Cardiovascular event risk prediction method, device, equipment, medium and product

Abstract

The invention discloses a cardiovascular event risk prediction method, device, equipment, medium and product, and relates to the technical field of medical auxiliary equipment.The method comprises the steps that single-factor Cox regression analysis and multi-factor Cox regression analysis are utilized, and multiple prediction factors are selected from all prediction factors to serve as covariables; on the basis of the data set and the multiple covariables, constructing a cardiovascular event risk prediction model by using an LASSO regression analysis principle; the cardiovascular event risk prediction model is used for completing cardiovascular event risk value prediction; the cardiovascular event risk prediction model is constructed and applied to medical auxiliary tools such as hemodialysis, the risk prediction precision of the medical auxiliary tools on cardiovascular events can be improved, and meanwhile the reasonability of assisting medical staff is improved.

Description

Technical Field

[0001] The present application relates to the technical field of medical auxiliary equipment, and in particular to a method, device, equipment, medium and product for predicting the risk of cardiovascular events. Background Art

[0002] International renal guidelines recommend early detection and prevention of cardiovascular events in maintenance hemodialysis patients. Identifying potential pathophysiological factors for cardiovascular events in maintenance hemodialysis patients is crucial to reducing the occurrence of cardiovascular events and mortality, specifying individualized treatment plans, and preventing clinically important complications. Existing clinical prediction models can effectively predict the risk of death in hemodialysis patients to a certain extent, but the sample size is small, single-center, and cross-sectional studies are mostly related to non-cardiovascular events of hemodialysis. Summary of the invention

[0003] The purpose of this application is to provide a cardiovascular event risk prediction method, device, equipment, medium and product, which can be applied to medical auxiliary tools such as hemodialysis, can improve the risk prediction accuracy of cardiovascular events by medical auxiliary tools, and improve the rationality of assisting medical staff.

[0004] To achieve the above objectives, this application provides the following solutions:

[0005] In a first aspect, the present application provides a method for predicting cardiovascular event risk, comprising:

[0006] Get multiple predictors from the dataset;

[0007] Using univariate Cox regression analysis and multivariate Cox regression analysis, multiple predictors were selected as covariates from all predictors;

[0008] Based on the data set and the multiple covariates, a cardiovascular event risk prediction model is constructed using the LASSO regression analysis principle;

[0009] Obtain all covariate measurement values ​​of the subjects to be tested to construct the test matrix;

[0010] The matrix to be tested is input into the cardiovascular event risk prediction model to determine the risk value of the subject to be tested for cardiovascular events.

[0011] Optionally, the data set includes predictor groups corresponding to multiple maintenance hemodialysis patients; the predictors in one predictor group include: age, dialysis mode, whether hypotension occurs during dialysis, waist-to-hip ratio, grip strength, extracellular water, lean tissue index, lymphocytes, platelet distribution density, total bilirubin, dialysis age, whether hypertension is present, and whether diabetes is present.

[0012] Optionally, after getting multiple predictors from the dataset, also include:

[0013] Deleting outliers in the data set;

[0014] Multiple imputation method was used to handle missing values ​​in the dataset;

[0015] The dataset is divided into a training set, an internal validation set, and an external validation set.

[0016] Optionally, using univariate Cox regression analysis and multivariate Cox regression analysis, multiple predictors are selected from all predictors as covariates, including:

[0017] Using R software to perform univariate Cox regression analysis on each of the predictors, and determining all predictors with a P value less than a first threshold as the first undetermined factors;

[0018] Using the Cox.zph function in the R software to perform a PH hypothesis analysis on the first undetermined factor, determining all first undetermined factors with a P value less than a second threshold as second undetermined factors;

[0019] Multivariate Cox regression analysis was performed on all the second undetermined factors to determine the covariates in the second undetermined factors.

[0020] Optionally, based on the data set and the multiple covariates, a cardiovascular event risk prediction model is constructed using the LASSO regression analysis principle, including:

[0021] All covariates corresponding to the same maintenance hemodialysis patient in the training set are constructed into a matrix to obtain multiple matrices;

[0022] Taking the matrix as input and the risk value of cardiovascular events in maintenance hemodialysis patients as output, the training set is fitted using the glmnet function to obtain a LASSO model;

[0023] Based on the objective function, the LASSO model is cross-validated using the internal validation set and the external validation set to obtain a cardiovascular event risk prediction model.

[0024] Optionally, the objective function is:

[0025]

[0026] Among them, L(β) is the objective function, X i is the covariate matrix of the ith observation, β j is the coefficient of the jth covariate, λ is the penalty parameter, n is the number of samples, y iis the actual value of the ith observation, β is the regression coefficient vector, and p is the number of variables.

[0027] In a second aspect, the present application provides a cardiovascular event risk prediction device, comprising:

[0028] A predictor acquisition module is used to obtain multiple predictors from a data set;

[0029] A covariate determination module is used to select multiple predictors as covariates from all predictors using univariate Cox regression analysis and multivariate Cox regression analysis;

[0030] A cardiovascular event risk prediction model construction module is used to construct a cardiovascular event risk prediction model based on the data set and the multiple covariates using the LASSO regression analysis principle;

[0031] The module for determining the matrix to be tested is used to obtain all the covariate measurement values ​​of the subject to be tested and construct the matrix to be tested;

[0032] The risk value prediction module is used to input the matrix to be tested into the cardiovascular event risk prediction model to determine the risk value of the subject to be tested for cardiovascular events.

[0033] 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 the above-mentioned cardiovascular event risk prediction method.

[0034] 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 the above-mentioned cardiovascular event risk prediction method.

[0035] In a fifth aspect, the present application provides a computer program product, including a computer program, which implements the above-mentioned cardiovascular event risk prediction method when executed by a processor.

[0036] According to the specific embodiments provided in this application, this application discloses the following technical effects:

[0037] The present application provides a method, device, equipment, medium and product for predicting the risk of cardiovascular events, which obtains multiple predictors from a data set; uses univariate Cox regression analysis and multivariate Cox regression analysis to select multiple predictors from all predictors as covariates; based on the data set and multiple covariates, uses the LASSO regression analysis principle to construct a cardiovascular event risk prediction model; obtains all covariate measurement values ​​of the subject to be tested to construct a test matrix; inputs the test matrix into the cardiovascular event risk prediction model to determine the risk value of the subject to be tested for cardiovascular events; establishes predictors through statistical analysis methods and multi-model comparison, and constructs and verifies a simple and easy-to-use cardiovascular event risk prediction model for maintenance hemodialysis patients, and explores the risk factors that lead to cardiovascular events in hemodialysis patients. Help clinicians develop individualized treatment and management strategies for patients with end-stage renal disease, reduce the risk of death, and prevent clinically important complications. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] 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.

[0039] Figure 1 This is a flow chart of a method for predicting cardiovascular event risk in one embodiment of the present application.

[0040] Figure 2 It is a structural schematic diagram of a computer device in one embodiment of the present application. DETAILED DESCRIPTION

[0041] 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.

[0042] 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.

[0043] In an exemplary embodiment, Figure 1 As shown, a method for predicting the risk of cardiovascular events is provided, comprising:

[0044] Step 101: Obtain multiple predictors from a data set. The data set includes predictor groups corresponding to multiple maintenance hemodialysis patients. The predictors in one predictor group include: age, dialysis mode, whether hypotension occurs during dialysis, waist-to-hip ratio, grip strength, extracellular water, lean tissue index, lymphocytes, platelet distribution density, total bilirubin, dialysis age, whether hypertension is present, and whether diabetes is present.

[0045] Data source and study population: 3488 maintenance hemodialysis patients from 20 hemodialysis centers from June 2015 to August 2020 were included and followed up until August 2021. Exclusion criteria: ① Hemodialysis patients under 18 years old; ② Previous cardiovascular events; ③ Combined with malignant tumors; ④ Mental illness or severe aphasia; ⑤ Patients who were unable to cooperate with the questionnaire and refused to participate in this study; ⑥ Patients with more than 20% missing total data. A total of 2812 patients met the criteria and were included in the study.

[0046] The model was constructed using physical measurement data and available laboratory indicators as predictors. The predictors included (13 predictors, including age, dialysis mode, whether hypotension occurred during dialysis, waist-to-hip ratio, grip strength, extracellular water, lean tissue index, lymphocytes, platelet distribution density, total bilirubin, dialysis age, and whether hypertension and / or diabetes were present). Before dialysis, researchers who had received unified training used tape measures, weight scales, and sphygmomanometers to collect baseline data from patients, including height, predialysis weight, waist circumference, hip circumference, calf circumference, and blood pressure before, during, and after dialysis. Bioelectrical impedance meter (Germany, Fresenius BCM-7BJA4951) was used to measure indicators such as extracellular and extracellular water, somatic cell index, and lean tissue index. Professional researchers uniformly distributed questionnaires and conducted one-on-one detailed inquiries, including the patient's gender, age, medical history, and medication status. Finally, the electronic medical record management system of each hospital was visited to collect laboratory test data from patients in the past 3 months, such as blood routine, serum creatinine, and electrolytes.

[0047] After step 101, the method further includes:

[0048] Step 106: Delete outliers in the data set.

[0049] Step 107: Use multiple interpolation to process missing values ​​in the data set.

[0050] Step 108: Divide the data set into a training set, an internal validation set, and an external validation set.

[0051] To derive and validate the model, the data were randomly split into a training set (70% of the total sample) and internal and external validation sets (20% and 10%, respectively). The model was constructed using the training set, and the performance of the model was evaluated by internal and external validation. For missing data, the potential bias of missing data was assessed by comparing the characteristics of patients with one or more missing predictor variables with those with complete data. When the data were missing more than 80%, they were deleted directly. Other missing data were assumed to be missing at random and imputed using multiple imputation. Due to the large sample size and less than 1% of the total outliers, the outliers were deleted and treated as missing values.

[0052] Step 102: Using univariate Cox regression analysis and multivariate Cox regression analysis, multiple predictors are selected from all predictors as covariates.

[0053] Step 102 includes:

[0054] Step 102-1: Use R software to perform univariate Cox regression analysis on each prediction factor, and determine all prediction factors with P values ​​less than the first threshold as the first undetermined factors.

[0055] Step 102-2: Use the Cox.zph function in the R software to perform a PH hypothesis analysis on the first undetermined factor, and determine that all first undetermined factors with a P value less than the second threshold are second undetermined factors.

[0056] Step 102-3: Perform multivariate Cox regression analysis on all the second factors to be determined to determine the covariates in the second factors to be determined.

[0057] First, univariate and multivariate survival Cox regression analysis was performed using R software.

[0058] After univariate Cox regression analysis, significant variables (P < 0.05) were analyzed using the Cox.zph function for PH assumption. The Cox.zph function comes from the survival package, which uses the Schoenfeld residual to test whether the Cox regression model meets the proportional hazard assumption. The key steps are as follows:

[0059] 1. Make sure the survival package is installed and loaded.

[0060] 2. You need to fit a Cox model first, using the coxph function.

[0061] 3. Perform proportional hazards assumption test: After fitting the Cox model, the cox.zph function can be used to check the PH assumption.

[0062] 4. Explain the test results The cox.zph function will output the following:

[0063] Globaltest: Whether the entire model violates the proportional hazards assumption.

[0064] The test results for each covariate include a test of the correlation between the Schoenfeld residual and time. If the p-value is very small (usually less than 0.05), it indicates that the corresponding covariate does not meet the PH assumption.

[0065] 5. Visualize the analysis results

[0066] You can further determine whether the proportional hazards assumption holds by visualizing the Schoenfeld residuals versus time. Generally speaking, the residual plot should be a horizontal line, indicating no trend over time.

[0067] Cox.zph is based on Schoenfeld residuals for testing, and its basic principles can be summarized as follows:

[0068] Schoenfeld residuals: The residuals represent the difference between the estimated risk factor contribution for each observation and the risk factor contribution predicted by the model. By examining the relationship between the residuals and time, we can see whether the hazard ratio of the covariate changes over time.

[0069] Formula: The test of the PH hypothesis is based on the following formula:

[0070] The Schoenfeld residual Rj is: Rj=Oj-Ej.

[0071] Among them, Oj represents the covariate value when the event occurs at time tj, and Ej represents the expected covariate value at that moment.

[0072] The Schoenfeld test determines whether the proportional hazard of the covariate remains constant by analyzing the time dependence of the residuals, that is, calculating the correlation between the residuals and time.

[0073] (P-value>0.05 meets the PH assumption) to ensure that the risk ratio of different individuals is a constant, that is, the risk ratio is constant over time.

[0074] Step 103: Based on the data set and multiple covariates, a cardiovascular event risk prediction model is constructed using the LASSO regression analysis principle.

[0075] Step 103 includes:

[0076] Step 103 - 1 : construct all covariates corresponding to the same maintenance hemodialysis patient in the training set into a matrix to obtain multiple matrices.

[0077] Step 103-2: Using the matrix as input and the risk value of cardiovascular events in maintenance hemodialysis patients as output, the training set is fitted using the glmnet function to obtain the LASSO model.

[0078] Step 103-3: Based on the objective function, the LASSO model is cross-validated using the internal validation set and the external validation set to obtain a cardiovascular event risk prediction model. The objective function is:

[0079] Among them, L(β) is the objective function, X i is the covariate matrix of the ith observation, β j is the coefficient of the jth covariate, λ is the penalty parameter, n is the number of samples, y i is the actual value of the ith observation, β is the regression coefficient vector, and p is the number of variables.

[0080] The independent variables that met the PH assumption were subjected to multivariate Cox regression analysis, and the survival data were subjected to Cross-LASSO regression analysis using the GlmnetR package. The key steps are as follows:

[0081] First install and load the glmnet package.

[0082] Data Preparation,The input of LASSO regression must be numerical matrices and vectors,,so the data needs to be converted into appropriate formats.

[0083] Use glmnet function to fit the LASSO model.

[0084] In order to choose the best penalty parameter λ (or lambdaλ) value, cross validation is usually used. You can use the cv.glmnet function to perform cross validation.

[0085] Refit the model using the optimal λ value. After finding the optimal λ value, you can refit the LASSO model with this value.

[0086] Prediction and evaluation: You can use the fitted model to predict new data. You can also evaluate the prediction effect of the model by calculating evaluation indicators such as mean square error (MSE).

[0087] In LASSO regression, the objective function is:

[0088]

[0089] λ is the penalty parameter, which is used to control the regularization strength.

[0090] Cross-LASSO regression combines the methods of cross-validation and LASSO regression, and can comprehensively consider the balance between feature selection and model complexity. This application performs multivariate Cox regression analysis based on Cross-LASSO regression analysis, which integrates the results of predictor feature selection and the effects of adjustment variables to more accurately estimate the risk ratio.

[0091] Finally, the optimal parameter set [mainly based on the consistency index (Harrell's C-index)] is determined by balancing the number of independent variables and the performance of the model (mainly through, Cross-LASSO regression combines the methods of cross-validation and LASSO regression, which can comprehensively consider the balance between feature selection and model complexity).

[0092] Based on the optimized parameters, the Cox proportional hazard model was used to draw a visual nomogram. The prediction performance of the model was evaluated by calculating the area under the training set operating characteristic curve (AUC) and the C index.

[0093]

[0094] Where: h i and h j is the model's predicted value for samples i and j (such as risk score or predicted survival time); y i and j is the actual observation time of samples i and j; I() is an indicator function, which takes the value of 1 when the condition is true and 0 otherwise. The C index indicates the consistency between the predicted and actual results, which is calibrated by comparing the average of the predicted survival rate with the actual survival rate. The reliability of the model was verified using 10-fold cross validation and bootstrap resampling (70% training set and 30% validation set, repeated 1000 times).

[0095] Step 104: Obtain all covariate measurement values ​​of the subjects to be tested to construct a test matrix.

[0096] Step 105: Input the matrix to be tested into the cardiovascular event risk prediction model to determine the risk value of the subject to be tested for cardiovascular events.

[0097] All analyses were performed using R version 4.3.0 (the R Foundation for Statistical Computing, version 4.3.0 https: / / www.r-project.org / ). Continuous variables were expressed as median [interquartile range (IQR)] and compared using the Mann-Whitney test. Categorical variables were expressed as frequency and percentage (%), and the Chi-square test was used for comparison between groups. P < 0.05 was considered statistically significant. Multicollinearity between predictor variables was assessed by correlation coefficient and variance inflation factor. Cross-LASSO regression analysis was performed using the survival, corrplot, glmnet, and foreignR packages. PH assumption analysis was performed using the Cox.zphR package. Cox proportional hazard model and consistency index analysis were performed using the RMS, survival, pec, and HmiscR packages. The timeROC package was used to draw the receiver operating characteristic (ROC) curve, the caret and riskRegressionR packages were used to perform 1000 10-fold cross validations, and the bootstrapR package was used to perform 1000 repeated bootstrap resampling validations.

[0098] Compared with existing models, the clinical prediction model of this application has many advantages. For example, Xu et al. included 388 regular hemodialysis patients recruited from the hemodialysis centers of the Third Hospital of Southern Medical University and the Third Hospital of Sun Yat-sen University from January 1, 2015 to September 30, 2019, and followed up until May 31, 2020. Through Cox proportional hazard regression analysis, age, hypertension, diabetes, and abnormal white blood cell count were screened as predictors among 26 clinical parameters, and the prediction score of cardiovascular event risk in Chinese hemodialysis patients was derived and verified. The results showed that 388 patients had no cardiovascular events at baseline, and 132 patients had their first cardiovascular event during an average follow-up of 3.27 years. In the end, a relatively simple clinical risk prediction model was presented, but this study only included data on dialysis patients from two Class A tertiary public hospitals in Guangzhou, lacked external data, and lacked many risk predictors for cardiovascular events in maintenance hemodialysis patients, such as volume status, residual renal function, nutritional status, and other information. In the study of Matsubara Y et al., the follow-up time of the model of this application was much longer than that of this model. It is important that this application conducted external validation of the model. Compared with the model of Guode Li et al., the sample size of the model of this application is large enough, the model is more concise, practical and stable, and the prediction indicators are intuitively derived from clinical work, which has high practicality.

[0099] In an exemplary embodiment, a cardiovascular event risk prediction device is provided, comprising:

[0100] The predictor acquisition module is used to obtain multiple predictors from the data set.

[0101] The covariate determination module is used to select multiple predictors as covariates from all predictors by using univariate Cox regression analysis and multivariate Cox regression analysis.

[0102] The cardiovascular event risk prediction model construction module is used to construct a cardiovascular event risk prediction model based on the data set and multiple covariates using the LASSO regression analysis principle.

[0103] The module for determining the matrix to be tested is used to obtain all covariate measurement values ​​of the subject to be tested to construct the matrix to be tested.

[0104] The risk value prediction module is used to input the matrix to be tested into the cardiovascular event risk prediction model to determine the risk value of the subject to be tested for cardiovascular events.

[0105] In an exemplary embodiment, a computer device is provided. The computer device may be a server or a terminal. The internal structure diagram thereof may be as follows: Figure 2 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through 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 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 through a network connection. When the computer program is executed by the processor, a method for predicting the risk of cardiovascular events is implemented.

[0106] Those skilled in the art will understand that Figure 2The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components. 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 processor implements the steps in the above-mentioned method embodiments when executing the computer program.

[0107] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0108] In an exemplary embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0109] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0110] Those of ordinary skill 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 (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access 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).

[0111] 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.

[0112] 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.

[0113] 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 method for predicting cardiovascular event risk, characterized in that: include: Get multiple predictors from the dataset; Using univariate Cox regression analysis and multivariate Cox regression analysis, multiple predictors were selected as covariates from all predictors; Based on the data set and the multiple covariates, a cardiovascular event risk prediction model is constructed using the LASSO regression analysis principle; Obtain all covariate measurement values ​​of the subjects to be tested to construct the test matrix; The matrix to be tested is input into the cardiovascular event risk prediction model to determine the risk value of the subject to be tested for cardiovascular events.

2. The method for predicting cardiovascular event risk according to claim 1, characterized in that: The data set includes a predictor group corresponding to a plurality of maintenance hemodialysis patients; The predictors in one predictor group included age, dialysis mode, intradialytic hypotension, waist-to-hip ratio, grip strength, extracellular water, lean tissue index, lymphocytes, platelet distribution density, total bilirubin, dialysis age, hypertension, and diabetes.

3. The method for predicting cardiovascular event risk according to claim 1, characterized in that: After getting multiple predictors from the dataset, also include: Removing outliers from the data set; Multiple imputation method was used to handle missing values ​​in the dataset; The dataset is divided into a training set, an internal validation set, and an external validation set.

4. The method for predicting cardiovascular event risk according to claim 1, characterized in that: Using univariate Cox regression analysis and multivariate Cox regression analysis, multiple predictors were selected as covariates from all predictors, including: Using R software to perform univariate Cox regression analysis on each of the predictors, and determining all predictors with a P value less than a first threshold as the first undetermined factors; Using the Cox.zph function in the R software to perform a PH hypothesis analysis on the first undetermined factor, determining all first undetermined factors with a P value less than a second threshold as second undetermined factors; Multivariate Cox regression analysis was performed on all the second undetermined factors to determine the covariates in the second undetermined factors.

5. The method for predicting cardiovascular event risk according to claim 3, characterized in that: Based on the data set and the multiple covariates, a cardiovascular event risk prediction model is constructed using the LASSO regression analysis principle, including: All covariates corresponding to the same maintenance hemodialysis patient in the training set are constructed into a matrix to obtain multiple matrices; Taking the matrix as input and the risk value of cardiovascular events in maintenance hemodialysis patients as output, the training set is fitted using the glmnet function to obtain a LASSO model; Based on the objective function, the LASSO model is cross-validated using the internal validation set and the external validation set to obtain a cardiovascular event risk prediction model.

6. The method for predicting cardiovascular event risk according to claim 5, characterized in that: The objective function is: Among them, L(β) is the objective function, X i is the covariate matrix of the ith observation, β j is the coefficient of the jth covariate, λ is the penalty parameter, n is the number of samples, y i is the actual value of the ith observation, β is the regression coefficient vector, and p is the number of variables.

7. A cardiovascular event risk prediction device, characterized in that: include: A predictor acquisition module is used to obtain multiple predictors from a data set; A covariate determination module is used to select multiple predictors as covariates from all predictors using univariate Cox regression analysis and multivariate Cox regression analysis; A cardiovascular event risk prediction model construction module is used to construct a cardiovascular event risk prediction model based on the data set and the multiple covariates using the LASSO regression analysis principle; The module for determining the matrix to be tested is used to obtain all covariate measurement values ​​of the subject to be tested and construct the matrix to be tested; The risk value prediction module is used to input the matrix to be tested into the cardiovascular event risk prediction model to determine the risk value of the subject to be tested for cardiovascular events.

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 cardiovascular event risk prediction method according to 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 predicting the risk of cardiovascular events according to 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 predicting the risk of cardiovascular events according to any one of claims 1 to 6 is implemented.

Images