Dynamic dot-spatial reference joint modulation ssvep bci stimulation method

Abstract

The application discloses a dynamic dot array-space reference joint modulation SSVEP brain-computer interface stimulation method, which solves the technical problems of low user comfort, few stimulation frequency resources, and great harmonic interference in the SSVEP brain-computer interface system. The method comprises the following steps: initializing a display environment, constructing a multi-position visual stimulation space layout, generating a flickering stimulation dynamic dot array texture sequence, integrating a visual focus guide structure, performing frame-level synchronization and frequency accurate control, performing joint modulation stimulation, and terminating the program. The multi-position space layout has a main stimulation area as a center and auxiliary points in the periphery, a sine function is used to modulate the lighting of the main stimulation area, the visual focus guide structure is integrated, and frame-level synchronization and frequency accurate control are realized. The method significantly reduces visual fatigue and improves gaze stability, guarantees high recognition performance, overcomes the frequency interference problem existing in the multi-frequency stimulation system, and enhances the comfort, practicality and scalability of the SSVEP brain-computer interface system.

Description

Technical Field

[0001] This invention belongs to the field of brain-computer interface technology, and mainly relates to visual stimulus presentation methods, specifically a dynamic dot matrix-spatial reference joint modulation SSVEP brain-computer interface stimulation method. Visual stimulation is achieved through the joint modulation of dynamic dot matrix textures and spatial reference markers, improving user comfort, gaze stability, and signal recognition accuracy of the SSVEP brain-computer interface system, and is applicable to the design and application of the SSVEP brain-computer interface system. Background Technology

[0002] Steady-state visual evoked potentials (SSVEPs) are electroencephalographic responses generated in the visual cortex of the brain when the human eye fixates on a visual stimulus flashing at a fixed frequency, synchronized with the stimulus frequency and its harmonics. SSVEP brain-computer interfaces have wide applications in neurorehabilitation, intelligent control, and human-computer interaction due to their advantages such as high signal strength, short training time, and fast information transmission rate. However, traditional SSVEP stimulation methods often employ large-area color blocks or checkerboard patterns with overall brightness and darkness reversal, which presents the following problems: First, drastic changes in light intensity easily cause visual fatigue and neurological discomfort; second, the lack of an effective visual guidance mechanism makes it easy for the user's gaze to drift, leading to signal attenuation; third, simultaneous flashing of multiple stimulus sources easily causes visual competition and distraction; and fourth, the stimulation frequency is limited by the display refresh rate, easily resulting in frequency drift and affecting decoding accuracy.

[0003] Suzhou Nianji Intelligent Technology Co., Ltd. disclosed an SSVEP brainwave signal recognition system in its patent application "Brain-Computer Interface System and Device Based on Steady-State Visual Evoked Potentials" (Application No.: CN202210594197.0, Publication No.: CN114916946 A). The method includes the following steps: (1) acquiring the brainwave signals of the subject through an electrode device; (2) displaying stimulation signals to the subject based on the SSVEP signals using a visual device, wherein the stimulation signals include at least the stimulation time frequency (3–15Hz) and stimulation pattern signals (such as random lines or checkerboard patterns); (3) receiving brainwave feedback and / or outputting SSVEP signals through a control device. This scheme improves visual comfort to a certain extent by optimizing the stimulation pattern (such as 1×6 pixel random lines or 6 pixel checkerboard grids) and contrast control (≥0.5). The shortcomings of this method are that it still relies on multi-frequency coding and overall area flickering stimulation, which fails to fundamentally solve the problems of visual fatigue and gaze stability. Furthermore, the stimulation frequency is limited by the display refresh rate, and the actual frequency is prone to drift, affecting recognition accuracy and system practicality.

[0004] In his paper “High-Frequency SSVEP Stimulation ParadigmBased On Dual Frequency Modulation” (DOI: 10.1109 / EMBC.2019.8856903), Liyan Liang proposed a high-frequency SSVEP stimulation paradigm based on dual-frequency modulation. The method includes the following steps: (1) presenting checkerboard visual stimuli using a 144Hz refresh rate display and encoding them using a dual-frequency combination (e.g., 30Hz and 34Hz) to expand the number of targets; (2) preloading stimulus sequences through the Psychtoolbox to achieve precise temporal control; (3) acquiring 9-channel occipital EEG signals and performing bandpass filtering (2–90 Hz); (4) using Task Relevant Component Analysis (TRCA) based on individual templates combined with a filter bank strategy for target recognition to address the frequency domain overlap and intermodulation interference problems caused by dual-frequency modulation. The shortcomings of this method are: although dual-frequency modulation expands the coding space, it leads to the aliasing of frequency templates between targets, and the recognition heavily relies on supervised training rather than unsupervised methods; at the same time, the overall flashing mode of the checkerboard pattern is still prone to causing visual fatigue, and the stimulation frequency is concentrated in the high-frequency band (30-38Hz), which has high requirements for individual response characteristics, limited applicability, and the system implementation is also more complicated.

[0005] In summary, existing SSVEP brain-computer interface stimulation methods generally employ large-area overall light-dark reversal or fixed checkerboard patterns, resulting in drastic changes in light intensity that easily induce visual fatigue, dizziness, and other discomfort in subjects, thus affecting the accuracy of the collected data. Traditional multi-target stimulation paradigms rely on the synchronous flickering of multiple independent frequency sources, which not only consumes significant frequency resources and easily induces harmonic interference but also leads to visual competition and attentional distraction, reducing signal quality and system stability. The presentation of stimulation frequencies heavily depends on integer frequency division of the display refresh rate; general-purpose graphics systems struggle to achieve precise frame synchronization, leading to frequency drift that affects the quality of the evoked signal and the accuracy of subsequent decoding algorithms. Summary of the Invention

[0006] The purpose of this invention is to address the shortcomings of the prior art by proposing a dynamic dot-spatial reference joint modulation SSVEP brain-computer interface stimulation method that offers high user comfort, strong gaze stability, and high signal recognition accuracy.

[0007] This invention is a dynamic dot-spatial reference joint modulation SSVEP brain-computer interface stimulation method. The stimulation program is developed using the Psychtoolbox platform in Matlab. After the stimulation program runs in Matlab, a stimulation image is presented on a display. The subject receives the visual stimulus and generates a response. The method is characterized by the following steps:

[0008] Step 1, Initialize the display environment: Call the Psychtoolbox toolkit to create a full-screen graphics window for the monitor; obtain the vertical refresh rate of the monitor; configure the stimulation parameters, including the target stimulation frequency, the size of the main stimulation region, the dot density, the probability of the center and periphery of the main stimulation region being lit, the brightness modulation period, and fix the random seed of the dot matrix of the main stimulation region;

[0009] Step 2, construct a multi-position visual stimulation spatial layout: set a main stimulation area at the geometric center of the display, and set auxiliary points in the four directions above, below, left and right of the main stimulation area. Each auxiliary point is marked with a circle, forming a multi-position visual stimulation spatial layout with the main stimulation area as the center and auxiliary points around it.

[0010] Step 3, Generate a dynamic dot matrix texture sequence for flickering stimulation: Divide the main stimulation area into a regular grid, and generate a binary mask matrix based on a random seed. The binary mask matrix is ​​the same size as the regular grid, and each cell of the binary mask matrix corresponds one-to-one with the regular grid. Each value at the corresponding position determines whether the corresponding grid point is lit. Generate a set of ordered dot matrix texture frames for the main stimulation area according to the target stimulation frequency and the vertical refresh rate of the display in the initial configuration. Modulate the effective radius of the main stimulation area by a sine function to generate a dynamic dot matrix texture sequence for flickering stimulation. The dynamic change of the effective radius of the lit area and the flickering of the dot matrix together realize the breathing-like brightness change of the dot matrix. Pre-cache the ordered dot matrix texture frame sequence.

[0011] Step 4, integrate the visual focus guidance structure: superimpose a red crosshair at the center of the main stimulus area, draw a square black border around the main stimulus area, with the red crosshair at the center of the border, and the two together form the visual focus guidance structure; the main stimulus area is set to a flashing stimulus; this guidance structure works synergistically with the "visual hotspot" effect of the main stimulus area dot matrix itself to guide and stabilize the subject's fixation point at the target center; the target center at this time is the red crosshair.

[0012] Step 5, Achieve Frame-Level Synchronization and Precise Frequency Control: Dynamically calculate the number of frames required for each stimulation cycle based on the target stimulation frequency and the vertical refresh rate of the display. Read and draw pre-buffered ordered dot matrix texture frames in sequence. Align the frame flipping timing with the vertical synchronization signal to form a complete stimulation cycle. Repeat this process for each stimulation cycle to achieve frame-level synchronization and precise frequency control. Start the stimulation program after completing frame-level synchronization and precise frequency control.

[0013] Step 6, Implementing the stimulation process of dynamic dot matrix-spatial reference joint modulation: After the stimulation program starts, a graphical interactive interface is provided, and it automatically enters the full-screen stimulation presentation state; the main stimulation area flashes in a breathing manner; a red cross is marked in the center of the main stimulation area, and a square black border is drawn around it to form a guide structure, with four static circular auxiliary marks around it; there are no interactive controls, at this time the subject looks at the stimulation target through the visual focusing guide structure, which greatly avoids external interference, while relevant data is collected at the same time, and the stimulation process is controlled by listening through an external keyboard;

[0014] Step 7, Program Termination and Data Collection: The stimulation program continuously listens for keyboard interrupt signals. Upon receiving a termination command, it closes the stimulation window and releases resources. The stimulation program ends and completes one subject's experiment. The relevant data is collected for subsequent analysis.

[0015] This invention addresses the problems of existing SSVEP stimulation methods, which commonly employ large-area overall light-dark reversal or fixed checkerboard patterns, resulting in drastic changes in light intensity that easily induce visual fatigue, dizziness, and other discomfort in users, thus affecting the accuracy of the collected data. Traditional multi-target stimulation paradigms rely on the synchronous flickering of multiple independent frequency sources, which not only consumes a large amount of frequency resources and easily causes harmonic interference, but also leads to visual competition and attention distraction, reducing signal quality and system stability.

[0016] Compared with the prior art, the present invention has the following advantages:

[0017] Improved safety, comfort, and user acceptance of the system during prolonged use: Because this invention uses dynamic dot matrix textures based on probability distribution and sinusoidal modulation, it achieves "breathing" local brightness gradients, completely avoiding the strong light intensity abrupt changes and visual flicker caused by the overall synchronous brightness and darkness flipping of large-area color blocks. From the perspective of physical stimulation, it significantly reduces the impact on the visual nervous system, effectively alleviating users' discomfort such as dry eyes, fatigue, and dizziness, and improving the safety, comfort, and user acceptance of the brain-computer interface system during prolonged use.

[0018] Effective suppression of attention drift and microsaccades: This invention innovatively integrates a red crosshair and a black border within the main stimulus area as a visual guidance structure, and places static spatial reference markers around the screen, providing users with clear and stable visual anchors and directional guidance. This effectively suppresses attention drift and microsaccades during fixation. Consequently, it helps improve the signal-to-noise ratio (SNR) and amplitude consistency of the induced SSVEP signal, providing a higher quality and more stable input signal for the backend decoding algorithm.

[0019] Solving the stimulus frequency drift problem: This invention employs dynamic frame count calculation based on the display's actual refresh rate and a frame flipping mechanism strictly aligned with the vertical sync signal, achieving microsecond-level high-precision timing control. This fundamentally solves the stimulus frequency drift problem caused by uncontrollable frame refresh timing in general graphics systems. It can control the error between the target frequency and the actual output frequency within an extremely small range, ensuring the accuracy and stability of the stimulus frequency, thereby significantly improving the reliability and robustness of decoding methods based on frequency domain analysis (such as the CCA algorithm).

[0020] Conserving stimulus frequency resources and avoiding harmonic interference and frequency selection challenges: This invention employs a "single stimulus-multiple gaze point" spatial coding paradigm, using only one dynamic stimulus frequency combined with the user's gaze at different spatial orientations (up, down, left, right) to recognize four different control commands. This significantly saves limited stimulus frequency resources, avoids the inherent harmonic interference and frequency selection challenges of multi-frequency systems, simplifies the system structure while maintaining high recognition performance, and reduces the requirements for hardware display performance. It provides a novel solution for developing low-cost, highly available, and easily scalable SSVEP-BCI systems. Attached Figure Description

[0021] Figure 1 This is a flowchart of the present invention;

[0022] Figure 2 This is a flowchart of the stimulation procedure of the present invention;

[0023] Figure 3 This is a diagram of the environmental platform of the present invention;

[0024] Figure 4 This is a schematic diagram illustrating the present invention;

[0025] Figure 5 To verify the method of the present invention, an experimental sample of the occipital electrode signal spectrum was collected and analyzed.

[0026] Figure 6 To verify the method of the present invention, an experiment was conducted to collect and analyze the spectrum of the ear electrode signal.

[0027] Figure 7 A graph showing the relationship between the accuracy and data length of classifying five targets (up, down, left, right, and center) using the TRCA algorithm on the collected occipital electrode signals.

[0028] Figure 8 A graph showing the relationship between the accuracy of classifying five targets (top, bottom, left, right, and center) using the TRCA algorithm on all collected electrode signals and the data length.

[0029] Figure 9A graph showing the relationship between the accuracy of classifying five targets (up, down, left, right, and center) using the TRCA algorithm on the collected ear electrode signals and the data length.

[0030] Figure 10 A graph showing the relationship between the accuracy of classifying four targets (up, down, left, and right) using the TRCA algorithm on the collected occipital electrode signals and the data length.

[0031] Figure 11 A graph showing the relationship between the accuracy of classifying four targets (up, down, left, and right) using the TRCA algorithm on all collected electrode signals and the data length.

[0032] Figure 12 The graph shows the relationship between the accuracy of classifying four targets (up, down, left, and right) using the TRCA algorithm on the collected ear electrode signals and the data length. Detailed Implementation

[0033] The present invention will now be described in detail with reference to the accompanying drawings.

[0034] Example 1: Existing SSVEP stimulation methods generally employ large-area overall light-dark reversal or fixed checkerboard patterns, resulting in drastic changes in light intensity that easily induce visual fatigue, dizziness, and other discomfort in users, leading to a poor user experience and consequently affecting the accuracy of the collected data. Furthermore, traditional multi-target stimulation methods rely on multiple independent frequencies, reducing signal quality and system stability. This invention addresses these issues and, through experiments and analysis, proposes a dynamic dot-matrix-spatial reference joint modulation SSVEP brain-computer interface stimulation method. This invention uses the Psychtoolbox platform in Matlab to develop the stimulation program. After the stimulation program runs in Matlab, the stimulation interface is displayed on the monitor as shown below. Figure 4 As shown; Figure 3 This is the environmental platform diagram of the present invention. In addition, a display screen is also required to present stimulating images. After the subject wears the EEG device correctly, he / she needs to sit quietly in front of the display screen. The subject is guided to receive visual stimulation and generate a response through the "visual hotspot" effect. During the experiment, the subject's EEG-related data needs to be continuously collected and used for subsequent data analysis.

[0035] After the subjects put on the EEG device, they sat quietly in front of the experimental equipment, ready to begin the experiment. (See also...) Figure 1 , Figure 1 The flowchart of this invention illustrates the SSVEP brain-computer interface stimulation method with dynamic dot matrix-spatial reference joint modulation, which includes the following steps:

[0036] Step 1, Initialize the display environment: Use the Psychtoolbox toolkit to create a full-screen graphics window for the monitor, where all subsequent stimulus interfaces will be displayed; obtain the monitor's vertical refresh rate; configure stimulus parameters, including target stimulus frequency, main stimulus region size, dot density, probability of illumination at the center and periphery of the main stimulus region, brightness modulation period, and fix the random seed of the main stimulus region dot matrix to ensure that the dot matrix texture sequence generated each time the stimulus program is run is completely consistent, thus ensuring the reproducibility of the experiment.

[0037] Step 2, Construct a multi-location visual stimulus spatial layout: Set a main stimulus area at the geometric center of the display (e.g., Figure 4 The size of the main stimulus area was calculated using the viewing angle conversion formula (pixel size = 2 × viewing distance × tan(viewing angle / 2) × screen DPI / 2.54). Auxiliary points were placed in the four directions above, below, left, and right of the main stimulus area. Each auxiliary point was marked with a circle, and the specific pixel coordinates of its center were determined by the viewing angle conversion. These four circular markers remained static throughout the experiment, and their properties did not change under any conditions. This created a multi-position visual stimulus spatial layout centered on the main stimulus area and surrounded by auxiliary points.

[0038] Step 3, Generate a dynamic dot matrix texture sequence for flickering stimulation: Divide the main stimulation area into a regular grid, and generate a binary mask matrix based on a random seed. The binary mask matrix is ​​the same size as the regular grid, and each cell of the binary mask matrix corresponds one-to-one with the regular grid. Each value at the corresponding position determines whether the corresponding grid point is lit. Generate a set of ordered dot matrix texture frames for the main stimulation area according to the target stimulation frequency and the vertical refresh rate of the display in the initial configuration. Modulate the effective radius of the main stimulation area by a sine function to generate a dynamic dot matrix texture sequence for flickering stimulation. The dynamic change of the effective radius of the lit area and the flickering of the dot matrix together realize the breathing-like brightness change of the dot matrix. Pre-cache the ordered dot matrix texture frame sequence.

[0039] Step 4, integrate the visual focus guidance structure: superimpose a red crosshair at the center of the main stimulus area, draw a square black border around the main stimulus area, with the red crosshair at the center of the border, and the two together form the visual focus guidance structure; the main stimulus area is set to a flashing stimulus; this guidance structure works synergistically with the "visual hotspot" effect of the main stimulus area dot matrix itself to guide and stabilize the subject's gaze point at the target center; the target center at this time is the red crosshair.

[0040] Step 5, Achieve Frame-Level Synchronization and Precise Frequency Control: Dynamically calculate the number of frames required for each stimulation cycle based on the target stimulation frequency and the vertical refresh rate of the display. Read and draw pre-buffered ordered dot matrix texture frames in sequence. Align the frame flipping timing through the vertical synchronization signal to form a complete stimulation cycle. Repeat this process for each stimulation cycle to achieve frame-level synchronization and precise frequency control. After completing frame-level synchronization and precise frequency control, start the stimulation program.

[0041] Step 6, Implementing the stimulation process of dynamic dot matrix-spatial reference joint modulation: After the stimulation program starts, a graphical interactive interface is provided, and it automatically enters the full-screen stimulation presentation state; the main stimulation area flashes in a breathing manner; a red cross is marked in the center of the main stimulation area, and a square black border is drawn around it to form a guide structure, with four static circular auxiliary markers around it; there are no interactive controls, at this time the subject looks at the stimulation target through the visual focusing guide structure, which greatly avoids external interference, while relevant data is collected at the same time, and the stimulation process is controlled by listening through an external keyboard.

[0042] Step 7, Program Termination and Data Collection: The stimulation program continuously listens for keyboard interrupt signals. Upon receiving a termination command, it closes the stimulation window and releases resources. The stimulation program ends and completes one subject's experiment. The relevant data is collected for subsequent analysis.

[0043] The core idea of ​​this invention is to abandon the traditional large-area overall brightness-darkness flipping stimulation method and instead set a dynamic dot matrix main stimulation area based on probability distribution and sinusoidal modulation in the center of the screen. This achieves localized, gradual "breathing" brightness changes, fundamentally avoiding visual discomfort caused by sudden changes in light intensity. Four static, non-flickering auxiliary circular markers are placed around the screen as spatial orientation references, providing clear directional guidance and assisting in intent recognition. A red crosshair and a black border are superimposed within the main stimulation area to form a composite visual guidance structure, further enhancing gaze stability and target boundary perception. A high-precision frame synchronization mechanism dynamically calculates the number of frames required per cycle based on the display's actual refresh rate and strictly controls the frame flipping timing in conjunction with the vertical synchronization signal, fundamentally eliminating frequency drift and ensuring the accuracy and stability of the stimulation frequency. A fixed random seed is used to initialize the random number generator, ensuring complete consistency of the dot matrix texture sequence and guaranteeing experimental repeatability and comparability of results. Ultimately, this invention achieves the recognition of multiple control commands using only a single stimulation frequency through a spatial coding strategy, greatly saving frequency resources and improving program usability and scalability.

[0044] The method proposed in this invention avoids the strong light intensity abrupt change and visual flicker caused by the overall synchronous light and dark flip of large-area color blocks. It significantly reduces the impact on the visual nervous system from the perspective of physical stimulation, effectively alleviates the user's discomfort such as dry eyes, fatigue, and dizziness, and thus effectively improves the quality of the data collected in the experiment.

[0045] Example 2: The overall scheme of the SSVEP brain-computer interface stimulation method with dynamic dot matrix-spatial reference joint modulation is the same as that in Example 1. In step 1 of this invention, the target stimulation frequency in the display environment is initialized to f0 = 6 Hz. The size of the main stimulation area is calculated according to the viewing angle conversion formula (pixel size = 2 × viewing distance × tan(viewing angle / 2) × screen DPI / 2.54). At a standard viewing distance of 70 cm, an 8° viewing angle corresponds to 316 pixels. Therefore, the main stimulation area is set to a square area of ​​316×316 pixels, with its center point coordinates coinciding with the center of the screen. The dot matrix density is set to 0.3. The lighting probability of the center area of ​​the main stimulation area is set to P_center = 0.4, and the lighting probability of the outer area is set to P_surround = 0.1. The brightness modulation period is consistent with the target stimulation frequency. The random seed is fixed using rng(12345). The target stimulation frequency, the lighting probability of the center area of ​​the main stimulation area, and the lighting probability of the outer area are all affected by the experimental conditions. A moderate dot density can reduce visual fatigue in subjects while ensuring stable experimental results, and a fixed random seed ensures the reproducibility of the experiment.

[0046] Example 3: SSVEP Brain-Computer Interface Stimulation Method with Dynamic Dot Matrix-Spatial Reference Joint Modulation. Similar to Examples 1-2, in step 2, the main stimulation region in constructing the multi-position visual stimulation spatial layout is a 316×316 pixel square region located in the center of the screen. Four static auxiliary circular markers are located at the top, bottom, left, and right of the screen, respectively. See [link to relevant documentation]. Figure 4The specific pixel coordinates of the center of each circle were determined through a perspective conversion: with the screen center as the origin, a 5° perspective corresponds to an offset of approximately 100 pixels at a viewing distance of 70 cm and a resolution of 1920×1080. Therefore, the coordinates of the four center points are screenCenter + [0, 100], screenCenter + [0, -100], screenCenter + [-100, 0], and screenCenter + [100, 0], respectively, where screenCenter is the exact center of the screen. Each auxiliary marker is drawn as a solid white circle with a diameter of 40 pixels using Screen('FillOval', window, [255 255 255], posRect), and a black outer ring with a line width of 2 pixels using Screen('FrameOval', window, [0 0 0], posRect, 2). These four markers remain static throughout the experiment, and their properties do not change under any conditions. A multi-position visual stimulation spatial layout is formed with the main stimulation area as the center and auxiliary points around it. This layout is the basis of the stimulation method of the present invention.

[0047] Example 4: SSVEP Brain-Computer Interface Stimulation Method with Dynamic Dot Matrix-Spatial Reference Joint Modulation. Similar to Examples 1-3, in step 3, the 316×316 pixel main stimulation region is divided into a 30×30 grid (each grid approximately 10.5×10.5 pixels) in the generated flashing stimulus dynamic dot matrix texture sequence. A fixed random seed (rng(12345)) is used to initialize the random state, generating a 30×30 binary mask matrix M. The value of each element is determined by the location of the grid point (center or periphery) and its corresponding probability (P_center or P_surround) through random sampling; 1 indicates on, and 0 indicates off. The number of frames required in one stimulation cycle is calculated as N_frames = ceil(refresh rate / f0) = ceil(60 / 6) = 10 frames. For each frame i (i ranges from 0 to N_frames-1), the current phase is calculated as phase = i / N_frames. The effective modulation radius of the current dot matrix is ​​calculated as r_current = 0.5 + 0.5 * sin(2 * π * phase) (range 0~1). Grid points within this radius are illuminated (set to white) if their corresponding value in the mask matrix is ​​1; otherwise, they are de-illuminated (set to the background color). Screen('MakeTexture', window, textureMatrix) is used to generate texture objects for each frame and store them in a texture handle array texSequence. This pre-generation and caching of all texture frames is used in the main loop to reduce runtime computation and rendering overhead, ensuring smooth screen updates and timing stability.

[0048] The sinusoidal modulation dynamic dot matrix flickering stimulation method avoids the strong light intensity abrupt changes and visual flicker caused by the synchronous brightness and darkness flipping of large-area color blocks, greatly improving user comfort and thus data quality. Simultaneously, the single-stimulus-multi-gaze stimulation method uses only one dynamic stimulation frequency to recognize multiple different control commands. This invention solves the problems of limited stimulation frequency resources and low user comfort, greatly saving limited stimulation frequency resources, avoiding the inherent harmonic interference and frequency selection difficulties of multi-frequency systems, simplifying the system structure and reducing the requirements for hardware display performance while maintaining high recognition performance.

[0049] Example 5: SSVEP brain-computer interface stimulation method with dynamic dot matrix-spatial reference joint modulation. The method is the same as in Examples 1-4. In step 4, the integrated visual focus guidance structure refers to superimposing a red crosshair at the center of the main stimulation area, drawing a square black border around the main stimulation area, and placing the red crosshair at the center of the border. The two together form the visual focus guidance structure. The RGB color value of the red crosshair is [255, 0, 0]. Screen('DrawLine', window, [255 0 0],screenCenter(1)-0.25*size, screenCenter(2), screenCenter(1)+0.25*size,screenCenter(2), 2) and

[0050] The `Screen('DrawLine', window, [255 0 0], screenCenter(1), screenCenter(2)-0.25*size, screenCenter(1), screenCenter(2)+0.25*size, 2)` method draws a line with a width of 2 pixels and a length approximately half the width of the main stimulus region in the center of the main stimulus region. The black border has an RGB color value of [0, 0, 0] and is drawn with a width of 2 pixels at the outer edge of the main stimulus region using `Screen('FrameRect', window, [0 0 0], targetRect, 2)`. Both of these methods persist throughout the experiment and their properties remain unchanged, enhancing gaze stability and boundary awareness, thereby improving data stability and data quality.

[0051] Example 6: The SSVEP brain-computer interface stimulation method with dynamic dot matrix-spatial reference joint modulation is the same as in Examples 1-5. In step 5, the frame-level synchronization and precise frequency control are achieved by dynamically calculating the number of frames required for each stimulation cycle based on the target stimulation frequency and the vertical refresh rate of the display, using an up-rounding method.

[0052] ;

[0053] in This indicates the number of frames required for each stimulation cycle. Indicates the monitor's vertical refresh rate. Indicates the frequency of the target stimulus.

[0054] Example 7: SSVEP Brain-Computer Interface Stimulation Method with Dynamic Dot Matrix-Spatial Reference Joint Modulation. Similar to Examples 1-6, step 5, which achieves frame-level synchronization and precise frequency control, involves recording the initial time t0 = GetSecs before the main loop begins. In each frame loop, the current running time t = GetSecs - t0 is calculated. The phase within the current stimulation cycle is calculated as current_phase = (t % T) / T. The texture frame index to be displayed is calculated as frame_index = floor(current_phase * N_frames) % N_frames. The corresponding texture is drawn onto the main stimulation area using Screen('DrawTexture', window, texSequence(frame_index+1), [], targetRect). The frame flip instruction vbl = Screen('Flip', window, vbl + (waitframes - 0.5) * ifi) is called to flip the frame. This instruction waits for the next vertical synchronization signal before refreshing the screen, thus achieving precise frame synchronization and ensuring the stimulation frequency remains stable at 6 Hz, avoiding any form of frequency drift.

[0055] Based on dynamic frame count calculation using the display's true refresh rate and a frame-flipping mechanism strictly aligned with the vertical synchronization signal, microsecond-level high-precision timing control is achieved. This allows the error between the target frequency and the actual output frequency to be controlled within an extremely small range, ensuring the accuracy and stability of the stimulus frequency, thereby significantly improving the reliability and robustness of decoding methods based on frequency domain analysis (such as the CCA algorithm).

[0056] Example 8: The SSVEP brain-computer interface stimulation method with dynamic dot matrix-spatial reference joint modulation is the same as in Examples 1-7, see [link to example]. Figure 2 , Figure 2This is a flowchart of the stimulation program of the present invention. In step 6, during the stimulation process of dynamic dot matrix-spatial reference joint modulation, the stimulation program executes the following sequentially after startup: initializing the display environment → multi-position visual stimulation spatial layout → generating a flashing stimulus dynamic dot matrix texture sequence → integrating a visual focusing guidance structure → continuously displaying the stimulus in the main stimulation area on the display and listening for keyboard interruption commands. The stimulation experiment is completed when the termination command is received. After the stimulation program starts, it automatically enters the full-screen stimulation presentation state. No interactive buttons, controls, or status labels are set in the program interface to maximize the display area and avoid distracting the user's attention. Keyboard events are listened for using the KbCheck or KbWait functions. Stimulation presentation begins when the experimenter or subject presses the 'space' key on the keyboard, and program termination is triggered when the 'escape' key is pressed. When the 'escape' key is detected, sca(Screen('CloseAll')) is first called to close all display windows, then ListenChar(0) is called to resume keyboard listening, and the program is safely exited, releasing all program resources.

[0057] Example 9:

[0058] The invention will be further illustrated by the following specific example:

[0059] Step 1: Initialize the PTB display environment.

[0060] The Psychtoolbox's PsychDefaultSetup class is used for basic setup, and Screen('Preference', 'SkipSyncTests', 0) is used to ensure rigorous frame synchronization testing. Then, a full-screen window is created using Screen('OpenWindow', screenNumber, [128 128 128]), with a background set to medium gray (RGB values ​​[128, 128, 128]). The precise refresh interval of the monitor is obtained using Screen('GetFlipInterval', window) (e.g., approximately 16.67 ms at a refresh rate of 60 Hz). The target stimulus frequency f0 = 6 Hz is set. The main stimulus region size was calculated using the viewing angle conversion formula (pixel size = 2 × viewing distance × tan(viewing angle / 2) × screen DPI / 2.54). At a standard viewing distance of 70 cm, an 8° viewing angle corresponds to 316 pixels. Therefore, the main stimulus region was set as a 316×316 pixel square region, with its center point coordinates coinciding with the screen center (calculated using [winWidth, winHeight] = Screen('WindowSize', window); screenCenter = [winWidth, winHeight] / 2;). The dot density was set to 0.3, and the lighting probability of the central region (with the center of the region as the center and a radius of 30%) was P_center = 0.4, while the lighting probability of the outer region was P_surround = 0.1. The period of the brightness modulation sine function was set to T = 1 / f0. A fixed random seed was set using rng(12345) to ensure that the generated dot texture sequence was completely consistent in each run, thus ensuring the reproducibility of the experiment.

[0061] Step 2: Construct a multi-location stimulus spatial layout.

[0062] The main stimulus region (texTarget) is a 316×316 pixel square area located in the center of the screen (screenCenter). Four static auxiliary circular markers (guidanceMarkers) are located at the top, bottom, left, and right of the screen. The specific pixel coordinates of their centers are determined through viewing angle calculation: with the screen center as the origin, a 5° viewing angle at a viewing distance of 70 cm and a resolution of 1920×1080 corresponds to an offset of approximately 100 pixels. Therefore, the coordinates of the four centers are screenCenter + [0, 100], screenCenter + [0, -100], screenCenter + [-100, 0], and screenCenter + [100, 0]. Each auxiliary marker is drawn as a solid white circle with a diameter of 40 pixels using `Screen('FillOval', window, [255 255 255], posRect)`, and as a black outer ring with a line width of 2 pixels using `Screen('FrameOval', window, [0 0 0],posRect, 2)`. These four markers remain static throughout the experiment, and their properties do not change under any conditions.

[0063] Step 3: Generate a dynamic raster texture sequence.

[0064] The 316×316 pixel main stimulus region is divided into a 30×30 grid (each grid is approximately 10.5×10.5 pixels). A fixed random seed (rng(12345)) is used to initialize the random state, generating a 30×30 binary mask matrix M. The value of each element is determined by the location of the grid point (center or periphery) and its corresponding probability (P_center or P_surround) through random sampling; 1 indicates illumination, and 0 indicates deactivation. The number of frames required for one stimulus cycle is calculated as N_frames = ceil(refresh rate / f0) = ceil(60 / 10) = 6 frames. For each frame i (i ranges from 0 to N_frames-1), the current phase is calculated as phase = i / N_frames. The effective modulation radius r_current = 0.5 + 0.5 * sin(2 * π * phase) (range 0~1) is calculated for the current dot matrix. Grid points within this radius are illuminated (set to white) if their corresponding value in the mask matrix is ​​1; otherwise, they are de-illuminated (set to the background color). Screen('MakeTexture', window, textureMatrix) is used to generate texture objects for each frame and store them in a texture handle array texSequence. This pre-generation and caching of all texture frames is used in the main loop to reduce runtime computation and rendering overhead, ensuring smooth screen updates and timing stability.

[0065] Step 4: Integrate the visual focus guidance structure.

[0066] The functions `Screen('DrawLine', window, [255 0 0], screenCenter(1)-0.25*size,screenCenter(2), screenCenter(1)+0.25*size, screenCenter(2), 2)` and `Screen('DrawLine', window, [255 0 0], screenCenter(1), screenCenter(2)-0.25*size,screenCenter(1), screenCenter(2)+0.25*size, 2)` were used to draw a red "+" crosshair (2 pixels wide) in the center of the main stimulus region. The function `Screen('FrameRect', window, [0 0 0], targetRect,2)` was used to draw a black border (2 pixels wide) around the outer edge of the main stimulus region. Both functions remained constant throughout the experiment to enhance fixation stability and boundary awareness.

[0067] Step 5: Implement frame-level synchronization and precise frequency control based on PTB.

[0068] Before the main loop begins, the initial time t0 = GetSecs is recorded. In each frame loop, the current running time t = GetSecs - t0 is calculated. The phase within the current stimulus cycle is calculated as current_phase = (t % T) / T. The texture frame index to be displayed is calculated as frame_index = floor(current_phase * N_frames) % N_frames. The corresponding texture is drawn onto the main stimulus area using Screen('DrawTexture', window, texSequence(frame_index+1), [],targetRect). Then, vbl = Screen('Flip', window,vbl + (waitframes - 0.5) * ifi) is called to perform a frame flip. This instruction waits for the next vertical sync signal (V-Sync) before refreshing the screen, thus achieving precise frame synchronization and ensuring the stimulus frequency remains stable at 6 Hz, avoiding any form of frequency drift.

[0069] Step 6: Provide a graphical interactive control interface and running status labels.

[0070] Upon startup, the program automatically enters full-screen stimulus presentation mode. No interactive buttons, controls, or status labels are set in the program interface to maximize the display area and avoid user distraction. Keyboard events are listened for using the KbCheck or KbWait functions. Stimulus presentation begins when the experimenter or subject presses the 'space' key, and program termination is triggered when the 'escape' key is pressed.

[0071] Step 7, the program terminates.

[0072] When the 'escape' key is detected, first call sca(Screen('CloseAll')) to close all display windows, then call ListenChar(0) to resume keyboard listening, and safely exit the program, releasing all program resources.

[0073] The SSVEP brain-computer interface stimulation method with dynamic dot-spatial reference joint modulation implemented through the above process not only effectively improves user comfort and gaze stability, but also solves the problems of limited stimulation frequency resources and low user comfort. It helps simplify the system structure and plays a substantial role in promoting the development of brain-computer interface systems.

[0074] The SSVEP brain-computer interface stimulation method of the present invention, characterized by dynamic dot matrix-spatial reference joint modulation, includes: initializing the display environment, constructing a multi-position visual stimulation spatial layout, generating a flashing stimulus dynamic dot matrix texture sequence, integrating a visual focusing guidance structure, achieving frame-level synchronization and precise frequency control, realizing the stimulation process of dynamic dot matrix-spatial reference joint modulation, completing the stimulation experiment, and collecting data. This invention, by combining a single dynamic stimulus source with spatial orientation reference, significantly reduces visual fatigue and improves fixation stability while ensuring high recognition performance. It effectively overcomes the problems of frequency interference, training dependence, and poor user experience inherent in multi-frequency stimulation systems, enhancing the comfort, practicality, and scalability of the SSVEP brain-computer interface system.

[0075] The technical effects of this invention will be verified below using simulations and data.

[0076] Example 10:

[0077] The SSVEP brain-computer interface stimulation method with dynamic dot matrix-spatial reference joint modulation is the same as in Examples 1-9.

[0078] Experimental conditions: Visual stimuli were presented using MATLAB and Psychtoolbox (PTB) V3. The stimulus source consisted of a rectangular main region (8-degree visual angle) and four surrounding circular markers (1-degree visual angle), with the center of each circular marker 5 degrees visually from the center of the rectangle. The stimulus frequency was set to 6 Hz.

[0079] The EEG signal acquisition channels include:

[0080] Occipital electrodes: PZ, POZ, PO3, PO4, PO7, PO8, OZ, O1, O2;

[0081] Electrodes around the ear and temporal region: FT9, FT10, T7, T8, TP7, TP8, TP9, TP10;

[0082] The reference electrode is placed at AFZ, and the ground electrode is at FCZ.

[0083] Experimental Procedure: Before starting, subjects sat 70 cm from the screen, correctly wore the EEG device, and began the experiment in an environment free from external light interference. Five target locations were set (center and four other locations: upper, lower, left, and right). After the stimulation program was initiated, subjects focused on each target, and data was collected six times. The acquired EEG data were preprocessed using a 4–55 Hz bandpass filter and classified using Task Relevant Component Analysis (TRCA) algorithm with leave-one-out cross-validation (LOOCV). By analyzing the classification performance under different electrode combinations (occipital region only, ear only, and all electrodes), the effectiveness of the stimulation method proposed in this invention was verified.

[0084] Experimental Results and Analysis:

[0085] See Figure 5 , 6 , Figure 5 To verify the method of the present invention, an experiment was conducted to collect and analyze the spectrum of signals from the occipital electrode. Figure 6 To verify the method of this invention, an experiment was conducted to collect and analyze the spectrum of the ear electrode signals. The horizontal axis of both graphs represents different frequencies (in Hz), and the vertical axis represents the amplitude of the corresponding frequency (in Hz). The spectra of both the occipital and ear signals showed a large amplitude at 6 Hz (the target stimulation frequency), while the amplitudes at other frequencies were significantly lower, indicating that the stimulation method of the present invention is highly effective.

[0086] Example 11:

[0087] The SSVEP brain-computer interface stimulation method with dynamic dot matrix-spatial reference joint modulation is the same as in Examples 1-9. The experimental conditions and experimental content are the same as in Example 10.

[0088] like Figure 7 , 8 As shown in Figure 9, Figure 7 The graph shows the relationship between the accuracy and data length of classifying five targets (up, down, left, right, and center) using the TRCA algorithm on the collected occipital electrode signals. Figure 8 The graph shows the relationship between the accuracy and data length of classifying five targets (up, down, left, right, and center) using the TRCA algorithm on all collected electrode signals. Figure 9This graph illustrates the relationship between the accuracy of classifying five targets (up, down, left, right, and center) using the TRCA algorithm on acquired ear electrode signals and data length. The x-axis represents data length (in seconds), and the y-axis represents the classification accuracy for different data lengths. In the classification task targeting five targets (center, up, down, left, and right), the experimental results clearly show the performance differences between different electrode combinations. Using only occipital electrodes (such as OZ, POZ, etc.) results in higher classification accuracy because these electrodes cover the primary visual cortex and visual association area of ​​the brain, making them most sensitive to dotted stimuli presented in the center of the monitor screen and effectively capturing changes in EEG signals caused by differences in stimulus location. Using only ear electrodes (such as T7, T8, TP7, TP8, etc.) results in significantly poorer classification performance. These areas are primarily responsible for auditory and some somatosensory spatial processing, and their response to purely visual spatial tasks is weak, providing limited information. The classification accuracy was optimal when all electrodes (occipital and auricular) were used in combination, indicating that although auricular electrodes are not effective when used alone, they still provide additional auxiliary information related to spatial processing (such as neural activity related to attentional allocation or spatial cognition), complementing the occipital signals and improving the overall resolution of the model. This result demonstrates that fusing multi-brain region response features can effectively improve the robustness of spatially encoded SSVEP paradigm recognition.

[0089] Example 12:

[0090] The SSVEP brain-computer interface stimulation method with dynamic dot matrix-spatial reference joint modulation is the same as in Examples 1-9. The experimental conditions and experimental content are the same as in Example 10.

[0091] like Figure 10 , 11 As shown in Figure 12, Figure 10 A graph showing the relationship between the accuracy of classifying four targets (up, down, left, and right) using the TRCA algorithm on the collected occipital electrode signals and the data length. Figure 11 A graph showing the relationship between the accuracy of classifying four targets (up, down, left, and right) using the TRCA algorithm on all collected electrode signals and the data length. Figure 12This graph shows the relationship between the accuracy of classifying four targets (up, down, left, right) using the TRCA algorithm on the acquired ear electrode signals and the data length. The horizontal axis represents data length (in seconds), and the vertical axis represents the classification accuracy for different data lengths. In the classification task excluding the central stimulus and focusing only on four targets (up, down, left, right), the experimental results generally follow the same trend as under the five-target condition, but with some subtle changes. Using only the occipital electrodes still showed high classification performance, indicating that the visual cortex still plays a dominant role in distinguishing peripheral spatial locations. Using only ear electrodes still resulted in poor performance, further confirming the limitations of these regions in purely visual spatial tasks. Using all electrodes still yielded the best classification results, highlighting the effectiveness of multi-region signal fusion. It is worth noting that excluding the central target may slightly simplify the task (reducing confusion between central and peripheral locations), but overall, the relative contribution pattern of the electrode combinations remained unchanged, indicating good stability of the experimental conclusions. This suggests that reducing the number of targets helps reduce classification difficulty, but the "single stimulus-multi-spatial reference" paradigm adopted in this invention still shows good discriminability even under the five-target condition.

[0092] The above experimental results confirm that this invention, through dynamic dot matrix stimulation combined with spatial orientation markers, can effectively induce SSVEP responses with high signal-to-noise ratio and spatial separability. The multi-channel EEG signal fusion strategy further improves the accuracy of intent recognition, providing a solid experimental basis for developing comfortable and efficient novel brain-computer interface systems.

[0093] In summary, this invention presents a dynamic dot-spatial reference joint modulation SSVEP brain-computer interface stimulation method, solving technical problems such as low user comfort, limited stimulation frequency resources, and high harmonic interference in the SSVEP brain-computer interface system. The implementation of this invention includes initializing the display environment, constructing a multi-position visual stimulation spatial layout, generating a flashing stimulation dynamic dot-matrix texture sequence, integrating a visual focus guidance structure, achieving frame-level synchronization and precise frequency control, realizing the stimulation process of dynamic dot-spatial reference joint modulation, completing the stimulation experiment, and collecting data. This invention uses a multi-position visual stimulation spatial layout centered on the main stimulation region with auxiliary points around it. The main stimulation region is illuminated by sinusoidal function modulation, and a visual focus guidance structure is integrated using a red cross and a black border. Frame-level synchronization and precise frequency control are also achieved. This effectively improves the stability and user comfort of the SSVEP brain-computer interface system, solves technical problems such as limited stimulation frequency resources and high harmonic interference, and can be used in the development of related brain-computer interface systems in military, medical, and entertainment fields.

Claims

1. A dynamic dot-matrix-spatial reference joint modulation SSVEP brain-computer interface stimulation method, wherein the stimulation program is developed using the Psychtoolbox platform of Matlab, and after the stimulation program is run in Matlab, the stimulation interface is presented on the display, and the subject is guided to accept visual stimulation and generate a response through the "visual hotspot" effect, characterized in that, It includes the following steps: Step 1, Initialize the display environment: Use the Psychtoolbox toolkit to create a full-screen graphics window for the monitor; Obtain the vertical refresh rate of the display; configure stimulation parameters, including target stimulation frequency, main stimulation area size, dot density, probability of the center and periphery of the main stimulation area being lit, brightness modulation period, and fix the random seed of the dot matrix of the main stimulation area; Step 2, construct a multi-position visual stimulation spatial layout: set a main stimulation area at the geometric center of the display, and set auxiliary points in the four directions above, below, left and right of the main stimulation area. Each auxiliary point is marked with a circle, forming a multi-position visual stimulation spatial layout with the main stimulation area as the center and auxiliary points around it. Step 3, Generate a dynamic dot matrix texture sequence for flickering stimulation: Divide the main stimulation area into a regular grid, and generate a binary mask matrix based on a random seed. The binary mask matrix is ​​the same size as the regular grid, and each cell of the binary mask matrix corresponds one-to-one with the regular grid. Each value at the corresponding position determines whether the corresponding grid point is lit. Generate a set of ordered dot matrix texture frames for the main stimulation area according to the target stimulation frequency and the vertical refresh rate of the display in the initial configuration. Modulate the effective radius of the main stimulation area by a sine function to generate a dynamic dot matrix texture sequence for flickering stimulation. The dynamic change of the effective radius of the lit area and the flickering of the dot matrix together realize the breathing-like brightness change of the dot matrix. Pre-cache the ordered dot matrix texture frame sequence. Step 4, integrate the visual focus guidance structure: superimpose a red crosshair at the center of the main stimulus area, draw a square black border around the main stimulus area, with the red crosshair at the center of the border, and the two together form the visual focus guidance structure; the main stimulus area is set to a flashing stimulus; this guidance structure works synergistically with the "visual hotspot" effect of the main stimulus area dot matrix itself to guide and stabilize the subject's fixation point at the target center; the target center at this time is the red crosshair. Step 5, Achieve Frame-Level Synchronization and Precise Frequency Control: Dynamically calculate the number of frames required for each stimulation cycle based on the target stimulation frequency and the vertical refresh rate of the display. Read and draw pre-buffered ordered dot matrix texture frames in sequence. Align the frame flipping timing with the vertical synchronization signal to form a complete stimulation cycle. Repeat this process for each stimulation cycle to achieve frame-level synchronization and precise frequency control. Start the stimulation program after completing frame-level synchronization and precise frequency control. Step 6, Implementing the stimulation process of dynamic dot matrix-spatial reference joint modulation: After the stimulation program starts, a graphical interactive interface is provided, and it automatically enters the full-screen stimulation presentation state; the main stimulation area flashes in a breathing manner; a red cross is marked in the center of the main stimulation area, and a square black border is drawn around it to form a guide structure, with four static circular auxiliary marks around it; there are no interactive controls, at this time the subject looks at the stimulation target through the visual focusing guide structure, which greatly avoids external interference, while relevant data is collected at the same time, and the stimulation process is controlled by listening through an external keyboard; Step 7, Program Termination and Data Collection: The stimulation program continuously listens for keyboard interrupt signals. Upon receiving a termination command, it closes the stimulation window and releases resources, ending the stimulation program and completing one subject's experiment. The program then terminates, and the relevant data is collected for subsequent analysis.

2. The SSVEP brain-computer interface stimulation method with dynamic dot matrix-spatial reference joint modulation according to claim 1, characterized in that, Step 1 initializes the target stimulus frequency in the display environment to 6Hz; the size of the main stimulus area is converted to pixel values ​​based on an 8° viewing angle. The dot density is set to 0.3; the illumination probability of the central region of the main stimulus area is set to 0.4, and the illumination probability of the peripheral region is set to 0.1; the brightness modulation period is consistent with the target stimulus frequency; the random seed is used to ensure that the ordered dot texture frame sequence is repeatable.

3. The SSVEP brain-computer interface stimulation method with dynamic dot matrix-spatial reference joint modulation according to claim 1 or 2, characterized in that, The center of the static auxiliary circular markers mentioned in step 2 is offset from the center of the display screen by 5°. Each marker consists of a white solid circle and a black outer ring, with a diameter of approximately 40 pixels and an outer ring line width of 2 pixels. All markers remain static throughout the entire operation.

4. The SSVEP brain-computer interface stimulation method with dynamic dot matrix-spatial reference joint modulation according to claim 1, characterized in that, In step 3, the regular grid divided by the main stimulus region is a 30×30 grid; the value of each element in the binary mask matrix is ​​determined by the location of the grid point and its corresponding probability through random sampling; the ordered dot matrix texture frame sequence achieves periodic contraction and expansion of the illumination range through sine function modulation.

5. The SSVEP brain-computer interface stimulation method with dynamic dot matrix-spatial reference joint modulation according to claim 1, characterized in that, In step 4, the RGB color value of the red crosshair is [255, 0, 0], the line width is 2 pixels, and the length is approximately half the width of the area; the RGB color value of the black border is [0, 0, 0], and the line width is 2 pixels.

6. The SSVEP brain-computer interface stimulation method with dynamic dot matrix-spatial reference joint modulation according to claim 1, characterized in that, The calculation of the number of frames required for each stimulus cycle, as described in step 5, based on the target stimulus frequency and the display's vertical refresh rate, is performed by rounding up. ； in This indicates the number of frames required for each stimulation cycle. Indicates the monitor's vertical refresh rate. Indicates the frequency of the target stimulus.

7. The SSVEP brain-computer interface stimulation method with dynamic dot matrix-spatial reference joint modulation according to claim 1, characterized in that, Step 5 describes achieving frame-level synchronization and precise frequency control, which means that in the main program loop, the corresponding frames are read sequentially from the pre-cached ordered texture frame sequence and drawn onto the main stimulus area through graphics drawing instructions; the frame flip instruction is called and the instruction is strictly aligned with the vertical refresh of the display.

8. The SSVEP brain-computer interface stimulation method with dynamic dot matrix-spatial reference joint modulation according to claim 1, characterized in that, The stimulation procedure described in step 6 is achieved by repeatedly executing steps 1 to 5, and the stimulation procedure ends upon receiving a termination command.

Images