Auxiliary diagnosis method and system for attention deficit hyperactivity disorder of children

By comprehensively evaluating multiple cognitive domains of children and using support vector machine models, the problem that existing diagnostic methods rely on subjective observation and complex processes is solved, improving the accuracy and operability of ADHD diagnosis, reducing misdiagnosis and Missing diagnosis enhances the sensitivity and specificity of the diagnosis.

CN118983076APending Publication Date: 2024-11-19BEIJING WISPIRIT TECH CO LTD +1
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202410968482.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-18
Publication Date
2024-11-19

AI Technical Summary

Technical Problem

The existing diagnostic methods for children with attention deficit hyperactivity disorder rely on clinical evaluation and subjective observation, are susceptible to bias, and the diagnosis process is complex and time-consuming, the accuracy is closely related to the clinical level of the doctor, and the existing continuous operational tests There are differences in diagnostic capabilities and cannot fully reflect the complexity of ADHD.

Method used

Comprehensively evaluate children's multiple cognitive functional domains, such as perception, attention, memory and inhibitory control, and assist in diagnosis through the support vector machine model, reduce subjective bias, improve diagnosis accuracy and operability, and build a support vector machine The model also improves the model's distinctive ability through feature selection and data preprocessing.

Benefits of technology

It significantly improves the accuracy and operability of the diagnosis, reduces the possibility of misdiagnosis and missed diagnosis, enhances the sensitivity and specificity of the diagnosis, and can effectively distinguish patients with attention deficit hyperactivity disorder from other groups with similar symptoms.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118983076A_ABST
    Figure CN118983076A_ABST
Patent Text Reader

Abstract

The invention discloses an auxiliary diagnosis method and system for children attention deficit hyperactivity disorder. The method comprises the following steps: setting a cognitive ability test according to cognitive characteristics and development conditions of a child group; collecting cognitive ability test data of subjects in the attention deficit hyperactivity disorder diagnosis group and cognitive ability test data of subjects in the normal control group; preprocessing the collected cognitive competence test data; and a support vector machine model is constructed, and the trained support vector machine model is utilized to implement auxiliary diagnosis of the children attention deficit hyperactivity disorder. According to the method, the support vector machine model is used for assisting in diagnosing the attention deficit hyperactivity disorder, and the accuracy and operability of diagnosis are remarkably improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to an auxiliary diagnosis method for attention deficit hyperactivity disorder in children, and also relates to a corresponding auxiliary diagnosis system, belonging to the technical field of medical care informatics. Background Art

[0002] Attention deficit hyperactivity disorder (ADHD) is a common neurodevelopmental disorder characterized by hyperactivity, impulsivity, and inattention. This disorder affects approximately 5% of children and adolescents worldwide, and has a significant impact on their mental health and social adaptability. Currently, the diagnosis of ADHD relies mainly on clinical assessment, reports from parents and teachers, and observation of children's behavior. However, this diagnostic method is susceptible to subjective bias, and the diagnostic process is complex and time-consuming, and its accuracy is closely related to the doctor's clinical level. The number of professional doctors worldwide is far from meeting the demand for ADHD diagnosis.

[0003] In order to improve the accuracy and operability of ADHD diagnosis, researchers are exploring a variety of diagnostic tools based on cognitive tests. These diagnostic tools provide a more objective and reliable diagnostic approach by assessing cognitive abilities such as perception, attention, memory, and inhibitory control. Nevertheless, existing continuous performance tests (CPTs) vary in diagnostic ability, and the cognitive abilities they involve are limited and cannot fully reflect the complexity of ADHD. Therefore, researchers recommend using a combination of multiple cognitive tests rather than a single cognitive test to more comprehensively diagnose ADHD.

[0004] In order to diagnose ADHD more effectively, it is necessary to develop and apply more objective and comprehensive cognitive testing methods. These cognitive testing methods should not only reduce subjective bias in the diagnostic process, but also improve the accuracy and reliability of the diagnosis. At the same time, considering the complexity of ADHD symptoms, the combination of multiple cognitive testing methods is a more scientific and effective diagnostic method. In the Chinese invention application with application number 202210191328.0, a deep learning-based algorithm for identifying behavioral characteristics of children with ADHD is disclosed. It fuses the attention fusion weights and the eigenvalues ​​of multiple models through inner product operations to improve the accuracy of recognition and assist medical workers in completing the preliminary diagnosis of children with ADHD. Summary of the invention

[0005] The primary technical problem to be solved by the present invention is to provide an auxiliary diagnosis method for attention deficit hyperactivity disorder in children.

[0006] Another technical problem to be solved by the present invention is to provide an auxiliary diagnosis system for attention deficit hyperactivity disorder in children.

[0007] In order to achieve the above technical objectives, the present invention adopts the following technical solutions:

[0008] According to a first aspect of an embodiment of the present invention, there is provided a method for auxiliary diagnosis of attention deficit hyperactivity disorder in children, comprising the following steps:

[0009] S1: Cognitive ability tests are set up according to the cognitive characteristics and development of the child group; wherein the cognitive ability tests include: perception test, attention test, memory test and inhibition control test;

[0010] S2: Collect cognitive ability test data of subjects in the attention deficit hyperactivity disorder diagnosis group and the cognitive ability test data of subjects in the normal control group;

[0011] S3: preprocessing the collected cognitive ability test data;

[0012] S4: Construct a support vector machine model and use the trained support vector machine model to implement auxiliary diagnosis of attention deficit hyperactivity disorder in children.

[0013] Preferably, in step S1, the perceptual test includes: a time selection task, a paper folding task, a random motion point task and a symbol search task;

[0014] The attention test includes: vigilance task, click / no-click task, visual search task and cancellation task;

[0015] The memory test includes: a digit span forward task and a digit span backward task;

[0016] The inhibitory control test includes: a flanking inhibition task and a number comparison task.

[0017] Preferably, in step S3, the preprocessing includes data cleaning and feature selection; the data cleaning includes: eliminating abnormal values ​​and / or missing values ​​that appear during the cognitive ability test; the feature selection includes: selecting cognitive ability tests that are highly correlated with attention deficit hyperactivity disorder by analyzing the correlation of the data.

[0018] Preferably, the feature selection includes the following sub-steps:

[0019] S31: performing descriptive statistical analysis on the collected cognitive ability test data, wherein the descriptive statistical analysis includes mean, median and standard deviation;

[0020] S32: Pearson correlation coefficients were calculated between each cognitive ability test and ADHD symptom scores to measure the linear correlation between continuous traits;

[0021] S33: Based on linear correlation, features highly correlated with ADHD symptoms were obtained.

[0022] Preferably, in the sub-step S32, the calculation formula is:

[0023]

[0024] Among them, X i is the cognitive ability paradigm score of the i-th sample; Y i The score of the Attention Deficit Hyperactivity Disorder Core Symptom Assessment Scale; It is the mean of cognitive ability test scores of all samples in the database; It is the mean score of the Attention Deficit Hyperactivity Disorder Core Symptom Assessment Scale for all samples in the database.

[0025] Preferably, in step S4, constructing a support vector machine model includes the following sub-steps:

[0026] S41: Standardize the features so that the value range of each feature is consistent;

[0027] S42: Construct new feature combinations based on the interactions between features;

[0028] S43: Apply radial basis function to construct support vector machine model;

[0029] S44: Adjust the hyperparameters C and γ through grid search and cross validation; where C is the penalty parameter and γ is an adjustable parameter;

[0030] S45: using the preprocessed data to train a support vector machine model, and determining whether the support vector machine model can effectively distinguish between the ADHD group subjects and the normal control group subjects;

[0031] If the support vector machine model can reach the preset effective distinction threshold, it is determined to be able to effectively distinguish, and the process proceeds to step S46; if the support vector machine model cannot reach the preset effective distinction threshold, it is determined to be unable to effectively distinguish, and steps S41 to S45 are repeated;

[0032] S46: Divide the preprocessed data into multiple folds through cross-validation, cyclically use each fold as a validation set, and use the remaining folds as training sets, and select the hyperparameter set that performs best in terms of model accuracy of the support vector machine model as the optimal parameter, thereby determining the trained support vector machine model.

[0033] Preferably, the expression of the support vector machine model is:

[0034] K(x i, x j)=exp(-γ∥x i -x j ∥ 2 )

[0035] Among them, x i and x j are all feature vectors in the input data set; ∥x i -x j ∥ 2 is the squared Euclidean distance between two eigenvectors.

[0036] Preferably, the standardization process includes the following sub-steps:

[0037] S411: Calculate the mean and standard deviation of each feature;

[0038] S412: The mean and standard deviation are applied to the data through the Z-score standardization method. The corresponding formula is:

[0039] X′=(X-μ) / σ

[0040] Among them, X is the original data; X′ is the standardized data; μ is the mean of the original data; σ is the standard deviation of the original data.

[0041] Preferably, the construction of a new feature combination includes the following sub-steps:

[0042] S421: Obtain nonlinear relationships between features and identify feature interactions through a correlation matrix and / or an information matrix;

[0043] S422: Construct statistical characteristics of the feature combination, including mean, standard deviation, maximum value and minimum value.

[0044] According to a second aspect of an embodiment of the present invention, there is provided an auxiliary diagnosis system for attention deficit hyperactivity disorder in children, comprising a processor and a memory; wherein the memory is coupled to the processor and is used to store a computer program, and when the computer program is executed by the processor, the processor implements the above-mentioned auxiliary diagnosis method.

[0045] Compared with the prior art, the present invention uses a support vector machine model to assist in the diagnosis of attention deficit hyperactivity disorder by comprehensively evaluating multiple cognitive function domains of children, such as perception, attention, memory and inhibitory control, and significantly improves the accuracy and operability of diagnosis. This method reduces the subjective bias in the traditional diagnostic process, enhances the sensitivity and specificity of diagnosis, effectively distinguishes patients with attention deficit hyperactivity disorder from other groups with similar symptoms, and reduces the possibility of misdiagnosis and missed diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1A flowchart of an auxiliary diagnosis method for attention deficit hyperactivity disorder in children provided by an embodiment of the present invention;

[0047] Figure 2 A schematic diagram of an auxiliary diagnosis system for attention deficit hyperactivity disorder in children provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0048] The technical content of the present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0049] First embodiment

[0050] like Figure 1 As shown, the first embodiment of the present invention provides an auxiliary diagnosis method for attention deficit hyperactivity disorder in children, which at least includes the following steps:

[0051] S1: Set up cognitive ability tests based on the cognitive characteristics and development of the child group.

[0052] Cognitive ability tests include: perception tests, attention tests, memory tests and inhibitory control tests.

[0053] Perceptual tests include: time selection task, paper folding task, random motion point task and symbol search task, which are used to test children's time perception, space perception, motion perception and processing speed.

[0054] The time selection task includes: the subject looks at the screen, and when a bomb icon appears on the screen, the subject starts to count in seconds. When the bomb icon turns into an explosion icon, the subject chooses the option closest to the number of seconds he counted in the task.

[0055] The origami task includes: the subject imagines a square piece of paper, folds the paper as required by the topic through imagination, imagines poking a hole in the paper after folding it, and finally unfolds it back to a square. After unfolding it back to a square, the subject is asked to choose how many holes there are in the paper at this time and the location of each hole.

[0056] The random moving point task includes: displaying multiple moving points on the screen, some of which move in a single direction to the left or right. The subjects observe the movement trajectory of each point, find the points that move in a single direction to the left or right, and click the arrow button of the corresponding direction.

[0057] The symbol search task includes: dividing the screen into two parts, the first part includes two figures, and the second part includes five figures. The subjects judge whether the five figures in the second part contain one of the two figures in the first part. If so, click the button with words; if not, click the button without words.

[0058] Attention tests include: alertness task, click / no-click task, visual search task and cancellation task. Its function is to test children's alertness, concentration, selective attention and sustained attention.

[0059] The alertness task included: the subject looked at the content on the screen, and when the content started to change, the subject clicked a stop button to stop the change.

[0060] For example, numbers are displayed on the screen. When the value of the number starts to increase (such as from 1 to 2, from 2 to 3... and so on), the subject clicks the stop button as quickly as possible to stop the increase in the value.

[0061] The click / no-click task includes: two different stimuli are presented continuously on the screen, and the subjects click on the stimulus that meets the requirements according to the requirements of the question.

[0062] For example: the subject clicks anywhere on the screen when he sees a tiger, but does not click when he sees a giraffe.

[0063] The visual search task includes: the subjects judge whether the target object appears in the displayed content and make a key response.

[0064] For example: determine whether there are green lines among the many red lines on the screen.

[0065] The cancellation task includes: the subject remembers the target graphics that appear in the graphic matrix and clicks on the target graphics in order from top to bottom and from left to right. The missed target graphics cannot be clicked back.

[0066] Preferably, the test time for the cancellation task is 3 minutes.

[0067] Preferably, the size of the graphic matrix is ​​15*10.

[0068] The memory test includes: digit span forward task and digit span backward task. Its function is to test children's verbal working memory and memory span.

[0069] The digit span forward-back task involves subjects memorizing sequentially presented digits and repeating them in chronological order.

[0070] The digit span backward task involves subjects memorizing digits presented sequentially and repeating them in reverse order.

[0071] The inhibitory control test includes the flanker task and the digital comparison task, which is used to test children's interference control and conflict inhibition.

[0072] The lateral inhibition task includes: the subject pays attention to the shape in the center of the screen, judges the direction of the shape, and clicks the corresponding direction button according to the direction.

[0073] The number comparison task involves the subjects repeatedly switching the selected criteria according to the rules to inhibit habitual thinking.

[0074] For example: The rules of the question are to pay attention to the values ​​and font sizes of the numbers on the left and right sides, and select the corresponding numbers according to the question.

[0075] S2: Collect cognitive ability test data of subjects in the attention deficit hyperactivity disorder diagnosis group and cognitive ability test data of subjects in the normal control group.

[0076] S3: Preprocess the collected cognitive ability test data.

[0077] Preprocessing includes data cleaning and feature selection.

[0078] Data cleaning includes: removing outliers and / or missing values ​​that appear during cognitive ability testing to ensure data integrity and reliability. Outliers include: taking the 3 standard deviations of the overall score data of the corresponding task score in the database as the boundary value, and data outside the mean ± 3 standard deviations are considered outliers. Missing values ​​include: the subject did not complete the corresponding task or the system did not record the score data of the corresponding task.

[0079] Feature selection includes: by analyzing the correlation of data, selecting cognitive ability tests that are highly correlated with ADHD, thereby improving the performance and interpretability of the model. The high correlation here refers to the psychological paradigm that has a statistically significant correlation (P<0.05) with the score of the ADHD Core Symptom Assessment Scale.

[0080] In one embodiment of the present invention, feature selection includes the following sub-steps:

[0081] S31: Conduct descriptive statistical analysis on the collected cognitive ability test data.

[0082] Descriptive statistical analysis includes mean, median and standard deviation, which is used to understand the basic characteristics and distribution of data.

[0083] S32: Pearson correlation coefficients were calculated between each cognitive ability test and ADHD symptom scores to measure the linear correlation between continuous traits.

[0084] The calculation formula is:

[0085]

[0086] Among them, X iY is the cognitive ability paradigm score of the i-th sample. i The score of the Attention Deficit Hyperactivity Disorder Core Symptom Assessment Scale. It is the mean of the cognitive ability test scores of all samples in the database. It is the mean score of the Attention Deficit Hyperactivity Disorder Core Symptom Assessment Scale for all samples in the database.

[0087] S33: Based on linear correlation, features highly correlated with ADHD symptoms were obtained.

[0088] These features can be used as input to a support vector machine (SVM) model to help the SVM model more accurately distinguish between patients with attention deficit hyperactivity disorder and normal control subjects.

[0089] S4: Construct a support vector machine model and use the trained support vector machine model to implement auxiliary diagnosis of attention deficit hyperactivity disorder in children.

[0090] In one embodiment of the present invention, constructing a support vector machine model includes the following sub-steps:

[0091] S41: Standardize the features so that the value range of each feature is consistent.

[0092] The standardization process includes the following sub-steps:

[0093] S411: Calculate the mean and standard deviation of each feature.

[0094] S412: The mean and standard deviation are applied to the data through the Z-score standardization method. The corresponding formula is:

[0095] X′=(X-μ) / σ

[0096] Where X is the original data. X′ is the standardized data. μ is the mean of the original data. σ is the standard deviation of the original data.

[0097] S42: Construct new feature combinations based on the interactions between features.

[0098] Constructing a new feature combination includes the following sub-steps:

[0099] S421: Obtain nonlinear relationships between features through correlation matrix and / or information matrix, and identify feature interactions.

[0100] S422: Construct statistical characteristics of the feature combination, including mean, standard deviation, maximum value and minimum value.

[0101] S43: Applying the radial basis function (RBF for short) to construct a support vector machine model.

[0102] The expression of the support vector machine model is:

[0103] K(x i, x j )=exp(-γ∥x i -x j ∥ 2 )

[0104] Among them, x i and x j are all feature vectors in the input data set. The feature vector is a vector composed of all features. For the i-th sample, its feature vector is represented by x i =[x i1 ,x i2 ,…,x in ], similarly, the feature vector of the jth sample is represented by x j =[x j1 ,x j2 ,…,x jn ]. ∥x i -x j ∥ 2 is the squared Euclidean distance between two eigenvectors.

[0105] S44: Adjust the hyperparameters C and γ through grid search (Grid Search CV) and cross-validation (Cross-Validation).

[0106] Among them, the parameter C is the penalty parameter, which determines the degree of penalty for misclassified samples during training. γ is an adjustable parameter used to control the width of the function, that is, the degree of "diffusion" of data points in high-dimensional space.

[0107] It should be noted that parameter C is a hyperparameter in the support vector machine model, and the optimal value can be selected through experiments and cross-validation. Although there is no fixed calculation formula, the optimal C value can be effectively found through tuning techniques such as grid search, random search, and Bayesian optimization.

[0108] S45: Using the preprocessed data to train the support vector machine model, and determining whether the support vector machine model can effectively distinguish between the ADHD group subjects and the normal control group subjects.

[0109] If the support vector machine model can reach the preset effective distinction threshold, it is determined to be effective, and the process proceeds to step S46. If the support vector machine model cannot reach the preset effective distinction threshold, it is determined to be ineffective, and steps S41 to S45 are repeated.

[0110] The preset effective distinction threshold is that the accuracy of the support vector machine model is greater than 80%.

[0111] S46: Divide the preprocessed data into multiple folds through cross-validation, cyclically use each fold as a validation set, and use the remaining folds as training sets, and select the hyperparameter set that performs best in terms of model accuracy of the support vector machine model as the optimal parameter, thereby determining the trained support vector machine model.

[0112] Next, the trained support vector machine model is applied to new subject data. Based on the classification results output by the support vector machine model, medical workers can make more accurate and reliable diagnosis of children's attention deficit hyperactivity disorder.

[0113] Second embodiment

[0114] Based on the above method, a second embodiment of the present invention provides an auxiliary diagnosis system for attention deficit hyperactivity disorder in children.

[0115] like Figure 2 As shown, the auxiliary diagnosis system includes one or more processors 41 and a memory 42. The memory 42 is used to store one or more programs. When the one or more programs are executed by the one or more processors 41, the auxiliary diagnosis method in the above embodiment is implemented.

[0116] Wherein, the processor is used to control the overall operation of the auxiliary diagnosis system to complete all or part of the steps of the above method. The processor can be a central processing unit (CPU), a graphics processing unit (GPU), a field programmable gate array (FPGA), an application-specific integrated circuit (ASIC), a digital signal processing (DSP) chip, etc. The memory is used to store various types of data to support the operation of the auxiliary diagnosis system, and these data may include, for example, instructions for any application or method used to operate on the auxiliary diagnosis system, and application-related data. The memory can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as a static random access memory (SRAM), an electrically erasable programmable read-only memory (EEPROM), an erasable programmable read-only memory (EPROM), a programmable read-only memory (PROM), a read-only memory (ROM), a magnetic memory, a flash memory, etc.

[0117] In an exemplary embodiment, the auxiliary diagnosis system can be implemented by a computer chip or entity, or by a product with a certain function, for executing the above method and achieving the same technical effect as the above method. A typical embodiment is a CT device, a security inspection device, a non-destructive testing device, and a computer. Specifically, the computer can be, for example, a personal computer, a laptop computer, a vehicle-mounted human-computer interaction device, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

[0118] In another exemplary embodiment, the present invention further provides a computer-readable storage medium including program instructions, which, when executed by a processor, implements the steps of the auxiliary diagnosis method in any of the above embodiments. For example, the computer-readable storage medium may be the above-mentioned memory including program instructions, and the above-mentioned program instructions may be executed by a processor to complete the above-mentioned method and achieve the same technical effect as the above-mentioned method.

[0119] It should be noted that the above-mentioned multiple embodiments are only examples, and the technical solutions of various embodiments can be combined, all of which are within the protection scope of the present invention.

[0120] Compared with the prior art, the present invention uses a support vector machine model to assist in the diagnosis of attention deficit hyperactivity disorder by comprehensively evaluating multiple cognitive function domains of children, such as perception, attention, memory and inhibitory control, and significantly improves the accuracy and operability of diagnosis. This method reduces the subjective bias in the traditional diagnostic process, enhances the sensitivity and specificity of diagnosis, effectively distinguishes patients with attention deficit hyperactivity disorder from other groups with similar symptoms, and reduces the possibility of misdiagnosis and missed diagnosis.

[0121] The auxiliary diagnosis method and system for attention deficit hyperactivity disorder in children provided by the present invention are described in detail above. For those skilled in the art, any obvious changes made to it without departing from the essence of the present invention will constitute an infringement of the patent right of the present invention and will bear corresponding legal responsibilities.

Claims

1. An auxiliary diagnostic method for attention deficit hyperactivity disorder in children. Features The steps include: S1: Cognitive ability tests are set up according to the cognitive characteristics and development of the child group; wherein the cognitive ability tests include: perception test, attention test, memory test and inhibition control test; S2: Collect cognitive ability test data of subjects in the attention deficit hyperactivity disorder diagnosis group and the cognitive ability test data of subjects in the normal control group; S3: preprocessing the collected cognitive ability test data; S4: Construct a support vector machine model and use the trained support vector machine model to implement auxiliary diagnosis of attention deficit hyperactivity disorder in children.

2. The auxiliary diagnosis method according to claim 1, Features In step S1, the perceptual test includes: a time selection task, a paper folding task, a random motion point task and a symbol search task; The attention test includes: vigilance task, click / no-click task, visual search task and cancellation task; The memory test includes: a digit span forward task and a digit span backward task; The inhibitory control test includes: a flanking inhibition task and a number comparison task.

3. The auxiliary diagnosis method according to claim 1, Features In the step S3, the preprocessing includes data cleaning and feature selection; The data cleaning includes: removing abnormal values ​​and / or missing values ​​that appear during the cognitive ability test; the feature selection includes: selecting cognitive ability tests that are highly correlated with attention deficit hyperactivity disorder by analyzing the correlation of the data.

4. The auxiliary diagnosis method according to claim 3, Features The feature selection includes the following sub-steps: S31: performing descriptive statistical analysis on the collected cognitive ability test data, wherein the descriptive statistical analysis includes mean, median and standard deviation; S32: Pearson correlation coefficients were calculated between each cognitive ability test and ADHD symptom scores to measure the linear correlation between continuous traits; S33: Based on linear correlation, features highly correlated with ADHD symptoms were obtained.

5. The auxiliary diagnosis method according to claim 4, Features In the sub-step S32, the calculation formula is: Among them, X i is the cognitive ability paradigm score of the i-th sample; Y i The score of the Attention Deficit Hyperactivity Disorder Core Symptom Assessment Scale; It is the mean of cognitive ability test scores of all samples in the database; It is the mean score of the Attention Deficit Hyperactivity Disorder Core Symptom Assessment Scale for all samples in the database.

6. The auxiliary diagnosis method according to claim 1, Features In step S4, constructing a support vector machine model includes the following sub-steps: S41: Standardize the features so that the value range of each feature is consistent; S42: Construct new feature combinations based on the interactions between features; S43: Apply radial basis function to construct support vector machine model; S44: Adjust the hyperparameters C and γ through grid search and cross validation; where C is the penalty parameter and γ is an adjustable parameter; S45: using the preprocessed data to train a support vector machine model, and determining whether the support vector machine model can effectively distinguish between the ADHD group subjects and the normal control group subjects; If the support vector machine model can reach the preset effective distinction threshold, it is determined to be able to effectively distinguish, and the process proceeds to step S46; if the support vector machine model cannot reach the preset effective distinction threshold, it is determined to be unable to effectively distinguish, and steps S41 to S45 are repeated; S46: Divide the preprocessed data into multiple folds through cross-validation, cyclically use each fold as a validation set, and use the remaining folds as training sets, and select the hyperparameter set that performs best in terms of model accuracy of the support vector machine model as the optimal parameter, thereby determining the trained support vector machine model.

7. The auxiliary diagnosis method according to claim 6, Features The expression of the support vector machine model is: K(x i, x j )=exp(-γ∥x i -x j ∥ 2 ) Among them, x i and x j are all feature vectors in the input data set; ∥x i -x j ∥ 2 is the squared Euclidean distance between two eigenvectors.

8. The auxiliary diagnosis method according to claim 6, Features The standardization process includes the following sub-steps: S411: Calculate the mean and standard deviation of each feature; S412: The mean and standard deviation are applied to the data through the Z-score standardization method. The corresponding formula is: X′=(X-μ) / σ Among them, X is the original data; X′ is the standardized data; μ is the mean of the original data; σ is the standard deviation of the original data.

9. The auxiliary diagnosis method according to claim 6, Features The construction of a new feature combination includes the following sub-steps: S421: Obtain nonlinear relationships between features and identify feature interactions through a correlation matrix and / or an information matrix; S422: Construct statistical characteristics of the feature combination, including mean, standard deviation, maximum value and minimum value.

10. An auxiliary diagnosis system for children's attention deficit hyperactivity disorder, Features It comprises a processor and a memory; wherein the memory is coupled to the processor and is used to store a computer program, and when the computer program is executed by the processor, the processor implements the auxiliary diagnosis method described in any one of claims 1 to 9.

Citation Information

Cited By

  • Method, device and system for assisting in recognizing attention deficit hyperactivity disorder based on electrocardiogram signals

    CN119314657A

  • Method, device and system for assisting in identifying attention deficit hyperactivity disorder based on electrocardiogram signals

    CN119314657B