Method and system for positioning brain morphological change epicentral region of patient with severe depression

Abstract

The invention relates to a method and a system for positioning a brain morphological change epicenter region of a patient with severe depression in the technical field of brain image analysis. The positioning method for the cerebral morphological change epicenter region of the patient with the major depression comprises the steps of calculating limitation of a structural connection matrix on cortex morphological change of the patient with the major depression, and performing cortex morphological change diffusion simulation on the basis of existence of the limitation. By comparing the magnetic resonance data of the patient with the severe depression with the magnetic resonance data of the healthy people and combining the biological mechanism of the brain structure connection network behind the cortical morphological change of the patient, the epicenter region of the cerebral morphological change of the patient with the severe depression can be accurately positioned. Besides, in the positioning process, compared with the prior art, the method can display and explain reasons for special spatial distribution of the cerebral region of the patient suffering from the severe depression due to cortical morphology change, and the key effect of the structural connection group in shaping the cortical atrophy of the patient suffering from the severe depression is proved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of brain image analysis, in particular to a method and system for locating the epicenter region of morphological changes in the brain of patients with major depressive disorder. BACKGROUND

[0002] Major depressive disorder (MDD) is a mental illness characterized by persistent abnormalities in cognition, emotion, and behavior, which not only brings physical and mental distress to patients and reduces their quality of life, but also causes heavy burden to families and society, including medical resource consumption and loss of labor. With the progress of neuroscience and the development of magnetic resonance technology, neuroimaging research has clearly revealed that patients with major depressive disorder have clear brain structural abnormalities. Although a large number of literatures have described the location and nature of these changes, the underlying neurobiological mechanisms that lead to the specific spatial pattern distribution of these changes are still unclear.

[0003] Different regions of the brain are connected by white matter axon fibers, and these fibers form brain structural connectivity groups that not only support the dynamic coordination between neurons, but also mediate the intracerebral transport of biomolecules. Therefore, brain connectivity groups may act as a channel for pathological transmission, allowing abnormalities in specific brain regions to spread and affect other regions, thereby leading to a specific spatial distribution pattern of abnormal regions. Although existing research has confirmed that abnormal changes in brain structure in certain neurodegenerative diseases and mental illnesses are limited to brain structural connectivity groups. However, there is still a gap in the research on whether the specific spatial pattern of morphological changes in the brain of patients with major depressive disorder is limited to structural connectivity groups. Which regions may become the potential origin of these changes, i.e., the potential epicenter of the disease, also needs to be explored. SUMMARY

[0004] In order to solve the technical problem that the prior art cannot accurately locate the epicenter region of morphological changes in the brain of patients with major depressive disorder, the present application provides a method and system for locating the epicenter region of morphological changes in the brain of patients with major depressive disorder.

[0005] In a first aspect, the present application provides a method for locating the epicenter region of morphological changes in the brain of patients with major depressive disorder, which comprises:

[0006] Collecting structural magnetic resonance data D1 of patients with major depressive disorder and healthy control subjects and processing to obtain a group difference t map representing cortical morphological changes in patients with major depressive disorder; extracting the value of the cortical morphological changes of the jth brain region d from the group difference t map jThe structural and diffusion magnetic resonance data D2 of the healthy population are collected and processed to obtain a group-level structural connection matrix C2, which is used to represent the neural connection relationship between I brain regions. The restrictiveness of C2 on the cortical morphological changes of the patients with severe depression is calculated. On the basis of the restrictiveness, the cortical morphological change diffusion simulation is performed: the i-th brain region is taken as a seed brain region, the diffusion amount of the seed brain region to other brain regions at each time step is calculated through a network diffusion model, and a cortical morphological change simulation diagram at each time step is obtained; the diffusion amount of the I brain regions and the corresponding cortical morphological change simulation diagram are obtained by traversing the I brain regions; i∈[1,I]. The cortical morphological change simulation diagram that best matches the real cortical morphological changes of the patients with severe depression is screened, and the seed brain region of the cortical morphological change simulation diagram is determined as the epicenter region of the brain morphological changes of the patients with severe depression.

[0007] As a preferred example, D1 of the patients with severe depression and D1 of the healthy control subjects are both collected from more than 1000 cases, and the patients with severe depression and the healthy control subjects are matched in terms of age, gender and education years.

[0008] As a preferred example, D1 is pretreated before use, which includes the following steps:

[0009] First, the raw data of D1 is converted into a magnetic resonance image in BIDS format, then the intensity inhomogeneity of the magnetic resonance T1 weighted image is corrected, and the skull is stripped and the cortical surface model is reconstructed after the corrected magnetic resonance image; the reconstructed individual cortical surface model is registered to a standard surface space, and a commonly used cortical morphological index map after registration is calculated; the spatial smoothing processing is performed on the cortical morphological index map to complete the pretreatment.

[0010] And / or, D2 is pretreated before use, which includes the following steps:

[0011] The raw data of D2 is pretreated: intensity standardization, distortion correction, image motion correction, gradient nonlinearity correction and brain tissue extraction; the pretreated D2 in the standard space brain template is converted to the individual native space by spatial registration and template inverse transformation to complete the pretreatment.

[0012] As a preferred example, D2 is first used to obtain an individual-level structural connection matrix C1 corresponding to a healthy population by using a probabilistic fiber tracking method, and then a group-level structural connection matrix C2 is obtained by using a consensus-based classification method for a plurality of C1s.

[0013] As a preferred example, the method for calculating the restrictiveness of C2 on the cortical morphological changes of the patients with severe depression includes the following steps:

[0014] First, the number N of brain regions structurally connected to the i-th brain region is calculated by C2i ; recalculate the mean value of d i of N j brain regions connected with the i-th brain region structure After traversing I brain regions, calculate the Pearson correlation coefficient r1 between the mean value of the cortical morphological change value of I brain regions and the mean value, so as to characterize the restrictiveness of C2 on the cortical morphological change of the severe depression patient; I represents the total number of brain regions.

[0015] As a preferred example, the restrictiveness is corrected by the spin test method and the rewire test method; if the result value of the spin test method and the rewire test method is less than a threshold value, it is proved that C2 has restrictiveness on the cortical morphological change of the severe depression patient.

[0016] As a preferred example, the expression of the network diffusion model is:

[0017] f(t)=e -aHt f0;

[0018] Wherein, t represents the diffusion time of the network diffusion model, f(t) represents the diffusion amount of the brain region at time t, a represents a constant for controlling the diffusion process, H represents the Laplace matrix of C2, f0 represents the initial distribution of the network diffusion model.

[0019] As a preferred example, the method for screening the cortical morphological change simulation graph that best matches the real cortical morphological change of the severe depression patient comprises:

[0020] Calculate the Pearson correlation coefficient r2 between the real cortical morphological change and the diffusion amount at each time step. Calculate the maximum Pearson correlation coefficient r2 max，1 、…、r2 max，i 、…、r2 max，I of I brain regions in all time steps of the diffusion simulation process max，1 、…、r max，i 、…、r max，I Compare the sizes of r

[0021] In the second aspect, the present application further provides a positioning system for the brain morphological change epicenter region of the severe depression patient, which uses the positioning method for the brain morphological change epicenter region of the severe depression patient in the first aspect. The positioning system comprises: an acquisition module, a preliminary calculation module, a restriction module, and a positioning module.

[0022] The acquisition module is used for acquiring structural magnetic resonance data D1 of patients with major depressive disorder and healthy control subjects and structural and diffusion magnetic resonance data D2 of a healthy population. j The restriction module is used for calculating the restriction of C2 on the cortical morphological change of the patient with major depressive disorder.

[0023] In a third aspect, the present application further provides a computer readable medium having a computer program stored thereon. The computer program is executed by a processor to implement the steps of the method for locating the epicenter of the brain morphological change of the patient with major depressive disorder according to the first aspect.

[0024] The present application has the following advantages:

[0025] 1. The present application can accurately locate the epicenter of the brain morphological change of the patient with major depressive disorder by comparing the magnetic resonance data of the patient with major depressive disorder and the healthy population and combining the biological mechanism of the brain structural connection network behind the cortical morphological change of the patient.

[0026] 2. The method for locating the epicenter of the brain morphological change of the patient with major depressive disorder has the advantages of clear principle, simple implementation, and strong statistical test method in the result explanation process, which reduces the false positive rate and greatly enhances the credibility of the result compared with the simple statistical method of the prior art. BRIEF DESCRIPTION OF DRAWINGS

[0027] Figure 1 is a flowchart of the method for locating the epicenter of the brain morphological change of the patient with major depressive disorder in the embodiment;

[0028] Figure 2is a schematic diagram of the structural connectivity matrix for the restriction of the cortical thickness changes of the patients with major depressive disorder;

[0029] Figure 3 is a schematic diagram of the brain region of the epicenter of the cortical thickness changes of the patients with major depressive disorder in the experimental example. DETAILED DESCRIPTION

[0030] The technical solutions in the embodiments of the present application will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all the other embodiments obtained by those skilled in the art without creative work are within the protection scope of the present application.

[0031] It should be noted that when a component is referred to as being "mounted on" another component, it can be directly on the other component or there can be a middle component. When a component is referred to as being "disposed on" another component, it can be directly disposed on the other component or there can be a middle component. When a component is referred to as being "fixed on" another component, it can be directly fixed on the other component or there can be a middle component.

[0032] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used in the description of the application herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. The terms "comprises", "comprising", "includes", "including" and the like are inclusive of the terms "consisting of" and "consisting essentially of".

[0033] The present application aims to explore the shaping ability of brain structural network on the morphological changes of brain cortex, and provides a better direction for clinical diagnosis and treatment. For this purpose, a method for positioning the epicenter region of the morphological changes of the brain of patients with major depressive disorder is provided. Please refer to Figure 1 The positioning method comprises the following steps:

[0034] S1, collecting structural magnetic resonance data D1 of patients with major depressive disorder and healthy control subjects.

[0035] S2, data preprocessing of D1.

[0036] S3, two-sample t-test is performed on the structural magnetic resonance data D1 of the patients with major depressive disorder and the healthy control subjects, and a group difference t map representing the morphological changes of the cortex of the patients with major depressive disorder is obtained.

[0037] S4, collecting structural and diffusion magnetic resonance data D2 of healthy people.

[0038] S5, data preprocessing of D2.

[0039] S6, obtain the group-level structural connectivity matrix C2 using the probabilistic fiber tracking method and the consensus-based classification method.

[0040] S7, calculate the restrictiveness of C2 to the cortical morphological changes in patients with major depression.

[0041] S8, on the basis of the restrictiveness, perform a diffusion simulation of the cortical morphological changes to infer the epicenter region of the brain morphological changes in patients with major depression.

[0042] Specifically, in S1, brain structural magnetic resonance images of 1660 patients with major depression and 1341 healthy control subjects matched therewith were collected from 23 research sites. The matching means that the patients with major depression and the healthy control subjects are matched in age, gender and education years, so as to reduce the interference of other factors on the sample. After image quality control and exclusion of subjects less than 18 years old and more than 65 years old, a total of 1422 qualified brain structural magnetic resonance images of patients with major depression and 1277 healthy control subjects from 22 research sites were included in D1 in the embodiment.

[0043] In S2, data preprocessing was performed using DPABISurf, a cortex-based resting-state magnetic resonance data analysis toolbox. DPABISurf calls fMRIPrep to preprocess the magnetic resonance structural imaging data and provides a set of statistical and viewing tools. The specific data preprocessing process includes the following steps:

[0044] S21, remove the site effect of the data: since D1 comes from multiple sites, different magnetic resonance imaging instruments and scanning sequences are used in each site, so the data from different sites need to be standardized before subsequent processing. This method uses the Combat algorithm based on empirical Bayes to standardize the data. Age, gender, education years and group information are included in the Combat model as reserved biological information to avoid over-correction, and the cortical morphological index map after standardization of the Combat data is used for subsequent calculation.

[0045] S22, convert the original data of the standardized D1 into a BIDS format magnetic resonance image.

[0046] S23, correct the intensity inhomogeneity of the T1-weighted image in the BIDS format magnetic resonance image using the N4BiasFieldCorrection algorithm.

[0047] S24, perform skull stripping on the corrected magnetic resonance image using a deformable template;

[0048] S25, Reconstruct the cortical surface model (gray-white matter boundary and pia matter surface) from the skull stripped MRI image by recon-all command of FreeSurfer 6.0.1.

[0049] S26, Nonlinearly register the reconstructed individual cortical surface model to the FreeSurfer fsaverage standard surface space.

[0050] S27, Calculate the common cortical morphological index maps: cortical thickness, cortical mean curvature, cortical surface area and cortical sulci depth.

[0051] S28, Spatially smooth the cortical morphological index maps using mri_surf2surf command of FreeSurfer: 15mm FWHM Gaussian kernel for cortical thickness and surface area. 25mm FWHM Gaussian kernel for cortical mean curvature and sulci depth.

[0052] In S3, two-sample t-test is performed using the statistical module of DPABI Surf. The effects of gender, age and education are removed as covariates, and then the group difference t map is obtained. After obtaining the group difference t map, the cortical morphological change value d j In this embodiment, based on the Desikan-Killiany atlas of random subdivision, the brain MRI image is divided into 1000 cortical regions in the cerebral cortex. The group difference t map is valued using the DPABI Surf software to obtain the cortical morphological change value d j (each cortical morphological index is a 1x1000 vector). In other embodiments, the cerebral cortex region can also be divided based on other standards, for example, the brain can be divided into 100 to 1000 regions based on the Schaefer atlas, the cerebral cortex can be divided into 360 regions according to the Glasser atlas, and other division methods.

[0053] D2 in S4 is taken from the human brain connection group database, representing the general healthy population.

[0054] The data preprocessing in S5 mainly includes two steps: (1) basic preprocessing, which includes: intensity normalization, distortion correction, image motion correction, gradient nonlinearity correction and brain tissue extraction; (2) converting the basic preprocessed D2 in the standard space brain template to the individual native space through spatial registration and template inverse transformation. Specifically, in the second step, since the structural connectivity is constructed in the individual brain space, it is necessary to convert the 1000 brain region mask in the Desikan-Killiany template in the standard space (Montreal Neurological Institute space) to the individual native space. The specific process is as follows:

[0055] S51, spatial registration: first, the individual structural magnetic resonance image is registered with the first b0 image, and then the registered structural magnetic resonance image is segmented and standardized to the standard space by using the Lie algebra-based differential homeomorphism registration algorithm (DARTEL).

[0056] S52, template inverse transformation: inverse the deformation parameter matrix generated by DARTEL. The 1000 brain region mask of Desikan-Killiany template is inversely mapped from the standard space to the individual native space.

[0057] In S6, D2 first uses a probabilistic fiber tracking method to obtain an individual-level structural connectivity matrix C1 corresponding to a healthy population, and then uses a consensus-based classification method to obtain a group-level structural connectivity matrix C2 for several C1s. The steps of constructing C1 by the probabilistic fiber tracking method include:

[0058] S61, use the probabilistic fiber tracking tool of FSL 6.0.7 to execute the probtrackx2 command (default parameters) for each brain region, and emit 5000 fibers from each voxel in the brain region to the whole brain.

[0059] S62, the structural connectivity strength between brain regions is quantified by the number of fibers (FN): the number of fibers located at both ends of two brain regions is counted to generate a 1000x1000 individual FN matrix, and the size of the brain region and the directional error of the connection are corrected.

[0060] S63, set a threshold for the individual FN matrix to avoid the influence of false connections, and the remaining individual FN matrix is C1.

[0061] The consensus-based classification method is used to obtain the group-level structural connectivity matrix for all C1s, and binarization can obtain C2.

[0062] In this embodiment, the limitation of C2 on the morphological changes of the cortex of patients with severe depression in S7 includes the following steps:

[0063] First, calculate the number of brain regions N connected to the structure of the ith brain region by C2 i Then, calculate the mean value of d i of N brain regions connected to the structure of the ith brain region j

[0064]

[0065] The mean value can be regarded as an estimate. After traversing I brain regions, the Pearson correlation coefficient r1 between the cortical morphological change value and the estimated value of the I brain regions is calculated to characterize the restrictiveness of C2 on the cortical morphological changes of patients with major depressive disorder. I represents the total number of brain regions. Further, considering the significant spatial autocorrelation effect between the cortical brain regions, r1 has a large false positive. In order to avoid and as far as possible to reduce the false positive rate, the spin test method and the rewire test method are used to correct r1. Spin test method is a non-parametric statistical method based on permutation test, which is specially used to solve the problem that traditional statistical methods (such as parametric test) fail due to the correlation between spatial adjacent voxels in brain imaging data. Its core principle is to generate zero distribution by rotating the spatial direction of brain image data, while preserving the spatial autocorrelation structure of the data, so as to realize accurate control of multiple comparison correction. Rewire test method is a non-parametric statistical method for brain network (connectome) analysis, which aims to evaluate whether the observed network topology properties (such as modularity, efficiency, centrality) significantly exceed the random chance level (especially the pseudo-correlation driven by spatial proximity) under the premise of controlling the spatial position or other basic attributes of brain regions. It achieves this goal by generating randomized networks (zero model) that preserve certain constraints (usually spatial distance-connection probability relationship). As shown in Figure 2 r1=0.78, P Spin <0.0001, P Rewire <0.0001, which indicates that the change of cortical thickness in patients with major depressive disorder is limited by the structural connection matrix C2.

[0066] In S8, based on the background that the structural connection restricts the change of cortical morphology, it can be concluded that the change of cortical morphology in patients with major depressive disorder follows the underlying structural connection framework, so the core brain regions of the development of the change of cortical morphology in patients with major depressive disorder can be deduced based on the network diffusion model. In this embodiment, the expression of the network diffusion model that can be used is:

[0067] f(t)=e -aHt f0.

[0068] ​where t represents the diffusion time of the network diffusion model (its unit can be any time unit). f(t) represents the diffusion amount of the brain region at time t. a represents a constant for controlling the diffusion process. H represents the Laplacian matrix of C2. f0represents the initial distribution of the network diffusion model. The steps of the cortical morphological change diffusion simulation include:

[0069] The i-th brain region is taken as a seed brain region. The value of the seed brain region is set to 1, and the values of all other brain regions are set to 0. The diffusion amount of the seed brain region to other brain regions at each time step is calculated by the network diffusion model, and a cortical morphological change simulation map at each time step is obtained. The diffusion amount of the I brain regions and the corresponding cortical morphological change simulation map are obtained by traversing the I brain regions (1000 brain regions in this embodiment). i ∈ [1, I]. This diffusion simulation can simulate the dynamic changes of the cortical morphological changes starting from each brain region of the disease. After obtaining 1000 results, it is necessary to test whether the diffusion starting from each brain region can produce a spatial pattern matching the real cortical morphological changes, that is, to screen the cortical morphological change simulation map that best matches the real cortical morphological changes of the severe depression patients. The screening method includes:

[0070] First, calculate the Pearson correlation coefficient r2between the real cortical morphological changes and the diffusion amount at each time step. Then, calculate the maximum Pearson correlation coefficient r2 max，1 , …, r2 max，i , …, r2 max，I in the diffusion simulation process of the I brain regions at all time steps. Compare the sizes of r max，1 , …, r max，i , …, r max，I , and the seed brain region of the cortical morphological change simulation map corresponding to the maximum Pearson correlation coefficient is determined as the epicenter region of the brain morphological changes of the severe depression patients, that is, the initial core brain region of the cortical morphological changes.

[0071] Similarly, the results obtained by the above diffusion simulation still have a false positive rate due to spatial autocorrelation, so the spin test is used to correct the results of the diffusion simulation in this embodiment. In addition, considering that a total of 1000 brain regions are included for cyclic calculation, the Bonferroni correction method is also used to correct the number of brain regions. As shown in Figure 3 , in one experimental example, the epicenter brain region of the cortical thickness changes of the severe depression patients was found by the diffusion simulation and screening method, and the maximum correlation coefficient after correction was 0.5.

[0072] In another embodiment, a system for locating the epicenter of brain morphological changes in patients with major depressive disorder is also proposed, which uses the method for locating the epicenter of brain morphological changes in patients with major depressive disorder in the above embodiment. The system comprises a collection module, a preliminary calculation module, a restriction module, and a locating module.

[0073] The collection module is configured to collect structural magnetic resonance data D1 of patients with major depressive disorder and healthy control subjects, and structural and diffusion magnetic resonance data D2 of a healthy population. The preliminary calculation module is configured to obtain a group difference t map representing cortical morphological changes of patients with major depressive disorder, and obtain a cortical morphological change value d j of the jth brain region by extracting values from the group difference t map. j The restriction module is configured to calculate the restriction of C2 on the cortical morphological changes of patients with major depressive disorder. The locating module is configured to perform a cortical morphological change diffusion simulation on the basis of the presence of the restriction: taking the ith brain region as a seed brain region, calculating the diffusion amount of the seed brain region to other brain regions at each time step by a network diffusion model, and obtaining a cortical morphological change simulation map at each time step. The diffusion amount of the I brain regions and the corresponding cortical morphological change simulation map are obtained by traversing the I brain regions. i∈[1,I]. The cortical morphological change simulation map that best matches the real cortical morphological changes of patients with major depressive disorder is screened, and the seed brain region of the cortical morphological change simulation map is determined as the epicenter of the brain morphological changes of patients with major depressive disorder.

[0074] In another embodiment, a computer readable medium having a computer program stored thereon is also proposed. The computer program, when executed by a processor, implements the steps of the method for locating the epicenter of brain morphological changes in patients with major depressive disorder in the above embodiment.

[0075] The technical features of the above embodiments can be combined in any manner. To make the description concise, not all possible combinations of the technical features in the above embodiments are described, but as long as the combinations of the technical features do not contradict, they should be considered within the scope of the present disclosure.

[0076] The above embodiments only express several embodiments of the present application, and the description is more specific and detailed, but it should not be understood as limiting the scope of the patent. It should be noted that for those skilled in the art, without departing from the concept of the present application, a number of modifications and improvements can be made, which are within the scope of the present application. Therefore, the scope of protection of the present application should be subject to the appended claims.

Claims

1. A method for locating the epicenter region of brain morphological changes in patients with major depressive disorder, characterized in that, It includes: Structural magnetic resonance imaging (SMRI) data (D1) of patients with major depressive disorder and healthy controls were collected and processed to obtain intergroup t-plots characterizing cortical morphological changes in patients with major depressive disorder. The t-plots were then used to extract the cortical morphological change value (d) for the j-th brain region. j ; Structural and diffusion magnetic resonance imaging (MRI) data D2 were collected from healthy individuals and processed to obtain a group-level structural connectivity matrix C2, which is used to characterize the neural connectivity relationships of I brain regions. Calculate the limiting effect of C2 on cortical morphological changes in patients with major depressive disorder; Based on the constraints, a simulation of the diffusion of cortical morphological changes is performed: the i-th brain region is taken as the seed brain region, and the diffusion amount from the seed brain region to other brain regions over time steps is calculated using a network diffusion model, and a simulation map of cortical morphological changes is obtained at each time step; the diffusion amount of I brain regions and the corresponding simulation map of cortical morphological changes are obtained by traversing I brain regions; i∈[1,I]. The simulation map of cortical morphological changes that best matches the actual cortical morphological changes in patients with major depressive disorder was selected, and the seed brain regions of the simulation map were identified as the epicenter regions of the brain morphological changes in patients with major depressive disorder.

2. The method for locating the epicenter region of brain morphological changes in patients with severe depression according to claim 1, characterized in that, More than 1,000 D1 samples were collected from both patients with major depressive disorder and healthy controls, and the patients with major depressive disorder and healthy controls were matched for age, sex and years of education.

3. The method for locating the epicenter region of brain morphological changes in patients with severe depression according to claim 1, characterized in that, D1 undergoes preprocessing before use, which includes the following steps: First, the raw D1 data is converted into BIDS format magnetic resonance images. Then, the intensity inhomogeneity of the T1-weighted magnetic resonance images is corrected, and the skull is dissected and the cortical surface model is reconstructed on the corrected magnetic resonance images. The reconstructed individual cortical surface model is registered to the standard surface space, and the commonly used cortical morphological index map after registration is calculated. The cortical morphological index map is spatially smoothed to complete the preprocessing. And / or, D2 is preprocessed before use, which includes the following steps: The raw D2 data undergoes basic preprocessing: intensity normalization, distortion correction, image motion correction, gradient nonlinearity correction, and brain tissue extraction. The preprocessed D2 data located in the standard spatial brain template is then transformed to the individual native space through spatial registration and template inverse transformation to complete the preprocessing.

4. The method for locating the epicenter of brain morphological changes in patients with severe depression according to claim 1, characterized in that, D2 first uses a probabilistic fiber tracing method to obtain the individual-level structural connectivity matrix C1 for a certain healthy population, and then uses a consensus-based classification method for several C1s to obtain the group-level structural connectivity matrix C2.

5. The method for locating the epicenter of brain morphological changes in patients with severe depression according to claim 1, characterized in that, The method for calculating the limiting effect of C2 on cortical morphological changes in patients with major depressive disorder includes the following steps: First, calculate the number N of brain regions connected to the i-th brain region structure using C2. i Then calculate N connected to the i-th brain region structure. i d in each brain region j mean After traversing I brain regions, the Pearson correlation coefficient r1 between the values ​​of cortical morphological changes in I brain regions and the mean was calculated to characterize the restrictive effect of C2 on cortical morphological changes in patients with major depressive disorder; I represents the total number of brain regions.

6. The method for locating the epicenter region of brain morphological changes in patients with major depressive disorder according to claim 5, characterized in that, The limitation was corrected using the spin test and rewire test. If the results of the spin test and rewire test are less than the threshold, it proves that C2 has a limiting effect on the cortical morphological changes in patients with major depressive disorder.

7. The method for locating the epicenter of brain morphological changes in patients with severe depression according to claim 1, characterized in that, The expression for the network diffusion model is: f(t)=e -aHt f0; Where t represents the diffusion time of the network diffusion model, f(t) represents the amount of diffusion in the brain region at time t, a represents a constant used to control the diffusion process, H represents the Laplace matrix of C2, and f0 represents the initial distribution of the network diffusion model.

8. The method for locating the epicenter region of brain morphological changes in patients with severe depression according to claim 1, characterized in that, Methods for selecting simulation images of cortical morphological changes that best match the actual cortical morphological changes in patients with major depressive disorder include: Calculate the Pearson correlation coefficient r² between the actual morphological changes in the cortex and the amount of diffusion at each time step; Calculate the maximum Pearson correlation coefficient r² across all time steps in the brain region diffusion simulation process. max，1 r2 max，i r2 max，I ; Compare r max，1 ... r max，i ... r max，I The size of the brain region, and the seed brain region corresponding to the largest Pearson correlation coefficient, were identified as the epicenter of brain morphological changes in patients with major depressive disorder.

9. A system for locating the epicenter region of brain morphological changes in patients with major depressive disorder, characterized in that, It uses the method for locating the epicenter of brain morphological changes in patients with major depressive disorder according to any one of claims 1 to 8; It includes: The acquisition module is used to acquire structural magnetic resonance data D1 from patients with major depressive disorder and healthy control subjects, and structural and diffusion magnetic resonance data D2 from healthy individuals. The preliminary calculation module is used to obtain inter-group t-plots representing cortical morphological changes in patients with major depressive disorder; the t-plots are then used to extract the cortical morphological change value d for the j-th brain region. j It is also used to obtain the group-level structural connectivity matrix C2; A constraint module is used to calculate the constraint of C2 on cortical morphological changes in patients with major depressive disorder; The localization module is used to simulate the diffusion of cortical morphological changes under constraints: the i-th brain region is used as the seed brain region, and the diffusion amount of the seed brain region to other brain regions over time steps is calculated through the network diffusion model, and the simulated cortical morphological changes at each time step are obtained; the diffusion amount of I brain regions and the corresponding simulated cortical morphological changes are obtained by traversing I brain regions; i∈[1,I]; the simulated cortical morphological changes that best match the actual cortical morphological changes of patients with major depressive disorder are selected, and the seed brain region of the simulated cortical morphological changes is determined as the epicenter of the brain morphological changes in patients with major depressive disorder.

10. A computer-readable medium, characterized in that, It stores a computer program that, when executed by a processor, implements the steps of the method for locating the epicenter region of brain morphological changes in patients with severe depression as described in any one of claims 1 to 8.

Images