Multimodal brain network fusion analysis method based on multilayer network

Abstract

The invention discloses a multi-modal network fusion analysis method based on a multi-layer network. The method comprises the following steps: 1, constructing a single-modal brain network based on canonical correlation analysis; 2, brain network structure-function coupling is extracted; and step 3, constructing and analyzing a high-order multi-mode brain network. The method has the advantages that the constructed high-order multi-mode brain network integrates brain structure and function information and structure-function coupling information, information loss caused by independent analysis of a structure network and a function network is overcome, multi-scale understanding of a brain mechanism and an abnormal mode is achieved, and a neural mechanism can be revealed more comprehensively; and 2, a core-peripheral tissue analysis method is introduced, the degeneration phenomenon of the core brain region of the brain is found from a multi-mode perspective, a new perspective is provided for analyzing the brain information processing process, and the understanding of the key brain region of the brain mechanism is further deepened.

Description

Technical Field

[0001] This invention relates to the interdisciplinary field of artificial intelligence and brain science based on brain networks and brain-like computing, specifically to a multimodal network fusion analysis method based on multilayer networks, which is mainly used for the joint analysis of brain structural and functional networks to explore the mechanisms of the human brain. Background Technology

[0002] With the increasing aging of society, the incidence of cognitive impairment diseases such as Alzheimer's disease and mild cognitive impairment is rising, placing a heavy burden on society and families. Currently, the prevention and treatment of these diseases requires a deeper understanding of the brain's cognitive and memory mechanisms. A typical method for studying brain mechanisms involves constructing and analyzing brain networks. However, brain network research methods mainly focus on single-modality brain imaging data, such as using only structural magnetic resonance imaging (sMRI) data to construct brain structural networks, or using only functional magnetic resonance imaging (fMRI) data to construct brain functional networks. These methods have certain limitations and cannot fully reflect the complex relationship between brain structure and function. A typical brain network analysis method is core-periphery analysis, but from a single-modality perspective, there are core-periphery divisions in a structural sense and core-periphery divisions in a functional sense, which cannot be effectively combined. This necessitates a joint analysis from a multimodal perspective. Summary of the Invention

[0003] This invention provides a multimodal network fusion analysis method based on multilayer networks, primarily used for the joint analysis of brain structural and functional networks to explore human brain mechanisms. The technical problem to be solved is that single-modal brain network analysis cannot fully reflect the complex relationship between brain structure and function, nor can it effectively combine the core-periphery division in both structural and functional senses.

[0004] The technical solution adopted by the present invention to solve the above-mentioned technical problems is as follows:

[0005] A multimodal brain network fusion analysis method based on multilayer networks is provided, including the following steps:

[0006] Step 1: Construction of a unimodal brain network based on canonical correlation analysis

[0007] Brain structural and functional networks were constructed using a data-driven canonical correlation analysis method. For the structural network, nodes were defined using the AAL116 template, and edge weights were calculated based on the canonical correlation of gray matter density between brain regions. Similarly, for the functional network, nodes were defined using the AAL116 template, and edge weights were calculated based on the canonical correlation of blood oxygenation signal time series between brain regions. The constructed networks were then sparsified to obtain binarized structural and functional networks.

[0008] Step 2: Extracting brain network structure-functional coupling

[0009] After constructing a unimodal brain network, this connection can be interpreted as the relationship between brain structural networks and brain functional networks. This intrinsic relationship is a key focus in brain network research and has received considerable attention, but it remains unresolved. For ease of expression, the term Structure-Function Coupling (SFC) has been proposed. SFC aims to express the statistical dependence between structure and function, i.e., how brain structure is statistically dependent on brain function, and how brain function is statistically dependent on brain structure. To quantify this statistical dependence, the most common approach is the statistical correlation method. This method uses the Pearson correlation coefficient to measure the statistical dependence between brain region structure and brain region function.

[0010] Step 3: Construction and Analysis of High-Order Multimodal Brain Networks

[0011] Based on a multi-layered network architecture, a high-order multimodal brain network is constructed by integrating brain structural and functional information as well as structure-function coupling information. Multi-layered network metrics such as node strength, betweenness centrality, proximity centrality, and local efficiency are used to analyze the multimodal brain network, further identifying differential brain regions associated with cognitive and memory impairments. Furthermore, core-periphery organizational analysis reveals the degeneration of the core brain regions and their crucial role in cognitive and memory mechanisms.

[0012] According to the multimodal brain network fusion analysis method based on multilayer networks described in this application, the method is characterized in that, in step 1, both the structural network and the functional network need to be sparsified by selecting appropriate thresholds, and the fully connected weighted network is transformed into a partially connected binary network; the threshold selection strategy is as follows: select a threshold such that the average degree value of the binary network is greater than twice the logarithm of N and the network density is less than 0.5.

[0013] According to the multi-modal brain network fusion analysis method based on multi-layer networks described in this application, the key feature is that, in step 2, extracting the brain network structure-function coupling, during implementation...

[0014] The structural-functional coupling of each brain region is defined by the following formula;

[0015] SFC i =|Pearson(SC) i FC i )|

[0016] Where i represents the brain region number, SC i and FC i SFC represents the structural and functional connectivity vectors of brain region i, respectively. iThis refers to the structure-function coupling of brain region i. For the sake of interpretation and convenience of subsequent calculations, the structure-function coupling values ​​here are all positive, that is, the value range is between 0 and 1. After calculating the coupling value of each brain region, a diagonal matrix is ​​constructed to obtain the whole brain structure-function coupling network, that is, the SFC network.

[0017] According to the multi-layer network-based multimodal brain network fusion analysis method described in this application, the multi-layer network architecture in step 3 can be of various types. Based on the multiplexed network architecture, a high-order multimodal brain network is constructed. The construction of the multimodal brain network includes four steps: node definition, intra-layer relationship definition, inter-layer relationship definition, and network representation.

[0018] The nodes of the multimodal brain network represent the brain regions of interest in the study. Ninety cerebral regions and 26 cerebellar regions were defined using the AAL116 template. The same node definition was used for both the structural and functional layers of the multimodal brain network. Intra-layer relationships refer to the connections between nodes within a layer. For the structural layer, the intra-layer relationship is defined as the absolute value of the correlation between gray matter densities in brain regions; for the functional layer, the intra-layer relationship is defined as the absolute value of the correlation between time series of blood oxygenation signals in brain regions. Therefore, the structural layer is equivalent to a non-sparsed unimodal brain structural connectivity network, and the functional layer is equivalent to a non-sparsed unimodal brain functional connectivity network, denoted as SC network and FC network, respectively. For inter-layer relationships, the structure-function coupling strength is used as the definition.

[0019] Specifically, the structure-function coupling of each brain region is defined by the following formula;

[0020] SFC i =|Pearson(SC) i FC i )|

[0021] Where i represents the brain region number, SC i and FC i SFC represents the structural and functional connectivity vectors of brain region i, respectively. i This represents the structure-function coupling of brain region i. For interpretability and ease of subsequent calculations, all structure-function coupling values ​​here are positive, ranging from 0 to 1. After calculating the coupling value for each brain region, a diagonal matrix is ​​constructed to obtain the whole-brain structure-function coupling network, i.e., the SFC network. SFC quantifies the statistical dependence between brain region structure and function, representing the correlation between structure and function. The multimodal network is represented using a superadjacency matrix. The superadjacency matrix is ​​represented by Super_Matrix.

[0022]

[0023] According to the multimodal brain network fusion analysis method based on multilayer networks described in this application, the core-periphery joint analysis method based on multimodal networks in step 3 is as follows:

[0024] For the adjacency matrix Super_Matri of a multimodal network, the analysis here only uses the SF network and the FC network; let SF = {a i,j} and FC={b i,j For each node i in SC and FC, calculate k. i SC and k i FC ;

[0025] k i SC =∑ j≠i a i,j

[0026] k i FC =∑ j≠i b i,j

[0027] Where j represents all other nodes in the network; it can be seen that k i SC and k i FC Let k represent the node strength of node i in the SC network and the FC network, respectively; i SC and k i FC Based on the magnitude of their values, they are divided into two subsets, namely k. i SC =k i SC+ +k i SC- and k i FC =k i FC+ +k i FC- ;

[0028] For SC and FC networks, calculate the coefficient c respectively. SC and c FC ;

[0029]

[0030] Where N represents the number of nodes in the SC network and the FC network; it can be seen that c SC and c FCThese are the reciprocals of the mean node strength in the SC and FC networks, respectively.

[0031] For each node, calculate μ. i Value and value;

[0032]

[0033] For all nodes i, rank i It represents its μ i The value is sorted among all nodes; calculation Regarding rank i Functional relationships, record max_rank determines the division between core nodes and peripheral nodes in a multimodal network; specifically, nodes with a rank value less than max_rank are core nodes, and nodes with a rank value greater than max_rank are peripheral nodes.

[0034] The beneficial effects of this invention are:

[0035] 1. The constructed high-order multimodal brain network integrates brain structural and functional information, as well as structure-function coupling information, overcoming the information loss caused by analyzing structural and functional networks separately. This enables a multi-scale understanding of brain mechanisms and abnormal patterns, contributing to a more comprehensive revelation of neural mechanisms. 2. The introduction of a core-periphery tissue analysis method revealed degeneration in the core brain regions from a multimodal perspective, providing a new perspective for analyzing brain information processing and further deepening our understanding of key brain regions involved in brain mechanisms. Attached Figure Description

[0036] Figure 1 These are schematic diagrams of three multi-layer network architectures in this application ((a) multiplexing network; (b) hypernetwork; (c) multidimensional network);

[0037] Figure 2 This is a schematic diagram of the multimodal brain network construction process of this application. Detailed Implementation

[0038] The following will be combined with the appendix Figures 1-2 This invention will be described in detail, and the technical solutions in the embodiments of this invention will be clearly and completely described. Obviously, the described embodiments are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0039] This invention provides a multimodal network fusion analysis method based on multilayer networks, primarily used for the joint analysis of brain structural and functional networks to explore human brain mechanisms. The technical problem to be solved is that single-modal brain network analysis cannot fully reflect the complex relationship between brain structure and function, nor can it effectively combine the core-periphery division in both structural and functional senses. The implementation is as follows;

[0040] 1. Construction of a unimodal brain network based on canonical correlation analysis

[0041] The construction of a unimodal structural network is based on sMRI structural images. After preprocessing, the sMRI structural image data is segmented into three parts: gray matter, white matter, and cerebrospinal fluid. The brain structural network is constructed based on the segmented gray matter images. Nodes in the structural network are defined using the AAL116 template. For defining structural connections, univariate analysis methods such as Pearson correlation coefficient or partial correlation coefficient are commonly used. However, necessary multivariate relationships may be overlooked by univariate analysis methods. Therefore, multivariate analysis is used to evaluate the correlation between nodes. Canonical correlation analysis is used to define the connections in the structural network.

[0042] For the functional network, the AAL116 template is also used to define nodes. Each brain region defined by the AAL116 template is designated as a Region of Interest (ROI). The time series of a voxel signal within an ROI constitutes a random variable, and the time series of all voxels constituting that ROI form a set of random variables. A linear combination of these random variable sets, i.e., canonical variables, is sought to best represent the set of random variables. By calculating the correlation between the canonical variables of two ROIs, the connection edges between the nodes of the two ROIs are defined. The connection edges of all ROIs in the network are calculated and sparsified to construct the functional connectivity network.

[0043] Both structural and functional networks require appropriate thresholding for sparsification, transforming fully connected weighted networks into partially connected binary networks. The threshold selection strategy is as follows: choose a threshold such that the average degree of the binary network is greater than twice the logarithm of N and the network density is less than 0.5.

[0044] 2. Extracting the structural-functional coupling of brain networks

[0045] The structure and function of the brain are naturally linked. After constructing brain networks, this link can be interpreted as the relationship between brain structural networks and brain functional networks. This intrinsic relationship is a key focus and has attracted much attention in brain network research, but it remains unresolved. For ease of expression, the term Structure-Function Coupling (SFC) has been proposed. SFC aims to express the statistical dependence between structure and function; that is, how brain structure statistically depends on brain function, and how brain function statistically depends on brain structure. To quantify this statistical dependence, a common approach is the statistical correlation method. This method uses the Pearson correlation coefficient to measure the statistical dependence between brain region structure and brain region function. Figure 2 This illustrates the construction process of the structure-function network. Specifically, the structure-function coupling of each brain region is defined by the following formula.

[0046] SFC i =|Pearson(SC) i FC i )|

[0047] Where i represents the brain region number, SC i and FC i SFC represents the structural and functional connectivity vectors of brain region i, respectively. i This represents the structure-function coupling of brain region i. For interpretability and ease of subsequent calculations, the structure-function coupling values ​​here are all positive, ranging from 0 to 1. After calculating the coupling value for each brain region, a diagonal matrix is ​​constructed to obtain the whole-brain structure-function coupling network, i.e., the SFC network.

[0048] Structure-function coupling (SC-FC) is a crucial characteristic of brain networks, and the degree of mismatch (i.e., decoupling) has significant physiological implications, representing an important feature of brain networks at physiological, developmental, cognitive, and pathological levels. Mismatch supports dynamic functional and cognitive flexibility. The degree of SC-FC coupling exhibits high spatiotemporal heterogeneity across brain regions, with mismatched areas often concentrated in the executive control and higher-order association cortex, facilitating dynamic network reorganization and information integration. Mismatch reflects the hierarchical organization and evolutionary expansion of the brain. On a macroscopic scale, SC-FC coupling shows a gradient from primary (visual, sensory) to higher-order (associative, executive) cortices, with higher-order cortices exhibiting greater decoupling, consistent with the expansion and increased complexity of functional areas in the cerebral cortex during evolution. Mismatch also reflects developmental and plasticity mechanisms and is related to cognitive reserves and compensation. Furthermore, abnormal SC-FC mismatch has become a potential biomarker for various neuropsychiatric disorders and traumatic brain injury, aiding in early diagnosis and intervention strategy development.

[0049] 3. Construction and analysis of multimodal brain networks

[0050] There are various architectures for multi-layer networks, such as... Figure 1 As shown. This method is based on a multiplexed network architecture to construct a high-order multimodal brain network, such as... Figure 2 As shown. The construction of a multimodal brain network includes four steps: node definition, intra-layer relation definition, inter-layer relation definition, and network representation. The nodes of the multimodal brain network represent the Region of Interest (ROI) under study, using 90 cerebral regions and 26 cerebellar regions defined by the AAL116 template. To better map the connection between structure and function, the same node definition is used for both the structural and functional layers of the multimodal brain network. Intra-layer relations refer to the connections between nodes within a layer. For the structural layer, intra-layer relations are defined as the absolute value of the correlation between gray matter densities in brain regions; for the functional layer, intra-layer relations are defined as the absolute value of the correlation between time series of blood oxygenation signals in brain regions. Thus, the structural layer is equivalent to a non-sparsed unimodal structural connectivity network, and the functional layer is equivalent to a non-sparsed unimodal functional connectivity network, denoted as SC network and FC network, respectively. For inter-layer relations, the structure-function coupling strength is used. Specifically, the structure-function coupling of each brain region is defined by the following formula.

[0051] SFC i =|Pearson(SC) i FC i )|

[0052] Where i represents the brain region number, SC i and FC i SFC represents the structural and functional connectivity vectors of brain region i, respectively. i This represents the structure-function coupling of brain region i. For interpretability and ease of subsequent calculations, the structure-function coupling values ​​here are all positive, ranging from 0 to 1. After calculating the coupling value for each brain region, a diagonal matrix is ​​constructed to obtain the whole-brain structure-function coupling network, i.e., the SFC network. SFC quantifies the statistical dependence between brain region structure and function, representing the correlation between structure and function. The multimodal network is represented using a superadjacency matrix. The superadjacency matrix is ​​denoted by Super_Matrix.

[0053]

[0054] Multimodal brain networks were analyzed using multi-layer network metrics such as node strength, betweenness centrality, proximity centrality, and local efficiency to reveal the important roles of abnormal brain regions and the cerebellum in brain mechanisms such as cognition and memory.

[0055] Further, a joint analysis of the core-periphery structure from a multimodal perspective is conducted. Core-periphery structure analysis of the brain aims to address the question: what is the core of the brain? By dividing brain regions into two disjoint subsets—core nodes and peripheral nodes—we can understand the key nodes in the brain network, thereby establishing a basic understanding of the brain's information processing processes. Analysis of the brain's core-periphery structure typically involves analyzing brain structural or functional information separately, using appropriate metrics to analyze key network nodes. However, this approach may lead to information loss. This method defines and constructs a multimodal brain network that incorporates both structural and functional information. Through a unified model, the core-periphery structure is obtained based on multimodal network analysis. This approach combines structural and functional information, overcoming the information loss caused by analyzing structure or function separately.

[0056] The specific core-periphery joint analysis method based on multimodal networks is as follows:

[0057] For the adjacency matrix Super_Matri of a multimodal network, the analysis here only uses the SF and FC networks. Let SF = {a i,j} and FC={b i,j For each node i in SC and FC, calculate k. i SC and k i FC .

[0058] k i SC =∑ j≠i a i,j

[0059] k i FC =∑ j≠i b i,j

[0060] Here, j represents all other nodes in the network. It can be seen that k i SC and k i FC Let k represent the node strength of node i in the SC network and the FC network, respectively. i SC and k i FC Based on the magnitude of their values, they are divided into two subsets, namely k. i SC =k i SC+ +k i SC- and k i FC =ki FC+ +k i FC- .

[0061] For SC and FC networks, calculate the coefficient c respectively. SC and c FC .

[0062]

[0063] Where N represents the number of nodes in the SC network and the FC network. As can be seen, c SC and c FC These are the reciprocals of the mean node strength in the SC and FC networks, respectively.

[0064] For each node, calculate μ. i Value and value.

[0065]

[0066] For all nodes i, rank i It represents its μ i The value is sorted among all nodes. Calculation Regarding rank i Functional relationships, record `max_rank` determines the distinction between core and peripheral nodes in a multimodal network. Specifically, nodes with a rank value less than `max_rank` are core nodes, and nodes with a rank value greater than `max_rank` are peripheral nodes.

[0067] Regarding the multimodal brain network fusion analysis method based on multilayer networks proposed in this application, the application areas are described, and this method has significant application value in the following fields: Through multimodal brain network fusion (structure, function, and structure-function coupling information) and core-periphery joint analysis, this method has significant application value in the following fields:

[0068] 1. Clinical diagnosis of neurodegenerative diseases:

[0069] It is suitable for early screening and dynamic monitoring of diseases such as Alzheimer's and Parkinson's. By fusing sMRI / fMRI data to construct a multimodal network and combining it with the structure-function coupling (SFC) mechanism, it can quantify the degeneration of core brain regions such as the hippocampus and prefrontal cortex, overcoming the limitations of single-modal analysis and improving diagnostic sensitivity. Its multiplexed network architecture ( Figure 1 It provides a flexible framework for pathological research:

[0070] 2. Cognitive Impairment Intervention and Brain Rehabilitation:

[0071] This approach provides target localization for personalized treatment of patients with cognitive and memory impairments (such as mild cognitive impairment). Based on core-periphery analysis, it accurately identifies functionally abnormal brain regions, guides targeted interventions using neuromodulation techniques such as transcranial magnetic stimulation (TMS), and dynamically assesses rehabilitation outcomes through network indicators (such as local efficiency).

[0072] 3. The field of understanding the pathological mechanisms of mental illness:

[0073] Advance the discovery of biomarkers for diseases such as depression and schizophrenia. Utilize structure-function mismatch (SFC network) to quantify the degree of decoupling in the limbic system (such as the amygdala) and correlate it with emotional or cognitive symptoms, providing a new perspective for drug development. Figure 2 The multimodal construction process shown intuitively supports the establishment of pathological models:

[0074] 4. Interdisciplinary applications of brain science and artificial intelligence:

[0075] This technology empowers the upgrade of neuromorphic computing and brain-computer interface (BCI) technologies. It utilizes the core nodes of multimodal networks as priority decoding channels for BCI signals, improving command recognition efficiency. Simultaneously, it simulates the multi-scale information processing mechanism of the human brain based on a multi-layered network architecture, providing a bio-inspired design paradigm for next-generation neural network algorithms.

[0076] Summary of application value

[0077] The core advantage of this method is:

[0078] Multi-scale fusion: Integrates structural, functional, and coupled information (SC / FC / SFC networks) to overcome the fragmentation problem of single-modal information;

[0079] Precise localization: Quantifying the degeneration of key brain regions through core-periphery analysis to advance precision medicine;

[0080] Technical scalability: Multi-layer network architecture Figure 1 It is compatible with a variety of brain imaging data types, supporting cross-domain applications.

[0081] These characteristics make it a universal analytical tool in the fields of neuromedicine, cognitive science, and artificial intelligence, and it has significant industrialization potential.

[0082] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A multimodal brain network fusion analysis method based on multilayer networks, characterized in that, Includes the following steps: Step 1: Construction of a unimodal brain network based on canonical correlation analysis; Brain structural and functional networks were constructed using a data-driven canonical correlation analysis method. For the brain structural network, nodes were defined using the AAL116 template, and edge weights were calculated based on the canonical correlation of gray matter density between brain regions. Similarly, nodes were defined using the AAL116 template, and edge weights were calculated based on the canonical correlation of blood oxygenation signal time series between brain regions. The constructed networks were then sparsified to obtain binarized structural and functional networks. Step 2: Extract brain network structure-functional coupling; After constructing a unimodal brain network, this connection can be interpreted as the relationship between brain structural networks and brain functional networks. The Pearson correlation coefficient is used to measure the statistical dependence between brain region structure and brain region function. Step 3: Construction and analysis of high-order multimodal brain networks; Based on a multi-layer network architecture, a high-order multimodal brain network is constructed by integrating brain structure and function information as well as structure-function coupling information. Multimodal brain networks were analyzed using multi-layer network metrics such as node strength, betweenness centrality, proximity centrality, and local efficiency to further identify differential brain regions associated with cognitive and memory impairments. Furthermore, core-periphery organization analysis revealed the degeneration of the core brain regions and their key role in cognitive and memory mechanisms.

2. The method for multimodal brain network fusion analysis based on multilayer networks according to claim 1, characterized in that, In step 1, both the structural network and the functional network need to be sparsified by selecting appropriate thresholds to transform the fully connected weighted network into a partially connected binary network. The specific strategy for selecting the threshold is as follows: select a threshold such that the average degree of the binary network is greater than twice the logarithm of N and the network density is less than 0.

5.

3. The method for multimodal brain network fusion analysis based on multilayer networks according to claim 1, characterized in that, Step 2 extracts the structure-function coupling of the brain network during implementation. The structural-functional coupling of each brain region is defined by the following formula; SFC i =|Pearson(SC i ,FC i )| Where i represents the brain region number, SC i and FC i SFC represents the structural and functional connectivity vectors of brain region i, respectively. i This refers to the structure-function coupling of brain region i. For the sake of interpretation and convenience of subsequent calculations, the structure-function coupling values ​​here are all positive, that is, the value range is between 0 and 1. After calculating the coupling value of each brain region, a diagonal matrix is ​​constructed to obtain the whole brain structure-function coupling network, that is, the SFC network.

4. The method for multimodal brain network fusion analysis based on multilayer networks according to claim 1, characterized in that, There are various architectures for multi-layer networks in step 3. Based on the multiplexed network architecture, a high-order multimodal brain network is constructed. The construction of the multimodal brain network includes four steps: node definition, intra-layer relationship definition, inter-layer relationship definition, and network representation. The nodes of the multimodal brain network represent the brain regions of interest in the study. Ninety cerebral regions and 26 cerebellar regions were defined using the AAL116 template. The same node definition was used for both the structural and functional layers of the multimodal brain network. Intra-layer relationships refer to the connections between nodes within a layer. For the structural layer, the intra-layer relationship is defined as the absolute value of the correlation between gray matter densities in brain regions; for the functional layer, the intra-layer relationship is defined as the absolute value of the correlation between time series of blood oxygenation signals in brain regions. Therefore, the structural layer is equivalent to a non-sparsed unimodal brain structural connectivity network, and the functional layer is equivalent to a non-sparsed unimodal brain functional connectivity network, denoted as SC network and FC network, respectively. For inter-layer relationships, the structure-function coupling strength is used as the definition. Specifically, the structure-function coupling of each brain region is defined by the following formula; SFC i =|Pearson(SC i ,FC i )| Where i represents the brain region number, SC i and FC i SFC represents the structural and functional connectivity vectors of brain region i, respectively. i This refers to the structure-function coupling of brain region i. For interpretability and convenience of subsequent calculations, the structure-function coupling values ​​here are all positive, i.e., the value range is between 0 and 1. After calculating the coupling value of each brain region, a diagonal matrix is ​​constructed to obtain the whole-brain structure-function coupling network, i.e., the SFC network. SFC quantifies the statistical dependence of brain region structure and function and can represent the correlation between structure and function. The multimodal network is represented using a superadjacency matrix, denoted by Super_Matrix.

5. The method for multimodal brain network fusion analysis based on multilayer networks according to claim 1, characterized in that, The specific core-periphery joint analysis method based on multimodal networks in step 3 is as follows: For the adjacency matrix Super_Matri of a multimodal network, the analysis here only uses the SF network and the FC network; let SF = {a i,j } and FC={b i,j For each node i in SC and FC, calculate k. i SC and k i FC ; Where j represents all other nodes in the network; it can be seen that k i SC and k i FC Let k represent the node strength of node i in the SC network and the FC network, respectively; i SC and k i FC Based on the magnitude of their values, they are divided into two subsets, namely... k i SC =k i SC+ +k i SC- and k i FC =k i FC+ +k i FC- ; For SC and FC networks, calculate the coefficient c respectively. SC and c FC ; Where N represents the number of nodes in the SC network and the FC network; it can be seen that c SC and c FC These are the reciprocals of the mean node strength in the SC and FC networks, respectively. For each node, calculate μ. i Value and value; For all nodes i, rank i It represents its μ i The value is sorted among all nodes; calculation Regarding rank i Functional relationships, record max_rank determines the division between core nodes and peripheral nodes in a multimodal network; specifically, nodes with a rank value less than max_rank are core nodes, and nodes with a rank value greater than max_rank are peripheral nodes.

Images