Alzheimer's disease virtual injury judgment method based on brain dynamics model

Abstract

The invention provides an Alzheimer's disease virtual injury judgment method based on a brain dynamics model, and the method comprises the steps: S1, obtaining a structural magnetic resonance image and a functional magnetic resonance image of a subject, and carrying out the preprocessing, and obtaining the preprocessed data; s2, calculating topological characteristic indexes of the structural network; s3, drawing a required topological feature profile map; s4, clustering is carried out on the topological feature profile map; s5, constructing a whole brain dynamic model for each subject; s6, constructing a node disturbance model of each subject, and analyzing a relationship among a structural network topology characteristic index, a cortical hierarchy gradient index and a whole brain dynamic index; s7, according to the cortex thickness and the cortex volume of the subject, in combination with the whole brain dynamics model, judging the influence of local brain region atrophy on whole brain metastable state dynamics; and S8, judging the influence of white matter fiber bundle connection disconnection on the whole brain metastable state dynamics. The invention provides a brain network dynamics model modeling method for different structure damage effects.

Description

Technical Field

[0001] The present invention relates to the technical field of virtual lesion judgment of Alzheimer's disease, and in particular to a virtual lesion judgment method of Alzheimer's disease based on a brain dynamics model. Background Art

[0002] The human brain is an extremely complex dynamic system, composed of relatively stable anatomical connections, which are separated and integrated through the coordination of neurodynamics to cope with the ever-changing external environment. With the aging of the population, Alzheimer's disease and other geriatric diseases are increasingly affecting the health of the people, and the incidence, prevalence and mortality of Alzheimer's disease continue to increase. However, there is still no effective treatment for Alzheimer's disease, and the development of brain science is urgently needed to provide new solutions for the prevention and treatment of Alzheimer's disease. Summary of the invention

[0003] The present invention provides a method for determining virtual damage of Alzheimer's disease based on a brain dynamics model, so as to determine virtual damage of Alzheimer's disease.

[0004] This specification discloses a method for determining virtual damage of Alzheimer's disease based on a brain dynamics model, including:

[0005] S1. Obtaining structural magnetic resonance images and functional magnetic resonance images of the subject, and performing preprocessing to obtain preprocessed data, wherein the preprocessed data includes the subject's cortical thickness, cortical volume, cortical level gradient index, white matter fiber bundle structural connection matrix, white matter fiber bundle fractional anisotropy data and functional connection matrix;

[0006] S2. Based on the preprocessed white matter fiber bundle structural connection matrix, calculating the structural network topological characteristic indicators, including node strength, degree centrality, eigenvector centrality, betweenness centrality, clustering coefficient, node shortest path length, participation coefficient and module intra-degree z score;

[0007] S3, based on the participation coefficient and module internality z score calculated above, jointly draw the required topological feature profile;

[0008] S4. Use k-means clustering method to cluster the topological feature profiles and obtain the whole brain dynamic index - state integration rate SOIR;

[0009] S5. Based on the preprocessed white matter fiber bundle structural connection matrix and functional connection matrix, a whole-brain dynamics model is constructed for each subject using the Desikan brain atlas proposed by American scientist Rahul S. Desikan;

[0010] S6. The atlas divides the brain into 68 cortical areas, and simulates the perturbation of each cortical area in turn. Each subject will get a 68-node perturbation model, and analyze the relationship between the structural network topological characteristic index, the cortical level gradient index and the whole brain dynamic index from a causal perspective;

[0011] S7. Based on the whole-brain dynamics model of the subject, combined with the subject's cortical thickness and cortical volume, determine the impact of local brain atrophy on the whole-brain metastable dynamics;

[0012] S8. Based on the whole-brain dynamics model of the subjects and the fractional anisotropy data of the white matter fiber bundles of the subjects, determine the impact of disconnection of white matter fiber bundles on the metastable dynamics of the whole brain.

[0013] In some embodiments of the present specification, in S1, the preprocessing step includes: time calibration, head motion calibration, spatial normalization and low-frequency filtering.

[0014] In some embodiments of this specification, in S2, the calculation of the structural network topology characteristic index includes:

[0015] The node strength S of the i-th node i The calculation formula is as follows:

[0016]

[0017] In formula (1), M ij represents the connection weight between node i and node j, and n represents the number of nodes;

[0018] The degree centrality D of the i-th node i The calculation formula is as follows:

[0019]

[0020] In formula (2), C ij represents the connection status between node i and node j, and n represents the number of nodes;

[0021] The eigenvector centrality E of the i-th node i The calculation formula is as follows:

[0022]

[0023] In formula (3), C ij represents the connection status between node i and node j, e j represents the importance measure of node j, λ represents the proportionality constant, and n represents the number of nodes;

[0024] The clustering coefficient CC of the i-th node i The calculation formula is as follows:

[0025]

[0026] In formula (4), R i Indicates the maximum number of triangles formed by node i and its neighboring nodes, and the denominator k i Indicates the maximum number of triangles that node i and its neighboring nodes form a complete graph (every two nodes are connected);

[0027] The betweenness centrality B of the i-th node i The calculation formula is as follows:

[0028]

[0029] In formula (5), W i (j,k) represents the number of shortest paths between nodes j and k through node i, and W(j,k) represents the number of all shortest paths between nodes j and k. A node with high betweenness centrality means that many or even all shortest paths between other nodes must pass through it.

[0030] In a network, the shortest path length between each node and all other nodes is the node shortest path length.

[0031] The participation coefficient P of the i-th node i The calculation formula is as follows:

[0032]

[0033] In formula (6), D is represents the degree centrality of node i in module s, D i represents the degree centrality of node i, s represents the module, and m represents the total number of modules in the network;

[0034] The module-internal degree z score Z of the i-th node i The calculation formula is as follows:

[0035]

[0036] In formula (7), D' i The degree centrality of node i within module s, is the average intra-module degree of all nodes in module s, is the standard deviation of the intra-module degree of all nodes in module s.

[0037] In some embodiments of this specification, in S4, the calculation formula of the state integration rate SOIR is as follows:

[0038]

[0039] In formula (8), T is the total number of time points, that is, the total number of TRs; T I Indicates the number of time points during the scan that the consolidation state was active.

[0040] In some embodiments of this specification, in S6, 68 nodes are disturbed in sequence under the same damage degree, and the disturbance method for node i is as follows:

[0041] C i =SC i *damage; (9)

[0042] In formula (9), SC i is all the connections of node i in the white matter fiber bundle structure connection matrix, damage is the damage factor, C i is the white matter fiber bundle structure connection matrix of node i after simulating damage.

[0043] In some embodiments of the present specification, in S6, in order to evaluate the effect of node perturbation on functional dynamics SOIR, a paired sample t-test is used to statistically compare the SOIR values ​​simulated by the whole brain dynamics model and the node perturbation model to determine whether there is a significant difference in the results of each node perturbation model;

[0044] The calculation method for evaluating the functional abnormality effect of damage to node i is as follows:

[0045] ΔSOIR i =SOIR(les i )-SOIR(nonles); (10)

[0046] In formula (10), ΔSOIR i is defined as the damage effect of the ith node, SOIR(les i ) is the state integration rate of the simulated data of the node i damage model, and SOIR(nonles) is the state integration rate of the simulated data of the complete whole-brain dynamics model. After calculating the damage effect of each node, the mean AVE(ΔSOIR i );

[0047] By calculating the Pearson correlation coefficient between AVE(ΔSOIR) and the structural network topological characteristic index and cortical level gradient index, we tested whether the cortical level gradient index and the structural network topological characteristic index of the node were related to the SOIR changes caused by the lesion.

[0048] In some embodiments of this specification, in S7, the effect of atrophy of different brain regions of Alzheimer's disease on metastable dynamics is determined, and virtual damage is performed on the node model of the abnormal brain region of the patient's cortical structure by changing the local recurrent synaptic coupling factor factor, and the specific steps are:

[0049] S71, virtually perturbing the nodes where the cortical thickness and cortical volume are significantly different between the healthy control group (CN) and the patient group, that is, virtually perturbing the local recurrent synaptic coupling factor of these nodes, thereby simulating the imbalance of the local circuit;

[0050] The group-level model of the healthy control group was constructed based on the white matter fiber bundle structure connection matrix and the optimal local recurrent synaptic coupling parameters. The group-level model of the healthy control group and the node perturbation model of local circuit imbalance used consistent simulation conditions to simulate brain activity, and the local recurrent synaptic coupling factor of abnormal node i was perturbed as follows:

[0051] factor i =factor*damage; (11)

[0052] In formula (11), factor i is the local recurrent synaptic coupling parameter after disturbance, factor is the optimal local recurrent synaptic coupling parameter of the group-level model of the healthy control group, and damage is the damage factor;

[0053] S72. Evaluate the effect of local recurrent synaptic coupling factor perturbation on metastable dynamics. Paired sample t-test was used to statistically compare the whole-brain metastability simulated by the group-level model of the healthy control group and the node perturbation model with local circuit imbalance to determine whether virtual damage to the node would affect the whole-brain metastable dynamics.

[0054] In some embodiments of this specification, S8 includes:

[0055] S81: Virtually perturb the edges with significant differences between the healthy control group and the patient group in the fractional anisotropy values ​​of white matter fiber bundles, that is, reduce the structural connection strength of these edges by 50%, and perform the above operation on all edges with significant differences; the construction method of the structural connection matrix l_C of the white matter fiber bundles after Alzheimer's disease (AD) and mild cognitive impairment (MCI) damage is as follows:

[0056]

[0057] In formula (12), l_C ij represents the connection weight between node i and node j in the white matter fiber bundle structure connection matrix l_C after damage, SC ij is the connection weight between node i and node j in the group-level average SC of the healthy control group, FA ij is the fractional anisotropy value of the white matter fiber bundle between node i and node j;

[0058] S82. Determine the effect of white matter connection damage of different resting state subnetworks (RSNs) on metastable dynamics, and perturb the connection of each resting state subnetwork to generate 7 resting state subnetwork white matter perturbation models, which are realized by the white matter fiber bundle structure connection matrix SC of the damage model; the construction method of the damaged white matter fiber bundle structure connection matrix l_Ck of different resting state subnetworks is as follows:

[0059]

[0060] In formula (13), is the connection weight between node i and node j in the white matter fiber bundle structure connection matrix l_Ck after the kth resting state subnetwork is damaged, SC ij is the connection weight between node i and node j in the group-level average white matter fiber bundle structure connection matrix of the healthy control group, FA ij is the fractional anisotropy value of the white matter fiber bundle between node i and node j, RSN k is the set of brain regions included in the kth subnetwork;

[0061] S83. Paired sample t-test was used to statistically compare the whole-brain metastable state simulated by the group-level model of the healthy control group and the different injury models to determine whether there were significant differences in the results of each white matter disturbance model.

[0062] The embodiments of this specification can at least achieve the following beneficial effects:

[0063] This paper proposes a brain network dynamics modeling method for different structural damage effects, exploring the relationship between anatomical structural abnormalities and dynamic functional changes from two aspects: long-range white matter connections and local electrophysiological circuit disturbances. Virtual injury studies using the damaged brain network dynamics model provide insights into the functional dynamics of cortical gradient shaping at the causal mechanism level. Furthermore, using multimodal imaging features of Alzheimer's disease to guide virtual injury, attempts to explore structural indicators that can lead to a significant decrease in whole-brain metastability and explore abnormal neural mechanisms of Alzheimer's disease. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0065] Figure 1 It is a schematic diagram of the framework of the virtual damage judgment method for Alzheimer's disease based on the brain dynamics model involved in the present invention.

[0066] Figure 2It is a schematic diagram of the local damage assessment results involved in the present invention.

[0067] Figure 3 It is a schematic diagram of the damage results of the model based on white matter connection disturbance involved in the present invention. DETAILED DESCRIPTION

[0068] In the following, only some exemplary embodiments are briefly described. As those skilled in the art will appreciate, the described embodiments may be modified in various ways without departing from the spirit or scope of the embodiments of the present invention. Therefore, the drawings and descriptions are considered to be exemplary and non-restrictive in nature.

[0069] The disclosure below provides many different embodiments or examples to implement different structures of the embodiments of the present invention. In order to simplify the disclosure of the embodiments of the present invention, the parts and settings of specific examples are described below. Of course, they are only examples, and the purpose is not to limit the embodiments of the present invention. In addition, the embodiments of the present invention can repeat reference numbers and / or reference letters in different examples, and this repetition is for the purpose of simplification and clarity, which itself does not indicate the relationship between the various embodiments and / or settings discussed.

[0070] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0071] like Figure 1 As shown, this specification discloses a method for determining virtual damage of Alzheimer's disease based on a brain dynamics model, comprising:

[0072] S1. Obtaining structural magnetic resonance images and functional magnetic resonance images of the subject, and performing preprocessing to obtain preprocessed data, wherein the preprocessed data includes the subject's cortical thickness, cortical volume, cortical level gradient index, white matter fiber bundle structural connection matrix, white matter fiber bundle fractional anisotropy data and functional connection matrix;

[0073] S2. Based on the preprocessed white matter fiber bundle structural connection matrix, calculating the structural network topological characteristic indicators, including node strength, degree centrality, eigenvector centrality, betweenness centrality, clustering coefficient, node shortest path length, participation coefficient and module intra-degree z score;

[0074] S3, based on the participation coefficient and module internality z score calculated above, jointly draw the required topological feature profile;

[0075] S4. Use k-means clustering method to cluster the topological feature profiles and obtain the whole brain dynamic index - state integration rate SOIR;

[0076] S5. Based on the preprocessed white matter fiber bundle structural connection matrix and functional connection matrix, a whole-brain dynamics model is constructed for each subject using the Desikan brain atlas proposed by American scientist Rahul S. Desikan;

[0077] S6. The atlas divides the brain into 68 cortical areas, and simulates the perturbation of each cortical area in turn. Each subject will get a 68-node perturbation model, and analyze the relationship between the structural network topological characteristic index, the cortical level gradient index and the whole brain dynamic index from a causal perspective;

[0078] S7. Based on the whole-brain dynamics model of the subject, combined with the subject's cortical thickness and cortical volume, determine the impact of local brain atrophy on the whole-brain metastable dynamics;

[0079] S8. Based on the whole-brain dynamics model of the subjects and the fractional anisotropy data of the white matter fiber bundles of the subjects, determine the impact of disconnection of white matter fiber bundles on the metastable dynamics of the whole brain.

[0080] In some embodiments, in S1, the preprocessing steps include: time calibration, head motion calibration, spatial normalization, and low-frequency filtering.

[0081] In some embodiments, in S2, the calculation of the structural network topology characteristic index includes:

[0082] Node strength refers to the sum of the weights of the edges connected to this node. The node strength S of the i-th node is i The calculation formula is as follows:

[0083]

[0084] In formula (1), M ij represents the connection weight between node i and node j, and n represents the number of nodes;

[0085] For a given node, the degree centrality is the number of neighbors connected to it. The degree centrality D of the i-th node i The calculation formula is as follows:

[0086]

[0087] In formula (2), C ij represents the connection status between node i and node j, and n represents the number of nodes. If there is a connection between node i and node j, then C ij =1, otherwise, C ij =0; This metric is only applicable to binary networks (when a weighted network is applied with this metric, it is considered as a binary network). It only considers the existence or non-existence of a connection, without considering any weight information;

[0088] Eigenvector centrality is an indicator used to evaluate the degree of a node's neighboring nodes. The importance of a node depends on both the number of its neighboring nodes (i.e., the degree of the node) and the importance of its neighboring nodes. i The calculation formula is as follows:

[0089]

[0090] In formula (3), C ij represents the connection status between node i and node j, e j represents the importance measure of node j, λ represents the proportionality constant, and n represents the number of nodes; a node with high degree centrality does not necessarily have high eigenvector centrality.

[0091] The clustering coefficient describes the degree to which a node in a graph and its connected nodes are clustered. The clustering coefficient CC of the i-th node is i The calculation formula is as follows:

[0092]

[0093] In formula (4), R i Indicates the maximum number of triangles formed by node i and its neighboring nodes, and the denominator k i Indicates the maximum number of triangles that node i and its neighboring nodes form a complete graph (every two nodes are connected);

[0094] Betweenness centrality is a measure of the connectivity between different nodes connected to a node. Mathematically, it relies on decomposing the graph into path lengths, which are calculated by the ratio of the number of shortest paths to a given node to the total number of shortest paths between all other nodes in the network. i The calculation formula is as follows:

[0095]

[0096] In formula (5), W i (j,k) represents the number of shortest paths between node j and node k through node i, and W(j,k) represents the number of all shortest paths between node j and node k. A node with a high betweenness centrality means that many or even all shortest paths between other points must pass through it. Therefore, in a network, the greater the betweenness centrality of a node, the greater its role in communication between nodes.

[0097] In a network, the shortest path length between each node and all other nodes is the node shortest path length.

[0098] The participation coefficient is a measure that describes the depth of a single node's embedding in its local module according to the optimal community structure. It is calculated as the ratio of the number of connections of the node in its module to the number of connections in the entire network. The larger the participation coefficient, the more connections this node has in the module. The participation coefficient P of the i-th node i The calculation formula is as follows:

[0099]

[0100] In formula (6), D is represents the degree centrality of node i in module s, D i represents the degree centrality of node i, s represents the module, and m represents the total number of modules in the network.

[0101] The intra-module degree z score is the intra-module version of degree centrality, quantifying the degree of connection between a node and other nodes in its module. i The calculation formula is as follows:

[0102]

[0103] In formula (7), D' i The degree centrality of node i within module s, is the average intra-module degree of all nodes in module s, is the standard deviation of the intra-module degree of all nodes in module s.

[0104] In some embodiments, in S4, the state integration ratio SOI R is calculated as follows:

[0105]

[0106] In formula (3), T is the total number of time points, that is, the total number of TRs; T I represents the number of time points during the scan where the integration state was active. Note that the state incidence of the separation state is the complement of the incidence of the integration state, as each time point belongs to only one of these two states.

[0107] In some embodiments, in S6, 68 nodes are disturbed in sequence under the same damage degree, and the disturbance method for node i is as follows:

[0108] C i =SC i *damage; (9)

[0109] In formula (9), SC i is all the connections of node i in the white matter fiber bundle structure connection matrix, damage is the damage factor, C iis the white matter fiber bundle structure connection matrix of node i after simulating damage.

[0110] In some embodiments, in S6, in order to evaluate the impact of nodal perturbations on functional dynamics SOIR, a paired sample t-test is used to statistically compare the SOIR values ​​simulated by the whole-brain dynamics model and the node perturbation model to determine whether there are significant differences in the results of each node perturbation model; 68 nodes are perturbed in turn, so 68 groups of paired t-test results will be generated under the same degree of damage.

[0111] The calculation method for evaluating the functional abnormality effect of damage to node i is as follows:

[0112] ΔSOIR i =SOIR(les i )-SOIR(nonles); (10)

[0113] In formula (10), ΔSOIR i is defined as the damage effect of node i, SOIR(les i ) is the state integration rate of the damage model simulation data of node i, and SOIR(nonles) is the state integration rate of the complete model simulation data; after calculating the damage effect of each node, the average AVE(ΔSOIR i ).

[0114] By calculating the Pearson correlation coefficient between AVE(ΔSOIR) and the structural network topological characteristic index and cortical level gradient index, we tested whether the cortical level gradient index and the structural network topological characteristic index of the node were related to the SOIR changes caused by the lesion.

[0115] After calculation, the results are as follows Figure 2 As shown: the cortical level gradient index is positively correlated with the change of SOIR; the participation coefficient and the shortest path length of the node are significantly negatively correlated with the change of SOIR after the lesion, the eigenvector centrality, clustering coefficient, node strength and module degree z-score are significantly positively correlated with the change of SOIR after the lesion, and the degree centrality and betweenness centrality are not correlated with SOIR.

[0116] In some embodiments, in S7, the effect of atrophy of different brain regions of Alzheimer's disease on metastable dynamics is determined, and a node model of the abnormal cortical structure of the patient is virtually damaged by changing the local recurrent synaptic coupling factor factor, and the specific steps are:

[0117] S71. Virtually perturb (halve the intensity) the nodes with significant differences in cortical thickness and cortical volume between the healthy control group (CN) and the patient group, i.e., virtually perturb the local recurrent synaptic coupling factor of these nodes to simulate the imbalance of the local circuit;

[0118] The group-level model of the healthy control group was constructed based on the white matter fiber bundle structure connection matrix and the optimal local recurrent synaptic coupling parameters. The group-level model of the healthy control group and the node perturbation model of local circuit imbalance used consistent simulation conditions to simulate brain activity, and the local recurrent synaptic coupling factor of abnormal node i was perturbed as follows:

[0119] factor i =factor*damage; (11)

[0120] In formula (11), factor i is the local recurrent synaptic coupling parameter after disturbance, factor is the optimal local recurrent synaptic coupling parameter of the group-level model of the healthy control group, damage is the damage factor; in this manual, the damage factor = 0.5;

[0121] S72. To evaluate the effect of local perturbation of recurrent synaptic coupling factors on metastable dynamics, paired sample t-tests were used to statistically compare the global metastable states simulated by the group-level model of the healthy control group and the node perturbation model with local circuit imbalance to determine whether virtual damage to the node would affect the global metastable dynamics.

[0122] Table 1 is a table of model damage results based on gray matter atrophy perturbation, showing the difference in cortical thickness and volume. The cortical thickness and volume of the healthy control group are significantly higher than those of the Alzheimer's disease group, and the difference brain areas between the two are basically similar. The brain areas with differences in both the three-sample permutation test and the two-sample permutation test (at least 1 p < 0.05 after FDR correction) are statistically significant between-group differences (p < 0.05). The sixth column of Table 1 shows the results of the perturbation of the local recurrent synaptic coupling factor in each brain area. Among them, all brain areas of the DMN have different degrees of atrophy, and the local circuit abnormalities of half of the nodes will affect the global metastability. In addition, the only 4 nodes of the FPN have significant atrophy, and their damage will have a greater impact on the metastability. The abnormality of the local circuit in this brain area is more likely to cause a decrease in the global metastability.

[0123] Table 1

[0124]

[0125]

[0126]

[0127]

[0128] In some embodiments, in order to study the effect of white matter connection damage in Alzheimer's disease on metastable dynamics, a virtual damage is performed on the structural connection of a group-level model of a healthy control group to simulate the functional effect of disconnection of the structural connection in Alzheimer's disease, S8 comprises:

[0129] S81: Virtually perturb the edges of the white matter fiber bundle fractional anisotropy data that are significantly different (after Bonferron i correction) between the healthy control group (CN) and the patient group, that is, reduce the structural connection strength of these edges by 50%, and perform the above operation on all edges with significant differences; the construction method of the structural connection l_C of Alzheimer's disease (AD) and mild cognitive impairment (MCI) damage is as follows:

[0130]

[0131] In formula (12), l_C ij represents the connection weight between node i and node j in the white matter fiber bundle structure connection matrix l_C after damage, SC ij is the connection weight between node i and node j in the group-level average SC of the healthy control group, FA ij is the fractional anisotropy value of the white matter fiber bundle between node i and node j;

[0132] S82. Determine the effect of white matter connection damage of different resting state subnetworks (RSNs) on metastable dynamics, and perturb the connection of each resting state subnetwork to generate 7 resting state subnetwork white matter perturbation models, which are realized by the white matter fiber bundle structure connection matrix SC of the damage model; the construction method of the damaged white matter fiber bundle structure connection matrix l_Ck of different resting state subnetworks is as follows:

[0133]

[0134] In formula (13), is the connection weight between node i and node j in the white matter fiber bundle structure connection matrix l_Ck after the kth resting state subnetwork is damaged, SC ij is the connection weight between node i and node j in the group-level average white matter fiber bundle structure connection matrix of the healthy control group, FA ij is the fractional anisotropy value of the white matter fiber bundle between node i and node j, RSN k is the set of brain regions included in the kth subnetwork;

[0135] S83. Paired sample t-test was used to statistically compare the global metastable states simulated by the group-level model of the healthy control group and the different injury models to determine whether there were significant differences in the results of each white matter perturbation model.

[0136] Figure 3 The results of paired sample t-test between intact model and injury model are shown (*P<0.05; **P<0.01; ***P<0.001), and CN bars all represent the simulated metastable state of the group-level model of the healthy control group. Figure 3 The results in (a) show that white matter structural damage in both mild cognitive impairment and Alzheimer's disease leads to a significant decrease in global metastability, indicating that the disconnection of long-range white matter connections leads to a decrease in global functional metastable dynamics, suggesting that changes in metastable dynamics in Alzheimer's patients are related to brain structural damage. Figure 3 (b) shows the effect of white matter connection damage in different resting-state subnetworks (RSN) on metastable dynamics. The results show that although the white matter connections of VIS, FPN and LIM are greatly weakened, it does not have a significant effect on the metastable neural dynamics of the whole brain. DMN and AN damage have sensitive effects on whole-brain metastability, and SMN only causes a significant decrease in metastability in the Alzheimer's disease damage model.

[0137] By exploring the relationship between anatomical abnormalities and functional dynamic changes from the perspective of local electrophysiological circuit disturbances, we found that the effect of node connection damage on functional dynamics showed regional dependence, and the graph theory topological characteristics of damaged nodes can predict the effect of node damage on functional changes. By exploring the relationship between anatomical abnormalities and functional dynamic changes from the perspective of gray matter atrophy, the results showed that all brain regions of the DMN underwent varying degrees of atrophy, and local circuit abnormalities in half of the nodes affected the metastability of the whole brain. Abnormalities in the local circuits of the FPN brain region are more likely to cause a decrease in metastability of the whole brain. By exploring the relationship between anatomical abnormalities and functional dynamic changes from the perspective of remote white matter connections, the results showed that white matter structural damage in mild cognitive impairment and Alzheimer's disease both led to a significant decrease in metastability of the whole brain, indicating that changes in metastable dynamics in patients with Alzheimer's disease are related to brain structural damage.

[0138] The above-described embodiments are used to illustrate the present invention, not to limit the present invention, so changes in the exemplified values ​​or replacement of equivalent elements should still fall within the scope of the present invention.

[0139] From the above detailed description, it can be understood by those skilled in the art that the present invention can indeed achieve the aforementioned objectives and is in compliance with the provisions of the Patent Law.

[0140] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they are aware of the basic creative concepts. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention. The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. It should be pointed out that any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

[0141] It should be noted that the above description of the relevant process is only for example and explanation, and does not limit the scope of application of this specification. For those skilled in the art, various modifications and changes can be made to the process under the guidance of this specification. However, these modifications and changes are still within the scope of this specification.

[0142] The basic concepts have been described above. Obviously, for those of ordinary skill in the art who have read this application, the above invention disclosure is only for example and does not constitute a limitation of this application. Although not explicitly stated here, those of ordinary skill in the art may make various modifications, improvements and amendments to this application. Such modifications, improvements and amendments are suggested in this application, so such modifications, improvements and amendments still belong to the spirit and scope of the exemplary embodiments of this application.

[0143] At the same time, the present application uses specific words to describe the embodiments of the present application. For example, "one embodiment", "an embodiment", and / or "some embodiments" refer to a certain feature, structure or characteristic related to at least one embodiment of the present application. Therefore, it should be emphasized and noted that "one embodiment" or "an embodiment" or "an alternative embodiment" mentioned twice or more in different positions in this specification does not necessarily refer to the same embodiment. In addition, some features, structures or characteristics in one or more embodiments of the present application can be appropriately combined.

[0144] In addition, it will be appreciated by those skilled in the art that various aspects of the present application may be illustrated and described by a number of patentable categories or situations, including any new and useful combination of processes, machines, products or substances, or any new and useful improvements thereto. Therefore, various aspects of the present application may be implemented entirely by hardware, entirely by software (including firmware, resident software, microcode, etc.), or by a combination of hardware and software. The above hardware or software may all be referred to as "units", "modules" or "systems". In addition, various aspects of the present application may take the form of a computer program product embodied in one or more computer-readable media, wherein computer-readable program code is contained therein.

[0145] The computer program code required for the operation of each part of the application can be written in any one or more programming languages, including object-oriented programming languages ​​such as Java, Scala, Smalltalk, Eiffel, JADE, Emerald, C++, C#, VB.NET, Python, etc., such as C programming language, VisualBasic, Fortran2103, Per l, COBOL2102, PHP, ABAP, such as Python, Ruby and Groovy dynamic programming language or other programming languages. The program code can be run completely on the user's computer, or run on the user's computer as an independent software package, or run partly on the user's computer and partly on the remote computer, or run completely on the remote computer or server. In the latter case, the remote computer can be connected to the user's computer through any network form, such as a local area network (LAN) or a wide area network (WAN), or connected to an external computer (e.g., via the Internet), or in a cloud computing environment, or used as a service such as software as a service (SaaS).

[0146] In addition, unless explicitly stated in the claims, the order of the processing elements and sequences described in this application, the use of alphanumeric characters, or the use of other names are not intended to limit the order of the processes and methods of this application. Although the above disclosure discusses some invention embodiments that are currently considered useful through various examples, it should be understood that such details are only for illustrative purposes, and the attached claims are not limited to the disclosed embodiments. On the contrary, the claims are intended to cover all modifications and equivalent combinations that are consistent with the essence and scope of the embodiments of this application. For example, although the implementation of the various components described above can be embodied in a hardware device, it can also be implemented as a pure software solution, for example, installation on an existing server or mobile device.

[0147] Similarly, it should be noted that in order to simplify the description disclosed in this application and thus help understand one or more embodiments of the invention, in the above description of the embodiments of the application, multiple features are sometimes combined into one embodiment, figure or description thereof. However, this method of the application should not be interpreted as reflecting the intention that the claimed object requires more features than those explicitly stated in each claim. On the contrary, the subject of the invention should have fewer features than the above single embodiment.

Claims

1. A virtual damage judgment method for Alzheimer's disease based on a brain dynamics model, characterized in that: include: S1. Obtaining structural magnetic resonance images and functional magnetic resonance images of the subject, and performing preprocessing to obtain preprocessed data, wherein the preprocessed data includes the subject's cortical thickness, cortical volume, cortical level gradient index, white matter fiber bundle structural connection matrix, white matter fiber bundle fractional anisotropy data and functional connection matrix; S2. calculating a structural network topological characteristic index based on the preprocessed white matter fiber bundle structural connection matrix; S3. Based on the structural network topological characteristic indicators, draw the required topological characteristic profile; S4. Use k-means clustering method to cluster the topological feature profiles and obtain the whole brain dynamic index - state integration rate SOIR; S5. constructing a whole-brain dynamics model for the subject based on the preprocessed white matter fiber bundle structural connection matrix and functional connection matrix; S6. Based on the whole-brain dynamics model, the cortical regions in the whole-brain dynamics model were simulated and perturbed in turn to obtain a node perturbation model, and the relationship between the structural network topological characteristic index, the cortical level gradient index and the whole-brain dynamic index was analyzed from a causal perspective; S7. Based on the whole-brain dynamics model of the subject, combined with the subject's cortical thickness and cortical volume, determine the impact of local brain atrophy on the whole-brain metastable dynamics; S8. Based on the whole-brain dynamics model of the subjects and the fractional anisotropy data of the white matter fiber bundles of the subjects, determine the impact of disconnection of white matter fiber bundles on the metastable dynamics of the whole brain.

2. The method for determining virtual damage of Alzheimer's disease based on a brain dynamics model according to claim 1, characterized in that: In S1, the preprocessing steps include: time alignment, head motion alignment, spatial normalization and low-frequency filtering.

3. The method for determining virtual damage of Alzheimer's disease based on a brain dynamics model according to claim 1, characterized in that: In S2, the calculation of structural network topology characteristic indicators includes: The node strength S of the i-th node i The calculation formula is as follows: In formula (1), M ij represents the connection weight between node i and node j, and n represents the number of nodes; The degree centrality D of the i-th node i The calculation formula is as follows: In formula (2), C ij represents the connection status between node i and node j, n represents the number of nodes; the eigenvector centrality E of the i-th node i The calculation formula is as follows: In formula (3), C ij represents the connection status between node i and node j, e j represents the importance measure of node j, λ represents the proportionality constant, and n represents the number of nodes; The clustering coefficient CC of the i-th node i The calculation formula is as follows: In formula (4), R i Indicates the maximum number of triangles formed by node i and its neighboring nodes, and the denominator k i Indicates the maximum number of triangles that node i and its neighboring nodes form a complete graph (every two nodes are connected); The betweenness centrality B of the i-th node i The calculation formula is as follows: In formula (5), W i (j,k) represents the number of shortest paths between node j and node k through node i, and W(j,k) represents the number of all shortest paths between node j and node k; In the network, the shortest path length between each node and all other nodes is the node shortest path length; the participation coefficient P of the i-th node i The calculation formula is as follows: In formula (6), D is represents the degree centrality of node i in module s, D i represents the degree centrality of node i, s represents the module, and m represents the total number of modules in the network; The module-internal degree z score Z of the i-th node i The calculation formula is as follows: In formula (7), D' i The degree centrality of node i within module s, is the average intra-module degree of all nodes in module s, is the standard deviation of the intra-module degree of all nodes in module s.

4. The method for determining virtual damage of Alzheimer's disease based on a brain dynamics model according to claim 1, characterized in that: In S4, the calculation formula of the state integration rate SOIR is as follows: In formula (8), T is the total number of time points, that is, the total number of TRs; T I Indicates the number of time points during the scan that the consolidation state was active.

5. The method for determining virtual damage of Alzheimer's disease based on a brain dynamics model according to claim 1, characterized in that: In S6, 68 cortical areas were perturbed in turn under the same degree of damage, and the perturbation method for node i was as follows: C i =SC i *damage;(9) In formula (9), SC i is all the connections of node i in the white matter fiber bundle structure connection matrix, damage is the damage factor, C i is the white matter fiber bundle structure connection matrix of node i after simulating damage.

6. The method for determining virtual damage of Alzheimer's disease based on a brain dynamics model according to claim 1, characterized in that: In S6, the impact of node perturbations on functional dynamics SOIR is evaluated. The specific steps are: S61. Paired sample t-test was used to statistically compare the SOIR values ​​simulated by the whole-brain dynamics model and the node perturbation model to determine whether there were significant differences in the results of each node perturbation model. The calculation method for evaluating the functional abnormality effect of damage to node i is as follows: ΔEVENING i =EVENING(the i )-EVENING(nonles); (10) In formula (10), ΔSOIR i is defined as the damage effect of the ith node, SOIR(les i ) is the state integration ratio of the simulated data of the damage model of node i, SOIR(nonles) is the state integration ratio of the simulated data of the complete whole-brain dynamics model; S62, after calculating the damage effect of each node, further calculate the mean AVE (ΔSOIR i ); the Pearson correlation coefficient between AVE(ΔSOIR) and the structural network topological characteristic index and cortical level gradient index was further calculated to test whether the cortical level gradient index and the structural network topological characteristic index of the node were related to the SOIR changes caused by the lesion.

7. The method for determining virtual damage of Alzheimer's disease based on a brain dynamics model according to claim 1, characterized in that: In S7, the effect of atrophy in different brain regions of Alzheimer's disease on metastable dynamics is determined. By changing the local recurrent synaptic coupling factor factor, the node model of the abnormal cortical structure of the patient is virtually damaged. The specific steps are as follows: S71, virtually perturbing the nodes where the cortical thickness and cortical volume are significantly different between the healthy control group and the patient group, that is, virtually perturbing the local recurrent synaptic coupling factor of these nodes, thereby simulating the imbalance of the local circuit; The group-level model of the healthy control group was constructed based on the white matter fiber bundle structure connection matrix and the optimal local recurrent synaptic coupling parameters. The group-level model of the healthy control group and the node perturbation model of local circuit imbalance used consistent simulation conditions to simulate brain activity, and the local recurrent synaptic coupling factor of abnormal node i was perturbed as follows: factor i =factor*damage;(11) In formula (11), factor i is the local recurrent synaptic coupling parameter after disturbance, factor is the optimal local recurrent synaptic coupling parameter of the group-level model of the healthy control group, and damage is the damage factor; S72. Evaluate the effect of local recurrent synaptic coupling factor perturbation on metastable dynamics. Paired sample t-test was used to statistically compare the whole-brain metastability simulated by the group-level model of the healthy control group and the node perturbation model with local circuit imbalance to determine whether virtual damage to the node would affect the whole-brain metastable dynamics.

8. The method for determining virtual damage of Alzheimer's disease based on a brain dynamics model according to claim 1, characterized in that: S8 includes: S81: Virtually perturb the edges with significant differences between the healthy control group and the patient group in the fractional anisotropy values ​​of white matter fiber bundles, that is, reduce the structural connection strength of these edges by 50%, and perform the above operation on all edges with significant differences; the construction method of the white matter fiber bundle structural connection matrix l_C after Alzheimer's disease and mild cognitive impairment damage is as follows: In formula (12), l_C ij represents the connection weight between node i and node j in the white matter fiber bundle structure connection matrix l_C after damage, SC ij is the connection weight between node i and node j in the group-level average SC of the healthy control group, FA ij is the fractional anisotropy value of the white matter fiber bundle between node i and node j; S82. Determine the effect of white matter connection damage of different resting state subnets on metastable dynamics, and perturb the connection of each resting state subnet to generate 7 resting state subnet white matter perturbation models, which are realized by the white matter fiber bundle structure connection matrix SC of the damage model; the construction method of the damaged white matter fiber bundle structure connection matrix l_Ck of different resting state subnets is as follows: In formula (13), is the connection weight between node i and node j in the white matter fiber bundle structure connection matrix l_Ck after the kth resting state subnetwork is damaged, SC ij is the connection weight between node i and node j in the group-level average white matter fiber bundle structure connection matrix of the healthy control group, FA ij is the fractional anisotropy value of the white matter fiber bundle between node i and node j, RSN k is the set of brain regions included in the k-th subnetwork; S83. Paired sample t-test was used to statistically compare the whole-brain metastable state simulated by the group-level model of the healthy control group and the different injury models to determine whether there were significant differences in the results of each white matter disturbance model.

Images