first multi-cluster variable attribute record data and the first multi-cluster preset detection attribute record data, and statistically analyze the proportion of multiple record data where the preset detection attribute changes with the variable attribute; when the proportion of multiple record data is greater than or equal to the ratio threshold, add the first fuzzy interference attribute to the associated attribute set.

[0106] Furthermore, the attribute association module 11 is also used to perform the following steps: When the preset detection attribute is a quantized attribute, the fuzzy interference attribute set is segmented to obtain a fuzzy interference type attribute set and a fuzzy interference quantized attribute set. The fuzzy interference type attribute set is kept unchanged. Based on the preset detection attribute record data and the fuzzy interference attribute record dataset, grey relational analysis is performed on the preset detection attribute and the fuzzy interference quantized attribute set. Fuzzy interference quantized attributes with a correlation degree greater than or equal to a correlation degree threshold are extracted and added to the relational attribute set. The first fuzzy interference type attribute of the fuzzy interference type attribute set is set as a variable attribute, and the remaining fuzzy interference type attributes and the fuzzy interference quantized attribute set are set as constant attributes. The preset detection attribute record data and the fuzzy interference attribute record dataset are integrated to obtain second multi-cluster variable attribute record data and second multi-cluster preset detection attribute record data. The average fluctuation of multiple preset detection attributes as the variable attribute type changes is statistically analyzed. Using the proportion of records within multiple clusters as weights, a weighted average is calculated on the average fluctuation of the multiple preset detection attributes to obtain a weighted average. When the weighted average is greater than or equal to a weighted average threshold, the first fuzzy interference type attribute is added to the relational attribute set.

[0107] Furthermore, the data mining module 12 is also used to perform the following steps: Using the multi-attribute historical time-series data as constraints, a consistent modality user sample set is retrieved; a detection attribute deviation threshold pre-configured by the health data management terminal is obtained; based on the detection attribute deviation threshold, cluster analysis is performed on the set of detection attribute record values ​​of the consistent modality user sample set to obtain multi-cluster detection attribute record values; clusters with the number of detection attribute record values ​​within a cluster less than or equal to a preset frequent number are deleted to obtain multi-cluster retained detection attribute record values, which are then added to the detection attribute confidence data.

[0108] Furthermore, the data mining module 12 is also used to perform the following steps: From the multi-cluster retention detection attribute record values, extract the maximum and minimum record values ​​of the first cluster retention detection attribute from the first cluster retention detection attribute record values; when the maximum and minimum record values ​​of the first cluster retention detection attribute are the same, set any record value of the first cluster retention detection attribute record values ​​as a detection attribute confidence discrete point; when the maximum and minimum record values ​​of the first cluster retention detection attribute are different, construct a detection attribute confidence continuous interval; add the detection attribute confidence discrete point and the detection attribute confidence continuous interval to the detection attribute confidence data.

[0109] Furthermore, the data mining module 12 is also used to perform the following steps: Using the multi-attribute historical time-series data as constraints, a user sample set with the same modality is retrieved. When the data volume of the user sample set with the same modality meets the convergence data volume, the user sample set with the same modality is set as the consistent modality user sample set. When the data volume of the user sample set with the same modality does not meet the convergence data volume, based on the multi-attribute historical time-series data, fault tolerance conditions are configured through the health data management terminal, and a similar modality user sample set that meets the fault tolerance conditions is retrieved. When the sum of the data volumes of the user sample set with the same modality and the similar modality user sample set meets the convergence data volume, the user sample set with the same modality and the similar modality user sample set are added to the consistent modality user sample set; otherwise, the retrieval is repeated.

[0110] Furthermore, the anomaly identification module 13 is also used to perform the following steps: From the confidence data of the detection attributes, extract several discrete confidence points and several continuous confidence intervals of the detection attributes; copy the values ​​of each of the discrete confidence points by a factor of 2 to construct several discrete point binary tree nodes; based on the several continuous confidence intervals of the detection attributes, construct several confidence interval binary tree nodes; combine the several discrete point binary tree nodes and the several confidence interval binary tree nodes to construct a multi-level binary tree; input the detection data to be verified into the multi-level binary tree, if no new binary tree nodes are generated during traversal, it is considered consistent; otherwise, it is considered inconsistent.

[0111] Furthermore, the anomaly identification module 13 is also used to perform the following steps: Constructing binary tree traversal rules: When a binary tree node cannot fall into a confidence interval or has a different value from a discrete binary tree node, generate a new binary tree node; when a binary tree node falls into a confidence interval or has the same value as a discrete binary tree node, proceed to the next level of traversal; when no new binary tree node is generated after traversing all levels, output 1, where 1 represents consistency; based on the binary tree traversal rules, combine the discrete binary tree nodes and the confidence interval binary tree nodes to construct the multi-level binary tree.

[0112] Through the foregoing detailed description of the intelligent health data management method, those skilled in the art can clearly understand the intelligent health data management system in this embodiment. As the system disclosed in the embodiment corresponds to the method disclosed in the embodiment, the description is relatively simple, and relevant parts can be referred to the method section.

[0113] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for intelligent management and control of health data, characterized in that, include: Input the detection attributes of the target user's unverified detection data into a predefined associated attribute mapping table, and output the associated attribute set; Based on the aforementioned set of associated attributes, download the multi-attribute historical time-series data of the target user that has been stored on the blockchain, perform frequency mining, and construct confidence data for the detection attributes; When the test data to be verified is inconsistent with the confidence data of the test attributes, a quality anomaly identifier is generated for the test data to be verified and sent to the health data management terminal.

2. The intelligent management and control method for health data as described in claim 1, characterized in that, Input the detection attributes of the target user's unverified detection data into a predefined associated attribute mapping table, and output an associated attribute set, including: Obtain preset detection attributes, as well as a set of fuzzy interference type attributes and a set of fuzzy interference quantification attributes set through the health data management terminal. Among them, the fuzzy interference attributes represent the attributes that do not have a clear degree of correlation and affect the preset detection attributes. Based on the preset detection attributes and the fuzzy interference attribute set, collect one-to-one corresponding preset detection attribute record data and fuzzy interference attribute record dataset, perform correlation analysis, and obtain the correlation attribute set; Based on the preset detection attributes and the associated attribute set, the associated attribute mapping table is constructed.

3. The intelligent management and control method for health data as described in claim 2, characterized in that, Based on the preset detection attributes and the fuzzy interference attribute set, one-to-one corresponding preset detection attribute record data and fuzzy interference attribute record dataset are collected, and correlation analysis is performed to obtain the correlation attribute set, including: Obtain the pre-configured fuzzy interference attribute deviation threshold set of the health data management terminal; When the preset detection attribute is a type attribute, the first fuzzy interference attribute is extracted from the fuzzy interference attribute set and set as a variable attribute. The remaining fuzzy interference attributes are set as constant attributes. A variable attribute refers to an attribute deviation greater than a deviation threshold, and a constant attribute refers to an attribute deviation less than or equal to a deviation threshold. Based on the fuzzy interference attribute deviation threshold set, and constrained by the variable attribute and the constant attribute, the preset detection attribute record data and the fuzzy interference attribute record dataset are integrated to obtain the first multi-cluster variable attribute record data and the first multi-cluster preset detection attribute record data. The proportion of multiple record data with the preset detection attribute changing with the variable attribute is statistically analyzed. When the proportion of multiple recorded data is greater than or equal to the ratio threshold, the first fuzzy interference attribute is added to the associated attribute set.

4. The intelligent management and control method for health data as described in claim 2, characterized in that, Based on the preset detection attributes and the fuzzy interference attribute set, one-to-one corresponding preset detection attribute record data and fuzzy interference attribute record dataset are collected, and correlation analysis is performed to obtain the correlation attribute set, including: When the preset detection attribute is a quantization attribute, the fuzzy interference attribute set is segmented to obtain a fuzzy interference type attribute set and a fuzzy interference quantization attribute set. Without changing the set of fuzzy interference type attributes, based on the preset detection attribute record data and the fuzzy interference attribute record dataset, perform gray relational analysis on the preset detection attributes and the fuzzy interference quantization attribute set, extract fuzzy interference quantization attributes with a correlation degree greater than or equal to the correlation degree threshold, and add them to the correlation attribute set; The first fuzzy interference type attribute of the fuzzy interference type attribute set is set as a variable attribute, and the other fuzzy interference type attributes and the fuzzy interference quantization attribute set are set as constant attributes. The preset detection attribute record data and the fuzzy interference attribute record dataset are integrated to obtain the second multi-cluster variable attribute record data and the second multi-cluster preset detection attribute record data. The average value of the fluctuation of multiple preset detection attributes as the variable attribute type changes is statistically analyzed. Using the proportion of records within multiple clusters as weights, a weighted average is calculated on the average fluctuation values ​​of the multiple preset detection attributes to obtain a weighted average. When the weighted mean is greater than or equal to the weighted mean threshold, the first fuzzy interference type attribute is added to the associated attribute set.

5. The intelligent management and control method for health data as described in claim 1, characterized in that, Based on the aforementioned set of associated attributes, download the multi-attribute historical time-series data of the target user stored on the blockchain, perform frequency mining, and construct confidence data for detection attributes, including: Using the aforementioned multi-attribute historical time-series data as constraints, retrieve a consistent modality user sample set; Obtain the pre-configured detection attribute deviation threshold of the health data management terminal; Based on the detection attribute deviation threshold, cluster analysis is performed on the set of detection attribute record values ​​of the consistent modality user sample set to obtain multi-cluster detection attribute record values; Clusters with fewer than or equal to a preset number of frequent detection attribute records are deleted to obtain multiple clusters of retained detection attribute records, which are then added to the detection attribute confidence data.

6. The intelligent management and control method for health data as described in claim 5, characterized in that, Clusters with fewer than or equal to a preset frequency of detected attribute record values ​​within a cluster are deleted, resulting in multiple clusters retaining the detected attribute record values. These are then added to the detected attribute confidence data, including: From the multi-cluster retention detection attribute record values, extract the maximum record value and the minimum record value of the first cluster retention detection attribute from the first cluster retention detection attribute record values; When the maximum record value of the first cluster retention detection attribute is the same as the minimum record value of the first cluster retention detection attribute, any record value of the first cluster retention detection attribute is set as the confidence discrete point of the detection attribute. When the maximum record value of the first cluster retained detection attribute and the minimum record value of the first cluster retained detection attribute are different, a continuous confidence interval for the detection attribute is constructed. The discrete points of the detection attribute confidence and the continuous intervals of the detection attribute confidence are added to the detection attribute confidence data.

7. The intelligent management and control method for health data as described in claim 5, characterized in that, Using the aforementioned multi-attribute historical time-series data as constraints, a consistent modality user sample set is retrieved, including: Using the aforementioned multi-attribute historical time-series data as constraints, retrieve a sample set of users with the same modality; When the amount of data in the same modal user sample set meets the convergence data requirement, the same modal user sample set is set as the consistent modal user sample set. When the amount of data in the same modal user sample set does not meet the convergence data amount, based on the multi-attribute historical time series data, fault tolerance conditions are configured through the health data management terminal, and similar modal user sample sets that meet the fault tolerance conditions are retrieved. When the sum of the data amounts of the same modal user sample set and the similar modal user sample set meets the convergence data amount, the same modal user sample set and the similar modal user sample set are added to the consistent modal user sample set; Otherwise, perform a loop search.

8. The intelligent management and control method for health data as described in claim 6, characterized in that, When the detection data to be verified is inconsistent with the confidence data of the detection attribute, including: From the confidence data of the detection attributes, extract several discrete confidence points and several continuous confidence intervals of the detection attributes; For each of the aforementioned discrete confidence points of the detection attributes, duplicate the value by a factor of 2 to construct a number of discrete point binary tree nodes; Based on the aforementioned continuous confidence intervals of several detection attributes, construct several confidence interval binary tree nodes; By combining the discrete point binary tree nodes and the confidence interval binary tree nodes, a multi-level binary tree is constructed; The detection data to be verified is input into the multi-level binary tree. If no new binary tree node is generated during traversal, it is considered consistent; otherwise, it is considered inconsistent.

9. The intelligent management and control method for health data as described in claim 8, characterized in that, By combining the discrete point binary tree nodes and the confidence interval binary tree nodes, a multi-level binary tree is constructed, including: Constructing binary tree traversal rules: When a binary tree node cannot fall into the confidence interval, or its value is different from that of the discrete point binary tree node, a new binary tree node is generated. When a node falls into the confidence interval binary tree node, or has the same value as the discrete point binary tree node, proceed to the next level of traversal; If no new binary tree node is generated after traversing all levels, output 1, where 1 represents consistency. Based on the binary tree traversal rules, and combining the discrete point binary tree nodes and the confidence interval binary tree nodes, the multi-level binary tree is constructed.

10. A health data intelligent management and control system, characterized in that, The system is used to implement the intelligent health data management and control method according to any one of claims 1-9, the system comprising: The attribute association module is used to input the detection attributes of the target user's detection data to be verified into a predefined association attribute mapping table and output an association attribute set. The data mining module is used to download multi-attribute historical time-series data of target users that are stored on the blockchain based on the associated attribute set, perform frequency mining, and construct confidence data for detection attributes. The anomaly identification module is used to generate a quality anomaly identification for the test data to be verified when the test data to be verified is inconsistent with the test attribute confidence data, and send it to the health data management terminal.

Images