Method, device, electronic device and storage medium for generating cerebral cortical morphological network

Abstract

The present application relates to a method, device, electronic device and storage medium for generating a cerebral cortical morphological network, wherein the method comprises: obtaining a cerebral cortical reconstructed image of a target object; segmenting the cerebral cortical reconstructed image to obtain multiple target areas of the cerebral cortex and a vertex set of each target area; connecting the vertices in the vertex set of each target area to obtain a sub-network corresponding to each target area; determining the state of the target leaf node based on whether the target leaf node is adjacent to other leaf nodes; if the state of the target leaf node is a to-be-connected state, determining a connecting leaf node corresponding to the target leaf node; connecting the leaf nodes in the to-be-connected state in each sub-network to the corresponding connecting leaf nodes to generate a cerebral cortical morphological network of the target object. Through the present application, the problem of low efficiency in generating a cerebral cortical morphological network is solved, and the efficiency of generating a cerebral cortical morphological network is improved.

Description

Technical Field

[0001] The present application relates to the field of medical image processing technology, and in particular to a method, device, electronic device and storage medium for generating a cerebral cortex morphology network. Background Art

[0002] Magnetic resonance imaging (MRI) technology enables the study of folding patterns in the cerebral cortex in vivo, attracting a growing number of researchers to the study of the brain. The brain is a complex network system whose developmental changes are closely related to human higher-order cognitive abilities. The cerebral cortex is typically composed of a folding pattern of concave sulci and convex gyri. Intuitively, the cerebral cortex is somewhat analogous to the "gullies" of nature, with sulci corresponding to concave valleys and gyri corresponding to ridges. Furthermore, sulci and gyri play different structural and functional roles in the brain. While some researchers have studied cerebral cortical morphology, most have analyzed brain morphological features in isolation, rather than analyzing them as "graphs" or "networks" in the form of structural or functional networks. Therefore, it is necessary to extract gyral ridges and sulci from the cerebral cortex and generate corresponding network morphological maps, thereby facilitating further study and analysis using graph theory.

[0003] Currently, there are related technologies that, when establishing a gyral network morphological graph, first segment the gyral ridges to obtain a set of gyral ridge vertices. Furthermore, multiple gyral ridge vertices are connected into subtrees through a tree marching algorithm. Furthermore, all subtrees are connected into a large tree, and then the leaf nodes of the large tree are pruned to obtain a gyral morphological network. In the above process, in order to avoid the generation of small loops after the subtrees are connected, all loops with a length less than a threshold need to be deleted, and only loops with a length greater than this threshold need to be retained. Furthermore, the leaf nodes of the connected large tree need to be pruned. However, the process of traversing the lengths of all loops and pruning the leaf nodes of the connected large tree consumes a lot of time, which ultimately leads to low efficiency in generating a cortical morphological network.

[0004] Regarding the low efficiency of generating cortical morphological networks in related technologies, no effective solution has been proposed so far. Summary of the Invention

[0005] In this embodiment, a method, device, electronic device, and storage medium for generating a cerebral cortical morphological network are provided to solve the problem of low efficiency in generating a cerebral cortical morphological network in related technologies.

[0006] In a first aspect, this embodiment provides a method for generating a cerebral cortical morphological network, comprising:

[0007] Acquire a reconstructed image of the cerebral cortex of the target object;

[0008] Segmenting the reconstructed cerebral cortex image to obtain a plurality of target regions of the cerebral cortex and a vertex set of each target region;

[0009] Connecting the vertices in the vertex set of each target area to obtain a subnetwork corresponding to each target area;

[0010] Determining a state of the target leaf node based on whether the target leaf node in the target subnetwork is adjacent to other leaf nodes, wherein the target subnetwork is any subnetwork among multiple subnetworks, the target leaf node is any leaf node in the target subnetwork, and the other leaf nodes are any leaf nodes in the multiple subnetworks other than the target leaf node, and the state of the target leaf node includes a pending connection state or a pending deletion state;

[0011] If the state of the target leaf node is a waiting state, determining a connection leaf node corresponding to the target leaf node;

[0012] The leaf nodes in the waiting-to-connect state in each sub-network are connected to the corresponding connected leaf nodes to generate the cerebral cortex morphological network of the target object.

[0013] In some embodiments, the segmenting of the cerebral cortex reconstructed image to obtain multiple target regions of the cerebral cortex and a vertex set of each target region includes:

[0014] Acquiring morphological characteristics of the cerebral cortex of the target object;

[0015] The cerebral cortex reconstructed image is segmented based on the cerebral cortex morphological features to obtain multiple target areas of the cerebral cortex and a vertex set of each target area.

[0016] In some embodiments, segmenting the reconstructed cerebral cortex image based on the morphological features of the cerebral cortex to obtain multiple target regions of the cerebral cortex and a vertex set for each target region includes:

[0017] Determining a flooding parameter of a watershed algorithm based on the morphological characteristics of the cerebral cortex;

[0018] The reconstructed image of the cerebral cortex is segmented based on the flooding parameter to obtain a plurality of target regions of the cerebral cortex and a vertex set of each target region.

[0019] In some embodiments, the cerebral cortex morphological features include N types, where N is a positive integer greater than or equal to 2, and determining the flooding parameter of the watershed algorithm based on the cerebral cortex morphological features includes:

[0020] A flooding parameter of a watershed algorithm is determined based on the feature values ​​of at least two types of cerebral cortex morphological features of the target object.

[0021] In some embodiments, the morphological features of the cerebral cortex include at least the depth of the sulcus, and the segmentation of the reconstructed cerebral cortex image based on the flooding parameter to obtain multiple target regions of the cerebral cortex and a vertex set of each target region includes:

[0022] Segmenting the reconstructed cerebral cortex image based on the flooding parameter to obtain a plurality of target regions of the cerebral cortex and a candidate vertex set for each target region;

[0023] Based on the characteristic value of the sulcus depth of the target object, the candidate vertex set of each target region is screened to obtain the vertex set of each target region.

[0024] In some embodiments, connecting the vertices in the vertex set of each target area to obtain a subnetwork corresponding to each target area includes:

[0025] Performing distance transformation on vertices in the vertex set of each target area to obtain a distance distribution value of each vertex in each target area;

[0026] Based on the distance distribution values ​​of the vertices in each target area, the vertices in the vertex set of the target area are connected in descending order to obtain a sub-network corresponding to each target area.

[0027] In some embodiments, the step of connecting vertices in the vertex set of the target area in descending order based on the distance distribution values ​​of the vertices in each target area to obtain a subnetwork corresponding to each target area includes:

[0028] Determining a travel threshold of a current batch and vertices of the current batch whose distance distribution values ​​are less than or equal to the travel threshold based on the distance distribution values ​​of each vertex in each target area;

[0029] The vertices of the current batch are connected in descending order according to the distance distribution value until the travel threshold of the current batch reaches zero, thereby obtaining a sub-network corresponding to each target area.

[0030] In some embodiments, determining the state of the target leaf node based on whether the target leaf node is adjacent to other leaf nodes in the target subnetwork includes:

[0031] If the target leaf node is not adjacent to any other leaf node, the state of the target leaf node is to be deleted;

[0032] If the target leaf node is adjacent only to leaf nodes in other subnetworks, the state of the target leaf node is a pending connection state, and the other subnetworks are any subnetworks in the plurality of subnetworks except the target subnetwork;

[0033] If the target leaf node is only adjacent to other target leaf nodes in the target sub-network, then determine whether the other target leaf nodes adjacent to the target leaf node have adjacent leaf nodes in other sub-networks; if the other target leaf nodes adjacent to the target leaf node do not have adjacent leaf nodes in other sub-networks, then the state of the target leaf node and the leaf node corresponding to the maximum value of the cerebral cortical morphological feature in the other target leaf nodes adjacent to the target leaf node is determined to be a to-be-connected state; and, the states of the target leaf node and the other target leaf nodes adjacent to the target leaf node, except for the leaf node corresponding to the maximum value of the cerebral cortical morphological feature, are all to be deleted; if the other target leaf nodes adjacent to the target leaf node have adjacent leaf nodes in other sub-networks, then the state of the target leaf node is to be deleted;

[0034] If the target leaf node is adjacent to other target leaf nodes, and the target leaf node is adjacent to leaf nodes in other subnetworks, then determine whether the other target leaf nodes adjacent to the target leaf node have adjacent leaf nodes in other subnetworks; if the other target leaf nodes adjacent to the target leaf node do not have adjacent leaf nodes in other subnetworks, then the state of the target leaf node is a to-be-connected state; if the other target leaf nodes adjacent to the target leaf node have adjacent leaf nodes in other subnetworks, then the state of the leaf node corresponding to the maximum value of the cerebral cortical morphological feature in the target leaf node and other target leaf nodes adjacent to the target leaf node is determined to be a to-be-connected state; and the state of the remaining leaf nodes in the target leaf node and other target leaf nodes adjacent to the target leaf node, except for the leaf node corresponding to the maximum value of the cerebral cortical morphological feature, are all in a to-be-deleted state.

[0035] In some embodiments, determining the connected leaf node corresponding to the target leaf node includes:

[0036] If the target leaf node is adjacent only to leaf nodes in other sub-networks, the connected leaf node is the leaf node corresponding to the maximum value of the cerebral cortical morphological feature among all leaf nodes adjacent to the target leaf node;

[0037] If the target leaf node is adjacent only to other target leaf nodes in the target subnetwork, and other target leaf nodes adjacent to the target leaf node do not have adjacent leaf nodes in other subnetworks, then the connected leaf node is the leaf node corresponding to the maximum value of the cerebral cortical morphological feature in the other subnetwork closest to the target subnetwork;

[0038] If the target leaf node is adjacent to other target leaf nodes, and the target leaf node is adjacent to leaf nodes in other sub-networks, then the connected leaf node is the leaf node corresponding to the maximum value of the cerebral cortical morphological feature in other sub-networks adjacent to the target leaf node.

[0039] In a second aspect, this embodiment provides a device for generating a cerebral cortical morphology network, including:

[0040] An acquisition module, used for acquiring a reconstructed image of the cerebral cortex of a target object;

[0041] a segmentation module, configured to segment the cerebral cortex reconstructed image to obtain a plurality of target regions of the cerebral cortex and a vertex set of each target region;

[0042] a connection module, configured to connect vertices in the vertex set of each target area to obtain a subnetwork corresponding to each target area;

[0043] A first determination module is configured to determine a state of a target leaf node based on whether the target leaf node is adjacent to other leaf nodes in a target subnetwork, wherein the target subnetwork is any subnetwork among multiple subnetworks, the target leaf node is any leaf node in the target subnetwork, and the other leaf nodes are any leaf nodes other than the target leaf node in the multiple subnetworks, and the state of the target leaf node includes a pending connection state or a pending deletion state;

[0044] A second determining module, if the state of the target leaf node is a waiting state, is used to determine a connecting leaf node corresponding to the target leaf node;

[0045] A generation module is used to connect the leaf nodes in the waiting-to-connect state in each sub-network with the corresponding connection leaf nodes to generate the cerebral cortex morphological network of the target object.

[0046] In a third aspect, an electronic device is provided in this embodiment, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein when the processor executes the computer program, the method for generating a cerebral cortical morphological network as described in any embodiment of the first aspect is implemented.

[0047] In a fourth aspect, a storage medium is provided in this embodiment, on which a computer program is stored. When the program is executed by a processor, the method for generating a cerebral cortical morphological network described in any embodiment of the first aspect above is implemented.

[0048] Compared with the related art, the cerebral cortex morphological network generation method provided in this embodiment obtains multiple target areas of the cerebral cortex and a vertex set of each target area by segmenting the cerebral cortex reconstructed image of the target object, and connects the vertices in the vertex set of each target area to obtain a corresponding sub-network. Furthermore, based on whether the target leaf node in the target sub-network is adjacent to other leaf nodes, it is determined whether the target leaf node is in a to-be-connected state or a to-be-deleted state. If the state of the target leaf node is in a to-be-connected state, the connection leaf node corresponding to the target leaf node is determined; further, the leaf nodes in the to-be-connected state in each sub-network are connected to the corresponding connection leaf nodes to generate the cerebral cortical morphological network of the target object. Therefore, before connecting the sub-networks, the state of the target leaf node is determined based on whether the target leaf node is adjacent to other leaf nodes. Furthermore, the connecting leaf node corresponding to the target leaf node in the to-be-connected state is determined, and the leaf nodes in the to-be-connected state in each sub-network are connected to the corresponding connecting leaf nodes. The cerebral cortical morphological network can be directly obtained. This scheme does not require the judgment of the loop length in the process of generating the cerebral cortical morphological network, and omits the process of pruning the leaf nodes of the connected network, thereby improving the efficiency of generating the cerebral cortical morphological network.

[0049] The details of one or more embodiments of the present application are set forth in the following drawings and description to make other features, objects, and advantages of the present application more readily apparent. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0051] Figure 1 This is a schematic diagram of an application scenario of a method for generating a cerebral cortical morphology network provided in an embodiment of the present application;

[0052] Figure 2 This is a flowchart of a method for generating a cerebral cortical morphology network provided in an embodiment of the present application;

[0053] Figure 3 This is a flowchart of an embodiment of a method for generating a cerebral cortical morphology network provided in an embodiment of the present application;

[0054] Figure 4This is a schematic diagram of the vertices of a gyrus region provided in an embodiment of the present application;

[0055] Figure 5 This is a schematic diagram of a gyrus region subnetwork provided in an embodiment of the present application;

[0056] Figure 6 This is a schematic diagram of a gyrus morphology network provided in an embodiment of the present application;

[0057] Figure 7 This is a structural block diagram of a cerebral cortical morphology network generation device provided in an example of the present application;

[0058] Figure 8 This is a schematic diagram of the structure of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0059] In order to more clearly understand the purpose, technical solutions and advantages of the present application, the present application is described and illustrated below in conjunction with the accompanying drawings and embodiments.

[0060] Unless otherwise defined, the technical terms or scientific terms involved in this application should have the general meaning understood by people with ordinary skills in the technical field to which this application belongs. The words "one", "an", "a", "the", "these" and the like in this application do not indicate quantitative restrictions, and they can be singular or plural. The terms "include", "comprise", "have" and any variants thereof involved in this application are intended to cover non-exclusive inclusions; for example, a process, method and system, product or device comprising a series of steps or modules (units) is not limited to the listed steps or modules (units), but may include unlisted steps or modules (units), or may include other steps or modules (units) inherent to these processes, methods, products or devices. The words "connect", "connected", "coupled" and the like involved in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The "plurality" involved in this application refers to two or more. "And / or" describes the relationship between related objects, indicating that three possible relationships exist. For example, "A and / or B" can represent: A exists alone, A and B exist simultaneously, and B exists alone. Generally, the character " / " indicates that the related objects are in an "or" relationship. The terms "first," "second," "third," etc. used in this application are only used to distinguish similar objects and do not represent a specific ordering of the objects.

[0061] The method for generating a cerebral cortical morphological network provided in the embodiment of the present application can be applied to Figure 1 In the application environment shown, Figure 1: This is a schematic diagram of an application scenario of a method for generating a cerebral cortical morphology network provided in an embodiment of the present application. The terminal 102 communicates with the server 104 via a network. The data storage system can store data that the server 104 needs to process. Specifically, the data stored in the data storage system can be data related to the cerebral cortex of the target object, such as structural magnetic resonance data, cerebral cortical morphology indicator data, etc. The data storage system can be integrated on the server 104, or it can be placed on the cloud or other network servers. The terminal 102 can be, but is not limited to, various personal computers, laptops, smart phones, tablets, Internet of Things devices and portable wearable devices. The Internet of Things devices can be smart speakers, smart TVs, smart air conditioners, smart car-mounted devices, etc. Portable wearable devices can be smart watches, smart bracelets, head-mounted devices, etc. The server 104 can be implemented as an independent server or a server cluster consisting of multiple servers.

[0062] Magnetic resonance imaging (MRI) technology enables the study of folding patterns in the cerebral cortex in vivo, attracting a growing number of researchers to the study of the brain. The brain is a complex network system whose developmental changes are closely related to human higher-order cognitive abilities. The cerebral cortex is typically composed of a folding pattern of concave sulci and convex gyri. Intuitively, the cerebral cortex is somewhat analogous to the "gullies" of nature, with sulci corresponding to concave valleys and gyri corresponding to ridges or peaks. Furthermore, sulci and gyri play different structural and functional roles in the brain. Currently, some researchers have studied cerebral cortical morphology, but most analyze brain morphological features in isolation, rather than analyzing them as a "graph" or "network" using structural or functional networks. Therefore, the goal is to extract gyral ridges and sulci from the cerebral cortex and form corresponding network morphological maps, thereby facilitating further study and analysis using graph theory.

[0063] Currently, related technologies for building gyral network morphological graphs first segment the gyral ridges to obtain a set of gyral ridge vertices. Furthermore, multiple gyral ridge vertices are connected into subtrees using a tree marching algorithm. Furthermore, all subtrees are connected into a large tree, and then the leaf nodes of the large tree are pruned to obtain a gyral morphological network. In the process of connecting subtrees, to avoid the generation of small loops, all loops with a length less than a threshold are deleted, and only loops with a length greater than this threshold are retained. However, traversing the lengths of all loops and pruning the leaf nodes of the connected large tree consumes a considerable amount of time, ultimately resulting in a low efficiency in generating cortical morphological networks.

[0064] Therefore, how to improve the efficiency of cortical morphological network generation is a problem that needs to be solved.

[0065] In the embodiment of the present application, a method for generating a cerebral cortical morphological network is provided. Figure 2 This is a flowchart of a method for generating a cerebral cortical morphology network provided by an embodiment of the present application. The execution subject of the method may be an electronic device. Optionally, the electronic device may be a server or a terminal device, but the present application is not limited thereto. Specifically, Figure 2 As shown, the process includes the following steps:

[0066] Step S201: Acquire a reconstructed image of the cerebral cortex of the target object.

[0067] Exemplarily, a structural magnetic resonance imaging (sMRI) of the target subject's brain, ie, a T1 structural image, is acquired. Further, gray-white matter segmentation and cortical surface reconstruction are performed on the T1 structural image to obtain a reconstructed image of the target subject's cerebral cortex.

[0068] Specifically, the FreeSurfer tool can be used to perform gray-white matter segmentation and cortical surface reconstruction on the T1 structural image of the target object to obtain a reconstructed image of the cerebral cortex of the target object, as well as the morphological characteristics of the cerebral cortex of the target object. The morphological characteristics of the cerebral cortex may include sulcus depth, cortical thickness, cortical curvature, volume, and surface area.

[0069] It should be noted that in the embodiments of the present application, only the FreeSurfer tool is used to perform cerebral cortex reconstruction and obtain a cerebral cortex reconstructed image as an example for explanation. In actual applications, FastSurfer can also be used to perform cerebral cortex reconstruction and obtain a cerebral cortex reconstructed image, and other software can also be used to perform cerebral cortex reconstruction and obtain a cerebral cortex reconstructed image. There is no limitation here.

[0070] Step S202 : segmenting the reconstructed cerebral cortex image to obtain a plurality of target regions in the cerebral cortex and a vertex set of each target region.

[0071] Furthermore, the reconstructed image of the cerebral cortex of the target object is segmented to obtain a plurality of target regions in the cerebral cortex of the target object and a vertex set of each target region.

[0072] Specifically, the cerebral cortex includes gyri and sulci, which correspond to valleys and ridges on the cerebral cortex, respectively. The target region can be a gyrus region or a sulcus region. In the embodiments of this application, the target region is described as a gyrus region. By segmenting the reconstructed cerebral cortex image, multiple gyrus regions of the target object and the vertex set corresponding to each gyrus region can be obtained.

[0073] Step S203 : Connect the vertices in the vertex set of each target area to obtain a sub-network corresponding to each target area.

[0074] Furthermore, the vertices in each target region are connected to obtain a subnetwork corresponding to each target region. Specifically, the vertices in multiple gyrus regions are connected to obtain a subnetwork corresponding to each gyrus region of the target object.

[0075] Step S204 : determining the state of the target leaf node based on whether the target leaf node is adjacent to other leaf nodes in the target sub-network.

[0076] Among them, the target subnetwork is any subnetwork among the multiple subnetworks, the target leaf node is any leaf node in the target subnetwork, and the other leaf nodes are any leaf nodes in the multiple subnetworks except the target leaf node. The status of the target leaf node includes a waiting-to-connect status or a waiting-to-delete status.

[0077] Exemplarily, among all leaf nodes in all sub-networks, it is determined whether there is a leaf node adjacent to the target leaf node in the target sub-network, thereby determining the state of the target leaf node.

[0078] A leaf node may refer to a node in a sub-network that has no child nodes, i.e., a terminal node. The status of a leaf node may include a pending connection state or a pending deletion state.

[0079] Specifically, in all sub-networks, it is determined whether there are leaf nodes adjacent to the target leaf node, so that the state of the target leaf node without adjacent leaf nodes can be determined as a state to be deleted, thereby determining the state of each leaf node in advance.

[0080] Step S205: If the state of the target leaf node is the waiting state, determine the connection leaf node corresponding to the target leaf node.

[0081] Exemplarily, if the state of the target leaf node is a waiting-to-connect state, a connection leaf node corresponding to the target leaf node is determined based on leaf nodes adjacent to the target leaf node, thereby facilitating the connection of the target leaf node.

[0082] Step S206: Connect the leaf nodes in the waiting-to-connect state in each sub-network with the corresponding connected leaf nodes to generate a cerebral cortical morphological network of the target object.

[0083] For example, in each subnetwork, leaf nodes in the pending connection state are connected to corresponding connected leaf nodes, and the leaf nodes in the pending deletion state are deleted, thereby completing the subnetwork connection and obtaining the target subject's cerebral cortical morphological network. Therefore, during the network connection process, there is no need to determine the loop length, and after the subnetwork is connected, there is no need to prune the leaf nodes of the connected network. Only the leaf nodes in the pending connection state are connected, further improving the cerebral cortical morphological network.

[0084] Specifically, if the target region is the gyrus region of the target object, the obtained cortical morphological network is the gyrus morphological network of the target object.

[0085] In the above implementation process, the reconstructed image of the cerebral cortex of the target object is segmented to obtain multiple target areas of the cerebral cortex and the vertex set corresponding to each target area, and the vertices in the vertex set of each target area are connected respectively to obtain the sub-network corresponding to each target area. Furthermore, according to whether the target leaf node is adjacent to other leaf nodes, it is determined whether the target leaf node is in a to-be-connected state or a to-be-deleted state. Therefore, before connecting the sub-network, the state of the target leaf node is determined according to whether the target leaf node is adjacent to other leaf nodes. Further, the connecting leaf node corresponding to the target leaf node in the to-be-connected state is determined, and the target leaf node in the to-be-connected state in each sub-network is connected to the corresponding connecting leaf node. Therefore, in the process of connecting the sub-networks, there is no need to traverse the loop length between the sub-networks, and after the sub-networks are connected, there is no need to perform the leaf node deletion process, which further improves the efficiency of generating the cerebral cortical morphological network.

[0086] In some embodiments, segmenting the cerebral cortex reconstructed image to obtain multiple target regions of the cerebral cortex and a vertex set of each target region may include the following steps:

[0087] Step 1: Obtain the morphological features of the cerebral cortex of the target object.

[0088] Step 2: Segment the reconstructed cerebral cortex image based on the morphological features of the cerebral cortex to obtain multiple target areas of the cerebral cortex and a vertex set of each target area.

[0089] For example, by performing cerebral cortex reconstruction on the structural magnetic resonance image of the target object, not only the cerebral cortex reconstruction image can be obtained, but also the morphological characteristics of the cerebral cortex of the target object can be obtained.

[0090] Furthermore, the reconstructed cortical image is segmented based on the acquired cortical morphological features to obtain multiple target regions and a vertex set for each target region. Specifically, taking the gyrus region as an example, the morphological features of the target object's cortex are used as segmentation parameters to segment the reconstructed cortical image of the target object, obtaining multiple gyrus regions and a vertex set corresponding to each gyrus region.

[0091] In the above implementation process, the cerebral cortex morphological features of the target object are used as segmentation parameters to segment the cerebral cortex reconstructed image, which can make the segmented gyrus area and the corresponding vertex set adapt to the cerebral cortex morphological features of the target object, avoiding the segmentation error caused by segmenting the cerebral cortex reconstructed images of different individuals by setting fixed segmentation parameters, thereby improving the flexibility of the cerebral cortex segmentation algorithm and the accuracy of the segmentation results.

[0092] In some embodiments, segmenting a reconstructed cerebral cortex image based on morphological features of the cerebral cortex to obtain multiple target regions of the cerebral cortex and a vertex set for each target region may include the following steps:

[0093] Step 1: Determine the flooding parameters of the watershed algorithm based on the morphological characteristics of the cerebral cortex.

[0094] Step 2: Based on the flooding parameter, the reconstructed image of the cerebral cortex is segmented to obtain multiple target areas of the cerebral cortex and a vertex set of each target area.

[0095] Exemplarily, the minimum value of the morphological features of the cerebral cortex is used as the initial value of the watershed algorithm, and the maximum value of the morphological features of the cerebral cortex is used as the end value of the watershed algorithm to obtain the watershed parameter of the watershed algorithm.

[0096] Furthermore, according to the determined flooding parameters, the reconstructed image of the cerebral cortex is segmented by a watershed algorithm to obtain vertex sets of multiple target areas in the cerebral cortex.

[0097] Specifically, the morphological characteristics of the cerebral cortex may include at least one of sulcus depth, cortical thickness, cortical curvature, volume, and surface area.

[0098] Taking cortical curvature as an example of the morphological feature of the cerebral cortex, the minimum value of the cortical curvature of the target object is used as the initial value of the watershed algorithm, and the maximum value of the cortical curvature of the target object is used as the end value of the watershed algorithm.

[0099] Furthermore, when segmenting the cerebral cortical reconstructed image using the watershed algorithm, the water level is increased starting from the minimum value of the cortical curvature of the target object and ending at the maximum value of the cortical curvature of the target object, thereby obtaining multiple gyrus regions of the target object and the vertex set of each gyrus region. For example, when segmenting the cerebral cortical reconstructed image using the watershed algorithm, the water level can be determined based on the percentage of the difference between the maximum and minimum values ​​of the morphological features of the cerebral cortex. For example, taking the cortical curvature as an example, 5% of the difference between the maximum and minimum values ​​of the cortical curvature can be determined as the water level for each segmentation. Then, when segmenting the cerebral cortical reconstructed image using the watershed algorithm, the water level is increased starting from the minimum value of the cortical curvature, and the entire water level increase process is performed 20 times until the maximum value of the cortical curvature is reached.

[0100] As another embodiment, the sulcus depth is taken as an example of a morphological feature of the cerebral cortex, the minimum value of the sulcus depth of the target object is used as the initial value of the watershed algorithm, and the maximum value of the sulcus depth of the target object is used as the end value of the watershed algorithm to obtain the watershed parameter of the watershed algorithm.

[0101] Furthermore, when the reconstructed cortical image is segmented by the watershed algorithm, the water level is increased from the minimum value of the sulcus depth of the target object until it reaches the maximum value of the sulcus depth of the target object, thereby obtaining multiple gyrus regions of the target object and a vertex set of each gyrus region.

[0102] As another embodiment, the reconstructed cortical image is segmented according to the cortical curvature and sulcus depth of the target object to obtain multiple target regions of the cortex and a vertex set for each target region. Specifically, a first flood parameter of a watershed algorithm is determined according to the cortical curvature of the target object, and the reconstructed cortical image is segmented according to the first flood parameter to obtain multiple first gyrus regions and a first vertex set corresponding to each first gyrus region; the reconstructed cortical image is segmented according to the sulcus depth of the target object to obtain multiple second gyrus regions and a second vertex set corresponding to each second gyrus region; the intersection of the first gyrus region and the corresponding second gyrus region is used as the corresponding gyrus region, and the intersection of the first vertex set and the second vertex set is used as the vertex set corresponding to the gyrus region; or, the union of the first gyrus region and the corresponding second gyrus region is used as the corresponding gyrus region, and the union of the first vertex set and the second vertex set is used as the vertex set corresponding to the gyrus region.

[0103] It should be noted that the embodiments of the present application only use the above-mentioned several cortical morphological features to segment the cortical reconstructed image as an example. When actually performing the segmentation of the cortical reconstructed image, various other cortical morphological features can also be incorporated, and there is no limitation here.

[0104] In the above implementation process, the rising water parameters of the watershed algorithm are determined according to the morphological characteristics of the cerebral cortex. Furthermore, the cerebral cortex reconstructed image is segmented by the watershed algorithm according to the determined rising water parameters. Therefore, when the watershed algorithm is used to segment the cerebral cortex reconstructed image, the rising water parameters can be adaptively adjusted for different individuals, thereby improving the accuracy of the cerebral cortex reconstructed image segmentation.

[0105] In some embodiments, the cerebral cortical morphological features include N types, N is a positive integer greater than or equal to 2, and the rising water parameters of the watershed algorithm are determined based on the cerebral cortical morphological features, including: determining the rising water parameters of the watershed algorithm based on the characteristic values ​​of at least two types of cerebral cortical morphological features of the target object.

[0106] For example, the morphological features of the cerebral cortex may include at least two types of features among sulcus depth, cortical thickness, cortical curvature, volume, and surface area. Further, based on the feature values ​​of the at least two types of morphological features of the cerebral cortex, a watershed parameter of the watershed algorithm is determined.

[0107] Specifically, taking the sulcus depth and cortical curvature as examples, the average of the minimum eigenvalue of the sulcus depth of the target object and the minimum eigenvalue of the cortical curvature is used as the initial water rise value of the watershed algorithm, and the average of the maximum eigenvalue of the sulcus depth of the target object and the maximum eigenvalue of the cortical curvature is used as the end water rise value of the watershed algorithm, thereby obtaining the water rise parameters of the watershed algorithm.

[0108] As another embodiment, the weighted average of the minimum eigenvalue of the sulcus depth of the target object and the minimum eigenvalue of the cortical curvature can be used as the initial water rise value of the watershed algorithm, and the weighted average of the maximum eigenvalue of the sulcus depth of the target object and the maximum eigenvalue of the cortical curvature can be used as the end water rise value of the watershed algorithm, thereby obtaining the water rise parameter of the watershed algorithm.

[0109] It should be noted that the embodiments of the present application are only described by taking the average value or weighted average value as an example to determine the initial value and the end value of the water rise. In actual applications, other multi-feature combination calculation methods can also be used to determine the initial value and the end value of the water rise, which is not limited here. In addition, the embodiments of the present application are only described by taking the two types of sulcus depth and cortical curvature as examples. In actual applications, the characteristic values ​​of 3 types or 4 types of cerebral cortical morphological features can also be used to determine the water rise parameters of the watershed algorithm, which is not limited here.

[0110] Furthermore, when the reconstructed cortical image is segmented by the watershed algorithm, the water level is increased from the initial water level of the target object until the water level reaches the end water level, thereby obtaining multiple gyrus regions of the target object and a vertex set of each gyrus region.

[0111] In the above implementation process, the rising water parameters of the watershed algorithm are determined according to the characteristic values ​​of at least two types of morphological features of the cerebral cortex, thereby improving the accuracy of determining the rising water parameters of the watershed algorithm, and further improving the accuracy of the division of the cerebral cortex area and the vertex set.

[0112] In some embodiments, the morphological features of the cerebral cortex include at least the depth of the sulci. Segmenting the reconstructed cerebral cortex image based on the flooding parameter to obtain multiple target regions of the cerebral cortex and a vertex set for each target region may include the following steps:

[0113] Step 1: Segment the reconstructed cerebral cortex image based on the flooding parameter to obtain multiple target regions of the cerebral cortex and a set of candidate vertices for each target region.

[0114] Step 2: Based on the characteristic value of the sulcus depth of the target object, the candidate vertex set of each target area is screened to obtain the vertex set of each target area.

[0115] Exemplarily, after determining the flooding parameter of the watershed algorithm according to the cortical curvature of the target object, the reconstructed image of the cerebral cortex is segmented according to the determined flooding parameter to obtain multiple target areas of the cerebral cortex and a set of candidate vertices for each target area.

[0116] Furthermore, the candidate vertex set of each target region is selected according to the characteristic value of the sulcus depth of the target object to obtain the vertex set of each target region.

[0117] Specifically, a sulcus depth threshold is determined. In the candidate vertex set of each target area, candidate vertices with sulcus depth values ​​greater than the sulcus depth threshold are retained, and candidate vertices with sulcus depth values ​​less than or equal to the sulcus depth threshold are deleted, thereby obtaining a vertex set corresponding to each gyrus area.

[0118] If the sulcus depth threshold is zero, then in the candidate vertex set of each gyrus region, the candidate vertices with sulcus depth values ​​greater than zero are retained, and the candidate vertices with sulcus depth values ​​less than or equal to zero are deleted, thereby obtaining the vertex set corresponding to each gyrus region.

[0119] As another example, to handle redundancy, 10% of the negative maximum value of the sulcus depth can be used as the sulcus depth threshold. Furthermore, within the set of candidate vertices for each target region, candidate vertices with sulcus depths greater than the threshold are retained, while candidate vertices with sulcus depths less than or equal to the threshold are deleted, thereby obtaining a vertex set corresponding to each target region. For example, if the range of sulcus depth is [-1.5, 1.5], the sulcus depth threshold is -1.5*10%=-0.15.

[0120] In the above implementation process, the vertices in the candidate vertex set are selected by the characteristic value of the sulcus depth, which can effectively eliminate the vertices in the non-target area and improve the accuracy of determining the vertex set in each target area.

[0121] In some embodiments, connecting the vertices in the vertex set of each target area to obtain a subnetwork corresponding to each target area may include the following steps:

[0122] Step 1: Perform distance transformation on the vertices in the vertex set of each target area to obtain the distance distribution value of each vertex in each target area.

[0123] Step 2: Based on the distance distribution value of each vertex in each target area, connect the vertices in the vertex set of the target area in order from large to small to obtain the sub-network corresponding to each target area.

[0124] Exemplarily, distance transformation is performed on vertices in the vertex set of each target region to obtain distance distribution values ​​of vertices in each target region, and the distance values ​​of vertices in each target region gradually decrease from the center to the periphery.

[0125] Furthermore, according to the distance distribution value of each vertex in each target area, the vertices in the vertex set of the target area are connected in order from large to small through a marching tree method to obtain a sub-network corresponding to each target area.

[0126] Exemplarily, performing distance transformation on vertices in the vertex set of each target region to obtain a distance distribution value of each vertex in each target region may include:

[0127] Step 1: integerize the distance value of the target vertex in each target area to obtain the integerized distance value of the target vertex, wherein the target vertex is any vertex in the target area.

[0128] Step 2: Based on the integerized distance value of the target vertex and the number of vertices adjacent to the target vertex, the integerized distance value of the target vertex is non-integerized to obtain the distance distribution value of each vertex in each target area.

[0129] Specifically, any one of the multiple gyral regions is selected as the target gyral region, and vertices in the target gyral region are sequentially determined as target vertices from the outside to the inside. Each vertex in the target gyral region is assigned an integer distance value, such that the vertex distance value decreases from the outside to the inside of the target gyral region. For example, the distance value of the vertex at the outermost edge of the target gyral region is 1, the distance value of the vertex closer to the center relative to the outermost vertex is 2, the distance value of the vertex even closer to the center is 3, and so on, until all vertices in the center of the target gyral region are assigned integer distance values.

[0130] Furthermore, the number of vertices adjacent to the target vertex and having distance values ​​greater than the distance value of the target vertex is divided by the number of all vertices adjacent to the target vertex to obtain a non-integer value of the target vertex, and the non-integer value of the target vertex is added to the difference between the integerized distance value of the target vertex and the integerized distance value of the outermost vertex to obtain a distance distribution value of the target vertex. Further, the distance distribution values ​​of all vertices in the target gyrus area are determined, thereby obtaining the distance distribution value of each vertex in each target area. Non-integerization of the integerized distance value of the target vertex can achieve stratification of the distances of each vertex in the target gyrus area, assign a distance distribution value of zero to isolated vertices, and facilitate deletion of isolated vertices.

[0131] In the above implementation process, the distance transformation is performed on the vertices in the target area to obtain the distance distribution value of each vertex in the target area. Furthermore, according to the distance distribution value of each vertex in the target area, the vertices in the vertex concentration of the target area are connected in descending order to obtain the sub-network corresponding to each target area, so that the sub-network corresponding to each target area is connected from the center to the edge to form a tree-like sub-network.

[0132] In some embodiments, based on the distance distribution values ​​of the vertices in each target area, the vertices in the vertex set of the target area are sequentially connected in descending order to obtain a subnetwork corresponding to each target area. The vertices in the target area can be connected in batches in descending order based on the distance distribution values ​​of the vertices in each target area to obtain a subnetwork corresponding to each target area. Specifically, the following steps may be included:

[0133] Step 1: Based on the distance distribution values ​​of each vertex in each target area, determine the current batch's travel threshold and the current batch's vertices whose distance distribution values ​​are greater than or equal to the current batch's travel threshold.

[0134] Step 2: Connect the vertices of the current batch in descending order according to the distance distribution value until the travel threshold of the current batch is zero, and obtain the sub-network corresponding to each target area.

[0135] For example, the vertices in each target region can be connected in batches, and the vertices in the current batch whose distance distribution values ​​are greater than or equal to the current batch travel threshold are selected. Furthermore, the vertices in the current batch are connected in descending order of distance distribution values ​​until the current batch travel threshold reaches zero, thereby obtaining a subnetwork corresponding to each target region.

[0136] When connecting vertices in each target area using the marching tree, in order to control the marching speed, the vertices in the target area can be connected in batches. Specifically, the marching threshold of the current batch can be determined from the distance distribution values ​​of the vertices in the target area in descending order, and the vertices in the target area with distance values ​​greater than or equal to the marching threshold of the current batch are determined as the vertices of the current batch. Furthermore, the vertices of the current batch are connected sequentially according to the distance distribution values ​​in descending order.

[0137] For example, the distance distribution values ​​of a target area are from large to small: 2.0, 1.833, 1.667, 1.5, 1.333, ..., 0.1667, 0. The travel threshold of the first batch of the target area can be determined as 1.833, then the current batch of vertices of the first batch includes vertices corresponding to 2.0 and 1.833. Furthermore, in the target area, starting from the vertex with a distance distribution value of 2, the breadth-first algorithm is used to traverse all vertices adjacent to the distance distribution value of 2, and the adjacent vertices with a distance distribution value of 1.833 are determined, and the vertex with a distance distribution value of 2 is connected to the determined adjacent vertices with a distance distribution value of 1.833, thereby completing the vertex connection of the first batch.

[0138] Furthermore, the travel thresholds of the second batch can be determined from the distance distribution values ​​of each vertex in the target area in descending order, that is, in the distance distribution values ​​of the target area, the value with a distance distribution value less than 1.833 can be determined as the travel threshold of the second batch. For example, 1.667 can be determined as the travel threshold of the second batch. Then, in the target area, the breadth-first algorithm is used to traverse the second batch of vertices with distance distribution values ​​greater than or equal to 1.667, including 2.0, 1.833 and 1.667. Further, the vertices corresponding to 2.0, 1.833 and 1.667 can be connected in sequence from large to small. However, since the vertices corresponding to 2.0 and 1.833 have been connected in the first batch, the vertices corresponding to 1.833 and 1.667 only need to be connected in descending order in the second batch, thereby completing the connection of the vertices of the second batch.

[0139] Furthermore, the travel thresholds of the third batch can be determined from the distance distribution values ​​of the vertices in the target area in descending order, that is, in the distance distribution values ​​of the target area, the values ​​with distance distribution values ​​less than 1.667 can be determined as the travel thresholds of the third batch. For example, 1.5 can be determined as the travel threshold of the third batch. Then, in the target area, the breadth-first algorithm is used to traverse the third batch of vertices with distance distribution values ​​greater than or equal to 1.5, including 2.0, 1.833, 1.667 and 1.5. Further, the vertices corresponding to 2.0, 1.833, 1.667 and 1.5 can be connected in descending order. However, since the connections of the vertices corresponding to 2.0 to 1.667 have been completed in the first and second batches, the vertices corresponding to 1.667 and 1.5 only need to be connected in descending order in the third batch, thereby completing the connection of the vertices in the third batch. This process is repeated until the progress threshold of the current batch is zero, and the connection of the vertices in the target area is finally completed, forming a subnetwork corresponding to the target area.

[0140] It should be noted that in the implementation of this application, the distance distribution value of the target area is selected continuously in order from large to small as the travel threshold of the current batch for illustration. In actual applications, the distance distribution value of the target area can also be selected at intervals in order from large to small as the travel threshold of the current batch. The number of intervals can be the same or different, and there is no restriction here.

[0141] As another embodiment, in order to improve the flexibility of setting the marching threshold, the marching threshold can be determined according to prior knowledge. Specifically, the distance distribution values ​​of the vertices of the target area are sorted in descending order to obtain: 2.0, 1.833, 1.667, 1.5, 1.333, ..., 0.1667, 0. Further, starting from the second maximum value of the distance distribution value (1.833), the values ​​are used in descending order as the marching thresholds of the current batch, and the vertex connection of the current batch is performed, that is, the vertex with the larger distance value is used as the root, and the breadth-first search algorithm is used to traverse and connect the adjacent current batch vertices to form a marching subtree. After completing the vertex connection of the current batch, the number of marching subtrees currently generated by the target object is counted. When the number of generated marching subtrees is the same as the number of target areas determined by prior knowledge, the marching threshold of the next batch can be directly set to zero, that is, the connection of the marching subtrees is terminated.

[0142] It should be noted that the number of target areas determined by prior knowledge may refer to the number of hub areas estimated from the regression curve of the number of hub areas changing with age, and there is no limitation here.

[0143] In the above implementation process, the travel threshold of the current batch and the vertices of the current batch are determined according to the distance distribution value of the target area. Furthermore, the vertices of the current batch are connected in order from large to small according to the distance distribution value. This not only realizes the batch connection of the vertices in the target area, but also can adaptively control the speed of vertex connection according to the distance distribution value of the target area. Therefore, there is no need to control the sub-network connection speed by setting a fixed threshold, which improves the flexibility and applicability of the sub-network generation algorithm.

[0144] In some embodiments, the target region is a gyrus region of the target object, and the state of the target leaf node is determined based on whether the target leaf node is adjacent to other leaf nodes in the target subnetwork, which may include the following situations:

[0145] Case 1: If the target leaf node is not adjacent to any other leaf node, the state of the target leaf node is to be deleted.

[0146] For example, if the target leaf node is not adjacent to any leaf node in the target subnetwork and other subnetworks, it indicates that the target leaf node is an isolated node and will not be connected to any other subnetwork during the subnetwork connection process. Therefore, the state of the target leaf node is determined to be a state to be deleted.

[0147] Case 2: If the target leaf node is only adjacent to leaf nodes in other subnetworks, the state of the target leaf node is to be connected, and the other subnetwork is any subnetwork in the multiple subnetworks except the target subnetwork.

[0148] For example, if the target leaf node is only adjacent to leaf nodes in other subnetworks, then during the subnetwork connection process, the connection between the target subnetwork and other subnetworks is through the target leaf node. Therefore, the state of the target leaf node is determined to be a pending connection state.

[0149] Case three: If the target leaf node is only adjacent to other target leaf nodes in the target sub-network, then determine whether the other target leaf nodes adjacent to the target leaf node have adjacent leaf nodes in other sub-networks. If the other target leaf nodes adjacent to the target leaf node do not have adjacent leaf nodes in other sub-networks, then the state of the target leaf node and the leaf node corresponding to the maximum value of the cerebral cortical morphological feature in the other target leaf nodes adjacent to the target leaf node is determined to be in a to-be-connected state. Moreover, the states of the remaining leaf nodes in the target leaf node and the other target leaf nodes adjacent to the target leaf node, except for the leaf node corresponding to the maximum value of the cerebral cortical morphological feature, are all in a to-be-deleted state. If the other target leaf nodes adjacent to the target leaf node have adjacent leaf nodes in other sub-networks, the state of the target leaf node is in a to-be-deleted state.

[0150] For example, if the target leaf node is only adjacent to other target leaf nodes in the target subnetwork, then during the subnetwork connection process, the connection between the target subnetwork and the other subnetworks may be through the target leaf node, or other target leaf nodes adjacent to the target leaf node are connected to leaf nodes in other subnetworks. Therefore, the leaf node that executes the connection between the target subnetwork and the other subnetworks can be determined among the target leaf node and other target leaf nodes adjacent to the target leaf node. Specifically, it can be further determined whether the other target leaf nodes adjacent to the target leaf node have adjacent leaf nodes in other subnetworks. If the other target leaf nodes adjacent to the target leaf node do not have adjacent leaf nodes in other subnetworks, then the target leaf node The target leaf node and other target leaf nodes adjacent to the target leaf node are all candidate leaf nodes, and the target area is the gyrus area of ​​the target object. What needs to be obtained in the end is the gyrus morphological network. Then, when the sub-networks are connected, the nodes located on the gyrus in the brain of the target object are connected. In order to make the connected leaf nodes accurately located on the gyrus, the states of the leaf nodes corresponding to the maximum values ​​of the morphological features of the cerebral cortex in the target leaf node and other target leaf nodes adjacent to the target leaf node are determined to be in the to-be-connected state, and the states of the remaining leaf nodes except the leaf nodes corresponding to the maximum values ​​of the morphological features of the cerebral cortex in the target leaf node and other target leaf nodes adjacent to the target leaf node are all in the to-be-deleted state.

[0151] If the target leaf node is only adjacent to other target leaf nodes in the target subnetwork, and other target leaf nodes adjacent to the target leaf node have adjacent leaf nodes in other subnetworks, then during the subnetwork connection process, the connection between the target subnetwork and other subnetworks is through the connection between other target leaf nodes adjacent to the target leaf node and leaf nodes in other subnetworks. Therefore, the status of the target leaf node is to be deleted.

[0152] Case 4: If the target leaf node is adjacent to other target leaf nodes, and the target leaf node is adjacent to leaf nodes in other sub-networks, then determine whether the other target leaf nodes adjacent to the target leaf node have adjacent leaf nodes in other sub-networks. If the other target leaf nodes adjacent to the target leaf node do not have adjacent leaf nodes in other sub-networks, then the state of the target leaf node is a to-be-connected state. If the other target leaf nodes adjacent to the target leaf node have adjacent leaf nodes in other sub-networks, then the state of the leaf node corresponding to the maximum value of the cerebral cortical morphological feature in the target leaf node and other target leaf nodes adjacent to the target leaf node is determined to be a to-be-connected state. In addition, the state of the remaining leaf nodes in the target leaf node and other target leaf nodes adjacent to the target leaf node, except for the leaf node corresponding to the maximum value of the cerebral cortical morphological feature, are all in a to-be-deleted state.

[0153] Exemplarily, if the target leaf node is adjacent to other target leaf nodes, and the target leaf node is also adjacent to leaf nodes in other subnetworks, then during the subnetwork connection process, the connection between the target subnetwork and the other subnetworks may be through the target leaf node, or other target leaf nodes adjacent to the target leaf node are connected to leaf nodes in other subnetworks. Therefore, the leaf node that executes the connection between the target subnetwork and the other subnetworks can be determined among the target leaf node and other target leaf nodes adjacent to the target leaf node. Specifically, it can be further determined whether other target leaf nodes adjacent to the target leaf node have adjacent leaf nodes in other subnetworks. If other target leaf nodes adjacent to the target leaf node do not have adjacent leaf nodes in other subnetworks, then during the subnetwork connection process, the connection between the target subnetwork and the other subnetworks is through the target leaf node and leaf nodes in other adjacent subnetworks. Therefore, the state of the target leaf node is a state to be connected.

[0154] If the target leaf node is adjacent to other target leaf nodes, and the target leaf node is also adjacent to leaf nodes in other sub-networks, and other target leaf nodes adjacent to the target leaf node have adjacent leaf nodes in other sub-networks, then during the sub-network connection process, the connection between the target sub-network and the other sub-networks may be through the target leaf node, or other target leaf nodes adjacent to the target leaf node are connected to leaf nodes in other sub-networks. Therefore, the leaf node that executes the connection between the target sub-network and the other sub-networks can be determined among the target leaf node and other target leaf nodes adjacent to the target leaf node. In order to make the connected leaf nodes accurately located on the gyrus, the state of the leaf node corresponding to the maximum value of the cerebral cortical morphological characteristics among the target leaf node and other target leaf nodes adjacent to the target leaf node is determined as a to-be-connected state, and the state of the remaining leaf nodes except the leaf node corresponding to the maximum value of the cerebral cortical morphological characteristics among the target leaf node and other target leaf nodes adjacent to the target leaf node are all to-be-deleted states.

[0155] In the above implementation process, in order to accurately determine the state of the target leaf node, based on the adjacent relationship between the target leaf node and the target sub-network and the leaf nodes in other sub-networks, the position relationship of the target leaf node in different situations is further analyzed, so as to accurately obtain the state of the target leaf node.

[0156] In some embodiments, determining the connected leaf node corresponding to the target leaf node may include the following situations:

[0157] Case 1: If the target leaf node is only adjacent to leaf nodes in other sub-networks, the connected leaf node is the leaf node corresponding to the maximum value of the cerebral cortical morphological feature among all leaf nodes adjacent to the target leaf node.

[0158] For example, if the target leaf node is only adjacent to leaf nodes in other sub-networks, in order to accurately locate the connected leaf node on the gyrus, the connected leaf node is the leaf node corresponding to the maximum value of the morphological feature of the cerebral cortex among all leaf nodes adjacent to the target leaf node.

[0159] Case 2: If the target leaf node is only adjacent to other target leaf nodes in the target sub-network, and there are no adjacent leaf nodes in other sub-networks adjacent to the target leaf node, then the connected leaf node is the leaf node corresponding to the maximum value of the cerebral cortical morphological characteristics in other sub-networks closest to the target sub-network.

[0160] For example, if the target leaf node is only adjacent to other target leaf nodes in the target sub-network, and there are no adjacent leaf nodes in other sub-networks to the other target leaf nodes adjacent to the target leaf node, in order to connect the sub-networks, the leaf node corresponding to the maximum value of the cerebral cortical morphological characteristics in other sub-networks closest to the target sub-network can be determined as the connected leaf node.

[0161] Case 3: If the target leaf node is adjacent to other target leaf nodes, and the target leaf node is adjacent to leaf nodes in other sub-networks, then the connected leaf node is the leaf node corresponding to the maximum value of the cerebral cortical morphological feature in other sub-networks adjacent to the target leaf node.

[0162] For example, if the target leaf node is adjacent to other target leaf nodes, and the target leaf node is adjacent to leaf nodes in other subnetworks, in order to achieve connection between the subnetworks and avoid the appearance of small loops within a subnetwork, the leaf node corresponding to the maximum value of the cerebral cortical morphological characteristics in other subnetworks adjacent to the target leaf node can be determined as the connected leaf node corresponding to the target subnetwork.

[0163] In the above implementation process, in order to realize the connection of each sub-network in the gyrus area and avoid the appearance of small loops within the sub-network, in other sub-networks, the leaf node corresponding to the maximum value of the morphological feature of the cerebral cortex among the leaf nodes adjacent to the target sub-network to be connected is determined as the connection leaf node, thereby ensuring that the leaf node performing the connection is located at the gyrus position, thereby facilitating the acquisition of the gyrus morphological network.

[0164] It should be noted that the implementation of this application and the generation of a gyrus morphology network are used as examples for illustration. In actual applications, the above method can also be used to generate a sulcus morphology map, and there is no limitation here. If the target area is a sulcus area, the maximum value of the cerebral cortical morphological feature can be replaced by the minimum value of the cerebral cortical morphological feature in the process of determining the state of the target leaf node and connecting the leaf nodes. Moreover, the cerebral cortical morphological feature in the process of determining the state of the target leaf node and connecting the leaf nodes may be consistent with or inconsistent with the cerebral cortical morphological feature for determining the water level parameters, and there is no limitation here.

[0165] In one embodiment, before connecting the leaf nodes in the waiting-to-connect state in each sub-network to the corresponding connecting leaf nodes, the method further includes:

[0166] Determine whether the two sub-networks where the leaf node in the waiting-to-connect state and the corresponding connecting leaf node are located are connected. If they are connected, delete the target leaf node and perform status determination on the next target leaf node.

[0167] For example, in order to avoid forming a short loop, when deciding to connect two leaf nodes, it is necessary to first determine whether the subnetworks where the two leaf nodes are located are already connected. If so, the target leaf node is set to the deleted state and the state of the next target leaf node is determined.

[0168] In the above implementation process, before two sub-networks are connected, the sub-network connection status is determined, so that other leaf nodes can be deleted when the two sub-networks are in a connected state, thereby avoiding the generation of small loops.

[0169] Figure 3 This is a flow chart of an embodiment of a method for generating a cerebral cortical morphology network provided in an embodiment of the present application, such as Figure 3 As shown, the flow chart includes the following steps:

[0170] Step S301: Acquire structural magnetic resonance data of a target object.

[0171] Specifically, a structural magnetic resonance imaging (sMRI) dataset from the Human Connectome Project (HCP) was collected, which contains data of approximately 1,200 healthy adults, with each healthy adult being a target subject.

[0172] Step S302 : performing cerebral cortex reconstruction on the structural magnetic resonance data of the target object to obtain a cerebral cortex reconstructed image of the target object and cerebral cortex morphological features.

[0173] Furthermore, the FreeSurfer toolset was used to perform gray-white matter segmentation and cerebral cortex reconstruction on the structural image of each target object, obtaining the cerebral cortex reconstruction image of each target object and its corresponding cerebral cortical morphological characteristic indicators, such as sulcus depth, cortical thickness and cortical curvature.

[0174] Step S303: determining the rising water parameter of the watershed algorithm based on the morphological characteristics of the cerebral cortex.

[0175] Furthermore, taking the cortical curvature as a morphological feature as an example, the minimum value of the cortical curvature of each target object is used as the initial value of the water rise of the corresponding target object, and the maximum value of the cortical curvature is used as the end value of the water rise of the corresponding target object, and the water rise parameters of the watershed algorithm relative to the target object are obtained.

[0176] As another example, the characteristic values ​​of at least two types of cerebral cortical morphological features, including sulcus depth, cortical thickness, cortical curvature, volume, and surface area, can be used to determine the initial and final values ​​of the water level rise through relevant calculation methods to obtain the water level rise parameters of the watershed algorithm relative to the target object.

[0177] Step S304 : segmenting the reconstructed cerebral cortex image according to the water level parameter to obtain a set of candidate vertices of the gyrus region of the target object.

[0178] Furthermore, according to the flooding parameter of each target object, the watershed algorithm is used to segment the corresponding cortical reconstructed image to obtain the target object in multiple gyrus regions and the candidate vertex set of each gyrus region.

[0179] Specifically, when using the watershed algorithm to segment the corresponding cerebral cortex reconstructed image, the amplitude of each water rise can be determined according to the percentage of the difference between the maximum and minimum values ​​of the cortical curvature. For example, each time the amplitude is 5% of the difference between the maximum and minimum values ​​of the cortical curvature. Then the entire water rise process will start from the minimum value of the cortical curvature, and will stop when it reaches the maximum value of the cortical curvature after 20 times.

[0180] Step S305 : Screen the candidate vertex set of the gyrus region according to the sulcus depth of the target object to obtain the vertex set of the gyrus region.

[0181] Furthermore, according to the sulcus depth of the target object, the corresponding gyrus region candidate vertex set is screened, thereby deleting the vertices located in the sulcus region to obtain the vertex set of the gyrus region.

[0182] Specifically, a sulcus depth value greater than 0 can be selected as a threshold, that is, among all candidate gyrus vertices, vertices whose sulcus depth values ​​are greater than 0 are retained, and vertices whose sulcus depths are less than or equal to 0 are deleted from the candidate gyrus vertices, thereby excluding vertices whose candidate vertices are concentrated in the sulcus area.

[0183] As another example, in order to retain as many candidate gyrus vertices as possible, the gyrus vertices are redundantly processed. For example, 10% of the negative maximum value of the sulcus depth is used as a threshold. Candidate vertices below this threshold are excluded, and the others are retained. For example, if the range of sulcus depth is [-1.5, 1.5], the sulcus depth threshold is -1.5*10%=-0.15. If an outlier occurs, for example, if the range of sulcus depth is [-9.0, 1.5], the sulcus depth threshold is -0.9 with a 10% constraint, which is equivalent to 60% of the maximum negative value of the sulcus depth under normal circumstances. This will cause many vertices located in the sulcus area to be mistaken for candidate gyrus ridge vertices, resulting in a large error.

[0184] Figure 4 is a vertex diagram of a gyrus region provided in an embodiment of the present application, such as Figure 4 As shown in the figure, the points distributed in each gyrus area are the vertices of the gyrus area.

[0185] Step S306 , performing distance transformation on the vertices in the vertex set of the gyrus region to obtain a distance distribution value of the gyrus region.

[0186] Furthermore, distance transformation is performed on the vertices in each gyrus region to obtain a distance distribution value of the gyrus region, so that the distance distribution value of the vertices in each gyrus region gradually increases from the outside to the inside.

[0187] Step S307 : Based on the distance distribution values, a breadth-first algorithm is used to connect the vertices in the multiple gyrus regions to obtain a sub-network corresponding to each gyrus region.

[0188] Specifically, according to the distance distribution value, starting from the vertex with the largest distance distribution value, a breadth-first search method is used to move toward the adjacent vertex with a smaller distance distribution value to generate subtrees. Each subtree corresponds to a subnetwork, thereby obtaining a subnetwork for each gyrus area.

[0189] For example, to control the speed of tree progression, the distance value in the distance distribution is used as the progression threshold for each tree progression. This means that the tree only progresses to a vertex greater than the progression threshold, stopping each time. Progress is then made to each vertex, connecting the vertices of each gyrus region into a corresponding subnetwork. For example, if the distances of vertices in a gyrus region are 2.0, 1.833, 1.667, 1.5, 1.333, ..., 0.1667, and 0, the progression thresholds can be set to 1.833, 1.667, ..., 0, respectively. To limit the number of generated progression trees, a threshold can be selected every three distance values, such as 1.833, 1.333, ..., 0, though this is not a limitation. Figure 5 This is a schematic diagram of a gyrus region subnetwork provided by an embodiment of the present application, such as Figure 5 As shown, the vertices in each gyrus area are connected to form a subnetwork similar to a tree structure.

[0190] Step S308: Determine the state of the target leaf node based on whether the target leaf node is adjacent to other leaf nodes in the target sub-network, and connect the sub-networks to obtain a gyrus morphology network.

[0191] Among them, the target subnetwork is any subnetwork among the multiple subnetworks, the target leaf node is any leaf node in the target subnetwork, and the other leaf nodes are any leaf nodes in the multiple subnetworks except the target leaf node. The status of the target leaf node includes a waiting-to-connect status or a waiting-to-delete status.

[0192] Furthermore, it is determined whether the target leaf node is adjacent to other leaf nodes, thereby determining the state of the target leaf node. Specifically, the adjacent relationship between the target leaf node and other leaf nodes may include the following situations:

[0193] Case 1: If the target leaf node is not adjacent to any other leaf node, the state of the target leaf node is to be deleted. Before the sub-network connection is made, the target leaf node is deleted.

[0194] Case 2: If the target leaf node is only adjacent to the leaf nodes in other sub-networks, the state of the target leaf node is to be connected, and among all the leaf nodes adjacent to the target leaf node, the leaf node corresponding to the maximum value of the cerebral cortical morphological features is determined as the connecting leaf node corresponding to the target leaf node, and the target leaf node is connected to the connecting leaf node to obtain a gyrus morphological network.

[0195] Case 3: If the target leaf node is only adjacent to other target leaf nodes in the target sub-network, determine whether the other target leaf nodes adjacent to the target leaf node have adjacent leaf nodes in other sub-networks.

[0196] If the target leaf node is only adjacent to other target leaf nodes in the target sub-network, and the other target leaf nodes adjacent to the target leaf node do not have adjacent leaf nodes in other sub-networks, then the target leaf node and the leaf nodes corresponding to the maximum value of the cerebral cortical morphological features in the other target leaf nodes adjacent to the target leaf node are determined to be in a to-be-connected state, and the states of the remaining leaf nodes except the leaf nodes corresponding to the maximum value of the cerebral cortical morphological features in the target leaf node and the other target leaf nodes adjacent to the target leaf node are all in a to-be-deleted state, and then the connecting leaf node corresponding to the target leaf node in the determined to-be-connected state can be the leaf node corresponding to the maximum value of the cerebral cortical morphological features in the other sub-networks closest to the target sub-network. Further, the target leaf node is connected to the connecting leaf node to obtain a gyrus morphological network.

[0197] If the target leaf node is adjacent only to other target leaf nodes in the target sub-network, and the other target leaf nodes adjacent to the target leaf node have adjacent leaf nodes in other sub-networks, the state of the target leaf node is to be deleted.

[0198] Case 4: If the target leaf node is adjacent to other target leaf nodes, and the target leaf node is adjacent to leaf nodes in other sub-networks, then determine whether the other target leaf nodes adjacent to the target leaf node have adjacent leaf nodes in other sub-networks.

[0199] If the target leaf node is adjacent to other target leaf nodes, the target leaf node is adjacent to leaf nodes in other sub-networks, and the target leaf nodes adjacent to the target leaf node do not have adjacent leaf nodes in other sub-networks, then the target leaf node is in a pending connection state, and the connecting leaf node corresponding to the target leaf node is the leaf node corresponding to the maximum value of the cerebral cortical morphological feature in the other sub-networks adjacent to the target leaf node. Further, the target leaf node is connected to the connecting leaf node to obtain a gyrus morphological network.

[0200] If the target leaf node is adjacent to other target leaf nodes, the target leaf node is adjacent to leaf nodes in other sub-networks, and other target leaf nodes adjacent to the target leaf node have adjacent leaf nodes in other sub-networks, then the target leaf node and the target leaf node, and the state of the leaf node corresponding to the maximum value of the cerebral cortical morphological feature in other target leaf nodes adjacent to the target leaf node, are determined to be in a to-be-connected state, and the target leaf node and the target leaf node, except for the leaf node corresponding to the maximum value of the cerebral cortical morphological feature in other target leaf nodes adjacent to the target leaf node, the states of the remaining leaf nodes are all in a to-be-deleted state. Furthermore, the connecting leaf node corresponding to the target leaf node in the to-be-connected state is the leaf node corresponding to the maximum value of the cerebral cortical morphological feature in other sub-networks adjacent to the target leaf node. Further, the target leaf node is connected to the connecting leaf node to obtain a gyrus morphological network.

[0201] In order to avoid forming a short loop, when deciding to connect two leaf nodes, it is necessary to first determine whether the subnetworks where the two leaf nodes are located are already connected. If so, the target leaf node is set to the deleted state and the status of the next target leaf node is determined.

[0202] Figure 6 This is a schematic diagram of a gyrus morphology network provided in an embodiment of the present application, such as Figure 6 The black lines shown form the gyral morphological network.

[0203] It should be noted that, although the various steps in the flowcharts involved in the various embodiments described above are displayed in sequence as indicated by the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and these steps can be executed in other orders. Moreover, at least a portion of the steps in the flowcharts involved in the various embodiments described above can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a portion of steps or stages in other steps.

[0204] In this embodiment, a device for generating a cerebral cortical morphology network is also provided, which is used to implement the above-mentioned embodiments and preferred embodiments, and the details that have been described will not be repeated. The terms "module", "unit", "sub-unit", etc. used below can implement a combination of software and / or hardware for a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceived.

[0205] Figure 7 This is a structural block diagram of a cerebral cortical morphology network generation device provided in an example of the present application. Figure 7 As shown, the device includes:

[0206] An acquisition module 701 is used to acquire a reconstructed image of the cerebral cortex of a target object;

[0207] a segmentation module 702 for segmenting the cerebral cortex reconstructed image to obtain multiple target regions of the cerebral cortex and a vertex set of each target region;

[0208] A connection module 703 is used to connect the vertices in the vertex set of each target area to obtain a sub-network corresponding to each target area;

[0209] A first determining module 704 is configured to determine a state of a target leaf node in a target subnetwork based on whether the target leaf node has adjacent leaf nodes in other subnetworks, where the target subnetwork is any subnetwork among the multiple subnetworks, the target leaf node is any leaf node in the target subnetwork, and the other subnetwork is any subnetwork other than the target subnetwork. The state of the target leaf node includes a pending connection state or a pending deletion state.

[0210] The second determining module 705 is configured to determine a connecting leaf node corresponding to the target leaf node if the target leaf node is in a waiting-to-connect state;

[0211] The generation module 706 is used to connect the sub-networks based on the status of each leaf node in each sub-network to generate a cerebral cortical morphological network of the target object.

[0212] In some embodiments, the segmentation module 702 is specifically configured to:

[0213] Acquire the morphological characteristics of the cerebral cortex of the target object;

[0214] The reconstructed cerebral cortex image is segmented based on the morphological features of the cerebral cortex to obtain multiple target areas of the cerebral cortex and a vertex set of each target area.

[0215] In some embodiments, the segmentation module 702 is specifically configured to:

[0216] Determine the rising water parameter of the watershed algorithm based on the morphological characteristics of the cerebral cortex;

[0217] Based on the flooding parameter, the reconstructed image of the cerebral cortex is segmented to obtain multiple target areas of the cerebral cortex and a vertex set of each target area.

[0218] In some embodiments, the cerebral cortical morphological features include N types, where N is a positive integer greater than or equal to 2, and the segmentation module 702 is specifically configured to:

[0219] A flooding parameter of a watershed algorithm is determined based on feature values ​​of at least two types of cerebral cortex morphological features of the target object.

[0220] In some embodiments, the morphological features of the cerebral cortex include at least the depth of the sulci, and the segmentation module 702 is specifically configured to:

[0221] Segment the reconstructed cerebral cortex image based on the flooding parameter to obtain multiple target regions of the cerebral cortex and a set of candidate vertices for each target region;

[0222] Based on the characteristic value of the sulcus depth of the target object, the candidate vertex set of each target area is screened to obtain the vertex set of each target area.

[0223] In some embodiments, the connection module 703 is specifically configured to:

[0224] Perform distance transformation on the vertices in the vertex set of each target area to obtain the distance distribution value of each vertex in each target area;

[0225] Based on the distance distribution values ​​of each vertex in each target area, the vertices in the vertex set of the target area are connected in order from large to small to obtain the sub-network corresponding to each target area.

[0226] In some embodiments, the connection module 703 is specifically configured to:

[0227] Determining a travel threshold for the current batch and vertices in the current batch whose distance distribution values ​​are greater than or equal to the travel threshold based on the distance distribution values ​​of the vertices in each target area;

[0228] The vertices of the current batch are connected in descending order according to the distance distribution value until the travel threshold of the current batch reaches zero, and the sub-network corresponding to each target area is obtained.

[0229] In some embodiments, the target region is a gyrus region of the target object, and the first determining module 704 is specifically configured to:

[0230] If the target leaf node is not adjacent to any other leaf node, the target leaf node is in the pending deletion state;

[0231] If the target leaf node is only adjacent to leaf nodes in other subnetworks, the state of the target leaf node is to be connected, and the other subnetworks are any subnetworks in the multiple subnetworks except the target subnetwork;

[0232] If the target leaf node is only adjacent to other target leaf nodes in the target sub-network, then determine whether the other target leaf nodes adjacent to the target leaf node have adjacent leaf nodes in other sub-networks. If the other target leaf nodes adjacent to the target leaf node do not have adjacent leaf nodes in other sub-networks, then the state of the target leaf node and the leaf node corresponding to the maximum value of the cerebral cortical morphological feature among the other target leaf nodes adjacent to the target leaf node is determined to be a to-be-connected state. Moreover, the states of the target leaf node and the other target leaf nodes adjacent to the target leaf node, except for the leaf node corresponding to the maximum value of the cerebral cortical morphological feature, are all to be deleted. If the other target leaf nodes adjacent to the target leaf node have adjacent leaf nodes in other sub-networks, the state of the target leaf node is to be deleted.

[0233] If the target leaf node is adjacent to other target leaf nodes, and the target leaf node is adjacent to leaf nodes in other sub-networks, then determine whether the other target leaf nodes adjacent to the target leaf node have adjacent leaf nodes in other sub-networks. If the other target leaf nodes adjacent to the target leaf node do not have adjacent leaf nodes in other sub-networks, then the state of the target leaf node is a to-be-connected state. If the other target leaf nodes adjacent to the target leaf node have adjacent leaf nodes in other sub-networks, then the state of the target leaf node and the leaf node corresponding to the maximum value of the cerebral cortical morphological feature among the other target leaf nodes adjacent to the target leaf node is determined to be a to-be-connected state. In addition, the state of the remaining leaf nodes except the leaf node corresponding to the maximum value of the cerebral cortical morphological feature among the target leaf node and the other target leaf nodes adjacent to the target leaf node is a to-be-deleted state.

[0234] In some embodiments, the second determining module 705 is specifically configured to:

[0235] If the target leaf node is only adjacent to leaf nodes in other sub-networks, the connected leaf node is the leaf node corresponding to the maximum value of the cerebral cortical morphological feature among all leaf nodes adjacent to the target leaf node;

[0236] If the target leaf node is only adjacent to other target leaf nodes in the target sub-network, and there are no adjacent leaf nodes in other sub-networks adjacent to the target leaf node, then the connected leaf node is the leaf node corresponding to the maximum value of the cerebral cortical morphological characteristics in the other sub-networks closest to the target sub-network.

[0237] If the target leaf node is adjacent to other target leaf nodes, and the target leaf node is adjacent to leaf nodes in other sub-networks, the connected leaf node is the leaf node corresponding to the maximum value of the cerebral cortical morphological feature in other sub-networks adjacent to the target leaf node.

[0238] It should be noted that the above modules can be functional modules or program modules, and can be implemented through software or hardware. For modules implemented through hardware, the above modules can be located in the same processor; or the above modules can be located in different processors in any combination.

[0239] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 8 shown. Figure 8 1 is a schematic diagram of the structure of a computer device provided in an embodiment of the present application, wherein the computer device includes a processor, a memory, and a network interface connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The database of the computer device is used to store data. The network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, a method for generating a cerebral cortical morphology network is implemented.

[0240] Those skilled in the art will understand that Figure 8 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0241] In one embodiment, an electronic device is further provided, including a memory and a processor. The memory stores a computer program, and the processor implements the steps in the above method embodiments when executing the computer program.

[0242] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0243] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0244] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, database or other media used in the embodiments provided in this application may include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory may include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in the various embodiments provided herein may include at least one of a relational database and a non-relational database. Non-relational databases may include, but are not limited to, distributed databases based on blockchains. The processor involved in the various embodiments provided herein may be, but are not limited to, a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic unit, a data processing logic unit based on quantum computing, and the like.

[0245] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0246] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of patent protection. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present application, and these modifications and improvements fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

Claims

1. A method for generating a cerebral cortical morphological network, characterized in that: include: Acquire a reconstructed image of the cerebral cortex of the target object; Segmenting the reconstructed cerebral cortex image to obtain a plurality of target regions of the cerebral cortex and a vertex set of each target region; Connecting the vertices in the vertex set of each target area to obtain a subnetwork corresponding to each target area; Determining a state of the target leaf node based on whether the target leaf node in the target subnetwork is adjacent to other leaf nodes, wherein the target subnetwork is any subnetwork among multiple subnetworks, the target leaf node is any leaf node in the target subnetwork, and the other leaf nodes are any leaf nodes in the multiple subnetworks other than the target leaf node, and the state of the target leaf node includes a pending connection state or a pending deletion state; If the state of the target leaf node is a waiting state, determining a connection leaf node corresponding to the target leaf node; The leaf nodes in the waiting-to-connect state in each sub-network are connected to the corresponding connected leaf nodes to generate the cerebral cortex morphological network of the target object.

2. The method for generating a cerebral cortical morphological network according to claim 1, wherein: The segmentation of the cerebral cortex reconstructed image to obtain multiple target areas of the cerebral cortex and a vertex set of each target area includes: Acquiring morphological characteristics of the cerebral cortex of the target object; The cerebral cortex reconstructed image is segmented based on the cerebral cortex morphological features to obtain multiple target areas of the cerebral cortex and a vertex set of each target area.

3. The method for generating a cerebral cortical morphological network according to claim 2, wherein: The segmentation of the cerebral cortex reconstructed image based on the cerebral cortex morphological features to obtain multiple target areas of the cerebral cortex and a vertex set of each target area includes: Determining a flooding parameter of a watershed algorithm based on the morphological characteristics of the cerebral cortex; The reconstructed image of the cerebral cortex is segmented based on the flooding parameter to obtain a plurality of target regions of the cerebral cortex and a vertex set of each target region.

4. The method for generating a cerebral cortical morphological network according to claim 3, wherein: The cerebral cortex morphological features include N types, where N is a positive integer greater than or equal to 2. The method of determining the watershed parameter of the watershed algorithm based on the cerebral cortex morphological features includes: A flooding parameter of a watershed algorithm is determined based on the feature values ​​of at least two types of cerebral cortex morphological features of the target object.

5. The method for generating a cerebral cortical morphological network according to claim 3, wherein: The morphological features of the cerebral cortex include at least the depth of the sulcus. The segmentation of the cerebral cortex reconstructed image based on the flooding parameter to obtain multiple target areas of the cerebral cortex and a vertex set of each target area includes: Segmenting the reconstructed cerebral cortex image based on the flooding parameter to obtain a plurality of target regions of the cerebral cortex and a candidate vertex set for each target region; Based on the characteristic value of the sulcus depth of the target object, the candidate vertex set of each target region is screened to obtain the vertex set of each target region.

6. The method for generating a cerebral cortical morphological network according to claim 1, wherein: Connecting the vertices in the vertex set of each target area to obtain a subnetwork corresponding to each target area includes: Performing distance transformation on vertices in the vertex set of each target area to obtain a distance distribution value of each vertex in each target area; Based on the distance distribution values ​​of the vertices in each target area, the vertices in the vertex set of the target area are connected in descending order to obtain a sub-network corresponding to each target area.

7. The method for generating a cerebral cortical morphological network according to claim 6, characterized in that: The method of connecting vertices in the vertex set of the target area in descending order based on the distance distribution values ​​of the vertices in each target area to obtain a subnetwork corresponding to each target area includes: Determining a travel threshold of a current batch and vertices of the current batch whose distance distribution values ​​are greater than or equal to the travel threshold based on the distance distribution values ​​of each vertex in each target area; The vertices of the current batch are connected in descending order according to the distance distribution value until the travel threshold of the current batch reaches zero, thereby obtaining a sub-network corresponding to each target area.

8. The method for generating a cerebral cortical morphological network according to claim 1, wherein: The target region is a gyrus region of the target object, and determining the state of the target leaf node based on whether the target leaf node in the target subnetwork is adjacent to other leaf nodes includes: If the target leaf node is not adjacent to any other leaf node, the state of the target leaf node is to be deleted; If the target leaf node is adjacent only to leaf nodes in other subnetworks, the state of the target leaf node is a pending connection state, and the other subnetworks are any subnetworks in the plurality of subnetworks except the target subnetwork; If the target leaf node is only adjacent to other target leaf nodes in the target sub-network, then determine whether the other target leaf nodes adjacent to the target leaf node have adjacent leaf nodes in other sub-networks; if the other target leaf nodes adjacent to the target leaf node do not have adjacent leaf nodes in other sub-networks, then the state of the target leaf node and the leaf node corresponding to the maximum value of the cerebral cortical morphological feature in the other target leaf nodes adjacent to the target leaf node is determined to be a to-be-connected state; and, the states of the target leaf node and the other target leaf nodes adjacent to the target leaf node, except for the leaf node corresponding to the maximum value of the cerebral cortical morphological feature, are all to be deleted; if the other target leaf nodes adjacent to the target leaf node have adjacent leaf nodes in other sub-networks, then the state of the target leaf node is to be deleted; If the target leaf node is adjacent to other target leaf nodes, and the target leaf node is adjacent to leaf nodes in other subnetworks, then determine whether the other target leaf nodes adjacent to the target leaf node have adjacent leaf nodes in other subnetworks; if the other target leaf nodes adjacent to the target leaf node do not have adjacent leaf nodes in other subnetworks, then the state of the target leaf node is a to-be-connected state; if the other target leaf nodes adjacent to the target leaf node have adjacent leaf nodes in other subnetworks, then the state of the leaf node corresponding to the maximum value of the cerebral cortical morphological feature in the target leaf node and other target leaf nodes adjacent to the target leaf node is determined to be a to-be-connected state; and the state of the remaining leaf nodes in the target leaf node and other target leaf nodes adjacent to the target leaf node, except for the leaf node corresponding to the maximum value of the cerebral cortical morphological feature, are all in a to-be-deleted state.

9. The method for generating a cerebral cortical morphological network according to claim 8, characterized in that: The determining of a connection leaf node corresponding to the target leaf node includes: If the target leaf node is adjacent only to leaf nodes in other sub-networks, the connected leaf node is the leaf node corresponding to the maximum value of the cerebral cortical morphological feature among all leaf nodes adjacent to the target leaf node; If the target leaf node is adjacent only to other target leaf nodes in the target subnetwork, and other target leaf nodes adjacent to the target leaf node do not have adjacent leaf nodes in other subnetworks, then the connected leaf node is the leaf node corresponding to the maximum value of the cerebral cortical morphological feature in the other subnetwork closest to the target subnetwork; If the target leaf node is adjacent to other target leaf nodes, and the target leaf node is adjacent to leaf nodes in other sub-networks, then the connected leaf node is the leaf node corresponding to the maximum value of the cerebral cortical morphological feature in other sub-networks adjacent to the target leaf node.

10. A device for generating a cerebral cortical morphological network, characterized in that: include: An acquisition module, used for acquiring a reconstructed image of the cerebral cortex of a target object; a segmentation module, configured to segment the cerebral cortex reconstructed image to obtain a plurality of target regions of the cerebral cortex and a vertex set of each target region; a connection module, configured to connect vertices in the vertex set of each target area to obtain a subnetwork corresponding to each target area; A first determination module is configured to determine a state of a target leaf node based on whether the target leaf node is adjacent to other leaf nodes in a target subnetwork, wherein the target subnetwork is any subnetwork among multiple subnetworks, the target leaf node is any leaf node in the target subnetwork, and the other leaf nodes are any leaf nodes other than the target leaf node in the multiple subnetworks, and the state of the target leaf node includes a pending connection state or a pending deletion state; A second determining module, if the state of the target leaf node is a waiting state, is used to determine a connecting leaf node corresponding to the target leaf node; A generation module is used to connect the leaf nodes in the waiting-to-connect state in each sub-network with the corresponding connection leaf nodes to generate the cerebral cortex morphological network of the target object.

11. An electronic device comprising a memory and a processor, characterized in that: A computer program is stored in the memory, and the processor is configured to run the computer program to execute the method for generating a cerebral cortical morphological network according to any one of claims 1 to 9.

12. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for generating a cerebral cortical morphological network according to any one of claims 1 to 9 are implemented.

Images