Motor imagery brain-computer interface user classification method and device, equipment and storage medium

Abstract

The invention discloses a motor imagery brain-computer interface user classification method, device and equipment and a storage medium, and the method comprises the steps: carrying out the decomposition processing of a target brain wave signal of a user through employing an independent component analysis algorithm, obtaining a plurality of initial independent components, respectively extracting the feature data of each initial independent component, and carrying out the classification of a target brain wave signal based on the feature data; and determining a first target independent component corresponding to the correlation desynchronization of the opposite-side event from the initial independent components, determining a second target independent component corresponding to the correlation synchronization of the same-side event, and finally determining a motor imagery brain-computer interface user classification result. According to the scheme, the users are classified through the independent component related brain dynamics modeling and the multi-dimensional neural indexes obtained based on the feature data, so that the classification accuracy and reliability are improved; besides, automatic screening of independent components is carried out in combination with various feature data, the subjectivity problem caused by manual intervention is avoided, and the applicability and expandability of the classification method are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of artificial intelligence, and in particular to a motor imagery brain-computer interface user classification method and device, equipment and a storage medium. BACKGROUND

[0002] A brain-computer interface (BCI) is an innovative control technology that enables human-computer interaction without peripheral nerve and muscle activity. Motor imagery (MI)-BCI is an important branch of it, which enables users to interact with the outside world by only imagining limb movement. However, there are significant individual differences in the motor imagery ability of BCI users, and some users cannot effectively produce distinguishable electroencephalogram patterns, resulting in poor BCI performance, which is referred to as "BCI illiteracy".

[0003] Currently, existing BCI illiteracy discrimination methods usually classify users into "excellent users" and "BCI illiterates" based on classification accuracy. However, a large number of studies have shown that about 70% of users are in the "intermediate user" interval of 60%-80% classification accuracy, and their performance is better than random level, but still cannot achieve reliable BCI control. In addition, existing studies have also found that such users show advantages in unilateral hand motor imagery, while the performance of the other side is poor. This unilateral advantage limits the training and classification effect of existing BCI models.

[0004] Therefore, the existing motor imagery brain-computer interface user classification method has the problem of low classification accuracy. SUMMARY

[0005] The present application aims to at least solve the technical problems existing in the prior art. To this end, the first aspect of the present application provides a motor imagery brain-computer interface user classification method, which comprises: For the left hand motor imagery task and the right hand motor imagery task of the user, an independent component analysis algorithm is used to decompose and process the target electroencephalogram signal of the user, to obtain a plurality of initial independent components; For each initial independent component, feature extraction processing is performed on the initial independent component to obtain feature data; wherein the feature data includes a dipole position corresponding to the initial independent component, a relative mu / beta power, a dynamic power change, a significant event-related desynchronization event map, and a significant event-related synchronization map; Based on the feature data, a first target independent component corresponding to a contralateral event-related desynchronization is determined from the initial independent components, and a second target independent component corresponding to an ipsilateral event-related synchronization is determined; The first target independent component, the second target independent component, the feature data, and a preset motor imagery brain-computer interface performance index are used to determine a motor imagery brain-computer interface user classification result, wherein the motor imagery brain-computer interface user classification result includes a first classification result for a left-hand motor imagery task and a second classification result for a right-hand motor imagery task.

[0006] In a possible implementation, the generation process of the dipole position includes: The initial independent component is subjected to equivalent current dipole fitting processing to generate a dipole position corresponding to the initial independent component.

[0007] In a possible implementation, the generation process of the relative mu / beta power includes: The neural power spectral density corresponding to the initial independent component is obtained. The neural power spectral density is subjected to parameterization processing to obtain at least one periodic component and an aperiodic component. After removing the interference of the aperiodic component, the component with the largest amplitude in the periodic component is extracted as the relative mu / beta power.

[0008] In a possible implementation, the generation process of the dynamic power change includes: The initial independent component is subjected to time-frequency decomposition processing by using a preset wavelet transform algorithm to obtain a decomposition result. The dynamic power change is calculated based on the power value at a preset time and the power mean value of a preset reference interval within a preset time window.

[0009] In a possible implementation, the generation process of the significant event-related desynchronization event map and the significant event-related synchronization map includes: Random sampling is performed in a preset reference interval and a preset task interval to calculate a target statistic and a target distribution. The significance level of each voxel unit in the brain is estimated based on the target statistic and the target distribution. Based on the significance level, the connected region corresponding to adjacent significant voxel units is determined. Based on the connected region, the significant event-related desynchronization event map and the significant event-related synchronization map are determined.

[0010] In a possible implementation, the method further includes: The acquired original brain wave signal is obtained, wherein the original brain wave signal is a signal related to motor imagery. The original brain wave signal is subjected to band-pass filtering processing to extract an oscillation signal in the mu frequency band and the beta frequency band. extract a time domain segment in a preset time period from the segmented oscillation signal; wherein, the preset time period comprises a preset reference interval and a preset task interval; perform artifact removal processing on the time domain segment to obtain a target brain wave signal.

[0011] In a possible implementation, the method further comprises: determining a corresponding target training strategy based on the first classification result and the second classification result; training the preset brain-computer interface model based on the target training strategy to obtain a training result.

[0012] The second aspect of the application provides a motor imagery brain-computer interface user classification device, which comprises: a decomposition module configured to perform decomposition processing on a target brain wave signal of a user by using an independent component analysis algorithm for a left-hand motor imagery task and a right-hand motor imagery task of the user, to obtain a plurality of initial independent components; an extraction module configured to perform feature extraction processing on the initial independent components to obtain feature data, wherein the feature data comprises a dipole position, a relative mu / beta power, a dynamic power change, a significant event-related desynchronization event map, and a significant event-related synchronization map corresponding to the initial independent components; a first determination module configured to determine a first target independent component corresponding to a contralateral event-related desynchronization and a second target independent component corresponding to an ipsilateral event-related synchronization from the initial independent components based on the feature data; a second determination module configured to determine a motor imagery brain-computer interface user classification result based on the first target independent component, the second target independent component, the feature data, and a preset motor imagery brain-computer interface performance indicator, wherein the motor imagery brain-computer interface user classification result comprises a first classification result for the left-hand motor imagery task and a second classification result for the right-hand motor imagery task.

[0013] In a possible implementation, the above motor imagery brain-computer interface user classification device is further configured to: The generation process of the dipole position comprises: performing equivalent current dipole fitting processing on the initial independent components to generate a dipole position corresponding to the initial independent components.

[0014] In a possible implementation, the above motor imagery brain-computer interface user classification device is further configured to: obtain a neural power spectral density corresponding to the initial independent components; perform parameterization processing on the neural power spectral density to obtain at least one periodic component and a non-periodic component; After removing the interference of the non-periodic component, the component with the largest amplitude in the periodic component is extracted as the relative mu / beta power.

[0015] In a possible implementation, the motion imagination brain-computer interface user classification apparatus is further configured to: perform time-frequency decomposition processing on the initial independent components by using a preset wavelet transform algorithm to obtain a decomposition result. In a preset time window, a dynamic power change is calculated based on a power value at a preset time and a power mean value of a preset reference interval.

[0016] In a possible implementation, the motion imagination brain-computer interface user classification apparatus is further configured to: random sampling is performed in a preset reference interval and a preset task interval, and a target statistic and a target distribution are calculated. Based on the target statistic and the target distribution, a significance level corresponding to each voxel unit in the brain is estimated. Based on the significance level, a connected region corresponding to adjacent significant voxel units is determined. Based on the connected region, a significant event-related desynchronization event map and a significant event-related synchronization map are determined.

[0017] In a possible implementation, the motion imagination brain-computer interface user classification apparatus is further configured to: obtain an original electroencephalogram signal collected; wherein the original electroencephalogram signal is a signal related to motion imagination; perform band-pass filtering processing on the original electroencephalogram signal to extract oscillation signals in mu and beta frequency bands; segmentally extract time domain segments in a preset time period from the oscillation signals; wherein the preset time period includes a preset reference interval and a preset task interval; perform artifact removal processing on the time domain segments to obtain a target electroencephalogram signal.

[0018] In a possible implementation, the motion imagination brain-computer interface user classification apparatus is further configured to: based on the first classification result and the second classification result, determine a corresponding target training strategy; based on the target training strategy, train a preset brain-computer interface model to obtain a training result.

[0019] A third aspect of the present application provides an electronic device, which includes a processor and a memory, the memory storing at least one instruction, at least one program, a code set or an instruction set, the at least one instruction, the at least one program, the code set or the instruction set being loaded and executed by the processor to implement the motion imagination brain-computer interface user classification method according to the first aspect.

[0020] The fourth aspect of the present application provides a computer readable storage medium, wherein the storage medium stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by a processor to implement the motor imagery brain-computer interface user classification method according to the first aspect.

[0021] The embodiments of the present application have the following beneficial effects: The motor imagery brain-computer interface user classification method provided by the embodiments of the present application comprises: for left-hand motor imagery tasks and right-hand motor imagery tasks of a user, respectively adopting an independent component analysis algorithm to perform decomposition processing on target brain wave signals of the user to obtain a plurality of initial independent components, for each initial independent component, performing feature extraction processing on the initial independent component to obtain feature data, wherein the feature data comprises a dipole position corresponding to the initial independent component, a relative mu / beta power, a dynamic power change, a significant event-related desynchronization event diagram and a significant event-related synchronization diagram, based on the feature data, determining a first target independent component corresponding to a contralateral event-related desynchronization from the initial independent components, and determining a second target independent component corresponding to an ipsilateral event-related synchronization, and based on the first target independent component, the second target independent component, the feature data and a preset motor imagery brain-computer interface performance index, determining a motor imagery brain-computer interface user classification result. The present scheme builds a brain dynamics model based on independent components, and classifies the user based on the multi-dimensional neural indicators obtained from the feature data, thereby improving the accuracy and reliability of the classification. In addition, the automatic screening of the independent components by combining a plurality of feature data avoids the subjective problems caused by manual intervention, thereby improving the applicability and scalability of the classification method. BRIEF DESCRIPTION OF DRAWINGS

[0022] Figure 1 A block diagram of a computer device provided by the embodiments of the present application is provided; Figure 2 A step flowchart of a motor imagery brain-computer interface user classification method provided by the embodiments of the present application is provided; Figure 3 A step flowchart of a method for generating target brain wave signals provided by the embodiments of the present application is provided; Figure 4 A step flowchart of a method for extracting a relative mu / beta power provided by the embodiments of the present application is provided; Figure 5 A step flowchart of a method for extracting a dynamic power change provided by the embodiments of the present application is provided; Figure 6 A step flowchart of a method for extracting a significant event-related desynchronization event diagram and a significant event-related synchronization diagram provided by the embodiments of the present application is provided; Figure 7 A structural block diagram of a motion imagination brain-computer interface user classification device is provided for the embodiments of the present application. DETAILED DESCRIPTION

[0023] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of protection of the present application.

[0024] A brain-computer interface (BCI) is an innovative control technology that enables human-computer interaction without peripheral nerve and muscle activity. Motor imagery (MI)-BCI is an important branch of it, which enables users to interact with the outside world by imagining limb movement only. However, there are significant individual differences in the motor imagery ability of BCI users, and some users cannot effectively produce distinguishable electroencephalogram patterns, resulting in poor BCI performance, which is called "BCI illiteracy". At present, the existing BCI illiteracy discrimination methods usually classify users into "excellent users" and "BCI illiterates" based on classification accuracy. However, a large number of studies have shown that about 70% of users are in the "intermediate user" interval of 60%-80% classification accuracy, and their performance is better than random level, but still cannot achieve reliable BCI control. In addition, existing studies have also found that such users show advantages in unilateral hand motor imagery, while the performance of the other side is poor. This unilateral advantage limits the training and classification effect of the existing BCI model.

[0025] In addition, Event-Related Desynchronization (ERD) and Event-Related Synchronization (ERS) reflect the task-related decrease and increase of EEG power, respectively, and are important indicators of cortical activation and deactivation. In actual and imagined movement, ERD usually appears in the mu (8-13 Hz) and beta (13-30 Hz) bands of the contralateral motor area, while ERS may appear in the ipsilateral area related to movement stop or attention shift, and is widely considered a reliable indicator of MI involvement. However, most existing studies rely on channel-level analysis, and some studies have introduced an entropy method based on vector quantization patterns to estimate ERD / ERS. In addition, researchers have attempted to combine sensory motor rhythm power with phase synchronization in the mu and beta bands to improve assessment accuracy, and have further demonstrated that directional connectivity patterns in the mu and beta bands are closely related to individual MI ability.

[0026] In recent years, some studies have attempted to apply Independent Component Analysis (ICA) to source separation of EEG signals and extract Independent Components (ICs) related to motor imagery by combining time-frequency features and power spectrum analysis. Studies have extracted the topological map, time-frequency features, and dipole location of each IC, and used clustering methods to divide them into multiple groups, with ERD-related groups selected through visual inspection. Some studies have ranked ICs according to their discriminability in left-hand imagined movement, right-hand imagined movement, and resting state, and selected several ICs with the best classification performance. Some studies have used single-trial coherence based on stimulation or response to divide ICs into sensory motor, sensory, motor, or unspecified groups, and have proposed multiple criteria for selecting sensory motor ICs: central scalp distribution, dipole positioning in the central pre- and post-central gyrus, and mu band suppression.

[0027] However, the existing motor imagery BCI evaluation mostly takes the overall classification accuracy as the standard, and fails to fully consider the neural characteristics of users in the intermediate performance interval (60%-80% accuracy), resulting in the inability to improve the training strategy specifically. Traditional ERD / ERS evaluation mostly relies on the electroencephalogram channel level and is affected by volume conduction and other interference, making it difficult to accurately reflect the real neural source activity of the cortex and affecting the accuracy of the evaluation. Most existing technologies fail to effectively distinguish between periodic neural oscillations and non-periodic background signals, resulting in interference with the ERD / ERS analysis results and affecting the extraction and classification of user characteristics. The existing ICA method mostly relies on manual experience or simple statistical indicators for IC selection, lacks an automated analysis process combining spatial positioning, ERD / ERS kinetic characteristics, and power spectrum estimation, and thus makes it difficult to stably extract independent components directly related to motor imagery, limiting the universality and interpretability of the method.

[0028] Based on this, the application provides a motor imagery brain-computer interface user classification method, which comprises: for the left hand motor imagery task and the right hand motor imagery task of a user, respectively adopting an independent component analysis algorithm to perform decomposition processing on the target electroencephalogram signal of the user, to obtain a plurality of initial independent components, for each initial independent component, performing feature extraction processing on the initial independent component to obtain feature data, wherein the feature data comprises a dipole position corresponding to the initial independent component, a relative mu / beta power, a dynamic power change, a significant event-related desynchronization event diagram, and a significant event-related synchronization diagram, based on the feature data, determining a first target independent component corresponding to a contralateral event-related desynchronization from the initial independent components, and determining a second target independent component corresponding to an ipsilateral event-related synchronization, and based on the first target independent component, the second target independent component, the feature data, and a preset motor imagery brain-computer interface performance indicator, determining a motor imagery brain-computer interface user classification result. The present scheme improves the accuracy and reliability of classification by modeling the brain dynamics related to independent components and classifying users based on the multi-dimensional neural indicators obtained from the feature data. In addition, by combining various feature data for automatic screening of independent components, the subjective problems caused by manual intervention are avoided, and the applicability and scalability of the classification method are improved.

[0029] Hereinafter, the terms "first" and "second" are only used for descriptive purposes, and cannot be understood to indicate or imply relative importance or implicitly indicate the number of indicated technical features. Therefore, the features defined with "first" and "second" can explicitly or implicitly include one or more of the features. In the description of the embodiments of the present disclosure, unless otherwise stated, the meaning of "a plurality of" is two or more. In addition, the use of "based on" or "according to" means openness and inclusiveness, because the process, step, calculation or other action "based on" or "according to" one or more stated conditions or values can be based on additional conditions or values beyond those stated in practice.

[0030] The motion imagination brain-computer interface user classification method provided by the application can be applied to a computer device (electronic device). The computer device can be a server or a terminal. The server can be a single server or a server cluster composed of multiple servers. The embodiments of the application do not make specific limitations on this. The terminal can be, but is not limited to, various personal computers, notebook computers, smart phones, tablet computers and portable wearable devices.

[0031] Taking the computer device as a server as an example, Figure 1 a block diagram of a server is shown, as Figure 1 shown, the server can include a processor and a memory connected through a system bus. The processor of the server is used to provide computing and control capabilities. The memory of the server includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The computer program is executed by the processor to implement a motion imagination brain-computer interface user classification method.

[0032] Those skilled in the art can understand, Figure 1 the structure shown in the figure, only the block diagram of part of the structure related to the scheme of the application, and does not constitute a limitation on the server to which the scheme of the application is applied. Alternatively, the server can include more or fewer components than those shown in the figure, or combine certain components, or have a different component arrangement.

[0033] It should be noted that the execution subject of the embodiments of the application can be a computer device or a motion imagination brain-computer interface user classification device. The following method embodiments are described with the computer device as the execution subject.

[0034] Figure 2 A step flowchart of a motion imagination brain-computer interface user classification method provided by the embodiments of the application is shown in Figure 2 The method includes the following steps: Step 202, for the left hand motion imagination task and the right hand motion imagination task of the user, respectively, an independent component analysis algorithm is used to decompose and process the target brain wave signal of the user, to obtain a plurality of initial independent components.

[0035] The target brain wave signal of the user is obtained by preprocessing the original brain wave signal. In some optional embodiments, as Figure 3 shown, Figure 3 A step flowchart for generating a target brain wave signal provided by the embodiments of the application includes: Step 302, obtaining the original brain wave signal collected.

[0036] Step 304, performing band-pass filtering processing on the original brain wave signal to extract the oscillation signal in the mu frequency band and the beta frequency band.

[0037] Step 306, segmenting and extracting the time domain segment in the preset time period from the oscillation signal.

[0038] Step 308, performing artifact removal processing on the time domain segment to obtain the target brain wave signal.

[0039] The original brain wave signal is a signal related to motor imagery, and the corresponding signal collected by the channel related to motor imagery such as frontal lobe, central parietal lobe, and posterior parietal lobe can be selected.

[0040] Then, the original brain wave signal can be subjected to band-pass filtering processing to extract the oscillation signal in the mu frequency band and the beta frequency band, and the time domain segment in the preset time period corresponding to each test is segmented and extracted, wherein the preset time period includes a preset reference interval and a preset task interval. Exemplarily, the preset time period can be -1 second to +4 seconds, the preset reference interval can be -1 second to 0 second, and the preset task interval can be 0 second to +4 second.

[0041] Finally, the time domain segment can be subjected to artifact removal processing to obtain the target brain wave signal. Optionally, the artifact removal processing process can include but is not limited to artifact rejection, trend correction, and outlier detection. In this way, the signal quality of the obtained target brain wave signal can be ensured.

[0042] Thus, for the left-hand motor imagery task and the right-hand motor imagery task of the user, the independent component analysis algorithm can be used to decompose and process the target brain wave signal of the user to obtain a plurality of initial independent components. Specifically, the independent component analysis algorithm can separate independent neural activities from highly overlapping scalp signals through statistical methods. Optionally, an algorithm based on maximum information transmission can be used to maximize the entropy of the signal to achieve signal separation. This algorithm calculates the demixing matrix W based on the information maximization principle to recover the potential source signal from the recorded target brain wave signal.

[0043] Then, for each user's two different tasks, i.e., the left-hand motor imagery task and the right-hand motor imagery task, the target brain wave signal of the selected preset number of channels is decomposed and processed to obtain a corresponding preset number of independent components. Optionally, the preset number is determined by the number of channels of the electroencephalogram acquisition device. Exemplarily, the preset number can be 30.

[0044] Step 204, for each initial independent component, performing feature extraction processing on the initial independent component to obtain feature data.

[0045] The feature data can include a dipole position corresponding to the initial independent component, a relative mu / beta power, a dynamic power change, a significant event related desynchronization event graph, and a significant event related synchronization graph.

[0046] In some optional embodiments, when extracting the dipole position, an equivalent current dipole fitting processing can be performed on the initial independent component to generate the dipole position corresponding to the initial independent component. Optionally, the DIPFIT2 plug-in can be used to perform the equivalent current dipole fitting processing on the initial independent component. The specific implementation process of the plug-in can use the prior art, which will not be described here. Thus, the dipole position corresponding to the initial independent component can be generated. By combining the dipole positioning technology, the automatic screening and positioning of the spatially independent motor imagery related neural sources are realized, the interference of volume conduction is weakened from the source, and the spatial resolution and interpretability of the signal are significantly improved.

[0047] In addition, the initial independent component with a residual variance greater than 15% can be removed, because these initial independent components are unlikely to come from a single neural source. In addition, the initial independent component with local activation limited to a single electrode (such as an electromyographic artifact, which is a high-frequency electrical activity in the form of a spike) can also be removed.

[0048] In some optional embodiments, when extracting the relative mu / beta power, as shown in Figure 4 , a step flow chart for extracting the relative mu / beta power provided by the embodiments of the present application includes: Figure 4 Step 402, obtaining a neural power spectral density corresponding to the initial independent component.

[0049] Step 404, performing parameterization processing on the neural power spectral density to obtain at least one periodic component and aperiodic component.

[0050] Step 406, after removing the interference of the aperiodic component, extracting the component with the largest amplitude in the periodic component as the relative mu / beta power.

[0051] ​The neural power spectral density can be parameterized by using a fitting oscillations & one-over-f (FOOOF) algorithm, at least one periodic component and aperiodic component are obtained, the oscillation peak value is extracted by identifying the spectral peak exceeding the modeled aperiodic background, and a relative narrowband power estimation without aperiodic interference is obtained to remove the interference of the aperiodic component.

[0052] Therefore, after removing the interference of the aperiodic component, the component with the largest amplitude in the extracted periodic component can be used as the relative mu / beta power. By effectively separating the periodic neural oscillation component from the aperiodic background signal, the extracted feature data is more real and stable, and the problem that the traditional narrowband analysis is easily disturbed by the background signal is solved.

[0053] In some optional embodiments, when extracting the dynamic power change, as shown in Figure 5 , a step flow chart for extracting the dynamic power change provided by the embodiment of the present application comprises: Figure 5 , a step flow chart for extracting the dynamic power change provided by the embodiment of the present application comprises: Step 502, a preset wavelet transform algorithm is used to perform time-frequency decomposition processing on the initial independent component to obtain a decomposition result.

[0054] Step 504, within a preset time window, a dynamic power change is calculated based on a power value at a preset time and a power mean value of a preset reference interval.

[0055] The dynamic power change can be denoted as ERD%, ERS%, the relative ERD power can be defined as the proportion of the power decrease relative to the preset reference interval, and the relative ERS power can be defined as the proportion of the power increase relative to the preset reference interval.

[0056] Optionally, the preset wavelet transform algorithm can be a Morlet wavelet transform algorithm, which can be set to 3 periods and use a Hanning window, so as to perform time-frequency decomposition processing on the initial independent component time series to obtain a decomposition result. The specific time-frequency decomposition processing process can refer to the prior art, and will not be described here.

[0057] Therefore, within the preset time window, a dynamic power change can be calculated based on a power value at a preset time and a power mean value of a preset reference interval. Optionally, ERD%, ERS% is equal to the power value at the preset time t minus the power mean value of the preset reference interval, and then divided by the power mean value of the preset reference interval, and then multiplied by 100%.

[0058] In some optional embodiments, when extracting the significant event related desynchronization event map and the significant event related synchronization map, as shown in Figure 6 , a step flow chart for extracting the significant event related desynchronization event map and the significant event related synchronization map provided by the embodiment of the present application comprises:Figure 6 A step flow chart for extracting a significant event related desynchronization event graph and a significant event related synchronization graph is provided for an embodiment of the present application, comprising: Step 602, random sampling is performed in the preset reference interval and the preset task interval, and a target statistic and a target distribution are calculated.

[0059] Step 604, based on the target statistic and the target distribution, the significance level corresponding to each voxel unit in the brain is estimated.

[0060] Step 606, based on the significance level, the connected region corresponding to adjacent significant voxel units is determined.

[0061] Step 608, based on the connected region, the significant event related desynchronization event graph and the significant event related synchronization graph are determined.

[0062] Wherein, after a large number of random sampling in the preset reference interval and the preset task interval, the target statistic and the target distribution can be calculated. The target statistic is a pseudo-t statistic, and the target distribution is the distribution of pseudo-t values.

[0063] Then, based on the target statistic and the target distribution, the significance level corresponding to each voxel unit in the brain can be estimated. For each significant voxel unit, its adjacent voxel connectivity in the horizontal direction (representing time) and the vertical direction (representing frequency) is further analyzed, and adjacent significant voxel units are classified into connected regions.

[0064] Then, based on the connected region, the average power change value of each 4-connected region can be calculated. For the region with an average value less than zero, it is defined as a significant event related desynchronization event region, i.e. a significant ERD region; for the region with an average value greater than zero, it is defined as a significant event related synchronization region, i.e. a significant ERS region.

[0065] The area of each significant ERD region or significant ERS region is represented by the number of significant voxel units it contains. For each initial independent component, the largest connected region can be selected as the representative significant ERD region or significant ERS region, i.e. the final significant event related desynchronization event graph and the significant event related synchronization graph are determined.

[0066] Step 206, based on the feature data, a first target independent component corresponding to the contralateral event related desynchronization is determined from the initial independent components, and a second target independent component corresponding to the ipsilateral event related synchronization is determined.

[0067] Wherein, in order to identify the independent component related to hand motor imagery, that is, the independent component capable of presenting contralateral ERD or ipsilateral ERS, the first target independent component corresponding to contralateral event-related desynchronization and the second target independent component corresponding to ipsilateral event-related synchronization can be determined from the initial independent component based on the feature data, so as to realize automatic screening of the initial independent component.

[0068] Optionally, the selection rule of the first target independent component corresponding to the contralateral event-related desynchronization comprises: 1. The dipole position corresponding to the initial independent component must be located in the brain area related to contralateral motor imagery, including contralateral somatosensory cortex (BA1-3 area), primary motor cortex (BA4 area), supramarginal gyrus (BA5, BA7 area), premotor area and supplementary motor area (BA6 area), anterior cingulate cortex (BA24, BA32 area), fusiform gyrus body selection area (BA37 area), inferior parietal lobule (BA40 area), insular cap area (BA44, BA45 area) and middle frontal gyrus (BA9, BA46 area), wherein the dipole position is usually represented by three-dimensional coordinates, and different brain areas also have corresponding three-dimensional coordinate ranges, so that the first independent component candidate set located in the brain area related to motor imagery can be determined.

[0069] 2. In the above-mentioned first independent component candidate set, the first three independent components with relatively high mu / beta power are selected to form a second independent component candidate set.

[0070] 3. In the above-mentioned second independent component candidate set, the independent component with the largest 4-connected area in the significant event-related desynchronization event map is selected as the first target independent component. If there is no corresponding significant event-related desynchronization event map in the above-mentioned second independent component candidate set, the independent component with the highest mu / beta power is selected as the first target independent component.

[0071] Optionally, the selection rule of the second target independent component corresponding to the ipsilateral event-related synchronization comprises: 1. The dipole position corresponding to the initial independent component must be located in the brain area related to ipsilateral motor imagery, to obtain a third independent component candidate set located in the brain area related to ipsilateral motor imagery.

[0072] 2. In the above-mentioned third independent component candidate set, the independent component with a significant oscillation peak in the mu / beta frequency range is selected as a fourth independent component candidate set. Wherein, the independent component with a significant oscillation peak has at least one periodic component.

[0073] 3. In the fourth independent component candidate set, the independent component with the largest 4-connected region area in the significant event-related synchronization graph is selected as the second target independent component. If there is no corresponding significant event-related synchronization graph in the fourth independent component candidate set, the independent component with the highest relative mu / beta power is selected as the second target independent component.

[0074] Step 208, based on the first target independent component, the second target independent component, the feature data and the preset motor imagery brain-computer interface performance index, determining a motor imagery brain-computer interface user classification result.

[0075] The motor imagery brain-computer interface user classification result can include a first classification result for a left-hand motor imagery task and a second classification result for a right-hand motor imagery task.

[0076] For the left-hand motor imagery task, the ERD% of the first target independent component corresponding to the contralateral event-related desynchronization during the left-hand motor imagery, the relative ERS power obtained from the feature data and the preset motor imagery brain-computer interface performance index ClassDis can be selected as input features, and then a k-means clustering algorithm (k=2) is used for clustering.

[0077] The preset motor imagery brain-computer interface performance index ClassDis can be used to quantify the performance of the motor imagery brain-computer interface, which is a class separability index. The class separability index between two electroencephalogram patterns is measured by the Fisher discriminant criterion suitable for the Riemannian geometry space of the covariance matrix. A higher ClassDis value indicates a greater degree of separation between classes and a smaller variability within classes, thereby reflecting better MI-BCI performance.

[0078] Therefore, users with a higher ClassDis value are classified as excellent performers of left-hand motor imagery, and users with a lower ClassDis value are classified as poor performers of left-hand motor imagery.

[0079] The same method is also used for the right-hand motor imagery task. According to the clustering results, users who perform well in both left-hand and right-hand motor imagery are marked as "excellent performers"; users who perform well in left-hand motor imagery but poorly in right-hand motor imagery are marked as "left-excellent right-poor users"; users who perform well in right-hand motor imagery but poorly in left-hand motor imagery are marked as "right-excellent left-poor users"; and users who perform poorly in both left-hand and right-hand motor imagery are marked as "poor performers".

[0080] In some optional embodiments, personalized training schemes can also be developed for different groups of users, i.e., based on the first classification result and the second classification result, a corresponding target training strategy can be determined, and then the preset brain-computer interface model can be trained based on the target training strategy to obtain a training result.

[0081] Among them, the target training strategy includes: 1, for the user with both hands, no additional training is needed to effectively control the MI-BCI system; 2, for the user with unilateral dominance, training is carried out for the weaker side hand, and the relative mu power of the weaker side hand is taken as the real-time feedback object; 2, for the user with both hands, it is recommended to use the emotional state BCI or visual evoked potential paradigm method. Finally, after the target training strategy is determined, the preset brain-computer interface model can be trained based on the target training strategy to obtain a training result. By formulating a personalized training scheme for different grouped users, the overall applicability and universality of the BCI system are improved.

[0082] In addition, the present application also carries out specific tests, and the test results show that a total of 27 users (accounting for 24.8% of the total number of people) are divided into the excellent performer "left optimal right poor user" group, which contains 30 users (27.5%), the "right optimal left poor user" group contains 13 users (11.9%), and the remaining 39 users (35.8%) are classified as poor performers. The excellent performers show higher values in all four dimensions, showing strong ERD% and relative ERS power response. In contrast, the poor performers have the lowest values in all dimensions. The "left optimal right poor user" group shows higher relative ERS power during the left hand motor imagination task, while the value is lower during the right hand motor imagination task; while the "right optimal left poor user" group shows the opposite pattern, with higher relative ERS power during the right hand task and lower relative ERS power during the left hand task.

[0083] The application provides a motor imagery brain-computer interface user classification method, which comprises the following steps: for a left-hand motor imagery task and a right-hand motor imagery task of a user, independently component analysis algorithm is used to decompose and process target brain wave signals of the user, to obtain a plurality of initial independent components; for each initial independent component, feature extraction processing is performed on the initial independent component, to obtain feature data, wherein the feature data comprises a dipole position corresponding to the initial independent component, a relative mu / beta power, a dynamic power change, a significant event-related desynchronization event diagram and a significant event-related synchronization diagram; based on the feature data, a first target independent component corresponding to a contralateral event-related desynchronization is determined from the initial independent components, and a second target independent component corresponding to ipsilateral event-related synchronization is determined; and based on the first target independent component, the second target independent component, the feature data and a preset motor imagery brain-computer interface performance index, a motor imagery brain-computer interface user classification result is determined. According to the scheme, independent component related brain dynamics modeling is performed, and multi-dimensional neural indexes obtained based on the feature data are used to classify the user, so that the accuracy and reliability of the classification are improved; in addition, automatic screening of the independent components is performed by combining a plurality of feature data, subjective problems caused by manual intervention are avoided, and the applicability and scalability of the classification method are improved.

[0084] It should be understood that, although each step in the flowchart involved in each embodiment as described above is displayed in sequence according to the arrow, these steps are not necessarily executed in sequence according to the arrow. Unless otherwise specified herein, the execution of these steps is not strictly limited in sequence, and these steps can be executed in other sequences. Moreover, at least part of the steps in the flowchart involved in each embodiment as described above can include a plurality of steps or stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution sequence of these steps or stages is not necessarily sequential, but can be executed in rotation or alternation with at least part of other steps or steps or stages in other steps.

[0085] Figure 7 A structural block diagram of a motor imagery brain-computer interface user classification device provided for an embodiment of the application.

[0086] As shown in Figure 7 , the motor imagery brain-computer interface user classification device 700 comprises: A decomposition module 702 is configured to, for a left-hand motor imagery task and a right-hand motor imagery task of a user, use independently component analysis algorithm to decompose and process target brain wave signals of the user, to obtain a plurality of initial independent components.

[0087] The extraction module 704 is configured to perform feature extraction processing on each initial independent component to obtain feature data, wherein the feature data comprises a dipole position corresponding to the initial independent component, a relative mu / beta power, a dynamic power change, a significant event-related desynchronization event map, and a significant event-related synchronization map.

[0088] The first determination module 706 is configured to determine, based on the feature data, a first target independent component corresponding to a contralateral event-related desynchronization from the initial independent components, and a second target independent component corresponding to an ipsilateral event-related synchronization.

[0089] The second determination module 708 is configured to determine, based on the first target independent component, the second target independent component, the feature data, and a preset motor imagery brain-computer interface performance index, a motor imagery brain-computer interface user classification result, wherein the motor imagery brain-computer interface user classification result comprises a first classification result for a left-hand motor imagery task and a second classification result for a right-hand motor imagery task.

[0090] As to the apparatus in the above-described embodiments, the specific manners in which various modules perform operations have been described in detail in the embodiments of the method, and thus will not be described in detail here. The various modules in the above-described motor imagery brain-computer interface user classification apparatus can be realized in whole or in part through software, hardware, or a combination thereof. The various modules described above can be embedded in or independent of a processor in a computer device in hardware form, or can be stored in a memory in a computer device in software form, so as to be called and executed by a processor to perform the operations of the various modules described above.

[0091] In an embodiment of the present application, a computer device is provided, which comprises a memory and a processor, the memory storing a computer program, and the processor implementing the following steps when executing the computer program: For a left-hand motor imagery task and a right-hand motor imagery task of a user, an independent component analysis algorithm is used to decompose and process target brain wave signals of the user to obtain a plurality of initial independent components; For each initial independent component, feature extraction processing is performed on the initial independent component to obtain feature data, wherein the feature data comprises a dipole position corresponding to the initial independent component, a relative mu / beta power, a dynamic power change, a significant event-related desynchronization event map, and a significant event-related synchronization map; Based on the feature data, a first target independent component corresponding to a contralateral event-related desynchronization and a second target independent component corresponding to an ipsilateral event-related synchronization are determined from the initial independent components; The first target independent component, the second target independent component, the feature data, and a preset motor imagery brain-computer interface performance index are used to determine a motor imagery brain-computer interface user classification result, wherein the motor imagery brain-computer interface user classification result includes a first classification result for a left-hand motor imagery task and a second classification result for a right-hand motor imagery task.

[0092] In an embodiment of the present application, the processor, when executing the computer program, further implements the following steps: The initial independent component is subjected to equivalent current dipole fitting processing to generate a dipole position corresponding to the initial independent component.

[0093] In an embodiment of the present application, the processor, when executing the computer program, further implements the following steps: The neural power spectral density corresponding to the initial independent component is obtained; The neural power spectral density is subjected to parameterization processing to obtain at least one periodic component and aperiodic component; After removing the interference of the aperiodic component, the component with the largest amplitude in the periodic component is extracted as the relative mu / beta power.

[0094] In an embodiment of the present application, the processor, when executing the computer program, further implements the following steps: The initial independent component is subjected to time-frequency decomposition processing using a preset wavelet transform algorithm to obtain a decomposition result; In a preset time window, a dynamic power change is calculated based on a power value at a preset time and a power mean value of a preset reference interval.

[0095] In an embodiment of the present application, the processor, when executing the computer program, further implements the following steps: Random sampling is performed in a preset reference interval and a preset task interval to calculate a target statistic and a target distribution; Based on the target statistic and the target distribution, a significance level corresponding to each voxel unit in the brain is estimated; Based on the significance level, a connected region corresponding to adjacent significant voxel units is determined; Based on the connected region, a significant event-related desynchronization event map and a significant event-related synchronization map are determined.

[0096] In an embodiment of the present application, the processor, when executing the computer program, further implements the following steps: An original electroencephalogram signal is obtained; wherein the original electroencephalogram signal is a signal related to motor imagery; The original electroencephalogram signal is subjected to band-pass filtering processing to extract oscillation signals in a mu frequency band and a beta frequency band; The time domain segment in a preset time period is extracted from the segmented oscillation signal; wherein, the preset time period comprises a preset reference interval and a preset task interval; The time domain segment is subjected to artifact removal processing to obtain a target brain wave signal.

[0097] In an embodiment of the present application, when the processor executes the computer program, the following steps are also implemented: Based on the first classification result and the second classification result, a corresponding target training strategy is determined; Based on the target training strategy, the preset brain-computer interface model is trained to obtain a training result.

[0098] The computer device provided by the embodiments of the present application has similar implementation principles and technical effects to the above-mentioned method embodiments, and will not be described here.

[0099] In an embodiment of the present application, a computer readable storage medium is provided, which stores a computer program. When the computer program is executed by a processor, the following steps are implemented: For the left-hand motor imagery task and the right-hand motor imagery task of the user, the independent component analysis algorithm is used to decompose and process the target brain wave signal of the user to obtain a plurality of initial independent components; For each initial independent component, the initial independent component is subjected to feature extraction processing to obtain feature data; wherein, the feature data comprises a dipole position corresponding to the initial independent component, a relative mu / beta power, a dynamic power change, a significant event-related desynchronization event map, and a significant event-related synchronization map; Based on the feature data, a first target independent component corresponding to the contralateral event-related desynchronization is determined from the initial independent components, and a second target independent component corresponding to the ipsilateral event-related synchronization is determined; Based on the first target independent component, the second target independent component, the feature data, and a preset motor imagery brain-computer interface performance indicator, a motor imagery brain-computer interface user classification result is determined; wherein, the motor imagery brain-computer interface user classification result comprises a first classification result for the left-hand motor imagery task and a second classification result for the right-hand motor imagery task.

[0100] In an embodiment of the present application, when the computer program is executed by the processor, the following steps are also implemented: The initial independent component is subjected to equivalent current dipole fitting processing to generate a dipole position corresponding to the initial independent component.

[0101] In an embodiment of the present application, when the computer program is executed by the processor, the following steps are also implemented: The neural power spectral density corresponding to the initial independent component is obtained; Parameterize the neural power spectral density to obtain at least one periodic component and aperiodic component; After removing the interference of the aperiodic component, extract the component with the maximum amplitude in the periodic component as the relative mu / beta power.

[0102] In an embodiment of the present application, the computer program, when executed by the processor, further implements the following steps: Perform time-frequency decomposition processing on the initial independent components using a preset wavelet transform algorithm to obtain a decomposition result. Within a preset time window, calculate the dynamic power change based on the power value at the preset time and the power mean of the preset reference interval.

[0103] In an embodiment of the present application, the computer program, when executed by the processor, further implements the following steps: Randomly sample within the preset reference interval and the preset task interval to calculate the target statistic and the target distribution. Based on the target statistic and the target distribution, estimate the significance level corresponding to each voxel unit in the brain. Based on the significance level, determine the connected region corresponding to adjacent significant voxel units. Based on the connected region, determine the significant event-related desynchronization event map and the significant event-related synchronization map.

[0104] In an embodiment of the present application, the computer program, when executed by the processor, further implements the following steps: Obtain the collected original brain wave signal; wherein the original brain wave signal is a signal related to motor imagery; Perform band-pass filtering processing on the original brain wave signal to extract the oscillation signal within the mu frequency band and the beta frequency band. From the oscillation signal, segmentally extract the time domain segment within the preset time period; wherein the preset time period includes a preset reference interval and a preset task interval. Perform artifact removal processing on the time domain segment to obtain the target brain wave signal.

[0105] In an embodiment of the present application, the computer program, when executed by the processor, further implements the following steps: Based on the first classification result and the second classification result, determine the corresponding target training strategy. Based on the target training strategy, train the preset brain-computer interface model to obtain a training result.

[0106] The computer readable storage medium provided in this embodiment has similar implementation principles and technical effects to the above method embodiments, which will not be described here.

[0107] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer readable storage medium. When the computer program is executed, the processes of the above-mentioned embodiments of the methods can be included. Any reference to memory, storage, databases, or other media in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration but not limitation, RAM is available in many forms such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0108] Other embodiments of the disclosure will be apparent to those skilled in the art from consideration of the specification and practice of the features disclosed herein. This application is intended to cover any variations, uses, or adaptations of the disclosure that are deemed to fall within the general principles of the disclosure and include commonly known or customary practice in the art. The specification and examples are to be considered exemplary only, with the true scope and spirit of the disclosure being indicated by the following claims.

[0109] It should be understood that the present disclosure is not limited to the precise structures herein described and illustrated in the drawings, and that various modifications and changes can be made without departing from the scope thereof. The scope of the present disclosure is limited only by the claims that follow.

Claims

1. A motor imagery brain-computer interface user classification method, characterized in that, The method comprises: For the left-hand motor imagery task and the right-hand motor imagery task of a user, an independent component analysis algorithm is used to decompose and process target brain wave signals of the user, to obtain a plurality of initial independent components; For each of the initial independent components, feature extraction processing is performed on the initial independent components to obtain feature data; wherein the feature data includes a dipole position corresponding to the initial independent component, a relative mu / beta power, a dynamic power change, a significant event-related desynchronization event map, and a significant event-related synchronization map; Based on the feature data, a first target independent component corresponding to a contralateral event-related desynchronization is determined from the initial independent components, and a second target independent component corresponding to an ipsilateral event-related synchronization is determined; Based on the first target independent component, the second target independent component, the feature data, and a preset motor imagery brain-computer interface performance indicator, a motor imagery brain-computer interface user classification result is determined; wherein the motor imagery brain-computer interface user classification result includes a first classification result for the left-hand motor imagery task and a second classification result for the right-hand motor imagery task.

2. The method of claim 1, wherein, The generation process of the dipole position comprises: The initial independent component is subjected to equivalent current dipole fitting processing to generate a dipole position corresponding to the initial independent component.

3. The method of claim 1, wherein, The generation process of the relative mu / beta power comprises: The neural power spectral density corresponding to the initial independent component is obtained; The neural power spectral density is subjected to parameterization processing to obtain at least one periodic component and a non-periodic component; After removing the interference of the non-periodic component, the component with the largest amplitude in the periodic component is extracted as the relative mu / beta power.

4. The method of claim 1, wherein, The generation process of the dynamic power change comprises: A preset wavelet transform algorithm is used to perform time-frequency decomposition processing on the initial independent component to obtain a decomposition result; Within a preset time window, the dynamic power change is calculated based on the power value at a preset time and the power mean value of a preset reference interval.

5. The method of claim 1, wherein, The generation process of the significant event-related desynchronization event map and the significant event-related synchronization map comprises: Random sampling is performed within a preset reference interval and a preset task interval to calculate a target statistic and a target distribution; Based on the target statistic and the target distribution, the significance level of each voxel unit in the brain is estimated; Based on the significance level, the connected region corresponding to adjacent significant voxel units is determined; Based on the connected region, the significant event-related desynchronization event map and the significant event-related synchronization map are determined.

6. The method according to any one of claims 1 to 5, characterized in that, The method further comprises: Obtaining the collected original brain wave signals; wherein the original brain wave signals are signals related to motor imagery; Band-pass filtering processing is performed on the original brain wave signals to extract oscillation signals in the mu frequency band and the beta frequency band; Time domain segments within a preset time period are extracted from the oscillation signals; wherein the preset time period includes a preset reference interval and a preset task interval; Artifact removal is performed on the time domain segment to obtain the target brain wave signal.

7. The method according to any one of claims 1 to 5, characterized in that, The method further includes: Based on the first classification result and the second classification result, a corresponding target training strategy is determined. Based on the target training strategy, a preset brain-computer interface model is trained to obtain a training result.

8. A motor imagery brain-computer interface user classification apparatus characterized by, The device includes: A decomposition module is configured to perform independent component analysis on target brain wave signals of a user for left and right hand motor imagery tasks to obtain a plurality of initial independent components. An extraction module is configured to perform feature extraction on each initial independent component to obtain feature data, wherein the feature data includes a dipole position, a relative mu / beta power, a dynamic power change, a significant event-related desynchronization event map, and a significant event-related synchronization map corresponding to the initial independent component. A first determination module is configured to determine a first target independent component corresponding to contralateral event-related desynchronization and a second target independent component corresponding to ipsilateral event-related synchronization from the initial independent components based on the feature data. A second determination module is configured to determine a motor imagery brain-computer interface user classification result based on the first target independent component, the second target independent component, the feature data, and a preset motor imagery brain-computer interface performance indicator, wherein the motor imagery brain-computer interface user classification result includes a first classification result for the left hand motor imagery task and a second classification result for the right hand motor imagery task.

9. An electronic device, comprising: The electronic device includes a processor and a memory, and the memory stores at least one instruction, at least one program, a code set, or an instruction set, which is loaded and executed by the processor to implement the motor imagery brain-computer interface user classification method of any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The storage medium stores at least one instruction, at least one program, a code set, or an instruction set, which is loaded and executed by the processor to implement the motor imagery brain-computer interface user classification method of any one of claims 1-7.

Images