A depression diagnosis method based on K-GAT network

Abstract

The present invention discloses a depression diagnosis method based on a K-GAT network, comprising the following steps: step 1, based on brain magnetic resonance images, obtaining the brain region thickness of each brain region; using brain regions as nodes, correlation connections between brain regions as edges between nodes, and the extracted brain region thickness as node features to construct a brain topology map; the brain topology map includes brain topology maps of patients with depression and brain topology maps of healthy people; step 2, constructing a K-GAT network for classifying brain magnetic resonance images; screening whether there are many brain region connections, and if there are many brain region connections, it is considered to be a brain magnetic resonance image of a patient with depression; otherwise, it is a brain magnetic resonance image of a healthy person; and inputting the brain topology map into the K-GAT network, training the K-GAT network, and obtaining a trained K-GAT network. Step 3, using the trained K-GAT network to classify the brain magnetic resonance images to distinguish between patients with depression and patients without depression.

Description

Technical Field

[0001] The present invention belongs to the technical field of image processing, and in particular relates to a depression diagnosis method based on a K-GAT network. Background Art

[0002] In recent years, with the in-depth study of depression, people have gradually realized that it is not just a simple problem of low mood, but a brain dysfunction involving structural and functional abnormalities in multiple brain regions. Neuroimaging studies have shown that patients with depression have obvious organic lesions in specific brain areas, such as areas related to emotional processing and cognitive executive functions. However, these changes can usually only be detected after the symptoms have persisted for a long time. Therefore, traditional detection methods based on structural magnetic resonance imaging (MRI) have significant limitations in the early identification of depression.

[0003] Functional magnetic resonance imaging (fMRI) provides a new approach to exploring depression. By detecting differences in functional connectivity in the brain during resting state, researchers can reveal the functional connectivity characteristics of the brain's neural networks. With the rapid development of deep learning technology, researchers are attempting to apply graph neural networks (GCNs) to the analysis of functional brain networks (FBNs) to extract underlying graph representations, thereby more effectively distinguishing patients with depression from healthy individuals.

[0004] Although the brain network analysis method based on graph neural network (GCN) has shown initial potential in depression diagnosis, it still has some shortcomings:

[0005] 1. Noise Issue: When constructing functional brain networks, the commonly used Pearson correlation coefficient (PC) method produces densely connected networks. However, these networks often contain a large amount of noise and spurious connections, primarily due to technical limitations, operator performance, and environmental factors. Unfiltered noise not only reduces the model's classification accuracy but can also lead to overfitting, further increasing computational complexity.

[0006] 2. Insufficient model interpretability: When using existing graph neural network (GCN) models for brain image classification, the model's multi-layered nonlinear transformations and complex graph structure representation make the model's decision-making process difficult to explain. This "black-box" feature aggregation process makes it difficult for clinicians to clearly understand which brain region connections are key to depression diagnosis, thus limiting the model's practical clinical application.

[0007] 3. Lack of training data: The effectiveness of deep learning models typically relies on large amounts of high-quality training data. However, currently, depression-related fMRI datasets for training are relatively limited. In particular, differences in brain region selection during fMRI data collection across different medical institutions increase data heterogeneity. This issue limits the model's generalization ability, making it difficult to demonstrate consistent performance across different datasets. Summary of the Invention

[0008] To address the shortcomings of the existing technology, the present invention proposes a depression diagnosis method based on a graph attention network. The method includes the following steps:

[0009] Step 1: Obtain the brain thickness of each brain region based on brain magnetic resonance images;

[0010] A brain topology map is constructed using brain regions as nodes, correlation connections between brain regions as edges between nodes, and the extracted thickness of each brain region as a feature of the node. The brain topology map includes brain topology maps of patients with depression and brain topology maps of healthy people; the healthy people refer to people who do not suffer from depression.

[0011] Step 2: Construct a K-GAT network to classify brain magnetic resonance images; input the brain topology map into the K-GAT network, train the K-GAT network, and obtain a trained K-GAT network.

[0012] Step 3: Use the trained K-GAT network to classify brain magnetic resonance images.

[0013] Furthermore, in step 1, the thickness of each brain region is obtained based on the brain magnetic resonance image, which specifically includes the following steps:

[0014] Step 1.1, obtaining a whole-brain gray matter image based on brain magnetic resonance imaging;

[0015] The gray matter image includes gray matter volume GMV and gray matter concentration GMC;

[0016] Step 1.2, based on the gray matter image, obtaining the gray matter thickness GMT and gray matter surface area SA corresponding to each brain region of the whole brain;

[0017] Step 1.3, constructing a first brain map based on the gray matter volume GMV, gray matter concentration GMC, gray matter thickness GMT and gray matter surface area SA;

[0018] Step 1.4: Calculate the low-frequency amplitude ALFF, fractional low-frequency amplitude fALFF, and local consistency ReHo for the first brain map to obtain the brain thickness of each brain region.

[0019] Furthermore, in the brain topology map, brain region thickness is used as a node feature Where n is the number of nodes, i.e. the number of brain regions, d in is the feature dimension; the connection relationship between nodes uses the adjacency matrix Indicates that A ij Represents the connection relationship between node i and node j, the adjacency matrix A ij Each element in is assigned a value according to the threshold, 1 if it is greater than the threshold and 0 if it is less than the threshold.

[0020] Furthermore, the K-GAT network includes a first linear layer, a K-GAT layer, a second linear layer

[0021] The node feature X in the brain topology map is input into the first linear layer of the K-GAT network to obtain the hidden feature d hidden Represents the feature dimension of the linear layer;

[0022] Hidden feature H (0) Expressed as:

[0023] H (0) =W in X

[0024] in, is the weight matrix of the linear layer, X represents the thickness of the brain area;

[0025] Hidden feature H (0) With the adjacency matrix A ij Input to the K-GAT layer, for node i and its neighbor node j, update node i and obtain the updated node feature H (L) .

[0026] The characteristics of each node H (L) After the second linear layer, the feature H (L) The dimension of d hidden Compress to the final output dimension d out ;

[0027] Z=W out H (L)

[0028] in, is the final output result, d out is the output category;

[0029] The output result Z is applied with log_softmax to convert the output of each node into the logarithmic probability P of the category;

[0030] P = log_softmax(Z)

[0031] in, is the log probability of the node in each category.

[0032] Furthermore, the hidden feature H (0) and adjacency matrix A ij Input to the K-GAT layer, for node i and its neighbor node j, update node i and obtain the updated node feature H (L) , specifically including the following steps:

[0033] Step 2.1 Linearly transform the hidden features

[0034] Hidden features for each node Perform a linear transformation:

[0035]

[0036] in, is the weight matrix, l represents the number of iterations;

[0037] Step 2.2 When A ij =1, calculate the attention coefficient e between node i and its neighbor node j ij :

[0038]

[0039] in is a learnable attention vector;

[0040] ‖ represents the concatenation operation of features;

[0041] LeakyReLU(·) represents the activation function;

[0042] Step 2.3, for the attention coefficient e ij Normalize and get the normalized attention weight α ij ;

[0043]

[0044] in Represents the neighbor set of node i

[0045] Step 2.4: Perform weighted summation on the neighbor node features of node i to obtain the new hidden feature representation of node i

[0046]

[0047] Where σ is a nonlinear activation function and W′ is a shared weight matrix;

[0048] Step 2.5: Use the multi-head attention mechanism to the new hidden features of node i Perform splicing, expressed as:

[0049]

[0050] Among them, K represents the number of attention heads, ‖ represents the splicing operation, and W (k) are the attention weight and linear transformation matrix of the k-th head respectively.

[0051] Step 2.6, the feature of node i Input to the KAN layer, further feature extraction, set l = l + 1, return to step 2.1, and finally get the feature H of each node (L) .

[0052] Furthermore, step 2 also includes pruning the trained K-GAT model to prune the network into a smaller sub-network.

[0053] Beneficial Effects: Compared with the traditional graph neural network (GCN) model, the K-GAT network-based depression diagnosis method of the present invention can more effectively deal with noise in brain networks by introducing an attention mechanism. The Kolmogorov-Arnold-Graph Attention Network (K-GAT) model assigns weights to different node connections. By learning the importance of connections between different brain regions, it more closely approximates the actual brain map structure, reduces the impact of noise, and thus improves the model's classification precision and accuracy.

[0054] The K-GAT model also offers significant advantages in improving model interpretability. Thanks to the introduction of the Kolmogorov-Arnold Network (KAN) layer, the model's classification phase uses frequency analysis of retained features after pruning to more intuitively demonstrate which brain regions play a key role in depression diagnosis, providing stronger decision support for clinicians.

[0055] Furthermore, this invention, through multimodal analysis of patients' functional MRI scans, effectively alleviates the limitations of existing datasets and improves the model's generalization and diagnostic accuracy. This not only enhances early diagnosis of depression but also provides new tools and methods for developing personalized treatment plans. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0057] Figure 1 This is a diagram of the K-GAT network architecture.

[0058] Figure 2 This is a schematic diagram of pruning after adding the KAN layer.

[0059] Figure 3 It is a diagram of the key brain areas that are preserved and their connections.

[0060] Figure 4 This is a comparison of the brain topology of a patient and a healthy person. DETAILED DESCRIPTION

[0061] In order to make the purpose, technical solutions and advantages of the present invention clearer, the specific embodiments of the present invention are described below in conjunction with the embodiments and drawings so that those skilled in the art can better understand the present invention. It should be noted that the described embodiments are part of the embodiments of the present invention, not all of the embodiments, and are not intended to limit the scope of the invention claimed for protection. All other embodiments obtained by those of ordinary skill in the art without making creative work are within the scope of protection of the present invention.

[0062] Step 1: Collaborating with a hospital, we acquired fMRI images from over a thousand real patients with depression. We then converted their formats, performed denoising, and registration, and obtained the patients' whole-brain gray matter volume (GMV) and gray matter concentration (GMC). ① Converted the original fMRI DICOM format files into the NIFTI format for data analysis.

[0063] ② Use SPM12 to divide, normalize, and modulate the structural phase, or only divide and normalize it, to obtain gray matter volume (GMV) and gray matter concentration (GMC) brain maps, i.e., gray matter images, and resample them to 1 mm x 1 mm x 1 mm.

[0064] ③ After visual inspection of the gray matter images confirmed no abnormalities, gray matter images of all subjects were collected and subjected to correlation analysis with the average gray matter image of all subjects. Images with a correlation coefficient less than 0.8 were considered abnormal and were excluded from the following analysis.

[0065] ④ Create a mask with a voxel value > 0.2 to filter and remove other possible noise signals.

[0066] ⑤ Registration. fMRI image registration is the process of aligning or overlaying one or more sets of fMRI images to spatially unify images from different time points or subjects. This process is crucial because fMRI images can exhibit inconsistencies in position and posture during acquisition due to factors such as subtle movements of the subject, different equipment, and differences in scanning angles, which can affect subsequent data analysis and interpretation. Image registration eliminates these spatial differences, allowing different fMRI images to accurately overlap within the same spatial reference frame, enabling time series analysis, comparisons of brain region function, and intergroup statistical analysis.

[0067] Step 2: Brain regions were divided according to the Desikan-Killiany (DK) template. Based on existing research, 68 brain regions with a high correlation with brain diseases such as depression and autism were selected to obtain gray matter thickness (GMT) and surface area (SA) of all 68 brain regions.

[0068] ① Refer to the Freesurfer script used and published by the ENIGMA working group to preprocess the gray matter images;

[0069] ② This script will generate a CSV file. The first line is a header describing the names of the extracted brain regions. Each line after the first line gives the gray matter thickness (GMT) and surface area (SA) of each brain region found in the FreeSurfer directory.

[0070] ③ Perform quality control on the gray matter thickness (GMT) and surface area (SA) results of different brain regions obtained above in R-studio, using a method similar to the aforementioned association analysis. Potential outliers will be marked to remind you to pay attention in further quality control.

[0071] ④ Perform inner surface quality control in Matlab and outer surface quality control in Freesurfer, referring to the relevant steps of ENIGMA.

[0072] Step 3: Troubleshoot interference factors.

[0073] The above-mentioned brain structural indices (gray matter volume (GMV), gray matter concentration (GMC), gray matter thickness, and gray matter surface area (SA)) were residualized. Regression analysis was first performed to remove the effects of age and sex on brain structure. A primary brain map was then constructed based on these factors to prepare for further analysis. This analysis was performed using custom scripts and programs in Matlab. Following the initial map, a whole-brain analysis was performed to account for disease duration, equivalent dose of psychiatric medications, and potential age-sex interactions to eliminate any significant influence of these factors on brain structure.

[0074] Step 4: Preprocessing and calculation of low frequency amplitude (ALFF), fractional low frequency amplitude (fALFF) and local consistency (ReHo)

[0075] ① The functional magnetic resonance imaging (fMRI) data of the first 10 time points were removed to avoid the influence of MRI machine field uniformity and subject discomfort, and the last 230 time points were included in the subsequent analysis;

[0076] ② Correct the acquisition time difference between different scanning slices at the same time point (temporal correction) and the subject's head motion effect at different time points during the entire data acquisition process (spatial correction). After spatial correction, check the head motion parameters of each subject and exclude subjects whose head movement during the scan was >3mm in any direction or >3° in any angle;

[0077] ③ After head motion correction, the brain images of each subject were normalized to the standardized brain MNI space developed by the Montreal Neurological Institute (MNI) according to the EPI template and an optimized 12-parameter affine transformation; the voxel size was further resampled to 3

[0078] ×3×3mm3;

[0079] ④ Smooth according to the 6 mm smoothing core;

[0080] ⑤ Remove data linear drift and perform 0.01-0.1 Hz bandpass filtering on the data to reduce the influence of high-frequency physiological noise;

[0081] ⑥ The white matter signal, cerebrospinal fluid signal, whole brain average signal and 24 head motion parameters were used as covariates and regressed out from the original time series voxel by voxel.

[0082] ⑦ Calculate the low-frequency amplitude ALFF, fractional low-frequency amplitude fALFF, and local consistency ReHo in the brain imaging data processing and analysis toolbox DPABI to obtain the specific thickness of each brain region.

[0083] Step 5: Organize the above specific brain map data, build brain area connections, and create a data set.

[0084] Brain thickness data were organized into CSV format, based on individual samples and different brain regions. Correlation analysis was performed on the data using correlation coefficients between brain regions, presenting them in a matrix format. Experiments were then conducted to set correlation thresholds to determine connectivity between brain regions and construct an adjacency matrix.

[0085] Brain regions are used as nodes, correlation connections between brain regions are used as edges between nodes, the extracted thickness of each brain region is used as the feature of the node, and the edge connections between nodes are constructed according to the Pearson correlation coefficient between each brain region, thereby realizing the construction of a brain topology map.

[0086] In the present invention, a single case sample is taken as an example of graph input. The input brain topology graph consists of node features and structural information. The node features are a 68×1 matrix, where 68 is the number of nodes (68 brain regions) and 1 is the feature dimension of the node (brain region thickness).

[0087] Step 6: Build a K-GAT network and use the brain map as the input for image classification

[0088] In response to the problems of low accuracy and poor interpretability of existing models that cannot focus on key brain areas, the following Figure 1 A graph attention network architecture with a KAN layer is shown.

[0089] First, brain region thickness is used as input node feature Where n is the number of nodes, i.e. the number of brain regions, d in As the input feature dimension, in this specific implementation, there is only one feature dimension, brain region thickness feature. The correlation between brain regions is used as the input edge, and the adjacency matrix is ​​used as the input edge. , which represents the connection relationship between nodes (brain regions), where A ij Represents the connection relationship between node i and node j. Each element in the adjacency matrix is ​​assigned a value according to the threshold (1 for elements greater than the threshold and 0 for elements less than the threshold). The element value can be regarded as whether it constitutes a connection. If it is set to 1, it is connected, and if it is set to 0, it is not connected. Through the linear layer (lin_in), the input feature X is mapped to the hidden feature space That is: H (0) =W in X, where is the weight matrix of the linear layer, H (0) is the initial hidden layer feature, d hidden represents the feature dimension of the linear layer, and the hidden layer is a linear layer;

[0090] Next, we enter the middle layer, the graph attention layer. The graph attention mechanism is the core of GAT. It performs weighted aggregation on the neighbors of each node, and controls the weights of different neighbors through the learned attention coefficients during aggregation. Specifically, given node i and its neighbor node j, the GAT update process is divided into the following steps:

[0091] (a) Linear transformation:

[0092] Hidden features for each node Perform a linear transformation:

[0093]

[0094] in, is the weight matrix, d ′ is the dimension after linear transformation, d represents the feature dimension, and l represents the number of iterations.

[0095] (b) Calculation of attention coefficient:

[0096] Next, the attention coefficient between node i and its neighbor j is calculated. Attention coefficient e ij Indicates the importance of node j to node i. The calculation formula is as follows:

[0097]

[0098] in, is a learnable attention vector, ‖ represents the concatenation operation of features, e ij is the unnormalized attention coefficient, and LeakyReLU(·) represents the activation function.

[0099] (c) Softmax normalization:

[0100] In order to obtain the normalized attention weight α ij , apply the Softmax function to all neighbor nodes j of each node i:

[0101]

[0102] in represents the neighbor set of node i.

[0103] (d) Weighted aggregation:

[0104] Finally, the weighted summation of the neighbor node features of node i is performed to obtain the new hidden feature representation of node i

[0105]

[0106] Where σ is a nonlinear activation function (such as ReLU) and W′ is a shared weight matrix.

[0107] (e) Multi-head attention mechanism:

[0108] To increase the expressive power of the model, GAT uses a multi-head attention mechanism. Through multiple independent attention mechanisms, each head generates a hidden feature representation, and then these representations are spliced ​​together:

[0109]

[0110] Among them, K represents the number of attention heads, ‖ represents the splicing operation, and W (k) are the attention weight and linear transformation matrix of the k-th head respectively;

[0111] After each GAT feature extraction, the output results Passed into KANLayer for further processing, the learnable single variable function G in KAN j,i (x) is parameterized by B-spline. Fourier functions are selected as spline functions, i.e., univariate functions; each univariate function operates on a predefined interval [a, b], which is evenly divided into g-1 subintervals (where g represents the number of grid points). In each subinterval, the function consists of q-order (i.e., q-1 degree) polynomial fragments, which are smoothly connected at the grid points. Each univariate function G j,i (x) can be expressed as a linear combination of g+q-1 B-spline basis functions:

[0112]

[0113] Among them, w l is a learnable weight parameter, B l,q (x) is the standard B-spline basis function. By learning these weight parameters, the model can adaptively adjust the shape of each univariate function to capture the complex nonlinear relationship between input features.

[0114] This B-spline-based representation method not only ensures the smoothness of the function, but also provides sufficient flexibility to approximate various nonlinear mappings. KANLayer takes the features of 68 brain regions extracted by graph neural network processing as input. After multiple layers of GAT and KANLayer, the final hidden feature H (L) After the second linear layer, the dimension is reduced from d hidden Compress to the final output dimension d out ; Z=W out H (L) ,in, is the final output result, d out is the number of output categories (for binary classification tasks, d out =2). Finally, log_softmax is applied to the output to convert the output of each node into the logarithmic probability of the category: P = log\_softmax(Z), where is the log probability of the node in each category.

[0115] like Figure 1As shown in Figure 1, the input includes node features and edge connection information. The edges are updated through the graph attention mechanism, where the thickness of the edge represents the weight of the edge. The larger the weight, the thicker the edge. Taking the directed edge CA as an example, the attention mechanism of the edge aggregation is determined by all the edges and node features connected to C. The update of node features is implemented through the K-GAT network, which is affected by all nodes directly connected to the node. Figure 1 Taking the update of node F as an example, we first perform neighborhood aggregation. F is directly connected to nodes E, D, and G, and the aggregation process is determined by these connections. Specifically, the contributions of edges FE, FD, and FG are and Where α is the weight in the normalized adjacency matrix, h (l) is the node feature after the lth update. These contributions are summed up to form the aggregated result Then, aggregate the results As input into the KAN layer, it passes through a series of learnable single variable functions G j,i (x) processing. Assume that the aggregation result is d hidden dimensional vector, the output of the KAN layer is m-dimensional. The KAN layer applies m functions to each dimension i. The j-th component of the KAN layer output is Finally, the new features of node F are determined by the output of the KAN layer: Among them, y m Represents the value of the mth dimension output by the KAN layer;

[0116] Where: i ranges from 1 to d hidden , represents the dimension of the input feature; is the i-th component of the aggregated input feature vector; G j,i is a learnable univariate function of the j-th output dimension on the i-th input dimension; therefore, y m Specifically refers to the last component of the KAN layer output vector, which is obtained by summing all input dimensions after the corresponding single variable function transformation. j Together they constitute the new feature representation of the node [y1,y2,…,y m ].

[0117] During the training process, the univariate function G in the KAN layer j,i The (x) parameter is optimized through backpropagation, enabling the model to learn the most appropriate nonlinear transformation. This architecture first aggregates the graph structure and then applies the nonlinear transformation, effectively combining graph structure information and node features, and enhancing the model's expressive power through learnable nonlinear functions.

[0118] Step 7: Set the threshold and calculate and verify the degree of association based on the pruning results (feature frequency).

[0119] After obtaining the trained K-GAT model, in order to visualize the importance of each brain region, the following steps need to be performed: Figure 2 The pruning operation shown is used to highlight key brain regions, set thresholds and neurons to be retained to prune the trained K-GAT model, prune nodes (brain regions) with smaller input or output connections, and retain important brain regions. Specifically: sparse the KAN at the node level. For example, for each node, define the incoming fraction I of the u-th neuron in the KAN layer at the w-th layer. w,u and outgoing score O l,i They are:

[0120] I w,u =max r (|φ w-1,r,u |1)

[0121] O w,u =max j (|φ w+1,j,u |1)

[0122] φ w-1,r,u represents the activation function connecting the rth and uth neurons in the w-1th layer;

[0123] φ w+1,j,u represents the activation function connecting the jth and uth neurons in layer l.

[0124] If the score I is passed in w,u and outgoing score O w,u If both are smaller than the set threshold, the neuron is considered to be unimportant and is pruned.

[0125] According to the characteristic frequencies of the brain regions that are finally retained, the above steps are combined to finally obtain the following Figure 3 The 17 key brain regions shown in the figure also showed that there were significant differences in the thickness and connectivity of key brain regions between patients (MDD) and healthy subjects (HC), such as Figure 4 As shown, the size of the node (ball) represents the thickness of the brain area, and the connection of the edge represents the connection between brain areas. It can be clearly observed that Figure 4 (a) The brain regions of patients with MDD were significantly more connected than those in Figure 4 Middle (b) HC of a healthy person, and there are also certain thickness differences in brain regions such as the left precuneus.

[0126] The K-GAT model proposed in this paper has significantly improved multiple indicators of graph classification (accuracy, recall, etc.) compared with existing models and traditional machine learning, as shown in Table 1.

[0127] Table 1 Evaluation indicators of different models

[0128]

[0129] like Figure 2 As shown, the model proposed in the present invention uses the graph attention network to extract features, and then inputs the extracted features into the KAN layer for classification. By setting the threshold and pruning operations, the node feature frequency is obtained. The node feature frequency refers to the number of times the node appears. The frequency (number) of the remaining brain regions after pruning, the more it appears, the higher the frequency, and the more important the brain region, thereby reflecting the degree of association between the brain region and depression.

[0130] According to the characteristic frequencies reflected by the final pruning results, the degree of association between each brain region and depression is shown in Table 2.

[0131] Table 2 Correlation between brain regions and disease

[0132]

[0133]

[0134]

[0135] By analyzing the disease associations and actual structural and functional characteristics of brain regions, the following conclusions were drawn:

[0136] ①Regional perspective.

[0137] The degree of association with the disease is from high to low: parietal lobe, temporal lobe, frontal lobe, limbic system, cingulate gyrus, parahippocampal gyrus, lingual gyrus, temporal pole, triangular inferior frontal gyrus, and paracentral lobule.

[0138] Among the above regions, only some sub-regions are related to the classification of depression. For example, the relevant regions in the parietal lobe are the left precuneus (L_precuneus), left postcentral gyrus (L_postcentral), left superior parietal lobule (L_superiorparietal), and right precuneus (R_precuneus); the relevant regions in the temporal lobe are the left fusiform gyrus (L_fusiform), left entorhinal cortex (L_entorhinal), and right parahippocampal gyrus (R_parahippocampal); and the relevant regions in the frontal lobe are the right frontal pole (R_frontalpole), right anterior middle frontal gyrus (R_rostralmiddlefrontal), and left orbital frontal cortex (L_parsorbitalis).

[0139] These associated subregions are mainly responsible for self-referential processing, introspection and episodic memory retrieval; facial recognition, emotion processing; higher cognitive functions, emotion regulation and reward processing; emotion processing, pain perception and cognitive control, etc., which are functions related to depression.

[0140] Abnormalities in these areas may lead to symptoms such as attention deficit disorder, difficulty forming and recalling positive memories, changes in body sensation, and psychomotor retardation.

[0141] The regions of the parietal lobe most strongly associated with the disease are the left precuneus (L_precuneus), left postcentral gyrus (L_postcentral), left superior parietal lobule (L_superiorparietal), and right precuneus (R_precuneus). The left precuneus is a core node of the default mode network (DMN), which is closely associated with self-referential processing, introspection, and episodic memory retrieval. Overactivity of the DMN may lead to excessive self-focus, rumination, negative self-evaluation, and difficulty disengaging from negative thoughts. Abnormalities in the postcentral gyrus and superior parietal lobule, which are involved in spatial cognition and sensory integration, may lead to impaired attention allocation and altered bodily sensation, potentially contributing to the attentional problems and somatic symptoms of depression.

[0142] The temporal lobe regions with the second strongest associations were the left fusiform gyrus, left entorhinal cortex, and right parahippocampal gyrus. The fusiform gyrus is involved in facial recognition and emotion processing, and abnormalities in it may lead to social-emotional processing disorders and difficulty recognizing facial expressions. The entorhinal cortex and parahippocampal gyrus are closely involved in memory formation and retrieval, and abnormalities in these areas may lead to problems with episodic and autobiographical memory, increased negative recollections of past experiences, and difficulty forming and recalling positive memories.

[0143] The third frontal lobe region most associated with disease is the right frontal pole, the right rostralmiddle frontal gyrus, and the left parsorbitalis. These regions are involved in higher-order cognitive functions, emotion regulation, and reward processing. Abnormalities in the frontal pole and the rostralmiddle frontal gyrus may lead to executive dysfunction, reduced emotion regulation, and decreased cognitive flexibility. Abnormalities in the parsorbitalis may also lead to a weakened response to rewards, causing anhedonia and difficulty regulating emotions.

[0144] The main implicated regions of the cingulate gyrus include the left caudal anterior cingulate gyrus (L_caudalanteriorcingulate) and the right rostralanteriorcingulate gyrus (R_rostralanteriorcingulate). The cingulate gyrus is an important component of the limbic system and is involved in emotion processing, pain perception, and cognitive control. Abnormalities in this region may lead to decreased emotion regulation, excessive reactivity to negative stimuli, an imbalance between cognitive control and emotion processing, and altered pain perception.

[0145] The parahippocampal gyrus includes the left and right parahippocampal gyri. It is closely associated with memory formation and spatial navigation. Abnormalities in the parahippocampal gyrus can lead to problems with episodic and situational memory, spatial cognition impairment, and abnormalities in emotion-related memory processing.

[0146] The lingual gyrus primarily involves the left lingual gyrus (L_lingual). This lingual gyrus is primarily involved in visual processing and visual memory. Abnormalities in this gyrus can lead to changes in visual information processing, problems with vision-related memory, and potentially affect the processing of visual stimuli related to emotions.

[0147] The temporal pole primarily involves the left temporal pole (L_temporalpole). This region is involved in social and emotional processing. Abnormalities in the temporal pole may lead to impairments in social cognition, changes in emotional processing, and problems with autobiographical memory retrieval.

[0148] The parstriangularis (PFG) primarily involves the left PFG. As part of Broca's area, it is involved in language processing and executive function. Abnormalities in the parstriangularis may lead to subtle changes in language expression, decreased executive control, and potentially affect internal language and thought processes.

[0149] The paracentral lobule primarily involves the left paracentral lobule (L_paracentral). The paracentral lobule is associated with motor and sensory function. Abnormalities in this lobule may lead to subtle changes in motor control, altered bodily sensation, and may be associated with psychomotor retardation, a symptom of depression.

[0150] ②Left and right brain angles.

[0151] From a quantitative analysis, there are 12 related areas in the left brain and 5 in the right brain. The left brain has significantly more than the right brain, which to a certain extent indicates that the left brain plays a more important role in the neuropathology of depression.

[0152] Table 3 Brain regions related to left and right brain diseases

[0153] Left Brain Key Brain Region L_Role_Brain_Region Right Brain Key Brain Region R_Role_Brain_Region Left precuneus Right frontal pole left fusiform gyrus right anterior cingulate gyrus Left entorhinal cortex Right middle frontal gyrus Left postcentral gyrus right parahippocampal gyrus Left superior parietal lobe Right precuneus Left caudal anterior cingulate gyrus Left parahippocampal gyrus left lingual gyrus Left temporal pole Left triangular gyrus Left paracentral lobule left orbital gyrus

[0154] Functionally, the left and right brain regions have different functions.

[0155] The relevant brain areas in the left brain space are mainly responsible for language processing (left triangular inferior frontal gyrus), logical thinking and analysis (left superior parietal lobule, left precuneus), memory formation (left entorhinal cortex, left parahippocampal gyrus), and emotion regulation (left orbital frontal cortex, left caudal anterior cingulate gyrus).

[0156] Therefore, the depressive symptoms that may be caused by changes in the left brain are mainly negative self-talk and thinking patterns, affecting logical reasoning and problem-solving abilities, affecting memory formation and retrieval, and may aggravate negative memory preferences. Abnormalities in the left emotion regulation network may lead to difficulties in emotion regulation.

[0157] The relevant brain areas in the right brain space are mainly responsible for processing attention (right precuneus), emotion processing and regulation (right frontal pole, right rostrum anterior cingulate gyrus), intuition and creative thinking (anterior part of the right middle frontal gyrus), and episodic memory (right parahippocampal gyrus).

[0158] Therefore, the depressive symptoms that may be caused by changes in the right brain are mainly problems with spatial cognition and attention allocation, affecting emotional processing and regulation, which may be the main source of core depressive symptoms, affecting intuition and creative thinking, which may lead to decreased problem-solving ability and rigid thinking, and affecting situational memory, which may lead to an increase in negative memories of past experiences.

[0159] Overall, more extensive left-brain abnormalities may lead to more cognitive symptoms, such as distorted thinking, attention, and memory problems. Right-brain abnormalities, while less common, may more directly impact core emotional symptoms. An imbalance between the left and right brain may lead to an imbalance in emotional and cognitive processing, exacerbating depressive symptoms.

[0160] ③Network perspective

[0161] The associated brain regions are important components of networks closely related to emotion regulation, cognitive control, and episodic memory retrieval.

[0162] The frontal lobe is an important component of the cognitive control network and the prefrontal-limbic system network. Dysfunction of these networks may lead to difficulties in emotion regulation and cognitive control in patients with depression.

[0163] The precuneus is a core node of the default mode network. This network is active during resting states and is associated with self-referential processing, theory of mind, and episodic memory retrieval. Abnormalities in the default mode network may contribute to excessive self-focus and rumination in patients with depression.

[0164] Parts of the temporal lobe are important components of emotion processing and memory networks, and abnormalities in these areas may contribute to the difficulties people with depression experience in experiencing emotions and retrieving memories.

[0165] The cingulate gyrus of the limbic system is an important node in the emotion processing network and the cognitive control network. Its abnormality may lead to an imbalance between the two networks, affecting the integration of emotion regulation and cognitive control.

[0166] In addition, the default mode network (DMN) involves areas such as the precuneus and cingulate gyrus. DMN abnormalities may lead to abnormal self-referential processing and increased rumination in patients with depression.

[0167] The emotion processing network includes temporal lobe structures and the limbic system. Abnormalities in this network may lead to difficulties experiencing and regulating emotions. The cognitive control network primarily involves frontal lobe regions. Dysfunction in this network may contribute to the difficulties with attention, decision-making, and emotion regulation seen in patients with depression.

[0168] In summary, the experimental results reveal the degree of correlation between different brain regions and depression, the functional abnormalities caused by different brain regions and their symptom reflections, and the impact on specific brain network structure and function.

Claims

1. A classification method for brain magnetic resonance imaging based on graph neural network, characterized in that: The steps include: Step 1: Obtain the brain thickness of each brain region based on brain magnetic resonance images; The brain regions are used as nodes, the correlation connections between brain regions are used as edges between nodes, and the extracted thickness of each brain region is used as the feature of the node to construct a brain topology map; The brain topology map includes brain topology maps of depression patients and brain topology maps of healthy people; Topography of the brain In the middle, brain region thickness is used as a node feature ,in is the number of nodes, that is, the number of brain regions, is the feature dimension; The connection relationship between nodes is based on the adjacency matrix express, Representation node and nodes The connection relationship between them, the adjacency matrix Each element in is assigned a value according to the threshold, if it is greater than the threshold, it is assigned 1, and if it is less than the threshold, it is assigned 0; Step 2: construct a K-GAT network for classifying brain magnetic resonance images, input the brain topology map into the K-GAT network, train the K-GAT network, and obtain a trained K-GAT network; The K-GAT network includes a first linear layer, a K-GAT layer, and a second linear layer; Node characteristics in brain topology maps Input to the first linear layer in the K-GAT network to obtain hidden features Represents the feature dimension of the linear layer; Hidden Features Expressed as: ； in, is the weight matrix of the linear layer, X represents the thickness of the brain area; Hidden Features With the adjacency matrix Input to the K-GAT layer, for node and its neighbor nodes , for nodes Update to get updated node features ; Characteristics of each node After the second linear layer, the features The dimension from Compress to the final output dimension ; ； in, is the final output, is the output category; Output Apply log_softmax to convert the output of each node into the logarithmic probability of the category ; ； in, is the log probability of the node in each category; Step 3: Use the trained K-GAT network to classify brain magnetic resonance images.

2. The method for classifying brain magnetic resonance imaging based on graph neural network according to claim 1, characterized in that: In step 1, the thickness of each brain region is obtained based on the brain magnetic resonance image, which specifically includes the following steps: Step 1.1, obtaining a whole-brain gray matter image based on brain magnetic resonance imaging; The gray matter image includes gray matter volume GMV and gray matter concentration GMC; Step 1.2, based on the gray matter image, obtaining the gray matter thickness GMT and gray matter surface area SA corresponding to each brain region of the whole brain; Step 1.3, constructing a first brain map based on the gray matter volume GMV, gray matter concentration GMC, gray matter thickness GMT and gray matter surface area SA; Step 1.4: Calculate the low-frequency amplitude ALFF, fractional low-frequency amplitude fALFF, and local consistency ReHo for the first brain map to obtain the brain thickness of each brain region.

3. The method for classifying brain magnetic resonance imaging based on graph neural network according to claim 1, characterized in that: Hidden Features Sum and adjacency matrix Input to the K-GAT layer, for node and its neighbor nodes , for nodes Update to get updated node features , specifically including the following steps: Step 2.1: Perform linear transformation on the hidden features; Hidden features for each node Perform a linear transformation: ； in, is the weight matrix, Indicates the number of iterations; Step 2.2 When When computing nodes and its neighbor nodes The attention coefficient between : ； in is a learnable attention vector; Represents the splicing operation of features; represents the activation function; Step 2.3, attention coefficient Normalize and get the normalized attention weight ; ； in, Representation node The set of neighbors of Step 2.4, for the node The neighbor node features of node New hidden feature representation of : ； in, is a nonlinear activation function, is the shared weight matrix; Step 2.5: Use multi-head attention mechanism to focus on nodes New hidden features Perform splicing, expressed as: ； in, represents the number of attention heads, Represents a splicing operation, and They are The attention weights and linear transformation matrices of each head; Step 2.6, the node Features Input to the KAN layer, after further feature extraction, let , return to step 2.1, and finally get the characteristics of each node .

4. The method for classifying brain magnetic resonance imaging based on graph neural network according to claim 1, characterized in that Step 2 also includes pruning the trained K-GAT model.

Images