Training method and device of mild cognitive impairment screening model, storage medium and electronic equipment

Abstract

The invention provides a training method and device for a mild cognitive impairment screening model, a storage medium and electronic equipment, and relates to the technical field of model training, and the method comprises the steps: collecting research sample data based on community natural population health data; acquiring characteristic variables related to mild cognitive impairment on the basis of the research sample data, and taking the characteristic variables as training data; and taking the optimal AUC as a training direction of model training, and training an extreme gradient lifting algorithm model based on the training data to obtain a mild cognitive impairment screening model. The mild cognitive impairment screening model obtained through training of the method is higher in screening precision and better in screening effect.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of model training, in particular to a mild cognitive impairment screening model training method and device, a storage medium and an electronic device. BACKGROUND

[0002] Mild cognitive impairment (MCI) is an early stage of dementia. How to early detect MCI patients in the apparent healthy community is the top priority of early prevention and control of dementia. Although large-scale population screening for mild cognitive impairment cannot be carried out due to various restrictions, the development and application of new screening tools to achieve rapid, accurate and intelligent screening of mild cognitive impairment in community residents has become the common goal of global scholars.

[0003] With the rapid development of artificial intelligence, using artificial intelligence models to screen mild cognitive impairment is one of the main development directions.

[0004] The existing mild cognitive impairment screening model has the following problems in the training process: The existing model does not consider the optimal combination of model training parameters in the training process, resulting in a model that is not good enough. SUMMARY

[0005] Embodiments of the present application provide a mild cognitive impairment screening model training method, device, storage medium and electronic device to solve the technical problems existing in the prior art.

[0006] Other characteristics and advantages of the present application will become apparent from the following detailed description, or will be learned by practice of the present application.

[0007] According to a first aspect of the embodiments of the present application, a mild cognitive impairment screening model training method is provided, characterized in that it comprises: Collecting research sample data based on community natural population health data; Obtaining feature variables related to mild cognitive impairment based on the research sample data as training data; Taking the best AUC as the training direction of model training, training an extreme gradient boosting algorithm model based on the training data to obtain a mild cognitive impairment screening model.

[0008] In some embodiments of the present application, based on the foregoing scheme, the best AUC is taken as the training direction of model training, and an extreme gradient boosting algorithm model is trained based on the training data to obtain a mild cognitive impairment screening model, comprising: The best AUC is taken as the training direction of model training, and the optimal parameter combination of the model in the hyperparameter space is calculated under the premise of expert opinion and 5-fold cross-validation; Under the optimal parameter combination, an extreme gradient boosting algorithm model is trained according to the training data to obtain a mild cognitive impairment screening model.

[0009] In some embodiments of the present application, based on the foregoing scheme, the optimal parameter combination of the model in the hyperparameter space under the premise of expert opinion and 5-fold cross-validation includes: Random seed, class balance, learning rate and maximum depth are taken as main target optimization parameters of the mild cognitive impairment screening model, and the search range of the hyperparameter space of each main target optimization parameter is determined; The optimal parameter configuration of the mild cognitive impairment screening model is searched and trained in the search range of the hyperparameter space using a grid search strategy, the optimal parameter combination is obtained, and the sensitivity, specificity, accuracy, AUC value and ROC curve are recorded.

[0010] In some embodiments of the present application, based on the foregoing scheme, the research sample data based on community natural population health data collection includes: Population samples are recruited in the natural population community, and health evaluation is performed on all population samples taking a population research strategy as the breakthrough point. The health evaluation data are collected as the research sample data.

[0011] In some embodiments of the present application, based on the foregoing scheme, the feature variables related to mild cognitive impairment are obtained based on the research sample data as training data, including: The feature variables related to mild cognitive impairment are screened from the evaluation data based on a literature review method, a professional field knowledge injection method, a field expert consultation method and a Shap interpretable machine learning method, and are taken as training data.

[0012] According to a second aspect of the embodiments of the present application, a mild cognitive impairment screening model training device is provided, including: The collection unit is configured to collect research sample data based on community natural population health data; The screening unit is configured to obtain feature variables related to mild cognitive impairment based on the research sample data as training data; The training unit is configured to take the best AUC as the training direction of model training, and train an extreme gradient boosting algorithm model based on the training data to obtain a mild cognitive impairment screening model.

[0013] According to a third aspect of the embodiments of the present application, a computer readable storage medium is provided, wherein the storage medium stores computer instructions, and the computer instructions, when executed on a computer, cause the computer to perform the method according to the first aspect.

[0014] According to a fourth aspect of the embodiments of the present application, an electronic device is provided, comprising: a memory and a processor; The memory is configured to store computer instructions. The processor is configured to invoke the computer instructions stored in the memory, so that the electronic device performs the method according to the first aspect.

[0015] The technical solution of the present application trains the model with the best AUC as the training target in the training process, so that the performance of the trained model is better, and the screening precision is higher.

[0016] It should be understood that the foregoing general description and the following detailed description are only exemplary and explanatory, and cannot limit the present application. BRIEF DESCRIPTION OF DRAWINGS

[0017] The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor. In the drawings: Figure 1 Fig. 1 shows a flowchart of a mild cognitive impairment screening model training method according to an embodiment of the present application; Figure 2 Fig. 2 shows a block diagram of a mild cognitive impairment screening model training device according to an embodiment of the present application; Figure 3 Fig. 3 shows a block diagram of an electronic device according to an embodiment of the present application; Figure 4 Fig. 4 shows a structural diagram of a computer system of an electronic device suitable for implementing the embodiments of the present application. DETAILED DESCRIPTION

[0018] Example implementations will now be described more fully with reference to the accompanying drawings. However, example implementations can be implemented in various forms and should not be construed as limited to the examples set forth herein; rather, these implementations are provided so that this disclosure will be more thorough and complete, and will fully convey the concept of example implementations to those skilled in the art.

[0019] Moreover, the described features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. In the following description, numerous specific details are provided to give a thorough understanding of embodiments of the application. One skilled in the relevant art will recognize, however, that the application can be practiced without one or more of the specific details, or with other methods, components, materials, and so forth. In other instances, well-known structures, devices, implementations, and operations have not been shown or described in detail to avoid obscuring aspects of the application.

[0020] The block diagrams in the drawings show only the functionality of the features and can not imply that the functions must be implemented in the particular order presented or by independent circuits or integrated circuits. These are functional block diagrams, and implementation of the described functionality can be performed with software, with hardware such as circuitry, with a combination of hardware and software, or with one or more hardware components or integrated circuits programmed with software to perform the functions. Embodiments are described herein with reference to block diagrams that show a functional overview of the basic components of embodiments of the present application.

[0021] The flow diagrams in the drawings show the functionality of the embodiments and do not imply that the functions must be performed in the particular order presented or by independent circuits or integrated circuits. These are functional flow diagrams, and implementations can differ depending on the implementation of the described features. For example, not all of the operations shown in the flow diagrams can be required, and the order of the operations can vary.

[0022] It should be noted that "a plurality" refers to two or more.

[0023] In order to make the objectives, technical solutions and advantages of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some 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 work fall within the scope of protection of the present application.

[0024] Some embodiments of the present application will be described in detail below with reference to the drawings. The following embodiments and features in the embodiments can be combined with each other without conflict.

[0025] Referring to Figure 1 , a flowchart of a training method of a mild cognitive impairment screening model according to an embodiment of the present application is shown.

[0026] As shown in Figure 1 , a training method of a mild cognitive impairment screening model is shown, which specifically includes steps S100 to S300.

[0027] Referring to Figure 1 , in step S100, sample data is collected based on community natural person health data.

[0028] In some possible embodiments, based on the foregoing scheme, the step S100 comprises: Recruit population samples in a natural population community, take the population research strategy as the breakthrough point, and conduct health evaluation on all population samples to collect health evaluation data as the research sample data.

[0029] It should be noted that in this embodiment, the way of collecting data from a real-world natural population community has more practical application value than the traditional way of obtaining data from hospital patients.

[0030] Continuing to refer to Figure 1 , step S200, obtaining feature variables related to mild cognitive impairment based on the research sample data as training data.

[0031] In some possible embodiments, based on the foregoing scheme, the step S200 comprises: Filtering feature variables related to mild cognitive impairment from the evaluation data based on literature review method, professional field knowledge injection method, field expert consultation method and Shap interpretable machine learning method as training data.

[0032] Exemplarily, the specific process of this step is as follows: First, based on the research literature evidence of "mild cognitive impairment" in the past, systematic literature retrieval, summary and carding work are carried out, and information related to "mild cognitive impairment" such as independent influencing factors, evaluation scales and biomarkers is collected to form a review list of mild cognitive impairment influencing factors; Using the review list of mild cognitive impairment influencing factors to filter feature variables closely related to cognitive impairment from the evaluation data.

[0033] Then, a feature screening model is developed around the review list of mild cognitive impairment influencing factors, and the performance of the feature screening model is verified.

[0034] Using the verified feature screening model to screen feature variables, eliminating irrelevant noise features, to obtain a first feature variable set.

[0035] Next, the data categories in the first feature variable set are uniformly classified, that is, the features in the first feature variable set are divided into multiple feature subsets, and each feature subset contains multiple features of different categories.

[0036] The multiple feature subsets are sequentially combined to obtain a feature variable set.

[0037] For example, four feature subsets A, B, C, and D are obtained after classification, and the four feature subsets A, B, C, and D are sequentially combined, and finally expanded into 15 data feature subsets composed of single mode and multi-mode, that is, A, B, C, D, AB, AC, AD, BC, BD, CD, ABC, ABD, ACD, BCD, and ABCD feature subsets, which form the feature variable set.

[0038] Then, the importance of the features in the feature variable set is calculated and ranked according to whether it is a mild cognitive impairment.

[0039] For example, using the SHAP interpretable machine learning technology, the contribution of the features in the A, B, C, D, AB, AC, AD, BC, BD, CD, ABC, ABD, ACD, BCD, and ABCD feature subsets to the label prediction is ranked according to the feature importance, and 15 sets of feature importance ranking feature distribution results are obtained.

[0040] Then, combined with expert experience, the top 10 features in each feature subset are retained as the feature pool for model training, which prepares for subsequent model training with different feature combinations. Of course, features ranked in the top 5 or top 15 can also be obtained, and the specific number is determined based on actual conditions.

[0041] Taking the top 10 features extracted from the A feature subset as an example, these features are systematically combined, i.e., "top 2 feature combination", "top 3 feature combination", "top 4 feature combination", "top 5 feature combination", "top 6 feature combination", "top 7 feature combination", "top 8 feature combination", "top 9 feature combination", and "top 10 feature combination", so that 9 combinations of features are obtained.

[0042] It should be noted that the "top x feature combination" described in the example refers to the combination of the top x features, and x is a positive integer related to the feature data with retained feature importance.

[0043] For example, "top 2 feature combination" refers to the combination of the first and second ranked features.

[0044] By analogy, after combining all feature subsets, 135 feature combinations are obtained, i.e., 15 groups * 9 groups = 135.

[0045] Then, the 135 feature combinations are used as model training data to train the model.

[0046] Continuing to refer to Figure 1, step S300, taking the optimal AUC as the training direction of model training, training the extreme gradient boosting algorithm model based on the training data to obtain the mild cognitive impairment screening model.

[0047] In some feasible embodiments, based on the foregoing scheme, step S300 comprises: Step S310, taking the optimal AUC as the training direction of model training, calculating the optimal parameter combination of the model in the hyperparameter space under the premise of expert opinion and 5-fold cross-validation. Step S320, under the optimal parameter combination, training the extreme gradient boosting algorithm model according to the training data to obtain the mild cognitive impairment screening model.

[0048] In some feasible embodiments, based on the foregoing scheme, the optimal parameter combination of the model in the hyperparameter space under the premise of expert opinion and 5-fold cross-validation comprises: Taking random seed, class balance, learning rate and maximum depth as the main target optimization parameters of the mild cognitive impairment screening model, and determining the hyperparameter space search range of each main target optimization parameter; Using a grid search strategy, searching and training the optimal parameter configuration of the mild cognitive impairment screening model in the hyperparameter space search range to obtain the optimal parameter combination, and recording the sensitivity, specificity, accuracy, AUC value and ROC curve.

[0049] It should be noted that in this embodiment, AUC refers to the area under the ROC curve, and ROC refers to the Receiver Operating Characteristic Curve, which stands for Receiver Operating Characteristic Curve.

[0050] It can be understood that in this step, the model is trained with AUC as the target, which can truly reflect the performance of the model under different data distributions, and the model is trained under the premise of the optimal parameter combination, further optimizing the performance of the model.

[0051] The following describes an apparatus embodiment of the present application, which can be used to execute the training method of the mild cognitive impairment screening model in one of the embodiments of the present application. For details not disclosed in the apparatus embodiment of the present application, please refer to the embodiments of the method of the present application.

[0052] Referring to Figure 2 Fig. 1, a training device 200 of a mild cognitive impairment screening model according to one embodiment of the present application comprises: The acquisition unit 201 is configured to acquire research sample data based on community natural population health data. The screening unit 202 is configured to acquire a feature variable related to mild cognitive impairment based on the research sample data as training data. The training unit 203 is configured to take the optimal AUC as a training direction of model training, train an extreme gradient boosting algorithm model based on the training data, and obtain a mild cognitive impairment screening model.

[0053] As shown in Figure 3 The electronic device 300 is further configured to implement the steps of the method for training a mild cognitive impairment screening model according to the first aspect.

[0054] Since the electronic device is the device used by the training device for the mild cognitive impairment screening model according to the embodiments of the present application, the specific implementation of the electronic device and various changes thereof can be understood by those skilled in the art based on the method according to the embodiments of the present application. Therefore, how the electronic device implements the method according to the embodiments of the present application will not be described in detail, and any device used by those skilled in the art to implement the method according to the embodiments of the present application is within the scope of the present application.

[0055] In the implementation process, the computer program 311 can implement any embodiment of the first aspect when executed by the processor.

[0056] Figure 4 A structural schematic diagram of a computer system of the electronic device suitable for implementing the embodiments of the present application is shown.

[0057] It should be noted that Figure 4 The computer system 400 of the electronic device shown is only an example and should not limit the functions and use range of the embodiments of the present application.

[0058] As shown in Figure 4 The computer system 400 includes a central processing unit 401, which can perform various appropriate actions and processes according to programs stored in a read-only memory 402 or programs loaded from a storage portion 408 to a random access memory 403, such as the method described in the above embodiments. Various programs and data required for system operation are also stored in the random access memory 403. The central processing unit 401, the read-only memory 402, and the random access memory 403 are connected to each other through a bus 404. An input / output interface 405 is also connected to the bus 404.

[0059] The following components are connected to the input / output interface 405: an input portion 406 including input devices such as a keyboard and a mouse; an output portion 407 including output devices such as a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), and a speaker; a storage portion 408 including a hard disk; and a communication portion 409 including a network interface card such as a LAN (Local Area Network) card, a modem, and the like. The communication portion 409 performs communication processing via a network such as the Internet. A drive 410 is also connected to the input / output interface 405 as necessary. A removable media 411 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, and the like is attached to the drive 410 as necessary, so that a computer program read therefrom is installed in the storage portion 408 as necessary.

[0060] In particular, according to embodiments of the present application, the processes described above with reference to the flowcharts can be implemented as a computer software program. For example, embodiments of the present application include a computer program product comprising a computer program carried on a computer readable medium, the computer program containing program code for executing the methods illustrated by the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via the communication portion 409, and / or installed from the removable media 411. When the computer program is executed by the central processing unit 401, various functions defined in the system of the present application are executed.

[0061] It should be noted that the computer-readable medium in the embodiments of the present application can be a computer-readable signal medium or a computer-readable storage medium or any combination thereof. The computer-readable storage medium may, for example, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or apparatus, or any combination thereof. More specific examples of the computer-readable storage medium can include, but are not limited to, an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (Compact Disc Read-Only Memory, CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present application, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device or apparatus. In the present application, the computer-readable signal medium can include a data signal carried in a baseband or as a part of a carrier wave, which carries computer-readable program code. Such a propagated data signal can take various forms, including but not limited to an electromagnetic signal, an optical signal, or any suitable combination thereof. The computer-readable signal medium can also be any computer-readable medium other than the computer-readable storage medium, which can send, transmit, propagate or transport a program for use by or in conjunction with an instruction execution system, device or apparatus. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to wireless, wired, or the like, or any suitable combination thereof.

[0062] The flowcharts and block diagrams in the drawings illustrate the possible implementation architectures, functions and operations of the systems, methods and computer program products according to various embodiments of the present application. In the flowcharts or block diagrams, each block can represent a module, a program segment or a part of code, which contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions noted in the blocks can occur in different orders than that shown in the drawings. For example, two blocks that are shown in succession can actually be executed substantially in parallel, and sometimes in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams or flowcharts, and the combination of blocks in the block diagrams or flowcharts, can be implemented by a dedicated hardware-based system that performs the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0063] The units described in the embodiments of the present application can be implemented in the form of software, or can be implemented in the form of hardware, and the described units can also be arranged in a processor. In some cases, the names of the units do not constitute a limitation on the units themselves.

[0064] As another aspect, the present application also provides a computer program product or computer program, which includes computer instructions stored in a computer readable storage medium. A processor of a computer device reads the computer instructions from the computer readable storage medium, and the processor executes the computer instructions to enable the computer device to perform the training method of the mild cognitive impairment screening model described in the above embodiments.

[0065] As another aspect, the present application also provides a computer readable medium, which can be included in the electronic device described in the above embodiments, or can exist separately without being assembled into the electronic device. The computer readable medium carries one or more programs, which, when executed by the electronic device, enable the electronic device to perform the training method of the mild cognitive impairment screening model described in the above embodiments.

[0066] It should be noted that although several modules or units of the device for action execution are mentioned in the above detailed description, such a division is not mandatory. In fact, according to the embodiments of the present application, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided into modules or units embodied by multiple modules or units.

[0067] From the above description of the embodiments, those skilled in the art can easily understand that the example embodiments described herein can be implemented by software, or by software in combination with necessary hardware. Therefore, the technical solutions according to the embodiments of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, U disk, mobile hard disk, etc.) or network, and includes several instructions to enable a computing device (which can be a personal computer, server, touch terminal, or network device, etc.) to perform the methods according to the embodiments of the present application.

[0068] Other embodiments of the application will be apparent to those skilled in the art from consideration of the specification and practice of the embodiments disclosed herein. It is intended that the application embrace any and all variations of the application that fall within the scope of the general description herein. It is to be understood that the application is not to be limited to the specific examples, methods, and procedures described herein, and that specific examples are to be considered as illustrative only. It is further understood that the application can encompass all such variations as fall within the scope of the application. It is intended that the scope of the application encompass all techniques capable of approximating the teachings provided herein.

Claims

1. A training method for a mild cognitive impairment screening model, characterized in that: include: Collect research sample data based on health data of natural population in the community; Acquiring characteristic variables related to mild cognitive impairment based on the research sample data as training data; The optimal AUC was used as the training direction for model training, and an extreme gradient boosting algorithm model was trained based on the training data to obtain a mild cognitive impairment screening model.

2. The method according to claim 1, characterized in that The method uses the optimal AUC as the training direction of the model training, trains the extreme gradient boosting algorithm model based on the training data, and obtains a mild cognitive impairment screening model, including: Taking the best AUC as the training direction of the model, under the premise of expert opinions and 5-fold cross-validation, the optimal parameter combination of the model in the hyperparameter space is calculated; Under the optimal parameter combination, the extreme gradient boosting algorithm model is trained according to the training data to obtain a mild cognitive impairment screening model.

3. The method according to claim 2, characterized in that Based on expert opinions and 5-fold cross-validation, the optimal parameter combination of the model in the hyperparameter space is calculated, including: Random seed, class balance, learning rate, and maximum depth are used as the main target optimization parameters of the mild cognitive impairment screening model, and the hyperparameter space search range of each main target optimization parameter is determined; A grid search strategy was used to search and train the optimal parameter configuration of the mild cognitive impairment screening model within the hyperparameter space search range to obtain the optimal parameter combination, and the sensitivity, specificity, accuracy, AUC value and ROC curve were recorded.

4. The method according to any one of claims 1 to 3, characterized in that The research sample data based on the community natural population health data collection includes: Recruit population samples from natural population communities, use a cohort research strategy as an entry point, conduct health assessments on all population samples, and collect health assessment data as the research sample data.

5. The method according to claim 4, characterized in that The step of obtaining characteristic variables related to mild cognitive impairment based on the research sample data as training data includes: Based on the literature review method, professional domain knowledge injection method, domain expert consultation method and Shap interpretable machine learning method, characteristic variables related to mild cognitive impairment were screened out from the evaluation data as training data.

6. A training device for a mild cognitive impairment screening model, characterized in that: include: Collection unit, used to collect research sample data based on health data of natural population in the community; a screening unit, configured to obtain characteristic variables related to mild cognitive impairment based on the research sample data as training data; A training unit is used to use the optimal AUC as a training direction for model training, train an extreme gradient boosting algorithm model based on the training data, and obtain a mild cognitive impairment screening model.

7. A computer-readable storage medium, characterized in that The storage medium stores computer instructions, which, when executed on a computer, enable the computer to execute the method according to any one of claims 1 to 5.

8. An electronic device, characterized in that: include: memory and processor; The memory is used to store computer instructions; The processor is configured to call the computer instructions stored in the memory so that the electronic device executes the method according to any one of claims 1 to 5.

Images