Brain structure-function-metabolism network correlation analysis method and system based on multi-modal synchronous imaging

Abstract

The invention provides a brain structure-function-metabolism network correlation analysis method and system based on multi-modal synchronous imaging, and relates to the technical field of brain network analysis, and the method comprises the steps: obtaining multi-modal data of a target subject by using nuclear magnetic photoelectric integrated equipment; performing modal specificity preprocessing on each piece of multi-modal data to obtain preprocessed data; determining a connection matrix of different brain regions according to the preprocessed data; and analyzing the mutual relation among different networks based on the connection matrix to obtain a correlation result of the structure network, the function network and the metabolism network. According to the invention, efficient integration and analysis of brain structure, function and metabolic network data are realized by using a multi-modal synchronous imaging technology, the mutual relation can be deeply revealed, and a brand new research normal form and technical means are provided for research and clinical evaluation of neurological diseases.

Description

Technical Field

[0001] The present invention relates to the technical field of brain network analysis, and particularly to a method and system for analyzing the association between brain structure-functional-metabolic networks based on multimodal synchronous imaging. Background Art

[0002] Brain activity is a complex and orderly process, involving the conduction of neuronal electrical activity, blood oxygen supply, and material metabolism (such as glucose decomposition). The pathogenesis of many neuropsychiatric diseases (such as schizophrenia and depression) is closely related to changes in brain networks. The analysis of brain networks mainly focuses on structural networks, functional networks, and metabolic networks. The analysis of structural networks is usually obtained by using diffusion tensor imaging (DTI) in MRI, and the potential structural connections of the human brain are inferred through the tracking of white matter fiber bundles. Functional connectivity networks mainly rely on technologies such as functional magnetic resonance imaging (fMRI), electroencephalogram (EEG), and functional near-infrared spectroscopy (fNIRS). All three technologies reflect the functional connectivity relationship of the brain by detecting the temporal correlation of functional activities in different brain regions, and their technical principles are different. fMRI measures a blood oxygenation level-dependent (BOLD) signal, which is related to both changes in cerebral blood oxygen and blood flow; EEG captures the electrical signals emitted by neurons to reflect the electrical activity of the brain; while fNIRS can directly detect changes in cerebral blood oxygen levels to evaluate brain activity. In addition, the brain metabolic network can be studied by dynamic positron emission tomography (fPET) technology to study the metabolic connection relationship between brain regions.

[0003] There is a close relationship between different modal networks. The structural network provides a physical basis for the functional network, and the activities of the functional network often depend on the corresponding structural connection relationship, but the two are not exactly the same. The functional network involves the cooperative activities between brain regions, and these activities require energy support, while the metabolic network is responsible for providing and regulating this energy. Brain metabolic activity affects the efficiency of brain function, and conversely, the activities of the functional network can also regulate metabolic demands.

[0004] The current technical bottleneck in brain science research is that due to the limitations of synchronous acquisition equipment, most studies can only achieve synchronous acquisition and analysis of two-modal data. For example, the correlation study of brain functional networks of synchronous fMRI and EEG. Although some studies have attempted to integrate multi-modal data collected at different time points for inter-network correlation analysis, due to the significant differences in the functional states and metabolic levels of the human brain at different time points, this analysis method of non-synchronous acquisition data is difficult to accurately reflect the internal relationships between various physiological processes of the brain. Summary of the Invention

[0005] In order to overcome the deficiencies of the prior art, the object of the present invention is to provide a method and system for analyzing the association of brain structure-function-metabolism networks based on multi-modal synchronous imaging, which uses multi-modal synchronous imaging technology to achieve efficient integration and analysis of brain structure, function, and metabolism network data, can deeply reveal their mutual relationships, and provides a new research paradigm and technical means for the research and clinical evaluation of neurological diseases.

[0006] To achieve the above object, the present invention provides the following solutions: A method for analyzing the association of brain structure-function-metabolism networks based on multi-modal synchronous imaging, comprising: Obtaining multi-modal data of a target subject by using a nuclear magnetic-photoelectric integrated device; the multi-modal data includes positron emission tomography data, magnetic resonance imaging data, electroencephalogram data, eye tracker data, and near-infrared spectroscopy imaging data; Performing modality-specific preprocessing on each of the multi-modal data to obtain preprocessed data; Determining the connection matrix of different brain regions according to the preprocessed data; Analyzing the mutual relationships between different networks based on the connection matrix to obtain the association results of the structure network, function network, and metabolism network.

[0007] Preferably, obtaining multi-modal data of a target subject by using a nuclear magnetic-photoelectric integrated device includes: Collecting T1-weighted structural images and diffusion tensor imaging data; Using 18 F-FDG tracer for dynamic data collection to obtain the positron emission tomography data; While performing dynamic data collection, starting synchronous data recording of the electroencephalogram data, the eye tracker data, the near-infrared spectroscopy imaging data, and the resting-state fMRI data to obtain the final multi-modal data; the magnetic resonance imaging data includes: resting-state fMRI data, T1-weighted structural images, and diffusion tensor imaging data.

[0008] Preferably, perform modality-specific preprocessing on each of the multimodal data to obtain preprocessed data, including: Perform 1Hz high-pass filtering, bad channel processing, downsampling, ICA denoising, rereferencing, source analysis, and eye movement data regression on the electroencephalogram data to obtain the preprocessed data corresponding to the electroencephalogram data; Perform head motion correction, eddy current correction, tractography, and spatial normalization on the diffusion tensor imaging data to obtain the preprocessed data corresponding to the diffusion tensor imaging data; Perform temporal correction, head motion correction, spatial normalization, nuisance regression, detrending, spatial smoothing, filtering, and eye movement data regression on the resting-state fMRI data to obtain the preprocessed data corresponding to the resting-state fMRI data; Perform artifact removal, filtering, blood oxygenation calculation, source analysis, and eye movement data regression on the near-infrared spectroscopy imaging data to obtain the preprocessed data corresponding to the near-infrared spectroscopy imaging data; Perform SUV calculation, spatial normalization, spatial smoothing, SUVr calculation, and eye movement data regression on the positron emission tomography data to obtain the preprocessed data corresponding to the positron emission tomography data.

[0009] Preferably, the connectivity matrix includes a functional connectivity matrix, a structural connectivity matrix, and a metabolic connectivity matrix.

[0010] Preferably, the method for determining the structural connectivity matrix includes: Extract the number of fiber bundles between each pair of brain regions in the preprocessed data corresponding to the diffusion tensor imaging data of each target subject; Calculate the average number of fiber bundles between the target subjects based on the number of fiber bundles to form an initial connectivity matrix of 78×78; Perform normalization processing on the initial connectivity matrix of each target subject by dividing the elements of each initial connectivity matrix by the total number of fibers in the whole brain to obtain a normalized matrix; Average the normalized matrices of all the target subjects to obtain the final structural connectivity matrix.

[0011] Preferably, the method for determining the functional connectivity matrix includes: Extract the time series of all voxels from the preprocessed data corresponding to the positron emission tomography data, the magnetic resonance imaging data, the electroencephalogram data, the eye tracker data, and the near-infrared spectroscopy imaging data to form the average time series of each brain region; Calculate the average value of all voxels in each brain region of the average time series and use each of the average values as the average time curve of each brain region; Based on the average time curves of each brain region, Pearson correlation analysis is performed on the average time curves of every two brain regions to generate a functional connectivity matrix.

[0012] Preferably, based on the connectivity matrix, the mutual relationships between different networks are analyzed to obtain the correlation results of the structural network, functional network, and metabolic network, including: Calculating the Spearman correlation coefficient between networks; the Spearman correlation coefficient The calculation formula is: ; where and are respectively the ranks of each point in a connectivity matrix and another connectivity matrix in the connectivity matrix, represents the average value of the ranks in one connectivity matrix; represents the average value of the ranks in another connectivity matrix; Using the Spearman correlation coefficient to determine the correlation results.

[0013] A brain structure-functional-metabolic network correlation analysis system based on multimodal synchronous imaging, comprising: A multimodal data acquisition unit for acquiring multimodal data of a target subject by using a nuclear magnetic photoelectric integrated device; the multimodal data includes positron emission tomography data, magnetic resonance imaging data, electroencephalogram data, eye tracker data, and near-infrared spectroscopy imaging data; A data preprocessing unit for performing modality-specific preprocessing on each of the multimodal data to obtain preprocessed data; A matrix determination unit for determining a connectivity matrix of different brain regions according to the preprocessed data; A result determination unit for analyzing the mutual relationships between different networks based on the connectivity matrix to obtain the correlation results of the structural network, functional network, and metabolic network.

[0014] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects: The present invention provides a method and system for analyzing the association between brain structure-function-metabolism networks based on multimodal synchronous imaging. The method includes: acquiring multimodal data of a target subject using a nuclear magnetic optoelectronic integrated device; the multimodal data includes positron emission tomography data, magnetic resonance imaging data, electroencephalogram data, eye tracker data, and near-infrared spectroscopy imaging data; respectively performing modality-specific preprocessing on each of the multimodal data to obtain preprocessed data; determining connection matrices of different brain regions according to the preprocessed data; and analyzing the mutual relationships between different networks based on the connection matrices to obtain the association results of the structural network, functional network, and metabolic network. The present invention uses multimodal synchronous imaging technology to achieve efficient integration and analysis of brain structure, function, and metabolic network data, can deeply reveal their mutual relationships, and provides a new research paradigm and technical means for the research and clinical evaluation of neurological diseases. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required to be used in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0016] Figure 1 It is a flowchart of the method provided by the embodiment of the present invention; Figure 2 It is the first correlation analysis schematic diagram provided by the embodiment of the present invention; Figure 3 It is the second correlation analysis schematic diagram provided by the embodiment of the present invention; Figure 4 It is the third correlation analysis schematic diagram provided by the embodiment of the present invention; Figure 5 It is the fourth correlation analysis schematic diagram provided by the embodiment of the present invention; Figure 6 It is the fifth correlation analysis schematic diagram provided by the embodiment of the present invention; Figure 7 It is the system structure schematic diagram provided by the embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0017] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0018] The object of the present invention is to provide a method and system for analyzing the association of brain structure-function-metabolism networks based on multimodal synchronous imaging, which uses multimodal synchronous imaging technology to achieve efficient integration and analysis of brain structure, function, and metabolism network data, can deeply reveal their mutual relationships, and provides a new research paradigm and technical means for the research and clinical evaluation of neurological diseases.

[0019] To make the above objects, features, and advantages of the present invention more obvious and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0020] Figure 1 The flowchart of the method provided by the embodiment of the present invention is as Figure 1 shown. The present invention provides a method for analyzing the association of brain structure-function-metabolism networks based on multimodal synchronous imaging, including: Step 100: Obtain multimodal data of a target subject using a nuclear magnetic-photoelectric integrated device; the multimodal data includes positron emission tomography data, magnetic resonance imaging data, electroencephalogram data (EEG), eye tracker data (ET), and near-infrared spectroscopy imaging data (NIRS); Step 200: Perform modality-specific preprocessing on each piece of multimodal data to obtain preprocessed data; Step 300: Determine the connection matrix of different brain regions according to the preprocessed data; Step 400: Analyze the mutual relationships between different networks based on the connection matrix to obtain the association results of the structural network, functional network, and metabolic network.

[0021] Preferably, obtaining the multimodal data of the target subject using a nuclear magnetic-photoelectric integrated device includes: Collecting T1-weighted structural images and diffusion tensor imaging data; Using 18 an 18F-FDG tracer for dynamic data acquisition to obtain the positron emission tomography data; While performing dynamic data acquisition, starting the synchronous data recording of the electroencephalogram data, the eye tracker data, the near-infrared spectroscopy imaging data, and the resting-state fMRI data to obtain the final multimodal data; the magnetic resonance imaging data includes: resting-state fMRI data, T1-weighted structural images, and diffusion tensor imaging data.

[0022] In the first step of this embodiment, data acquisition is performed, and five-modal data are synchronously collected using a nuclear magnetic-photoelectric integrated real-time synchronous brain imaging device (PMEEN).

[0023] (1) For positron emission tomography data (PET), an 18F-FDG tracer is used for dynamic PET data acquisition.

[0024] (2) Magnetic resonance imaging data (MRI) includes T1-weighted structural images, DTI (diffusion tensor imaging data), and resting-state fMRI data. The purpose of acquiring T1-weighted imaging is to exclude anatomical abnormalities and perform spatial normalization. Given the relative stability of the white matter fiber bundle structure, the T1 structural image and DTI scans are preferentially completed.

[0025] After completing the structural MRI scans (T1, DTI), the multi-modal joint acquisition of dynamic PET (fPET), resting-state fMRI, EEG, ET, and fNIRS is synchronously initiated.

[0026] Preferably, modal-specific preprocessing is performed on each of the multi-modal data to obtain preprocessed data, including: Performing 1Hz high-pass filtering, bad channel processing, downsampling, ICA denoising, re-referencing, source analysis, and eye movement data regression on the electroencephalogram data to obtain the preprocessed data corresponding to the electroencephalogram data; Performing head motion correction, eddy current correction, fiber bundle tracking, and spatial normalization on the diffusion tensor imaging data to obtain the preprocessed data corresponding to the diffusion tensor imaging data; Performing temporal correction, head motion correction, spatial normalization, nuisance regression, detrending, spatial smoothing, filtering, and eye movement data regression on the resting-state fMRI data to obtain the preprocessed data corresponding to the resting-state fMRI data; Performing artifact removal, filtering, blood oxygen calculation, source analysis, and eye movement data regression on the near-infrared spectroscopy imaging data to obtain the preprocessed data corresponding to the near-infrared spectroscopy imaging data; Performing SUV calculation, spatial normalization, spatial smoothing, SUVr calculation, and eye movement data regression on the positron emission tomography data to obtain the preprocessed data corresponding to the positron emission tomography data.

[0027] 2. Modal-specific preprocessing is performed based on the fPET, MRI data (T1, DTI, fMRI), EEG, ET, and fNIRS data in Step 1: The data of different modalities collected are preprocessed. For EEG data, the scalp potential distribution is mapped to the cortical source space through inverse modeling; specifically, for DTI data, motion correction, eddy current correction, fiber tractography, and spatial normalization are performed; for resting-state fMRI data, temporal correction, motion correction, spatial normalization, nuisance regression, detrending, spatial smoothing, and filtering are performed; for EEG data, filtering, interpolation of bad channels, rejection of bad segments, downsampling, ICA denoising, rereferencing, and source analysis are performed; for fNIRS data, artifact removal, filtering, Beer Lambert calculation, and source analysis are performed; for fPET data, standardized uptake value calculation, spatial normalization, spatial smoothing, and Standard Uptake Value Ratio (SUVr) calculation are performed. Among them, the spatial normalization in DTI, fMRI, and fPET refers to registering the spatial positions of the original data to the standard brain template space (such as MNI152), and then using 78 cortical brain regions of the AAL (Automated Anatomical Labeling) atlas as analysis nodes. EEG and fNIRS data are segmented into 78 brain regions using source analysis. Subsequently, for BOLD-fMRI, EEG, fNIRS, and fPET data, eye movement data regression is performed to eliminate the interference of fixation behavior on the brain function and metabolic network connectivity.

[0028] Specifically, for DTI data, it is processed using the FSL toolbox. It includes: (1) Head motion correction: Rigid body transformation is used to compensate for the head displacement during scanning; (2) Eddy current correction: Correct the geometric distortion caused by magnetic field inhomogeneity; (3) Fiber tractography: Apply a deterministic algorithm (such as FACT) to reconstruct the white matter fiber pathway (4) Spatial normalization: The data is first registered from the individual space to the T1 data, and then registered to the MNI152 standard space.

[0029] Furthermore, for resting-state fMRI data, it is processed using the DPARSF toolbox. It includes: (1) Temporal correction: Correct the slice acquisition time difference with the middle layer as the reference; (2) Head motion correction: Eliminate the head motion artifacts through rigid body transformation; (3) Spatial normalization: The individual fMRI data is first rigidly registered to the T1 image, and then nonlinearly registered to the MNI152 space; (4)Interference regression: Remove the influence of white matter, cerebrospinal fluid, and global signal; (5)Detrending: Eliminate the linear trend introduced by scanner drift; (6)Spatial smoothing: Use an 8mm FWHM Gaussian kernel to improve the signal-to-noise ratio; (7)Filtering: Retain the low-frequency oscillation signal of 0.01 - 0.1Hz.

[0030] Furthermore, for EEG data, the EEGLAB toolbox is used for processing. It includes: (1)1Hz high-pass filtering: Eliminate baseline drift; (2)Bad channel processing: Repair abnormal channels by interpolating adjacent electrodes; (3)Downsampling: Reduce the sampling rate to 200Hz to reduce the computational load; (4)ICA denoising: Identify and remove artifact components such as eye movement and electromyogram; (5)Re-reference: Convert to the average reference electrode; (6)Source analysis: Use the sLORETA method to reconstruct cortical source activity.

[0031] Even further, for fNIRS data, the FC-NIRS toolbox is used for processing. Specifically, it includes: (1)Artifact removal: Detect and remove motion artifacts by moving standard deviation; (2)Filtering: Use a 0.01 - 0.2Hz band-pass filter to retain the functional blood flow oscillation signal; (3)Blood oxygen calculation: Apply the modified Beer-Lambert law to convert HbO2 / HbR concentration; (4)Source analysis: Invert the neural activity source based on the finite element model.

[0032] Optionally, for fPET data, the SPM toolbox is used for processing. Specifically, it includes: (1)SUV calculation: Quantify the standardized uptake value; (2)Spatial normalization: Register to the MNI152 standard space; (3)Spatial smoothing: Smooth with an 8mm FWHM Gaussian kernel; (4)SUVr calculation: Standardize with the cerebellum as the reference region.

[0033] Optionally, cross-modal normalization and eye movement correction are also performed in this embodiment, as follows: (1)Spatial unification: Map all modal data to 78 cortical brain regions of the AAL atlas; (2)Eye movement data regression: The fixation coordinates and pupil diameter data are obtained through an eye tracker. After removing outliers and filtering, the velocity-threshold algorithm is used to classify fixation events; the fixation point dispersion, pupil diameter change, and saccade events are extracted as stability indicators; a generalized linear model (GLM) is constructed to regress the variance related to eye movement. (3)Output the preprocessed data for subsequent analysis.

[0034] Preferably, the connection matrix includes a functional connection matrix, a structural connection matrix, and a metabolic connection matrix.

[0035] Preferably, the method for determining the structural connection matrix includes: Extract the number of fiber bundles between each pair of brain regions in the preprocessed data corresponding to the diffusion tensor imaging data of each target subject. Calculate the average number of fiber bundles between the target subjects based on the number of fiber bundles to form an initial connection matrix of 78×78. Normalize the initial connection matrix of each target subject by dividing the elements of each initial connection matrix by the total number of fibers in the whole brain to obtain a normalized matrix. Average the normalized matrices of all the target subjects to obtain the final structural connection matrix.

[0036] Optionally, the structural connection matrix (Msc) of this embodiment is constructed as follows: For DTI data, for a certain subject, use the above fiber number tracking to obtain the number of fiber bundles between every two brain regions in 78 brain regions of this subject (78×78 matrix). Then divide by the total number of fibers in the whole brain of this subject (normalize), and average the matrices obtained from all subjects to obtain the structural connection matrix Msc between brain regions.

[0037] Preferably, the method for determining the functional connection matrix includes: Extract the time series of all voxels from the preprocessed data corresponding to the positron emission tomography data, the magnetic resonance imaging data, the electroencephalogram data, the eye tracker data, and the near-infrared spectroscopy imaging data to form the average time series of each brain region. Calculate the average value of all voxels in each brain region in the average time series, and use each average value as the average time curve of each brain region. Based on the average time curves of each brain region, perform Pearson correlation analysis on the average time curves of every two brain regions to generate a functional connection matrix.

[0038] Furthermore, for fMRI, EEG, fNIRS, and fPET data, for a certain subject, MATLAB was used to extract the time series of all voxels in 78 brain regions, and the average value of all voxels in each brain region was calculated as the average time curve of the brain region. Subsequently, the Pearson correlation analysis of the time curves of every two brain regions was performed using the average time curve of each brain region to generate a functional connectivity matrix. The functional connectivity matrix of the subjects was averaged to obtain M fc-fmri 、M fc-eeg , Mf c-fnirs and M mc .

[0039] Among them, Pearson correlation analysis The formula is as follows:

[0040] in, represents the signal intensity at the i-th time point in brain region X, represents the average signal intensity of the time curve of brain area X; represents the signal intensity of brain region Y at the i-th time point, Represents the average signal intensity of the time curve of brain region Y. n represents all the acquired time points.

[0041] Preferably, the relationships between different networks are analyzed based on the connection matrix to obtain the correlation results of the structural network, the functional network and the metabolic network, including: Calculate the Spearman correlation coefficient between networks; the Spearman correlation coefficient The calculation formula is: ;in, and are the positions of each point in a connection matrix and another connection matrix respectively. represents the average value of the positions in a connection matrix; represents the average value of the positions in another connection matrix; The association result is determined using the Spearman correlation coefficient.

[0042] Specifically, based on the above connection matrix, this embodiment performs correlation analysis and calculates the Spearman correlation coefficient between different networks to analyze the relationship between networks. The value range of the Spearman correlation coefficient is -1 to 1. Generally speaking, 0.8 and above: strong correlation; 0.6 to 0.8: moderate correlation; 0.4 to 0.6: weak correlation; below 0.4: basically no correlation.

[0043] Optionally, in this embodiment, the correlations between the three functional networks obtained by calculating fMRI, EEG, and fNIRS are calculated to reveal the differences in brain functional network construction among different technologies. When M fc-fmri and M fc-eeg the correlation coefficient ρ between them is > 0.8, it indicates a strong correlation between the functional networks constructed by fMRI and EEG; when 0.6 < ρ < 0.8, it indicates a medium correlation; when 0.4 < ρ < 0.6, it indicates a weak correlation; when ρ < 0.4, it indicates basically no correlation. Similarly, the relationships between the EEG and fNIRS, and fMRI and fNIRS functional networks can be obtained. In addition, when the correlation coefficient between M fc-fmri and M fc-eeg is significantly higher than that between M fc-fmri and M fc-fnirs , it indicates that the association between the functional networks constructed by fMRI and EEG is stronger than that between fMRI and fNIRS.

[0044] Furthermore, in this embodiment, the correlations between the structural connection network, functional connection network, and metabolic connection network are calculated to reveal the mutual associations among structure, function, and metabolism. The correlation coefficients between M fc-fmri and M sc , and between M fc-fmri and M mc are calculated respectively. When the Spearman correlation coefficient ρ between M fc-fmri and M sc is > 0.8, it indicates a strong correlation between the fMRI functional connection network and the structural connection network; when 0.6 < ρ < 0.8, it indicates a medium correlation; when 0.4 < ρ < 0.6, it indicates a weak correlation; when ρ < 0.4, it indicates basically no correlation. Similarly, the relationships between the fMRI functional connection network and the metabolic network, and between the metabolic network and the structural network can be obtained.

[0045] As Figure 2 shown, the anatomical structure connection is significantly correlated with the magnetic resonance brain functional connection, and the correlation coefficient is 0.37. As Figure 3 shown, the magnetic resonance functional connection is significantly correlated with the positron emission tomography metabolic parameters, and the correlation coefficient is 0.39. As Figure 4 shown, the anatomical structure connection is significantly correlated with the positron emission tomography metabolic parameters, and the correlation coefficient is 0.38. As Figure 5 shown, the electroencephalogram functional connection is significantly correlated with the magnetic resonance functional connection, and the correlation coefficient is 0.35. As Figure 6 shown, the anatomical structure connection is significantly correlated with the positron emission tomography metabolic parameters, and the correlation coefficient is 0.45.

[0046] Corresponding to the above method, as Figure 7As shown in the figure, this embodiment also provides a brain structure-function-metabolism network correlation analysis system based on multimodal synchronous imaging, including: A multimodal data acquisition unit for acquiring multimodal data of a target subject by using a nuclear magnetic optoelectronic integrated device; the multimodal data includes positron emission tomography data, magnetic resonance imaging data, electroencephalogram data, eye tracker data, and near-infrared spectroscopy imaging data; A data preprocessing unit for performing modality-specific preprocessing on each of the multimodal data to obtain preprocessed data; A matrix determination unit for determining a connection matrix of different brain regions according to the preprocessed data; A result determination unit for analyzing the mutual relationship between different networks based on the connection matrix to obtain the correlation results of the structural network, functional network, and metabolic network.

[0047] The core innovation of the present invention lies in the adoption of an advanced PMEEN multimodal synchronous acquisition system, which realizes the synchronous acquisition of five-modal data of PET, MRI, EEG, ET, and NIRS. This technological breakthrough effectively solves the key problem of asynchronous multimodal data in traditional methods. Based on the high spatio-temporal resolution data obtained by synchronous acquisition, the present invention can acquire the functional activities (BOLD signal, EEG signal, and fNIRS signal) and metabolic state (dynamic PET signal) of the brain in the same state, so as to accurately analyze the correlation between the metabolic network, functional network, and structural network, and deeply reveal the neurovascular function coupling mechanism and brain energy metabolism law. It provides a new research paradigm for understanding the brain function mechanism. The implementation of the present invention will provide innovative technical means for brain science research and clinical brain function evaluation, and provide new ideas and solutions for multimodal brain information integration.

[0048] The beneficial effects of the present invention are as follows: The present invention combines the synchronously acquired fMRI, EEG, fNIRS, fPET, and ET data, as well as the DTI data of the corresponding subject, to provide an effective method for exploring the correlation between the brain structure, function, and metabolism network. Specifically, by performing correlation analysis between functional networks on the synchronous data of three different technologies, fMRI, EEG, and fNIRS, the mutual relationship between the functional networks reflected by the BOLD signal, brain electrical activity, and cerebral blood oxygenation change can be revealed. At the same time, by analyzing the correlation between DTI and the functional and metabolic networks, the interaction between the brain structure, function, and metabolism can be deeply explored.

[0049] In the present specification, the various embodiments are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. For the same or similar parts among the various embodiments, reference can be made to each other. For the system disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple. For the relevant parts, reference can be made to the description in the method section.

[0050] In this article, specific examples are used to elaborate on the principles and implementation manners of the present invention. The descriptions of the above embodiments are only used to help understand the method and its core idea of the present invention. At the same time, for those of ordinary skill in the art, according to the idea of the present invention, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation to the present invention.

Claims

1. A method for analyzing the association between brain structure-functional-metabolic networks based on multimodal synchronous imaging, characterized in that Including: Obtaining multimodal data of a target subject by using a nuclear magnetic, photoelectric and integrated device; the multimodal data includes positron emission tomography data, magnetic resonance imaging data, electroencephalogram data, eye tracker data and near-infrared spectroscopy imaging data; Performing modality-specific preprocessing on each of the multimodal data to obtain preprocessed data; Determining a connection matrix of different brain regions according to the preprocessed data; Analyzing the mutual relationship between different networks based on the connection matrix to obtain the correlation results of the structural network, functional network and metabolic network; Obtaining multimodal data of a target subject by using a nuclear magnetic, photoelectric and integrated device, including: Collecting T1-weighted structural images and diffusion tensor imaging data; Use 18 Dynamic data acquisition is performed using the F-FDG tracer to obtain the positron emission tomography data; While collecting dynamic data, starting synchronous data recording of the electroencephalogram data, the eye tracker data, the near-infrared spectroscopy imaging data and resting-state fMRI data to obtain the final multimodal data; the magnetic resonance imaging data includes: resting-state fMRI data, T1-weighted structural images and diffusion tensor imaging data.

2. The method for analyzing the association between brain structure-function-metabolism networks based on multimodal synchronous imaging according to claim 1, wherein Performing modality-specific preprocessing on each of the multimodal data to obtain preprocessed data, including: Performing 1 Hz high-pass filtering, bad channel processing, downsampling, ICA denoising, rereferencing, source analysis and eye movement data regression on the electroencephalogram data to obtain the preprocessed data corresponding to the electroencephalogram data; Performing head motion correction, eddy current correction, fiber tractography and spatial normalization on the diffusion tensor imaging data to obtain the preprocessed data corresponding to the diffusion tensor imaging data; Performing time correction, head motion correction, spatial normalization, nuisance regression, detrending, spatial smoothing, filtering and eye movement data regression on the resting-state fMRI data to obtain the preprocessed data corresponding to the resting-state fMRI data; Performing artifact removal, filtering, blood oxygen calculation, source analysis and eye movement data regression on the near-infrared spectroscopy imaging data to obtain the preprocessed data corresponding to the near-infrared spectroscopy imaging data; Performing SUV calculation, spatial normalization, spatial smoothing, SUVr calculation and eye movement data regression on the positron emission tomography data to obtain the preprocessed data corresponding to the positron emission tomography data.

3. The method for analyzing the association between brain structure-function-metabolism networks based on multimodal synchronous imaging according to claim 1, wherein The connection matrix includes a functional connection matrix, a structural connection matrix and a metabolic connection matrix.

4. The method for analyzing the association between brain structure-function-metabolism networks based on multimodal synchronous imaging according to claim 3, wherein The method for determining the structural connection matrix includes: Extracting the number of fiber tracts between each pair of brain regions in the preprocessed data corresponding to the diffusion tensor imaging data of each target subject; Calculating the average number of fiber tracts between the target subjects according to the number of fiber tracts to form an initial connection matrix of 78×78; Performing normalization processing on the initial connection matrix of each target subject, that is, dividing the elements of each initial connection matrix by the total number of fibers in the whole brain to obtain a normalized matrix; Averaging the normalized matrices of all the target subjects to obtain the final structural connection matrix.

5. The method for analyzing the association between brain structure-function-metabolism networks based on multimodal synchronous imaging according to claim 3, wherein The method for determining the functional connection matrix includes: Extract the time series of all voxels from the preprocessed data corresponding to the positron emission tomography data, the magnetic resonance imaging data, the electroencephalogram data, the eye tracker data, and the near-infrared spectroscopy imaging data, and form the average time series of each brain region; Calculate the average value of all voxels in each brain region of the average time series, and use each of the average values as the average time curve of each brain region; Based on the average time curves of each brain region, perform Pearson correlation analysis on the average time curves of every two brain regions to generate a functional connectivity matrix.

6. The method for analyzing the association between brain structure-function-metabolism networks based on multimodal synchronous imaging according to claim 3, wherein, Analyze the mutual relationship between different networks based on the connectivity matrix, and obtain the correlation results of the structural network, the functional network, and the metabolic network, including: Calculate the Spearman correlation coefficient between networks; the Spearman correlation coefficient is calculated by the formula: ; where and are the ranks of each point in a connection matrix and another connection matrix in the connection matrix respectively, represents the average value of the ranks in a connection matrix; represents the average value of the ranks in another connection matrix; Use the Spearman correlation coefficient to determine the correlation results.

7. A brain structure-function-metabolism network association analysis system based on multimodal synchronous imaging, characterized in that, Including: A multimodal data acquisition unit for acquiring multimodal data of a target subject using a nuclear magnetic optoelectronic integrated device; the multimodal data includes positron emission tomography data, magnetic resonance imaging data, electroencephalogram data, eye tracker data, and near-infrared spectroscopy imaging data; A data preprocessing unit for performing modality-specific preprocessing on each of the multimodal data to obtain preprocessed data; A matrix determination unit for determining the connectivity matrix of different brain regions according to the preprocessed data; A result determination unit for analyzing the mutual relationship between different networks based on the connectivity matrix to obtain the correlation results of the structural network, the functional network, and the metabolic network.

Images