Construction method and application of full life cycle growth model of human cerebral cortex form similar network

Abstract

The invention relates to a method for constructing a growth model of a brain in a life cycle, and the method comprises the steps: obtaining brain MRI images of individuals of different age groups, and dividing the cerebral cortex into n regions; constructing a morphological feature network for sMRI of the brain MRI image of the individual, and performing morphological feature similarity degree analysis to obtain a first phenotype; constructing a functional network for the fMRI of the brain MRI image of the individual; constructing a structure-function coupling matrix for the morphological feature network and the function network, and calculating to obtain a second phenotype; selecting the first phenotype and the second phenotype of each individual as dependent variables and the age of each individual as a smooth item to carry out GAMLSS modeling, and selecting JSU distribution to respectively fit the change of the first phenotype and the second phenotype along with the age to generate a reference growth curve. The method can be used for studying the growth mode of the brain network in the whole life cycle, studying life cycle association between morphological characteristics and functional characteristics of different areas and the like.

Description

Technical Field

[0001] This application relates to the technical field of magnetic resonance brain image analysis, and particularly to a method for constructing a full-life-cycle growth model of a human cerebral cortex morphological similarity network and its application. Background Art

[0002] The morphological characteristics of the cerebral cortex structure (such as cortical thickness, surface area, and cortical folding) undergo a complex, genetically regulated, and regionally heterogeneous growth process, representing the basic structural framework of the cerebral cortex throughout the entire human life cycle.

[0003] More and more studies have shown that cerebral cortex growth is not limited to a single region (the brain is divided into multiple brain regions, and here the region refers to the brain region), but shows structural covariation between regions, forming an interconnected network. Based on this, there are currently some network models for studying different regions of the cerebral cortex, such as the morphological similarity network (MSN), the morphometric inverse divergence (MIND) network, etc., which can be used to study the correlation between different regions of the brain.

[0004] Most of the current studies on the relationship between different regions of the brain are limited to a narrow age window, such as the fetal period, neonatal period, childhood and adolescence, and adulthood, and rarely involve the study of brain networks throughout the entire life cycle.

[0005] Based on this, this application provides a solution for constructing a full-life-cycle growth model of a human cerebral cortex morphological similarity network and its application, which can be used for studying the growth patterns of the brain network at the whole-brain level, regional level, cytoarchitectonic class, etc. throughout the entire life cycle, and for studying the life-cycle correlation between the morphological and functional characteristics of different regions throughout the entire life cycle of the brain network, etc., which can help better understand the life-cycle development and degradation patterns of the brain structural network and already has potential analytical value for evaluating brain health. Summary of the Invention

[0006] In view of the above problems in the prior art, this application provides a method for constructing a full-life-cycle growth model of a human cerebral cortex morphological similarity network and its application, which is used for studying the morphological and functional characteristics of the brain network throughout the entire life cycle.

[0007] The first aspect of this application provides a method for constructing a full-life-cycle growth model of a human cerebral cortex morphological similarity network, including:

[0008] Obtain brain MRI images of individuals of different ages and perform quality control operations;

[0009] Based on the brain MRI images of different individuals after quality control operations, divide the brain MRI images into n regions in the cerebral cortex;

[0010] For each region of the structural image of the brain MRI image of each individual, m cortical morphological features are determined to form m eigenvalues; according to the m eigenvalues included in n regions, a morphological feature network of the individual's brain network is constructed, where the morphological feature network is represented as an n×n matrix, and each element value in the matrix is obtained by performing a similarity calculation on the eigenvalues of the corresponding two regions; a morphological feature similarity degree analysis and calculation is performed on the morphological feature network of the individual's brain network to obtain at least one first phenotype;

[0011] For each region of the functional image of the brain MRI image of each individual, the time series of the blood oxygenation level-dependent signal of the region is determined and used as an eigenvalue; according to the respective eigenvalues included in n regions, a functional network of the individual's brain network is constructed, where the functional network is represented as an n×n matrix, and each element value in the matrix is obtained by performing a correlation calculation on the eigenvalues of the corresponding two regions;

[0012] For the morphological feature network and the functional network of an individual, a structure-functional coupling matrix is constructed, and the value of each element of the structure-functional coupling matrix is obtained by performing a correlation calculation on the elements at the corresponding positions of the morphological feature network and the functional network; at least one second phenotype is obtained by calculating the structure-functional coupling matrix

[0013] Select at least one first phenotype and at least one second phenotype of each individual as dependent variables respectively, use the age of each individual as a smoothing term for GAMLSS modeling, and select the JSU distribution to respectively fit the age-related changes of the first phenotype and the second phenotype to generate a reference growth curve.

[0014] As a possible implementation manner of the first aspect, the quality control operation includes at least one of the following: performing quality assessment based on standard quality assessment criteria to exclude problematic MRI images; for the MRI images of individuals of different ages, selecting different pipelines for preprocessing the MRI images, including preprocessing of the structural image and preprocessing of the functional image; adopting a surface quality control method to evaluate the Euler number for the structural image, and selecting MRI images within a conforming range based on the Euler number; screening MRI images for the functional image based on head motion control.

[0015] As a possible implementation manner of the first aspect, the determining m cortical morphological features to form m eigenvalues includes: determining five cortical morphological features, including: gray matter volume, cortical thickness, surface area, mean curvature, and sulcus depth; respectively performing standardization processing on each morphological feature to form five eigenvalues for each region.

[0016] As a possible implementation of the first aspect, the morphological feature network includes a MIND network; each element value in the matrix of the morphological feature network is obtained by calculating the similarity of the feature values of the corresponding two regions, including: each element value in the MIND network is obtained by calculating the multivariate KL divergence between the five feature values of the corresponding two regions.

[0017] As a possible implementation of the first aspect, at least one first phenotype includes: whole-brain level connection parameter values, including: whole-brain network mean or whole-brain network variance value; class-level connection parameter values, including: within-class connection parameter value or between-class connection parameter value; region-level connection parameter values; at least one first phenotype further includes the modularity index Q or small-world property metric parameter obtained by performing graph-theoretic metric analysis on the morphological feature network.

[0018] As a possible implementation of the first aspect, each element value in the matrix of the functional network is obtained by calculating the correlation of the feature values of the corresponding two regions, including: each element value in the matrix of the functional network is obtained by calculating the correlation using the Pearson correlation coefficient for the blood oxygenation level-dependent signal sequences of the corresponding two regions.

[0019] As a possible implementation of the first aspect, at least one second phenotype includes: the mean, variance, median, or standard deviation of the calculated structure-functional coupling matrix.

[0020] As a possible implementation of the first aspect, the GAMLSS modeling further includes: the gender and Euler number of each individual as covariates, and the site as a random effect; wherein, the Euler number is obtained by evaluating the structural image of the brain MRI image, and the site corresponds to the source location of the brain MRI image of the individual.

[0021] The second aspect of the present application provides an application of a full-life cycle growth model of a human cerebral cortex morphological similarity network, the growth model is constructed according to the construction method of the full-life cycle growth model of the human cerebral cortex morphological similarity network according to any one of the first aspect, and the application includes at least one of the following: applied to the analysis of the age-related trend of the morphological network of the cerebral cortex; applied to the analysis of the different patterns shown by the structural network of the cerebral cortex between the sensory region and the paralimbic region with age; applied to the analysis of the impact of the structural changes of the cerebral cortex on the function; applied to the individual heterogeneity analysis of the brain networks of patients with Alzheimer's disease, major depressive disorder or autism spectrum disorder.

[0022] A third aspect of the present application provides a computing device, including: a processor, and a memory storing program instructions thereon, and when the program instructions are executed by the processor, the processor is caused to execute the method for constructing the full-life-cycle growth model of the human cerebral cortex morphological similarity network according to any one of the first aspect.

[0023] As described above, the present application provides a solution for constructing the full-life-cycle growth model of the human cerebral cortex morphological similarity network, which can be used for the research on the growth patterns of the brain network at the whole-brain level, regional level, cytoarchitectonic class, etc. during the entire life cycle, and for the research on the life-cycle association between morphological features (corresponding to the MIND network) and functional features (corresponding to the functional network) of the brain network during the entire life cycle, etc., so as to promote the understanding of human brain growth. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 is a flowchart of the method for constructing the full-life-cycle growth model of the human cerebral cortex morphological similarity network provided by the first embodiment of the present application;

[0025] Figure 2 is Figure 1 a schematic diagram of the construction method shown;

[0026] Figure 3 is a reference growth curve at the whole-brain level and its related schematic diagram provided by the embodiment of the present application;

[0027] Figure 4 is a reference growth curve at the regional level and its related schematic diagram provided by the embodiment of the present application;

[0028] Figure 5 is a reference growth curve of the phenotype related to the structure-function coupling matrix and its related schematic diagram provided by the embodiment of the present application;

[0029] Figure 6 is a structural schematic diagram of a computing device provided by the embodiment of the present application.

[0030] It should be understood that in the above structural schematic diagram, the sizes and shapes of the respective block diagrams are for reference only and should not constitute an exclusive interpretation of the embodiments of the present invention. The relative positions and inclusion relationships between the respective block diagrams presented by the structural schematic diagram only schematically represent the structural associations between the block diagrams, rather than limiting the physical connection manners of the embodiments of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0031] The following is a further description of the technical solutions provided in this application with reference to the accompanying drawings and by way of examples. It should be understood that the system structures and business scenarios provided in the embodiments of this application are mainly for illustrating possible implementation manners of the technical solutions of this application, and should not be construed as the only limitation to the technical solutions of this application. As is known to those of ordinary skill in the art, with the evolution of system structures and the emergence of new business scenarios, the technical solutions provided in this application are equally applicable to similar technical problems.

[0032] It should be understood that the solution for constructing the full-life cycle growth model of the human cerebral cortex morphological similarity network provided in the embodiments of this application includes a method, device, computing device, storage medium, and application for constructing the full-life cycle growth model of the human cerebral cortex morphological similarity network. Since the principles of these technical solutions for solving problems are the same or similar, in the following introduction of specific embodiments, some repetitive parts may not be elaborated again, but it should be regarded that there are mutual references between these specific embodiments and they can be combined with each other.

[0033] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which this application belongs. In case of inconsistency, the meaning described in this specification or the meaning derived from the content recorded in this specification shall prevail. In addition, the terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application. In order to accurately describe the technical content in this application and to accurately understand the present invention, the following explanations or definitions are given for the terms used in this specification before the description of the specific embodiments:

[0034] 1) Magnetic Resonance Imaging (MRI) image: An MRI image refers to medical imaging data obtained through magnetic resonance imaging technology. In this application, MRI applied to the head is used to obtain brain MRI. For the sake of simplicity in description, unless otherwise specified, the MRI images in this application all refer to brain MRI.

[0035] 2) T1 and T2 weighted images in brain MRI: T1 and T2 weighted images are two basic image types in MRI, and they each have their own characteristics in showing tissue structures and lesions. Among them, T1 weighted images are mainly used to show the skeletal structure and anatomical morphology of tissues, and are commonly used to evaluate the morphological structures of organs and tissues. T2 weighted images are mainly used to show the water content and tissue spaces of tissues, and are commonly used to evaluate lesions such as inflammation, edema, and tumors.

[0036] 3) Structural MRI (or Structural MRI) and Functional MRI (or Functional MRI) of MRI.

[0037] Structural MRI (sMRI) and Functional MRI (fMRI) are two major categories in MRI. sMRI is mainly used to display the internal anatomical structures of the human body, especially the detailed information of the brain tissue, spinal cord, joints, muscles, and internal organs. fMRI is mainly used to show the brain activities when performing specific tasks or at rest.

[0038] Task-free fMRI images: These images are obtained when an individual is at rest, so they are called task-free fMRI images, and their purpose is to capture the spontaneous neural activities of the brain.

[0039] 4) Brain research tools:

[0040] HCP (Human Connectome Project) pipeline: A standardized process for processing and analyzing high-resolution neuroimaging data, aiming to promote the research on the human brain connectome.

[0041] iBEAT V2.0 pipeline: A preprocessing pipeline specifically designed for neuroimaging data of early ages (0 - 24 months), which optimizes the preprocessing of infant MRI images and performs excellently in tissue segmentation and cortical reconstruction.

[0042] dHCP (The Developing HCP) pipeline: A set of MRI image preprocessing and analysis processes specifically designed for newborns and infants (especially full-term infants at 37 - 42 weeks after birth), aiming to handle the unique characteristics of newborn MRI images, including high-resolution structural images and functional images.

[0043] 5) Euler number: Used to quantify the topological complexity of a surface. For example, MRI images corresponding to an overly small Euler number (such as too many holes or handles on the cortical surface) can be considered abnormal and excluded.

[0044] 6) Interquartile Range (IQR): A statistic for measuring the dispersion of data, used to describe the distribution of the middle 50% of the data. Defined by calculating the difference between the first quartile (Q1) and the third quartile (Q3), IQR can be used to identify outliers and describe the data distribution.

[0045] 7) fs_LR_32k standard space: A standardized cortical surface template used in HCP to align cortical surface data of different individuals to a common space.

[0046] 8) Desikan-Killiany Atlas: A widely used brain cortical parcellation atlas, initially developed by Desikan and Killiany. This atlas divides the cerebral cortex into multiple anatomical regions, with each region corresponding to specific anatomical structures and functions.

[0047] Randomly Subdivided Desikan-Killiany Atlas: To improve the applicability and flexibility of the atlas, researchers may randomly subdivide the original Desikan-Killiany atlas, such as adding or merging certain regions, or adjusting the boundaries of regions. Random subdivision can increase the diversity and robustness of the atlas, making it better applicable to different types of data and research goals.

[0048] 9) MIND Network (Morphometric INverse Divergence Network): A method for analyzing brain connectivity. By dividing the cerebral cortex into multiple regions, a connection matrix between each region can be constructed, and then the patterns and characteristics of these connections can be analyzed.

[0049] 10) Standardization of Features (z-score): In neuroimaging and brain network research, standardizing (z-score) each feature across all vertices is a common data preprocessing method. The purpose of standardization is to eliminate the scale differences between different features, making the data comparable. The z-score is a standardization method that is achieved by converting each feature value into units relative to the mean and standard deviation of that feature.

[0050] 11) Morphometric Similarity Strength Analysis (MSS): A method for evaluating the morphological similarity between brain cortical regions. By quantifying the similarity between cortical regions, a morphometric connectome can be constructed and its global and local properties can be further analyzed.

[0051] 12) Functional Network: Refers to the functional connections between different regions in the brain. By analyzing the BOLD signal time series of each region, a functional connection matrix can be constructed to reflect the synchrony and interaction between different regions.

[0052] 13) GAMLSS (Generalized Additive Models for Location, Scale and Shape) Model: A generalized additive model used to establish non-linear relationships for response variables.

[0053] The core idea of the GAMLSS model is to decompose the distribution of data into three parts: location, scale, and shape, and model them through a generalized linear model. Among them, the location part represents the mean or median of the variable, the scale part represents the degree of dispersion or standard deviation of the variable, and the shape part represents the distribution shape of the variable. By modeling these three parts, the characteristics of the data can be described more comprehensively.

[0054] 14) JSU distribution: A distribution form in the Johnson distribution family, used to describe the skewness and kurtosis of data, and can be applied to the modeling of distributions related to time series.

[0055] 15) Reference growth curve: Refers to the normal development trend of certain phenotypes with age in the healthy population. Among them, the phenotype usually refers to the measurement parameters extracted from imaging data that can reflect the functional, structural, or morphological characteristics of the brain.

[0056] By establishing a reference growth curve, the measured values of an individual can be compared with the reference growth curve to identify possible abnormalities or deviations from the normal development trajectory. It can also be used to understand the normal development range at different ages, evaluate the developmental status of an individual, and can also be applied in clinical research to compare the phenotype curves of patients and healthy control groups and explore the potential mechanisms and impacts of diseases.

[0057] This application provides a construction scheme for the full-life-cycle growth model of the human cerebral cortex morphological similarity network, which can be used for the study of the growth patterns of the brain network at the whole-brain level, regional level, cytoarchitectonic class, etc. during the entire life cycle, and for the study of the life-cycle association between morphological characteristics (corresponding to the MIND network) and functional characteristics (corresponding to the functional network) of the brain network during the entire life cycle, etc., to promote the understanding of human brain growth. Based on the reference growth model (i.e., the reference growth curve) of the brain network obtained from the solution of this application, the applicant further obtained the following research conclusions:

[0058] 1) From birth to early adulthood, the morphological network of the cerebral cortex gradually becomes more modular and the organizational structure becomes more efficient. This means that as humans age, different regions of the brain begin to divide labor and cooperate more clearly, forming an efficient communication network.

[0059] 2) The growth of the structural network in the cerebral cortex during the life cycle shows different patterns between the sensory regions and the paralimbic regions: for example, the similarity in the sensory regions is decreasing, while the similarity in the cingulate gyrus and insular regions is increasing. The gradual decline in morphological similarity in the sensory regions with age may reflect the gradual functional specialization and optimization of the sensory system during maturation. The gradual increase in morphological similarity in the cingulate gyrus and insular regions with age may be related to the gradual specialization and optimization of these regions in emotional regulation and cognitive functions.

[0060] 3) The life cycle growth of the morphological network is closely related to the functional network, especially in the sensory regions. This indicates a tight connection between structure and function, and changes in structure may affect the performance of function.

[0061] 4) The growth model based on the brain network reveals significant individual heterogeneity in patients with Alzheimer's disease, major depressive disorder, and autism spectrum disorder. This provides an important reference for understanding the pathological mechanisms of these diseases and developing personalized treatment plans.

[0062] Next, the embodiments provided by the present application will be introduced in detail with reference to the respective drawings.

[0063] The first embodiment of the present application provides a method for constructing a full-life cycle growth model of a morphological similarity network of the human cerebral cortex. Referring to Figure 1 and Figure 2 the flow chart and schematic diagram shown, this method includes the following steps S1 - S12:

[0064] S1: Obtain brain MRI images of individuals of different ages and perform quality control (QC) operations. In some embodiments, this step S1 may include the following sub-steps S110 - S120:

[0065] S110: Obtain the original MRI images of individuals of different ages from multiple sites.

[0066] In this example, the original MRI images are from 183 sites, and the total number of individual samples is 45,833, with ages ranging from 0 to 80 years old.

[0067] S120: Perform quality control operations on the obtained original MRI images, including: removing data that does not meet the requirements, screening to obtain valid MRI images, and performing preprocessing of the MRI images. In some embodiments, this step S120 may include the following sub-steps S121 - S124:

[0068] S121: Perform a preliminary quality assessment of the original MRI images to exclude MRI images with obvious acquisition problems. Specifically, it may include the following:

[0069] a) For datasets such as dHCP (Developing Human Connectome Project), HCP-Development, HCP-Aging, HCP-Young Adult, and ABCD (Adolescent Brain Cognitive Development), these datasets provide standard quality assessment guidelines, and preliminary quality control can be carried out according to the inclusion criteria recommended by these datasets.

[0070] For example, in the HCP-Young Adult dataset, if the signal-to-noise ratio (SNR) of a T1-weighted image is below the recommended threshold, then the image will be excluded.

[0071] b) For the BCP (Brain Connectome Project) dataset, manual review can be performed by experienced pediatric neuroradiologists.

[0072] For example, if two (or a specified number of) radiologists both believe that there are obvious motion artifacts or signal inhomogeneity in an image, then the image will be excluded.

[0073] c) For other datasets that do not provide standard quality assessment guidelines, automated assessment is performed using the MRI Quality Control (MRIQC) tool to generate quality metrics for the structural and functional images of each individual sample, and quality control is carried out based on these metrics.

[0074] Among them, the quality metrics for structural images include: Entropy Focus Criterion (EFC), Foreground-Background Energy Ratio (FBER), Combined Coefficient of Variation (CJV), Contrast-to-Noise Ratio (CNR), Signal-to-Noise Ratio (SNR), Dietrich's Signal-to-Noise Ratio (SNRd).

[0075] Among them, the quality metrics for functional images include: AFNI's Outlier Ratio (AOR), AFNI's Quality Index (AQI), DVARS Standard Deviation (DVARS_std), DVARS Variable Standard Deviation (DVARS_vstd), Signal-to-Noise Ratio (SNR), Temporal Signal-to-Noise Ratio (tSNR).

[0076] Among them, the assessment method includes: if the structural image of an individual is identified as an outlier in three or more of the quality metrics for structural images (for example, exceeding 1.5 times the interquartile range IQR), or if the functional image of the individual is identified as an outlier in three or more of the quality metrics for functional images, then the image will be excluded.

[0077] S122: For the MRI images after preliminary quality assessment, preprocessing is performed, and this preprocessing includes:

[0078] a) MRI images of individuals of different ages are preprocessed using different pipelines, including:

[0079] MRI images of individuals two years old and above: Preprocess using the HCP pipeline;

[0080] MRI images of individuals aged 0 - 24 months: Preprocess using the iBEAT V2.0 pipeline to ensure high-quality cortical reconstruction;

[0081] MRI images of 0-year-old individuals (full-term infants with a gestational age of 37 - 42 weeks, defined as 0 years): Process using the dHCP pipeline for neonates.

[0082] b) Among them, different pipelines preprocess MRI images, including preprocessing of structural images and functional images. The following is an example:

[0083] Preprocessing of structural images includes: brain extraction, denoising, and bias field correction; generating cortical surfaces, including segmenting brain tissue, reconstructing cortical surfaces, and aligning these surfaces with a standard template; converting the processed data to the HCP format (CIFTI), aligning the volume data to the MNI standard space, and mapping the surface data to the fs_LR_32k standard space.

[0084] Preprocessing of functional images includes: interslice time correction, motion correction, and geometric distortion correction; mapping the volume data to the fs_LR_32k standard space.

[0085] S123: After preprocessing, further quality control is performed on the structural images (sMRI) and functional images (fMRI) of individual MRIs, which may specifically include the following:

[0086] a) For structural images, the Surface Quality Control method is used to evaluate the Euler number, and individual MRI images are screened based on the Euler number. For example, if the Euler number of an individual MRI image is lower than 1.5 times the interquartile range (IQR) of the scan site, that is, lower than the research range, then the MRI image will be excluded. Additionally, all MRI images with an Euler number lower than a certain threshold (such as -217) will be excluded.

[0087] b) For functional images, head motion control is a key step to ensure data quality. Excessive head motion can lead to image blurring and signal distortion. Therefore, individual MRI images are screened by evaluating the average frame shift, the proportion of frames exceeding the threshold, or the number of available time points (the final time point of the scan is less than the threshold number), and valid MRI images are retained.

[0088] S124: Additionally, it may further include visual inspection by an expert group. For example, multiple experts trained in anatomy can be formed into an expert group to visually inspect MRI images, record and discuss the problems found, and screen the MRI images.

[0089] In this example, after the quality control in step 120, the effective MRI images obtained are: a total of 33,937 individual samples from 141 sites, and the age distribution of the individual samples is between 0 and 80 years old. As shown in Figure 2 a in Figure 2 in a, the horizontal axis represents the names of each site, the vertical axis represents the age distribution of the subjects, and the age distribution of individuals at each site is shown. Among them, these 33,937 samples have high-quality structural images and can be used to construct a lifespan growth model of the morphometric network; 32,887 sub-samples of these 33,937 samples also have high-quality functional images and can be used to study the structural and functional images of the structure-function coupling throughout the life cycle.

[0090] In addition, MRI images of 1,202 patients have been screened out from the original MRI images, including 180 patients with Alzheimer's disease (AD), 622 patients with major depressive disorder (MDD), and 400 patients with autism spectrum disorder (ASD). The MRI images of these patients can be used to verify the clinical relevance of the network-based reference growth curve.

[0091] Through sufficient individual samples from different sites and different age groups, the research on the human brain throughout the complete life cycle is realized.

[0092] S2: Based on the brain MRI images of different individuals obtained after quality control, and based on the randomly subdivided Desikan-Killiany atlas, the brain MRI images are divided into 318 cortical regions in the cerebral cortex.

[0093] Among them, 219 of these 318 regions are used for verification. The effectiveness of these 219 regions can be verified through independent datasets or methods to ensure the reliability of dividing 318 cortical regions based on the randomly subdivided Desikan-Killiany atlas.

[0094] In some other embodiments, the division of cortical regions of the brain can also be based on other criteria. For example, the brain can be divided into 100 to 1,000 regions based on the Schaefer atlas, 360 regions based on the Glasser atlas, 180 regions based on the HCP-MMP1.0 atlas, and other division methods.

[0095] S3: For each of the 318 regions of the structural image (sNRI) of each individual, five cortical morphological features of that region are determined, including: gray matter volume (Vol), cortical thickness (CT), surface area (SA), mean curvature (MC), and sulcal depth (SD), and the features are respectively normalized (Z-score) to form 5 eigenvalue corresponding to each region. It is not difficult to understand that eigenvalues of more or fewer morphological features can also be used.

[0096] S4: For the MRI of each individual, based on the 318 regions, a MIND network of the individual is constructed based on the multivariate KL (Kullback-Leibler) divergence between each two regions. The elements in the MIND network matrix represent the MIND value between a pair of regions, reflecting the degree of similarity in morphological features between these two regions. MIND (the degree of similarity in morphological features between two regions) can be calculated as follows:

[0097]

[0098] where a and b respectively represent the eigenvalue of any two regions in the 318 regions of the individual's MRI. Figure 2 For example, ROIa (region of interest a) and ROIb (region of interest b) are used to represent. For example, a is (Vola, CTa, SAa, Sda, MCa) after normalization, and b is (Volb, CTb, SAb, SDb, MCb) after normalization. MIND(a, b) represents the value of the degree of similarity in morphological features between region a and region b.

[0099] In some other embodiments, the values of the elements in the MIND network, in addition to being calculated by the above KL method, can also be obtained by other methods. For example: based on the above 5 eigenvalues of two regions, calculate the Euclidean Distance, Manhattan Distance or Mahalanobis Distance between the two regions, etc.

[0100] S5: According to the obtained morphological similarity network of each individual (in this example, the above MIND network), perform morphological feature similarity degree analysis (MSS) to obtain the global level connectivity parameter value, class level connectivity parameter value, and regional level connectivity parameter value corresponding to the MIND network of the individual. These parameters or the sub-parameters included can be called phenotypes. Figure 2 b in the figure shows these phenotypes, and these parameters are described as follows:

[0101] Whole-brain level connection parameter values, including: the whole-brain network mean value, obtained by calculating the average of all parameter values of the MIND network; the whole-brain network variance value, obtained by calculating the variance of all parameter values of the MIND network;

[0102] Class-level connection parameter values, including: the intra-class connectivity parameter value, which refers to the average of the morphological similarities between all regions within the same cytoarchitectonic class, obtained by calculating the average of the morphological similarities between all region pairs within this cytoarchitectonic class; the inter-class connectivity parameter value, which refers to the average of the morphological similarities between all regions between different cytoarchitectonic classes. For each cytoarchitectonic class, it is obtained by calculating the average of the morphological similarities between all region pairs between this class and other classes. Among them, the way of dividing the regions of the cerebral cortex into different cytoarchitectonic classes can be based on cytoarchitectonic characteristics (such as the von Economo atlas) for division.

[0103] The regional-level connection parameter value refers to the average of the morphological similarities between each region and all other regions. For each region, it is obtained by calculating the average of the morphological similarities between this region and all other regions.

[0104] It should be noted that it is not limited to the phenotypes obtained by the calculation methods mentioned above. For example, in addition to the above mean and variance calculation methods, other methods including standard deviation, median, mode, etc. can also be used to count the corresponding phenotypes.

[0105] S6: Based on the obtained MIND network of each individual, perform graph-theoretic metric analysis to obtain the modular structure and small-world characteristics. The graph-theoretic metric analysis can include the following steps:

[0106] S601: For the current individual's MIND network, perform binarization. For example, a 10% density threshold can be applied to convert the MIND network into a binary network (i.e., a network composed of 0 and 1);

[0107] S602: Based on the binary network, calculate the modularity index Q, which reflects the ability of the network to be divided into several modules.

[0108] S603: Based on the binary network, perform small-world characteristic analysis, including: calculating the clustering coefficient and the shortest path length of the binary network, and further calculating the small-world characteristic metric parameters Gamma, Lamda, and Sigma, which are used to evaluate the small-world characteristics of the network. Figure 2b in it shows Gamma, Lamda, and Sigma of the Q and the small-world characteristic metric parameters.

[0109] Among them, graph theory metric analysis can be implemented using GRETNA software (GRETNA software is a graph theory-based complex brain network analysis software developed for magnetic resonance imaging data and has graph theory metric analysis functions).

[0110] S7: Among all individual MRI images, including the resting-state fMRI images, in this example, these fMRI images are the MRI images of individuals with high-quality functional images (for example, the 32,887 sub-samples with high-quality functional images obtained in step S1). For each of the 318 regions of each fMRI image of these individuals, the time series of the blood oxygenation level-dependent (BOLD) signal of that region (abbreviated as BOLD signal) is determined as the eigenvalue.

[0111] S8: For the fMRI of each individual, based on the 318 regions, the correlation of the BOLD signals between each two regions is calculated to obtain a functional network. Each element of this functional network represents the correlation of the BOLD signals of a pair of regions, that is, it represents the synchronization between different regions.

[0112] Specifically, the calculation method for each element of the functional network can be calculated using calculation methods such as Pearson correlation coefficient, Mutual Information, Coherence, Phase Locking Value (PLV), etc. Taking the calculation method of Pearson correlation coefficient as an example, the method is as follows: Extract the BOLD time series from each region of the fMRI image, use the Pearson correlation coefficient to calculate the correlation between the BOLD time series of two regions, and based on this, calculate the correlation values between all pairs of regions and fill them into a matrix to form a functional network.

[0113] In some other embodiments, the construction of the functional network can be based on other data in addition to being based on BOLD data. For example, it can be based on the data of the blood flow and oxygenation levels in each region of the cerebral cortex detected by near-infrared spectroscopy imaging. The data in each region obtained using techniques that can record brain electrical activity (EEG) and magnetic field activity (MEG).

[0114] S9: Obtain the respective MIND networks corresponding to these individuals with fMRI (already calculated in step S4 above); for each of these individuals, calculate the correlation value between the corresponding elements of its functional network and the MIND network, and use it as the value of the corresponding element in the structure-functional coupling matrix, thereby constructing the structure-functional coupling matrix for this individual.

[0115] Among them, for the calculation of the correlation value between two corresponding elements of the functional network and the MIND network, the Pearson correlation coefficient can be used for calculation (for example Figure 2 the value R = 0.58 obtained by calculating the Pearson correlation coefficient of the corresponding two elements is shown in c of

[0116] S10: According to the structure-functional coupling matrix of each individual, calculate the whole-brain level structure-functional coupling phenotype value. This phenotype value can be calculated by means of calculating the average value, variance value, or median, standard deviation, etc. of the structure-functional coupling matrix. The algorithm can refer to the algorithm for calculating the whole-brain level connection parameter value mentioned in step S5 above, and will not be elaborated here.

[0117] In some other embodiments, for the structure-functional coupling matrix, in addition to the above-mentioned whole-brain level structure-functional coupling phenotype value (corresponding to whole-brain level connection), the class-level structure-functional coupling phenotype value and the region-level structure-functional coupling phenotype value can also be calculated according to the methods for calculating the class-level connection parameter value and the region-level connection parameter value mentioned in step S5, and will not be elaborated here.

[0118] S11: Based on the phenotypes obtained in the above steps S5 and S10, select the target phenotypes. The three phenotypes selected and used in this embodiment are: the whole-brain network mean value, the whole-brain network variance value, and the whole-brain level structure-functional coupling phenotype value in the whole-brain level connection parameter value. Based on these phenotypes at the whole-brain level, the characteristics of the whole brain can be reflected, and can be used to study the changes in the whole-brain characteristics of the brain network with age.

[0119] Among them, the selected phenotypes may also include the class-level connection parameter values, such as the within-class connection and between-class connection parameter values obtained in step S5 above, or the class-level structure-functional coupling phenotype value mentioned in step S10. These phenotypes can reflect the characteristics of different cytoarchitectonic classes in the brain network and can be used to study the changes in different cytoarchitectonic classes of the brain with age.

[0120] Among them, the selected phenotypes may also include the region-level connection parameter values obtained in step S5 above, or the region-level structure-functional coupling phenotype value mentioned in step S10. These phenotypes can reflect the characteristics at the region level of the brain network (such as between regions) and can be used to study the changes at the region level of the brain network with age.

[0121] Among them, the selected phenotypes can also include the modularity index Q and the small-world property metric parameter obtained in step S6 above. The modularity index Q can reflect the degree of modularity of the brain network divided into each module, and the small-world property can reflect the local connection density and path of the nodes in the brain network, reflecting the ability in information transmission and processing. These phenotypes can be used to study the changes in the functional partitioning (modularity) of the brain network and the changes in the brain communication network with age.

[0122] S12: Based on the selected target phenotypes in this embodiment, perform GAMLSS modeling to establish a reference growth curve (normative), and select the Johnson's Su distribution (JSU distribution) to respectively fit the age-related changes of each selected target phenotype, generating a reference growth curve (or called a growth model) for each target phenotype. The reference growth curve refers to the normal development trend of the target phenotype with age in the healthy population.

[0123] Among them, in the specific implementation process, the fitting of multiple distributions was evaluated, and finally, the JSU distribution (Johnson's Su distribution) was preferably used for fitting, and JUS showed better fitting performance in the evaluation.

[0124] Among them, when modeling, each metric (denoted by t, the metric corresponding to the phenotype) is used as the dependent variable, age is used as the smoothing term, sex and Euler number are used as covariates, and site is used as a random effect. Therefore, the GAMLSS model is defined as follows:

[0125] t = JSU(μ, σ, ν, τ),

[0126]

[0127] σ = f σ (age) + β σ (sex),

[0128] v = β v

[0129] τ = β τ

[0130] The specific description of the above formula is as follows: y corresponds to the target phenotype. In this example, there are three phenotypes, namely the mean value of the whole-brain network, the variance value of the whole-brain network, and the whole-brain level structure-function coupling phenotype. The dependent variable y uses the JSU distribution including four parameters to respectively model the three phenotypes. The four parameters are the median (μ), the coefficient of variation (σ), the skewness (ν), and the kurtosis (τ).

[0131] Corresponding to μ: fμ is a smooth function of age, usually fitted using a smoothing function (such as B-spline). is the regression coefficient for sex is the regression coefficient for the Euler number, z μ is the random effect for site; corresponding to σ: f σ is a smooth function of age, β σ is the regression coefficient for sex; corresponding to v: β v is the regression coefficient, a constant; corresponding to τ: β τ is the regression coefficient, a constant.

[0132] Among them, in the process of using the GAMLSS model to separately fit the age-related changes of the above-mentioned target phenotypes (such as the three different phenotypes in this example) to generate the reference growth curve, the location parameter μ is smoothed using cubic spline functions, and the degrees of freedom (df) of the smoothing function are set to 3 to 5. The degrees of freedom of the scale parameter σ are set to the default value of 3, and v and τ only include the intercept term (intercept term: in the regression model, the intercept term represents the expected value of the dependent variable when all independent variables are zero. It is a constant).

[0133] The maximum number of iterations in the fitting process can be set to 200, and the convergence criterion is that the logarithmic likelihood difference can be 0.001.

[0134] In addition, as described in the previous step S11, other phenotypes can also be used for GAMLSS modeling and JSU distribution fitting to generate the reference growth curve of the corresponding phenotype. For example, when the phenotypes used include regional connection parameter values, a reference growth curve reflecting the regional-level changes of the brain network with age can be obtained. In Figures 3 - 5 some of the figures include the reference growth curves of other phenotypes.

[0135] In addition, after constructing the reference growth curve of the target phenotype (referring to the phenotype selected in step S12), the following steps can also be included:

[0136] Match the brain maps corresponding to the target phenotype at each age or age range (which can be simulation diagrams, rendering diagrams, etc. of the brain maps reflecting the corresponding target phenotype) with the reference growth curve based on age, so as to form the relationship between the reference growth curve of the target phenotype and the series of brain maps, so that the corresponding relationship can be intuitively displayed, and then applied to teaching and research.

[0137] In addition, the growth rate curve, etc. can also be calculated according to the reference growth curve of the target phenotype to study the growth rate, etc. of the target phenotype at different ages. Among them,Figures 3 - 5 The reference growth curves of some target phenotypes, the growth rate curves calculated based on the reference growth curves, the changes in brain maps related to the target phenotypes with age, and the related analyses are shown. Some of them are introduced below.

[0138] Figure 3 Panel a in [reference] shows the reference growth curve (upper panel) and the growth rate curve (lower panel) of the whole-brain network variance value. Among them, the solid line (median) in the reference growth curve of the whole-brain network variance value represents the 50th percentile, and the dashed lines represent the 5th, 25th, 75th, and 95th percentiles. The growth rate of the growth rate curve is evaluated using the first derivative of the median line, and the 1000% confidence interval is estimated through 95 bootstrap analyses ( Figure 3 the shaded gray part in panel a of [reference] is the confidence interval). Figure 3 Panel b in [reference] shows the reference growth curve (upper panel) and the growth rate curve (lower panel) of the whole-brain network mean. Figure 3 Panel c in [reference] shows the reference growth curve (upper panel) and the growth rate curve (lower panel) of the average MSS in different distance regions. Figure 3 Panel d in [reference] shows the reference growth curve (upper panel) and the growth rate curve (lower panel) of the modularity index Q and the small-world property metric parameters Gamma, Lamda, and Sigma. Figure 3 CI in [reference] represents the confidence interval, and yr represents year.

[0139] Figure 4 Panel a in [reference] shows the reference growth curve (upper panel), the growth rate curve (middle panel), and the evolution of the brain map of the central region (lower panel) at the regional level. Among them, the central region is defined as the region where the MSS value exceeds the mean plus 1.5 times the standard error. Figure 4 Panel b in [reference] shows the hierarchical cluster analysis of the pairwise correlations between MSS for each age interval. The resulting dendrogram is divided into four clusters. Figure 4 Panel c in [reference] shows the number distribution of Hubs (Hubs refer to some nodes that play a central role in the brain network. These nodes have a high degree of connectivity in the brain functional network) in 7 cytoarchitectonic classes.

[0140] Figure 5 Panel a in [reference] shows the reference growth curve (left panel) and the growth rate curve (right panel) of the whole-brain level (mean value) corresponding to the structure-function coupling matrix. The gray shading in the reference growth curve represents the 95% confidence interval. Figure 5 Panel b in [reference] shows the reference growth curve (left panel) and the growth rate curve (right panel) calculated from the mean value of each region of the structure-function coupling matrix. Figure 5 Panel c in [reference] shows the reference growth curve (left panel) and the growth rate curve (right panel) calculated from the mean value of the regional coupling in each cytoarchitectonic class. Figure 5The d in [description] shows the correlation between the structure-function coupling matrix and the MSS matrix. The lower line in the left figure represents the significant correlation. The points on the line represent the correlations observed at the corresponding ages, and the boxes represent the correlations obtained from the spin tests. The several scatter plots on the right describe the Pearson correlations between the structure-function coupling matrix and the MSS matrix at different ages (0, 20 years old, 40 years old, 50 years old, 80 years old).

[0141] Figures 3 - 5 The reference growth curves and related graphs for some phenotypes obtained based on this application are shown. It is not difficult to understand that reference growth curves and related graphs for other phenotypes can also be obtained, which will not be elaborated here.

[0142] This application also provides an application of the growth model constructed by the method for constructing a full-life cycle growth model of a human cerebral cortex morphological similarity network. The growth model is constructed according to any one of the methods for constructing a full-life cycle growth model of a human cerebral cortex morphological similarity network in the first embodiment. The application includes at least one of the following:

[0143] Applied to the analysis of the trend of the morphological network of the cerebral cortex changing with age;

[0144] Applied to the analysis of the different patterns shown by the structural network of the cerebral cortex between the sensory region and the paralimbic region as it changes with age;

[0145] Applied to the analysis of the impact of the changes in the structure of the cerebral cortex on its function;

[0146] Applied to the analysis of the individual heterogeneity of the brain networks of patients with Alzheimer's disease, major depressive disorder, or autism spectrum disorder.

[0147] There may be other applications, some of which have been described in the first embodiment or its alternative embodiments, and will not be elaborated here.

[0148] Figure 6 is a structural schematic diagram of a computing device 900 provided by an embodiment of this application. This computing device can execute the above method or each alternative embodiment of the above method. This computing device can be a terminal, or a chip or chip system inside the terminal. As Figure 6 shown, the computing device 900 includes: a processor 910, a memory 920, and a communication interface 930.

[0149] It should be understood that Figure 6 the communication interface 930 in the computing device 900 shown can be used for communication with other devices, and specifically can include one or more transceiver circuits or interface circuits.

[0150] Among them, the processor 910 can be connected to the memory 920. The memory 920 can be used to store the program code and data. Therefore, the memory 920 can be a storage unit inside the processor 910, an external storage unit independent of the processor 910, or a component including a storage unit inside the processor 910 and an external storage unit independent of the processor 910.

[0151] Optionally, the computing device 900 may further include a bus. Among them, the memory 920 and the communication interface 930 can be connected to the processor 910 through the bus. The bus can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For the sake of convenience of representation, Figure 6 a line without an arrow is used in the figure, but it does not mean that there is only one bus or one type of bus.

[0152] It should be understood that in the embodiments of the present application, the processor 910 can adopt a central processing unit (CPU). The processor can also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. Or the processor 910 adopts one or more integrated circuits to execute relevant programs to implement the technical solutions provided by the embodiments of the present application.

[0153] The memory 920 can include a read-only memory and a random access memory, and provide instructions and data to the processor 910. A part of the processor 910 can also include a non-volatile random access memory. For example, the processor 910 can also store information about the device type.

[0154] When the computing device 900 is running, the processor 910 executes the computer-executable instructions in the memory 920 to execute any operation step of the above method and any optional embodiment thereof.

[0155] It should be understood that the computing device 900 according to the embodiments of the present application may correspond to the corresponding subject executing the methods according to the various embodiments of the present application, and the above and other operations and / or functions of each module in the computing device 900 respectively implement the corresponding processes of the methods of each of the present embodiments. For the sake of brevity, they will not be described in detail herein.

[0156] Those of ordinary skill in the art can realize that the units and algorithm steps of the examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present application.

[0157] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments and will not be described in detail herein.

[0158] In several embodiments provided by the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces. The indirect coupling or communication connection of the devices or units may be in an electrical, mechanical, or other form.

[0159] The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the present embodiment.

[0160] In addition, the functional units in the various embodiments of the present application can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit.

[0161] When the above-mentioned functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art or a part of this technical solution can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of this application. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs.

[0162] An embodiment of this application also provides a computer-readable storage medium, on which a computer program is stored. When the program is executed by a processor, it is used to execute the above method, and the method includes at least one of the solutions described in the above various embodiments.

[0163] The computer storage medium of the embodiments of this application can adopt any combination of one or more computer-readable media. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples (non-exhaustive list) of the computer-readable storage medium include: an electrical connection with one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM 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 of the above. In this document, the computer-readable storage medium can be any tangible medium that contains or stores a program, and this program can be used by or combined with an instruction execution system, apparatus, or device.

[0164] The computer-readable signal medium can include a data signal propagated in a baseband or as part of a carrier wave, which carries the computer-readable program code. Such a propagated data signal can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer-readable signal medium can also be any computer-readable medium other than the computer-readable storage medium, and this computer-readable medium can send, propagate, or transmit a program for use by or in combination with an instruction execution system, apparatus, or device.

[0165] The program code contained on a computer-readable medium can be transmitted using any appropriate medium, including but not limited to wireless, wire, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

[0166] The computer program code for performing the operations of this application can be written in one or more programming languages or combinations thereof. The programming languages include object-oriented programming languages such as Java, Smalltalk, C++, and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, executed as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or, alternatively, can be connected to an external computer (e.g., by connecting through the Internet using an Internet service provider).

[0167] In addition, the terms "first," "second," "third," etc. or similar terms such as module A, module B, module C, etc. in the specification and claims are only used to distinguish similar objects and do not represent a specific order for the objects. Understandably, the specific order or sequence can be interchanged when permitted so that the embodiments of this application described herein can be implemented in an order other than that illustrated or described herein.

[0168] In the above description, the reference numerals representing steps, such as S110, S120, etc., do not necessarily indicate that the steps will be executed in this order. The order of the front and back steps can be interchanged when permitted, or they can be executed simultaneously.

[0169] The term "comprising" used in the specification and claims should not be construed as limited to the content listed thereafter; it does not exclude other elements or steps. Therefore, it should be interpreted as specifying the presence of the stated features, wholes, steps, or components, but does not exclude the presence or addition of one or more other features, wholes, steps, or components and their groups. Therefore, the expression "a device comprising device A and B" should not be limited to a device consisting only of components A and B.

[0170] As used herein, the term "one embodiment" or "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the present application. Thus, the appearances of the phrase "in one embodiment" or "in an embodiment" in various places in this specification are not necessarily all referring to the same embodiment, but may refer to the same embodiment. Additionally, in one or more embodiments, the various particular features, structures, or characteristics can be combined in any suitable manner, as will be apparent to those of ordinary skill in the art from the present disclosure.

[0171] Note that the above is only a preferred embodiment of the present application and the technical principles applied. Those skilled in the art will understand that the present application is not limited to the specific embodiments described herein, and various obvious changes, re-adjustments, and substitutions can be made by those skilled in the art without departing from the scope of protection of the present application. Therefore, although the present application has been described in more detail through the above embodiments, the present application is not limited to the above embodiments. Without departing from the concept of the present application, more other equivalent embodiments can be included, all of which fall within the scope of protection of the present application.

Claims

1. A method for constructing a full-life cycle growth model of a human brain cortex morphological similarity network, characterized in that, Including: Obtaining brain MRI images of individuals of different ages and performing quality control operations; Based on the brain MRI images of different individuals after quality control operations, dividing the brain MRI images into n regions in the cerebral cortex; For each region of the structural image of the brain MRI image of each individual, determining m cortical morphological features to form m eigenvalue; constructing a morphological feature network of the individual's brain network according to the m eigenvalues included in the n regions, wherein the morphological feature network is represented as an n*n matrix, and each element value in the matrix is obtained by performing similarity calculation on the eigenvalues of the corresponding two regions; performing morphological feature similarity degree analysis calculation on the morphological feature network of the individual's brain network to obtain at least one first phenotype; For each region of the functional image of the brain MRI image of each individual, determining the time series of the blood oxygenation level-dependent signal of the region and using it as an eigenvalue; constructing a functional network of the individual's brain network according to the respective eigenvalues included in the n regions, wherein the functional network is represented as an n*n matrix, and each element value in the matrix is obtained by performing correlation calculation on the eigenvalues of the corresponding two regions; For the morphological feature network and functional network of an individual, constructing a structure-function coupling matrix, and each element value of the structure-function coupling matrix is obtained by performing correlation calculation on the elements at the corresponding positions of the morphological feature network and the functional network; calculating at least one second phenotype for the structure-function coupling matrix Selecting at least one first phenotype and at least one second phenotype of each individual as dependent variables respectively, using the age of each individual as a smoothing term to perform GAMLSS modeling, and selecting the JSU distribution to respectively fit the age-related changes of the first phenotype and the second phenotype to generate a reference growth curve.

2. The method according to claim 1, wherein The quality control operation includes at least one of the following: Performing quality assessment based on standard quality assessment criteria and excluding problematic MRI images; For the MRI images of individuals of different ages, selecting different pipelines for preprocessing the MRI images, including preprocessing of the structural image and preprocessing of the functional image; Adopting a surface quality control method to evaluate the Euler number for the structural image, and selecting MRI images within a conforming range based on the Euler number; Screening MRI images for the functional image based on head motion control.

3. The method according to claim 1, wherein The determination of m cortical morphological features to form m eigenvalues includes: Determining five cortical morphological features, including: gray matter volume, cortical thickness, surface area, mean curvature, and sulcus depth; Performing standardization processing on each morphological feature respectively to form five eigenvalues for each region.

4. The method according to claim 3, wherein The morphological feature network includes the MIND network; Each element value in the matrix of the morphological feature network is obtained by performing similarity calculation on the eigenvalues of the corresponding two regions, including: Each element value in the MIND network is obtained by performing multivariate KL divergence calculation between the five eigenvalues of the corresponding two regions.

5. The method according to claim 4, characterized in that At least one first phenotype includes: Whole-brain level connection parameter values, including: whole-brain network mean or whole-brain network variance value; Class-level connection parameter values, including: intra-class connection parameter values or inter-class connection parameter values; Region-level connection parameter values; At least one first phenotype further includes a modularity index Q obtained by performing graph-theoretic metric analysis on the morphological feature network, or a small-world property metric parameter.

6. The method according to claim 1, characterized in that Each element value in the matrix of the functional network is obtained by calculating the correlation of the eigenvalue corresponding to two regions, including: Each element value in the matrix of the functional network is obtained by calculating the correlation of the blood oxygenation level-dependent signal sequences corresponding to two regions using the Pearson correlation coefficient.

7. The method according to claim 1, wherein At least one second phenotype includes: The mean value, variance value, median, or standard deviation of the calculated structure-functional coupling matrix.

8. The method according to claim 1, wherein The performing of GAMLSS modeling further includes: the gender and Euler number of each individual as covariates, and the site as a random effect; Wherein, the Euler number is obtained by evaluating the structural image of the brain MRI image, and the site corresponds to the source location of the brain MRI image of the individual.

9. Application of a full-life-cycle growth model for a human cerebral cortex morphological similarity network, characterized in that, The growth model is constructed according to the method for constructing the full-life cycle growth model of the human cerebral cortex morphological similarity network according to any one of claims 1-8, and the applications include at least one of the following: Applied to the analysis of the trend of change of the morphological network of the cerebral cortex with age; Applied to the analysis of the different patterns shown between the sensory region and the paralimbic region in the change of the structural network of the cerebral cortex with age; Applied to the analysis of the influence of the change of the structure of the cerebral cortex on the function; Applied to the analysis of individual heterogeneity of the brain network of patients with Alzheimer's disease, major depressive disorder or autism spectrum disorder.

10. A computing device, characterized in that, Including: A processor, and A memory, on which program instructions are stored, and when the program instructions are executed by the processor, the processor executes the method for constructing the full-life cycle growth model of the human cerebral cortex morphological similarity network according to any one of claims 1 to 8.

Images