Brain-computer interface experiment method and device for multi-task teaching demonstration, edge computing equipment and storage medium

Abstract

The invention provides a brain-computer interface experiment method and device for multi-task teaching demonstration, edge computing equipment and a storage medium, and relates to the technical field of brain-computer interfaces, and the method comprises the steps: loading a brain-computer interface decoding model corresponding to a teaching normal form based on a currently selected teaching normal form; the obtained original electroencephalogram signals are input into a brain-computer interface decoding model for decoding processing, and real-time electroencephalogram data are generated; when it is judged that the similarity between the real-time electroencephalogram data and the preset electroencephalogram feature template is larger than or equal to a similarity threshold value, teaching commentary content corresponding to the preset electroencephalogram feature template is triggered; and the teaching explanation content is visually demonstrated. By adopting the method, the original electroencephalogram signal decoding process and the typical electroencephalogram characteristic teaching explanation can be combined, the visual presentation of the brain-computer interface key processing process is realized, and the intuition of teaching demonstration and the learning understanding effect can be improved.

Description

Technical Field

[0001] This disclosure relates to the field of brain-computer interface technology, and in particular to a brain-computer interface experimental method, apparatus, edge computing device and storage medium for multi-task teaching demonstration. Background Technology

[0002] Brain-computer interface (BCI) technology is a human-computer interaction technology that enables information exchange between the brain and external devices by collecting and analyzing human brain electrical signals. In recent years, it has received widespread attention and application in fields such as rehabilitation medicine, neuroscience research, and intelligent control. With the development of related theories and technologies, the demand for its application in teaching demonstrations and popular science education is gradually emerging.

[0003] However, from the perspective of teaching demonstration methods, brain-computer interface-related teaching methods still mainly rely on fixed experimental procedures or pre-set demonstration scripts. The teaching process often focuses on demonstrating the final decoding results or control effects. However, the key links of EEG signals, such as acquisition, processing, and decoding output, are still highly concealed for learners. This makes it difficult for learners to understand the background reasons for the appearance of different EEG characteristics and their mechanisms of action in different brain-computer interface paradigms during teaching demonstrations. This can easily lead to a "black box" cognitive experience, thereby affecting the in-depth understanding of relevant theoretical knowledge and technical principles.

[0004] Meanwhile, in multi-paradigm brain-computer interface teaching scenarios, there are significant differences in the types of EEG features and decoding logic corresponding to different paradigms. However, most teaching methods use a single brain-computer interface paradigm as the demonstration object, lacking a teaching support environment that provides a unified display and comparison of different brain-computer interface paradigms. This makes it difficult for learners to form an overall understanding of the brain-computer interface technology system.

[0005] Therefore, how to improve the interactivity and teaching effectiveness of brain-computer interface teaching demonstrations remains a pressing technical problem to be solved in existing teaching demonstration methods. Summary of the Invention

[0006] One of the technical problems this disclosure aims to solve is: how to effectively link the teaching explanation content with the actual processing of EEG signals during brain-computer interface teaching demonstrations, thereby improving the relevance and comprehensibility of the teaching demonstration process.

[0007] To address the aforementioned technical problems, this disclosure provides a brain-computer interface experimental method for multi-task teaching demonstrations, comprising: loading a brain-computer interface decoding model corresponding to a currently selected teaching paradigm; inputting the acquired raw EEG signals into the brain-computer interface decoding model for decoding processing to generate real-time EEG data; when the similarity between the real-time EEG data and a preset EEG feature template is determined to be greater than or equal to a similarity threshold, triggering teaching explanation content corresponding to the preset EEG feature template; and visually demonstrating the teaching explanation content.

[0008] In some embodiments, when the similarity between real-time EEG data and a preset EEG feature template is determined to be greater than or equal to a similarity threshold, the teaching explanation content corresponding to the preset EEG feature template is triggered, including: performing sliding time window analysis on the real-time EEG data to divide the real-time EEG data and generate EEG data segments; calculating the similarity between a preset number of consecutive EEG data segments and any preset EEG feature template in a preset feature library, and summing the calculated similarities; when the summed similarity value is determined to be greater than or equal to the similarity threshold, highlighting the identified EEG data segments in the visualization interface and simultaneously pushing the associated teaching explanation content.

[0009] In some embodiments, when visualizing the teaching explanation content, the method further includes: collecting user interaction behavior data when visualizing the teaching explanation content; identifying the user's learning style type based on the interaction behavior data; adaptively adjusting the layout template and interactive presentation mode of the visualization interface according to the identified learning style type, and pushing teaching guidance prompts that match the learning style type.

[0010] In some embodiments, pushing instructional guidance prompts that match the learning style type includes: calculating the user's performance score during the instructional demonstration and using piecewise function matching to assess the user's mastery level of the knowledge points involved in the current teaching paradigm; and dynamically pushing matched personalized instructional content and / or advanced experimental tasks based on a preset brain-computer interface instructional knowledge graph, combined with the mastery level and the identified learning style type.

[0011] In some embodiments, when visualizing the teaching and explanation content, the method further includes: displaying the spatial distribution of real-time EEG data in different brain regions of the three-dimensional head model; when the similarity between the real-time EEG data and the preset EEG feature template is determined to be greater than or equal to a similarity threshold, spatially highlighting the corresponding brain region; and responding to user operations by presenting relevant neurophysiological function descriptions.

[0012] In some embodiments, when visualizing the teaching and explanation content, the method further includes: visualizing the decoding process of the original EEG signal; receiving a real-time algorithm parameter configuration instruction from the user for the brain-computer interface decoding model when visualizing the teaching and explanation content; responding to the algorithm parameter configuration instruction, synchronously updating the algorithm parameters in the decoding process, and based on the updated algorithm parameters, rendering the feature spectrum changes of real-time EEG data and the corresponding decoding performance changes in real-time during the visual demonstration of the decoding process.

[0013] In some embodiments, the method further includes: constructing the decoding process of the raw EEG signal into a hierarchical tree structure of nodes; recording the parameter configuration and intermediate output information of each processing node when visualizing the teaching explanation content in real time; and generating a decoding process status analysis report based on the recorded parameter configuration and intermediate output information.

[0014] In some embodiments, before loading the brain-computer interface decoding model corresponding to the teaching paradigm locally, the method includes: acquiring raw brain signals collected by a new user in a preset basic brain-computer task; and using a meta-learning model, determining initial values ​​of brain-computer interface decoding model parameters suitable for the new user based on the raw brain signals.

[0015] This disclosure also provides a brain-computer interface experimental device for multi-task teaching demonstrations, comprising: a signal acquisition module for acquiring raw EEG signals; a data processing and analysis module for loading a brain-computer interface decoding model corresponding to a currently selected teaching paradigm; inputting the acquired raw EEG signals into the brain-computer interface decoding model for decoding processing to generate real-time EEG data; an AI engine module for triggering teaching explanation content corresponding to the preset EEG feature template when the similarity between the real-time EEG data and the preset EEG feature template is determined to be greater than or equal to a similarity threshold; and a visual teaching interaction module for visually demonstrating the teaching explanation content.

[0016] This disclosure also provides an edge computing device, including a processor, a memory, and a computer program / instructions stored in the memory. The processor is used to execute the computer program / instructions, and when the computer program / instructions are executed, the above-described brain-computer interface experimental method for multi-task teaching demonstration is implemented.

[0017] This disclosure also provides a computer-readable storage medium storing a computer program, which, when executed, implements the above-described brain-computer interface experimental method for multi-task teaching demonstration.

[0018] Through the above technical solutions, the brain-computer interface experimental method, device, edge computing equipment, and storage medium for multi-task teaching demonstrations provided in this disclosure can dynamically present teaching explanations corresponding to the current processing state based on the actual changes in the characteristics of the brain signals during the brain signal decoding process. This allows the teaching demonstration process to be effectively linked to the real-time processing results of the brain signals, thereby enhancing the relevance and timeliness of the teaching explanations. Simultaneously, it can adapt to multiple brain-computer interface teaching paradigms within the same teaching process, facilitating the intuitive demonstration of brain signal characteristics and decoding differences under different paradigms. This helps learners build a holistic understanding of the brain-computer interface technology system, thereby improving the depth of understanding and teaching effectiveness of brain-computer interface teaching demonstrations. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of this disclosure or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart illustrating the brain-computer interface experimental method for multi-task teaching demonstration disclosed in this embodiment. Figure 2 This is a flowchart illustrating the teaching explanation triggering process based on sliding time window and similarity accumulation judgment disclosed in this embodiment of the present disclosure; Figure 3 This is a flowchart illustrating the adaptive optimization of the teaching demonstration presentation method based on user interaction behavior disclosed in this embodiment. Figure 4 This is a schematic diagram of the visual interface disclosed in the embodiments of this disclosure; Figure 5 It is the paradigm workflow interface in the visual interface disclosed in the embodiments of this disclosure; Figure 6 This is a schematic diagram of the incremental learning framework disclosed in an embodiment of this disclosure; Figure 7 This is a structural block diagram of the brain-computer interface experimental device for multi-task teaching demonstration disclosed in this embodiment; Figure 8 This is a schematic diagram of the workflow of the edge computing device disclosed in this embodiment. Detailed Implementation

[0021] The embodiments of this disclosure will be further described in detail below with reference to the accompanying drawings and examples. The detailed description of the embodiments and the accompanying drawings are used to illustrate the principles of this disclosure by way of example, but should not be used to limit the scope of this disclosure. This disclosure can be implemented in many different forms and is not limited to the specific embodiments disclosed herein, but includes all technical solutions falling within the scope of the claims.

[0022] These embodiments are provided to make the disclosure thorough and complete, and to fully express the scope of the disclosure to those skilled in the art. It should be noted that, unless otherwise specifically stated, the relative arrangement of components and steps, material composition, numerical expressions, and values ​​set forth in these embodiments should be interpreted as exemplary only and not as limiting.

[0023] All terms used in this disclosure have the same meaning as understood by one of ordinary skill in the art to which this disclosure pertains, unless otherwise specifically defined. It should also be understood that terms defined in general dictionaries should be interpreted as having meanings consistent with their meanings in the context of the relevant art, and not as idealized or highly formalized, unless expressly defined herein.

[0024] Techniques, methods, and equipment known to those skilled in the art may not be discussed in detail, but where appropriate, they should be considered part of the specification.

[0025] Figure 1 A flowchart illustrating a brain-computer interface experimental method for multi-task teaching demonstrations, provided for a disclosed embodiment, may include the following steps: Step S101: Based on the currently selected teaching paradigm, load the brain-computer interface decoding model corresponding to the teaching paradigm.

[0026] Before or during a teaching demonstration, determine the brain-computer interface teaching paradigm to be demonstrated. The teaching paradigm can be preset by the teacher or selected by other users (such as students) through a visual interface, indicating the brain-computer interface signal processing approach and decoding logic corresponding to this teaching demonstration.

[0027] After determining the teaching paradigm, a brain-computer interface (BCI) decoding model matching the teaching paradigm is loaded from a pre-configured set of BCI decoding models. These BCI decoding models are constructed according to different BCI paradigms, such as steady-state visual evoked potentials (SSVEP), event-related potentials (P300), and motor imagery (MI). Each model integrates a signal processing algorithm chain adapted to its corresponding paradigm, including a signal preprocessing unit for suppressing noise and interference, a feature extraction unit for highlighting target EEG patterns, and a decision-making unit for state or category determination, thus forming a complete decoding process.

[0028] Step S102: The acquired raw EEG signals are input into the brain-computer interface decoding model for decoding processing to generate real-time EEG data.

[0029] During the teaching demonstration, raw EEG signals generated by the user are continuously acquired. These raw EEG signals can be obtained in real time using an EEG acquisition device (such as an EEG cap). In practical applications, the raw EEG signals can be continuously acquired multi-channel EEG time-series data, with sampling frequency and signal format matching the loaded brain-computer interface decoding model.

[0030] The raw EEG signals are fed into the brain-computer interface decoding model in chronological order as input data for decoding processing. During the decoding process, based on the internal structure and parameters of the loaded decoding model, signal preprocessing, feature extraction, and discriminant analysis are performed on the raw EEG signals sequentially, thereby generating real-time EEG data reflecting the current state of brain electrical activity.

[0031] Step S103: When it is determined that the similarity between the real-time EEG data and the preset EEG feature template is greater than or equal to the similarity threshold, the teaching explanation content corresponding to the preset EEG feature template is triggered.

[0032] A pre-built library of EEG feature templates for teaching demonstrations is constructed. This library stores representative EEG feature templates from various typical brain-computer interface teaching paradigms, used to characterize EEG activity features with pedagogical significance in different paradigms. Examples include positive peak waveforms in the event-related potential paradigm, specific frequency energy spectra in the steady-state visual evoked potential paradigm, and μ / β rhythms in the motor imagery paradigm. Each EEG feature template is associated with its corresponding teaching explanation content for automatic triggering during teaching demonstrations.

[0033] After generating real-time EEG data, a similarity calculation is performed between the real-time EEG data and EEG feature templates in the EEG feature template library to assess the degree of matching between the current EEG activity state and preset typical EEG features. Similarity calculation can be based on time-domain waveform similarity, frequency-domain feature distribution consistency, or statistical feature correlation. When the similarity calculation result reaches or exceeds a preset similarity threshold, it is determined that a typical EEG feature with pedagogical significance has appeared in the current teaching demonstration, and the teaching explanation content associated with that EEG feature template is triggered.

[0034] The instructional explanations are used to help learners understand the background of the current EEG characteristics and their significance in the corresponding teaching paradigm. The explanations may include textual descriptions of the relevant neurophysiological mechanisms, brief explanations of the causes of EEG characteristics, and an index of course knowledge points related to the EEG characteristics, thereby providing learners with timely and targeted teaching information support without interrupting the teaching demonstration process.

[0035] Step S104: Visualize the teaching and explanation content.

[0036] After triggering instructional explanations corresponding to typical EEG features, these explanations are visualized, allowing learners to simultaneously acquire targeted instructional information during the EEG signal decoding and demonstration process. For example, the instructional explanations are displayed in a combined text and image format in the auxiliary display area of ​​the visualization interface. The text information explains the neuroscience principles, physiological significance, and role of the currently detected EEG features within the current instructional paradigm.

[0037] In practical applications, the explanation content corresponding to the current EEG characteristics can be presented in audio form according to the needs of teaching demonstrations. By playing pre-configured audio explanation segments, key concepts or core knowledge points can be emphasized and explained, thereby reducing learners' reliance on textual information and improving the efficiency of understanding the teaching demonstration process without affecting the main process of EEG signal demonstration.

[0038] The aforementioned visualization demonstration process can be carried out while maintaining the continuous presentation of the EEG signal processing and display main area, so as to avoid interfering with the teaching demonstration process and thus ensure the continuity and integrity of the brain-computer interface teaching demonstration.

[0039] In practical applications, after visualizing the teaching content, instructional demonstration commands generated based on real-time EEG decoding results can be output to external interactive devices to enhance the immersion and interactivity of the demonstrations. For example, control commands can be generated based on the current decoding state of EEG characteristics to drive changes in virtual reality scenes, allowing learners to intuitively perceive the mapping relationship between EEG activity and external behavioral feedback within the virtual environment. Alternatively, demonstration commands can be generated to control robotic arm models, virtual limb movements, or the playback rhythm of multimedia courseware, thereby transforming EEG decoding results into observable and understandable external representations.

[0040] Through the above steps, the multi-paradigm brain-computer interface teaching demonstration method disclosed herein can automatically trigger and visually present teaching explanations corresponding to the current brain-computer interface features based on the matching results between real-time brain-computer data and preset typical brain-computer interface features during the brain-computer signal decoding process, ensuring that teaching information is presented synchronously with the brain-computer signal processing process. This method helps learners intuitively understand the generation mechanism and teaching significance of brain-computer interface features under different teaching paradigms during actual demonstrations, and realizes the demonstration and explanation of multi-paradigm brain-computer interface features under a unified teaching process, thereby improving the overall effect and teaching value of brain-computer interface teaching demonstrations.

[0041] It should be noted that the original EEG signal in this embodiment can also be a simulated EEG signal generated by a relevant EEG generation circuit, so as to correspond one-to-one with the corresponding paradigm and match the teaching and explanation content.

[0042] In some embodiments of this disclosure, during teaching demonstrations where the original EEG signals change continuously, the EEG feature matching results at a single time point are easily affected by instantaneous noise or occasional fluctuations, leading to unstable triggering of the teaching explanation content. Therefore, embodiments of this disclosure also provide a teaching explanation triggering method based on a sliding time window and similarity accumulation judgment, such as... Figure 2 As shown, the following steps may be included: Step S201: Perform sliding time window analysis on the real-time EEG data to divide the real-time EEG data and generate EEG data segments.

[0043] Real-time EEG data is divided into sliding time windows of preset duration to form continuous EEG data segments. The sliding time window has a fixed time span (e.g., 2 seconds) and slides forward on the time axis according to a preset step size, so that there is temporal overlap between adjacent EEG data segments, thereby ensuring the integrity of EEG features during continuous temporal changes.

[0044] Step S202: Calculate the similarity between a preset number of consecutive EEG data segments and any preset EEG feature template in the preset feature library, and sum the calculated similarities.

[0045] After obtaining EEG data segments, a predetermined number of consecutive EEG data segments are selected as analysis units. The similarity of each EEG data segment in the analysis unit with a predetermined EEG feature template is calculated to evaluate the degree of matching between real-time EEG activity and typical EEG features.

[0046] In practical applications, similarity calculation can be achieved based on time series matching, frequency domain feature consistency, or dynamic time warping. For multiple EEG data segments in the analysis unit, the calculated similarity results are summed to form a cumulative similarity value that reflects the overall matching degree of EEG features within a continuous time period, thereby reducing the impact of abnormal fluctuations in a single time segment on the judgment result.

[0047] Step S203: When the sum of similarities is greater than or equal to the similarity threshold, the identified EEG data fragments are highlighted in the visualization interface, and related teaching explanations are pushed simultaneously.

[0048] After obtaining the cumulative similarity value, it is compared with a preset similarity threshold. When the cumulative similarity value reaches or exceeds the similarity threshold, it is determined that a typical EEG feature with educational significance has been stably observed within the current time interval.

[0049] If the determination is successful, the identified EEG data segment is highlighted in the visualization interface to intuitively present the timing of the appearance of typical EEG features in the decoding process. Simultaneously, instructional explanations associated with the EEG feature template are triggered to guide learners in understanding the formation mechanism of the EEG feature and its role in the current teaching paradigm.

[0050] In practical applications, to prevent teaching explanations from being repeatedly triggered in a short period of time, time suppression conditions or status flags can be set for already triggered teaching explanation events. This will reduce the interference of redundant prompts on the learning process while ensuring the continuity of the teaching rhythm.

[0051] In the above embodiments, by dividing real-time EEG data into sliding time windows and performing cumulative similarity judgment on continuous EEG data segments and preset EEG feature templates, the interference of occasional noise or transient fluctuations in a single time segment on the feature recognition results can be effectively suppressed, thereby improving the stability and reliability of typical EEG feature trigger judgment. Simultaneously, when the similarity reaches a threshold, the identified signal segments and the teaching explanation content are associated and presented at the corresponding algorithm processing nodes in the visualization interface, enabling learners to intuitively understand the generation location and processing logic of specific EEG features in the complete decoding process, thereby improving the coherence and teaching effectiveness of brain-computer interface teaching demonstrations.

[0052] In some embodiments of this disclosure, different learners exhibit significant differences in their understanding of the instructional content and their interaction preferences when learning brain-computer interface (BCI) teaching demonstrations. Using a uniform visualization method could easily lead to excessive comprehension burden or low learning efficiency for some learners. Therefore, embodiments of this disclosure also provide a multi-paradigm BCI teaching demonstration method that adaptively optimizes the teaching demonstration presentation based on user interaction behavior, such as... Figure 3 As shown, the following steps may be included: Step S301: Collect user interaction data when visualizing the teaching explanation content.

[0053] During the visualization demonstration of teaching content, user interactions are implicitly recorded, forming corresponding interactive behavior sequence data. This interactive behavior data reflects user operational preferences and learning habits during the teaching demonstration. It may include the proportion of time spent adjusting parameters, the number and frequency of directly initiating preset teaching demonstration flows, the number of clicks and dwell time on related knowledge point prompts in the teaching explanation, and the order and frequency of adjustments, replacements, or repetitions of processing steps during the decoding process of constructing or adjusting the teaching paradigm.

[0054] Step S302: Identify the user's learning style type based on interaction behavior data.

[0055] After collecting user interaction data during the visualization demonstration of teaching and explanation content, the interaction data can be preprocessed to eliminate abnormal operation records and invalid interaction information. Time normalization and frequency statistics processing can be performed on different types of interaction data to form a multi-dimensional set of behavioral features that can reflect user operation preferences.

[0056] Based on a multidimensional behavioral feature set, the interaction habits of users during teaching demonstrations are analyzed to characterize the degree of emphasis users place on different learning behaviors such as experimental operation, theoretical research, and following examples. In practical applications, behavioral features can be pre-analyzed offline based on historical user samples, and various representative learning style prototypes can be summarized through cluster analysis or similarity measurement.

[0057] In the actual identification process, the behavioral characteristics of the current user are matched with the learning style prototype to determine the user's learning style type. For example, it can identify an experimental exploration learning style that tends to frequently adjust parameters and actively construct experimental procedures, a theory-led learning style that tends to consult knowledge point descriptions and focus on principle explanations, and a follower-and-imitator learning style that mainly operates according to existing examples.

[0058] Step S303: Based on the identified learning style type, adaptively adjust the layout template and interactive presentation of the visualization interface, and push teaching guidance prompts that match the learning style type.

[0059] After identifying the user's learning style, the layout and interactive presentation of the visualization interface are adaptively adjusted according to the pre-set presentation strategies for different learning styles. This ensures that the information presentation areas, interactive control distribution, and explanatory content display methods match the current learning style. Specifically, for experimental and exploratory learners, the visualization interface prioritizes adjustable parameter areas and interactive content related to experimental process adjustments, allowing users to easily modify key parameters and observe changes in EEG decoding results. For theory-driven learners, the interface prioritizes presenting knowledge point indexes, principle explanations, and visual representations of algorithm processes related to the current EEG characteristics, enabling them to establish a complete theoretical understanding before operation. For imitative learners, the interface strengthens example process prompts and step-by-step guidance information to reduce operational complexity.

[0060] Simultaneously, during user interaction, matching instructional guidance prompts are dynamically pushed based on the user's learning style and current operation status. When a user's actions significantly change the decoding effect, instructional guidance prompts are output to explain the current operation result or guide optimization. For example, for experimental exploratory learners, when parameter settings cause a decrease in decoding performance, they can be prompted that the current parameter range may increase the probability of false triggers, and suggested to conduct comparative experiments within adjacent parameter ranges. For theory-oriented learners, under the same circumstances, they can be prompted that the current settings may cause a decrease in the model's generalization ability, and guided to consult theoretical knowledge points related to decision boundaries or feature distributions.

[0061] Through the technical solution, the multi-paradigm brain-computer interface teaching demonstration method disclosed herein can dynamically perceive the interactive behavior characteristics of users during the visualization demonstration of teaching explanation content, and identify the user's learning style type accordingly. This allows for adaptive adjustments to the layout of the teaching interface, the interactive presentation format, and the teaching guidance information, making the teaching demonstration process more in line with the cognitive habits and learning needs of different learners. This is beneficial for improving the relevance, participation, and overall teaching effectiveness of brain-computer interface teaching demonstrations.

[0062] In some embodiments of this disclosure, during multi-paradigm brain-computer interface (BCI) teaching demonstrations, different learners exhibit significant differences in their depth of understanding, operational proficiency, and mastery of key EEG features within the same teaching paradigm. If a uniform teaching guide and experimental task arrangement are consistently used, a mismatch between the teaching pace and the learners' actual abilities can easily occur. Therefore, the BCI experimental method of this disclosure further provides a method that, based on identifying learning style types, further refines the delivery of teaching guide content by combining learning performance. By quantitatively assessing the learner's current mastery level and combining it with a BCI teaching knowledge graph, personalized teaching content and advanced experimental tasks matching the learner's ability stage and learning style are dynamically generated, thereby improving the targeting and effectiveness of teaching guidance.

[0063] In this embodiment of the disclosure, pushing teaching guidance prompts that match the learning style type includes: calculating the user's performance score during the teaching demonstration process, and using a piecewise function matching method to assess the user's mastery level of the knowledge points involved in the current teaching paradigm; and dynamically pushing matched personalized teaching content and / or advanced experimental tasks based on a preset brain-computer interface teaching knowledge graph, combined with the mastery level and the identified learning style type.

[0064] Specifically, during the teaching demonstration, learners' operational results and interactive feedback under different teaching paradigms are continuously collected, and their learning performance is quantitatively evaluated based on the collected data. For example, a comprehensive analysis of multiple performance parameters, such as learners' completion of experimental tasks, EEG signal quality indicators, stability of decoding results, classification accuracy, or target recognition success rate, can be used to form a performance score reflecting the learner's current learning status. This performance score can be mapped using a pre-defined piecewise function to convert the continuous numerical performance results into discrete mastery level intervals, which characterize the learner's degree of mastery of the current teaching paradigm and its related knowledge points.

[0065] After assessing the learner's mastery level, the learning content is further matched using a pre-constructed brain-computer interface (BCI) teaching knowledge graph. This knowledge graph describes the relationships between different concepts, principles, EEG characteristics, experimental tasks, and key skills in the BCI field, enabling the organization and retrieval of knowledge content from different teaching paradigms within a unified structure. Based on this, and according to the learner's mastery level and identified learning style, teaching content or experimental tasks matching the current learning status are selected from the teaching knowledge graph and pushed out as personalized teaching content and / or advanced experimental tasks.

[0066] For example, when the assessment results show that learners have a low completion rate of relevant experimental tasks under the motor imagery paradigm and that the μ rhythm features in the EEG signals are not obvious, targeted supplementary teaching content can be pushed by combining the correlation between the physiological mechanisms of motor imagery, attention regulation methods and feature enhancement principles in the teaching knowledge graph. This includes explanations of the neurophysiological basis of motor imagery, or interactive practice tasks to improve concentration and consistency of imagination, so as to help learners consolidate their basic cognition.

[0067] For example, when learners perform well in the steady-state visual evoked potential paradigm and are proficient in target recognition tasks under basic stimulus frequency conditions, more challenging experimental tasks can be pushed to learners based on the association paths of paradigm progression experiments in the knowledge graph. These tasks could include increasing the stimulus frequency range, introducing multi-target selection scenarios, or comparing decoding effects under different frequency conditions, thereby guiding learners to expand their abilities based on their existing knowledge.

[0068] Furthermore, based on the cross-paradigm relationships in the teaching knowledge graph and the learner's learning style, comparative teaching content between different paradigms can be pushed. For example, after learners have completed the basic learning of a certain paradigm, they can be guided to further understand the similarities and differences between different brain-computer interface paradigms in terms of attention mechanisms, stimulus response characteristics, or decoding principles, so as to promote their systematic understanding of the overall theoretical system of brain-computer interfaces.

[0069] Through the above technical solutions, the multi-paradigm brain-computer interface teaching demonstration method provided in this disclosure can, on the basis of identifying learning styles, further quantify the learner's mastery level by combining the learner's actual performance, and realize the dynamic matching and push of teaching guidance content based on the brain-computer interface teaching knowledge graph, so that the teaching content can be adaptively adjusted according to the learner's ability changes, thereby avoiding the problem of mismatch between teaching difficulty and learning status, which is conducive to improving the teaching effect and learning efficiency of multi-paradigm brain-computer interface teaching demonstration.

[0070] In some embodiments of this disclosure, addressing the difficulty learners face in intuitively understanding the brain region activity distribution and corresponding neurophysiological significance reflected by real-time EEG data during brain-computer interface teaching demonstrations, this disclosure also provides a teaching demonstration method that dynamically presents the decoding and processing flow of raw EEG data in a flowchart format within a visual interface. Figure 4 The teaching demonstration method that incorporates a three-dimensional human head model in the visualization interface shown in this disclosure also includes: displaying the spatial distribution of real-time EEG data in different brain regions in the three-dimensional human head model; highlighting the corresponding brain region spatially when the similarity between the real-time EEG data and the preset EEG feature template is greater than or equal to a similarity threshold; and responding to user operations by presenting relevant neurophysiological function descriptions.

[0071] Specifically, in the process of visualizing the teaching and explanation content, based on a pre-constructed standard 3D human head model, real-time EEG data is mapped to different brain regions within the 3D human head model to demonstrate the spatial distribution of real-time EEG data. The activity intensity of different brain regions is differentiated and displayed through color changes, brightness differences, or texture overlays to intuitively show the spatial variations of EEG signals.

[0072] During the teaching demonstration, when the similarity between real-time EEG data and a preset EEG feature template is greater than or equal to a similarity threshold, it is considered that a typical EEG feature with teaching significance has stably appeared within the current time interval. When this condition is met, and a teaching event corresponding to a particular typical EEG feature is detected, the target brain region primarily involved in that typical EEG feature is spatially highlighted in a 3D human head model. This highlights the neural activity state of that brain region under the current teaching paradigm, thereby guiding learners to associate the recognition result of the typical EEG feature with its corresponding brain region location for understanding.

[0073] Simultaneously, based on the highlighted brain regions, the system responds to learners' actions on the visualization interface by presenting explanations of the neurophysiological functions related to the highlighted brain regions. These explanations may include the functional role of the brain region within the current teaching paradigm, explanations of the neural activity mechanisms related to the identified EEG characteristics, and association prompts with corresponding knowledge points in the course syllabus. This provides learners with theoretical support information directly related to spatial brain region activity without interrupting the teaching demonstration process.

[0074] In the above embodiments, while visually demonstrating the teaching and explanation content, a real-time spatial distribution display of EEG data based on a three-dimensional human head model is introduced, and the corresponding brain regions are spatially highlighted when typical EEG features are judged, so that learners can intuitively perceive the distribution characteristics of EEG features in different brain regions and their neurophysiological significance, thereby improving the intuitiveness and teaching effect of multi-paradigm brain-computer interface teaching demonstrations.

[0075] In some embodiments of this disclosure, during brain-computer interface (BCI) teaching demonstrations, algorithm parameter configuration is typically done offline or based on preset parameters. Learners find it difficult to intuitively perceive the real-time impact of algorithm parameter changes on the real-time EEG data processing results and decoding performance, thus limiting their in-depth understanding of the BCI algorithm mechanism. Therefore, embodiments of this disclosure provide support for real-time algorithm parameter interaction and result feedback during the visualization demonstration of teaching content. The method further includes: visualizing the decoding process of the original EEG signal; receiving real-time algorithm parameter configuration instructions from the user for the BCI decoding model during the visualization demonstration of the teaching content; responding to the algorithm parameter configuration instructions, synchronously updating the algorithm parameters in the decoding process; and, based on the updated algorithm parameters, rendering the feature spectrum changes of the real-time EEG data and the corresponding decoding performance changes in real-time during the visualization demonstration of the decoding process.

[0076] Specifically, when visualizing teaching and explanation content, the decoding and processing flow of raw EEG signals can also be visualized, so as to form a visual interface such as Figure 4 The system provides a panoramic visualization of the algorithm flow, enabling learners to intuitively understand the decoding process structure and data flow relationships. The visualization of the decoding process is used to dynamically display the complete data processing pipeline from raw EEG signal input to decoded result output in real time. In its implementation, a graphics rendering engine can be used to render and display the decoding process in real time, showing the real-time EEG data representation at each process node. For example, the time-domain waveform of real-time EEG data can be plotted in real time at the raw data or preprocessing node; the changes in feature vectors or feature spectra can be dynamically displayed at the feature extraction node; and the current decoding confidence or output status can be presented as a progress bar or numerical value at the decoding and discrimination node.

[0077] In practical applications, users are allowed to use methods such as... during teaching demonstrations. Figure 5The visualization interface shown, featuring a paradigm workflow, allows for interactive adjustments to the decoding process steps. For example, by dragging or reconnecting process nodes, the order or combination of raw EEG signal processing can be changed, or a specific processing step can be enabled, disabled, or have its parameters modified interactively. After the workflow adjustment, the updated decoding process immediately takes effect on subsequent real-time EEG data analysis, enabling learners to visually observe the impact of changes in the workflow structure on similarity assessment results and the timing of instructional explanations.

[0078] During the teaching demonstration, users are allowed to, for example... Figure 4 The parameter interaction sandbox in the visualization interface allows for interactive configuration of decoding-related algorithm parameters for the currently selected brain-computer interface teaching paradigm. These parameter configuration operations can include adjusting signal preprocessing parameters, feature extraction parameters, or decoding discrimination parameters. For example, parameters such as the filtering bandwidth range of the original EEG signal, the length of the feature calculation time window, the classification threshold, or the discrimination weight can be modified.

[0079] When a user performs a parameter configuration operation, the corresponding parameter change content is parsed into an algorithm parameter configuration instruction, and the time sequence of parameter adjustments and the corresponding parameter values ​​are recorded.

[0080] Upon receiving the algorithm parameter configuration instruction, the system performs real-time calculations on the real-time EEG data feature results generated in key stages of the decoding process based on the updated algorithm parameters, and synchronously renders the feature spectrum changes in the visualization demonstration of the decoding process. Simultaneously, corresponding to the updated decoding results, the system calculates and displays the changing trends of relevant decoding performance indicators to reflect the impact of parameter configuration on decoding accuracy or stability.

[0081] In practical applications, while enabling real-time parameter interaction and result feedback, the system can simultaneously monitor the real-time EEG data quality, the rationality of parameter values, and the operational status of the decoding process. When increased noise, decreased signal-to-noise ratio, or enhanced interference is detected in the real-time EEG data, corresponding prompts are triggered without affecting the continuity of the demonstration. This guides learners to pay attention to changes in signal quality and their impact on the decoding results, and allows for appropriate adjustments to preprocessing parameters to improve signal stability. For example, when the user adjusts algorithm parameters beyond the preset reasonable range, resulting in significantly abnormal decoding results, the relevant parameter values ​​are highlighted, and targeted adjustment suggestions are provided to help learners understand the relationship between parameter settings and algorithm performance.

[0082] If data interruption or operational abnormality occurs during the teaching demonstration, the update based on real-time EEG data can be paused and the demonstration can be switched to preset example data to continue. At the same time, learners can be prompted to pay attention to the cause of the abnormality, thereby ensuring the continuity of the teaching process while enhancing their understanding of the actual application scenario.

[0083] In the above embodiments, by allowing users to configure key algorithm parameters of the brain-computer interface decoding model in real time during the teaching demonstration, and synchronously applying parameter changes to the decoding process, while simultaneously presenting the EEG feature spectrum of real-time EEG data and the results of decoding performance changes, learners can intuitively observe the impact of algorithm parameters on the decoding process. This approach helps transform abstract algorithmic principles into a perceptible and comparable dynamic demonstration process, enhancing learners' understanding of the brain-computer interface algorithm mechanism and their sense of participation in experiments, thereby improving the interactivity and teaching effectiveness of brain-computer interface teaching demonstrations.

[0084] In some embodiments of this disclosure, learners often only observe the processing results of the original EEG signals in a single experimental process, making it difficult to systematically retrospectively analyze and compare intermediate results under different algorithm parameter configurations or different processing paths. This hinders their understanding of the impact of each processing step on the final decoding effect. Therefore, embodiments of this disclosure also provide a brain-computer interface experimental method. During the teaching demonstration, the decoding processing flow of the original EEG signals is constructed into a hierarchical tree structure of nodes, and the parameter configuration and intermediate output information of each processing node are recorded in real time. This supports the generation of a decoding processing flow status analysis report after the teaching demonstration, thereby enhancing the analyzability and reproducibility of the teaching demonstration. The method further includes: constructing the decoding processing flow of the original EEG signals into a hierarchical tree structure of nodes; recording the parameter configuration and intermediate output information of each processing node in real time when visualizing the teaching explanation content; and generating a decoding processing flow status analysis report based on the recorded parameter configuration and intermediate output information.

[0085] Specifically, the decoding process used in the current teaching paradigm is broken down according to the processing order and dependencies, and organized hierarchically into a tree-like structure. Upper-level nodes represent the main stages of signal processing, while lower-level nodes represent the specific processing steps within those stages. For example, the tree-like structure may sequentially include acquisition nodes representing the raw EEG signal input, filtering nodes representing bandwidth limiting or noise suppression operations, feature nodes representing feature extraction operations, and classification nodes representing the discrimination or decision-making process. The parent-child relationships between nodes reflect the dependencies within the decoding process.

[0086] During the teaching demonstration, for each processing node in the tree structure, the parameter configuration used by that node in the current decoding process and the corresponding intermediate output information are recorded in real time. The parameter configuration includes, but is not limited to, the filtering bandwidth range, feature calculation window length, threshold settings, or discrimination strategy selection. The intermediate output information includes, but is not limited to, the filtered signal waveform, feature vector distribution, or intermediate discrimination results. Through this method, a traceable status record is created for each processing node during the teaching demonstration.

[0087] After the teaching demonstration, based on the parameter configurations and intermediate output information recorded in the tree structure nodes, the process status of the decoding process under different parameter settings or different processing paths is summarized and analyzed, and a decoding process status analysis report is generated. The output decoding process status analysis report presents the impact of changes in parameters at different processing nodes on intermediate results and the final decoding effect in a comparative form, thereby intuitively demonstrating the correlation between each link in the decoding process and its contribution to the teaching paradigm demonstration results.

[0088] Based on tree-structured modeling and state recording of the decoding process, the process state information generated during the teaching demonstration can be structurally archived to support subsequent review and reuse in teaching. For each teaching experiment, a unified structured data record is constructed, whose data structure includes at least four elements: experimental objective, operation step sequence, state data snapshot, and learning effect index. Among them, the experimental objective is used to characterize the teaching paradigm and teaching intention corresponding to the current teaching demonstration; the operation step sequence is used to describe the actual decoding process and parameter adjustment behavior executed during the teaching demonstration; the state data snapshot is used to record the state of the original EEG signal or intermediate analysis results at each key processing node; and the learning effect index is used to reflect the decoding effect or teaching feedback results.

[0089] When the teaching experiment process is started, the corresponding structured record is automatically created, and the user's operation behavior is continuously tracked during the teaching demonstration. Operations such as parameter modification and process switching are transformed into reproducible step descriptions. At the same time, relevant status data is synchronously collected and stored at preset processing nodes to form a snapshot of status data corresponding to each operation step, thereby ensuring the complete retention of the experimental process and intermediate states.

[0090] After the structured records are completed, the structured records are automatically labeled based on the experimental objectives, process characteristics, and status data content, and an index relationship is established so that experimental processes formed under different teaching paradigms, different parameter configurations, or different signal quality conditions can be effectively distinguished and retrieved.

[0091] After the teaching demonstration, based on structured records and tag indexes, different experimental procedures formed under the same experimental objective are compared and analyzed, generating reproducible process analysis results. For example, teacher users can select typical cases, insert them into lesson plans with one click, and directly replay the complete process and intermediate states of the experiment during explanation. Student users can compare state data snapshots and learning effect indicators under different parameters in their personal learning history to generate visual comparison reports to assist in debriefing.

[0092] Through the above technical solution, the brain-computer interface experimental method provided in this disclosure can express the decoding process in a hierarchical and structured manner, and record the parameter configuration and intermediate state of each processing node in a complete manner. This expands the teaching demonstration process from a single result display to a traceable and comparable analysis process, which is conducive to learners' systematic understanding of the impact mechanism of different parameter settings and processing strategies on the decoding performance of brain-computer interfaces, thereby improving the interpretability, reproducibility and teaching depth of brain-computer interface teaching demonstrations.

[0093] In some embodiments of this disclosure, to address the issue that decoding models typically require a lengthy individual calibration process and struggle to quickly adapt to individual EEG differences when new users participate in multi-paradigm brain-computer interface (BCI) teaching demonstrations for the first time, thus affecting teaching continuity and user experience, this disclosure also introduces a meta-learning mechanism for rapid individual initialization of model parameters. This mechanism infers initial values ​​of model parameters suitable for new users using a small amount of basic EEG task data, enabling rapid startup and individual adaptation of BCI teaching demonstrations. Before loading the BCI decoding model corresponding to the teaching paradigm, the method includes: acquiring raw EEG signals collected by the new user in a preset basic EEG task; and using a meta-learning model, determining initial values ​​of BCI decoding model parameters suitable for the new user based on the raw EEG signals.

[0094] Specifically, before loading the brain-computer interface decoding model corresponding to the teaching paradigm, new users are first guided to complete a preset basic EEG task to obtain raw EEG signals for individual adaptation. In practical applications, the basic EEG task can be a standardized task format related to the target teaching paradigm, such as a simple target stimulus identification task for the event-related potential paradigm, or a short-term visual fixation task for the steady-state visual evoked potential paradigm, thereby acquiring raw EEG signals that can reflect individual EEG characteristics in a relatively short time.

[0095] After obtaining the raw EEG signals, a pre-trained meta-learning model is used to perform rapid inference processing on the raw EEG signals. The meta-learning model is trained offline based on a large number of raw EEG signals from historical users. During its training, it uses a cross-user, multi-task learning approach to extract the intrinsic correlation between the differences in EEG features between different individuals and the model parameters, thereby forming a parameter initialization mapping capability with rapid adaptability.

[0096] During the actual adaptation phase, a small amount of raw EEG signals collected from the new user during a basic EEG task are input into the meta-learning model. This model infers the initial parameter values ​​of the brain-computer interface decoding model suitable for the new user. The output initial parameter values ​​are used as the starting parameter configuration when loading the brain-computer interface decoding model later, so that the decoding model has a certain degree of individual matching ability in the initial state, without having to go through a complete and time-consuming traditional calibration process.

[0097] After parameter initialization is completed, the brain-computer interface decoding model corresponding to the current teaching paradigm is loaded, and the raw EEG signals collected subsequently are decoded based on the inferred initial parameter values, thereby ensuring the accuracy of decoding while enabling the rapid start and continuous conduct of the teaching demonstration process.

[0098] In practical applications, after initializing the parameters for new users and loading the brain-computer interface decoding model corresponding to the teaching paradigm, the EEG feature data generated during subsequent teaching demonstrations can be continuously accumulated for incremental optimization of the decoding model. Figure 6 A schematic diagram of the incremental learning framework is shown. In this framework, the EEG feature representations generated during the teaching demonstration are anonymized and characterized. Under the premise of meeting data security and privacy protection requirements, the anonymized feature data is uploaded to a cloud server. The cloud server performs aggregate analysis on the anonymized feature data and iteratively updates the parameters of the decoding model to form a parameter configuration with stronger generalization capabilities. After the model parameters are updated, they are distributed, allowing the updated parameters to serve as a new initialization reference. This enables the model to gradually absorb statistical characteristics from multiple users and scenarios, thereby forming a decoding parameter configuration with greater generalization capabilities.

[0099] Through the above implementation methods, the method provided by this disclosure can achieve rapid individualized initialization of the brain-computer interface decoding model by utilizing the meta-learning mechanism before new users participate in teaching demonstrations, effectively shortening individual calibration time, reducing the learning threshold for new users, and improving the stability and availability of decoding processing in the early stages of teaching. Thus, in multi-paradigm brain-computer interface teaching demonstration scenarios, it takes into account teaching continuity, individual adaptability, and overall teaching efficiency.

[0100] This disclosure also provides a brain-computer interface experimental device for multi-task teaching demonstrations, used to demonstrate and explain the decoding process under different brain-computer interface teaching paradigms. Its overall structure is as follows: Figure 7 As shown, it includes a signal acquisition module, a data processing and analysis module, an AI engine module, and a visual teaching interaction module.

[0101] The signal acquisition module is used to acquire raw EEG signals generated by the user during the teaching demonstration. The acquired raw EEG signals can be multi-channel continuous time-series data, and their sampling method and signal format meet the requirements of subsequent decoding processing, providing a basic data source for the teaching demonstration.

[0102] The data processing and analysis module is used to load the corresponding brain-computer interface decoding model according to the currently selected teaching paradigm and perform decoding processing on the acquired raw EEG signals. Specifically, this module can sequentially perform preprocessing, feature extraction, and discriminant analysis on the raw EEG signals according to the processing logic corresponding to different teaching paradigms, thereby generating real-time EEG data reflecting the current state of brain activity and providing a basis for triggering subsequent teaching explanations.

[0103] The AI ​​engine module is used to analyze and judge real-time EEG data. When the similarity between real-time EEG data and a preset EEG feature template reaches or exceeds a preset threshold, the teaching explanation content corresponding to that EEG feature template is triggered, thereby automatically identifying typical EEG features with teaching significance during the teaching demonstration.

[0104] The visual teaching interaction module is used to visualize the triggered teaching explanations, enabling learners to simultaneously acquire teaching information related to the current EEG characteristics during the decoding process and demonstration, thereby enhancing their understanding of the principles of brain-computer interfaces and the meaning of EEG characteristics.

[0105] In addition, in some embodiments, the device may also include a control output module for outputting control commands or demonstration results generated during the teaching demonstration to external devices, such as virtual reality scenes, interactive teaching terminals, or multimedia courseware, to support further presentation or extended application of the teaching demonstration results.

[0106] For specific limitations of each module in the brain-computer interface experimental device for multi-task teaching demonstrations, please refer to the limitations of the brain-computer interface experimental method for multi-task teaching demonstrations mentioned above, which will not be repeated here.

[0107] This disclosure also provides an edge computing device, including a processor, a memory, and a computer program / instructions stored in the memory. The processor is used to execute the computer program / instructions, and when the computer program / instructions are executed, a brain-computer interface experimental method for multi-task teaching demonstration is implemented.

[0108] Based on the specific implementation of the aforementioned modules, the operation of the edge computing device can be achieved through... Figure 8 The workflow shown is clearly presented, and it can be summarized as the following continuous work process.

[0109] During the device startup phase, once the user wears and connects to the user-end brain-computer interface (BCI) sensing device, the device is triggered to start and loads the BCI decoding model corresponding to the current teaching paradigm. This BCI decoding model is continuously trained and optimized by a cloud server and can be distributed to the terminal for use after model version updates, thus ensuring the dynamic evolution of the model's capabilities.

[0110] After the device enters the operating state, it first determines whether the current user is a new user. If it is determined to be a new user, the device automatically enters the meta-learning rapid calibration process, which guides the user to complete preset basic EEG tasks to quickly establish the initial parameters of the brain-computer interface decoding model adapted to individual characteristics; if it is determined to be a non-new user, it directly loads the established personalized model parameters as the starting configuration of the brain-computer interface decoding model and enters the formal interaction stage.

[0111] Then, the core teaching demonstration begins. During the interaction, the user performs brain-computer interface tasks according to the teaching or training requirements, such as performing brain cognitive activities like motor imagery. The device simultaneously collects the corresponding raw EEG signals and decodes the raw EEG signals on the terminal side based on the EEG interface decoding model.

[0112] After completing signal acquisition and processing, the device uses its loaded AI engine to perform online analysis of real-time EEG data and provides users with real-time teaching explanations, task feedback, or knowledge point pushes based on the analysis results. Simultaneously, the device displays the decoding and processing flow and the corresponding real-time EEG data changes through a visual interface to enhance the user's understanding of the interaction process and learning content.

[0113] The device then evaluates the confidence level of the current decoding result. If the confidence level is below a preset threshold, the device outputs a status message to the user, indicating that the current state is unsatisfactory and suggesting adjustments. If the confidence level reaches or exceeds the threshold, the device generates corresponding control outputs based on the decoding result, such as triggering interactive feedback in the virtual reality scene or executing corresponding teaching demonstration actions. This completes an effective human-computer interaction loop.

[0114] During the aforementioned operation, when the device is idle or reaches a scheduled time, it uploads anonymized feature data to the cloud server. The cloud server integrates and analyzes the collected anonymized feature data, continuously runs incremental learning algorithms to generate optimized new brain-computer interface decoding models, and distributes the updated models to various terminals, thereby achieving continuous improvement in model performance.

[0115] When the course ends or the user actively logs out, the device terminates the current session and releases resources, ending the current workflow.

[0116] The embodiments of this disclosure have now been described in detail. To avoid obscuring the concept of this disclosure, some details known in the art have not been described. Those skilled in the art can fully understand how to implement the technical solutions disclosed herein based on the above description.

[0117] While specific embodiments of this disclosure have been described in detail by way of examples, those skilled in the art should understand that the examples are for illustrative purposes only and not intended to limit the scope of this disclosure. Those skilled in the art should understand that modifications can be made to the above embodiments or equivalent substitutions can be made to some technical features without departing from the scope and spirit of this disclosure. In particular, as long as there is no structural conflict, the technical features mentioned in the various embodiments can be combined in any manner.

Claims

1. A brain-computer interface experimental method for multi-task teaching demonstrations, characterized in that, include: Based on the currently selected teaching paradigm, load the brain-computer interface decoding model corresponding to the teaching paradigm; The acquired raw EEG signals are input into the brain-computer interface decoding model for decoding processing to generate real-time EEG data; When the similarity between the real-time EEG data and the preset EEG feature template is determined to be greater than or equal to the similarity threshold, the teaching and explanation content corresponding to the preset EEG feature template is triggered. The teaching explanation content is presented visually.

2. The method according to claim 1, characterized in that, When the similarity between the real-time EEG data and the preset EEG feature template is determined to be greater than or equal to a similarity threshold, the teaching explanation content corresponding to the preset EEG feature template is triggered, including: A sliding time window analysis is performed on the real-time EEG data to divide the real-time EEG data and generate EEG data segments; A preset number of consecutive EEG data segments are compared with any preset EEG feature template in a preset feature library to calculate similarity, and the calculated similarities are summed. When the sum of similarities is greater than or equal to the similarity threshold, the identified EEG data fragments are highlighted in the visualization interface, and related teaching explanations are pushed simultaneously.

3. The method according to claim 1, characterized in that, When visually demonstrating the teaching and explanation content, the method further includes: Collect user interaction data when visualizing the teaching and explanation content; Identify the user's learning style type based on the aforementioned interactive behavior data; Based on the identified learning style type, the layout template and interactive presentation of the visualization interface are adaptively adjusted, and teaching guidance prompts matching the learning style type are pushed.

4. The method according to claim 3, characterized in that, The pushed instructional guidance prompts that match the learning style type include: Calculate user performance scores during the teaching demonstration process, and use piecewise function matching to assess users' mastery of the knowledge points involved in the current teaching paradigm; Based on a pre-defined brain-computer interface teaching knowledge graph, combined with the mastery level and the identified learning style type, the system dynamically pushes matching personalized teaching content and / or advanced experimental tasks.

5. The method according to claim 1, characterized in that, When visually demonstrating the teaching and explanation content, the method further includes: Based on a three-dimensional human head model, the spatial distribution of the real-time EEG data in different brain regions of the three-dimensional human head model is shown. When the similarity between the real-time EEG data and the preset EEG feature template is determined to be greater than or equal to the similarity threshold, the corresponding brain region is spatially highlighted. In response to user actions, present relevant neurophysiological function descriptions.

6. The method according to claim 1, characterized in that, When visually demonstrating the teaching and explanation content, the method further includes: A visual demonstration of the decoding process of the raw EEG signals is provided. Receive real-time algorithm parameter configuration instructions for the brain-computer interface decoding model from the user when visually demonstrating the teaching and explanation content; In response to the algorithm parameter configuration command, the algorithm parameters in the decoding process are updated synchronously, and based on the updated algorithm parameters, the feature spectrum changes of the real-time EEG data and the corresponding decoding performance changes are rendered in real time in the visualization demonstration of the decoding process.

7. The method according to claim 1 or 6, characterized in that, The method further includes: The decoding process of the original EEG signal is constructed as a hierarchical tree structure node; Real-time recording of parameter configurations and intermediate output information of each processing node when visualizing the teaching and explanation content; Based on the recorded parameter configurations and intermediate output information, a decoding processing flow status analysis report is generated.

8. The method according to claim 1, characterized in that, Before loading the brain-computer interface decoding model corresponding to the teaching paradigm, the method includes: Acquire raw EEG signals from new users during preset basic EEG tasks; Using a meta-learning model, initial values ​​for the brain-computer interface decoding model parameters suitable for the new user are determined based on the original EEG signals.

9. A brain-computer interface experimental device for multi-task teaching demonstration, characterized in that, include: The signal acquisition module is used to acquire raw electroencephalogram (EEG) signals. The data processing and analysis module is used to load the brain-computer interface decoding model corresponding to the currently selected teaching paradigm; input the acquired raw EEG signals into the brain-computer interface decoding model for decoding processing, and generate real-time EEG data; The AI ​​engine module is used to trigger teaching and explanation content corresponding to the preset EEG feature template when the similarity between the real-time EEG data and the preset EEG feature template is greater than or equal to a similarity threshold. The visual teaching interaction module is used to visually demonstrate the teaching explanation content.

10. An edge computing device, characterized in that, The method includes a processor, a memory, and a computer program / instructions stored in the memory, characterized in that the processor is configured to execute the computer program / instructions, and when the computer program / instructions are executed, implement the steps of the method as described in any one of claims 1 to 8.

11. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 8.

Images