Brain-computer interface visual stimulation brightness adjusting system and method

Abstract

The invention discloses a brain-computer interface visual stimulation brightness adjustment system and method. The system comprises a stimulation display module, an electroencephalogram acquisition module, an electroencephalogram classification module, a feature extraction module and a brightness determination module. Wherein the stimulus display module displays target stimulus corresponding to a calibration task to a user; the electroencephalogram acquisition module acquires electroencephalogram data when a user executes a calibration task; the electroencephalogram acquisition module performs classification and recognition based on the electroencephalogram data to obtain a classification result; the feature extraction module performs feature extraction according to the classification result to obtain an aggregation feature group, and the number of elements in the aggregation feature group corresponds to the number of subtask groups; if the brightness determination module judges that a certain element in the aggregated feature group meets a preset condition, the proper brightness of visual stimulation for the user is determined according to the target stimulation brightness of the subtask corresponding to the element. According to the method, the electroencephalogram data when the user executes the calibration task of the target stimulation with different brightness is collected, the appropriate brightness of the visual stimulation of each user is determined, and the individual difference can be met.

Description

Technical Field

[0001] The present application relates to the technical field of brain-computer interfaces, and specifically relates to a brain-computer interface visual stimulation brightness adjustment system and method. Background Art

[0002] Brain-computer interface technology is a control method that does not rely on the periphery and muscles. Users can directly send control commands to the outside world through brain activities. Brain-computer interface technology has a large number of application scenarios in scientific research, rehabilitation, medical treatment and other fields, such as motor rehabilitation, mind typing machine, electroencephalogram-controlled wheelchair, etc.

[0003] Visual stimulation is a main stimulation method, and the corresponding paradigms include P300, Visual Evoked Potential (VEP), steady state visual evoked potential (SSVEP), etc. Taking SSVEP as an example, when the user gazes at a visual stimulus with a fixed frequency, the brain visual cortex will generate continuous spectral response characteristics. The brain-computer interface system can determine the visual information gazed by the user by identifying the above-mentioned brain activities. The visual stimulation brightness affects the performance of this type of brain-computer interface. Inappropriate stimulation brightness will interfere with the user, resulting in their inability to maintain concentration for a long time and reducing the system performance.

[0004] At present, the visual stimulation methods of brain-computer interface systems cannot determine the appropriate brightness for different users, resulting in visual discomfort for users and affecting the recognition accuracy of the electroencephalogram data of users by the brain-computer interface system. Summary of the Invention

[0005] The technical purpose of the present application is to provide a brain-computer interface visual stimulation brightness adjustment system and method for the technical problem that the visual stimulation methods of brain-computer interface systems cannot determine the appropriate brightness for different users.

[0006] To achieve the above technical purpose, the embodiments of the present application adopt the following technical solutions.

[0007] In the first aspect, the embodiments of the present application provide a brain-computer interface visual stimulation brightness adjustment system, including:

[0008] A stimulation display module configured to display a target stimulus corresponding to a calibration task to the user, where the calibration task includes multiple groups of subtasks, the target stimuli in the same group of subtasks have the same brightness, and the target stimuli in different subtasks have different brightnesses;

[0009] An electroencephalogram acquisition module configured to acquire the electroencephalogram data of the user when performing the calibration task;

[0010] An electroencephalogram classification module, configured to perform electroencephalogram classification and recognition on the electroencephalogram data, determine a recognized target stimulus, compare the recognized target stimulus with a true target stimulus, and obtain a classification result;

[0011] A feature extraction module, configured to perform feature extraction according to the classification result to obtain an aggregated feature group, where the number of elements in the aggregated feature group corresponds to the number of groups of subtasks;

[0012] A brightness determination module, configured to determine that if an element in the aggregated feature group meets a preset condition, then determine a suitable brightness of the visual stimulus for the user according to the target stimulus brightness of the subtask corresponding to the element.

[0013] In a second aspect, an embodiment of the present application provides a method for adjusting the brightness of a visual stimulus of a brain-computer interface, including:

[0014] Presenting a target stimulus corresponding to a calibration task to the user, where the calibration task includes multiple groups of subtasks, the target stimulus brightness in the same group of subtasks is the same, and the target stimulus brightness in different subtasks is different;

[0015] Collecting electroencephalogram data of the user when performing the calibration task;

[0016] Performing electroencephalogram classification and recognition on the electroencephalogram data, determining a recognized target stimulus, comparing the recognized target stimulus with a true target stimulus, and obtaining a classification result;

[0017] Performing feature extraction according to the classification result to obtain an aggregated feature group, where the number of elements in the aggregated feature group corresponds to the number of groups of subtasks;

[0018] Determining that if an element in the aggregated feature group meets a preset condition, then determining a suitable brightness of the visual stimulus for the user according to the target stimulus brightness of the subtask corresponding to the element.

[0019] Compared with the prior art, the beneficial technical effects achieved by the visual stimulus brightness adjustment method and the brain-computer interface system provided by the embodiments of the present application are as follows: By collecting electroencephalogram data of the user when performing calibration tasks with target stimuli of different brightnesses and performing feature extraction, an objective and reliable data-driven method for adjusting the brightness of visual stimuli is provided. The suitable brightness of the visual stimulus is determined based on the unique electroencephalogram response of each user, which can meet individual differences. Compared with the traditional fixed brightness setting, it is more in line with the visual needs of different people, and improves the comfort and pertinence of the visual experience. Description of the Drawings

[0020] The accompanying drawings described herein are for illustrative purposes only and are not intended to limit the scope of the disclosure of the present application in any way. Additionally, the shapes and proportional dimensions of the components in the figures are merely schematic and are used to assist in understanding the present application, rather than specifically defining the shapes and proportional dimensions of the components of the present application. Those skilled in the art can, under the teachings of the present application, select various possible shapes and proportional dimensions according to specific circumstances to implement the present application. In the accompanying drawings:

[0021] Figure 1 Schematic structural diagram of the brain-computer interface system provided for the embodiment;

[0022] Figure 2 Schematic flow diagram of feature extraction according to the classification result in the embodiment;

[0023] Figure 3 Schematic flow diagram of the visual stimulus brightness adjustment method provided for the embodiment. Detailed implementation manners

[0024] In order to enable those skilled in the art to better understand the technical solutions in the present application, the following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a 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 efforts shall fall within the protection scope of the present application.

[0025] In the description of the present application, the terms "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include one or more features.

[0026] Embodiment 1

[0027] This embodiment provides a brain-computer interface visual stimulus brightness adjustment system, including a stimulus display module, an electroencephalogram (EEG) acquisition module, an EEG classification module, a feature extraction module, and a brightness determination module. Among them:

[0028] The stimulus display module is configured to present the target stimulus corresponding to the calibration task to the user. The calibration task includes multiple groups of subtasks, and the target stimulus brightness in the same group of subtasks is the same, while the target stimulus brightness in different subtasks is different.

[0029] The EEG acquisition module is configured to acquire the EEG data of the user when performing the calibration task.

[0030] The EEG classification module is configured to perform EEG classification and recognition on EEG data, determine the recognized target stimulus, and obtain a classification result based on comparing the recognized target stimulus with the true target stimulus.

[0031] The feature extraction module is configured to perform feature extraction according to the classification result to obtain an aggregated feature group, and the number of elements in the aggregated feature group corresponds to the number of subtask groups.

[0032] The brightness determination module is configured to determine that if a certain element in the aggregated feature group meets a preset condition, then determine the appropriate brightness of the visual stimulus for the user according to the target stimulus brightness of the subtask corresponding to the element.

[0033] In an embodiment, the visual stimulus brightness adjustment system of the brain-computer interface may further include a stimulus capture module. The stimulus capture module is configured to collect the optical signal output by the stimulus display module. If the intensity of the optical signal exceeds a threshold, then determine the current timestamp and transmit the current timestamp to the EEG acquisition module to achieve time synchronization between the calibration task and the EEG data.

[0034] The EEG acquisition module includes corresponding acquisition hardware and data processing software for collecting the electrophysiological signals at the scalp position of the user. The EEG acquisition module is a multi-channel EEG cap, and the EEG position layout conforms to the international 10-20 standard. The electrodes must include a ground electrode, a reference electrode, and electrodes at the occipital region position.

[0035] In some embodiments, the EEG acquisition module simultaneously receives the signal of the stimulus capture module and records it in the label leads of the EEG data to mark the exact time point when the stimulus starts.

[0036] The stimulus display module is a key hardware part in the visual evoked potential (VEP) brain-computer interface system, responsible for presenting visual stimuli with different brightnesses, frequencies, and positions to the user to evoke stable EEG signals (such as SSVEP). This device needs to ensure the accuracy and consistency of the visual stimulus, and at the same time avoid visual fatigue caused by long-term use, which affects the system performance. This module is usually a high-definition liquid crystal display or an LED screen, with a refresh rate of at least 120Hz and supports gradual brightness adjustment.

[0037] The targets of the visual stimulus of SSVEP are usually customizable graphics such as square color blocks, letter characters, virtual buttons, etc. After the visual stimulus starts, each graphic flashes at a fixed frequency, and they are different from each other. The selectable range of the flashing frequency is 4Hz to 50Hz. In addition, different graphics can select different phase differences to increase the difference of the targets. When the system is a dialing device for the brain-computer interface, there can be 10 targets, representing the numbers 0-9 respectively. During use, there can be arrow prompts to guide the user to focus on a specific target, and then all the targets flash at a fixed frequency. The user stares at a certain target according to the instruction to complete the specific EEG induction.

[0038] The EEG classification module determines whether the user focuses on the target as required from the EEG data. The EEG classification method is related to specific tasks. In some examples, a hybrid CCA method is used as the classification method for identifying the EEG features of SSVEP to identify the target stimulus that the user focuses on from the EEG data.

[0039] In some embodiments, the brain-computer interface visual stimulus brightness adjustment system further includes a result feedback module, and the EEG classification module sends the identified target stimulus to the result feedback module. The result feedback module is configured to receive the target stimulus processed by the EEG classification module and display it on the display screen. When the system is a dialing device of the brain-computer interface, the EEG recognition module returns specific numerical values from 0 to 9. The result feedback module reads the target serial number, and the result value is displayed at the top of the screen for the user to view.

[0040] After the brain-computer interface visual stimulus brightness adjustment system is started, the user needs to annotate the visual stimulus displayed by the stimulus display module to complete the calibration task.

[0041] In the embodiments, the calibration task includes several groups of subtasks, and there is a certain rest time between groups. The target stimulus brightness of the subtasks in the same group is the same, and the target stimulus brightness of the subtasks in different groups is different. The stimulus paradigms included in different groups of subtasks are the same. The stimulus paradigm can be the same as the paradigm of the brain-computer interface system. The presentation time of the stimulus and the rest time of the user can be adjusted according to the actual situation. The total presentation time of each group of subtasks should not be less than 10 seconds.

[0042] In one embodiment, the brain-computer interface system adopts the SSVEP paradigm, and the stimuli are four square color blocks flashing at different frequencies, with frequencies of 4Hz, 6Hz, 8Hz, and 10Hz in turn, presented at the four positions of the upper, lower, left, and right of the screen. In the calibration task, the target stimulus of each group of subtasks includes the above 4 square stimuli, and the stimuli are presented at the corresponding positions on the screen in turn. Each stimulus is presented for 5 seconds and then disappears. The user needs to gaze at the above stimuli in turn. After one group ends, the user rests for 5 seconds, and then starts the next group of subtasks. The total number of groups of subtasks in the calibration task is related to the display brightness. If the brightness adjustment range is between 300 and 400 lumens, a group of brightness can be set every 50 lumens, with a total of 3 groups of subtasks. The stimulus brightness of each group of subtasks is 300, 350, and 400 respectively. Empirically, the number of groups of subtasks should be greater than or equal to 5 groups.

[0043] In some embodiments, the target stimulus of each group of subtasks includes multiple candidate targets, and the frequencies of each candidate target are the same or different, and the positions of each candidate target are different.

[0044] If there are many candidate targets in the brain-computer interface system, such as a typewriter based on SSVEP. In the above typewriter system, dozens of candidate stimuli are formed to constitute a virtual screen keyboard by adjusting the frequencies, phases, coding sequences, etc. of different stimuli. When adjusting the visual stimulus brightness, each group of subtasks can select 3-6 fixed candidate targets as target stimuli, and the target stimuli should cover as many frequencies and screen positions as possible to ensure the applicability of the adjusted brightness.

[0045] After the calibration task is completed, the EEG acquisition module collects all EEG data to form a first data set.

[0046] In some embodiments, such as Figure 1 As shown, the brain-computer interface visual stimulus brightness adjustment system further includes an EEG preprocessing module. The EEG preprocessing module is a software module for processing EEG data. This module is used to read the EEG data and label data of the EEG acquisition module. Through the label data, after each group of stimuli starts, the user's EEG data is obtained, and then the data is processed through filtering, artifact removal, segmentation and other processing steps. After the original data is sorted out, it is used for the feature extraction process.

[0047] As an example, the EEG preprocessing module performs data preprocessing on the first data set, and the processing content includes:

[0048] (1) Perform eighth-order Butterworth filtering, 1-35 Hz band-pass filtering, and 50 Hz notch filtering on the first data set in sequence;

[0049] (2) For the stimulation duration period of each subtask, according to an n 1 -second time window duration and an n 2 -second overlap duration, the data is segmented into several time windows. The EEG data corresponding to each group of subtasks is divided into several time windows in chronological order. In the EEG data corresponding to each group of subtasks, the interception time of the time windows is the same;

[0050] (3) Check the data of each time window. If the standard deviation of the EEG data of one lead exceeds a preset difference value (such as 100 μV) in a single time window, discard all time windows corresponding to this subtask;

[0051] (4) If the proportion of discarded time windows in the first data set exceeds a preset ratio (such as 30%), abandon the current test data, and the user needs to be asked to re-complete the SSVEP brightness adjustment and re-collect the EEG data when the user performs a preset multiple groups of calibration tasks respectively.

[0052] Specifically, n 1 is determined by following the data method of the original system, and n 2 is determined according to the specific situation, and n 2Try to ensure that there are enough time windows and they cover the entire stimulus presentation stage. Empirically, in the EEG data corresponding to each group of subtasks, the total number of time windows is not less than 10. If not satisfied, n can be adjusted appropriately. 2 and the stimulus presentation time.

[0053] In one embodiment, the SSVEP paradigm adopted by the brain-computer interface system is square stimuli flashing at four different frequencies. In each group of subtasks, the stimuli are sequentially presented at specific positions on the screen, and each stimulus is presented for 5 seconds. During online processing of the system, the time window is 2 seconds of data to determine the specific target that the user is looking at. In the data preprocessing stage of SSVEP brightness adjustment, which is the same as the system processing method, the time window duration is 2 seconds. When the overlapping duration is 1 second, the time windows are 0 - 2, 1 - 3, 2 - 4, 3 - 5, a total of 4, and each group of data has a total of 16 time windows. When the overlapping duration is 0.5 second, the time windows 0.5 - 2.5, 1.5 - 3.5, 2.5 - 4.5 are added, and the total number of time windows is 7, and each group of data has a total of 28 time windows.

[0054] In the embodiment, the EEG classification module classifies the data of each time window using a classification method, identifies the target stimulus recognized in each time window, compares the difference between the recognized target stimulus and the target stimulus, and obtains the classification result of each group of data. Different brain-computer interface systems have different processing methods. After processing, an identification result can be obtained, indicating the target stimulus recognized by the system for the user in that time window. The recognized target stimulus judged by the system is compared with the actual target stimulus. If they are the same, the result of this time window is 1, otherwise it is 0. Each element of the classification result is an array, corresponding to a group of subtasks in a calibration task. The length of a single-element array is the same as the number of time windows, and the content is 1 or 0.

[0055] In the embodiment, feature extraction is performed according to the classification result to obtain aggregated features, including: determining a first feature group, a second feature group, and a third feature group according to the classification result; determining a first aggregation group according to the first feature group; determining a second aggregation group according to the second feature group; determining a third aggregation group according to the third feature group; combining the first aggregation group, the second aggregation group, and the third aggregation group to determine a fourth aggregation group, and taking the fourth aggregation group as the aggregated feature group.

[0056] In the embodiment, as Figure 2 shown, a first feature group is obtained according to the classification result, which is used to determine the accuracy rate of the elements in each classification result. The number of elements in the first feature group is the same as the number of elements in the classification result. Each value in each element of the classification result is a binary value of 0 or 1.

[0057] The expression of the first feature group is as follows:

[0058]

[0059] Among them, represents the i-th element in the first feature group. Each element in the first feature group corresponds to each element in the classification result, that is, the number of elements in the first feature group is the same as the number of elements in the classification result. n is the length of each element in the classification result. is the k-th value in the i-th element of the classification result. If the recognized target stimulus is the same as the true target stimulus, then is 1; if the recognized target stimulus is different from the true target stimulus, then is 0.

[0060] In the embodiment, determining the second feature group according to the classification result includes: for each element a in the classification result i performing window partitioning to divide it into several windows, each window having a length of n 3 , and the overlapping length between windows is n 4 ; the k-th window data in the i-th element of the classification result is The total number of windows is K, each window contains n 3 elements, and each element in the window is 0 or 1.

[0061] For example, please continue to refer to Figure 2 , the length of each element in the classification result is 10, the length of each window is 3, the overlapping length is 1, there are a total of 8 windows, and each window contains n 3 elements.

[0062] Determine the serial number imax of the maximum value element in the first feature group; determine the distance group between each window in each element of the classification result and the corresponding window of the element with serial number imax in the classification result. The calculation method of a single distance in the group is as follows:

[0063]

[0064] Among them, represents the i-th element in the second feature group, and the length of the i-th element in the second feature group is K. is the k-th window of the element with serial number imax in the classification result. is the k-th window data of the i-th element of the classification result, k = 1, 2, 3,..., n 3 .

[0065] It can be understood that each element in the second feature group also corresponds to each element in the classification result, that is, the number of elements in the second feature group is the same as the number of elements in the classification result. For example, if the serial number of the maximum value element in the first feature group is determined to be 2, then for each element in the classification result, determine the distance group between each window in the element and the corresponding window of the element with serial number 2 in the classification result.

[0066] D is a distance calculation method, and the calculation method of D is as follows:

[0067]

[0068] Among them, b 1 and b 2 are both arrays of length n 3 The i-th element of b 1 and b 2 is and

[0069] is the absolute value of the difference between the corresponding elements in the two groups. The larger the obtained value, the greater the difference between the b 1 and b 2 arrays.

[0070] Specifically, the distance calculation method D can also be replaced with other records, such as Manhattan distance, Euclidean distance, etc., to obtain better results.

[0071] In some embodiments, as Figure 2 shown, to obtain the third feature group according to the classification result, the following steps are included: Perform window partitioning on each element in the classification result using the above window partitioning method. Take the average of the data within each window, and then discretize it to obtain the third feature group:

[0072]

[0073] Among them is the i-th element of the third feature group, and the element length is K. is the k-th window data of the i-th element of the classification result, mean is to calculate the mean value, int is to take the integer, and each value of the third feature group ranges from one of 0, 1, 2, 3, 4.

[0074] In some embodiments, to process the first feature group to obtain the first aggregation group, the following steps are included:

[0075] Calculate the standard deviation s 1 of the first feature group;

[0076] Calculate each element in the first aggregation group, and the calculation method is as follows:

[0077]

[0078] is the i-th element in the first aggregation group.

[0079] In the embodiment, processing the second feature group to obtain a second aggregation group includes the following steps: Each element of the second feature group is an array. Calculate the sum of each element array, where the serial number of the element with the maximum sum is imax2, and the serial number of the element with the minimum sum is imin2. Determine the element corresponding to the serial number imax2 in the second feature group Determine the element corresponding to the serial number imin2 in the second feature group

[0080] Perform clustering processing on each element in the second feature group, with the number of categories being 2. The first category c 1 The initial center is The second category c 2 The initial center is Subsequently, perform the clustering algorithm. After clustering, obtain two categories c 1 and c 2 , c 1 is the category with a smaller sum of class center elements.

[0081] For each element in the second aggregation group, the calculation method is as follows:

[0082]

[0083] is the i-th element in the first aggregation group.

[0084] As an example, according to the classification result of the third feature group, obtain a third aggregation group, including the following steps:

[0085] For each element in the third feature group, calculate the frequency of occurrence of each in each category to form a fourth feature group Take as the expected frequency, and perform a chi-square (χ ) test with others 2 to calculate the significance.

[0086] Calculate each element in the third aggregation group, and the calculation method is as follows:

[0087]

[0088] is the i-th element in the first aggregation group.

[0089] Integrate the first, second, and third aggregation groups to obtain a fourth aggregation group, and the calculation method is as follows:

[0090]

[0091] As an example, among the elements where the fourth aggregation group is 1, find the lowest brightness value as the most appropriate brightness.

[0092] Whether the classification result is correct is a performance indicator of the brain-computer interface system. The elements with the fourth aggregation group value of 1 represent that the classification accuracy of this group is relatively high. The most comfortable brightness can be found in the above group. Generally, the higher the brightness, the more likely it is to attract people's attention and the more likely it is to cause fatigue. Therefore, the most comfortable brightness should be the lowest brightness among those with comparable performance.

[0093] In the embodiments of the present application, feature extraction is performed on the classification result from different perspectives. The first aggregation group reflects the information of the overall recognition accuracy dimension based on the matching situation between the recognized target stimulus and the true target stimulus in the classification result; the second aggregation group considers the local structure and distribution characteristics of the data through window segmentation and distance calculation; the third aggregation group uses the frequency of occurrence to detect the significance of differences and provides new feature information from the perspective of statistical significance. Then, these three groups of features are fused to form the fourth aggregation group, making the feature description of the classification result more comprehensive and rich, and being able to more accurately reflect the potential relationship between the electroencephalogram data and the brightness of visual stimuli. Different feature groups depict the data from different aspects and complement each other. For example, the first aggregation group focuses on the overall matching situation, the second aggregation group pays attention to the local data structure differences, and the third aggregation group emphasizes the statistical significance differences. The fused feature group can integrate these advantages, avoid the limitations of a single feature extraction method, and thus provide a more reliable basis for determining the appropriate brightness of visual stimuli.

[0094] Combined with the above embodiments, the brain-computer interface visual stimulus brightness adjustment system usually includes a series of software and hardware and usage processes. A common usage process example based on the SSVEP dial system is given, and it can be replaced according to the product design requirements in actual use. The process includes:

[0095] Step 1: The user sits on a comfortable seat and wears the electroencephalogram acquisition module. The stimulus display module is placed directly in front of the user, and 10 targets representing 0 - 9 are displayed in the center of the stimulus display module. The top is the feedback area for feedback results.

[0096] Step 2: After the user is ready, the calibration task starts.

[0097] Step 2.1: The targets on the screen are displayed with the first brightness in the candidate brightnesses.

[0098] Step 2.2: An arrow appears on the screen to indicate the target, and then the arrow disappears, and all the targets flash simultaneously. After seeing the prompt, the user immediately focuses on the target pointed by the arrow. After the target flashes for 5 seconds, the flashing stops.

[0099] Step 2.3: The arrow appears again, pointing to a new target, and repeat Step 2.2 until all candidate targets are completed.

[0100] After all the targets in step 3 are completed, the targets on the screen are displayed at the next brightness in the candidate brightness, and steps 2.1, 2.2, and 2.3 are repeated.

[0101] After all the targets at all brightness levels are completed in step 4, the calibration task is completed.

[0102] In step 5, the data passes through the EEG preprocessing module, the EEG classification module, and the brightness determination module to determine the appropriate target flashing brightness.

[0103] In step 6, the user uses the EEG dialing system

[0104] In step 6.1, the user focuses on specific targets according to their own needs.

[0105] In step 6.2, the system collects EEG in real time, sends the EEG to the data EEG preprocessing module and the EEG classification module, and obtains the recognition result.

[0106] In step 6.3, the system uses the result feedback module to display the value represented by the recognized target at the top of the screen and feedback it to the user.

[0107] In step 6.4, after a certain amount of targets are completed, the user ends this use.

[0108] This application proposes a brain-computer interface visual stimulus brightness adjustment system independent of the stimulus paradigm, which can automatically confirm the most suitable brightness for different users, in different usage scenarios, and for different visual paradigms, ensuring both system performance and minimizing the stimulation and discomfort to users.

[0109] Embodiment 2

[0110] Based on the same inventive concept as the brain-computer interface visual stimulus brightness adjustment system provided in the above embodiments, the embodiments of this application also provide a brain-computer interface visual stimulus brightness adjustment method, as Figure 3 shown, including:

[0111] Showing the target stimuli corresponding to the calibration task to the user, the calibration task includes multiple groups of subtasks, the target stimuli in the same group of subtasks have the same brightness, and the target stimuli in different subtasks have different brightness; collecting the EEG data of the user when performing the calibration task; performing EEG classification and recognition on the EEG data to determine the recognized target stimuli, comparing the recognized target stimuli with the real target stimuli to obtain the classification result; performing feature extraction according to the classification result to obtain an aggregated feature group, and the number of elements in the aggregated feature group corresponds to the number of subtask groups; judging that if a certain element in the aggregated feature group meets the preset condition, then determine the appropriate brightness of the visual stimulus for the user according to the target stimulus brightness of the subtask corresponding to the element.

[0112] Combined with the above embodiments, for the specific implementation manners of each step of the brain-computer interface visual stimulation brightness adjustment method, reference can be made to the implementation steps of each module of the brain-computer interface visual stimulation brightness adjustment system provided in the above embodiments, which will not be elaborated in this embodiment.

[0113] This application proposes a brain-computer interface visual stimulation brightness adjustment method. After the user completes a series of visual stimulation tasks, the system analyzes the collected electroencephalogram (EEG) data, balances the system performance and the user's comfort, and automatically determines the optimal display brightness for the user.

[0114] The above has introduced in detail the brain-computer interface visual stimulation brightness adjustment system and method provided in this application. Specific examples are used in this article to elaborate on the principle and implementation manner of this application. The description of the above embodiments is only used to help understand the concept of this application and should not be construed as a limitation on the protection scope of this application.

Claims

1. A brain-computer interface visual stimulation brightness adjustment system, characterized in that: include: A stimulus display module is configured to display a target stimulus corresponding to a calibration task to a user, wherein the calibration task includes a plurality of groups of subtasks, the target stimulus brightness in the subtasks of the same group is the same, and the target stimulus brightness in different subtasks is different; An EEG acquisition module, configured to acquire EEG data of the user when performing a calibration task; An EEG classification module is configured to perform EEG classification and recognition on the EEG data, determine a recognition target stimulus, compare the recognition target stimulus with a real target stimulus, and obtain a classification result; A feature extraction module is configured to perform feature extraction according to the classification result to obtain an aggregate feature group, wherein the number of elements in the aggregate feature group corresponds to the number of subtask groups; The brightness determination module is configured to determine if an element in the aggregate feature group meets a preset condition, and then determine the appropriate brightness of the visual stimulation for the user according to the target stimulation brightness of the subtask corresponding to the element.

2. The brain-computer interface visual stimulation brightness adjustment system according to claim 1, characterized in that: The system further comprises an EEG preprocessing module, which is configured to perform the following steps: Applying eighth-order Butterworth filtering, bandpass filtering and notch filtering to the EEG data in sequence; For each group of the filtered EEG data corresponding to the subtask, the data is divided into a number of time windows according to a time window length of n1 seconds and an overlap time of n2 seconds; each time window data is checked, and if there is a lead whose standard deviation in a single time window exceeds a preset difference, all time window data corresponding to the subtask are discarded; If the lost time window data in the EEG data exceeds a preset ratio, the EEG data collected this time is abandoned, and the EEG data of the user performing the preset calibration task is collected again.

3. The brain-computer interface visual stimulation brightness adjustment system according to claim 1, characterized in that: The system also includes a stimulus capture module; The stimulus capture module is configured to collect the light signal output by the stimulus display module. If the intensity of the light signal exceeds a threshold, a current timestamp is determined and the current timestamp is transmitted to the EEG acquisition module to achieve time synchronization between the calibration task and the EEG data.

4. The brain-computer interface visual stimulation brightness adjustment system according to claim 1, characterized in that: The feature extraction module extracts features according to the classification results to obtain aggregate features, including: Determine a first feature group, a second feature group and a third feature group according to the classification result; determine a first aggregation group according to the first feature group; determine a second aggregation group according to the second feature group; determine a third aggregation group according to the third feature group; determine a fourth aggregation group by combining the first aggregation group, the second aggregation group and the third aggregation group, and use the fourth aggregation group as the aggregation feature group.

5. The brain-computer interface visual stimulation brightness adjustment system according to claim 4, characterized in that: The expression for determining the first feature group according to the classification result is as follows: in, represents the i-th element in the first feature group, n is the length of each element in the classification result, is the kth value in the i-th element in the classification result; if the identified target stimulus is the same as the true target stimulus, then is 1; if the identified target stimulus is different from the real target stimulus, then is 0; The expression for determining the first aggregation group according to the first feature group is as follows: in, is the i-th element in the first aggregation group, is the maximum value in the first feature group, and s1 is the standard deviation of the first feature group.

6. The brain-computer interface visual stimulation brightness adjustment system according to claim 4, characterized in that: Determining a second feature group according to the classification result includes: For each element a in the classification result i , divided into K windows, each window length is n3, the overlapping length between windows is n4, and each element in the window is 0 or 1; The k-th window data in the i-th element in the classification result is Determine the serial number imax of the maximum value element in the first feature group; determine the distance group between each window in each element in the classification result and the corresponding window of the serial number imax element in the classification result, and the single distance in the group is calculated as follows: in, represents the i-th element in the second feature group. The length of the i-th element in the second feature group is K. is the kth window of the sequence number imax element in the classification result, is the kth window data of the i-th element of the classification result, k = 1, 2, 3, ..., n3; b1 and b2 are both arrays of length n3; the i-th elements of b1 and b2 are and 7. The brain-computer interface visual stimulation brightness adjustment system according to claim 4, characterized in that: Determining a second aggregation group according to the second feature group includes: Determine the sum of the arrays of each element in the second feature group, wherein the sequence number of the element with the minimum sum value imin2 is determined; and the sequence number of the element with the maximum sum value imax2 is determined; Determine the element with sequence number imax2 in the second feature group Determine the element with the serial number imin2 in the second feature group Using the clustering algorithm, each element in the second feature group is clustered. The number of categories is 2, and the initial center of the first category c1 is The initial center of the second type c2 is Each element in the second aggregation group is calculated as follows: is the i-th element in the first aggregation group.

8. The brain-computer interface visual stimulation brightness adjustment system according to claim 4, characterized in that: Determining a third aggregation group according to the third feature group includes: Calculate the frequency of each category in each data in the third feature group to form a fourth feature group; Take the fourth feature group as the expected frequency, perform a chi-square test with other fourth feature groups, and calculate the significance; Calculate each element in the third aggregation group as follows: is the i-th element in the first aggregation group.

9. The brain-computer interface visual stimulation brightness adjustment system according to claim 1, characterized in that: The fourth aggregation group expression is: in, is the i-th element in the fourth aggregation group, is the i-th element in the first aggregation group, is the i-th element in the first aggregation group, is the i-th element in the first aggregation group, each element in the first aggregation group, the second aggregation group and the third aggregation group corresponds to a group of subtasks; the element corresponding to the first aggregation group, the element corresponding to the second aggregation group and the element corresponding to the third aggregation group are all binary parameters; If a certain element in the aggregated feature group satisfies a preset condition, determining the target stimulus brightness of the calibration task corresponding to the element as a suitable brightness for visual stimulation of the user includes: The lowest value of the target stimulus brightness in each group of subtasks corresponding to the i-th element being 1 in the fourth aggregation group is determined as the appropriate brightness of the visual stimulus.

10. A method for adjusting the brightness of visual stimulation of a brain-computer interface, characterized in that: include: Displaying a target stimulus corresponding to a calibration task to a user, wherein the calibration task includes a plurality of groups of subtasks, the target stimulus brightness in the subtasks of the same group is the same, and the target stimulus brightness in different subtasks is different; Collecting EEG data of the user when performing a calibration task; Performing EEG classification and recognition on the EEG data, determining a recognition target stimulus, and comparing the recognition target stimulus with a real target stimulus to obtain a classification result; Perform feature extraction according to the classification result to obtain an aggregated feature group, wherein the number of elements in the aggregated feature group corresponds to the number of subtask groups; If it is determined that an element in the aggregate feature group meets a preset condition, a suitable brightness of visual stimulation for the user is determined according to the target stimulation brightness of the subtask corresponding to the element.

Images