Brain network node hub index detection method and device, equipment and medium

Abstract

The invention provides a brain network node hub index detection method and device, equipment and a medium, and relates to the field of medical image processing. The method comprises the following steps: constructing an undirected brain network and a directed brain network through Pearson's correlation analysis and bivariate Granger causality analysis based on a resting state functional magnetic resonance image; screening out an undirected effective connection graph in the undirected brain network, and screening out a directed effective connection graph in the directed brain network; extracting an average value of the undirected centrality of voxels in the region of interest from the undirected effective connection graph, and taking the average value as the undirected centrality of the region of interest; extracting an average value of the directed out-degree centrality and an average value of the directed in-degree centrality of voxels in the region of interest from the directed effective connection graph, and respectively taking the average value and the average value as the directed out-degree centrality and the directed in-degree centrality of the region of interest; and calculating the role tendency of the region of interest based on the directed out-degree centrality of the region of interest and the directed in-degree centrality of the region of interest.

Description

Technical Field

[0001] This application relates to the field of medical image processing, and in particular to a method, apparatus, computer device, and readable storage medium for detecting pivotal indicators of brain network nodes. Background Technology

[0002] The pathological mechanisms of depression are closely related to abnormal brain network topology. Current research techniques mainly utilize data such as functional magnetic resonance imaging (fMRI) or electroencephalography (EEG) to construct and extract overall brain network features, train classification models to identify characteristics of depression, or predict treatment effects. The core of these techniques lies in analyzing the macroscopic and overall properties of the brain network, thereby playing a role in assisting diagnosis and predicting treatment efficacy.

[0003] The aforementioned existing technologies have significant limitations: they rely too heavily on the overall characteristics of the brain network and neglect targeted analysis of the characteristics of key nodes, resulting in the inability to quantify the potential damage to local nodes during network information integration and transmission. Summary of the Invention

[0004] In view of this, this application provides a method, apparatus, computer device, and readable storage medium for detecting pivotal indicators of brain network nodes to solve the above-mentioned technical problems.

[0005] In a first aspect, embodiments of this application provide a method for detecting pivotal indicators of brain network nodes, including: Based on resting-state functional magnetic resonance imaging, time series of all voxels in the whole brain were extracted. Based on the voxel time series, undirected brain networks and directed brain networks were constructed by Pearson correlation analysis and bivariate Granger causality analysis, respectively. The undirected brain network represents the undirected functional connections between voxels, and the directed brain network represents the directed functional connections between voxels. The undirected functional connections reflect the synchronicity of neural activities between voxels, and the directed functional connections reflect the causal regulatory effects between voxels. Based on a first preset threshold, select undirected effective connection graphs covering the entire brain in an undirected brain network, and based on a second preset threshold, select directed effective connection graphs covering the entire brain in a directed brain network. In the undirected effective connectivity graph, the average value of the undirection centrality of the voxels within the region of interest is extracted as the undirection centrality of the region of interest; In the directed effective connectivity graph, the average value of the directed out-degree centrality and the average value of the directed in-degree centrality of the voxels within the region of interest are extracted, and used as the directed out-degree centrality and the directed in-degree centrality of the region of interest, respectively. Based on the directed out-degree centrality and the directed in-degree centrality of the region of interest, the role tendency of the region of interest is calculated; The undirected centrality, directed out-degree centrality, and role-biasedness of the region of interest are configured as the pivotal indicators of the brain network.

[0006] The method described in the embodiments of this application may also have the following additional technical features: Optionally, in the above technical solution, the method further includes: The pivotal indices of the brain networks of individuals with the target disease were analyzed and compared with those of healthy individuals to determine the statistical results of the differences between the pivotal indices of the brain networks of individuals with the target disease and healthy individuals.

[0007] Optionally, in any of the above technical solutions, the method further includes: The correlation between the pivotal indicators of the brain network of the target patients and their pathological symptoms was determined by analyzing and comparing the pivotal indicators of the brain network of the target patients and their pathological symptoms.

[0008] In any of the above technical solutions, optionally, before constructing the undirected brain network and the directed brain network based on resting-state functional magnetic resonance imaging using Pearson correlation analysis and bivariate Granger causality analysis, respectively, the method further includes: Acquire resting-state functional magnetic resonance images and preprocess the resting-state functional magnetic resonance images; wherein, the preprocessing includes at least one of the following: removing resting-state functional magnetic resonance images from a preset time period, head motion correction, structural image registration to functional image, structural image tissue segmentation, delinearization, regression covariates, standardization, resampling, and bandpass filtering.

[0009] In any of the above technical solutions, optionally, in the undirected effective connectivity graph, the average value of the undirected centrality of voxels within the region of interest is extracted as the undirected centrality of the region of interest, including: Calculate the undirection centrality of each voxel based on the undirected effective connectivity graph; Calculate the average value of the undirected centrality of voxels within the region of interest (ROI), and use this average value as the undirected centrality of the ROI; wherein, the undirected centrality of the ROI is used to characterize the connectivity breadth of the ROI in the undirected brain network.

[0010] In any of the above technical solutions, optionally, in the directed effective connectivity graph, the average value of the directed out-degree centrality and the average value of the directed in-degree centrality of voxels within the region of interest are extracted, and used as the directed out-degree centrality and the directed in-degree centrality of the region of interest, respectively, including: The directed out-degree centrality and directed in-degree centrality of each voxel are calculated based on the directed effective connectivity graph. The average value of the directed out-degree centrality of voxels within the region of interest (ROI) is calculated, and this average value is taken as the directed out-degree centrality of the ROI. The average value of the directed in-degree centrality of voxels within the ROI is also calculated, and this average value is taken as the directed in-degree centrality of the ROI. The directed out-degree centrality of the ROI is used to characterize the information output capability of the ROI in the directed brain network, and the directed in-degree centrality of the ROI is used to characterize the information input capability of the ROI in the directed brain network.

[0011] In any of the above technical solutions, optionally, the role tendency of the region of interest (ROI) is calculated based on its directed out-degree centrality and directed in-degree centrality, including: Based on the directed out-degree centrality and the directed in-degree centrality of the region of interest (ROI), the role tendency of the ROI is calculated according to the first formula; wherein the first formula is:

[0012] Where rDC represents the role orientation of the region of interest, oDC represents the directed out-degree centrality of the region of interest, and iDC represents the directed in-degree centrality of the region of interest.

[0013] Secondly, embodiments of this application provide a device for detecting pivotal indicators of brain network nodes, comprising: The network construction module is used to: extract the time series of all voxels in the whole brain based on resting-state functional magnetic resonance imaging; and construct undirected and directed brain networks based on the voxel time series through Pearson correlation analysis and bivariate Granger causality analysis, respectively. The undirected brain network represents the undirected functional connections between voxels, and the directed brain network represents the directed functional connections between voxels. The undirected functional connections reflect the synchronicity of neural activities between voxels, and the directed functional connections reflect the causal regulatory effects between voxels. The filtering module is used to: filter out undirected effective connection graphs covering the entire brain in an undirected brain network based on a first preset threshold, and to filter out directed effective connection graphs covering the entire brain in a directed brain network based on a second preset threshold. The calculation module is used to: extract the average value of the undirected centrality of voxels within the region of interest (ROI) in an undirected effective connectivity graph, as the undirected centrality of the ROI; extract the average value of the directed out-degree centrality and the average value of the directed in-degree centrality of voxels within the ROI in a directed effective connectivity graph, as the directed out-degree centrality and the directed in-degree centrality of the ROI, respectively; and calculate the role tendency of the ROI based on the directed out-degree centrality and the directed in-degree centrality of the ROI. The indicator configuration module is used to configure the undirected centrality, directed out-degree centrality, and role tendency of the region of interest as the node hub indicators of the brain network.

[0014] Thirdly, embodiments of this application provide a computer device including a processor and a memory, the memory storing programs or instructions that can run on the processor, the programs or instructions implementing the steps of the method as described in the first aspect when executed by the processor.

[0015] Fourthly, embodiments of this application provide a readable storage medium on which a program or instructions are stored, which, when executed by a processor, implement the steps of the method as described in the first aspect.

[0016] This application discloses a method, apparatus, computer device, and readable storage medium for detecting pivotal indicators of brain network nodes. Resting-state functional magnetic resonance imaging (fMRI) images are acquired and preprocessed. Undirected and directed brain networks are constructed using Pearson correlation analysis and bivariate Granger causality analysis, respectively. Thresholds are set to filter out effective undirected / directed connectivity maps across the entire brain. The average value of voxels within a region of interest (ROI) is then extracted as the pivotal indicator for that ROI. This application, by identifying and quantifying the state of key information pivotal nodes in the brain network, enables a more precise characterization of the localization of network damage.

[0017] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0018] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 A flowchart illustrating a method for detecting pivotal indicators of brain network nodes according to an embodiment of this application is shown. Figure 2 This illustration shows one of the brain network node pivotality detection indicators according to an embodiment of this application; Figure 3 This is a second schematic diagram illustrating the brain network node pivotality detection index according to an embodiment of this application; Figure 4 This is the third schematic diagram illustrating the brain network node pivotality detection index according to an embodiment of this application; Figure 5 This illustration shows one of the schematic diagrams of abnormal changes in nodal pivotal detection indicators of the temporal visual cortex in human patients with depression, according to an embodiment of this application; Figure 6 This is the second schematic diagram illustrating abnormal changes in nodal pivotal detection indicators of the temporal visual cortex in human patients with depression, according to an embodiment of this application. Figure 7 This is the third schematic diagram illustrating abnormal changes in nodal pivotal detection indicators of the temporal visual cortex in human patients with depression, according to an embodiment of this application. Figure 8 This illustration shows one of the schematic diagrams relating nodal pivotal detection indicators of the temporal visual cortex to pathological symptoms in human patients with depression according to an embodiment of this application; Figure 9 This is the second schematic diagram illustrating the association between nodal pivotal detection indicators of the temporal visual cortex and pathological symptoms in human patients with depression according to an embodiment of this application; Figure 10 This is the third schematic diagram illustrating the association between nodal pivotal detection indicators of the temporal visual cortex and pathological symptoms in human patients with depression according to an embodiment of this application; Figure 11 A structural block diagram of a brain network node pivotality index detection device according to an embodiment of this application is shown; Figure 12 A structural block diagram of a computer device according to an embodiment of this application is shown. Detailed Implementation

[0019] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0020] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class and the number of objects is not limited; for example, a first object can be one or more. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0021] The following description, in conjunction with the accompanying drawings, details the method, apparatus, computer equipment, and readable storage medium for detecting pivotal indicators of brain network nodes provided in this application, through specific embodiments and application scenarios. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0022] This application provides a method for detecting pivotal indicators of brain network nodes, such as... Figure 1 As shown, the method includes: Step 101: Based on resting-state functional magnetic resonance imaging, extract the time series of all voxels in the whole brain. Based on the voxel time series, construct undirected brain networks and directed brain networks through Pearson correlation analysis and bivariate Granger causality analysis, respectively. The undirected brain network represents the undirected functional connections between voxels, and the directed brain network represents the directed functional connections between voxels. The undirected functional connections reflect the synchronicity of neural activities between voxels, and the directed functional connections reflect the causal regulatory effects between voxels.

[0023] In this step, resting-state functional magnetic resonance imaging (fMRI) images are acquired, and based on these images, undirected and directed brain networks are constructed. The resting-state fMRI images can be those of the brain networks of individuals with the target condition (e.g., patients with depression) or those of healthy individuals.

[0024] In one embodiment, after acquiring resting-state functional magnetic resonance imaging (fMRI) images and before constructing undirected and directed brain networks, the method further includes: preprocessing the resting-state fMRI images; wherein the preprocessing includes at least one of the following: removing resting-state fMRI images from a pre-preset time period (e.g., the first 10 seconds), head motion correction, structural image registration to functional image, structural image tissue segmentation, delinearization, regression of covariates (e.g., Friston-24 head motion parameters, white matter and cerebrospinal fluid signals), standardization, resampling, and bandpass filtering.

[0025] The removal of the first 10 seconds of resting-state functional magnetic resonance imaging (fMRI) images refers to the process where, at the initial stage of the scan, the magnetic field of the MRI equipment is not yet fully stable, and the subject has not yet entered a stable resting state. Removing the data from the first 10 seconds eliminates interference from initial unstable signals. This process removes noise data caused by magnetic field drift and the subject's adaptation period, ensuring the baseline stability of subsequent data analysis.

[0026] Head motion correction refers to the process of aligning functional images at each time point with reference images using registration algorithms (such as rigid body registration) to correct signal deviations caused by minute head movements during scanning. This eliminates artifacts caused by head motion, avoids voxel signal misalignment due to head displacement, and ensures the spatial consistency of time-series data.

[0027] Structural-to-functional image registration refers to the spatial registration of high-resolution T1-weighted structural images with functional images, establishing a spatial correspondence between the two images and providing anatomical reference for subsequent tissue segmentation and standardization. This process enables spatial alignment between structural and functional images, ensuring that subsequent structural-based analyses accurately correspond to functional voxels.

[0028] Structural image tissue segmentation refers to the process of dividing brain tissue into three types of tissues—gray matter, white matter, and cerebrospinal fluid (CSF)—based on the registered structural image and using segmentation algorithms, thereby obtaining masks for pure gray matter, white matter, and CSF regions.

[0029] Delinearization refers to the process of removing slow linear signal trends in functional image time series caused by factors such as device hardware drift and temperature changes through linear fitting methods. This eliminates non-physiological linear drift noise and restores the true brain functional activity time series signal.

[0030] Regression covariates refer to the construction of a regression model that includes Friston-24 head movement parameters, white matter average signal, and cerebrospinal fluid average signal. Removing these covariates from the functional image time series can effectively remove residual head movement noise, non-neuronal spontaneous fluctuation interference from white matter and cerebrospinal fluid, and improve the signal-to-noise ratio of brain functional signals.

[0031] Standardization refers to mapping the functional and structural images of an individual's brain space to a standard brain template space through nonlinear registration, thereby achieving spatial normalization among different subjects. This eliminates differences in brain anatomy between individuals, allowing functional data from different subjects to be statistically analyzed between groups within a unified space.

[0032] Resampling refers to adjusting the registered images to the standard voxel size during the standardization process to ensure consistent spatial resolution across all voxels. By standardizing the image resolution across different subjects and scanning sequences, the data format requirements for subsequent statistical analysis are met.

[0033] Bandpass filtering refers to using filtering algorithms to retain signals in a specific frequency range related to spontaneous brain activity in a functional image time series, while filtering out high-frequency physiological noise (such as heartbeat and respiration) and low-frequency drift noise, further purifying signals related to resting-state brain networks, and reducing the interference of physiological noise on functional connectivity analysis.

[0034] After the above processing steps, the signal-to-noise ratio of resting-state functional magnetic resonance imaging is significantly improved, and non-specific interferences such as head movement, equipment drift, and physiological noise are removed; the normalization of individual data to a standard space is achieved, providing high-quality and comparable standardized data for subsequent studies such as brain functional connectivity analysis.

[0035] Furthermore, for the preprocessed resting-state functional magnetic resonance imaging (fMRI), time series data of all voxels in the whole brain were first extracted. Then, Pearson correlation analysis and bivariate Granger causality analysis were performed on the time series data of each voxel and other voxels in the whole brain, yielding Pearson correlation coefficient (R-value) maps and directed path coefficient maps at the voxel level, respectively. The Pearson correlation coefficient (R-value) map represents the undirected functional connectivity between the voxel and other voxels in the whole brain, serving as an undirected brain network. This undirected functional connectivity reflects the synchronicity of neural activity between the voxel and other voxels. The directed path coefficient map represents the directed functional connectivity between the voxel and other voxels in the whole brain, serving as a directed brain network. This directed functional connectivity reflects the output of the voxel to other voxels and the input received from other voxels, i.e., the causal regulatory effects between voxels.

[0036] Step 102: Based on a first preset threshold, select undirected valid connection graphs covering the entire brain in the undirected brain network, and based on a second preset threshold, select directed valid connection graphs covering the entire brain in the directed brain network.

[0037] In this step, based on the undirected brain network, the R-value threshold (i.e., the first preset threshold) is set to 0.2-0.3 to obtain an undirected effective connection graph, which can also be called a weighted undirected adjacency matrix. Similarly, based on the directed brain network, the directed path coefficient threshold (i.e., the second preset threshold) is set to 0.2-0.3 to obtain a directed effective connection graph, which can also be called a weighted directed adjacency matrix.

[0038] Step 103: In the undirected effective connectivity graph, extract the average value of the undirected centrality of the voxels within the region of interest, and use it as the undirected centrality of the region of interest.

[0039] In one embodiment of this application, in an undirected effective connectivity graph, the average value of the undirected centrality of voxels within the region of interest is extracted as the undirected centrality of the region of interest, including: Calculate the undirection centrality of each voxel based on the undirected effective connectivity graph; Calculate the average value of the undirected centrality of voxels within the region of interest (ROI), and use this average value as the undirected centrality of the ROI; wherein, the undirected centrality of the ROI is used to characterize the connectivity breadth of the ROI in the undirected brain network.

[0040] In this embodiment, based on the undirected effective connectivity graph, the undirected centrality (uDC) of each voxel is calculated. uDC is defined as the number of meaningful positively correlated functional connections between that voxel and all other voxels in the whole brain. For example, as... Figure 2As shown, higher undirected centrality (uDC) indicates a larger number of connections, while lower undirected centrality (uDC) indicates a smaller number of connections. Furthermore, voxel-level uDC maps are generated across the entire brain, and z-score normalization and spatial smoothing are performed. Then, the average uDC of voxels within the region of interest (i.e., the node) is extracted as the uDC of that region of interest. The uDC of the region of interest represents the connection breadth of the node in the undirected brain network.

[0041] Step 104: In the directed effective connectivity graph, extract the average value of the directed out-degree centrality and the average value of the directed in-degree centrality of the voxels within the region of interest, and use them as the directed out-degree centrality and the directed in-degree centrality of the region of interest, respectively.

[0042] In one embodiment of this application, in a directed effective connectivity graph, the average value of the directed out-degree centrality and the average value of the directed in-degree centrality of voxels within the region of interest are extracted, and used as the directed out-degree centrality and the directed in-degree centrality of the region of interest, respectively, including: The directed out-degree centrality and directed in-degree centrality of each voxel are calculated based on the directed effective connectivity graph. The average value of the directed out-degree centrality of voxels within the region of interest (ROI) is calculated, and this average value is taken as the directed out-degree centrality of the ROI. The average value of the directed in-degree centrality of voxels within the ROI is also calculated, and this average value is taken as the directed in-degree centrality of the ROI. The directed out-degree centrality of the ROI is used to characterize the information output capability of the ROI in the directed brain network, and the directed in-degree centrality of the ROI is used to characterize the information input capability of the ROI in the directed brain network.

[0043] In this embodiment, based on the directed effective connectivity graph, the directed out-degree centrality (oDC) of each voxel's output direction is calculated. oDC is defined as the number of meaningful positive-acting functional connections from that voxel to all other voxels in the whole brain. For example, as... Figure 3 As shown, a higher directed out-degree centrality (oDC) indicates a larger number of output connections, while a lower directed out-degree centrality (oDC) indicates a smaller number of output connections. Furthermore, a voxel-level oDC map is generated across the entire brain, and z-score normalization and spatial smoothing are performed. Then, the average oDC of voxels within the region of interest (i.e., the node) is extracted as the oDC value of that region of interest. The oDC value of the region of interest represents the node's ability to output information globally.

[0044] Furthermore, based on the directed effective connectivity graph, the directed in-degree centrality (iDC) of each voxel's output direction is calculated. iDC is defined as the number of meaningful positive functional connections from all other voxels in the brain to that region. For example, a higher iDC indicates a larger number of input connections, while a lower iDC indicates a smaller number of input connections. Subsequently, a voxel-level iDC map is generated across the entire brain, and z-score normalization and spatial smoothing are performed. Then, the average iDC of voxels within a region of interest (i.e., a node) is extracted as the iDC value for that region of interest. The iDC value of a region of interest represents the node's ability to receive global information input.

[0045] Step 105: Calculate the role tendency of the region of interest based on its directed out-degree centrality and directed in-degree centrality.

[0046] In this step, the role bias (rDC) of each region of interest is defined as the difference between the oDC and iDC of that region of interest divided by the sum of the oDC and iDC, as shown in the first formula:

[0047] Where rDC represents the role orientation of the region of interest, oDC represents the directed out-degree centrality of the region of interest, and iDC represents the directed in-degree centrality of the region of interest.

[0048] rDC represents whether a region of interest is biased towards being an information source or an information receiver in the neural activity of brain networks. For example... Figure 4 As shown, a role bias (rDC) > 0 indicates that output exceeds input, favoring the role of information source; a role bias (rDC) < 0 indicates that input exceeds output, favoring the role of information receiver.

[0049] Step 106: Configure the undirected centrality, directed out-degree centrality, and role bias of the region of interest as the node hub index of the brain network.

[0050] In this step, the undirected centrality (uDC), directed out-degree centrality (oDC), and role-biased centrality (rDC) of the region of interest are the pivotal indicators of the brain network. This application quantifies pivotal damage, solving the problem of the lack of specific detection methods in related technologies, and achieving precise quantification of the degree of damage to key pivotal nodes in the brain network of patients with depression.

[0051] In other embodiments, betweenness centrality can be used to reflect the pivotal nature of nodes during the indicator extraction stage after brain network construction. Betweenness centrality is good at identifying key nodes in cross-module communication. However, compared to degree centrality, betweenness centrality tends to overlook local loop center nodes, making it unsuitable for locating local loop targets in depression. Furthermore, the calculation of the indicator is more complex and the results are less stable.

[0052] In one embodiment of this application, the method further includes: analyzing and comparing the pivotal indicators of the brain networks of the target disease patient with the pivotal indicators of the brain networks of healthy patients, and determining the statistical results of the differences between the pivotal indicators of the brain networks of the target disease patient and the brain networks of healthy patients.

[0053] In this embodiment, the undirected centrality (uDC), directed out-degree centrality (oDC), and role bias (rDC) of the region of interest are statistically analyzed between the target disease group and the healthy control group to obtain the magnitude difference results of undirected centrality (uDC), directed out-degree centrality (oDC), and role bias (rDC) between the target disease group and the healthy control group.

[0054] In one embodiment of this application, the method further includes: analyzing and comparing the pivotal indicators of the brain network of the target patient with the pathological symptoms of the target patient, and determining the correlation results between the pivotal indicators of the brain network of the target patient and the pathological symptoms of the target patient.

[0055] In this embodiment, the undirected centrality (uDC), directed out-degree centrality (oDC), and role bias (rDC) indices of the region of interest are correlated with the pathological symptoms of the target patient to establish pathophysiological correlation results.

[0056] This application provides a mechanism for verifying symptom association, which addresses the problem that related technologies have not verified the association between node characteristics and clinical symptoms. It establishes a quantitative model between damage to key hub nodes and the core symptoms of the disease, revealing the physiological significance of the method.

[0057] This application provides a specific embodiment of a method for precise detection of pivotal nodes in the brain network of depression based on resting-state functional magnetic resonance imaging, as follows: (1) Subjects An analysis was conducted on patients with depression and healthy controls.

[0058] (2) Testing process Data acquisition was performed using a whole-body MRI scanner equipped with a 32-channel array head coil. To reduce the effects of noise and head movement, earplugs and foam pads were used for fixation during the scan.

[0059] The scanning sequences included resting-state functional magnetic resonance imaging (fMRI) and structural magnetic resonance imaging (SMRI). Resting-state fMRI used an echo-planar imaging (EPI) sequence with the following parameters: isotropic resolution 1.5 mm, repetition time (TR) 2000 ms, echo time (TE) 20.6 ms, 160 time points, 110 slices, flip angle 70°, and multi-band acceleration factor 5. Subjects kept their eyes closed during the scan, which lasted 6 minutes and 24 seconds. Structural SMRI used the MP2RAGE sequence with the following parameters: isotropic resolution 0.7 mm, TR 5000 ms, flip time TI1 901 ms, TI2 3200 ms, and a scan duration of 8 minutes and 42 seconds.

[0060] (3) Calculation of nodal pivot index Using the human middle temporal visual cortex (hMT+ area) as the region of interest, the undirected centrality (uDC), directed out-degree centrality (oDC), and role bias (rDC) of the region of interest were calculated using the above-mentioned index calculation method of this application.

[0061] (4) Test results Figure 5 , Figure 6 , Figure 7 These are the statistical results of inter-group differences in undirected centrality (uDC), directed out-centrality (oDC), and role bias (rDC) in the temporal visual cortex (hMT+ area) of humans in the embodiments of this application. MDD represents the depressed patient group, HC represents the healthy control group, uDC represents undirected centrality, oDC represents directed out-centrality, and rDC represents role bias. " indicates: P < 0.05, " "P < 0.01" indicates significance. When comparing the pivotal indices of brain network nodes in the hMT+ region between healthy subjects and patients with depression, it was found that the undirected centrality (uDC) of patients with depression was significantly lower than that of healthy subjects, while the directed out-degree centrality (oDC) and role bias (rDC) of patients with depression were significantly higher than those of healthy subjects.

[0062] Figure 8 , Figure 9 , Figure 10 These are the pathological symptom correlation results of undirected centrality (uDC), directed out-centrality (oDC), and role bias (rDC) in the temporal visual cortex (hMT+ area) of humans in the embodiments of this application. MDD represents the depressed patient group, uDC represents undirected centrality, oDC represents directed out-centrality, and rDC represents role bias. " indicates: P < 0.05, " "P < 0.01" indicates significance. Correlation analysis of the pivotal indicators of brain network nodes in the hMT+ region with the pathological symptoms of patients with depression revealed that the undirected centrality (uDC), directed out-degree centrality (oDC), and role bias (rDC) of patients with depression were all significantly negatively correlated with psychomotor retardation scores.

[0063] (5) Summary and analysis of test results 1) In the detection of pivotal nodes in brain networks, taking the human middle temporal visual cortex as an example, it was observed that the undirected centrality (uDC) of depressed patients was significantly lower than that of healthy subjects, while the directed out-degree centrality (oDC) and role bias (rDC) were significantly higher. This indicates that the pivotal nature of the human middle temporal visual cortex is affected by depression, with a significant decrease in global neural activity synchronicity, an increase in local area output to the global level, and this node being more inclined to act as an information source in brain network neural activity. The changes in these three dimensions of pivotality may reflect the important role of the human middle temporal visual cortex in the abnormal brain mechanisms of depression.

[0064] 2) In the association analysis between pivotal role and pathological symptoms, in the temporal visual cortex of humans, the undirected centrality (uDC), directed out-centrality (oDC), and role bias (rDC) of patients with depression all showed a significant negative correlation with psychomotor retardation scores. That is, the lower the pivotal role in the temporal visual cortex of humans, the higher the psychomotor retardation score and the more severe the symptoms. This significant association between pivotal role and psychomotor retardation supports the possibility that the temporal visual cortex in humans with depression makes a specific contribution to psychomotor retardation and may be a key pathological node in this symptom.

[0065] This application proposes a method for detecting the pivotal nature of brain network nodes in a target disease. This method overcomes the limitations of related technologies that only analyze the overall characteristics of the brain network, achieving precise detection and quantification of the degree of pivotal damage to specific key nodes from three dimensions: undirected degree centrality (uDC), directed out-degree centrality (oDC), and role-biased centrality (rDC). Furthermore, it establishes a quantitative mapping relationship between the pivotal nature of key nodes and the clinical symptoms of the target disease, revealing the driving force of node damage on specific symptoms. This provides direct evidence for understanding pathological mechanisms and developing personalized diagnostic and therapeutic targets, promoting a shift in treatment models from "symptom-driven" to "precision-targeted."

[0066] As a specific implementation of the above-mentioned method for detecting pivotal indicators of brain network nodes, this application provides a device for detecting pivotal indicators of brain network nodes. For example... Figure 11 As shown, the detection device 1100 for the pivotal indicators of brain network nodes includes: a network construction module 1101, a screening module 1102, a calculation module 1103, and an indicator configuration module 1104.

[0067] The network construction module 1101 is used to: extract the time series of all voxels in the whole brain based on resting-state functional magnetic resonance imaging; and construct undirected brain networks and directed brain networks based on the voxel time series through Pearson correlation analysis and bivariate Granger causality analysis, respectively. The undirected brain network represents the undirected functional connections between voxels, and the directed brain network represents the directed functional connections between voxels. The undirected functional connections reflect the synchronicity of neural activities between voxels, and the directed functional connections reflect the causal regulatory effects between voxels. The filtering module 1102 is used to: filter out undirected effective connection graphs of the whole brain in an undirected brain network based on a first preset threshold, and filter out directed effective connection graphs of the whole brain in a directed brain network based on a second preset threshold. The calculation module 1103 is used to: extract the average value of the undirected centrality of voxels within the region of interest in an undirected effective connectivity graph, as the undirected centrality of the region of interest; extract the average value of the directed out-degree centrality and the average value of the directed in-degree centrality of voxels within the region of interest in a directed effective connectivity graph, as the directed out-degree centrality and the directed in-degree centrality of the region of interest, respectively; and calculate the role tendency of the region of interest based on the directed out-degree centrality and the directed in-degree centrality of the region of interest. The indicator configuration module 1104 is used to configure the undirected centrality, directed out-degree centrality, and role tendency of the region of interest as the node hub indicators of the brain network.

[0068] Furthermore, the device also includes: The difference statistics module is used to analyze and compare the pivotal indicators of the brain networks of individuals with target diseases with those of healthy individuals, and to determine the statistical results of the differences between the pivotal indicators of the brain networks of individuals with target diseases and those of healthy individuals.

[0069] Furthermore, the device also includes: The correlation analysis module is used to analyze and compare the pivotal indicators of the brain network of the target patient with the pathological symptoms of the target patient, and to determine the correlation results between the pivotal indicators of the brain network of the target patient and the pathological symptoms of the target patient.

[0070] Furthermore, the device also includes: The preprocessing module is used to: acquire resting-state functional magnetic resonance images and preprocess the resting-state functional magnetic resonance images; wherein, the preprocessing includes at least one of the following: removing resting-state functional magnetic resonance images from a preset time period, head motion correction, structural image registration to functional image, structural image tissue segmentation, delinearization, regression covariates, standardization, resampling, and bandpass filtering.

[0071] Furthermore, the calculation module 1103 is specifically used for: Calculate the undirection centrality of each voxel based on the undirected effective connectivity graph; Calculate the average value of the undirected centrality of voxels within the region of interest (ROI), and use this average value as the undirected centrality of the ROI; wherein, the undirected centrality of the ROI is used to characterize the connectivity breadth of the ROI in the undirected brain network.

[0072] Furthermore, the calculation module 1103 is specifically used for: The directed out-degree centrality and directed in-degree centrality of each voxel are calculated based on the directed effective connectivity graph. The average value of the directed out-degree centrality of voxels within the region of interest (ROI) is calculated, and this average value is taken as the directed out-degree centrality of the ROI. The average value of the directed in-degree centrality of voxels within the ROI is also calculated, and this average value is taken as the directed in-degree centrality of the ROI. The directed out-degree centrality of the ROI is used to characterize the information output capability of the ROI in the directed brain network, and the directed in-degree centrality of the ROI is used to characterize the information input capability of the ROI in the directed brain network.

[0073] Furthermore, the calculation module 1103 is specifically used for: Based on the directed out-degree centrality and the directed in-degree centrality of the region of interest (ROI), the role tendency of the ROI is calculated according to the first formula; wherein the first formula is:

[0074] Where rDC represents the role orientation of the region of interest, oDC represents the directed out-degree centrality of the region of interest, and iDC represents the directed in-degree centrality of the region of interest.

[0075] The brain network node pivotality indicator detection device 1100 in this application embodiment can be a computer device or a component within a computer device, such as an integrated circuit or a chip. The brain network node pivotality indicator detection device 1100 provided in this application embodiment can achieve... Figure 1 The various processes implemented in the method embodiment for detecting pivotal indicators of brain network nodes will not be described again here to avoid repetition.

[0076] This application also provides a computer device, such as... Figure 12As shown, the computer device 1200 includes a processor 1201 and a memory 1202. The memory 1202 stores a program or instruction that can run on the processor 1201. When the program or instruction is executed by the processor 1201, it implements the various steps of the above-described method embodiment for detecting the pivotal indicators of brain network nodes and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0077] The memory 1202 can be used to store software programs and various data. The memory 1202 may primarily include a first storage area for storing programs or instructions and a second storage area for storing data. The first storage area may store the operating system, application programs or instructions required for at least one function (such as sound playback, image playback, etc.). Furthermore, the memory 1202 may include volatile memory or non-volatile memory, or both. The non-volatile memory may be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory can be random access memory (RAM), static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct memory bus RAM (DRRAM). The memory 1202 in this embodiment includes, but is not limited to, these and any other suitable types of memory.

[0078] Processor 1201 may include one or more processing units; optionally, processor 1201 integrates an application processor and a modem processor, wherein the application processor mainly handles operations involving the operating system, user interface, and applications, and the modem processor mainly handles wireless communication signals, such as a baseband processor. It is understood that the aforementioned modem processor may also not be integrated into processor 1201.

[0079] This application also provides a readable storage medium storing a program or instructions. When the program or instructions are executed by a processor, they implement the various processes of the above-described method for detecting pivotal indicators of brain network nodes and achieve the same technical effect. To avoid repetition, they will not be described again here.

[0080] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. Furthermore, it should be noted that the scope of the methods and apparatuses in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0081] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

Claims

1. A method for detecting a brain network node hubness index, characterized in that, The method comprises the following steps: Based on resting-state functional magnetic resonance images, the time series of all voxels in the whole brain are extracted, and based on the time series of the voxels, the undirected brain network and the directed brain network are constructed by Pearson correlation analysis and bivariate Granger causality analysis respectively; the undirected brain network represents the undirected functional connection between voxels, and the directed brain network represents the directed functional connection between voxels; the undirected functional connection reflects the synchronism of neural activities between voxels, and the directed functional connection reflects the causal regulation between voxels; Based on the first preset threshold, the undirected effective connection graph in the whole brain is screened out in the undirected brain network, and based on the second preset threshold, the directed effective connection graph in the whole brain is screened out in the directed brain network; In the undirected effective connection graph, the average value of the undirected degree centrality of the voxels in the region of interest is extracted as the undirected degree centrality of the region of interest; In the directed effective connection graph, the average value of the directed out-degree centrality and the average value of the directed in-degree centrality of the voxels in the region of interest are extracted as the directed out-degree centrality of the region of interest and the directed in-degree centrality of the region of interest respectively; Based on the directed out-degree centrality of the region of interest and the directed in-degree centrality of the region of interest, the role tendency of the region of interest is calculated; The undirected degree centrality, the directed out-degree centrality and the role tendency of the region of interest are configured as the node hubness index of the brain network.

2. The method of claim 1, wherein, The method further comprises: The node hubness index of the brain network of the target disease personnel is analyzed and compared with the node hubness index of the brain network of the healthy personnel, and the difference statistical result of the node hubness index of the brain network of the target disease personnel and the node hubness index of the brain network of the healthy personnel is determined.

3. The method of claim 1, wherein, The method further comprises: The node hubness index of the brain network of the target disease personnel is analyzed and compared with the pathological symptoms of the target disease personnel, and the correlation result of the node hubness index of the brain network of the target disease personnel and the pathological symptoms of the target disease personnel is determined.

4. The method of claim 1, wherein, Before constructing the undirected brain network and the directed brain network based on the resting-state functional magnetic resonance images by Pearson correlation analysis and bivariate Granger causality analysis respectively, the method further comprises: Collecting the resting-state functional magnetic resonance images and preprocessing the resting-state functional magnetic resonance images; wherein the preprocessing comprises at least one of the following: removing the resting-state functional magnetic resonance images of the first preset time period, head motion correction, registering the structural image to the functional image, structural image organization segmentation, linear trend removal, regression covariate, standardization, resampling and band pass filtering.

5. The method of claim 1, wherein, In the undirected effective connection graph, the average value of the undirected degree centrality of the voxels in the region of interest is extracted as the undirected degree centrality of the region of interest, which comprises: Based on the undirected effective connection graph, the undirected degree centrality of each voxel is calculated; The average value of the undirected degree centrality of the voxels in the region of interest is calculated, and the average value is taken as the undirected degree centrality of the region of interest; wherein the undirected degree centrality of the region of interest is used to represent the connection breadth of the region of interest in the undirected brain network.

6. The method of claim 1, wherein, In the directed effective connection graph, the average of the directed out-degree centrality of the voxels in the region of interest and the average of the directed in-degree centrality of the voxels in the region of interest are extracted as the directed out-degree centrality of the region of interest and the directed in-degree centrality of the region of interest, respectively, comprising: calculating the directed out-degree centrality and the directed in-degree centrality of each voxel based on the directed effective connection graph; calculating the average of the directed out-degree centrality of the voxels in the region of interest and taking the average as the directed out-degree centrality of the region of interest, and calculating the average of the directed in-degree centrality of the voxels in the region of interest and taking the average as the directed in-degree centrality of the region of interest; wherein the directed out-degree centrality of the region of interest is used to represent the information output capability of the region of interest in the directed brain network, and the directed in-degree centrality of the region of interest is used to represent the information input capability of the region of interest in the directed brain network.

7. The method of claim 1, wherein, Based on the directed out-degree centrality of the region of interest and the directed in-degree centrality of the region of interest, the role tendency of the region of interest is calculated, comprising: Based on the directed out-degree centrality of the region of interest and the directed in-degree centrality of the region of interest, the role tendency of the region of interest is calculated according to a first formula; wherein the first formula is: wherein rDC represents the role tendency of the region of interest, oDC represents the directed out-degree centrality of the region of interest, and iDC represents the directed in-degree centrality of the region of interest.

8. A device for detecting a brain network node hubness index, comprising: comprising: a network construction module, configured to: extract time series of all voxels in the whole brain based on resting-state functional magnetic resonance images, and construct an undirected brain network and a directed brain network based on the time series of the voxels through Pearson correlation analysis and bivariate Granger causality analysis respectively; the undirected brain network represents undirected functional connections between voxels, and the directed brain network represents directed functional connections between voxels; the undirected functional connections reflect the synchronicity of neural activities between voxels, and the directed functional connections reflect the causal regulation between voxels; a screening module, configured to: screen an undirected effective connection graph in the whole brain in the undirected brain network based on a first preset threshold, and screen a directed effective connection graph in the whole brain in the directed brain network based on a second preset threshold; a calculation module, configured to: in the undirected effective connection graph, extract the average of the undirected degree centrality of the voxels in the region of interest as the undirected degree centrality of the region of interest; and in the directed effective connection graph, extract the average of the directed out-degree centrality of the voxels in the region of interest and the average of the directed in-degree centrality of the voxels in the region of interest as the directed out-degree centrality of the region of interest and the directed in-degree centrality of the region of interest respectively; based on the directed out-degree centrality of the region of interest and the directed in-degree centrality of the region of interest, the role tendency of the region of interest is calculated; an index configuration module, configured to: configure the undirected degree centrality, the directed out-degree centrality and the role tendency of the region of interest as a node hub index of the brain network.

9. A computer device, comprising: comprising a processor and a memory, the memory storing programs or instructions running on the processor, the programs or instructions being executed by the processor to implement the steps of the brain network node hub index detection method according to any one of claims 1 to 7.

10. A readable storage medium, on which a program or instructions are stored, characterized in that, The program or the instruction realizes the steps of the brain network node hubness index detection method as claimed in any one of claims 1 to 7 when executed by the processor.

Images