Cognitive control ability improving system based on n-back psychological training and fmθ neurofeedback-social reward coupling

Abstract

The application belongs to the technical field of neurofeedback training, and particularly relates to a cognitive control ability improvement system based on N-Back psychological training and Fm theta neurofeedback-social reward coupling. A display module is used for presenting serialized stimulation and collecting reaction time and accuracy data. A collection module is used for collecting electroencephalogram signals and outputting preprocessed signals. A signal processing and analysis module is used for determining Fm theta relative power and comparing the Fm theta relative power with a dynamic threshold to generate a compliance determination result. The display module is also used for receiving the Fm theta relative power and presenting in real time. A social reward triggering module is used for outputting audio and video reward content simulating social interaction. A central control module is used for updating the dynamic threshold according to historical statistical values of the compliance determination result, and issuing a triggering instruction when compliance is achieved and the compliance duration reaches a preset length. The application monitors Fm theta power in real time in the N-Back task, and presents social rewards in real time when compliance is achieved, so as to improve the training effect of cognitive control ability.

Description

Technical Field

[0001] This invention belongs to the field of neurofeedback training technology, specifically relating to a cognitive control ability enhancement system based on N-BACK psychological training and Fmθ neurofeedback-social reward coupling. Background Technology

[0002] Cognitive control refers to the psychological function of an individual adaptively mobilizing cognitive resources to adjust cognitive processes and cope with environmental changes in goal-oriented behavior. Neuroscience research shows that the power of the frontal midline theta rhythm increases during cognitive control tasks, and there is a positive correlation between the power level of this rhythm and cognitive control ability. Based on this physiological law, existing technologies have attempted to regulate the frontal midline theta rhythm through neurofeedback training methods, such as inducing this rhythmic activity through acoustic stimulation or modulating the cognitive control system using a real-time functional magnetic resonance imaging neurofeedback platform. In the field of cognitive behavioral training, the N-BACK working memory update task has been shown to robustly induce an increase in the power of the frontal midline theta rhythm, and this effect increases with increasing task load. Based on this, computer-based methods for training cognitive abilities using the N-BACK task have been developed.

[0003] However, in the aforementioned existing technologies, the modulation of theta rhythm in the prefrontal midline and cognitive behavioral training are technically independent, and the neurofeedback process and the cognitive task execution process are not linked in real time. Users cannot obtain immediate feedback on the correspondence between their EEG state and task performance during cognitive task execution, and there is a lack of synchronous coupling between the neuromodulation process and cognitive behavior. Furthermore, the reward mechanisms in existing neurofeedback schemes are mostly low-value feedback forms such as points or simple animations, which differ from the cognitive reinforcement path of high-value social rewards in daily social interaction situations. The enhancing effect of reward signals on cognitive control motivation is not fully utilized. In addition, there is a time delay or logical disconnect between the generation of neural activity achievement signals and the presentation of reward signals in existing systems. The immediate connection between neural activity and behavioral reinforcement has not been established, resulting in low efficiency of neurofeedback training and limited transfer of training effects to daily cognitive abilities. Summary of the Invention

[0004] The purpose of this invention is to provide a cognitive control ability enhancement system based on N-BACK psychological training and Fmθ neurofeedback-social reward coupling, in order to solve the problems in existing cognitive training systems where neurofeedback and cognitive behavioral tasks are independent of each other, and there is a lack of immediate connection between reward mechanisms and the achievement of neural activity targets, resulting in low training efficiency and limited transferability of effects.

[0005] The present invention achieves the above objectives through the following technical solutions: A cognitive control ability enhancement system based on N-BACK psychological training and Fmθ neurofeedback-social reward coupling includes a display module, a data acquisition module, a signal processing and analysis module, a social reward triggering module, and a central control module. The display module is used to present sequential stimuli to the user and collect reaction time data and accuracy data; The acquisition module is used to acquire the EEG signals of the user at the Fz site, and outputs the preprocessed signal after preprocessing. The signal processing and analysis module is used to perform sliding window power spectrum analysis on the preprocessed signal, determine the Fmθ relative power of the set frequency band in the current window, and compare the Fmθ relative power with the current dynamic threshold to generate a compliance judgment result. The central control module is used to receive the reaction time data, the accuracy data, the Fmθ relative power, and the achievement judgment result, update the dynamic threshold according to the historical statistical value of the achievement judgment result, and issue the trigger command to the social reward trigger module when the achievement judgment result indicates that the target has been achieved and the achievement duration reaches the preset duration. The social reward triggering module is used to respond to the triggering command issued by the central control module and output audio and video reward content simulating social interaction to the user through the display module.

[0006] Furthermore, the set frequency band is the 4-8Hz frequency band corresponding to the theta rhythm of the anterior midline of the frontal lobe.

[0007] Furthermore, the N-BACK task parameters used by the display module to present the sequential stimuli are set by the central control module. The task parameters include the n value, stimulus type, stimulus presentation duration, stimulus interval, and the total number of trials for each training block. The n value increases from the initial value to a high load value as the training progresses. The serialized stimulus generated by the display module is a pseudo-random sequence. In this pseudo-random sequence, adjacent trials do not repeat, the proportion of matching trials to the total number of trials is a preset percentage, and the positions of matching trials under the same conditions are pseudo-randomly distributed. The display module presents each trial in the following sequence of stages: initial stimulus presentation stage, where the current stimulus is presented in the center of the screen; fixation point stage, where the fixation point is presented in the center of the screen; target stimulus presentation stage, where the current stimulus is presented again; and response window stage, where the user's key press response is collected. The display module is also used to receive the Fmθ relative power and present it to the user in real time in a visual form.

[0008] Furthermore, the reaction time data collected by the display module is the time interval between the start of stimulus presentation and the button response by the user, and the display module only retains the reaction time data corresponding to the trials with correct responses; the accuracy data collected by the display module is the proportion of correct judgments of whether the user's button response matches the current stimulus under the current N-BACK condition.

[0009] Furthermore, the acquisition module employs a non-invasive EEG acquisition electrode, which is positioned at the Fz site. The acquisition module also includes a reference electrode and a ground electrode. The acquisition module is used to perform differential amplification, analog-to-digital conversion, and hardware filtering on the EEG signal before outputting the preprocessed signal.

[0010] Furthermore, the signal processing and analysis module uses a sliding window to segment the preprocessed signal and updates the Fmθ relative power according to the sliding step size of the sliding window; The signal processing and analysis module is also used to remove artifacts from the preprocessed signal before power spectrum analysis. The artifact removal includes: removing electrooculogram artifacts using independent component analysis; calculating the power ratio of the signal in the electromyography interference band, and removing the current window if the power ratio exceeds a set threshold; and calculating the signal variance, and removing the current window if the signal variance exceeds a set range. The signal processing and analysis module uses the periodogram method to estimate the power spectral density of the preprocessed signal in the set frequency band, obtains the Fmθ relative power by integrating the power spectral density, and converts the Fmθ relative power into a relative percentage form based on the resting-state baseline average power.

[0011] Furthermore, after the training session is started, the central control module first controls the acquisition module to acquire the Fmθ EEG signal of the user in the resting state, sets the initial value of the dynamic threshold according to the sum of the average power of Fmθ in the resting state and the standard deviation of the set multiple, and then starts the cyclic execution of each training block. The central control module updates the dynamic threshold based on the historical statistical values ​​of the compliance determination results, specifically as follows: After the current training block ends, the pass rate of each sliding window in the training block is calculated. If the pass rate of two consecutive training blocks is higher than the first percentage threshold, the dynamic threshold is increased by one step. If the pass rate of two consecutive training blocks is lower than the second percentage threshold, the dynamic threshold is decreased by the second step.

[0012] Furthermore, the preset duration includes multiple values, and the central control module determines the current target achievement duration from the multiple values ​​according to the training stage; wherein, the value corresponding to the initial training stage is less than the value corresponding to the middle training stage, and the value corresponding to the middle training stage is less than the value corresponding to the later training stage.

[0013] Furthermore, the social reward triggering module has a built-in reward content library, and the audio and video reward content is a randomly selected item from the reward content library; the reward content library includes peer recognition content, cooperative celebration content and / or social approval content, and the audio and video reward content is presented from a first-person perspective; The peer recognition content includes virtual peer recognition animations and accompanying voiceovers; the collaborative celebration content includes team high-five celebration animations and accompanying background music; and the social approval content includes virtual audience applause and cheering animations and text encouragement.

[0014] Furthermore, the central control module is also used to control the training process, including starting and stopping training sessions, switching training blocks, and arranging rest intervals; monitoring the electrode impedance and signal quality of the acquisition module; performing multimodal data time synchronization on the reaction time data, the accuracy data, the Fmθ relative power, and the triggering events of the social reward triggering module; and recording training process data, which includes the Fmθ power change trend, pass rate, behavioral performance, and number of reward triggers for each training block.

[0015] The beneficial effects of this invention are as follows: 1. This invention integrates N-BACK cognitive training tasks, real-time monitoring of Fmθ rhythm, and social reward triggering into a single system. This allows users to simultaneously receive visual feedback on their own neural activity state during cognitive task execution and instantly receive social reward signals when Fmθ power reaches a target. This mechanism correlates real-time changes in neural activity with specific cognitive behavioral performance. Users can intuitively perceive changes in their own neural state and their correspondence with reward acquisition during cognitive task execution, thereby enhancing the training effect of cognitive control.

[0016] 2. This invention uses adaptive adjustment of dynamic thresholds to automatically change the training difficulty standard as the user's ability improves, maintaining training within a moderately challenging range and avoiding a decline in training effectiveness due to standards that are too low or too high. Simultaneously, by instantly linking the presentation of social rewards to the achievement of neural activity targets, the invention leverages the enhancing effect of social reward signals on cognitive control motivation, providing positive reinforcement at the neural level and contributing to improved training effectiveness. Attached Figure Description

[0017] Figure 1 This is a system block diagram of a cognitive control ability enhancement system proposed in one embodiment of the present invention; Figure 2 This is a flowchart of signal power spectrum analysis and feedback proposed in one embodiment of the present invention; Figure 3 This is a flowchart illustrating the achievement determination and reward triggering process according to an embodiment of the present invention; Figure 4 This is a graph showing the trend of the Fmθ power achievement rate of the three groups as a function of the training process in the experimental verification of this invention. Figure 5 This is a comparison chart of the N-BACK response time differences before and after the intervention in the three groups during the experimental verification of this invention; Figure 6 This is a comparison chart of the differences in executive function scale scores before and after the intervention in the experimental verification of this invention. Detailed Implementation

[0018] The present application will now be described in further detail with reference to the accompanying drawings. It should be noted that the following specific embodiments are only used to further illustrate the present application and should not be construed as limiting the scope of protection of the present application. Those skilled in the art can make some non-essential improvements and adjustments to the present application based on the above application content.

[0019] Before describing the specific implementation methods of this application, for ease of understanding, the relevant terms involved in this application will be explained as follows: Fmθ rhythm, also known as the anterior midline theta rhythm, refers to an EEG rhythm signal located in the midline region of the frontal lobe with a frequency range of 4-8 Hz, acquired by Fz electrode sites. This rhythm shows a significant increase in power during cognitive control tasks, and its power level is positively correlated with cognitive control ability, exhibiting neuroplasticity.

[0020] N-BACK task: A working memory update paradigm that requires users to determine whether the currently presented stimulus is consistent with the nth previous stimulus, used to induce an increase in Fmθ rhythm power. The larger the value of n, the higher the task load and the greater the demand on cognitive control.

[0021] Training block: The basic training unit of the N-BACK task, which contains a preset number of consecutive trials. After each training block, the pass rate is statistically analyzed as the basis for dynamic threshold adjustment.

[0022] Social rewards: Positive feedback signals that simulate real-life social interaction scenarios, presented in audio and video format, including peer recognition, cooperative celebration, and social approval, used to enhance the training motivation of the users.

[0023] First Embodiment Please see Figure 1 A specific embodiment of this invention proposes a cognitive control ability enhancement system based on N-BACK psychological training and Fmθ neurofeedback-social reward coupling. The system includes a display module, a data acquisition module, a signal processing and analysis module, a social reward triggering module, and a central control module. This system enhances the cognitive control ability of users by integrating cognitive training tasks, real-time EEG monitoring, and social rewards.

[0024] In this application, the target users are individuals who need to improve their cognitive control abilities through cognitive training. The training objective of this system is to enhance the power of the theta rhythm in the prefrontal midline of the target users through the coupling effect of neurofeedback and social reward, thereby improving their executive functions such as working memory updating ability, inhibitory control ability, and cognitive flexibility.

[0025] Specifically, cognitive control refers to an individual's ability to adaptively mobilize cognitive resources to cope with environmental changes in goal-oriented behavior. Its core neurophysiological indicator is the frontal midline theta rhythm, or Fmθ rhythm. The Fmθ rhythm shows a significant increase in power during the execution of cognitive control tasks and exhibits neuroplasticity, which can be enhanced through training to improve an individual's ability to autonomously regulate it.

[0026] The display module presents sequential stimuli to users and collects reaction time (RT) and accuracy (ACC) data. The sequential stimuli, based on the classic N-BACK paradigm, are presented in the center of the screen. The N-BACK task robustly induces an increase in Fmθ power, and this effect strengthens with increasing task load; therefore, it can serve as a behavioral vehicle for Fmθ neurofeedback training. Reaction time data reflects the user's information processing speed, and accuracy data reflects the user's task execution accuracy; together, they constitute behavioral indicators for assessing cognitive control abilities.

[0027] Specifically, the accuracy rate is the proportion of correct judgments made by the subject regarding whether their key press response matches the current stimulus in the current n-back condition. The calculation formula is: ACC = N total ×N correct ×100%; where N correct To correctly respond to the number of trials, N total The total number of valid response trials.

[0028] Reaction time is specifically the time interval (ms) from the onset of stimulus presentation to the subject pressing a key. Only the RT of correct trials is included in the analysis, calculated using the formula: RT = T response -T stimulus_onset Among them, T response T represents the key press response time (system clock timestamp). stimulus_onsetThe stimulus is presented at the beginning moment.

[0029] The acquisition module is used to acquire the EEG signals of the user at the Fz site, and outputs a preprocessed signal after preprocessing. The Fz site is the optimal recording location for theta rhythm in the anterior midline of the frontal lobe. The acquisition module uses a non-invasive EEG acquisition electrode placed at the Fz site. The acquisition module also includes a reference electrode and a ground electrode. The reference electrode can be the bilateral mastoid site or the Cz site, and the ground electrode can be the AFz or FPz site. The acquisition module performs differential amplification, analog-to-digital conversion, and hardware filtering on the raw EEG signal before outputting the preprocessed signal.

[0030] In practical applications, before signal acquisition, the scalp skin at the Fz site, reference electrode site, and ground electrode site should be cleaned with exfoliating scrub and alcohol swabs to reduce scalp impedance to below 5kΩ. During electrode installation, the electrodes are positioned according to the international 10-20 system, and conductive gel is used to fix the electrodes to the Fz site, ensuring good contact between the electrodes and the scalp. During signal acquisition, the system continuously acquires raw EEG signals at a sampling rate of no less than 500Hz. This raw signal is recorded in voltage-time series form; that is, the EEG device measures the voltage value at fixed time intervals, and the time points and corresponding voltage values ​​are recorded sequentially as a data sequence. The raw EEG signal is calculated by subtracting the potential recorded by the reference electrode from the potential recorded by the Fz electrode. The acquisition module performs differential amplification, analog-to-digital conversion, and hardware filtering on the raw EEG signal before outputting a preprocessed signal.

[0031] The signal processing and analysis module performs sliding window power spectrum analysis on the preprocessed signal to determine the relative power of Fmθ within the set frequency band of the current window. The relative power of Fmθ is then compared with the current dynamic threshold to generate a pass / fail judgment result. The set frequency band is the 4-8 Hz band corresponding to the theta rhythm in the anterior midline of the frontal lobe. The relative power of Fmθ reflects the intensity of the cognitive control neural activity of the user at the current moment. By comparing it with the dynamic threshold, it can be determined whether the current EEG state has reached the training goal.

[0032] After receiving the Fmθ relative power, the display module presents it to the user in real time as a vertical energy bar. The height of the energy bar corresponds to the current value of the Fmθ relative power; the higher the power, the higher the energy bar fills, and the lower the power, the lower the energy bar. A marker line is located at the top or middle of the energy bar, corresponding to the current dynamic threshold value, allowing the user to visually compare the current power value with the threshold. When the Fmθ relative power reaches or exceeds the threshold, the energy bar fills above the marker line, and its color changes to indicate that the power has reached the target, for example, from blue to green. The refresh rate of the energy bar is consistent with the update rate of the sliding window in the signal processing module, ensuring a real-time correspondence between the visual feedback received by the user and the current EEG state. Through this visualization method, the user can simultaneously observe the instantaneous changes in their own Fmθ power during the N-BACK task and adjust their cognitive control strategies accordingly to attempt to increase the power level, thereby achieving the operant conditioning process of neurofeedback training.

[0033] The central control module is used to receive reaction time data and accuracy data, Fmθ relative power and compliance judgment results, update dynamic thresholds based on historical statistical values ​​of compliance judgment results, and issue trigger commands to the social reward trigger module when the compliance judgment result indicates compliance and the compliance duration reaches the preset duration.

[0034] The social reward triggering module responds to trigger commands issued by the central control module and outputs audio-visual reward content simulating social interaction to the user through the display module. Social rewards are designed based on the anticipatory control value theory. By simulating recognition and approval in real social interaction scenarios, it enhances the dorsal anterior cingulate cortex's valuation of cognitive control behavior, thereby strengthening the user's intrinsic motivation.

[0035] In practical applications, the display module should preferably use an LCD monitor with a refresh rate of at least 60Hz and a resolution of at least 1920×1080. The distance between the user and the screen should be maintained at 60-80 cm, and the field of view should be controlled within the range of 15-20 degrees. The acquisition module uses a non-invasive EEG acquisition device with Ag / AgCl sintered electrodes positioned at the Fz site according to the international 10-20 system. The reference electrodes are positioned at the bilateral mastoid sites or Cz sites, and the grounding electrodes are positioned at the AFz or FPz sites. The bioelectric amplifier of the acquisition module has a sampling rate of at least 500Hz, an input impedance of at least 100MΩ, a common-mode rejection ratio of at least 100dB, and an analog-to-digital converter resolution of at least 24bit. The signal processing and analysis module, the social reward triggering module, and the central control module can be integrated into a general-purpose computer, a dedicated embedded processor, or a cloud server. The signal processing, threshold management, process control, and reward triggering logic are implemented through software programs. Data transmission between modules can be achieved through wired data interfaces or wireless communication.

[0036] In this embodiment, the dynamic threshold update allows the training difficulty standard to adaptively adjust as the user's ability improves, avoiding training saturation due to excessively low standards or frustration due to excessively high standards. The issuance of trigger commands establishes a correlation between achieving neural activity targets and social rewards, enabling the user to receive positive reinforcement at the neural level.

[0037] In some embodiments, the frequency band is set to the 4-8 Hz band corresponding to the theta rhythm in the anterior midline of the frontal lobe. This frequency band is a characteristic band where Fmθ power increases significantly during the execution of cognitive control tasks. According to the definition in electroencephalology, the natural frequency range of theta rhythm is 4-8 Hz. This invention specifically targets this frequency band for neurofeedback training, which differs from previous attention training methods that target frequency bands such as alpha waves, and achieves precise targeting of the cognitive control neural mechanism.

[0038] In some embodiments, the N-BACK task parameters used by the display module to present sequential stimuli are set by the central control module. These parameters include the n-value, stimulus type, stimulus presentation duration, stimulus interval, and the total number of trials for each training block. The n-value increases progressively from an initial value to a higher load value as training progresses. Specifically, the n-value is initially set to 1-back, and gradually increased to 2-back and 3-back as training progresses. This progressive increase in task load ensures that the training difficulty matches the user's ability improvement, avoiding frustration caused by excessively high initial difficulty.

[0039] The sequential stimuli generated by the display module are pseudo-random sequences. Adjacent trials in this pseudo-random sequence do not repeat, and the proportion of matching trials to the total number of trials is a preset percentage. Furthermore, the positions of matching trials under the same conditions are pseudo-randomly distributed. The design of the pseudo-random sequence avoids the subject forming expectations about the stimulus sequence, ensuring that the task can accurately measure working memory update capabilities. The preset percentage of matching trials to the total number of trials can be any value between 30% and 40%. This proportion ensures effective measurement of cognitive control ability while avoiding task difficulty imbalance caused by too few or too many matching trials.

[0040] The display module presents each trial in the following sequence: Initial stimulus presentation phase, where the current stimulus is presented in the center of the screen; Fixation phase, where the fixation point is presented in the center of the screen; Target stimulus presentation phase, where the current stimulus is presented again; Response window phase, where the stimulus disappears and the system waits for the user's key press response. These four phases constitute a complete trial cycle. The initial stimulus presentation phase presents the stimulus information to be memorized to the user; the fixation point phase prompts the user to prepare a response; the target stimulus presentation phase presents the target stimulus that needs to be judged; and the response window phase collects the user's key press response.

[0041] In some embodiments, the reaction time data collected by the display module is the time interval between the start of stimulus presentation and the button response by the user, and the display module only retains the reaction time data corresponding to trials with correct responses. The accuracy data collected by the display module is the proportion of correct judgments of whether the user's button response matches the current stimulus under the current N-BACK condition.

[0042] Understandably, reaction time and accuracy are two core behavioral indicators for measuring cognitive control ability in cognitive control tasks. Reaction time reflects the speed of information processing, while accuracy reflects the accuracy of task execution. Including only the reaction time of correct trials in valid records can exclude invalid reaction time data caused by guessing or distraction in erroneous responses, making the behavioral assessment results more accurate and reliable. A higher accuracy indicates that the user performs the cognitive control task more accurately; a shorter reaction time indicates that the user's information processing efficiency is higher.

[0043] In some embodiments, see Figure 2 The signal processing and analysis module uses a sliding window to segment the preprocessed signal and updates the relative power of Fmθ according to the sliding step size of the sliding window.

[0044] Specifically, the signal processing and analysis module segments the continuous EEG signal using a sliding window method. This sliding window extracts data from the signal along the time dimension, and the signal data within each window is used to calculate the relative power of Fmθ at that moment. The sliding step size determines the update frequency of the power estimate; a shorter step size results in more frequent updates and higher real-time feedback.

[0045] The signal processing and analysis module is also used to remove artifacts from the preprocessed signal before power spectrum analysis. Artifact removal includes: removing electrooculography artifacts using independent component analysis; calculating the power proportion of the signal in the electromyography interference band, and removing the current window if the power proportion exceeds a set threshold; and calculating the signal variance, and removing the current window if the signal variance exceeds a set range.

[0046] Electrooculography (EOG) artifacts are interference signals generated by blinking and eye movements, with amplitudes much larger than those of electroencephalogram (EEG) signals. Failure to remove these artifacts will severely impact the accuracy of Fmθ power estimation. Independent component analysis (ICA), a blind source separation method based on joint analysis of the Fz electrode and the vertical EEO channel, can identify and eliminate independent components related to blinking, thereby effectively removing EEO artifacts.

[0047] Electromyographic artifacts are high-frequency interference signals generated by facial and neck muscle activity. The interference frequency band is typically 20-100Hz. By calculating the power proportion of this frequency band, the degree of electromyographic contamination can be quantified. If high-frequency noise accounts for more than 30% of the total power, it indicates severe electromyographic contamination in that window and should be removed.

[0048] Poor electrode contact or signal saturation can also render data invalid. A flat signal indicates poor electrode contact, while signal saturation indicates amplifier overload; both render EEG data unusable. These abnormal states can be effectively identified by calculating the signal variance and comparing it to a predetermined interval.

[0049] Artifact removal employs a cascaded strategy, which executes the process sequentially in the order of electrooculography artifact removal, electromyography artifact detection, and electrode contact failure detection. If any artifact is detected, the current window is removed.

[0050] The signal processing and analysis module uses the periodogram method to estimate the power spectral density of the preprocessed signal in a set frequency band. The relative power Fmθ is obtained by integrating the power spectral density, and then converted into a relative percentage based on the resting-state baseline average power. The periodogram method obtains the power distribution of the signal in the frequency domain by performing a Fourier transform on the signal and calculating the square of its amplitude. Integrating the power spectral density of the 4-8Hz frequency band yields the total power value for that band. Converting the real-time power value into a percentage relative to the baseline eliminates differences in resting-state power levels between individuals, making the training effects comparable across different users.

[0051] In some embodiments, after the training session is started, the central control module first controls the acquisition module to acquire the Fmθ EEG signal of the user in the resting state, sets the initial value of the dynamic threshold according to the sum of the average power of Fmθ in the resting state and the standard deviation of the set multiple, and then starts the cyclic execution of each training block.

[0052] The resting-state Fmθ EEG signal reflects the baseline neural activity level of the user in a task-free state. The resting-state recording duration can be 2 minutes. During this time period, the acquisition module continuously acquires EEG signals at the Fz site, and the signal processing and analysis module calculates the average power and standard deviation of Fmθ during this period. The multiplier is set to 1.5 times the standard deviation. This value is based on the following considerations: the Fmθ power in the resting state follows an approximately normal distribution, and 1.5 times the standard deviation corresponds to medium to high effect magnitudes, which ensures both the challenge of the threshold to promote neural plasticity and the attainability of the threshold to avoid excessive frustration.

[0053] The central control module updates the dynamic threshold based on historical statistics of the compliance determination results. Specifically, after the current training block ends, the compliance rate of each sliding window within that training block is calculated. If the compliance rate of two consecutive training blocks is higher than the first percentage threshold, the dynamic threshold is increased by one step; if the compliance rate of two consecutive training blocks is lower than the second percentage threshold, the dynamic threshold is decreased by a second step. The compliance rate is defined as the proportion of sliding windows whose relative power Fmθ reaches or exceeds the current dynamic threshold within a training block, relative to the total number of sliding windows in that training block.

[0054] The stepped adaptive threshold algorithm dynamically adjusts the training difficulty standard through statistical analysis of historical pass rates. The first percentage threshold can be 70%, the second percentage threshold can be 30%, and both the first and second step lengths can be 0.1 times the resting-state Fmθ standard deviation. When the pass rate of two consecutive training blocks exceeds 70%, it indicates that the current standard is too easy for the user, so the threshold is increased to enhance the training challenge. When the pass rate of two consecutive training blocks is below 30%, it indicates that the current standard is too difficult for the user, so the threshold is decreased to reduce the training difficulty and prevent users from becoming frustrated due to frequent failures. Through the above adaptive adjustment, the pass rate during the training process can be maintained at a moderate interval of 50% to 60%, ensuring that the training is always in a moderately challenging range and maximizing learning efficiency.

[0055] In some embodiments, the preset duration includes multiple values, and the central control module determines the current target achievement duration from these values ​​based on the training stage. The value corresponding to the initial training stage is less than the value corresponding to the middle training stage, and the value corresponding to the middle training stage is less than the value corresponding to the later training stage.

[0056] Specifically, the duration of achieving the target in the initial stage can be 3 seconds, in the intermediate stage it can be 5 seconds, and in the later stage it can be 7 seconds. In the initial stage of training, the users have not yet mastered the skill of autonomously controlling Fmθ power. Using a shorter duration of achievement can reduce the difficulty of starting and help users quickly establish a successful training experience. As training progresses, the users' control ability gradually increases, and gradually extending the duration of achievement can increase the training challenge and continuously promote neuroplasticity.

[0057] A duration of 3 seconds is sufficient to eliminate instantaneous random fluctuations in Fmθ power. A duration of 5 seconds matches the single trial cycle of the N-BACK task, ensuring that social rewards are tied to task-related sustained increases in Fmθ power, rather than to transient event-related potentials. A duration of 7 seconds further enhances training on sustained cognitive control states. If the duration is too short, such as below 3 seconds, it will introduce a higher risk of false alarms due to random fluctuations; if the duration is too long, such as above 10 seconds, it will weaken the correlation between behavior and reward, reducing learning efficiency.

[0058] In some embodiments, the social reward triggering module has a built-in reward content library, and the audio / video reward content is randomly selected from the reward content library. The reward content library includes peer recognition content, cooperative celebration content, and / or social approval content, and the audio / video reward content is presented from a first-person perspective.

[0059] Each time the reward is triggered, the social reward trigger module randomly selects an item from the reward content library for playback. This random selection mechanism avoids habituation caused by repeatedly playing the same content, maintaining the freshness and motivational effect of the reward. The playback duration of the audio / video reward content can be 3-5 seconds, which is sufficient for users to fully perceive social recognition without excessively interfering with the training process. The first-person perspective presentation enhances the users' sense of immersion and perception of social value, making the simulated social interaction more realistic and effective.

[0060] Peer recognition content includes virtual peer recognition animations and accompanying voiceovers. Virtual peers give a thumbs-up with encouraging voiceovers such as "Great job!" or "Keep it up!", simulating peer recognition in real-life social scenarios. Collaboration celebration content includes team high-five celebration animations and accompanying background music. Animated scenes of team members high-fiving in celebration, accompanied by upbeat background music, create an atmosphere of successful collaboration. Social approval content includes virtual audience applause and cheering animations and encouraging text visuals. Virtual audiences applaud and cheer along with text messages such as "Your brain is getting stronger," conveying signals of social approval.

[0061] The three types of rewards described above simulate positive feedback in real-world social interactions from different dimensions: peer recognition focuses on individual achievement, collaborative celebration focuses on team belonging, and social approval focuses on social recognition. The combined use of these three types can comprehensively stimulate the intrinsic motivation of the users.

[0062] In some embodiments, the central control module is also used to control the training process, including starting and stopping training sessions, switching training blocks, and scheduling rest intervals. By controlling the training process, the central control module ensures that training is executed in an orderly manner according to a preset plan. A training session may contain 10 training blocks, each training block containing 20 trials. A short rest can be scheduled after each training block is completed to prevent user fatigue.

[0063] The central control module is also used to monitor the electrode impedance and signal quality of the acquisition module. By monitoring the electrode impedance and signal quality in real time, the central control module can output prompts in a timely manner when there is poor electrode contact or a decline in signal quality, ensuring the reliability of EEG data acquisition.

[0064] The central control module is also used for multimodal data time synchronization of reaction time data, accuracy data, Fmθ relative power, and trigger events from the social reward trigger module. Multimodal data time synchronization is a fundamental guarantee for the closed-loop training system. EEG signals, behavioral data, and trigger events come from different data sources and each has its own independent timestamp. Through the central control module's synchronization and storage function, these data can be aligned according to a unified time benchmark, providing a complete data foundation for subsequent training effect evaluation and algorithm optimization.

[0065] The central control module also records training process data, including the Fmθ power change trend, achievement rate, behavioral performance, and reward trigger count for each training block. This complete recording of training process data enables the quantitative evaluation of training effectiveness. Analysis of the Fmθ power change trend can assess whether the user's ability to autonomously regulate Fmθ power improves with training; analysis of the achievement rate trend can assess the effectiveness of the dynamic threshold adjustment strategy; analysis of behavioral performance can assess whether the training effect has transferred to cognitively controlled behavior; and analysis of the reward trigger count can assess the user's training engagement.

[0066] According to the above embodiments, the complete workflow of the system is as follows.

[0067] First, the central control module initiates the training session, and the control display module presents sequential stimuli, starting the N-BACK task. During task execution, the acquisition module continuously acquires the user's EEG signals at the Fz locus, and outputs them to the signal processing and analysis module after preprocessing. The signal processing and analysis module performs sliding window segmentation and artifact removal on the EEG signals, estimates the power spectral density of the 4-8Hz frequency band using the periodogram method, and integrates to obtain the Fmθ relative power. This power value is output to the display module and the central control module as a relative percentage. The display module presents the current Fmθ relative power level and its relative position to the dynamic threshold to the user in real time in the form of a vertical energy bar. The user attempts to adjust their own Fmθ power by observing the changes in the energy bar. (See also...) Figure 3 The central control module receives the achievement judgment results output by the signal processing and analysis module. When the achievement judgment result indicates that the relative power of the current window Fmθ reaches or exceeds the dynamic threshold, the central control module begins to accumulate the achievement duration. If this continuous achievement state is maintained for a preset duration, the central control module issues a trigger command to the social reward trigger module. If, during the accumulation process, the achievement judgment result of any sliding window indicates that the achievement has not been achieved, the central control module resets the achievement duration count to zero and continues to monitor subsequent sliding windows. The social reward trigger module responds to the trigger command, randomly selects an audio-visual reward from the reward content library, and outputs it to the user through the display module. At the same time, the central control module records the achievement rate of each training block and adaptively adjusts the dynamic threshold based on the historical achievement rate at the end of the training block. Throughout the process, the central control module performs multimodal time-synchronous storage of EEG signals, behavioral data, and trigger events, forming a complete training process data record.

[0068] Thus, during the N-BACK task, users learn about their own Fmθ power state through real-time visual feedback, consciously adjust and increase their Fmθ power using cognitive strategies, and receive social rewards after reaching the target power level and maintaining it for a sufficient duration. These mechanisms work together to gradually enhance users' ability to autonomously regulate their Fmθ power as training progresses, thereby improving their cognitive control capabilities.

[0069] Experimental verification To verify the effectiveness of this system, the following experiments were conducted.

[0070] I. Verification Objective The objectives of the experiments included: whether the subjects' self-regulation ability of Fmθ power improved with the training process; whether cognitive control behavior performance, i.e., reaction time and accuracy of the N-BACK task, was improved; and whether the training effect could be transferred to untrained cognitive tasks.

[0071] II. Experimental Design A randomized controlled trial design was used, with eligible participants randomly assigned to three groups, each receiving a different intervention protocol. The experimental group received training using the system, specifically a complete protocol combining Fmθ neurofeedback with an N-BACK task and social reward. Control group 1 received Fmθ neurofeedback combined with the N-BACK task but without social reward. Control group 2 received sham feedback, where a pseudo-random sham feedback signal was played during the N-BACK task; this signal was unrelated to the participants' actual Fmθ power. Except for the intervention protocol, all groups maintained the same training frequency, training period, and assessment points. Training was conducted four times per week, with each session lasting approximately 30 minutes, for a four-week period. Assessment points included baseline assessment before training, assessment within one week after training, and a follow-up assessment one month after training.

[0072] III. Experimental Subjects The inclusion criteria for the participants were: age 18 to 60 years; subjective complaints about cognitive control function, such as self-reported inattention, impulsivity, or difficulty completing complex tasks; scores on the Adult Executive Function Scale exceeding the clinical threshold of 65; no history of serious neurological or mental illness; no history of epilepsy; and voluntary signing of informed consent.

[0073] IV. Evaluation Indicators The evaluation indicators included the time trend of Fmθ power achievement rate during training, the changes in reaction time and accuracy of the N-BACK task before and after intervention, and the changes in scores on the Adult Executive Function Scale before and after intervention. The Fmθ power achievement rate reflects the change in the user's ability to autonomously regulate the Fmθ rhythm as training progresses. The N-BACK task reaction time and accuracy reflect the direct impact of training on cognitive control behavior performance. The Adult Executive Function Scale scores reflect the degree of transfer of training effects to daily executive function.

[0074] V. Data Analysis Methods Repeated measures ANOVA was used to examine the time trend of Fmθ power achievement rate in the three groups. The analysis indicators included group main effect, time main effect, and group-time interaction effect. One-way ANOVA was used to compare the differences in N-BACK task response time before and after intervention in the three groups, as well as the differences in adult executive function scale questionnaire scores before and after intervention in the three groups. Pairwise comparisons between groups were performed using post-hoc tests.

[0075] VI. Experimental Results Please see Figure 4 , Figure 4The graph shows the trend of Fmθ power achievement rate in the three groups as the training progresses. Repeated measures ANOVA results showed a significant interaction effect between group and time (p<0.05), indicating that the trends of Fmθ power achievement rate in the three groups differed significantly over time. The Fmθ power achievement rate in the experimental group showed a significant upward trend as the training progressed, while control group 1 also showed a certain degree of increase, but the increase was smaller than that in the experimental group. Control group 2 showed no significant change during the training process. These results indicate that in the groups trained using this system, the self-regulation ability of users to control Fmθ power significantly increased as the training progressed, and its growth pattern was substantially different from that of the two control groups that did not receive complete intervention.

[0076] Please see Figure 5 , Figure 5 This is a comparison of the N-back reaction time differences before and after the intervention in the three groups. One-way ANOVA showed significant differences in reaction time differences among the three groups. The difference in N-back reaction time before and after intervention in the experimental group was significantly higher than that in the two control groups, and the reaction time difference in control group one was also significantly higher than that in control group two. These results indicate that in the groups trained using this system, the reduction in N-back reaction time was significantly greater than in the other two groups, meaning that information processing efficiency was significantly improved. The control group without social reward also showed some improvement, but the degree of improvement was lower than that in the experimental group, indicating that the introduction of social reward enhanced the training effect.

[0077] Please see Figure 6 , Figure 6 This is a comparative graph showing the differences in scores on the Adult Executive Function Scale before and after intervention in the three groups. One-way ANOVA results showed significant differences in scores among the three groups. The experimental group showed significantly greater improvement in scores than the two control groups, and the improvement in control group one was also significantly greater than that in control group two. These results indicate that in the groups trained using this system, the daily executive function of the participants was significantly improved, and this improvement remained stable at the one-month follow-up after training, suggesting that the training effect has a certain long-term sustainability and transferability to daily cognitive function.

[0078] Based on the above experimental results, this system demonstrates significant superiority over the control group in enhancing Fmθ power self-regulation ability, improving N-BACK task performance, and promoting executive function improvement, thus verifying the effectiveness of the system in enhancing cognitive control ability.

[0079] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated.

[0080] The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions may be transmitted from one website, computer, server, or data center to another via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can access, or a data storage device such as a server or data center that integrates one or more available media. The available medium may be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., a solid-state drive (SSD)).

[0081] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0082] In addition, the functional modules in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0083] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A cognitive control ability enhancement system based on N-BACK psychological training and Fmθ neurofeedback-social reward coupling, characterized in that, It includes a display module, a data acquisition module, a signal processing and analysis module, a social reward triggering module, and a central control module; The display module is used to present sequential stimuli to the user and collect reaction time data and accuracy data; The acquisition module is used to acquire the EEG signals of the user at the Fz site, and outputs the preprocessed signal after preprocessing. The signal processing and analysis module is used to perform sliding window power spectrum analysis on the preprocessed signal, determine the Fmθ relative power of the set frequency band in the current window, and compare the Fmθ relative power with the current dynamic threshold to generate a compliance judgment result. The central control module is used to receive the reaction time data, the accuracy data, the Fmθ relative power, and the achievement judgment result, update the dynamic threshold according to the historical statistical value of the achievement judgment result, and issue the trigger command to the social reward trigger module when the achievement judgment result indicates that the target has been achieved and the achievement duration reaches the preset duration. The social reward triggering module is used to respond to the triggering command issued by the central control module and output audio and video reward content simulating social interaction to the user through the display module.

2. The cognitive control ability enhancement system according to claim 1, characterized in that, The set frequency band is the 4-8Hz band corresponding to the theta rhythm of the anterior midline of the frontal lobe.

3. The cognitive control ability enhancement system according to claim 1, characterized in that, The N-BACK task parameters used by the display module to present the sequential stimuli are set by the central control module. The task parameters include the n value, stimulus type, stimulus presentation duration, stimulus interval, and the total number of trials for each training block. The n value increases from the initial value to a high load value as the training progresses. The serialized stimulus generated by the display module is a pseudo-random sequence. In this pseudo-random sequence, adjacent trials do not repeat, the proportion of matching trials to the total number of trials is a preset percentage, and the positions of matching trials under the same conditions are pseudo-randomly distributed. The display module presents each trial in the following sequence of stages: initial stimulus presentation stage, where the current stimulus is presented in the center of the screen; fixation point stage, where the fixation point is presented in the center of the screen; target stimulus presentation stage, where the current stimulus is presented again; and response window stage, where the user's key press response is collected. The display module is also used to receive the Fmθ relative power and present it to the user in real time in a visual form.

4. The cognitive control ability enhancement system according to claim 3, characterized in that, The reaction time data collected by the display module is the time interval between the start of stimulus presentation and the button response by the user, and the display module only retains the reaction time data corresponding to the trials with correct responses; the accuracy data collected by the display module is the proportion of correct judgments of whether the user's button response matches the current stimulus and the current N-BACK condition.

5. The cognitive control ability enhancement system according to claim 1, characterized in that, The acquisition module uses a non-invasive EEG acquisition electrode, which is located at the Fz site. The acquisition module also includes a reference electrode and a ground electrode. The acquisition module is used to perform differential amplification, analog-to-digital conversion and hardware filtering on the EEG signal before outputting the preprocessed signal.

6. The cognitive control ability enhancement system according to claim 3, characterized in that, The signal processing and analysis module uses a sliding window to segment the preprocessed signal and updates the Fmθ relative power according to the sliding step size of the sliding window. The signal processing and analysis module is also used to remove artifacts from the preprocessed signal before power spectrum analysis. The artifact removal includes: removing electrooculogram artifacts using independent component analysis; calculating the power ratio of the signal in the electromyography interference band, and removing the current window if the power ratio exceeds a set threshold; and calculating the signal variance, and removing the current window if the signal variance exceeds a set range. The signal processing and analysis module uses the periodogram method to estimate the power spectral density of the preprocessed signal in the set frequency band, obtains the Fmθ relative power by integrating the power spectral density, and converts the Fmθ relative power into a relative percentage form based on the resting-state baseline average power.

7. The cognitive control ability enhancement system according to claim 6, characterized in that, After the training session is started, the central control module first controls the acquisition module to acquire the Fmθ EEG signal of the user in the resting state, sets the initial value of the dynamic threshold according to the sum of the average power of Fmθ in the resting state and the standard deviation of the set multiple, and then starts the cyclic execution of each training block. The central control module updates the dynamic threshold based on the historical statistical values ​​of the compliance determination results, specifically as follows: After the current training block ends, the pass rate of each sliding window in the training block is calculated. If the pass rate of two consecutive training blocks is higher than the first percentage threshold, the dynamic threshold is increased by one step. If the pass rate of two consecutive training blocks is lower than the second percentage threshold, the dynamic threshold is decreased by the second step.

8. The cognitive control ability enhancement system according to claim 7, characterized in that, The preset duration includes multiple values, and the central control module determines the current target achievement duration from the multiple values ​​according to the training stage; wherein, the value corresponding to the initial training stage is less than the value corresponding to the middle training stage, and the value corresponding to the middle training stage is less than the value corresponding to the later training stage.

9. The cognitive control ability enhancement system according to claim 7, characterized in that, The social reward triggering module has a built-in reward content library, and the audio and video reward content is a randomly selected item from the reward content library; the reward content library includes peer recognition content, cooperative celebration content and / or social approval content, and the audio and video reward content is presented from a first-person perspective; The peer recognition content includes virtual peer recognition animations and accompanying voiceovers; the collaborative celebration content includes team high-five celebration animations and accompanying background music; and the social approval content includes virtual audience applause and cheering animations and encouraging text visuals.

10. The cognitive control ability enhancement system according to claim 1, characterized in that, The central control module is also used to control the training process, including starting and stopping training sessions, switching training blocks, and arranging rest intervals; monitoring the electrode impedance and signal quality of the acquisition module; performing multimodal data time synchronization on the reaction time data, the accuracy data, the Fmθ relative power, and the triggering events of the social reward triggering module; and recording training process data, which includes the Fmθ power change trend, pass rate, behavioral performance, and number of reward triggers for each training block.

Images