WO2023093470A1

WO2023093470A1 - Brain map drawing method and system, and related device

Info

Publication number: WO2023093470A1
Application number: PCT/CN2022/128965
Authority: WO
Inventors: 张长征; 朱森华; 涂丹丹
Original assignee: 华为云计算技术有限公司
Priority date: 2021-11-27
Filing date: 2022-11-01
Publication date: 2023-06-01
Also published as: CN116188669A

Abstract

A brain map drawing method and system, and a related device. The method may comprise the following steps: acquiring a first brain map; providing a user with a plurality of nerve fiber segments where key points in the first brain map are located, and acquiring correction information input by the user, wherein the key points comprise intersections of the plurality of nerve fiber segments in the first brain map or bifurcations of a single nerve fiber segment; and correcting the connection relationship among the plurality of nerve fiber segments according to the correction information, so as to output a second brain map. By means of the method, when a user manually corrects neural fibers in a brain map, the user can correct the connection relationship between the neural fibers near key points, without the need for distinguishing twisted, crisscrossed and intertwined nerve fibers, such that the efficiency of manually correcting the connection relationship between neural fibers in a brain map can be improved, thereby improving the efficiency of drawing a brain map.

Description

A method, system, and related equipment for drawing a brain map

This application claims the priority of the Chinese patent application with the application number 202111426615.7 and the application title "A method, system and related equipment for drawing a brain map" submitted to the China Patent Office on November 27, 2021, the entire contents of which are incorporated by reference in this application.

technical field

The present application relates to the field of artificial intelligence, in particular to a method, system and related equipment for drawing a brain map.

Background technique

Brain atlas is a model that expresses human brain tissue structure, brain functional structure, and brain nerve structure. Usually, after observing brain slices through microscopic imaging technology, multiple brain cross-sectional images can be obtained, and then three-dimensional reconstruction technology can be used to obtain Brain Atlas of the Brain. However, the shape of neurons in the brain is very complex, the nerve fibers of neurons are densely distributed, and the nerve fibers are twisted, criss-crossed, and intertwined. The brain atlas obtained after 3D reconstruction will have the problem of connection errors in the internal nerve fibers.

Usually, after the three-dimensional reconstruction of the brain, it is necessary to manually correct the shape of neurons in the brain atlas, and judge the connection relationship between nerve fibers through human eyes. This method of labeling is time-consuming and labor-intensive, which greatly hinders the The development of mapping analysis techniques.

Contents of the invention

The present application provides a brain atlas drawing method, system and related equipment, which can improve the efficiency of manual correction of the shape of neurons in the brain atlas, thereby improving the efficiency of brain atlas drawing and analysis.

In the first aspect, a method for drawing a brain atlas is provided, and the method includes the following steps: acquiring a first brain atlas, wherein the first brain atlas is obtained after three-dimensional reconstruction of the brain, and providing the user with the first brain atlas A plurality of nerve fiber segments where the key points are located obtains correction information input by the user, wherein the key points include intersection points of multiple nerve fiber segments or bifurcation points of a single nerve fiber segment in the first brain atlas, according to the correction information The connection relationship between multiple nerve fiber segments is corrected, and the second brain atlas is output.

In a specific implementation, the brain reconstruction data uploaded by the user may be obtained first, and then three-dimensional reconstruction is performed on the brain reconstruction data to obtain the above-mentioned first brain atlas. Wherein, the brain reconstruction data includes multiple cross-sectional images or multiple longitudinal-sectional images of the brain, and the above-mentioned brains can be the brains of living organisms, such as the brains of rodents such as monkey brains, zebrafish brains, and mice. Multiple cross-sectional images can be obtained by digital photography or microscope imaging of multiple thin-layer continuous longitudinal slices of the brain, and multiple longitudinal-sectional images can be multiple thin-layer continuous longitudinal slices of the brain obtained by digital photography or microscope imaging acquired afterwards. The brain reconstruction data may be uploaded by users or downloaded from other servers, which is not limited in this application. The first brain atlas includes the neuron morphology after three-dimensional reconstruction, and can also include brain structure information, functional connection information, etc., such as marking the prefrontal cortex, motor language area, and premotor cortex of the brain. This application is wrong. The brain map is specifically defined.

Implement the method described in the first aspect, by determining key points between nerve fibers, wherein the key points are intersection points between nerve fibers or bifurcation points of nerve fibers, and determine multiple nerve fiber segments according to the key points, the user When manually correcting the nerve fibers in the brain atlas, the user only needs to correct the connection relationship between the nerve fibers near the key points, and there is no need to distinguish the twisted, intertwined, and intertwined nerve fibers, which can improve the manual correction. The efficiency of the connection relationship between nerve fibers in the brain map, thereby improving the efficiency of brain map drawing.

In a possible implementation, the key points in the first brain atlas are determined by a machine learning method, or the key points in the first brain atlas are determined by a geometric topology algorithm to analyze the relationship between nerve fibers in the first brain atlas The geometric relationship is determined.

Optionally, since the nerve fibers of different neurons in the brain are different in thickness, in order to improve the accuracy of key point determination, the reconstructed neuron morphology can be skeletonized, and the diameters of multiple nerve fibers in the first brain atlas Unification, such as replacing itself with the central axis of each nerve fiber, obtains the first skeletonized brain atlas, and determines the key points of the skeletonized first brain atlas, which can improve the accuracy of key point determination.

Optionally, the geometric relationship between nerve fibers in the first brain atlas can be analyzed to determine the key points by means of a geometric topology algorithm. In the specific implementation, after obtaining the skeletonized first brain atlas, the geometric topological algorithm can be used to first determine the overlapping, adjacent or bifurcated areas of the nerve fibers in the first brain atlas, and then perform further processing on the nerve fibers in the area. The analysis determines the coordinates of the key points.

Optionally, the key points can also be determined by machine learning, and the first brain atlas can be input into the key point determination model to obtain the coordinates of the key points in the first brain atlas, wherein the key point determination model can use a sample set pair Obtained after the neural network is trained, the sample set includes known brain atlases and corresponding known key point coordinates. The above-mentioned neural network can be a convolutional neural network (convolutional neural networks, CNN), a recurrent neural network (recurrent neural network, RNN) ), deep neural networks (deep neural networks, DNN), etc., this application does not limit the types of neural networks.

Implementing the above implementation method, the key points determined by the machine learning method are more accurate, but occupy a large amount of computing resources. The accuracy of using the set topology algorithm is not high, but the computing resources are low, and the key point determination efficiency is high. Users The way to determine the key points can be selected according to the actual situation.

In a possible implementation, each nerve fiber segment includes two key points, or each nerve fiber segment includes a key point and the starting point of a neuron, or each nerve fiber segment includes a key point and a neuron end point.

Optionally, a nerve fiber segment may include two key points. For example, nerve fiber segment BC refers to a segment of nerve fiber segment intercepted by key point B and key point C. Then the multiple nerve fiber segments where key point B is located include the above-mentioned nerve fiber segment fiber segment BC; or, a nerve fiber segment can also include a key point and a neuron starting point; or, a nerve fiber segment can include a key point and a neuron end point (also called terminal or terminal), such as nerve Unit O includes nerve fiber segments XY and YZ, where Y is the intersection point with other nerve fibers, X is the starting point of the neuron, and Z is the end point of the neuron. It should be understood that the above examples are for illustration, and the present application does not specifically limit them.

It can be understood that the number of key points in the first brain atlas is multiple. When providing the user with multiple nerve fiber segments where the key points are located, the image area where the corresponding key points are located can be provided according to the user's operation. For example, the user can select the first Enlarge the first brain map, select the key point B that needs to be corrected, and provide the user with multiple nerve fiber segments where the key point B is located in response to the user's operation; or provide the user with a complete first brain map and multiple key points. The point coordinate list, after the user selects the coordinates of the key point B, provides the user with multiple nerve fiber segments where the key point B is located. It should be understood that the above examples are for illustration, and this application does not make specific limitations.

Implement the above implementation method, determine the nerve fiber segment through the key point, and then provide the user with multiple nerve fiber segments where the key point is located. Distinguishing the intertwined, crisscross, and intertwined nerve fibers can improve the efficiency of artificially correcting the connection relationship between nerve fibers in the brain atlas, thereby improving the efficiency of brain atlas drawing.

In a possible implementation, the nerve fiber segments where different key points are located can be continuously provided to the user, and the corresponding correction information input by the user is continuously received, and the nerve fiber segments where different key points are located are continuously corrected according to the correction information, until all The nerve fiber segment corresponding to the key point is corrected, and the second brain atlas is obtained.

Understandably, according to the user's business needs, usage habits and other personal information, different interaction modes can be designed to obtain the user's correction information. After determining the key points, the first brain map is provided to the user, and then the interaction mode selected by the user is obtained. Perform human-computer interaction with the user according to the interaction logic corresponding to the mode selected by the user, obtain correction information, and then correct the first brain atlas according to the correction information to obtain the second brain atlas.

Optionally, the above-mentioned interaction mode may include a single-body modification mode. Among them, the monomer correction mode refers to that the user corrects the connection relationship between multiple nerve fiber segments of one neuron at the granularity of neurons, and then corrects the connection relationship between multiple nerve fiber segments of another neuron. The connection relationship is corrected.

Specifically, multiple nerve fiber segments where each key point of the target neuron is located can be sequentially provided to the user to obtain the first correction information input by the user, wherein one key point corresponds to one first correction information, and the first correction information is used It is used to modify the connection relationship between the nerve fiber segments of the target neuron. That is to say, after the first brain atlas is provided to the user, the user can select the key point A on the target neuron that needs to be corrected on the first brain atlas, and in response to the user's operation, provide the user with multiple neurons where the key point A is located. Fiber segment, and receive the first correction information input by the user, the first correction information is used to correct the connection relationship between the nerve fiber segments of the target neuron in the multiple nerve fiber segments where the key point A is located. Then, the user can be provided with multiple nerve fiber segments where the key point B of the target neuron is located, wherein key point A and key point B are key points on the same nerve fiber segment of the target neuron, and receive user input The first correction information, the first correction information is used to correct the connection relationship between the nerve fiber segments of the target neuron in the multiple nerve fiber segments where the key point B is located, and then, the key point C can continue to be provided. Multiple nerve fiber segments, key point C and key point B are key points on the same nerve fiber segment of the target neuron, and so on until all the nerve fiber segments of the target neuron are corrected.

At the same time, the coordinates of the key points not marked by the user can be recorded, for example, the coordinates of the key points not marked by the user can be recorded in the unlabeled list. When the key point provided to the user is the starting point or end point of the target neuron, the The multiple nerve fiber segments where the key points recorded in the unlabeled list are provided to the user in sequence.

Among them, the key points not marked by the user may also include bifurcation points. After providing the user with multiple nerve fiber segments where each key point of each target neuron is located, the user may also be provided with the target neuron in sequence. For multiple nerve fiber segments where each bifurcation point is located, second correction information input by the user is obtained, and the second correction information is used to correct the connection relationship between the uncorrected nerve fiber segments of the target neuron. It can be understood that for the bifurcation point, when the user corrects the connection relationship between the multiple nerve fiber segments where the bifurcation point is located, he can only select one of the fork roads for correction, so the coordinates of the bifurcation point are recorded in the In the label list, when the user corrects a branch until it reaches the end point of the nerve fiber of the target neuron, the user can be provided with multiple nerve fiber segments where the branch point is located according to the coordinates in the unlabeled list, and the user can make another The fork is corrected, and so on, until the nerve fiber segment of the target neuron is corrected. If the first key point provided to the user is not the starting point or the end point of the target neuron, then the key points not marked by the user may include the above-mentioned first key point.

Optionally, the above-mentioned interaction mode may also include a group correction mode. Among them, the group correction mode refers to that the user corrects the connection relationship between all nerve fiber segments in one image area, and then corrects the connection relationship between all nerve fiber segments in the next image area. The relationship is corrected.

Specifically, the first image area may be provided to the user, wherein the first image area includes a plurality of nerve fiber segments where the first key point is located, and the first key point may be a key point selected by the user, or a key point randomly provided. point, this application does not specifically limit it. The third correction information input by the user may be received, and the third correction information is used to correct the connection relationship between all the nerve fiber segments in the first image area.

In specific implementation, the first brain atlas and key point coordinate list can be provided to the user, wherein the key point coordinate list includes multiple key point coordinates in the first brain atlas, and the user can select a key point in the key point coordinate list to provide The multiple nerve fiber segments where the key points selected by the user are located, the user corrects the connection relationship between multiple nerve fiber segments, and so on, whichever key point the user selects, the corresponding key point will be displayed in the image display area .

It should be noted that the above-mentioned individual correction mode and group correction mode are used for illustration. More correction modes can be set according to the user's business needs. Under different correction modes, different display interfaces can be designed to respond to different types of user requests. The operation steps are to realize the correction of the first brain atlas, which are not illustrated here one by one.

It should be noted that after the nerve fiber segments corresponding to all key points have been corrected, the second brain atlas can be obtained through deskeletonization. Referring to the above content, it can be seen that since the nerve fibers of different neurons in the brain are different in thickness, in order to improve the key To determine the accuracy of the point, the reconstructed neuron morphology is skeletonized at step S610, so after the first brain atlas is corrected at step S630, the deskeletonization operation can be performed to restore the original diameter of each nerve fiber.

In the above implementation manner, a complete target neuron can be corrected through the individual correction mode, and the connection relationship between all nerve fiber segments in the region can be corrected through the group correction mode. Users can choose the correction mode according to their own business needs. If the user needs to observe and analyze a certain neuron, the single correction mode can be used. If the user needs to observe and analyze the shape of neurons in a large area, the group correction can be used. The model meets the business needs of users in all aspects and improves the user experience.

In a second aspect, a brain atlas drawing system is provided, the system includes an acquisition unit for acquiring a first brain atlas, wherein the first brain atlas is obtained after three-dimensional reconstruction of the brain, and a key point determination unit for Provide the user with multiple nerve fiber segments where the key points in the first brain atlas are located, and obtain correction information input by the user, wherein the key points include intersections of multiple nerve fiber segments or a single nerve fiber segment in the first brain atlas The bifurcation point, the correction unit, is used to correct the connection relationship between multiple nerve fiber segments according to the correction information, and output the second brain atlas.

Implementing the system described in the second aspect, by determining key points between nerve fibers, wherein the key points are intersection points between nerve fibers or bifurcation points of nerve fibers, and determining a plurality of nerve fiber segments according to the key points, the user When manually correcting the nerve fibers in the brain atlas, the user only needs to correct the connection relationship between the nerve fibers near the key points, and there is no need to distinguish the twisted, intertwined, and intertwined nerve fibers, which can improve the manual correction. The efficiency of the connection relationship between nerve fibers in the brain map, thereby improving the efficiency of brain map drawing.

In a possible implementation, the key point determination unit is configured to sequentially provide the user with multiple nerve fiber segments where each key point of the target neuron is located, and obtain the first correction information input by the user, wherein one key point corresponds to A first correction information, the first correction information is used to correct the connection relationship between the nerve fiber segments of the target neuron.

In a possible implementation manner, the key point determining unit is configured to sequentially provide the user with each bifurcation point of the target neuron after providing the user with a plurality of nerve fiber segments where each key point of the target neuron is located The plurality of nerve fiber segments where the user is located obtains the second correction information input by the user, wherein a bifurcation point corresponds to a second correction information, and the second correction information is used for the uncorrected nerve fiber segments of the target neuron. The connection relationship is corrected.

In a possible implementation, the key point determination unit is configured to provide the user with a first image area, the first image area includes a plurality of nerve fiber segments where the first key point is located, and the first key point is a key point selected by the user a key point determining unit, configured to receive third correction information input by the user, and the third correction information is used to correct the connection relationship between all the nerve fiber segments in the first image area.

In a possible implementation, the system further includes a three-dimensional reconstruction unit, an acquisition unit, configured to acquire brain reconstruction data uploaded by users, wherein the brain reconstruction data includes multiple cross-sectional images or multiple longitudinal images of the brain; the three-dimensional The reconstruction unit is used to perform three-dimensional reconstruction on the brain reconstruction data to obtain the first brain atlas.

A third aspect provides a computing device, the computing device includes a processor and a memory, the memory stores codes, and the processor is configured to execute the method described in the first aspect or any possible implementation manner of the first aspect.

In a fourth aspect, a computer-readable storage medium is provided, where instructions are stored in the computer-readable storage medium, and when the computer-readable storage medium is run on a computer, the computer is made to execute the methods described in the above aspects.

On the basis of the implementation manners provided in the foregoing aspects, the present application may further be combined to provide more implementation manners.

Description of drawings

Figure 1 is a schematic diagram of the distribution of nerve fibers inside a brain atlas;

Fig. 2 is a structure diagram of a brain atlas drawing system provided by the present application;

Fig. 3 is a skeletonized example diagram provided by the present application;

Fig. 4 is an interface example diagram of a monomer correction mode provided by the present application;

Fig. 5 is an interface example diagram of a group correction mode provided by the present application;

Fig. 6 is an example diagram of the steps of a brain atlas drawing method provided by the present application;

Fig. 7 is a kind of human-computer interaction interface provided by the present application;

FIG. 8 is a schematic structural diagram of a computing device provided by the present application.

Detailed ways

The technical solutions in the embodiments of the present invention will be described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.

First, the application scenarios involved in this application are described.

Brain atlas is a model that expresses human brain tissue structure, brain functional structure, and brain nerve structure. Usually, after observing brain slices through microscopic imaging technology, multiple brain cross-sectional images can be obtained, and then three-dimensional reconstruction technology can be used to obtain The brain map of the brain, through the analysis of the brain map, can provide the necessary support for the analysis of the neural circuit principles of advanced cognitive functions, and provide precise neural circuit targets for the diagnosis and treatment of serious brain diseases.

However, the shape of neurons in the brain is very complex, and the nerve fibers of neurons are densely distributed. There is only one connection mode in the brain, so it is difficult for the current 3D reconstruction technology to identify the correct connection relationship among the densely distributed nerve fibers. After the 3D reconstruction of the brain, it is necessary to manually correct the neuron morphology in the brain atlas.

For example, Figure 1 is a schematic diagram of the distribution of nerve fibers inside a brain atlas, where the nerve fibers of different neurons can be marked with different colors or different shades, these reconstructed nerve fibers are twisted, criss-crossed, intertwined, and artificially corrected When it comes to the connection relationship between nerve fibers, it is very difficult for the staff to judge by human eyes. Moreover, if some part of the reconstructed nerve fiber has a wrong connection relationship, then the nerve fibers in other parts will have a wrong connection relationship. Therefore, the brain atlas The manual correction of the brain map is very time-consuming and labor-intensive, which greatly hinders the development of the drawing and analysis technology of the brain atlas.

In order to solve the time-consuming and labor-intensive problem of manually correcting the nerve fibers in the brain atlas, this application provides a brain atlas drawing system. After obtaining the first brain atlas after three-dimensional reconstruction, the system first determines the key points in the first brain atlas point, the key point includes the intersection of multiple nerve fiber segments or the bifurcation point of a single nerve fiber segment, and then displays the multiple nerve fiber segments where the key point is located to the user, obtains the correction information of the user, and performs multiple The connection relationship between the nerve fiber segments is corrected, so that the user only needs to correct the nerve fiber segments near the key points, and no longer needs to distinguish the twisted, intertwined, and intertwined nerve fibers. The efficiency of the connection relationship between fibers can improve the efficiency of brain map drawing.

FIG. 2 is an architecture diagram of a brain atlas drawing system provided by the present application. As shown in FIG. 2, the architecture includes a server 200 and a client 300, wherein there is a communication connection between the server 200 and the client 300, specifically It may be a wired connection or a wireless connection, which is not specifically limited in this application. It should be understood that FIG. 2 is only an exemplary division manner, and various systems, devices, and units may be combined or split into more or fewer systems, devices, and units, which are not specifically limited in this application.

The client 300 can be deployed on a terminal device held by a user, which is a terminal device with an image display function and a human-computer interaction function, such as a computer, a smart phone, a handheld processing device, a tablet computer, a mobile notebook, an augmented reality ( Augmented reality (AR) devices, virtual reality (virtual reality, VR) devices, integrated handhelds, wearable devices, vehicle-mounted devices, smart conference devices, smart advertising devices, smart home appliances, etc., are not specifically limited here. In a specific implementation, the client 300 may be an application program, a browser, an application program interface (application program interface, API), etc., which are not specifically limited in this application. In the public cloud scenario, the client 300 can be constructed based on common web development technologies such as html, javascript, and css, and the client 300 can interact with the server 200 through the websocket protocol or other communication protocols, which is not specifically limited in this application.

The server 200 may be a bare metal server (Bare Metal Server, BMS), a virtual machine or a container. Among them, BMS refers to a general-purpose physical server, such as an ARM server or an X86 server; a virtual machine refers to a network function virtualization (Network Functions Virtualization, NFV) technology that has complete hardware system functions through software simulation. A complete computer system running in a completely isolated environment, a container refers to a group of processes that are resource-constrained and isolated from each other. Optionally, the server 200 can be deployed in a cloud data center, and users can purchase corresponding services in the cloud data center to obtain the operation authority of the server 200 .

It should be noted that the client 300 and the server 200 in FIG. 2 are two different devices. In some embodiments, the client 300 can also be deployed in the server 200, that is, the server 200 has the function of image display. It can directly interact with users.

The server 200 can be further divided into a plurality of module units. For example, as shown in FIG. The client 300 may include a display unit 310. It should be understood that the division shown in FIG. 2 is for illustration, and the server 200 may also include more or less unit modules, which are not specifically limited in this application.

The acquisition unit 210 is used to acquire brain reconstruction data, and the brain reconstruction data includes multiple cross-sectional images or multiple longitudinal section images of the brain, wherein the above-mentioned brain can be the brain of an organism, such as a monkey brain, a zebrafish brain, a rodent such as a mouse, etc. animal brains. It should be understood that the brain reconstruction data may also include other data or information for three-dimensional reconstruction of the brain, which will not be illustrated here one by one. Brain reconstruction data can be uploaded by users or downloaded from other servers. For example, users store brain reconstruction data in the cloud data center's distributed file system (hadoop distributed file system, HDFS), object storage service In paths such as (object storage service, OBS), the acquisition unit 210 may download brain reconstruction data from the path, which is not limited in this application.

In a specific implementation, the multiple cross-sectional images can be obtained by digital photography or microscope imaging of multiple thin-layer continuous cross-sectional slices of the brain, and the multiple longitudinal sectional images can be multiple thin-layer continuous longitudinal slices of the brain obtained through digital imaging. Obtained by photography or microscopy imaging.

The three-dimensional reconstruction unit 220 is used to perform three-dimensional reconstruction on the brain reconstruction data to obtain a first brain atlas, wherein the first brain atlas includes neuron morphology after three-dimensional reconstruction, such as the example diagram of the brain atlas shown in Figure 1, the first brain atlas It can also include brain structure information, functional connection information, etc., such as marking the prefrontal cortex, motor language area, and premotor cortex of the brain. This application does not specifically limit the brain atlas.

In the specific implementation, since the internal structure of the brain is very complex, direct three-dimensional reconstruction of the entire brain requires a large amount of computing resources. Therefore, the obtained brain reconstruction data can be divided into data blocks to obtain the brain reconstruction data of multiple brain partitions, and then separately The brain reconstruction data of each partition were subjected to three-dimensional reconstruction. Optionally, after data slicing is performed on the brain reconstruction data, preprocessing operations such as data quality control may be performed on the slicing brain reconstruction data to further reduce the amount of data and improve the efficiency of 3D reconstruction.

In some embodiments, the server 200 may not include the 3D reconstruction unit 220, and the acquisition unit 210 may directly acquire the first brain atlas reconstructed by other devices in 3D, which is not specifically limited in this application. For example, the user can store the reconstructed first brain atlas in a path such as HDFS, OBS, etc., and the server 200 can download the reconstructed first brain atlas from the path. It should be understood that the above examples are for illustration, and the present application does not specifically limit them.

The key point determination unit 230 is used to determine the key points according to the first brain atlas, and provide the user with multiple nerve fiber segments where the key points in the first brain atlas are located, wherein the key points include intersection points of multiple nerve fibers in the brain Or the bifurcation point of a single nerve fiber. The key point determination unit 230 can obtain the coordinates of each key point in the first brain atlas, and store the type and coordinates of the key points. Here, the coordinates can be image coordinates of the first brain atlas, specifically a three-dimensional coordinate.

Optionally, since the thickness of nerve fibers of different neurons in the brain is different, in order to improve the accuracy of key point determination, the 3D reconstruction unit 220 can skeletonize the reconstructed neuron morphology, and combine multiple neurons in the first brain atlas The diameter of nerve fibers is unified, for example, the central axis of each nerve fiber is used to replace itself, and the first skeletonized brain atlas is obtained.

Exemplarily, FIG. 3 is an example diagram of skeletonization provided by the present application. FIG. 3 takes a partial region in the first brain atlas as an example. After the nerve fibers in this region are skeletonized, the skeletonized first A brain atlas, the key point determining unit 230 can determine the key points A˜H in FIG. 3 . It is understandable that the diameter of nerve fibers in the skeletonized first brain atlas is uniform, and the connection relationship between nerve fibers is clearly visible, which can not only improve the accuracy of key point determination, but also improve the processing efficiency of subsequent manual corrections.

Optionally, the key point determination unit 230 may determine the key points by analyzing the geometric relationship between nerve fibers in the first brain atlas through a geometric topology algorithm. In the specific implementation, after obtaining the skeletonized first brain atlas, the geometric topological algorithm can be used to first determine the overlapping, adjacent or bifurcated areas of the nerve fibers in the first brain atlas, and then perform further processing on the nerve fibers in the area. The analysis determines the coordinates of the key points.

Optionally, the key point determination unit 230 may also determine the key points by means of machine learning, input the first brain atlas into the key point determination model, and obtain the coordinates of the key points in the first brain atlas, wherein the key point determination model may It is obtained after training the neural network using a sample set. The sample set includes known brain maps and corresponding known key point coordinates. The above-mentioned neural network can be CNN, RNN, RCNN, DNN, etc. This application does not apply to the type of neural network To limit.

In a specific implementation, the key point determining unit 230 may provide the user with multiple nerve fiber segments where the key point is located through the display unit 310 , acquire correction information input by the user, and feed it back to the correction unit 240 . The display unit 310 may also display the first brain atlas and the skeletonized first brain atlas to the user.

Optionally, each nerve fiber segment may include two key points. For example, nerve fiber segment BC refers to a segment of nerve fiber segment intercepted by key point B and key point C. Then the multiple nerve fiber segments where key point B is located include the above-mentioned Nerve fiber fragment BC. For example, the image area shown in Figure 3 may include 7 nerve fiber segments, the nerve fiber segments where the key point B is located include AB, BE and BC, and the nerve fiber segments where the key point C is located include BC, GC, DC, HC and FC, it should be understood that FIG. 3 is used for illustration, and the present application does not specifically limit it.

Optionally, each nerve fiber segment may include a key point and a neuron starting point, or each nerve fiber segment may include a key point and a neuron end point (also called terminal or terminal), such as a neuron O includes nerve fiber segments XY and YZ, where Y is the intersection point with other nerve fibers, X is the starting point of the neuron, and Z is the end point of the neuron. It should be understood that the above examples are for illustration, and the present application does not specifically limit them.

It can be understood that the number of key points determined by the key point determination unit 230 is multiple, and when the key point determination unit 230 provides the user with multiple nerve fiber segments where the key points are located through the display unit 310, the display unit 310 can display according to the user's operation Display the image area where the corresponding key points are located. For example, the display unit 310 can mark all the key points in the first brain atlas (such as highlighting the key points in the first brain atlas), and display the key points to the user. The first brain atlas, the user can zoom in on the first brain atlas, select the key point B that needs to be corrected, and the display unit 310 will then display to the user the multiple nerve fiber segments where the key point B is located; or, the display unit 310 will display to the user A complete first brain atlas and a list of coordinates of multiple key points. After the user selects the coordinates of key point B, the display unit 310 displays to the user the multiple nerve fiber segments where key point B is located. It should be understood that the above examples are for illustration. This application does not make specific limitations.

The correction unit 240 is used for correcting the connection relationship among the above-mentioned multiple nerve fiber segments according to the correction information, so as to obtain the second brain atlas.

In an embodiment, the display unit 310 can continuously display the nerve fiber segments where different key points are located to the user, and continuously receive the corresponding correction information input by the user, and the correction unit 240 can continuously perform corrections on the nerve fiber segments where different key points are located according to the correction information. Correction until the nerve fiber segments corresponding to all key points are corrected, and the second brain atlas is obtained.

In the specific implementation, after the correction unit corrects the nerve fiber segments corresponding to all the key points, it can obtain the second brain atlas through deskeletonization. Referring to the above content, it can be seen that since the nerve fibers of different neurons in the brain are different in thickness, in order to To improve the accuracy of key point determination, the 3D reconstruction unit 220 skeletonizes the reconstructed neuron morphology, and unifies the diameters of multiple nerve fibers in the first brain atlas, so the correction unit 240 corrects the first brain atlas Afterwards, deskeletalization can be performed to restore the original diameter of each nerve fiber.

It can be understood that according to the user's business needs, usage habits and other personal information, different interaction modes can be designed to obtain the user's correction information. After the key point determination unit 230 determines the key points, the display unit 310 can display the first brain atlas to the user. , and then obtain the interaction mode selected by the user, perform human-computer interaction with the user according to the interaction logic corresponding to the mode selected by the user, obtain correction information, and then correct the first brain atlas according to the correction information to obtain the second brain atlas.

In an embodiment, the above-mentioned interactive mode may include a single-body modification mode. Among them, the monomer correction mode refers to that the user corrects the connection relationship between multiple nerve fiber segments of one neuron at the granularity of neurons, and then corrects the connection relationship between multiple nerve fiber segments of another neuron. The connection relationship is corrected.

Specifically, the key point determination unit 230 may sequentially provide the user with a plurality of nerve fiber segments where each key point of the target neuron is located through the display unit 310, and obtain the first correction information input by the user, wherein one key point corresponds to a first One correction information, the first correction information is used to correct the connection relationship between the nerve fiber segments of the target neuron. That is to say, after the display unit 310 displays the first brain atlas to the user, the user can select the key point A on the target neuron that needs to be corrected on the first brain atlas, and in response to the user's operation, the display unit 310 displays the key point to the user The multiple nerve fiber segments where A is located, and receive the first correction information input by the user, the first correction information is used to determine the connection relationship between the nerve fiber segments of the target neuron in the multiple nerve fiber segments where the key point A is located Make corrections. Next, the display unit 310 can display to the user a plurality of nerve fiber segments where key point B of the target neuron is located, wherein key point A and key point B are key points on the same nerve fiber segment of the target neuron, and receive The first correction information input by the user, the first correction information is used to correct the connection relationship between the nerve fiber segments of the target neuron in the multiple nerve fiber segments where the key point B is located, and then, the key point can continue to be displayed The multiple nerve fiber segments where C is located, key point C and key point B are key points on the same nerve fiber segment of the target neuron, and so on until all the nerve fiber segments of the target neuron are corrected.

For example, FIG. 4 is an example interface diagram of a monomer correction mode provided by the present application. Assuming that the user zooms in on the first brain atlas, the display unit 310 displays the interface 1 shown in FIG. 4 to the user in response to the user's operation. , interface 1 includes a plurality of nerve fiber segments, if the user selects key point A, display unit 310 can display interface 21 to the user, that is, the next key point of the nerve fiber segment where key point A is located (ie, key point B) as the center The interface 21 includes a plurality of nerve fiber segments where the key point B is located. If the user connects the nerve fiber segment AB to the nerve fiber segment BC, the display unit 310 can display the interface 22 to the user, that is, the nerve fiber segment BC. The image area centered on the next key point (i.e. key point C), the interface 22 includes a plurality of nerve fiber segments where the key point C is located, if the user connects the nerve fiber segment BC with the nerve fiber segment CD, the display unit 310 can display The user displays the interface 23, which is the image area centered on the next key point (i.e. key point D) of the nerve fiber segment CD. The interface 23 includes a plurality of nerve fiber segments where the key point D is located. If the user connects the nerve fiber segment CD and After the nerve fiber segment IJ is connected, the display unit 310 can display the interface 24 to the user, that is, the image area centered on the next key point (ie, key point J) of the nerve fiber segment IJ, and the interface 24 includes a plurality of key points where the key point J is located. For the nerve fiber segment, the user can connect the nerve fiber IJ to the nerve fiber JK. If the key point K is the end point or starting point of the nerve fiber of the target neuron, then the labeling operation of this monomer correction mode can end, and the display unit 310 can show the user The display interface 3 shows the marked nerve fibers of the target neuron to the user. If key point A and key point K are not the starting point and end point of the target neuron, other nerve fiber segments of the target neuron can be marked according to the above process. After the marking is completed, the display unit 310 can display interface 4 to the user. The target neuron has been corrected.

It should be understood that the above examples are for illustration, and this application does not limit the display interface during the labeling process. After the user selects a key point, as shown in the above example, the display unit 310 can display the image area centered on the next key point, or , if the next key point is still in the currently displayed image area, the display unit 310 can also highlight the next key point, or highlight the corrected nerve fiber segment without changing the displayed image area. When a key point is not in the currently displayed image area, a new image area centered on the next key point is displayed again, and after the user corrects it, the display unit 310 can directly display the next key point as shown in the above example as The image area in the center may also trigger a jump button (such as a "next" button), and then display the image area with the next key point as the center to the user, which is not limited in this application.

Optionally, the key point determination unit 230 may record the coordinates of the key points not marked by the user, for example, record the coordinates of the key points not marked by the user into the unmarked list, when the key point displayed to the user by the display unit 310 is the target When the neuron starts or ends, the display unit 310 can sequentially display the multiple nerve fiber segments where the key points recorded in the unmarked list are located to the user.

In a specific implementation, the key points not marked by the user may also include bifurcation points. After sequentially displaying to the user the multiple nerve fiber segments where each key point of each target neuron is located, the target neuron can also be provided to the user in sequence. The plurality of nerve fiber segments where each bifurcation point of the neuron is located obtains the second correction information input by the user, and the second correction information is used to correct the connection relationship between the uncorrected nerve fiber segments of the target neuron. It can be understood that for the bifurcation point, when the user corrects the connection relationship between the multiple nerve fiber segments where the bifurcation point is located, he can only select one of the fork roads for correction, so the coordinates of the bifurcation point are recorded in the In the marked list, when the user corrects a branch until it reaches the end point of the nerve fiber of the target neuron, the display unit 310 can display to the user the multiple nerve fiber segments where the branch point is located according to the coordinates in the unmarked list, and the user can Correct another fork, and so on, until the nerve fiber segment of the target neuron is corrected.

In a specific implementation, if the first key point displayed to the user by the display unit 310 is not the starting point or end point of the target neuron, then the key points not marked by the user may include the above-mentioned first key point, which is the key point in Figure 4 Point A. It should be understood that, in the example shown in FIG. 4, if the key point A is not the starting point or the end point of the target neuron, there is still a nerve fiber segment that needs to be corrected on the left side of the key point A, so when the interface 21 (key point B) is displayed to the user is the central image area), the display unit 310 may record the coordinates of the key point A in the unmarked list.

In an embodiment, the above-mentioned interaction mode may also include a group correction mode. Among them, the group correction mode refers to that the user corrects the connection relationship between all nerve fiber segments in one image area, and then corrects the connection relationship between all nerve fiber segments in the next image area. The relationship is corrected.

Specifically, the key point determination unit 230 can provide the user with the first image area through the display unit 310, wherein the first image area includes a plurality of nerve fiber segments where the first key point is located, and the first key point can be selected by the user. The key points may also be randomly displayed key points, which is not specifically limited in this application. The display unit 310 may receive third correction information input by the user, and the third correction information is used to correct the connection relationship between all the nerve fiber segments in the first image area.

In a specific implementation, the display unit 310 can display the first brain atlas and a key point coordinate list to the user, wherein the key point coordinate list includes a plurality of key point coordinates in the first brain atlas, and the user can select in the key point coordinate list Key points, the display unit 310 displays the multiple nerve fiber segments where the key points selected by the user are located, the user corrects the connection relationship between the multiple nerve fiber segments displayed on the display unit 310, and so on, which key point the user selects point, the display unit 310 displays the corresponding key point in the image display area.

For example, as shown in FIG. 5, FIG. 5 is an interface example diagram of a group correction mode provided by the present application. Assuming that the user selects a key point C from the key point coordinate list, the display unit 310 can display the key point C to the user. C is the image area centered, which is the image area shown in interface 1 in Figure 5. This image area includes key points A~H, and the user can modify the connection relationship between all nerve fibers in this image area. After correcting the nerve fiber segment where a key point is located, it can be deleted in the key point coordinate list. Specifically, the display unit 310 can delete the corrected key point in the key point coordinate list according to user operations, or it can be deleted by the user , this application does not specifically limit. After the user corrects all the nerve fiber segments in the interface 1, the display unit 310 can display the interface 2 in FIG. The coordinates of a key point, that is, coordinates I～N, the user can select any key point in the coordinates I～N, such as key point I, and the display unit 310 can display the image area centered on key point I to the user, for the user Multiple nerve fiber segments in the image area are corrected, and so on, and examples are not given here.

It should be understood that FIG. 5 is used for illustration, and the present application does not limit the specific interaction logic of human-computer interaction. For example, as shown in Figure 5, the user can splice all the nerve fiber segments in the image area displayed by the display unit 310, or use other methods such as auditing, and directly connect the corresponding nerve fiber segments without correcting them. The key point coordinates of are removed from the key point coordinate list, the wrongly connected nerve fiber segments are corrected, and the corrected key point coordinates are removed from the key point coordinate list.

In summary, the brain atlas drawing system provided by the present application determines the key points between nerve fibers, wherein the key points are intersection points between nerve fibers or bifurcation points of nerve fibers, and multiple key points are determined according to the key points. Nerve fiber fragments, when the user manually corrects the nerve fibers in the brain atlas, the user only needs to correct the connection relationship between the nerve fibers near the key points, and does not need to distinguish the twisted, intertwined, and intertwined nerve fibers , can improve the efficiency of manual correction, and then improve the efficiency of drawing brain maps.

The method for drawing a brain map provided by the present application will be explained below in conjunction with the accompanying drawings.

As shown in Figure 6, Figure 6 is an example diagram of the steps of a brain atlas drawing method provided by the present application. This method can be applied to the brain atlas drawing system shown in Figures 2 to 5, and the method may include the following steps :

S610: Obtain a first brain atlas, where the first brain atlas is obtained after three-dimensional reconstruction of the brain. This step can be implemented by the acquisition unit 210 and the three-dimensional reconstruction unit 220 in FIG. 2 .

Specifically, the brain reconstruction data uploaded by the user may be obtained first, and then three-dimensional reconstruction is performed on the brain reconstruction data to obtain the above-mentioned first brain atlas. Wherein, the brain reconstruction data includes multiple cross-sectional images or multiple longitudinal-sectional images of the brain, and the above-mentioned brains can be the brains of living organisms, such as the brains of rodents such as monkey brains, zebrafish brains, and mice. Multiple cross-sectional images can be obtained by digital photography or microscope imaging of multiple thin-layer continuous longitudinal slices of the brain, and multiple longitudinal-sectional images can be multiple thin-layer continuous longitudinal slices of the brain obtained by digital photography or microscope imaging acquired afterwards. The brain reconstruction data may be uploaded by users or downloaded from other servers, which is not limited in this application.

In a specific implementation, the first brain atlas includes the neuron morphology after three-dimensional reconstruction, such as the example map of the brain atlas shown in Figure 1, the first brain atlas can also include brain structure information, functional connection information, etc., such as marking the brain The application does not specifically limit the brain map for the prefrontal cortex area, motor language area, premotor cortex area, etc.

It should be understood that for the content not described in step S610, reference may be made to the descriptions of the acquisition unit 210 and the three-dimensional reconstruction unit 220, and details are not repeated here.

S620: Provide the user with multiple nerve fiber segments where the key points in the first brain atlas are located, and obtain correction information input by the user, where the key points include intersections of multiple nerve fiber segments or a single nerve in the first brain atlas Bifurcation points of fiber segments. This step can be implemented by the key point determination unit 230 and the display unit 310 in FIG. 2 . The key point determination unit 230 can obtain the coordinates of each key point in the first brain atlas, and store the type and coordinates of the key points. Here, the coordinates can be image coordinates of the first brain atlas, specifically a three-dimensional coordinate. The display unit 310 displays multiple nerve fiber segments where key points are located to the user, and obtains correction information input by the user.

Optionally, since the nerve fibers of different neurons in the brain are different in thickness, in order to improve the accuracy of key point determination, the reconstructed neuron morphology can be skeletonized, and the diameters of multiple nerve fibers in the first brain atlas Unify, for example, replace itself with the central axis of each nerve fiber to obtain the first skeletonized brain atlas. For the skeletonized description, reference may be made to the embodiment in FIG. 3 , and examples are not repeated here.

Optionally, the key points can also be determined by machine learning, and the first brain atlas can be input into the key point determination model to obtain the coordinates of the key points in the first brain atlas, wherein the key point determination model can use a sample set pair The neural network is obtained after training. The sample set includes known brain atlases and corresponding known key point coordinates. The above-mentioned neural network can be CNN, RNN, RCNN, DNN, etc. This application does not limit the type of neural network.

Optionally, each nerve fiber segment may include two key points. For example, nerve fiber segment BC refers to a segment of nerve fiber segment intercepted by key point B and key point C. Then the multiple nerve fiber segments where key point B is located include the above-mentioned Nerve fiber segment BC; or, each nerve fiber segment may also include a key point and a neuron starting point; or, each nerve fiber segment may include a key point and a neuron end point (also called terminal or terminal) , for example, neuron O includes nerve fiber segments XY and YZ, where Y is the intersection point with other nerve fibers, X is the starting point of the neuron, and Z is the end point of the neuron. It should be understood that the above examples are for illustration, and the present application does not specifically limit them.

It should be understood that for the content not described in step S620, reference may be made to the description of the aforementioned key point determination unit 230 and the display unit 310, which will not be repeated here.

S630: Correct the connection relationship between the multiple nerve fiber segments according to the correction information, and output the second brain atlas. This step can be implemented by the correction unit 240 in FIG. 2 .

In one embodiment, step S620 and step S630 can be repeatedly executed to continuously provide the user with nerve fiber segments where different key points are located, and continuously receive corresponding correction information input by the user, and continuously adjust the nerve fiber segments where different key points are located according to the correction information. The fragments are corrected until the nerve fiber fragments corresponding to all the key points are corrected, and the second brain atlas is obtained.

In the specific implementation, after the nerve fiber segments corresponding to all the key points are corrected, the second brain atlas can be obtained through deskeletonization. Referring to the above content, it can be seen that since the nerve fibers of different neurons in the brain are different in thickness, in order to improve the key point To determine the accuracy of the point, the reconstructed neuron morphology is skeletonized at step S610, so after the first brain atlas is corrected at step S630, the deskeletonization operation can be performed to restore the original diameter of each nerve fiber.

Specifically, multiple nerve fiber segments where each key point of the target neuron is located can be sequentially provided to the user to obtain the first correction information input by the user, wherein one key point corresponds to one first correction information, and the first correction information is used It is used to modify the connection relationship between the nerve fiber segments of the target neurons. That is to say, after the first brain atlas is provided to the user, the user can select the key point A on the target neuron that needs to be corrected on the first brain atlas, and in response to the user's operation, provide the user with multiple neurons where the key point A is located. Fiber segment, and receive the first correction information input by the user, the first correction information is used to correct the connection relationship between the nerve fiber segments of the target neuron in the multiple nerve fiber segments where the key point A is located. Then, the user can be provided with multiple nerve fiber segments where the key point B of the target neuron is located, wherein key point A and key point B are key points on the same nerve fiber segment of the target neuron, and receive user input The first correction information, the first correction information is used to correct the connection relationship between the nerve fiber segments of the target neuron in the multiple nerve fiber segments where the key point B is located, and then, the key point C can continue to be provided. Multiple nerve fiber segments, key point C and key point B are key points on the same nerve fiber segment of the target neuron, and so on until all the nerve fiber segments of the target neuron are corrected. For details, reference may be made to the description of the foregoing embodiment in FIG. 4 , and examples are not repeated here.

Optionally, the coordinates of the key points not marked by the user can be recorded, for example, the coordinates of the key points not marked by the user are recorded in the unmarked list. When the key point provided to the user is the starting point or end point of the target neuron, Multiple nerve fiber segments where the key points recorded in the unlabeled list are located may be sequentially provided to the user.

In a specific implementation, the key points not marked by the user may also include bifurcation points. After sequentially providing the user with multiple nerve fiber segments where each key point of each target neuron is located, the user may also be sequentially provided with the target neuron The plurality of nerve fiber segments where each bifurcation point of the neuron is located obtains the second correction information input by the user, and the second correction information is used to correct the connection relationship between the uncorrected nerve fiber segments of the target neuron. It can be understood that for the bifurcation point, when the user corrects the connection relationship between the multiple nerve fiber segments where the bifurcation point is located, he can only select one of the fork roads for correction, so the coordinates of the bifurcation point are recorded in the In the label list, when the user corrects a branch until it reaches the end point of the nerve fiber of the target neuron, the user can be provided with multiple nerve fiber segments where the branch point is located according to the coordinates in the unlabeled list, and the user can make another The fork is corrected, and so on, until the nerve fiber segment of the target neuron is corrected.

In a specific implementation, if the first key point provided to the user is not the starting point or end point of the target neuron, then the key points not marked by the user may include the first key point, that is, key point A in FIG. 4 .

In specific implementation, the first brain atlas and key point coordinate list can be provided to the user, wherein the key point coordinate list includes multiple key point coordinates in the first brain atlas, and the user can select a key point in the key point coordinate list to provide The multiple nerve fiber segments where the key points are selected by the user, the user corrects the connection relationship between the multiple nerve fiber segments, and so on. For details, reference may be made to the description in the embodiment in FIG. 5 , and examples are not repeated here.

FIG. 7 is a human-computer interaction interface provided by the present application. This interface can be displayed to the user after the first brain atlas is obtained in step S610. Referring to the foregoing, it can be seen that this interface can be realized by the display unit 310 shown in FIG. 2 . As shown in Figure 7, the interface may include a brain atlas display interface 710, a partial display interface 720, a setting interface 730, an unmarked list 740, and a key point coordinate list 750. It should be understood that Figure 7 is used for illustration, the human-computer interaction The interface may also include more or less content, and this application does not specifically limit the interface of human-computer interaction.

The brain atlas display interface 710 is used to display the first brain atlas to the user. Specifically, it can be the skeletonized first brain atlas as shown in FIG. 7 , or the first brain atlas without skeletonization processing. This application is not correct. This is limited.

The local display interface 720 is used to display the local area of the first brain atlas to the user, and the user can modify the connection relationship between the nerve fibers on this interface.

Optionally, the local area may be an area selected by the user from the brain atlas display interface 710. For example, as shown in FIG. , the local area selected by the marquee tool 711 can be displayed in the partial display interface 720 .

Optionally, the local area can also be an image area centered on a key point, and the key point can be a key point selected by the user from the unmarked list 740 or the key point coordinate list 750. For details, refer to the aforementioned FIG. 4 and FIG. 5. The description about the unmarked list and the key point coordinate list in the embodiment will not be repeated here.

Optionally, the partial display interface 720 may include a button 721. When the user clicks the button 721, the partial display interface 720 may display other nerve fiber segments to the user, which may be determined according to the correction mode selected by the user. For example, in the monomer correction mode, when the user clicks the button 721, the local display interface 720 can display to the user a plurality of nerve fiber segments where the next key point of the target neuron is located; in the group correction mode, when the user clicks the button 721, the local display interface 720 The display interface 720 can display multiple nerve fiber segments where the next key point in the key point coordinate list 750 is located to the user. For details, reference may be made to the descriptions in the foregoing embodiments, and details are not repeated here.

The setting interface 730 is used to provide the user with a setting interface. Exemplarily, as shown in FIG. Add the coordinates of the key points that need to be corrected in the unlabeled list 740 or the key point coordinate list 750, and also include a setting button, which is used for the user to set the image resolution, the color of the key point or the nerve fiber segment, etc. This application does not make any Specific limits.

For the unmarked list 740 and the key point coordinates list 750, reference may be made to the description of the foregoing embodiments, and details are not repeated here.

In this embodiment of the application, the user can select the required correction mode through the setting interface 730. If the correction mode selected by the user is a single correction mode, the user can first select the correction mode to be modified from the brain map display interface 710 using the frame selection tool 711. Area, the local display interface 720 displays the corresponding local area, and the user can modify the connection relationship between the nerve fibers of the target neuron in the local display interface 720. For example, in FIG. After the connection is made and the button 721 is clicked, the partial display interface 720 can display the image area centered on the key point D to the user. For details, please refer to the description in the embodiment in FIG. 4 , which will not be repeated here.

It should be noted that in the single correction mode, the brain map drawing system can automatically add the coordinates of the bifurcation point to the unlabeled list 740 during the user's annotation process, or the user can manually add the coordinates of the bifurcation point to the unlabeled list 740 , this application does not limit it. After the user finishes correcting the nerve fiber of the target neuron, the user can click on the coordinates in the unmarked list 740 to continue to correct another nerve fiber branch at the bifurcation point.

If the correction mode selected by the user is the group correction mode, first use the marquee tool 711 to select the area to be modified from the brain atlas display interface 710, and the local display interface 720 displays the corresponding local area, or select coordinates from the key point coordinate list 750, The local display interface 720 displays the local area where the coordinates are located, and the user can modify the connection relationship between all nerve fibers in the local area, and then click the button 721, and the local display interface 720 can display the coordinates of the key points in the list 750 to the user. The multiple nerve fiber segments where the next key point is located, and so on, for details, refer to the description in the embodiment in FIG. 5 , which will not be repeated here.

In summary, the brain atlas drawing system provided by the present application determines the key points between nerve fibers, wherein the key points are intersection points between nerve fibers or bifurcation points of nerve fibers, and multiple key points are determined according to the key points. Nerve fiber fragments, when the user manually corrects the nerve fibers in the brain atlas, the user only needs to correct the connection relationship between the nerve fibers near the key points, and does not need to distinguish the twisted, intertwined, and intertwined nerve fibers , can improve the efficiency of artificially correcting the connection relationship between nerve fibers in the brain atlas, thereby improving the efficiency of brain atlas drawing.

FIG. 8 is a schematic structural diagram of a computing device provided by the present application. The computing device 800 may be a client or a server in the embodiments of FIG. 1 to FIG. 7, and the computing device may be a physical server, a virtual machine or a server cluster, It can also be a chip (system) or other components or components that can be set on a physical server or a virtual machine, which is not limited in this application.

Further, the computing device 800 includes a processor 801, a memory 802, and a communication interface 803, where the processor 801, the memory 802, and the communication interface 803 communicate through a bus 805, and may also communicate through other means such as wireless transmission.

The processor 801 may be composed of at least one general-purpose processor, such as a CPU, an NPU, or a combination of a CPU and a hardware chip. The aforementioned hardware chip may be an application-specific integrated circuit (Application-Specific Integrated Circuit, ASIC), a programmable logic device (Programmable Logic Device, PLD) or a combination thereof. The above-mentioned PLD can be a complex programmable logic device (Complex Programmable Logic Device, CPLD), a field programmable logic gate array (Field-Programmable Gate Array, FPGA), a general array logic (Generic Array Logic, GAL) or any combination thereof. The processor 801 executes various types of digitally stored instructions, such as software or firmware programs stored in the memory 802, which enable the computing device 800 to provide a wide variety of services.

In a specific implementation, as an embodiment, the processor 801 may include one or more CPUs, such as CPU0 and CPU1 shown in FIG. 8 .

In a specific implementation, as an embodiment, the computing device 800 may also include multiple processors, such as the processor 801 and the processor 804 shown in FIG. 8 . Each of these processors can be a single-core processor (single-CPU) or a multi-core processor (multi-CPU). A processor herein may refer to one or more devices, circuits, and/or processing cores for processing data (eg, computer program instructions).

The memory 802 is used to store program codes, which are executed under the control of the processor 801, so as to execute the processing steps of the workflow system in any of the above-mentioned embodiments in FIGS. 1-7. One or more software modules may be included in the program code.

When the computing node is a server, the above one or more software modules can be the acquisition unit, the key point determination unit and the correction unit in the embodiment of Fig. 2, wherein the acquisition unit is used to obtain the first brain atlas, and the key point determination unit uses To determine the key points of the first brain atlas, provide the user with multiple nerve fiber segments where the key points are located, the correction unit is used to receive the correction information uploaded by the user, and correct the first brain atlas according to the correction information to obtain the second brain atlas . For the above specific implementation manner, reference may be made to the embodiments in FIG. 2 to FIG. 7 , which will not be repeated here.

When the computing node is a client, the above one or more software modules may be the display unit in the embodiment of FIG. 2, wherein the display unit is used to display to the user the multiple nerve fiber segments where the key points sent by the server are located, and The correction information fed back by the user is received to the server. For the above specific implementation manner, reference may be made to the method embodiments in FIG. 2 to FIG. 7 , which will not be repeated here.

The memory 802 may include read-only memory and random-access memory, and provides instructions and data to the processor 801 . Memory 802 may also include non-volatile random access memory. For example, memory 802 may also store device type information.

Memory 802 can be volatile memory or nonvolatile memory, or can include both volatile and nonvolatile memory. Among them, the non-volatile memory can be read-only memory (read-only memory, ROM), programmable read-only memory (programmable ROM, PROM), erasable programmable read-only memory (erasable PROM, EPROM), electronically programmable Erases programmable read-only memory (electrically EPROM, EEPROM) or flash memory. Volatile memory can be random access memory (RAM), which acts as external cache memory. By way of illustration and not limitation, many forms of RAM are available such as static random access memory (static RAM, SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (synchronous DRAM, SDRAM), Double data rate synchronous dynamic random access memory (double data date SDRAM, DDR SDRAM), enhanced synchronous dynamic random access memory (enhanced SDRAM, ESDRAM), synchronous connection dynamic random access memory (synchlink DRAM, SLDRAM) and direct Memory bus random access memory (direct rambus RAM, DR RAM). It can also be a hard disk (hard disk), U disk (universal serial bus, USB), flash memory (flash), SD card (secure digital memory Card, SD card), memory stick, etc., and the hard disk can be a hard disk drive (hard disk drive). , HDD), solid state disk (solid state disk, SSD), mechanical hard disk (mechanical hard disk, HDD), etc., which are not specifically limited in this application.

The communication interface 803 can be a wired interface (such as an Ethernet interface), an internal interface (such as a high-speed serial computer expansion bus (Peripheral Component Interconnect express, PCIe) bus interface), a wired interface (such as an Ethernet interface) or a wireless interface ( For example, a cellular network interface or a wireless local area network interface) is used to communicate with other servers or modules. In specific implementation, the communication interface 803 can be used to receive a message for the processor 801 or processor 804 to process the message.

The bus 805 can be a peripheral component interconnection standard (Peripheral Component Interconnect Express, PCIe) bus, or an extended industry standard architecture (extended industry standard architecture, EISA) bus, unified bus (unified bus, Ubus or UB), computer fast link (compute express link, CXL), cache coherent interconnect for accelerators (CCIX), etc. The bus 805 can be divided into an address bus, a data bus, a control bus, and the like.

It should be noted that FIG. 8 is only a possible implementation of the embodiment of the present application. In practical applications, the computing device 800 may include more or fewer components, which is not limited here. Regarding the content not shown or described in the embodiment of the present application, reference may be made to the related explanations in the foregoing embodiments of FIGS. 1-7 , which will not be repeated here.

It should be understood that the computing device 800 shown in FIG. 8 can also be a computer cluster composed of at least one physical server. For details, reference can be made to the specific description of the brain map drawing system in the embodiments of FIGS. 1 to 7. In order to avoid repetition, no Let me repeat.

An embodiment of the present application provides a chip, which can be specifically used in a server where a processor of the X86 architecture resides (also may be called an X86 server), a server where a processor of the ARM architecture resides (also may be referred to as an ARM server for short), etc. The chip may include the above-mentioned devices or logic circuits, and when the chip is run on the server, the server is made to execute the brain atlas drawing method described in the above method embodiments.

An embodiment of the present application provides a computer-readable storage medium, including: computer instructions are stored in the computer-readable storage medium; when the computer instructions are run on a computer, the computer is made to execute the brain map described in the above method embodiment drawing method.

An embodiment of the present application provides a computer program product containing instructions, including a computer program or instruction, when the computer program or instruction is run on a computer, it causes the computer to execute the brain atlas drawing method described in the method embodiment above.

The above-mentioned embodiments may be implemented in whole or in part by software, hardware, firmware or other arbitrary combinations. When implemented using software, the above-described embodiments may be implemented in whole or in part in the form of computer program products. A computer program product comprises at least one computer instruction. When the computer program instructions are loaded or executed on the computer, the processes or functions according to the embodiments of the present invention will be generated in whole or in part. The computer can be a general purpose computer, a special purpose computer, a computer network, or other programmable devices. Computer instructions may be stored in or transmitted from one computer-readable storage medium to another computer-readable storage medium, e.g. Coaxial cable, optical fiber, digital subscriber line (digital subscriber line, DSL)) or wireless (such as infrared, wireless, microwave, etc.) transmission to another website site, computer, server or data center. A computer-readable storage medium may be any available medium that can be accessed by a computer, or a data storage node such as a server or a data center that includes at least one set of available media. Available media may be magnetic media (eg, floppy disks, hard disks, tapes), optical media (eg, high-density digital video discs (DVD), or semiconductor media. The semiconductor media may be SSDs.

The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of various equivalent repairs or repairs within the technical scope disclosed in the present invention. Replacement, these repairs or replacements should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

A method for drawing a brain map, characterized in that the method comprises:

Obtaining a first brain atlas, wherein the first brain atlas is obtained after three-dimensional reconstruction of the brain;

providing the user with multiple nerve fiber segments where the key points in the first brain atlas are located, and obtaining correction information input by the user, wherein the key points include multiple nerve fiber segments in the first brain atlas The intersection point or bifurcation point of a single nerve fiber segment;

Correcting the connection relationship among the plurality of nerve fiber segments according to the correction information, and outputting a second brain atlas.
The method according to claim 1, wherein the key points in the first brain atlas are determined by a machine learning method, or the key points in the first brain atlas are analyzed by a geometric topology algorithm The geometric relationship between nerve fibers in the first brain atlas is determined.
The method according to claim 2, wherein the providing the user with a plurality of nerve fiber segments where the key point is located, and obtaining the correction information input by the user includes:

sequentially provide the user with multiple nerve fiber segments where each key point of the target neuron is located, and obtain multiple first correction information input by the user, wherein one key point corresponds to one first correction information, and the first correction The information is used to modify the connection relationship between the nerve fiber segments of the target neuron.
The method according to claim 3, characterized in that, after the multiple nerve fiber segments where each key point of the target neuron is provided to the user in turn, the method further comprises:

sequentially provide the user with a plurality of nerve fiber segments where each bifurcation point of the target neuron is located, and obtain a plurality of second correction information input by the user, wherein one bifurcation point corresponds to one second correction information, so The second correction information is used to correct the connection relationship between the uncorrected nerve fiber segments of the target neuron.
The method according to any one of claims 1 to 4, wherein providing the user with a plurality of nerve fiber segments where the key point is located, and obtaining the correction information input by the user includes:

providing the user with a first image area, the first image area including a plurality of nerve fiber segments where the first key point is located, the first key point being the key point selected by the user;

The third correction information input by the user is received, and the third correction information is used to correct the connection relationship between all the nerve fiber segments in the first image area.
The method according to any one of claims 1 to 5, wherein said obtaining the first brain atlas comprises:

Obtaining brain reconstruction data uploaded by users, wherein the brain reconstruction data includes multiple cross-sectional images or multiple longitudinal images of the brain;

performing three-dimensional reconstruction on the brain reconstruction data to obtain the first brain atlas.
The method according to any one of claims 1 to 6, wherein each nerve fiber segment comprises two key points, or each nerve fiber segment comprises a key point and the starting point of a neuron, or each A nerve fiber segment consists of a key point and a neuron terminal.
A brain map drawing system, characterized in that the system includes:

An acquisition unit, configured to acquire a first brain atlas, wherein the first brain atlas is obtained after three-dimensional reconstruction of the brain;

A key point determination unit, configured to provide the user with a plurality of nerve fiber segments where the key points in the first brain atlas are located, and obtain correction information input by the user, wherein the key points include the first brain atlas The intersection point of multiple nerve fiber segments or the bifurcation point of a single nerve fiber segment;

The correction unit is configured to correct the connection relationship between the plurality of nerve fiber segments according to the correction information, and output the second brain atlas.
The system according to claim 8, wherein the key points in the first brain atlas are determined by a machine learning method, or the key points in the first brain atlas are analyzed by a geometric topology algorithm. The geometric relationship between nerve fibers in the first brain atlas is determined.
The system according to claim 9, wherein the key point determination unit is configured to sequentially provide the user with a plurality of nerve fiber segments where each key point of the target neuron is located, and obtain a plurality of nerve fiber segments input by the user. The first correction information, wherein one key point corresponds to one first correction information, and the first correction information is used to correct the connection relationship between the nerve fiber segments of the target neuron.
The system according to claim 10, wherein the key point determination unit is configured to sequentially provide the user with a plurality of nerve fiber segments where each key point of the target neuron is located. A plurality of nerve fiber segments where each bifurcation point of the target neuron is located obtains a plurality of second correction information input by the user, wherein one bifurcation point corresponds to one second correction information, and the second correction The information is used to modify the connection relationship between the uncorrected nerve fiber segments of the target neuron.
A system according to any one of claims 8 to 11, characterized in that,

The key point determination unit is configured to provide the user with a first image area, the first image area includes a plurality of nerve fiber segments where the first key point is located, and the first key point is the key point selected by the user ;

The key point determining unit is configured to receive third correction information input by the user, and the third correction information is used to correct connection relationships between all nerve fiber segments in the first image area.
The system according to any one of claims 8 to 12, wherein the system further comprises a three-dimensional reconstruction unit,

The acquisition unit is configured to acquire brain reconstruction data uploaded by users, wherein the brain reconstruction data includes multiple cross-sectional images or multiple longitudinal sectional images of the brain;

The three-dimensional reconstruction unit is configured to perform three-dimensional reconstruction on the brain reconstruction data to obtain the first brain atlas.
The system according to any one of claims 8 to 13, wherein each nerve fiber segment comprises two key points, or each nerve fiber segment comprises a key point and the starting point of a neuron, or each A nerve fiber segment consists of a key point and a neuron terminal.
A computing device, characterized by comprising a processor and a memory, the memory is used to store codes, and the processor is used to execute the codes to implement the method according to any one of claims 1 to 7.