Method, device and equipment for analyzing the correlation between brain and behavior of people with mental illness

Abstract

The present invention is applicable to the field of neuroscience technology and provides a method, device and equipment for analyzing the association between brain and behavior in a group of mental illnesses. The method includes: collecting brain structural image data and behavioral data of a target group, processing the brain structural image data and behavioral data separately to obtain a GMV matrix and a behavior matrix, performing a latent component analysis of the brain-behavior association on the GMV matrix and the behavior matrix to obtain a GMV score matrix and a behavior score matrix, and based on the GMV score matrix and the behavior score matrix, performing a mediation analysis of social support factors on the brain-behavior association of the target group to determine the significant mediating path of the brain-behavior association of the target group, thereby significantly improving the statistical power of neuroimaging analysis of mental illness, accurately revealing the intrinsic connection between brain structure and behavior, deeply analyzing the pathogenesis of mental illness, and providing more targeted targets for early intervention and personalized treatment of mental illness.

Description

Technical Field

[0001] The present invention belongs to the field of neuroscience technology, and in particular relates to a method, device and equipment for analyzing the correlation between the brain and behavior of a group of mental illnesses. Background Art

[0002] Structural Magnetic Resonance Imaging (sMRI), a non-invasive, high-spatial-resolution brain structural imaging technique, can precisely observe brain anatomy. Research is conducted using indicators such as gray matter volume (reflecting changes in neuronal density and tissue volume in specific brain regions), cortical thickness (reflecting the development and degeneration of cortical structure), and the volume of specific brain regions. Among these, gray matter volume (GMV), as a stable and reliable indicator of brain structure, is widely used in neurological disease research. For example, a study extracted whole-brain GMV from patients with Major Depressive Disorder (MDD) and Bipolar Disorder (BD), providing further evidence for brain structural abnormalities in these two disorders and aiding the development of diagnostic biomarkers for psychiatric disorders.

[0003] Existing research on non-suicidal self-injury (NSSI) in adolescents with depression based on magnetic resonance imaging (MRI) has the following shortcomings:

[0004] Sample size limitations: Most studies have small sample sizes, which limits the general applicability and reliability of the research results and makes it difficult to support cross-group generalization;

[0005] Single diagnostic dimension: Some studies lack patients with clinically diagnosed depression, and the severity of depression is assessed only by scales, which cannot accurately reflect the deep pathological characteristics of brain activity during the disease process;

[0006] Lack of correlation analysis: Existing studies have mostly explored behavioral performance, brain structure, or brain function in isolation, without systematically exploring the potential correlation pathways between behavior, brain structure, and brain function.

[0007] Mechanistic gaps: No research has yet explored the factors (such as social environment) that influence the brain-behavior patterns of depression patients, making it difficult to reveal the complete mechanism of disease development.

[0008] These deficiencies have led to insufficient explanation of the pathological mechanisms of adolescent depression NSSI in current research. Therefore, a new method for analyzing the association between brain and behavior in groups with mental illness is urgently needed to achieve a more in-depth and comprehensive analysis of the disease mechanism. Summary of the Invention

[0009] The purpose of the present invention is to provide a method, device and equipment for analyzing the correlation between brain and behavior in a group of mental illnesses, aiming to solve the problem that existing technologies cannot effectively reveal the intrinsic connection between brain structure and behavior.

[0010] In a first aspect, the present invention provides a method for analyzing the association between brain and behavior in a group of people with mental illness, the method comprising the following steps:

[0011] Collecting brain structural image data and behavioral data of the target group, and processing the collected brain structural image data and behavioral data respectively to obtain corresponding GMV matrix and behavior matrix;

[0012] Performing a latent component analysis of brain-behavior association on the GMV matrix and the behavior matrix to obtain a GMV score matrix and a behavior score matrix of the target group on significant latent components;

[0013] Based on the GMV score matrix and the behavior score matrix, a mediation analysis of social support factors is performed on the brain-behavior association of the target group to determine the significant mediating path of the brain-behavior association of the target group.

[0014] In some embodiments, the step of performing latent component analysis of brain-behavior association on the GMV matrix and the behavior matrix comprises:

[0015] Performing singular value decomposition on the covariance matrix of the GMV matrix and the behavior matrix to obtain a GMV significance matrix, a behavior significance matrix, and a singular value matrix;

[0016] Evaluating the statistical significance of each singular value in the singular value matrix through a permutation test, screening out significant singular values ​​that meet a preset significance threshold, and determining the potential components corresponding to the significant singular values ​​as the significant potential components;

[0017] Extracting column vectors corresponding to the significant latent components from the GMV significance matrix and the behavior significance matrix respectively to form a significant GMV weight matrix and a significant behavior weight matrix;

[0018] The GMV score matrix is ​​calculated based on the GMV matrix and the significant GMV weight matrix, and the behavior score matrix is ​​calculated based on the behavior matrix and the significant behavior weight matrix.

[0019] In some embodiments, the step of performing a mediation analysis of social support factors on the brain-behavior association of the target group based on the GMV score matrix and the behavior score matrix includes:

[0020] A mediation model was constructed with the GMV score matrix as the independent variable, the behavior score matrix as the dependent variable, and the social support factor as the mediating variable. The mediation model is: ,in, represents the GMV score matrix, represents the social support factors, represents the behavior score matrix, represents the first intercept, represents the first residual term, represents the second intercept, represents the second residual term, represents the regression coefficient of the independent variable on the mediating variable, represents the regression coefficient of the mediating variable on the dependent variable, It represents the regression coefficient of the direct effect of the independent variable on the dependent variable after controlling the mediating variable;

[0021] The indirect effect value of the social support factor between the GMV score matrix and the behavior score matrix is ​​calculated through the mediation model, and the significance test of the indirect effect value is performed based on the Bootstrap sampling method. The significant mediation path of the target group's brain and behavior association is determined according to the test results.

[0022] In some embodiments, the steps of processing the collected brain structural image data and the collected behavioral data separately include:

[0023] The brain structural image data are preprocessed using voxel-based morphological analysis to obtain whole-brain GMV data;

[0024] The whole-brain GMV data and the behavioral data were respectively subjected to Z-score normalization to obtain the GMV matrix and the behavioral matrix.

[0025] In some embodiments, the social support factor is a family support score and / or a friend support score, wherein the family support score is obtained through the family support subscale in the social support scale, and the friend support score is obtained through the friend support subscale in the social support scale.

[0026] In a second aspect, the present invention provides a device for analyzing the association between brain and behavior in a group of people with mental illnesses, the device comprising:

[0027] A target data processing unit is used to collect brain structural image data and behavioral data of the target group, and process the collected brain structural image data and behavioral data respectively to obtain corresponding GMV matrix and behavior matrix;

[0028] a latent component analysis unit, configured to perform a latent component analysis of brain-behavior association on the GMV matrix and the behavior matrix to obtain a GMV score matrix and a behavior score matrix of the target group on significant latent components;

[0029] A mediation effect analysis unit is used to perform a mediation analysis of social support factors on the brain-behavior association of the target group based on the GMV score matrix and the behavior score matrix, and determine the significant mediation path of the brain-behavior association of the target group.

[0030] In some embodiments, the latent component analysis unit includes:

[0031] a covariance decomposition unit, configured to perform singular value decomposition on the covariance matrix of the GMV matrix and the behavior matrix to obtain a GMV significance matrix, a behavior significance matrix, and a singular value matrix;

[0032] a singular value screening unit, configured to evaluate the statistical significance of each singular value in the singular value matrix by a permutation test, screen out significant singular values ​​that meet a preset significance threshold, and determine the potential components corresponding to the significant singular values ​​as the significant potential components;

[0033] A column vector extraction unit is used to extract column vectors corresponding to the significant latent components from the GMV significance matrix and the behavior significance matrix respectively, to form a significant GMV weight matrix and a significant behavior weight matrix;

[0034] A score matrix calculation unit is used to calculate the GMV score matrix based on the GMV matrix and the significant GMV weight matrix, and to calculate the behavior score matrix based on the behavior matrix and the significant behavior weight matrix.

[0035] In some embodiments, the mediation effect analysis unit includes:

[0036] The mediation model construction unit is used to construct a mediation model with the GMV score matrix as the independent variable, the behavior score matrix as the dependent variable, and the social support factor as the mediating variable. The mediation model is ,in, represents the GMV score matrix, represents the social support factors, represents the behavior score matrix, represents the first intercept, represents the first residual term, represents the second intercept, represents the second residual term, represents the regression coefficient of the independent variable on the mediating variable, represents the regression coefficient of the mediating variable on the dependent variable, It represents the regression coefficient of the direct effect of the independent variable on the dependent variable after controlling the mediating variable;

[0037] A mediation effect testing unit is used to calculate the indirect effect value of the social support factor between the GMV score matrix and the behavior score matrix through the mediation model, and perform a significance test on the indirect effect value based on the Bootstrap sampling method, and determine the significant mediation path of the target group's brain and behavior association based on the test results.

[0038] In a third aspect, the present invention further provides a computing device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-described method when executing the computer program.

[0039] In a fourth aspect, the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method described above are implemented.

[0040] The embodiment of the present invention collects brain structural image data and behavioral data of the target group, processes the collected brain structural image data and behavioral data respectively to obtain corresponding GMV matrix and behavior matrix, performs latent component analysis of brain-behavior association on the GMV matrix and behavior matrix, obtains GMV score matrix and behavior score matrix of the target group on significant latent components, and performs mediation analysis of social support factors on the brain-behavior association of the target group based on the GMV score matrix and behavior score matrix, determines the significant mediating path of the brain-behavior association of the target group, thereby significantly improving the statistical power of neuroimaging analysis of mental illness, accurately revealing the intrinsic connection between brain structure and behavior, deeply analyzing the pathogenesis of mental illness, and providing more targeted targets for early intervention and personalized treatment of mental illness. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 This is a flowchart of the method for analyzing the association between brain and behavior in a group of mental illnesses provided in Example 1 of the present invention;

[0042] Figure 2 This is a schematic diagram of the structure of the device for analyzing the association between brain and behavior of a group of people with mental illnesses provided in the second embodiment of the present invention;

[0043] Figure 3 It is a structural diagram of the computing device provided in Example 3 of the present invention. DETAILED DESCRIPTION

[0044] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0045] It should be understood that when used in this specification and the appended claims, the term "comprising" indicates the presence of the described features, integers, steps, operations, elements, and / or components, but does not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or combinations thereof. Furthermore, the terminology used in this specification is for the purpose of describing specific embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms unless the context clearly indicates otherwise. The terms "first," "second," and similar terms do not denote any order, quantity, or importance, but are simply used to distinguish one component from another. Terms such as "connected" or "connected" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positions; changes in the absolute position of the described objects may also change the relative positions of the objects. The term "plurality" refers to two or more, and other quantifiers are used similarly.

[0046] In order to keep the following description of the embodiments of the present invention clear and concise, detailed descriptions of some known functions and components are omitted in this specification.

[0047] The following describes the specific implementation of the present invention in detail with reference to specific embodiments:

[0048] Example 1:

[0049] Figure 1 The following illustrates the implementation process of the method for analyzing the association between brain and behavior of a group of patients with mental illnesses provided in Example 1 of the present invention. For ease of illustration, only the portion related to the embodiment of the present invention is shown, which is described in detail as follows:

[0050] In step S101 , the brain structure image data and behavioral data of the target group are collected, and the collected brain structure image data and behavioral data are processed respectively to obtain corresponding GMV matrix and behavior matrix.

[0051] The embodiment of the present invention is applicable to computing devices, such as personal computers, servers, etc. In the embodiment of the present invention, the target group is patients with mental illness, including but not limited to patients with depression, anxiety or bipolar disorder. Here, a high spatial resolution T1 structural image scan is performed on each subject in the target group. After the scan, the original DICOM (Digital Imaging and Communications in Medicine) format brain structural image data of each subject is obtained. At the same time, the behavioral data of the target group is collected through a series of standardized questionnaire assessments in behavioral scales covering multiple behavioral dimensions (such as emotions, cognition, social function, traumatic experience, etc.). The dimensions of the behavioral data are expressed as ,in, represents the total number of subjects in the target group, Indicates the number of behavioral indicators. Specifically, all participants in the target group completed the following questionnaire assessment:

[0052] Patient Health Questionnaire (PHQ): used to assess the overall health status of participants, including physical function, psychological status, etc.;

[0053] Generalized Anxiety Disorder (GAD): used to screen for generalized anxiety disorder and assess the severity of anxiety symptoms;

[0054] Beck Hopelessness Scale (BHS): used to measure an individual's level of hopelessness and predict depression and suicide risk;

[0055] UCLA Loneliness Scale (UCLA): used to assess an individual's loneliness and reflect the quality of social relationships;

[0056] Connor-Davidson Resilience Scale (CD-RISC): used to assess an individual's psychological adaptability in the face of stress and / or adversity;

[0057] Toronto Alexithymia Scale (TAS): used to measure an individual's difficulty in identifying and expressing emotions;

[0058] Rosenberg Self-Esteem Scale (RSES): used to assess an individual's self-worth and self-esteem;

[0059] Ruminative Responses Scale (RRS): This includes the symptom rumination (RRS-sr) subscale, the reflective pondering (RRS-rp) subscale, and the rumination-brooding (RRS-b) subscale, which assesses an individual's tendency to repeatedly think about negative emotions and events.

[0060] Emotion Regulation Questionnaire (ERQ): includes the Cognitive Reappraisal (ERQ-cr) subscale and the Expression Suppression (ERQ-es) subscale, which are used to assess individuals' emotion regulation strategies;

[0061] Borderline Personality Features Scale for Children (BPFS-C): includes the Affective Instability (BPFS-C-ai) subscale, the Negative Relationships (BPFS-C-nr) subscale, the Identity Problems (BPFS-C-ip) subscale, and the Self-Harm (BPFS-C-sh) subscale, used to assess borderline personality traits in children;

[0062] Childhood Trauma Questionnaire (CTQ): includes the Emotional Abuse (CTQ-ea) subscale, the Physical Abuse (CTQ-pa) subscale, the Sexual Abuse (CTQ-sa) subscale, the Emotional Neglect (CTQ-en) subscale, and the Physical Neglect (CTQ-pn) subscale, used to assess the type and degree of trauma experienced during childhood;

[0063] Perceived Stress Scales (PSS): used to assess the level of stress an individual experiences over a period of time;

[0064] Social Support Scale (SSS): As an intermediate factor, it is used to assess the level of social support received by an individual, including family support, friend support, etc. It includes the Family Social Support Subscale (FSSS) and the Friend Support Subscale (FSS);

[0065] Through the above questionnaire assessment, comprehensive behavioral data of the target group will be collected to provide a basis for subsequent analysis. These data will be combined with brain structural image data to conduct in-depth research on the relationship between the brain and behavior of the mental illness group. Among them, the standardized questionnaire assessment follows the following rules: using the above scale with clinical validity verification, according to the unified instructions, scoring rules and quality control standards of the scale manual, and conducted and recorded by professionals.

[0066] Each participant in the target group underwent a high-resolution T1 structural imaging scan. Specifically, before data acquisition began, the subject's head was fixed to minimize head movement, and a mute device was used to reduce noise interference during the scan. Subsequently, a high-resolution T1 structural imaging scan was performed using a Siemens 3T Prisma magnetic resonance imaging system. The parameters were: repetition time (TR) of 2530 milliseconds, echo time (TE) of 2.27 milliseconds, inversion time (TI) of 1100 milliseconds, flip angle of 7°, slice thickness of 1 mm, acquisition matrix of 256 × 256, and a total of 144 slices. During the scan, the subjects were instructed to close their eyes, remain relaxed but not asleep, and maintain a resting state. Finally, raw DICOM-formatted brain structural imaging data were obtained for all subjects, representing three-dimensional T1-weighted MRI images.

[0067] In a feasible embodiment, the collected brain structural image data and behavioral data are processed separately by the following steps:

[0068] (S101.1) Preprocess the brain structural image data using voxel-based morphological analysis to obtain whole-brain GMV data;

[0069] In an embodiment of the present invention, voxel-based morphometry (VBM) is used to preprocess the brain structural image data to obtain whole-brain GMV data, wherein the preprocessing operations include but are not limited to segmentation, spatial registration, smoothing, etc. Specifically, the brain structural image data is preprocessed by voxel-based morphometry using the SPM12 and CAT12 toolkits based on the MATLAB platform. The preprocessing includes the following sub-steps:

[0070] (S101.1.1) Using the SPM12 toolkit on the MATLAB platform, convert the original DICOM format brain structural imaging data to NIFTI (Neuroimaging Informatics Technology Initiative) format structural imaging data. Perform a data quality check on the NIFTI format structural imaging data to exclude subjects with low data quality.

[0071] (S101.1.2) Perform tissue segmentation on the filtered, NIFTI-formatted structural image data using the CAT12 toolkit. For each subject, obtain NIFTI-formatted data for gray matter (GM), white matter (WM), and cerebrospinal fluid (CSF).

[0072] (S101.1.3) Due to individual differences in brain anatomy among subjects, standardized comparisons and statistical inferences are difficult to achieve when performing direct group analysis. Therefore, spatial registration processing was performed on the gray matter (GM) data in NIFTI format (GM data) of all subjects. This involved mapping individual GM data to a unified standard spatial coordinate system (such as the Montreal Neurological Institute (MNI) coordinate system) to eliminate the interference of anatomical differences in group analysis. This resulted in standardized GM data, providing a standardized data foundation for subsequent studies such as brain volume statistics and inter-group comparisons.

[0073] (S101.1.4) Smooth the registered GM data to obtain smoothed GM data. Smoothing can improve the signal-to-noise ratio, suppress high-frequency noise, enhance the stability of the true signal, and strengthen the robustness of statistical tests, facilitating group comparison and analysis.

[0074] (S101.1.5) For the smoothed GM data, use the AAL3 template to extract the gray matter volume of the whole brain region of interest (ROI). Each subject will obtain several (in Represents) GMV data of ROI, and finally get dimensional whole-brain GMV data, where represents the total number of subjects in the target group, Indicates the number of ROIs, specifically, Set to 166, Set to 221.

[0075] (S101.2) Perform Z-score normalization on the whole-brain GMV data and behavioral data to obtain the GMV matrix and behavioral matrix, respectively.

[0076] In the embodiment of the present invention, since the dimensions of GMV and behavioral indicators are very different (for example, GMV is in mm³ and the behavioral scale is 0-100 points), the data needs to be standardized to eliminate the dimension effect and make the data comparable. Here, the whole brain GMV data is normalized. Each GMV indicator (column) is Z-score standardized to obtain the normalized GMV matrix , the formula is expressed as ,in, represents the column mean of the whole-brain GMV data, Represents the column standard deviation of the whole brain GMV data; similarly, for behavioral data Each behavioral indicator (column) is Z-score standardized to obtain the normalized behavioral matrix .

[0077] The above steps (S101.1) to (S101.2) are used to process the collected brain structural image data and behavioral data, thereby improving the data quality and ensuring the stability and accuracy of subsequent analysis.

[0078] In step S102, a latent component analysis of brain-behavior association is performed on the GMV matrix and the behavior matrix to obtain a GMV score matrix and a behavior score matrix of the target group on the significant latent components.

[0079] In an embodiment of the present invention, Partial Least Squares (PLS) is a multivariate statistical technique designed to identify latent components (LCs) that can best explain the covariance between two data sets. Here, PLS is used to perform latent component analysis of brain-behavior association on the GMV matrix and the behavior matrix to screen out significant latent components. Based on the projections of the GMV matrix and the behavior matrix on the significant latent components, a GMV score matrix and a behavior score matrix are generated. The GMV score matrix represents the comprehensive score of the GMV indicators of each subject on the significant latent components, and the behavior score matrix represents the comprehensive score of the behavioral indicators of each subject on the significant latent components, thereby helping to reveal the covariation pattern of brain structure and behavior in the sample and providing standardized data after dimensionality reduction for subsequent mediation analysis.

[0080] In one embodiment, latent component analysis of brain-behavior correlation is performed by the following steps:

[0081] (S102.1) Perform singular value decomposition on the covariance matrix of the GMV matrix and the behavior matrix to obtain a GMV significance matrix, a behavior significance matrix, and a singular value matrix;

[0082] In the embodiment of the present invention, first, the GMV matrix is ​​calculated and Behavior Matrix The correlation between and The covariance matrix of the synergistic relationship between , the calculation formula of the covariance matrix is , The dimension is , then, the covariance matrix Perform singular value decomposition (SVD) to obtain three low-dimensional matrices. Specifically, , these three low-dimensional matrices are GMV significance matrix , behavioral saliency matrix and the singular value matrix ,in, The dimension is , represents the contribution weight of the GMV indicator on the potential component (the larger the absolute value, the stronger the contribution), represents the number of potential components obtained by PLS analysis, The dimension is , represents the contribution weight of the behavioral indicator on the potential component, The dimension is , whose diagonal elements For the The singular value corresponding to the potential component represents the potential components (i.e., potential components )right and The explanatory power of the covariance of , The larger the value, the more potential the component right and The stronger the covariance explanatory power of the component, the stronger the brain-behavior association signal carried by the component.

[0083] (S102.2) Evaluate the statistical significance of each singular value in the singular value matrix through a permutation test, screen out significant singular values ​​that meet a preset significance threshold, and identify the latent components corresponding to the significant singular values ​​as significant latent components;

[0084] In the embodiment of the present invention, maintaining Unchanged, yes Perform random permutation to obtain the permuted behavior matrix , after each replacement, recalculate and The covariance matrix of the α is decomposed and the singular value is decomposed to obtain the permuted singular value matrix. Repeat this process Second-rate( ≥1000, preferred =5000), for each potential component, collect the permuted singular values, and use the formula according to the permuted singular values. Compute the true singular values ​​of each latent component The significance probability (i.e., obtained without permutation) ,when Less than the significance threshold When The singular value corresponding to the value is statistically significant, and the latent component corresponding to the singular value is determined as a significant latent component (i.e., a significant latent component), where Indicates the indicator function, which takes 1 when the condition is met, otherwise it takes 0. Indicates that statistics After random permutation Among all the singular values ​​corresponding to the potential components, the The number of Indicates the After the first replacement The singular values ​​corresponding to the latent components, Set to 0.05.

[0085] (S102.3) Extracting column vectors corresponding to significant latent components from the GMV significance matrix and the behavior significance matrix respectively, to form a significant GMV weight matrix and a significant behavior weight matrix;

[0086] In this embodiment of the present invention, the GMV significance matrix is ​​extracted The column vectors corresponding to the significant latent components in constitute the significant GMV weight matrix , expressed as , extract the behavioral saliency matrix The column vectors corresponding to the significant latent components in constitute the significant behavior weight matrix , expressed as ,in, The set of indices representing the significant latent components.

[0087] As an example, is a 100×5 matrix (i.e., 100 brain regions, 5 potential components), and the index set of significant potential components is , indicating that the first and third potential components are significant potential components, then .

[0088] (S102.4) Calculate a GMV score matrix based on the GMV matrix and the significant GMV weight matrix, and calculate a behavior score matrix based on the behavior matrix and the significant behavior weight matrix.

[0089] In the embodiment of the present invention, the GMV score matrix is ​​calculated based on the GMV matrix and the significant GMV weight matrix, and the behavior score matrix is ​​calculated based on the behavior matrix and the significant behavior weight matrix. Specifically, , ,in, represents the GMV score matrix, Each element in Indicates the The subjects in The brain structure projection value on the first significant latent component (i.e., the brain structure characteristics of the subject (such as gray matter volume) and the the degree of matching of potential components), represents the behavior score matrix, Each element in Indicates the The subjects in The behavioral projection value on the significant latent component (i.e., the behavioral characteristics of the subject are related to the the degree of matching of the underlying components).

[0090] Through the above steps (S102.1) to (S102.4), the high-dimensional GMV data and behavioral data are compressed into a low-dimensional saliency space, thereby more accurately capturing the potential brain-behavior patterns and facilitating in-depth exploration of the patient's potential neural phenotype.

[0091] In step S103, based on the GMV score matrix and the behavior score matrix, a mediation analysis of social support factors is performed on the brain-behavior association of the target group to determine the significant mediating path of the brain-behavior association of the target group.

[0092] In this embodiment of the present invention, the GMV score matrix is the independent variable and behavior score matrix The dependent variable, social support factor Using social mediators as mediating variables, mediation analysis was used to verify the social regulation mechanism of brain-behavior associations and identify the significant mediating paths of brain-behavior associations in the target group.

[0093] In a feasible embodiment, the social support factor is a family support score and / or a friend support score, wherein the family support score is obtained through the family support subscale in the social support scale, and the friend support score is obtained through the friend support subscale in the social support scale, which are used to analyze the impact mechanism of social support on brain-behavior association.

[0094] In a feasible embodiment, the following steps are performed to implement the mediation analysis of social support factors on the brain-behavior relationship of the target group:

[0095] (S103.1) Using the GMV score matrix as the independent variable, the behavior score matrix as the dependent variable, and the social support factor as the mediating variable, a mediation model was constructed. The mediation model is: ,in, represents the GMV score matrix, Indicates social support factors, represents the behavior score matrix, represents the first intercept, represents the first residual term, represents the second intercept, represents the second residual term, represents the regression coefficient of the independent variable on the mediating variable, represents the regression coefficient of the mediating variable on the dependent variable, It represents the regression coefficient of the direct effect of the independent variable on the dependent variable after controlling the mediating variable;

[0096] In this embodiment of the present invention, all social support factors 、 and Perform mean centering to eliminate potential influences. 、 and , construct the mediating variable regression equation and the dependent variable regression equation , the mediation model is composed of the mediating variable regression equation and the dependent variable regression equation, among which the mediating variable regression equation is used to test the influence of the independent variable on the mediating variable (path → ), the dependent variable regression equation is used to test the influence of independent variables and mediating variables on the dependent variable (path → and → → ), represents the first intercept, represents the first residual term, represents the second intercept, represents the second residual term, Represents the regression coefficient of the independent variable on the mediating variable (i.e., path → The path coefficient of right The direct effect of Represents the regression coefficient of the mediating variable on the dependent variable (i.e., path → The path coefficient of right The direct effect of It represents the regression coefficient of the direct effect of the independent variable on the dependent variable after controlling the mediating variable, reflecting the direct effect of brain characteristics on behavior.

[0097] (S103.2) Calculate the indirect effect value of social support factors between the GMV score matrix and the behavior score matrix through the mediation model, and perform a significance test on the indirect effect value based on the Bootstrap sampling method. Determine the significant mediating path of the target group's brain-behavior association based on the test results.

[0098] In the embodiment of the present invention, linear regression is first performed on the mediating variable and the independent variable through the mediating variable regression equation to calculate the regression coefficient Then, the dependent variable, mediating variable and independent variable are subjected to multiple linear regression through the dependent variable regression equation to calculate the direct effect and mediation path coefficients , after which, by and Calculate the indirect effect value (IE), that is , and conduct a significance test on the indirect effect value based on the Bootstrap sampling method, and determine the significant mediating path of the association between the brain and behavior of the target group based on the test results. When conducting a significance test on the indirect effect value based on the Bootstrap sampling method, specifically, the original data ( , , ) Perform K random samplings with replacement (usually K=5000), refit the above two regression equations after each sampling, and obtain K groups 、 、 and The estimated value of each coefficient (i.e. 、 、 or ), calculate its bias-corrected confidence interval, if the 95% confidence interval does not contain 0, then the coefficient is considered significant. When significant, it is considered that there is a significant mediation path: → → , combined with The significance of the significant mediation path is determined to determine whether the mediation type of the significant mediation path is a partial mediation path or a complete mediation path. The mediation type is classified as follows:

[0099] when 、 If both are significant, the mediation type of the significant mediation path is determined to be a partial mediation path, that is, right The impact is partly through Transfer (i.e. indirect path → → ), partly direct action (i.e. direct path → );

[0100] when Significant, If it is not significant, the mediation type of the significant mediation path is determined to be a complete mediation path, that is, right The impact is entirely through Transfer, no direct path;

[0101] In addition, when If significant, confirm → The path exists, when If significant, confirm → The path exists, when and When at least one is not significant, regardless of Whether the effect is significant or not, there is no effective mediation path.

[0102] The identification of significant mediating pathways of the target group's brain-behavior associations through the above steps (S103.1) to (S103.2) will help explore the impact of multidimensional social support factors on potential brain-behavior patterns.

[0103] In a feasible embodiment, social support factors are used as moderator variables, and a regression model including interaction terms between the moderator variables and the independent variables is constructed, which is expressed as , based on the Bootstrap sampling method for the interaction term (i.e. )coefficient Perform a significance test. If the interaction coefficient If the 95% confidence interval does not contain 0, it is determined that social support factors have a significant moderation effect (Moderation), that is, the strength / direction of the relationship between the brain and behavior can be changed through social support factors.

[0104] In an embodiment of the present invention, brain structural image data and behavioral data of a target group are collected, and the brain structural image data and behavioral data are processed separately to obtain a GMV matrix and a behavior matrix, and a latent component analysis of brain-behavior association is performed on the GMV matrix and the behavior matrix to obtain a GMV score matrix and a behavior score matrix. Based on the GMV score matrix and the behavior score matrix, a mediation analysis of social support factors is performed on the brain-behavior association of the target group, and a significant mediating path of the brain-behavior association of the target group is determined, thereby achieving causal mechanism modeling of the correlation between high-dimensional imaging data and behavioral symptoms, significantly improving the statistical power of neuroimaging analysis of mental illness, accurately revealing the intrinsic connection between brain structure and behavior, deeply analyzing the pathogenesis of mental illness, and providing more targeted targets for early intervention and personalized treatment of mental illness.

[0105] Example 2:

[0106] Figure 2 The structure of the device for analyzing the association between brain and behavior of a group of people with mental illnesses provided by the second embodiment of the present invention is shown. For ease of explanation, only the parts related to the embodiment of the present invention are shown, including:

[0107] The target data processing unit 21 is used to collect brain structural image data and behavioral data of the target group, and process the collected brain structural image data and behavioral data to obtain corresponding GMV matrix and behavior matrix;

[0108] A latent component analysis unit 22 is used to perform latent component analysis of brain-behavior correlation on the GMV matrix and the behavior matrix to obtain a GMV score matrix and a behavior score matrix of the target group on significant latent components;

[0109] The mediation effect analysis unit 23 is used to perform mediation analysis of social support factors on the brain-behavior association of the target group based on the GMV score matrix and the behavior score matrix, and determine the significant mediation path of the brain-behavior association of the target group.

[0110] Preferably, the latent component analysis unit 22 includes:

[0111] A covariance decomposition unit is used to perform singular value decomposition on the covariance matrix of the GMV matrix and the behavior matrix to obtain the GMV significance matrix, the behavior significance matrix and the singular value matrix;

[0112] a singular value screening unit, configured to evaluate the statistical significance of each singular value in the singular value matrix through a permutation test, screen out significant singular values ​​that meet a preset significance threshold, and determine the potential components corresponding to the significant singular values ​​as significant potential components;

[0113] A column vector extraction unit is used to extract column vectors corresponding to significant latent components from the GMV significance matrix and the behavior significance matrix respectively, to form a significant GMV weight matrix and a significant behavior weight matrix;

[0114] The score matrix calculation unit is used to calculate the GMV score matrix based on the GMV matrix and the significant GMV weight matrix, and to calculate the behavior score matrix based on the behavior matrix and the significant behavior weight matrix.

[0115] Preferably, the mediation effect analysis unit 23 includes:

[0116] The mediation model construction unit is used to construct a mediation model with the GMV score matrix as the independent variable, the behavior score matrix as the dependent variable, and the social support factor as the mediating variable. The mediation model is ,in, represents the GMV score matrix, Indicates social support factors, represents the behavior score matrix, represents the first intercept, represents the first residual term, represents the second intercept, represents the second residual term, represents the regression coefficient of the independent variable on the mediating variable, represents the regression coefficient of the mediating variable on the dependent variable, It represents the regression coefficient of the direct effect of the independent variable on the dependent variable after controlling the mediating variable;

[0117] The mediation effect test unit is used to calculate the indirect effect value of the social support factor between the GMV score matrix and the behavior score matrix through the mediation model, and perform a significance test on the indirect effect value based on the Bootstrap sampling method. According to the test results, the significant mediation path of the target group's brain and behavior association is determined.

[0118] In the embodiments of the present invention, for the convenience and brevity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In actual applications, the above-mentioned functional distribution can be implemented by different functional units and modules as needed, that is, the internal structure of the device is divided into different functional units or modules to implement all or part of the functions described above. The various units and modules of the device can be implemented by corresponding hardware or software units. Each unit and module can be an independent software or hardware unit, or can be integrated into a software or hardware unit, which is not intended to limit the present invention. In addition, the specific names of the various functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the device can refer to the corresponding description in the aforementioned method embodiment, which will not be repeated here.

[0119] Example 3:

[0120] Figure 3 The structure of a computing device provided by the third embodiment of the present invention is shown. For ease of description, only the parts related to the embodiment of the present invention are shown.

[0121] The computing device 3 of the embodiment of the present invention includes a processor 30, a memory 31, and a computer program 32 stored in the memory 31 and executable on the processor 30. When the processor 30 executes the computer program 32, the steps of the embodiment of the method for analyzing the association between brain and behavior of a group of mental illnesses are implemented, such as Figure 1 Alternatively, when the processor 30 executes the computer program 32, the functions of each unit in the above-mentioned device embodiments are realized, for example Figure 2 Function of the unit shown.

[0122] In an embodiment of the present invention, brain structural image data and behavioral data of a target group are collected, and the collected brain structural image data and behavioral data are processed separately to obtain corresponding GMV matrices and behavioral matrices, and a latent component analysis of brain-behavior association is performed on the GMV matrix and the behavior matrix to obtain a GMV score matrix and a behavior score matrix of the target group on significant latent components. Based on the GMV score matrix and the behavior score matrix, a mediation analysis of social support factors is performed on the brain-behavior association of the target group, and a significant mediating path of the brain-behavior association of the target group is determined, thereby significantly improving the statistical power of neuroimaging analysis of mental illness, accurately revealing the intrinsic connection between brain structure and behavior, deeply analyzing the pathogenesis of mental illness, and providing more targeted targets for early intervention and personalized treatment of mental illness.

[0123] The computing device of the embodiment of the present invention may be a personal computer. The steps implemented when the processor 30 of the computing device 3 executes the computer program 32 to implement the method for analyzing the association between brain and behavior of a group of mental illnesses can be referred to the description of the aforementioned method embodiment and will not be repeated here.

[0124] Example 4:

[0125] In an embodiment of the present invention, a computer-readable storage medium is provided, which stores a computer program. When the computer program is executed by a processor, the steps in the embodiment of the above-mentioned method for analyzing the association between brain and behavior of a group of mental illnesses are implemented, for example, Figure 1 Alternatively, when the computer program is executed by a processor, the functions of each unit in the above-mentioned device embodiments are realized, for example Figure 2 Function of the unit shown.

[0126] In an embodiment of the present invention, brain structural image data and behavioral data of a target group are collected, and the collected brain structural image data and behavioral data are processed separately to obtain corresponding GMV matrices and behavioral matrices, and a latent component analysis of brain-behavior association is performed on the GMV matrix and the behavior matrix to obtain a GMV score matrix and a behavior score matrix of the target group on significant latent components. Based on the GMV score matrix and the behavior score matrix, a mediation analysis of social support factors is performed on the brain-behavior association of the target group, and a significant mediating path of the brain-behavior association of the target group is determined, thereby significantly improving the statistical power of neuroimaging analysis of mental illness, accurately revealing the intrinsic connection between brain structure and behavior, deeply analyzing the pathogenesis of mental illness, and providing more targeted targets for early intervention and personalized treatment of mental illness.

[0127] The computer-readable storage medium of the embodiments of the present invention may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to, an electrical connection having one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EEPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the embodiments of the present invention, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0128] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the scope of disclosure involved in the above embodiments is not limited to the technical solutions formed by the specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above disclosed concepts. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

[0129] In addition, although adopting specific order to describe each operation, this should not be interpreted as requiring these operations to be executed in the specific order shown or in sequential order.Under certain environment, multitasking and parallel processing may be advantageous.Similarly, although comprising some specific implementation details in the above discussion, these should not be interpreted as limiting the scope of the present invention.Some features described in the context of independent embodiment can also be implemented in single embodiment in combination.On the contrary, the various features described in the context of independent embodiment also can be implemented in multiple embodiments individually or in the mode of any suitable subcombination.

Claims

1. A method for analyzing the association between brain and behavior in a group of patients with mental illness, characterized by: The method comprises the following steps: Collecting brain structural image data and behavioral data of the target group, and processing the collected brain structural image data and behavioral data respectively to obtain corresponding GMV matrix and behavior matrix; Performing a latent component analysis of brain-behavior association on the GMV matrix and the behavior matrix to obtain a GMV score matrix and a behavior score matrix of the target group on significant latent components; Based on the GMV score matrix and the behavior score matrix, a mediation analysis of social support factors is performed on the brain-behavior association of the target group to determine the significant mediating path of the brain-behavior association of the target group; The step of performing latent component analysis of brain-behavior association on the GMV matrix and the behavior matrix includes: Performing singular value decomposition on the covariance matrix of the GMV matrix and the behavior matrix to obtain a GMV significance matrix, a behavior significance matrix, and a singular value matrix; Evaluating the statistical significance of each singular value in the singular value matrix through a permutation test, screening out significant singular values ​​that meet a preset significance threshold, and determining the potential components corresponding to the significant singular values ​​as the significant potential components; Extracting column vectors corresponding to the significant latent components from the GMV significance matrix and the behavior significance matrix respectively to form a significant GMV weight matrix and a significant behavior weight matrix; Calculate the GMV score matrix according to the GMV matrix and the significant GMV weight matrix, and calculate the behavior score matrix according to the behavior matrix and the significant behavior weight matrix; The step of performing a mediation analysis of social support factors on the brain-behavior association of the target group based on the GMV score matrix and the behavior score matrix includes: A mediation model was constructed with the GMV score matrix as the independent variable, the behavior score matrix as the dependent variable, and the social support factor as the mediating variable. The mediation model is: ,in, represents the GMV score matrix, represents the social support factors, represents the behavior score matrix, represents the first intercept, represents the first residual term, represents the second intercept, represents the second residual term, represents the regression coefficient of the independent variable on the mediating variable, represents the regression coefficient of the mediating variable on the dependent variable, It represents the regression coefficient of the direct effect of the independent variable on the dependent variable after controlling the mediating variable; Calculating the indirect effect value of the social support factor between the GMV score matrix and the behavior score matrix through the mediation model, performing a significance test on the indirect effect value based on the Bootstrap sampling method, and determining the significant mediation path of the association between the brain and behavior of the target group according to the test results; The steps of processing the collected brain structure image data and the collected behavioral data respectively include: The brain structural image data are preprocessed using voxel-based morphological analysis to obtain whole-brain GMV data; The whole-brain GMV data and the behavioral data were respectively subjected to Z-score normalization to obtain the GMV matrix and the behavioral matrix.

2. The method according to claim 1, wherein The social support factor is a family support score and / or a friend support score, wherein the family support score is obtained through the family support subscale in the social support scale, and the friend support score is obtained through the friend support subscale in the social support scale.

3. A device for analyzing the correlation between brain and behavior of a group of people with mental illness, characterized by: The device comprises: a target data processing unit, configured to collect brain structural image data and behavioral data of a target group, and process the collected brain structural image data and behavioral data respectively to obtain corresponding GMV matrices and behavioral matrices, including: preprocessing the brain structural image data using voxel-based morphological analysis to obtain whole-brain GMV data; and performing Z-score normalization on the whole-brain GMV data and the behavioral data respectively to obtain the GMV matrix and the behavioral matrix; a latent component analysis unit, configured to perform a latent component analysis of brain-behavior association on the GMV matrix and the behavior matrix to obtain a GMV score matrix and a behavior score matrix of the target group on significant latent components; a mediation effect analysis unit, configured to perform a mediation analysis of social support factors on the brain-behavior association of the target group based on the GMV score matrix and the behavior score matrix, and determine a significant mediation path of the brain-behavior association of the target group; Wherein, the potential component analysis unit includes: a covariance decomposition unit, configured to perform singular value decomposition on the covariance matrix of the GMV matrix and the behavior matrix to obtain a GMV significance matrix, a behavior significance matrix, and a singular value matrix; a singular value screening unit, configured to evaluate the statistical significance of each singular value in the singular value matrix through a permutation test, screen out significant singular values ​​that meet a preset significance threshold, and determine the potential components corresponding to the significant singular values ​​as the significant potential components; a column vector extraction unit, configured to extract column vectors corresponding to the significant latent components from the GMV significance matrix and the behavior significance matrix, respectively, to form a significant GMV weight matrix and a significant behavior weight matrix; a score matrix calculation unit, configured to calculate the GMV score matrix based on the GMV matrix and the significant GMV weight matrix, and calculate the behavior score matrix based on the behavior matrix and the significant behavior weight matrix; The mediation effect analysis unit includes: The mediation model construction unit is used to construct a mediation model with the GMV score matrix as the independent variable, the behavior score matrix as the dependent variable, and the social support factor as the mediating variable. The mediation model is ,in, represents the GMV score matrix, represents the social support factors, represents the behavior score matrix, represents the first intercept, represents the first residual term, represents the second intercept, represents the second residual term, represents the regression coefficient of the independent variable on the mediating variable, represents the regression coefficient of the mediating variable on the dependent variable, It represents the regression coefficient of the direct effect of the independent variable on the dependent variable after controlling the mediating variable; A mediation effect testing unit is used to calculate the indirect effect value of the social support factor between the GMV score matrix and the behavior score matrix through the mediation model, and perform a significance test on the indirect effect value based on the Bootstrap sampling method, and determine the significant mediation path of the target group's brain and behavior association based on the test results.

4. The device according to claim 3, wherein The social support factor is a family support score and / or a friend support score, wherein the family support score is obtained through the family support subscale in the social support scale, and the friend support score is obtained through the friend support subscale in the social support scale.

5. A computing device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 2 are implemented.

6. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 2 are implemented.

Images