Post-stroke cognitive impairment prediction model construction method and device, equipment and medium
By constructing a prediction model for acute and early onset poststroke cognitive impairment based on historical data, the problem of lack of effective prediction models in the existing technology is solved, and prediction accuracy is improved, and early intervention and personalized treatment is supported.
Patent Information
- Application Number
- CN202510276359.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-10
- Publication Date
- 2025-06-10
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The existing technology lacks effective objective prediction models and cannot achieve acute and medium- and long-term early warnings for post-stroke cognitive impairment, which limits the formulation of early intervention and personalized treatment strategies.
By collecting historical data of stroke patients, statistical methods and machine learning methods were used to construct predictive models of acute and early-onset cognitive impairment after stroke, which were used to predict the probability of cognitive impairment within 2 weeks and 3 to 6 months, respectively.
Improves the prediction accuracy of post-stroke cognitive impairment, reduces prediction errors due to uncertain patient demonstration behavior, and supports early intervention and personalized treatment.
Smart Images
Figure CN120126784A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of intelligent medical technology, and particularly to a method, device, equipment and medium for constructing a prediction model for post-stroke cognitive impairment. Background Art
[0002] Common problems of Post-Stroke Cognitive Impairment (PSCI) are memory loss and decreased computing ability. In some patients, the situation may be more serious, with a decline in executive function, which will affect the quality of life of the patients. In related technologies, the diagnosis of PSCI mainly uses screening scales. By asking patients to perform a series of demonstrations, scores are given to various functions, and combined with the clinical diagnosis results of stroke, the development of cognitive impairment in patients is comprehensively evaluated. There is a problem that the accuracy of judgment is affected by artificial and objective factors. Currently, the research lacks an effective objective prediction model and cannot achieve early warning of the occurrence of PSCI in the acute phase, medium and long term, which limits the formulation of early intervention and personalized treatment strategies. Summary of the Invention
[0003] The purpose of the present application is to provide a method, device, equipment and medium for constructing a prediction model for post-stroke cognitive impairment, which can improve the accuracy of predicting post-stroke cognitive impairment.
[0004] To achieve the above purpose, the present application provides the following solutions:
[0005] In the first aspect, the present application provides a method for constructing a prediction model for post-stroke cognitive impairment, including:
[0006] Collecting a historical data set of a target population, where each sample in the historical data set includes multiple parameters of a stroke patient;
[0007] Using statistical methods to conduct importance tests on the parameters in the historical data set, and screening out the parameters that meet the set conditions; the parameters that meet the set conditions and the label data form a training set, and the label data is the occurrence probability of acute post-stroke cognitive impairment or the occurrence probability of early-onset post-stroke cognitive impairment; the acute post-stroke cognitive impairment is cognitive impairment occurring within 2 weeks after stroke, and the early-onset post-stroke cognitive impairment is cognitive impairment occurring within 3 to 6 months after stroke;
[0008] Dividing the training set into a first training subset and a second training subset according to the label data; the first training subset is composed of samples with the label data being the occurrence probability of acute post-stroke cognitive impairment, and the second training subset is composed of samples with the label data being the occurrence probability of early-onset post-stroke cognitive impairment;
[0009] Construct an acute post-stroke cognitive impairment prediction model using machine learning methods based on the first training subset, and construct an early-onset post-stroke cognitive impairment prediction model using machine learning methods based on the second training subset.
[0010] Optionally, perform an importance test on each parameter in the historical dataset using statistical methods, and screen out the parameters that meet the set conditions, specifically including:
[0011] Perform a t-test, Mann-Whitney U test, chi-square test, or Fisher test on each parameter in the historical dataset using statistical methods to obtain the P-value of each parameter, and use the parameters with a P-value less than the set value as the parameters after P-value screening;
[0012] Use the random forest method to screen out the top N parameters with the highest importance from the parameters after P-value screening in the first training subset to obtain the first set of parameters;
[0013] Use the random forest method to screen out the top N parameters with the highest importance from the parameters after P-value screening in the second training subset to obtain the second set of parameters;
[0014] Take the union of the first set of parameters and the second set of parameters to obtain the parameters that meet the set conditions.
[0015] Optionally, the set value is 0.05 and the N value is 10.
[0016] Optionally, the parameters include age, years of education, volume of white matter hyperintensities, presence or absence of diabetes history, presence or absence of stroke history, presence or absence of cortical infarction, presence or absence of lacunar lesions, and presence or absence of infarcts in key regions; the key regions include the basal ganglia, thalamus, hippocampus, medial inferior temporal gyrus, and angular gyrus.
[0017] Optionally, construct an acute post-stroke cognitive impairment prediction model using machine learning methods based on the first training subset, specifically including:
[0018] Construct multiple first prediction models using different machine learning methods based on the first training subset;
[0019] Use the first test set to test each of the first prediction models to obtain the AUC value corresponding to each of the first prediction models;
[0020] Take the first prediction model with the largest AUC value as the acute post-stroke cognitive impairment prediction model.
[0021] Optionally, construct an early-onset post-stroke cognitive impairment prediction model using machine learning methods based on the second training subset, specifically including:
[0022] Construct multiple second prediction models using different machine learning methods based on the second training subset;
[0023] Use a second test set to test each of the second prediction models to obtain the AUC value corresponding to each of the second prediction models;
[0024] Take the second prediction model with the largest AUC value as the prediction model for post - stroke cognitive impairment in the early - onset type.
[0025] Optionally, the machine learning method includes gradient boosting algorithm, K - nearest neighbor algorithm, logistic regression algorithm, random forest algorithm, artificial neural network, support vector machine, and XGBoost.
[0026] In a second aspect, the present application provides a device for constructing a post - stroke cognitive impairment prediction model. The device for constructing a post - stroke cognitive impairment prediction model applies the method for constructing a post - stroke cognitive impairment prediction model described above. The device for constructing a post - stroke cognitive impairment prediction model includes:
[0027] A historical data set acquisition module, configured to acquire a historical data set of a target population, where each sample in the historical data set includes multiple parameters of a stroke patient;
[0028] A training set determination module, configured to perform an importance test on each parameter in the historical data set by using a statistical method, and screen out parameters that meet the set conditions; the parameters that meet the set conditions and the label data form a training set, and the label data is the probability of occurrence of post - acute - stroke cognitive impairment or the probability of occurrence of post - early - onset - stroke cognitive impairment; the post - acute - stroke cognitive impairment is cognitive impairment occurring within 2 weeks after stroke, and the post - early - onset - stroke cognitive impairment is cognitive impairment occurring within 3 to 6 months after stroke;
[0029] A training subset determination module, configured to divide the training set into a first training subset and a second training subset according to the label data; the first training subset is composed of samples with the label data being the probability of occurrence of post - acute - stroke cognitive impairment, and the second training subset is composed of samples with the label data being the probability of occurrence of post - early - onset - stroke cognitive impairment;
[0030] A prediction model construction module, configured to construct a prediction model for post - acute - stroke cognitive impairment by using a machine learning method according to the first training subset, and construct a prediction model for post - early - onset - stroke cognitive impairment by using a machine learning method according to the second training subset.
[0031] In a third aspect, the present application provides a computer device, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor executes the computer program to implement the steps of the method for constructing a post - stroke cognitive impairment prediction model described in any one of the above.
[0032] Fourthly, the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method for constructing a post-stroke cognitive impairment prediction model described in any one of the above are implemented.
[0033] According to the specific embodiments provided by the present application, the following technical effects are disclosed:
[0034] The present application provides a method, device, equipment and medium for constructing a post-stroke cognitive impairment prediction model. By performing importance screening, parameters meeting the set conditions are obtained, and a training set is composed of the parameters meeting the set conditions and label data. The label data includes the occurrence probability of acute PSCI and the occurrence probability of early-onset PSCI. It does not rely on the data generated by the patient's active demonstration, reduces the prediction error caused by the uncertainty of the patient's demonstration behavior, and improves the PSCI prediction accuracy. In addition, the present application creates prediction models for acute and early-onset PSCI respectively, further improving the PSCI prediction accuracy. BRIEF DESCRIPTION OF THE DRAWINGS
[0035] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.
[0036] Figure 1 It is a schematic flowchart of a method for constructing a post-stroke cognitive impairment prediction model provided by an embodiment of the present application;
[0037] Figure 2 It is a schematic diagram for comparing the performance of an acute post-stroke cognitive impairment prediction model provided by an embodiment of the present application;
[0038] Figure 3 It is a schematic diagram for comparing the performance of an early-onset post-stroke cognitive impairment prediction model provided by an embodiment of the present application;
[0039] Figure 4 It is a schematic diagram for assigning importance to model variables constructed based on the random forest algorithm provided by an embodiment of the present application;
[0040] Figure 5 It is a schematic diagram of the structure of a computer device provided by an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS
[0041] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.
[0042] To make the above objects, features, and advantages of the present application more obvious and understandable, the present application will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.
[0043] The present application provides a method for constructing a prediction model for post-stroke cognitive impairment, as Figure 1 shown, the method for constructing a prediction model for post-stroke cognitive impairment includes steps 101 to 104.
[0044] Step 101: Collect the historical data set of the target population, and each sample in the historical data set includes multiple parameters of a stroke patient.
[0045] Step 102: Use statistical methods to conduct importance tests on the parameters in the historical data set, and screen out the parameters that meet the set conditions; the parameters that meet the set conditions and the label data form a training set, and the label data is the probability of occurrence of post-acute stroke cognitive impairment or the probability of occurrence of early-onset post-stroke cognitive impairment; the post-acute stroke cognitive impairment is cognitive impairment occurring within 2 weeks after stroke, and the early-onset post-stroke cognitive impairment is cognitive impairment occurring within 3 to 6 months after stroke.
[0046] Step 103: Divide the training set into a first training subset and a second training subset according to the label data; the first training subset is composed of samples with the label data being the probability of occurrence of post-acute stroke cognitive impairment, and the second training subset is composed of samples with the label data being the probability of occurrence of early-onset post-stroke cognitive impairment.
[0047] Step 104: Construct a prediction model for post-acute stroke cognitive impairment using machine learning methods according to the first training subset, and construct a prediction model for early-onset post-stroke cognitive impairment using machine learning methods according to the second training subset.
[0048] In an exemplary embodiment, step 101 specifically includes: The target population is 243 stroke patients statistically from December 2022 to April 2024. Cognitive assessments and structural image acquisitions are completed for the 243 stroke patients 14 days after admission and 3 to 6 months after onset to obtain the historical data set.
[0049] Definition of acute PSCI: Within 2 weeks after stroke, the score of the Mini-Mental State Examination (MMSE) or the Montreal Cognitive Assessment (MoCA) is lower than the cut-off value of the elderly population within the set range.
[0050] Definition of early-onset PSCI: 3 - 6 months after stroke, the score of MMSE or MoCA is lower than the cut-off value of the elderly population within the set range.
[0051] In an exemplary embodiment, step 102 specifically includes:
[0052] Using statistical methods to perform t-tests, Mann-Whitney U-tests, chi-square tests, or Fisher tests on each parameter in the historical dataset to obtain the P-value of each parameter. The parameters with P-values less than the set value are used as the parameters after P-value screening. More specifically, python is used to perform statistical processing on the data. For clinical variable normal distribution measurement data, it is expressed as mean ± standard deviation, and t-tests are used for between-group comparisons; for non-normal distribution measurement data, it is expressed as M(Q1, Q3), and Mann-Whitney U-tests are used for between-group comparisons; for count data, it is expressed as frequency and relative number composition (%), and chi-square tests or Fisher tests are used for between-group comparisons. M(Q1, Q3) represents the combination of the median and the interquartile range.
[0053] Using the random forest method to screen out the top N parameters with the highest importance from the parameters after P-value screening in the first training subset to obtain the first set of parameters.
[0054] Using the random forest method to screen out the top N parameters with the highest importance from the parameters after P-value screening in the second training subset to obtain the second set of parameters.
[0055] Take the union of the first set of parameters and the second set of parameters to obtain the parameters that meet the set conditions. The random forest method is used to assign importance to the parameters as Figure 4 shown.
[0056] In an exemplary embodiment, the set value is 0.05 and the N value is 10. Variables with P < 0.05 in the univariate analysis are included in the random forest variable importance ranking.
[0057] Among the 243 study subjects, 189 were male and 54 were female, and the age at enrollment was 61 (53, 70) years. There were 112 patients with PSCI in the acute phase. The age, volume of white matter hyperintensities, diabetes, hyperlipidemia, incidence of stroke history, lacunes, burden of cerebral small vessel disease, incidence of infarcts in key areas and cortex in the PSCI group were significantly higher than those in the group without post-stroke cognitive impairment (PSNCI), and the years of education were lower than those in the PSNCI group (all P<0.05); at 3-6 months after stroke, the age, volume of white matter hyperintensities, incidence of hyperlipidemia, imaging burden of cerebral small vessel disease, and incidence of cortical infarcts in the PSCI group were significantly higher than those in the PSNCI group (all P<0.05).
[0058] The data in the historical dataset include imaging data, and the processing methods of the imaging data include:
[0059] (1) Calculation of the volume of white matter hyperintensities: Based on the T2-FLAIR sequence, the MATLAB-SPM12-LST toolkit was used to segment the baseline volume of white matter hyperintensities to obtain the volume of white matter hyperintensities.
[0060] (2) Calculation of the infarct volume: Based on the Diffusion-Weighted Imaging (DWI) and Apparent Diffusion Coefficient (ADC) sequences, the MATLAB-SPM12-LST toolkit was used for baseline segmentation to obtain the infarct volume.
[0061] (3) Infarcts in key areas: Based on the interpretation of the DWI and ADC sequences, determine whether there are infarcts in the basal ganglia, thalamus, hippocampus, medial inferior temporal gyrus and angular gyrus.
[0062] In an exemplary embodiment, the parameters include age, years of education, volume of white matter hyperintensities, presence or absence of diabetes history, presence or absence of stroke history, presence or absence of cortical infarcts, presence or absence of lacunar lesions, and presence or absence of infarcts in key areas; the key areas include the basal ganglia, thalamus, hippocampus, medial inferior temporal gyrus and angular gyrus.
[0063] In an exemplary embodiment, a prediction model for post-stroke cognitive impairment is constructed using a machine learning method according to the first training subset, specifically including:
[0064] Multiple first prediction models are constructed using different machine learning methods according to the first training subset.
[0065] Each of the first prediction models is tested using the first test set to obtain the area under the curve (AUC) value corresponding to each of the first prediction models. The data structure of the first test set is the same as that of the first training subset.
[0066] The first prediction model with the largest AUC value is used as the prediction model for cognitive impairment after acute stroke.
[0067] In an exemplary embodiment, a prediction model for cognitive impairment after early-onset stroke is constructed by using a machine learning method according to a second training subset, specifically including:
[0068] Construct a plurality of second prediction models by using different machine learning methods according to the second training subset.
[0069] Test each of the second prediction models by using a second test set to obtain the AUC value corresponding to each of the second prediction models. The data structure of the second test set is the same as that of the second training subset.
[0070] The second prediction model with the largest AUC value is used as the prediction model for cognitive impairment after early-onset stroke.
[0071] The parameters that meet the set conditions screened out in the steps are used as target parameters.
[0072] The prediction model for cognitive impairment after acute stroke is used to input the target parameters of the patient, especially the target parameters at the time of admission, and output the occurrence probability of acute PSCI.
[0073] The prediction model for cognitive impairment after early-onset stroke is used to input the target parameters of the patient, especially the target parameters at the time of admission, and output the occurrence probability of early-onset PSCI.
[0074] The machine learning methods include Gradient Boost, K_neighbors, Logistic regression, Artificial Neural Network (ANN), Random_forest, Support_Vector_Machines, and XGBoost.
[0075] Among them, the AUC of the prediction models for cognitive impairment after acute stroke constructed by Gradient Boost, K_neighbors, Logistic regression, Random_forest, Support_Vector_Machines, and XGBoost all exceed 0.75, as Figure 2 shown. The AUC of the prediction models for cognitive impairment after early-onset stroke constructed by Logistic regression, Support_Vector_Machines, and Gradient Boost all exceed 0.75, as Figure 3 shown.
[0076] Based on the same inventive concept, an embodiment of the present application further provides a device for constructing a post-stroke cognitive impairment prediction model for implementing the method for constructing a post-stroke cognitive impairment prediction model involved above. The solution provided by this device for solving problems is similar to the solution described in the above method. Therefore, the specific limitations in one or more embodiments of the device for constructing a post-stroke cognitive impairment prediction model provided below can refer to the limitations on the method for constructing a post-stroke cognitive impairment prediction model in the above text, and will not be repeated here.
[0077] In an exemplary embodiment, a device for constructing a post-stroke cognitive impairment prediction model is provided. The device for constructing a post-stroke cognitive impairment prediction model applies the method for constructing a post-stroke cognitive impairment prediction model described above. The device for constructing a post-stroke cognitive impairment prediction model includes:
[0078] A historical data set acquisition module, configured to acquire a historical data set of a target population, and each sample in the historical data set includes multiple parameters of a stroke patient.
[0079] A training set determination module, configured to perform an importance test on each parameter in the historical data set by using a statistical method, and screen out parameters that meet the set conditions; the parameters that meet the set conditions and the label data form a training set, and the label data is the occurrence probability of post-stroke cognitive impairment within 2 weeks after stroke or the occurrence probability of early-onset post-stroke cognitive impairment; the post-stroke cognitive impairment within 2 weeks after stroke is cognitive impairment occurring within 2 weeks after stroke, and the early-onset post-stroke cognitive impairment is cognitive impairment occurring within 3 to 6 months after stroke.
[0080] A training subset determination module, configured to divide the training set into a first training subset and a second training subset according to the label data; the first training subset is composed of samples with the label data being the occurrence probability of post-stroke cognitive impairment within 2 weeks after stroke, and the second training subset is composed of samples with the label data being the occurrence probability of early-onset post-stroke cognitive impairment.
[0081] A prediction model construction module, configured to construct a post-stroke cognitive impairment prediction model within 2 weeks after stroke by using a machine learning method according to the first training subset, and construct an early-onset post-stroke cognitive impairment prediction model by using a machine learning method according to the second training subset.
[0082] In an exemplary embodiment, a computer device is provided. The computer device can be a server or a terminal, and its internal structure diagram can be as Figure 5As shown in the figure. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O), and a communication interface. Among them, 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. Among them, 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 data for constructing a prediction model for post-stroke cognitive impairment. The input / output interface of the computer device is used to exchange information between the processor and external devices. 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, it implements a method for constructing a prediction model for post-stroke cognitive impairment.
[0083] Those skilled in the art can understand that Figure 5 The structure shown in the figure is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have different component arrangements. In an exemplary embodiment, a computer device is provided, including a memory and a processor. A computer program is stored in the memory, and when the processor executes the computer program, the steps in the above method embodiments are implemented.
[0084] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by the processor, the steps in the above method embodiments are implemented.
[0085] In an exemplary embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by the processor, the steps in the above method embodiments are implemented.
[0086] 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 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 need to comply with relevant regulations.
[0087] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. 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 methods. Among them, any reference to a memory, database, or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memories. Non-volatile memories can include read-only memory (ROM), magnetic tapes, floppy disks, flash memories, optical memories, high-density embedded non-volatile memories, resistive random access memories (ReRAM), magnetoresistive random access memories (MRAM), ferroelectric random access memories (FRAM), phase change memories (PCM), graphene memories, etc. Volatile memories can include random access memories (RAM) or external cache memories, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.
[0088] The databases involved in the embodiments provided in the present application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., without limitation. The processors involved in the embodiments provided in the present application can be general-purpose processors, central processors, graphics processors, digital signal processors, data processing logics of programmable logics, etc., without limitation.
[0089] The technical features of the above embodiments can be combined arbitrarily. For the sake of concise description, 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, it should be considered as the scope described in this specification.
[0090] Specific examples are used in this article to elaborate on the principles and implementation manners of the present application. The descriptions of the above embodiments are only used to help understand the methods and core ideas of the present application; at the same time, for those of ordinary skill in the art, according to the ideas of the present application, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation to the present application.
Claims
1. A method for constructing a prediction model for post-stroke cognitive impairment, characterized in that: The method for constructing a prediction model for post-stroke cognitive impairment comprises: Collecting a historical data set of a target population, wherein each sample in the historical data set includes multiple parameters of a stroke patient; A statistical method is used to perform an importance test on each parameter in the historical data set to screen out parameters that meet the set conditions; the parameters that meet the set conditions and label data constitute a training set, and the label data is the probability of occurrence of acute post-stroke cognitive impairment or the probability of occurrence of early-onset post-stroke cognitive impairment; the acute post-stroke cognitive impairment is cognitive impairment occurring within 2 weeks after the stroke, and the early-onset post-stroke cognitive impairment is cognitive impairment occurring within 3 to 6 months after the stroke; Dividing the training set into a first training subset and a second training subset according to the label data; the first training subset is composed of samples whose label data is the probability of occurrence of cognitive impairment after acute stroke, and the second training subset is composed of samples whose label data is the probability of occurrence of cognitive impairment after early-onset stroke; A prediction model for cognitive impairment after acute stroke was constructed using machine learning methods based on the first training subset, and a prediction model for cognitive impairment after early-onset stroke was constructed using machine learning methods based on the second training subset.
2. The method for constructing a prediction model for post-stroke cognitive impairment according to claim 1, characterized in that: Statistical methods are used to test the importance of each parameter in the historical data set to screen out parameters that meet the set conditions, including: Using statistical methods to perform t-test, Mann-Whitney U test, chi-square test or Fisher test on each parameter in the historical data set to obtain the P value of each parameter, and taking the parameter with a P value less than the set value as the parameter after P value screening; The random forest method is used to select the top N parameters in importance from the parameters after the P value screening of the first historical data subset to obtain a first group of parameters; the first historical data subset is composed of samples in the historical data set whose label data is the probability of occurrence of cognitive impairment after acute stroke; The random forest method is used to select the top N parameters in importance from the parameters after the P value screening of the second historical data subset to obtain a second group of parameters; the second historical data subset is composed of samples in the historical data set whose label data is the probability of occurrence of cognitive impairment after early-onset stroke; The union of the first set of parameters and the second set of parameters is taken to obtain parameters that meet the set conditions.
3. The method for constructing a prediction model for post-stroke cognitive impairment according to claim 2, characterized in that: The set value is 0.05 and the N value is 10.
4. The method for constructing a prediction model for post-stroke cognitive impairment according to claim 1, characterized in that: The parameters include age, years of education, volume of white matter high signals, presence or absence of a history of diabetes, presence or absence of a history of stroke, presence or absence of cortical infarction, presence or absence of lacunar foci, and presence or absence of infarction in key areas; key areas include the basal ganglia, thalamus, hippocampus, medial inferior temporal gyrus, and angular gyrus.
5. The method for constructing a prediction model for post-stroke cognitive impairment according to claim 1, characterized in that: Based on the first training subset, a prediction model for cognitive impairment after acute stroke was constructed using machine learning methods, including: Constructing a plurality of first prediction models using different machine learning methods according to the first training subset; Using a first test set to test each of the first prediction models, to obtain an AUC value corresponding to each of the first prediction models; The first prediction model with the largest AUC value was taken as the prediction model for cognitive impairment after acute stroke.
6. The method for constructing a prediction model for post-stroke cognitive impairment according to claim 1, characterized in that: A prediction model for early-onset post-stroke cognitive impairment was constructed using machine learning methods based on the second training subset, including: Building a plurality of second prediction models using different machine learning methods based on the second training subset; Using a second test set to test each of the second prediction models, to obtain an AUC value corresponding to each of the second prediction models; The second prediction model with the largest AUC value was used as the prediction model for early-onset post-stroke cognitive impairment.
7. The method for constructing a prediction model for post-stroke cognitive impairment according to claim 5 or 6, characterized in that: The machine learning methods include gradient boosting algorithm, K nearest neighbor algorithm, logistic regression algorithm, random forest algorithm, artificial neural network, support vector machine and XGBoost.
8. A device for constructing a prediction model for cognitive impairment after stroke, characterized in that: The device for constructing a prediction model for cognitive impairment after stroke applies the method for constructing a prediction model for cognitive impairment after stroke according to any one of claims 1 to 7, and the device for constructing a prediction model for cognitive impairment after stroke comprises: A historical data set acquisition module, used to acquire historical data sets of a target population, wherein each sample in the historical data set includes multiple parameters of a stroke patient; A training set determination module is used to use statistical methods to perform importance tests on each parameter in the historical data set, and screen out parameters that meet the set conditions; the parameters that meet the set conditions and label data constitute the training set, and the label data is the probability of occurrence of acute post-stroke cognitive impairment or the probability of occurrence of early-onset post-stroke cognitive impairment; the acute post-stroke cognitive impairment is cognitive impairment occurring within 2 weeks after the stroke, and the early-onset post-stroke cognitive impairment is cognitive impairment occurring within 3 to 6 months after the stroke; A training subset determination module, configured to divide the training set into a first training subset and a second training subset according to the label data; the first training subset is composed of samples whose label data is the probability of occurrence of cognitive impairment after acute stroke, and the second training subset is composed of samples whose label data is the probability of occurrence of cognitive impairment after early-onset stroke; The prediction model construction module is used to construct a prediction model for acute post-stroke cognitive impairment using a machine learning method based on the first training subset, and to construct a prediction model for early-onset post-stroke cognitive impairment using a machine learning method based on the second training subset.
9. 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 method for constructing a post-stroke cognitive impairment prediction model according to any one of claims 1 to 7.
10. 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 constructing a prediction model for post-stroke cognitive impairment according to any one of claims 1 to 7 is implemented.