Node selection method, multi-brain collaborative brain-computer interface system, storage medium and equipment

Abstract

The invention provides a node selection method, a multi-brain collaborative brain-computer interface system, a storage medium and equipment. The node selection method comprises the following steps: S11, acquiring data of each node in a node set, wherein the node set comprises a plurality of nodes in the brain networking system; s12, traversing all nodes in the node set, and calculating a redundant node measurement index of each node according to the data; s13, selecting redundant nodes from the node set according to the redundant node measurement indexes; s14, deleting the redundant nodes from the node set; and S15, repeatedly executing the steps S12 to S14 until the nodes in the node set meet the termination condition. The node selection method can reduce the number of nodes and save the labor cost on the premise of ensuring the system performance.

Description

Technical Field

[0001] This application belongs to the technical field of multi-brain collaboration, and relates to a node selection method, in particular to a node selection method, a multi-brain collaborative brain-computer interface system, a storage medium, and a device. Background Art

[0002] Brain-Computer Interfaces (BCI) is a technology that establishes direct communication between the human (or animal) brain and the external world. Through this technology, human (or animal) brain signals can be used to express sensory feelings or control external devices.

[0003] From "single brain" to "multi-brain", that is, to achieve the socialization of brain-computer interfaces, is one of the important development trends of brain-computer interface technology. The Internet of Brains (IoB) is a technology that connects "multi-brains" to the Internet, where each independent "brain" is a node of the Internet of Brains. Compared with current communication systems used for human interaction with the physical world, such as the Internet of Things (IoT), etc., the Internet of Brains can provide a more direct interaction channel between users and between users and the external environment.

[0004] Collaborative Brain-Computer Interfaces (cBCI) is a type of Internet of Brains technology. This technology can fuse electroencephalograms (EEGs) from multiple nodes in the Internet of Brains that perform the same task to improve the performance of the overall brain-computer interface system. The performance of the multi-brain collaborative brain-computer interface system will improve as the number of nodes increases. In some actual application scenarios, it is often necessary to consume more labor (increase the number of nodes in the Internet of Brains system) to achieve more reliable collaborative work. Internet of Brains node selection is a technology for selecting a small number of nodes from multiple nodes within the Internet of Brains system. Through this technology, costs can be significantly saved while ensuring the overall performance of the system. However, the effects of existing node selection methods are often not ideal. Summary of the Invention

[0005] Embodiments of this application provide a node selection method, a multi-brain collaborative brain-computer interface system, a storage medium, and a device, which are used to reduce the number of nodes and save labor costs while ensuring the performance of the system.

[0006] In a first aspect, an embodiment of the present application provides a node selection method applied to a multi-brain collaborative brain-computer interface system, characterized in that the node selection method includes: S11, obtaining data of each node in a node set, where the node set includes multiple nodes in a brain network system; S12, traversing all nodes in the node set and calculating a redundancy node metric for each node according to the data; S13, selecting redundancy nodes from the node set according to the redundancy node metric; S14, deleting the redundancy nodes from the node set; S15, repeating steps S12 to S14 until the nodes in the node set meet a termination condition.

[0007] In one implementation manner of the first aspect, the redundancy node metric is constructed based on the minimum redundancy maximum correlation criterion.

[0008] In one implementation manner of the first aspect, for a node in the node set, calculating the redundancy node metric of the node includes: calculating the mutual information between the electroencephalogram data of the node and its corresponding label as a first mutual information; calculating the mutual information between the electroencephalogram data of the node and the electroencephalogram data of other nodes in the node set as a second mutual information; based on the minimum redundancy maximum correlation criterion, obtaining the redundancy node metric of the node by using the first mutual information and the second mutual information.

[0009] In one implementation manner of the first aspect, the redundancy node metric is shown as the following formula:

[0010]

[0011] where J mRMR (X m ) represents the redundancy node metric of the m-th node in the b-th round of loop, X m and Y m are respectively the electroencephalogram data of the m-th node and its corresponding label, I(X m ,Y m ) is the first mutual information, I(X m ,X i ) is the second mutual information, S a-b is the node set in the b-th round of loop, and a is the number of nodes included in the node set obtained in step S11.

[0012] In one implementation manner of the first aspect, the first mutual information and the second mutual information are obtained based on the KL divergence by a mutual information neural estimator.

[0013] In an implementation of the first aspect, selecting redundant nodes from the node set according to the redundant node metric includes: selecting 1 node with the smallest redundant node metric from the node set as the redundant node; and / or the termination condition is that the number of nodes in the node set is equal to the preset number of nodes.

[0014] In an implementation of the first aspect, the multi-brain collaborative brain-computer interface system is a centralized multi-brain collaborative brain-computer interface system.

[0015] In a second aspect, an embodiment of the present application provides a multi-brain collaborative brain-computer interface system, which includes: a synchronous acquisition module for synchronously acquiring electroencephalogram signals of multiple nodes; a data processing module for preprocessing and / or feature extraction of the electroencephalogram signals of the multiple nodes to obtain processed node data; a node selection module for selecting several nodes from the node set as target nodes according to the processed node data; a data fusion module for fusing the node data of the target nodes to obtain fusion data; and a classifier module for processing the fusion data using a classifier model to obtain a collaborative processing result.

[0016] In a third aspect, an embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the method described in any item of the first aspect of the embodiments of the present application is implemented.

[0017] In a fourth aspect, an embodiment of the present application provides an electronic device, which includes: a memory storing a computer program; a processor communicatively connected to the memory and executing the method described in any item of the first aspect of the embodiments of the present application when calling the computer program.

[0018] An embodiment of the present application provides a node selection method applied to a multi-brain collaborative brain-computer interface system. This node selection method can select the optimal node combination based on information redundancy metrics, thereby reducing the number of nodes while ensuring system performance and saving labor costs.

[0019] An embodiment of the present application also provides a multi-brain collaborative brain-computer interface system, which includes a node selection module that can screen nodes performing the same task in the system, which is beneficial to saving labor costs and computing costs. Description of the Drawings

[0020] Figure 1 It shows a schematic diagram of an application scenario of an embodiment of the present application.

[0021] Figure 2 It shows a flowchart of the node selection method provided by an embodiment of the present application.

[0022] Figure 3 It shows a flowchart for obtaining redundant node metric indicators in an embodiment of the present application.

[0023] Figure 4 It shows a flowchart of the operation of a multi-brain collaborative brain-computer interface system provided in an embodiment of the present application.

[0024] Figure 5 It shows a schematic structural diagram of an electronic device provided in an embodiment of the present application.

[0025] Description of component labels

[0026] 110 Synchronous acquisition module

[0027] 120 Data processing module

[0028] 130 Node selection module

[0029] 140 Data fusion module

[0030] 150 Classifier module

[0031] 500 Electronic device

[0032] 510 Memory

[0033] 520 Processor

[0034] 530 Display

[0035] Steps S21 to S25

[0036] Steps S31 to S33 Detailed implementation manners

[0037] The following uses specific specific examples to illustrate the implementation manners of the present application. Those skilled in the art can easily understand other advantages and effects of the present application from the content disclosed in this specification. The present application can also be implemented or applied through other different specific implementation manners. Various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present application. It should be noted that, without conflict, the following embodiments and the features in the embodiments can be combined with each other.

[0038] It should be noted that the diagrams provided in the following embodiments only illustrate the basic concept of the present application in a schematic manner. Therefore, only the components related to the present application are shown in the diagrams, rather than being drawn according to the number, shape, and size of the components in actual implementation. The type, quantity, and ratio of each component in actual implementation can be arbitrarily changed, and the component layout type may also be more complex.

[0039] Under the paradigm of a collaborative brain-computer interface system, brain signals from multiple nodes in the brain network can be used to achieve collaborative processing, which includes collaborative perception, collaborative control, etc. The collaborative perception task refers to multiple nodes in the brain network simultaneously annotating a static image or a video to more effectively detect and identify relevant information and events (such as specific objects, scenes, people) in the static image and video. Representative brain-computer interface paradigms include rapid serial visual presentation (RSVP) and video detection. The collaborative control task, on the one hand, refers to using multiple brain network nodes that perform the same operation to better decode active brain-computer interfaces, such as the steady-state visual evoked potential (SSVEP) paradigm, the motor imagery (MI) paradigm, and the speech imagery (SI), etc., corresponding control instructions; on the other hand, it refers to assigning different roles and tasks to multiple nodes and achieving complex goals through the mutual cooperation between nodes.

[0040] The performance of a multi-brain collaborative brain-computer interface system will improve with the increase in the number of nodes. In some practical application scenarios, it is often necessary to consume more labor (increase the number of nodes in the brain network system) to achieve more reliable collaborative work. Brain network node selection is a technique for selecting a small number of nodes from multiple nodes within the brain network system. Through this technique, costs can be significantly saved while ensuring the overall performance of the system.

[0041] There are differences in the signal quality of each node in the brain network. On the one hand, this is because the attention and cognitive states of different nodes when performing tasks are different, and on the other hand, it is because the adaptation degrees of different nodes to different brain-computer interface paradigms are different. For example, nodes with higher event-related potential (ERP) induction intensity in the collaborative brain-computer interface system are used to construct the source domain in the domain adaptation algorithm; the response time of the node event-related potential is used to estimate the node confidence, and weights are assigned to the decoding results according to the magnitudes of the confidences of different nodes. Some technical solutions optimize the collaborative brain-computer interface system according to the characteristics of different nodes with different induction intensities and response times, but these technical solutions more emphasize fusing the electroencephalogram signals or features of individual nodes with the best decoding capabilities, without considering the mutual relationships between multiple nodes, such as information redundancy, resulting in the performance of these technical solutions being less than ideal.

[0042] At least for the above problems, the embodiments of the present application provide a node selection method based on information redundancy measurement. Figure 1 Shown is a schematic diagram of the application scenario of the node selection method provided by the embodiments of the present application.Figure 2 Shown is a flowchart of the node selection method provided by an embodiment of the present application. This node selection method can be applied to Figure 1 the node selection module 130 shown. As Figure 2 shown, the node selection method provided by an embodiment of the present application includes the following steps S21 to S25.

[0043] S21, obtain the data of each node in the node set S a where the node set S a includes multiple nodes in the brain networking system.

[0044] S22, traverse all the nodes in the node set S a and calculate the redundant node metric for each node according to the data.

[0045] In some implementation manners, the redundant node metric is constructed based on the minimal redundancy maximal relevance (mRMR) criterion.

[0046] S23, select redundant nodes from the node set S a according to the redundant node metric. For example, in each round of loop, 1 node can be selected from the node set S a as a redundant node, but the present application is not limited thereto.

[0047] S24, delete the redundant nodes from the node set S a Taking the selection of 1 redundant node in each round of loop as an example, after deleting the redundant node in the b-th round of loop, the number of nodes in the node set is a - b.

[0048] S25, repeatedly execute steps S22 to S24 until the nodes in the node set meet the termination condition. Wherein, the termination condition can be set according to actual requirements. For example, it can be that the number of nodes in the node set is equal to the preset number of nodes

[0049] According to the above description, it can be known that an embodiment of the present application provides a node selection method based on information redundancy metric, and this node selection method is a backward search method. Through the node selection method provided by an embodiment of the present application, the number of nodes can be reduced while maintaining the system performance, thereby saving costs. This method can not only improve the system efficiency, but also provide a faster response time and a better user experience.

[0050] In addition, the node selection method provided by an embodiment of the present application does not fuse the electroencephalogram signals or features of multiple single nodes with the optimal decoding ability, but considers the mutual relationship between multiple nodes, so as to select the optimal node combination.

[0051] For any node A in the node set, Figure 3 It shows the flowchart for calculating the redundancy node metric of node A in the embodiments of the present application. As Figure 3 shown, this calculation process includes the following steps S31 to S33.

[0052] S31, calculate the mutual information between the electroencephalogram data of node A and its corresponding label as the first mutual information.

[0053] S32, calculate the mutual information between the electroencephalogram data of node A and the electroencephalogram data of other nodes in the node set as the second mutual information.

[0054] S33, based on the minimum redundancy maximum correlation criterion, use the first mutual information and the second mutual information to obtain the redundancy node metric of node A.

[0055] In a similar way through the above steps S31 to S33, the redundancy metrics of all nodes in the node set can be obtained, and then redundant nodes can be selected from them.

[0056] In some implementation manners, the redundancy node metric is shown as the following formula:

[0057]

[0058] where J mRMR (X m ) represents the redundancy node metric of the m-th node in the b-th round of loop, X m and Y m are respectively the electroencephalogram data of the m-th node and its corresponding label, I(X m ,Y m ) is the first mutual information between the m-th node and its corresponding label, I(X m ,X i ) is the second mutual information between the electroencephalogram data of the m-th node and the i-th node, S a-b is the node set in the b-th round of loop, a is the number of nodes included in the node set obtained in step S21, that is, the number of nodes in the original node set.

[0059] Electroencephalogram data usually has high-dimensional characteristics and the true data distribution is unknown. To address this problem, in some implementation manners, a mutual information neural estimator (MINE) is adopted and the first mutual information and / or the second mutual information is estimated based on the KL divergence.

[0060] Specifically, for two random distributions X and Y, X and Y are respectively X mand Y m or X respectively m and X i Their marginal distributions are p(X) and p(Y) respectively, and the joint distribution is p(X, Y). The mutual information between X and Y is equivalent to the KL divergence between the joint distribution p(X, Y) and the product of the marginal distributions p(X)×p(Y). Therefore, the following formula can be obtained:

[0061]

[0062] where is an arbitrary function from the Cartesian product of the two distributions X and Y to the real number set that satisfies the integrability constraint of the Donsker-Varadhan representation. sup represents the supremum, represents the mathematical expectation.

[0063] Exemplarily, the function T can be approximated using a neural network. Let be the function class of , θ∈Θ is the neural network parameter, and the mutual information estimate is defined as follows:

[0064]

[0065] where k is the number of samples in the random distribution.

[0066] In some implementation manners, selecting redundant nodes from the node set according to the redundant node metric includes: selecting 1 node with the smallest redundant node metric from the node set as the redundant node, but the present application is not limited thereto.

[0067] In some implementation manners, the multi-brain collaborative brain-computer interface system is a centralized multi-brain collaborative brain-computer interface system. In the centralized collaborative brain-computer interface system, electroencephalograms from multiple nodes are fused at the signal level or the feature level, and the fused signals or features are decoded.

[0068] The protection scope of the node selection method provided by the embodiments of the present application is not limited to the execution order of the steps listed in this embodiment. Any solution implemented by adding or subtracting steps of the prior art and replacing steps according to the principle of the present application is included in the protection scope of the present application.

[0069] The embodiments of the present application also provide a multi-brain collaborative brain-computer interface system. Please continue to refer to Figure 1 , the multi-brain collaborative brain-computer interface system provided by the embodiments of the present application may include a synchronous acquisition module 110, a data processing module 120, a node selection module 130, a data fusion module 140, and a classifier module 150.

[0070] The synchronous acquisition module 110 is used to synchronously acquire electroencephalogram signals of multiple nodes, where these nodes receive the same stimulus. In some implementation manners, the stimulus paradigms are, for example, RSVP and video detection for collaborative perception tasks, or SSVEP and motor imagery for collaborative control tasks.

[0071] The data processing module 120 is used to preprocess and / or extract features from the electroencephalogram signals of multiple nodes acquired by the synchronous acquisition module 110, so as to obtain processed node data. In some implementation manners, the data processing module 120 can select corresponding data preprocessing and / or feature extraction schemes for different stimulus paradigms, and these schemes include but are not limited to data interception schemes, filtering schemes or feature extraction schemes.

[0072] The node selection module 130 is used to select several nodes from the node set as target nodes according to the node data processed by the data processing module 120. In some implementation manners, the node selection module 130 can adopt Figure 2 the node selection method based on information redundancy shown in the figure to select the target nodes as the optimal nodes.

[0073] The data fusion module 140 is used to fuse the node data of the target nodes to obtain fused data. Among them, the data fusion module 140 can perform data fusion at the signal level or the feature level for different stimulus paradigms.

[0074] In some implementation manners, the data fusion module 140 can splice signals or features from multiple target nodes based on different dimensions (such as space and time) to obtain fused data.

[0075] In other implementation manners, the data fusion module 140 can average the signals or features from multiple target nodes to obtain fused data.

[0076] The classifier module 150 is used to process the fused data by using a classifier model to obtain a collaborative processing result, and the collaborative processing result is, for example, a collaborative perception result or a collaborative control state. In some implementation manners, the classifier module 150 can adjust the classifier model for different stimulus paradigms. In the training stage, the training data of the classifier model can include the fused target node data. In the testing stage or the actual application stage, the classifier module 150 can use the classifier model to process the fused data of the target nodes to obtain a collaborative processing result.

[0077] Next, a specific example will be used to introduce the multi-brain collaborative brain-computer interface system provided by the embodiments of the present application in detail. It should be noted that this example is only used to illustrate the working process of the multi-brain collaborative brain-computer interface system, and does not limit the protection scope of the present application. As Figure 4As shown, this example is implemented based on the rapid serial visual presentation paradigm of the collaborative perception task of the collaborative brain-computer interface system. Among them, rapid serial visual presentation is the process of sequentially displaying images at a high presentation rate of multiple images per second at the same spatial position. Based on the brain-computer interface system of rapid serial visual presentation, the target images and non-target images in the rapidly presented image sequence can be detected by decoding electroencephalograms.

[0078] As Figure 4 shown, the brain network system in this example includes n nodes, and n users in the system simultaneously gaze at the rapidly presented image sequence. Among them, the image sequence includes target pictures and non-target pictures, and the ratio between the two can be set according to actual needs.

[0079] In this example, the synchronous acquisition module 110 is used to acquire the electroencephalogram data of n users, which includes electrodes / electrode caps and a signal synchronous acquisition unit. The data processing module 120 is used to receive the electroencephalogram data acquired by the synchronous acquisition module 110, intercept the signal for a certain number of milliseconds after the event label, and send the signal into a FIR band-pass filter to obtain the processed node data.

[0080] The node selection module 110 uses a node selection method based on information redundancy to select target nodes. Specifically, the node set is denoted as s n . Starting from s n , a loop is carried out. In each round of the loop, a redundant node is deleted until the number of nodes in the node set is reduced to the preset number of nodes Among them, the metric index of the redundant node is constructed based on the minimum redundancy maximum correlation criterion. In addition, the node selection module 110 estimates the first mutual information through the first fully connected neural network MINE1 and estimates the second mutual information by using the second fully connected neural network MINE2.

[0081] The data fusion module 140 is used to fuse the data of the target nodes at the signal level. The fusion method it adopts can be to splice the signals or features from multiple target nodes based on different dimensions (such as space and time) to obtain the fusion data, or to average the signals or features from multiple target nodes to obtain the fusion data.

[0082] The classifier module 150 is used to train the classifier model with the fused target node data during the training phase. Among them, the classifier can adopt the Hierarchical Discriminant Component Analysis (HDCA) algorithm. The HDCA algorithm uses linear discriminant analysis to divide the data of a single trial according to time windows, performs spatial and temporal projections on each time window, and finally integrates the decision values in time to decode the electroencephalogram to obtain the target detection result (target image / non-target image).

[0083] According to the above description, an embodiment of the present application provides a multi-brain collaborative brain-computer interface system with a node selection module. This system can screen out the subset of optimal nodes (target nodes) that execute the same task in the system through the node selection module. In addition, the node selection module has the characteristics of plug-and-play and can be easily integrated into different multi-brain collaborative brain-computer interface systems without complex custom development. This provides flexibility for various application scenarios, including collaborative perception tasks and collaborative control tasks, etc. Through the multi-brain collaborative brain-computer interface system provided by the embodiment of the present application, users can interact with the external environment more efficiently and achieve goals such as collaborative perception and collaborative control. This not only broadens the application field of brain-computer interface technology but also provides important support for future multi-brain collaboration, providing a more efficient and economical solution for multi-brain collaboration research and applications.

[0084] In several embodiments provided by the present application, it should be understood that the disclosed system, device or method can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of modules / units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple modules or units can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection between each other can be through some interfaces, and the indirect coupling or communication connection of devices or modules or units can be in electrical, mechanical or other forms.

[0085] The modules / units described as separate components may or may not be physically separated. The components displayed as modules / units may or may not be physical modules, that is, they can be located in one place or distributed to multiple network units. Some or all of the modules / units can be selected according to actual needs to achieve the purpose of the embodiments of the present application. For example, in each embodiment of the present application, the various functional modules / units can be integrated in a processing module, or each module / unit can exist physically alone, or two or more modules / units can be integrated in one module / unit.

[0086] Those of ordinary skill in the art should also be able to further realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, computer software, or a combination of the two. To clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described according to functions in the above description. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0087] The embodiments of this application also provide a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the node selection method provided by the embodiments of this application. Those of ordinary skill in the art can understand that all or part of the steps in the methods of the above embodiments can be completed by instructing a processor through a program. The described program can be stored in a computer-readable storage medium. The storage medium is a non-transitory medium, such as random access memory, read-only memory, flash memory, hard disk, solid-state drive, magnetic tape, floppy disk, optical disc, and any combination thereof. The above storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more integrated available media. The available medium can be a magnetic medium (such as a floppy disk, hard disk, magnetic tape), an optical medium (such as a digital video disc (DVD)), or a semiconductor medium (such as a solid-state disk (SSD)), etc.

[0088] The embodiments of this application can also provide an electronic device. Figure 5 Shown is a schematic structural diagram of an electronic device 500 in an embodiment of this application. As Figure 5 shown, in this embodiment, the electronic device 500 includes a memory 510 and a processor 520.

[0089] The memory 510 is used to store a computer program. In some possible implementation manners, the memory 510 may include various media that can store program codes, such as ROM, RAM, magnetic disk, USB flash drive, memory card, or optical disc.

[0090] In the embodiments of the present application, the memory 510 may include a computer system-readable medium in the form of volatile memory, such as RAM and / or cache memory. The electronic device 500 may further include other removable / non-removable, volatile / non-volatile computer system storage media. The memory 510 may include at least one program product having a set (e.g., at least one) of program modules configured to perform the functions of the embodiments of the present application.

[0091] The processor 520 is connected to the memory 510 and is configured to execute the computer program stored in the memory 510, so that the electronic device 500 executes the node selection method.

[0092] Exemplarily, the processor 520 may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc. In other embodiments, the processor 520 may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0093] In some implementations, the electronic device 500 provided by the embodiments of the present application may further include a display 530. The display 530 is communicatively connected to the memory 510 and the processor 520 and is configured to display a relevant graphical user interface (GUI) of the node selection method.

[0094] In the embodiments of the present application, the display 530 may include a display screen (display panel). In some implementations, the display panel may be configured in the form of a liquid crystal display (LCD), an organic light-emitting diode (OLED), etc. In addition, the display 530 may also be a touch panel (touch screen, touch-sensitive screen), and the touch panel may include a display screen and a touch-sensitive surface. When the touch-sensitive surface detects a touch operation on or near it, it is transmitted to the processor 520 to determine the type of touch event, and then the processor 520 provides a corresponding visual output on the display device according to the type of touch event.

[0095] The descriptions of the processes or structures corresponding to the above-mentioned respective drawings each have their own focuses. For parts not detailed in a certain process or structure, reference may be made to the relevant descriptions of other processes or structures.

[0096] The above embodiments are only illustrative of the principles and effects of the present application and are not intended to limit the present application. Any person familiar with this technology can modify or change the above embodiments without departing from the spirit and scope of the present application. Therefore, all equivalent modifications or changes made by those with ordinary knowledge in the technical field without departing from the spirit and technical ideas disclosed in the present application should still be covered by the claims of the present application.

Claims

1. A node selection method applied to a multi-brain collaborative brain-computer interface system, characterized in that The node selection method includes: S11. Obtain the data of each node in the node set, where the node set includes multiple nodes in the brain networking system; S12. Traverse all nodes in the node set and calculate the redundancy node metric for each node according to the data; S13. Select redundant nodes from the node set according to the redundancy node metric; S14. Delete the redundant nodes from the node set; S15. Repeat steps S12 to S14 until the nodes in the node set meet the termination condition.

2. The node selection method according to claim 1, wherein The redundancy node metric is constructed based on the minimum redundancy maximum relevance criterion.

3. The node selection method according to claim 2, wherein For a node in the node set, calculating the redundancy node metric of this node includes: Calculating the mutual information between the electroencephalogram data of this node and its corresponding label as the first mutual information; Calculating the mutual information between the electroencephalogram data of this node and the electroencephalogram data of other nodes in the node set as the second mutual information; Based on the minimum redundancy maximum relevance criterion, use the first mutual information and the second mutual information to obtain the redundancy node metric of this node.

4. The node selection method according to claim 3, wherein The redundancy node metric is shown as the following formula: Among them, J mRMR (X m ) represents the redundant node metric of the m-th node in the b-th round of loop, X m and Y m are the electroencephalogram data of the m-th node and its corresponding label respectively, I(X m , Y m ) is the first mutual information, I(X m , X i ) is the second mutual information, S a-b is the set of nodes in the b-th round of loop, and a is the number of nodes included in the set of nodes obtained in step S11.

5. The node selection method according to claim 2, wherein The first mutual information and the second mutual information are obtained based on the KL divergence through a mutual information neural estimator.

6. The node selection method according to claim 1, wherein: Selecting redundant nodes from the node set according to the redundancy node metric includes: selecting 1 node with the smallest redundancy node metric from the node set as the redundant node; and / or The termination condition is that the number of nodes in the node set is equal to the preset number of nodes.

7. The node selection method according to claim 1, characterized in that The multi-brain collaborative brain-computer interface system is a centralized multi-brain collaborative brain-computer interface system.

8. A multi-brain collaborative brain-computer interface system, characterized in that, The multi-brain collaborative brain-computer interface system includes: A synchronous acquisition module, configured to synchronously acquire electroencephalogram signals of multiple nodes; A data processing module, configured to preprocess and / or extract features from the electroencephalogram signals of the multiple nodes to obtain processed node data; A node selection module, configured to select several nodes from the node set as target nodes according to the processed node data; A data fusion module, configured to fuse the node data of the target nodes to obtain fusion data; A classifier module, configured to process the fusion data using a classifier model to obtain a collaborative processing result.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method according to any one of claims 1 to 7.

10. An electronic device, characterized in that, The electronic device includes: A memory, storing a computer program; A processor, communicatively connected to the memory, and when calling the computer program, executes the method according to any one of claims 1 to 7.

Images