Health condition assessment method and device based on electronic health record, equipment, medium and product

Abstract

The invention discloses a health condition assessment method and device based on electronic health records, equipment, a medium and a product, and relates to the technical field of machine learning, the method comprises the following steps: preprocessing historical EHR data and real-time EHR data, and determining the preprocessed EHR data; performing feature selection on the preprocessed EHR data through a machine learning algorithm, determining selected EHR data, and taking the processed EHR data as a data set to be clustered; updating a plurality of cluster center positions in the to-be-clustered data set according to the constructed fitness function, and determining an optimal cluster center position corresponding to each cluster center position based on an iterative convergence mechanism; the cluster center position is the position of a clustering center; clustering processing is carried out on the data points corresponding to the optimal cluster center position, and an EHR clustering result is determined; according to the method and the device, the convergence speed is improved, and the health condition is evaluated more accurately.

Description

Technical Field

[0001] The present application relates to the field of machine learning technology, and in particular to a method, device, equipment, medium and product for assessing health status based on electronic health records. Background Art

[0002] With the rapid development of information technology and medical technology, Electronic Health Records (EHR) have gradually become a vital data source in the medical system. The EHR system can integrate a variety of medical information such as patients' medical history, examination reports, diagnosis records, treatment plans, etc., providing strong support for clinical diagnosis, treatment decisions and health management. However, due to the large amount and complex structure of EHR data, how to efficiently extract useful information from it and conduct health status assessment has become the focus of current research.

[0003] At present, most health ratings rely on supervised learning methods. Although supervised learning can provide a certain degree of accuracy for health prediction, its widespread application faces several challenges. First, supervised learning relies heavily on large-scale labeled data, and the data labeling process is time-consuming and error-prone. Second, the existing feature selection methods are relatively simple, which often leads to low accuracy, robustness and stability of feature selection, which in turn limits the model's ability to handle high-dimensional, collinear and noisy data, affecting the reliability and practicality of the rating results.

[0004] As an unsupervised learning technique, clustering methods can automatically group and classify similar samples by analyzing the internal structure of data, avoiding the problem of over-reliance on labels. However, general clustering methods have problems such as slow convergence due to improper selection of initial cluster centers when clustering data sets, and low accuracy of evaluation results due to the uniform distribution of data points. Summary of the invention

[0005] The purpose of this application is to provide a health status assessment method, device, equipment, medium and product based on electronic health records to solve the problems of slow convergence caused by improper selection of initial cluster centers during data set clustering, and low accuracy of assessment results caused by uniform data point distribution.

[0006] To achieve the above objectives, this application provides the following solutions:

[0007] In a first aspect, the present application provides a method for assessing health status based on electronic health records, comprising:

[0008] Preprocessing historical EHR data and real-time EHR data to determine preprocessed EHR data;

[0009] Performing feature selection on the preprocessed EHR data by a machine learning algorithm, determining the selected EHR data, and using the processed EHR data as a data set to be clustered;

[0010] According to the constructed fitness function, multiple cluster center positions in the data set to be clustered are updated, and based on an iterative convergence mechanism, the optimal cluster center position corresponding to each cluster center position is determined; the cluster center position is the position of the cluster center;

[0011] Performing clustering processing on the data points corresponding to the optimal cluster center position to determine the EHR clustering result;

[0012] The user's health status is evaluated according to the EHR clustering result.

[0013] In a second aspect, the present application provides a health status assessment device based on electronic health records, comprising:

[0014] A data preprocessing module, used for preprocessing historical EHR data and real-time EHR data, and determining preprocessed EHR data;

[0015] A feature selection module, for performing feature selection on the preprocessed EHR data through a machine learning algorithm, determining the selected EHR data, and using the processed EHR data as a data set to be clustered;

[0016] An optimal cluster center position determination module is used to update multiple cluster center positions in the data set to be clustered according to the constructed fitness function, and determine the optimal cluster center position corresponding to each cluster center position based on an iterative convergence mechanism; the cluster center position is the position of the cluster center;

[0017] A clustering processing module, used to perform clustering processing on the data points corresponding to the optimal cluster center position to determine the EHR clustering result;

[0018] An evaluation module is used to evaluate the health status of the user according to the EHR clustering result.

[0019] In a third aspect, the present application provides a computer device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement any of the above-mentioned methods for assessing health conditions based on electronic health records.

[0020] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the above-mentioned methods for assessing health conditions based on electronic health records.

[0021] In a fifth aspect, the present application provides a computer program product, including a computer program, which, when executed by a processor, implements any of the above-mentioned methods for assessing health conditions based on electronic health records.

[0022] According to the specific embodiments provided in this application, this application discloses the following technical effects:

[0023] The present application preprocesses historical EHR data and real-time EHR data to determine the preprocessed EHR data; performs feature selection on the preprocessed EHR data through a machine learning algorithm to determine the selected EHR data, and uses the processed EHR data as the data set to be clustered, and updates the cluster center position through a fitness function, obtains the optimal cluster center position based on an iterative convergence mechanism, improves the stability and accuracy of clustering, and increases the convergence speed; clusters the data points corresponding to the above-mentioned optimal cluster center position to determine the EHR clustering result, thereby providing a more detailed clustering result, so that each data point can be reasonably assigned to an appropriate cluster according to its relative position, thereby improving the classification accuracy of the cluster, and thus being able to more accurately assess the health status. By performing cluster analysis on the patient's EHR data, health risk groups can be better identified, thereby achieving personalized health management and preventive measures. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0025] Figure 1 This is a flowchart of a health status assessment method based on electronic health records in one embodiment of the present application. DETAILED DESCRIPTION

[0026] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0027] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0028] The embodiment of the present application provides a health status assessment method based on electronic health records. The method is executed by a computer device, and can be executed by a computer device such as a terminal or a server alone, or by a terminal and a server together. In the embodiment of the present application, Figure 1 As shown, the method includes the following steps 101-105.

[0029] Step 101: pre-process historical EHR data and real-time EHR data to determine pre-processed EHR data.

[0030] Step 102: Perform feature selection on the preprocessed EHR data through a machine learning algorithm, determine the selected EHR data, and use the processed EHR data as the data set to be clustered.

[0031] Step 103: updating multiple cluster center positions in the data set to be clustered according to the constructed fitness function, and determining the optimal cluster center position corresponding to each cluster center position based on an iterative convergence mechanism; the cluster center position is the position of the cluster center.

[0032] Step 104: clustering the data points corresponding to the optimal cluster center position to determine the EHR clustering result.

[0033] Step 105: Evaluate the user's health status according to the EHR clustering result.

[0034] In an exemplary embodiment, before step 101 , the process further includes: a data collection process.

[0035] The data collection is to collect historical EHR data and real-time EHR data; the historical EHR data and real-time EHR data both include basic information data, health status, functional data and physical examination data.

[0036] The collected historical EHR data and real-time EHR data include the subjects' basic information data, health status and functional data, and physical examination data as the original data sets. Among them, the basic information data is automatically entered into the system through questionnaires, including gender, age, date of birth, whether married, education, and income, etc.; the health status and functional data are generally entered into the system through the clinical diagnosis results of clinicians, including whether the patient suffers from hypertension, whether the patient suffers from other diseases such as diabetes, whether the motor function is good, whether the cognitive function is reduced, etc. The physical examination data are mainly routine physical examinations, including height, weight, systolic blood pressure, diastolic blood pressure, and body mass index.

[0037] In an exemplary embodiment, the data preprocessing in step 101 specifically includes: data cleaning, data conversion and standardization of the collected historical EHR data and real-time EHR data; wherein, data cleaning is to process missing values, duplicate values ​​and outliers; for missing values, multiple interpolation is performed using MissForest to make the data meet the distribution of the original data. If a piece of data is missing or the proportion of missing values ​​of a feature is too high, the data is deleted or the feature is deleted; duplicate data is deleted to prevent it from affecting the performance of the model; Isolation Forest is used for outliers; data conversion is to convert the data into vector form; standardization is to standardize the data based on the maximum and minimum normalization method.

[0038] The main steps of Isolation Forest include the following:

[0039] 1. Randomly select features and split points: Isolation forest randomly selects features and performs random splits on these features to generate multiple trees. The construction of each tree is based on splitting the data points into different regions.

[0040] 2. Split depth indicates isolation: For a point, if it can be isolated earlier in the tree segmentation process (the number of split layers is shallower), it means that it is more likely to be isolated and may be an outlier. The shallower the depth of the segmentation, the higher the abnormality of the data point.

[0041] 3. Anomaly score: The segmentation depth of each data point is calculated through the construction of the tree, and its anomaly score is calculated through these depths. Points with high scores are considered to be anomalies.

[0042] In an exemplary embodiment, step 102 may be replaced by the following steps 201 to 204 .

[0043] Step 201: Filter the pre-processed EHR data using the Least absolute shrinkage and selection operator (Lasso) algorithm to determine filtered EHR data.

[0044] Step 202: Using an adaptive least absolute shrinkage and selection operator (Adaptive lasso) algorithm to perform elimination processing on the pre-processed EHR data to determine the eliminated EHR data.

[0045] Step 203: using a generalized elastic net algorithm to evaluate the stability of the preprocessed EHR data, and determining the evaluated EHR data.

[0046] Step 204: Filter out the common EHR data among the filtered EHR data, the eliminated EHR data, and the evaluated EHR data, and use the common EHR data as the selected EHR data.

[0047] Since high-dimensional data will reduce the accuracy and efficiency of the model, three robust variable selection methods are used to reduce the dimension, namely Lasso, Adaptive lasso and Generalized elastic net.

[0048] Specifically, 1. A penalty term is added to the model estimation through the Lasso algorithm to compress the regression coefficient that is too small to zero. While compressing the coefficient, it plays the role of variable filtering, which can effectively avoid the problem of insufficient model probability due to overfitting.

[0049] 2. Use Adaptive lasso to remove unimportant variables more easily, thereby effectively reducing the bias of coefficient estimation. This method improves the Lasso algorithm by adding adaptive weights before the penalty term. The adaptive weight coefficient depends on the data itself. The degree of penalty for the regression coefficients of different variables is different. The weight coefficient of the variable with a larger initial coefficient estimate is smaller, so the degree of penalty is also smaller. On the contrary, the weight coefficient corresponding to the variable with a smaller initial coefficient estimate is larger, and the degree of penalty is also larger, thereby achieving better variable selection.

[0050] 3. The generalized elastic net is used to improve the stability of coefficient estimation (reduce variance). In addition, this method also has the stability of ridge regression and the interpretability of subset selection. The common parts selected by the three methods are sent to the model as the result of feature selection and used as the feature set of the model.

[0051] Through the above operations, the feature selection method is relatively simple, which often leads to low accuracy, robustness and stability of feature selection, making the model insufficient in processing high-dimensional, collinear and noisy data. This application uses Lasso, Adaptive lasso and Generalized elastic net to select features from multiple angles, improve the generalization ability and interpretability of the model, make the processing of electronic health record data more flexible and efficient, and thus improve the adaptability and robustness of health status classification.

[0052] In an exemplary embodiment, step 103 may be replaced by the following steps 301 to 305 .

[0053] Step 301: randomly select multiple data points from the data set to be clustered as initial cluster centers, and determine the positions of the initial cluster centers.

[0054] Step 302: Based on the initial cluster center positions, define the intra-cluster cohesion and inter-cluster separation metrics.

[0055] Step 303: construct a fitness function according to the intra-cluster cohesion and the inter-cluster separation metric.

[0056] Step 304: Calculate the fitness value of each cluster center according to the fitness function, and update the access level and cluster center position of the cluster center.

[0057] Step 305: Based on the iterative convergence mechanism, the cluster center position with the maximum fitness value is used as the optimal cluster center position.

[0058] In an exemplary embodiment, step 305 may be replaced by the following steps 401 to 404 .

[0059] Step 401: for each cluster center, determine whether the fitness value of the updated cluster center is higher than the fitness value before the update; if so, execute step 402; if not, execute step 403.

[0060] Step 402: Continue to update the access level of the cluster center position and the cluster center position.

[0061] Step 403: The access level of the cluster center position and the cluster center position are not updated.

[0062] Step 404: until the iterative update of the cluster center position reaches convergence, the cluster center position with the maximum fitness value is taken as the optimal cluster center position.

[0063] In an exemplary embodiment, the fitness function fit AHA (CS) is: fit AHA (CS) = Ces(CS) - Ss(CS); where Ces(CS) is the minimum intra-cluster cohesion, CS is the cluster set, Ce(c k ) is cluster c k The cohesion of k is the cluster number, K is the total number of clusters; Ss(CS) is the separation measure between clusters.

[0064] In an exemplary embodiment, the present application updates the cluster center positions by designing a fitness function, and obtains the optimal cluster center set based on an iterative convergence mechanism; the optimal cluster center set includes the optimal cluster center position corresponding to each cluster center position; different types of data are allocated to the data points based on the relative distance set and the minimum distance of the data points, thereby obtaining EHR clustering results.

[0065] The clustering classification specifically includes the following steps 501 to 503.

[0066] Step 501: Initialization: randomly select K data points from the data set as initial cluster centers.

[0067] Step 502: Define a fitness function.

[0068] Step 5021: Define the intra-cluster cohesion.

[0069]

[0070] Among them, Ce(c k ) is cluster c k The cohesion degree; x i is the data point; dist(·)· is the Euclidean distance; k is the cluster number.

[0071] Step 5022: Define the inter-cluster separation metric.

[0072] Ss(CS)=maxdist(c i3 ,c j3 );

[0073] Among them, Ss(CS) is the inter-cluster separation measure; CS is the cluster set; c i3 ,c j3 are the two cluster centers.

[0074] Step 5023: Define the fitness function.

[0075] fit AHA (CS) = Ces(CS) - Ss(CS);

[0076]

[0077] Among them, Ces(CS) is the cohesion of the entire cluster set; fit AHA (CS) is the fitness value function; K is the total number of clusters.

[0078] Step 503: Optimization.

[0079] For each cluster center, the fitness value of the location is calculated, and the visit level and location are updated; the location with the maximum fitness value is taken as the optimal cluster center.

[0080]

[0081] x new =x old +φ(x old -x target );

[0082] In the formula, C opt is the optimal cluster center set; x new and x oldare the updated position and the updated position respectively; φ is the update coefficient; x target is the location of the optimal cluster center.

[0083] Step 503: Iteration.

[0084] For each cluster center, if the updated fitness value is higher than the fitness value before the update, the position is updated; otherwise, the position is not updated; the position update is iterated repeatedly until convergence.

[0085] In an exemplary embodiment, step 104 may be replaced by the following steps 601 to 603 .

[0086] Step 601: Divide the core area and the edge area based on the optimal cluster center position.

[0087] Step 602: For the core data points of the core area, determine the relative distance set between the core data points, and classify the core data points into an upper approximate set of the cluster where the core data points are located based on the relative distance set, and determine the EHR clustering results corresponding to the core data points.

[0088] Step 603: For edge data points in the edge area, determine the minimum distance between edge data points, classify the edge data points into a cluster with the minimum distance from the cluster where the edge data points are located, and determine the EHR clustering result corresponding to the edge data points.

[0089] In an exemplary embodiment, the clustering process is to perform a k-means clustering on the data set using the converged initial cluster centers in advance, and the subsequent clustering process specifically includes the following steps 701 to 703.

[0090] Step 701: Region division.

[0091] Divide the core area, expressed as:

[0092]

[0093] Divide the edge area, expressed as:

[0094]

[0095] in,

[0096] The upper approximation set of the cluster is defined as in, is the core area of ​​the kth cluster; x is a data point; TypeA and TypeB are two types of data point sets; is the upper approximation set of the kth cluster; is an expansion set; is the number of data points in the expansion set; is the optimal cluster center of the kth cluster; is the updated cluster center; β is the core area threshold; is the distance between cluster centers before and after the update; x 1 、x 2 and is the extended data point; is the upper approximate set of the jth cluster. The upper approximate set refers to the data points that may belong to a certain class or cluster, including points that may be in the boundary area. Although these points may not necessarily belong to a certain cluster, they have a greater connection with the cluster. The upper approximate set usually includes data points that are in fuzzy boundaries or have uncertain affiliation; K is the total number of clusters; is the edge region of the kth cluster.

[0097] Step 702: Data point allocation.

[0098] For the data points in the core area, a relative distance set is defined, expressed as Classify data points into the upper approximate set of clusters, expressed as

[0099] For data points in the edge area, the minimum distance is defined as Assign data to the cluster with the smallest distance; where Z x is the relative distance set of x; is the cluster center of the jth cluster; is the nearest cluster center; α is the decision threshold.

[0100] Step 703: Cluster determination.

[0101] When clustering reaches convergence or a maximum number of iterations, clustering ends; clustering results of historical electronic health record data and real-time electronic health record data are obtained.

[0102] The features selected from historical EHR data and real-time EHR data are input into the model, and different levels of populations of concern are divided through statistical results of the data within the cluster. Different levels of populations of concern are the clustering results of historical electronic health record data and real-time electronic health record data.

[0103] By performing the above operations, the present application can update the cluster center position through the fitness function, obtain the optimal cluster center set based on the iterative convergence mechanism, thereby improving the stability and accuracy of clustering. The present application can provide more detailed clustering results, so that each data point can be reasonably assigned to an appropriate cluster according to its relative position, improve the classification accuracy of the cluster, and thus make the accuracy of the final health status classification better.

[0104] Based on the same inventive concept, the embodiment of the present application also provides a health status assessment device based on electronic health records for implementing the above-mentioned health status assessment method based on electronic health records. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above-mentioned method, so the specific limitations in one or more embodiments of the health status assessment device based on electronic health records provided below can refer to the limitations of the health status assessment method based on electronic health records above, and will not be repeated here.

[0105] In an exemplary embodiment, a health status assessment device based on electronic health records is provided, comprising:

[0106] The data preprocessing module is used to preprocess the historical EHR data and the real-time EHR data and determine the preprocessed EHR data.

[0107] The feature selection module is used to perform feature selection on the preprocessed EHR data through a machine learning algorithm, determine the selected EHR data, and use the processed EHR data as a data set to be clustered.

[0108] The optimal cluster center position determination module is used to update multiple cluster center positions in the data set to be clustered according to the constructed fitness function, and determine the optimal cluster center position corresponding to each cluster center position based on an iterative convergence mechanism; the cluster center position is the position of the cluster center.

[0109] The clustering processing module is used to perform clustering processing on the data points corresponding to the optimal cluster center position to determine the EHR clustering result.

[0110] An evaluation module is used to evaluate the health status of the user according to the EHR clustering result.

[0111] In an exemplary embodiment, a computer device is provided, which may be a server or a terminal. The computer device includes a processor, a memory, an input / output interface (I / O for short) and a communication interface. The processor, the memory and the input / output interface are connected via a system bus, and the communication interface is connected to the system bus via the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store health status assessment data based on electronic health records. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, a health status assessment method based on electronic health records is implemented.

[0112] In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the above method is implemented when the processor executes the computer program.

[0113] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, which implements the above method when executed by a processor.

[0114] In an exemplary embodiment, a computer program product is provided, including a computer program, which implements the above method when executed by a processor.

[0115] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ReadOnlyMemory, ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (Magnetoresistive RandomAccess Memory, MRAM), ferroelectric random access memory (Ferroelectric RandomAccess Memory, FRAM), phase change memory (Phase Change Memory, PCM), graphene memory, etc. Volatile memory can include random access memory (RandomAccess Memory, RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0116] In this application, all actions to obtain signals, information or data are carried out in compliance with the relevant data protection laws and policies of the country where they are located and with the authorization given by the owner of the corresponding device.

[0117] The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on blockchain, etc., but is not limited thereto. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but is not limited thereto.

[0118] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0119] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A health status assessment method based on electronic health records, characterized in that: The health status assessment method based on electronic health records includes: Preprocessing historical EHR data and real-time EHR data to determine preprocessed EHR data; Performing feature selection on the preprocessed EHR data by a machine learning algorithm, determining the selected EHR data, and using the processed EHR data as a data set to be clustered; According to the constructed fitness function, multiple cluster center positions in the data set to be clustered are updated, and based on an iterative convergence mechanism, the optimal cluster center position corresponding to each cluster center position is determined; the cluster center position is the position of the cluster center; Performing clustering processing on the data points corresponding to the optimal cluster center position to determine the EHR clustering result; The user's health status is evaluated according to the EHR clustering result.

2. The method for health status assessment based on electronic health records according to claim 1, characterized in that: Performing feature selection on the preprocessed EHR data by a machine learning algorithm to determine the selected EHR data specifically includes: Using the Lasso algorithm to filter the preprocessed EHR data to determine filtered EHR data; Using the Adaptive lasso algorithm to perform elimination processing on the pre-processed EHR data to determine the eliminated EHR data; Using a generalized elastic net algorithm to evaluate the stability of the preprocessed EHR data, and determining the evaluated EHR data; The common EHR data among the filtered EHR data, the eliminated EHR data and the evaluated EHR data are screened out, and the common EHR data are used as the selected EHR data.

3. The method for health status assessment based on electronic health records according to claim 1, characterized in that: The plurality of cluster center positions in the to-be-clustered data set are updated according to the constructed fitness function, and the optimal cluster center position corresponding to each cluster center position is determined based on an iterative convergence mechanism, specifically including: Randomly select multiple data points from the data set to be clustered as initial cluster centers, and determine the positions of the initial cluster centers; Based on the initial cluster center positions, defining intra-cluster cohesion and inter-cluster separation metrics; Constructing a fitness function based on the intra-cluster cohesion and the inter-cluster separation metric; Calculate the fitness value of each cluster center according to the fitness function, and update the access level and cluster center position of the cluster center; Based on the iterative convergence mechanism, the cluster center position with the maximum fitness value is taken as the optimal cluster center position.

4. The method for health status assessment based on electronic health records according to claim 3, characterized in that: Based on the iterative convergence mechanism, the cluster center position with the maximum fitness value is taken as the optimal cluster center position, which includes: For each cluster center, determine whether the fitness value of the updated cluster center is higher than the fitness value before the update; If so, continue to update the access level and cluster center position of the cluster center; If not, the access level and cluster center position of the cluster center are not updated; Until the cluster center position is iteratively updated to converge, the cluster center position with the maximum fitness value is taken as the optimal cluster center position.

5. The method for health status assessment based on electronic health records according to claim 3, characterized in that: The fitness function fit AHA (CS) is: fit AHA (CS)=Ces(CS)-Ss(CS); Among them, Ces(CS) is the minimum intra-cluster cohesion, CS is the cluster set, Ce(c k ) is cluster c k The cohesion of k is the cluster number, K is the total number of clusters; Ss(CS) is the separation measure between clusters.

6. The method for health status assessment based on electronic health records according to claim 1, characterized in that: Clustering the data points corresponding to the optimal cluster center position to determine the EHR clustering result specifically includes: Dividing the core area and the edge area based on the optimal cluster center position; For the core data points in the core area, determine a relative distance set between the core data points, and classify the core data points into an upper approximate set of the cluster where the core data points are located according to the relative distance set, and determine the EHR clustering results corresponding to the core data points; For the edge data points in the edge area, the minimum distance between the edge data points is determined, and the edge data points are classified into a cluster having the minimum distance from the cluster where the edge data points are located, and the EHR clustering result corresponding to the edge data points is determined.

7. A health status assessment device based on electronic health records, characterized in that: The health status assessment device based on electronic health records includes: A data preprocessing module, used for preprocessing historical EHR data and real-time EHR data, and determining preprocessed EHR data; A feature selection module, for performing feature selection on the preprocessed EHR data through a machine learning algorithm, determining the selected EHR data, and using the processed EHR data as a data set to be clustered; An optimal cluster center position determination module is used to update multiple cluster center positions in the data set to be clustered according to the constructed fitness function, and determine the optimal cluster center position corresponding to each cluster center position based on an iterative convergence mechanism; the cluster center position is the position of the cluster center; A clustering processing module, used to perform clustering processing on the data points corresponding to the optimal cluster center position to determine the EHR clustering result; An evaluation module is used to evaluate the health status of the user according to the EHR clustering result.

8. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the health status assessment method based on electronic health records as described in any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for assessing health status based on electronic health records described in any one of claims 1 to 6 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method for assessing health status based on electronic health records described in any one of claims 1 to 6 is implemented.

Images