Method and system for dynamically adjusting individual word memory quantity based on EEG (electroencephalogram)

Abstract

The invention discloses an individual word memory quantity dynamic adjustment method and system based on EEG, and belongs to the crossing field of education technology and electroencephalogram signal processing. The method comprises the steps that S1, dry electrode EEG equipment is used, an n-back task (changeable number / letter breadth task) is used for stimulation, an n value is added, and EEG signals (key P300 waves and theta waves), the accuracy rate and the reaction time are recorded; s2, preprocessing EEG data by using software such as EEGLAB and the like, and establishing a regression model to obtain basic working memory capacity; s3, setting two-dimensional monitoring, regularly monitoring a 2-back task analysis state rule at 3-4 time points per day, and monitoring a resting state EEG and a self-evaluation scale division state level 1-2 minutes before memory borrowing in real time; the system comprises four modules and software and hardware combination for dynamically adjusting the individual word memory quantity of the EEG, the system adapts to individual difference and state fluctuation, memory efficiency is improved, fatigue feeling is reduced, the system is easy to landing, and data safety is guaranteed.

Description

Technical Field

[0001] This invention relates to the interdisciplinary field of educational technology and EEG signal processing, and in particular to a method and system for dynamically adjusting the number of words memorized by an individual based on EEG. It is applicable to language learning, personalized education and other scenarios where the efficiency of word memorization is improved by scientifically adjusting the parameters of the memorization task. Background Technology

[0002] Currently, existing technologies in the field of word memorization training have many shortcomings.

[0003] 1. Ignoring the differences in individual working memory capacity, most solutions use a fixed number of words to memorize per group (such as a uniform setting of 4 to 6 words per group), without taking into account the natural differences in working memory capacity among individuals. This leads to some individuals having low memorization efficiency or being prone to fatigue, decreased memorization accuracy (usually below 60%), prolonged reaction time, and even developing a resistance to memorizing words.

[0004] 2. Lack of dynamic adaptation to individual state fluctuations. Existing technologies rely solely on the subject's subjective feelings (e.g., "reduce the number if you feel tired") or fixed time periods (e.g., "remember 5 in the morning and 3 in the afternoon") for adjustment. An objective, real-time state monitoring mechanism has not been established, and it is impossible to accurately capture the fluctuations in working memory state caused by factors such as rest, emotions, and physiological rhythms. The adjustment is lagging or inaccurate.

[0005] 3. The application of EEG technology is insufficient. Although a few technologies have attempted to use EEG to measure working memory capacity, they are only used as "static labels" (such as using the capacity result for a long time after a single measurement). The memory task parameters are not dynamically adjusted in conjunction with real-time EEG monitoring data, and a quantitative correspondence between "EEG indicators ~ status level ~ memory quantity" has not been established, resulting in poor technology implementation.

[0006] 4. The effectiveness evaluation dimensions are too limited. Existing solutions mostly evaluate effectiveness only through "immediate memory accuracy" and do not include key factors such as "long-term forgetting rate" and "fatigue quantification score", which cannot comprehensively measure the effectiveness and comfort of memory training.

[0007] In summary, existing technologies struggle to achieve personalized word memorization adjustments "based on objective physiological indicators and adapted to individual differences and state fluctuations," resulting in room for improvement in both memorization efficiency and user experience. Summary of the Invention

[0008] To overcome the shortcomings mentioned above, this invention provides a method and system for dynamically adjusting the number of words memorized based on EEG. Through a process of "measurement-monitoring-adjustment-evaluation," it achieves personalized dynamic optimization of the number of words memorized. The specific technical solution is as follows: The method for dynamically adjusting the number of words memorized by an individual based on EEG includes the following steps: Step S1: Select a dry electrode EEG device, use the n-back task paradigm in the device, present stimuli to the subject, gradually increase the n value of the stimulus, and record the EEG signal, task accuracy and reaction time under different n values. Step S2: Use EEGLAB software or BrainVision Analyzer software to preprocess EEG data, establish a regression model, and obtain the individual's basic working memory capacity based on EEG measurements. Step S3: Set up real-time and periodic monitoring of individual status fluctuations; Regular monitoring involves 3 to 4 fixed time points each day, with each monitoring session lasting 5 to 10 minutes. The device uses a 2-back task paradigm to collect EEG data and task performance, and analyzes the patterns of status fluctuations. Real-time monitoring involves collecting resting-state EEG signals for 1-2 minutes before memorizing words, and combining this with the subject's self-assessment state scale to classify the state into levels, with each level corresponding to the degree of deviation from the baseline state.

[0009] Furthermore, in step S3, the real-time monitoring divides the subject's self-assessment status scale into four status levels: excellent, good, average, and poor. When in optimal condition, the deviation from the individual's basic working memory capacity is ≤10%. When in good condition, the deviation from an individual's basic working memory capacity is 10-20%. The deviation between the actual working memory capacity and the individual's baseline working memory capacity during the state of being in a state of being is 20-30%. When in poor condition, the deviation from the individual's basic working memory capacity is >30%.

[0010] Furthermore, a table corresponding to the number of basic memories is established based on the basic working memory capacity, and the number of memories is adjusted according to the state level; When the condition is optimal, the initial quantity is increased by 1 to 2 per group, and the adjusted quantity does not exceed the maximum tolerable quantity corresponding to the basic working memory capacity; Maintain the initial quantity when the condition is good; When in the current state, the initial quantity is ~1 per group; If you are in poor condition, pause the task, rest for 15 to 30 minutes, and then retest. If you are still in poor condition, reduce the total number of sets for the day by 20% to 30%, and reduce the number of sets to 50% to 70% of the initial number.

[0011] Furthermore, step S4 involves collecting EEG signals for 1 minute after every 2-3 sets of memory. If an abnormal increase in theta wave or a decrease in P300 amplitude is detected, the number of sets is reduced by 1, and the rest period between sets is extended to 8-10 minutes. The initial rest period between sets before the extension is 5 minutes.

[0012] Furthermore, it also includes step S5, effect evaluation and program optimization: short-term evaluation involves recording the memory accuracy rate after 10 minutes, the forgetting rate after 24 hours, and the fatigue score daily; long-term evaluation involves calculating the total cumulative memory volume and memory efficiency per unit time monthly; and collecting feedback from participants every two weeks, and adjusting the state level threshold and the increase or decrease in the number of memories based on the evaluation data.

[0013] Furthermore, before step S1 is performed, subjects are required to avoid ingesting neuroactive substances for 1 hour before the measurement, rest for more than 30 minutes, and wear clothing without metal decorations. The error of the dry electrode EEG device must meet industry standards.

[0014] Furthermore, in step S1, the EEG signal focuses on monitoring the P300 amplitude, P300 latency, and theta power. The task accuracy threshold is set to 70%, and the n-back task paradigm can be replaced with a number or letter breadth task.

[0015] Furthermore, when the accuracy rate is below 70%, the capacity corresponding to the previous n value is the basic working memory capacity, and each n value is tested 3 times and the average value is taken. The table of basic memory quantity correspondence is as follows: 2-3 basic working memory units correspond to 3 initial memory units per group, 4-5 units correspond to 5 initial memory units per group, and 6-7 units correspond to 7 initial memory units per group.

[0016] The EEG-based system for dynamically adjusting the number of words a person can remember includes an EEG data acquisition module, a status analysis module, a memory adjustment module, and a data management module. The EEG data acquisition module is connected to the EEG device to receive raw signals in real time and complete preprocessing. The preprocessing process includes removing electrooculography and electromyography artifacts by combining independent component analysis (ICA) with filtering, completing data segmentation and baseline correction, and performing preprocessing operations using EEGLAB software or BrainVision Analyzer software. The state analysis module automatically classifies the state level and outputs an analysis report based on the preprocessed data and self-assessment scale. The state level classification combines resting-state EEG features and self-assessment scores. The resting-state EEG features include alpha wave power. The memory adjustment module generates memory quantity adjustment suggestions based on the status level, synchronously controls the output of the word library for each group of words, and receives signals transmitted in real time from the EEG data acquisition module during the memorization process to trigger temporary adjustment instructions. The data management module stores EEG data, memory process data, and evaluation data. It supports data encryption and access control. Access control includes administrator permissions and subject viewing permissions, each corresponding to different data operation scopes.

[0017] Furthermore, the EEG data acquisition module also includes a device adaptation unit, which supports establishing data communication with different models of dry electrode EEG devices.

[0018] Compared with the prior art, the beneficial effects of the present invention are: 1. Improve memory efficiency and effectiveness: By accurately capturing individual capacity differences and state fluctuations through EEG, the amount of memory can be dynamically adjusted to avoid "capacity waste" or "overload". In the short term, the memory accuracy rate can be improved by ≥10% and the forgetting rate can be reduced by ≥8%. In the long term, the total amount of memory can be increased by ≥15% and the efficiency per unit time can be increased by ≥12%.

[0019] 2. Reduce fatigue and improve user experience: Based on objective EEG indicators and subjective self-assessment status monitoring, fatigue signals can be identified in advance and task parameters can be adjusted, resulting in a decrease of ≥2 points in the fatigue score of the subjects and reducing resistance caused by inappropriate memory load.

[0020] 3. Strong technology applicability: Portable dry electrode EEG equipment is used (lowering the barrier to entry), a quantitative correspondence between "capacity, status and quantity" is established, and the supporting system realizes full-process automation (no need for professional personnel to operate in real time), which can be directly integrated into language learning APP and educational auxiliary equipment.

[0021] 4. Security and privacy protection: The EEG device meets safety standards (avoiding the risk of electric shock), and the system uses encrypted data storage and access control to prevent the leakage of the subject's EEG data and personal information.

[0022] 5. Continuous optimization capability: Through the "evaluation-feedback-adjustment" closed loop, parameters can be optimized according to the characteristics of different groups (such as students and adults), adapting to more usage scenarios and extending the technology life cycle.

[0023] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This is the individual working memory status level classification table based on EEG in this invention; Figure 2This is a table showing the correspondence between basic working memory capacity and the number of words memorized in this invention; Figure 3 This is a comparison table of memory effects in the experimental results of this invention; Figure 4 This is a reference table comparing fatigue and user experience in the experimental results of this invention. Detailed Implementation

[0026] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to specific implementation methods. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Example 1

[0027] The method for dynamically adjusting the number of words memorized by an individual based on EEG includes the following steps: Step S1: Select a dry electrode EEG device, use the n-back task paradigm in the device, present stimuli to the subject, gradually increase the n value of the stimulus, and record the EEG signal, task accuracy and reaction time under different n values. Step S2: Use EEGLAB software or BrainVision Analyzer software to preprocess EEG data, establish a regression model, and obtain the individual's basic working memory capacity based on EEG measurements. Step S3: Set up real-time and periodic monitoring of individual status fluctuations; Regular monitoring involves 3 to 4 fixed time points each day, with each monitoring session lasting 5 to 10 minutes. The device uses a 2-back task paradigm to collect EEG data and task performance, and analyzes the patterns of status fluctuations. Real-time monitoring involves collecting resting-state EEG signals for 1-2 minutes before memorizing words, and combining this with the subject's self-assessment state scale to classify the state into levels, with each level corresponding to the degree of deviation from the baseline state.

[0028] Furthermore, in step S3, the real-time monitoring divides the subject's self-assessment status scale into four status levels: excellent, good, average, and poor. When in optimal condition, the deviation from the individual's basic working memory capacity is ≤10%. When in good condition, the deviation from an individual's basic working memory capacity is 10-20%. The deviation between the actual working memory capacity and the individual's baseline working memory capacity during the state of being in a state of being is 20-30%. When in poor condition, the deviation from the individual's basic working memory capacity is >30%.

[0029] Furthermore, a table corresponding to the number of basic memories is established based on the basic working memory capacity, and the number of memories is adjusted according to the state level; When the condition is optimal, the initial quantity is increased by 1 to 2 per group, and the adjusted quantity does not exceed the maximum tolerable quantity corresponding to the basic working memory capacity; Maintain the initial quantity when the condition is good; When in the current state, the initial quantity is ~1 per group; If you are in poor condition, pause the task, rest for 15 to 30 minutes, and then retest. If you are still in poor condition, reduce the total number of sets for the day by 20% to 30%, and reduce the number of sets to 50% to 70% of the initial number.

[0030] Furthermore, step S4 involves collecting EEG signals for 1 minute after every 2-3 sets of memory. If an abnormal increase in theta wave or a decrease in P300 amplitude is detected, the number of sets is reduced by 1, and the rest period between sets is extended to 8-10 minutes. The initial rest period between sets before the extension is 5 minutes.

[0031] Furthermore, it also includes step S5, effect evaluation and program optimization: short-term evaluation involves recording the memory accuracy rate after 10 minutes, the forgetting rate after 24 hours, and the fatigue score daily; long-term evaluation involves calculating the total cumulative memory volume and memory efficiency per unit time monthly; and collecting feedback from participants every two weeks, and adjusting the state level threshold and the increase or decrease in the number of memories based on the evaluation data.

[0032] Furthermore, before step S1 is performed, subjects are required to avoid ingesting neuroactive substances for 1 hour before the measurement, rest for more than 30 minutes, and wear clothing without metal decorations. The error of the dry electrode EEG device must meet industry standards.

[0033] Furthermore, in step S1, the EEG signal focuses on monitoring the P300 amplitude, P300 latency, and theta power. The task accuracy threshold is set to 70%, and the n-back task paradigm can be replaced with a number or letter breadth task.

[0034] Furthermore, when the accuracy rate is below 70%, the capacity corresponding to the previous n value is the basic working memory capacity, and each n value is tested 3 times and the average value is taken. The table of basic memory quantity correspondence is as follows: 2-3 basic working memory units correspond to 3 initial memory units per group, 4-5 units correspond to 5 initial memory units per group, and 6-7 units correspond to 7 initial memory units per group.

[0035] The EEG-based system for dynamically adjusting the number of words a person can remember includes an EEG data acquisition module, a status analysis module, a memory adjustment module, and a data management module. The EEG data acquisition module is connected to the EEG device to receive raw signals in real time and complete preprocessing. The preprocessing process includes removing electrooculography and electromyography artifacts by combining independent component analysis (ICA) with filtering, completing data segmentation and baseline correction, and performing preprocessing operations using EEGLAB software or BrainVision Analyzer software. The state analysis module automatically classifies the state level and outputs an analysis report based on the preprocessed data and self-assessment scale. The state level classification combines resting-state EEG features and self-assessment scores. The resting-state EEG features include alpha wave power. The memory adjustment module generates memory quantity adjustment suggestions based on the status level, synchronously controls the output of the word library for each group of words, and receives signals transmitted in real time from the EEG data acquisition module during the memorization process to trigger temporary adjustment instructions. The data management module stores EEG data, memory process data, and evaluation data. It supports data encryption and access control. Access control includes administrator permissions and subject viewing permissions, each corresponding to different data operation scopes.

[0036] Furthermore, the EEG data acquisition module also includes a device adaptation unit, which supports data communication with different models of dry electrode EEG devices. The encryption method of the data management module includes symmetric encryption or asymmetric encryption to prevent the leakage of the subject's EEG data and personal information. Example 2

[0037] An EEG-based method for dynamically adjusting the number of words memorized by an individual includes: Step S1: EEG measurement of individual basic working memory capacity: Through standardized EEG measurements and data processing, the stable baseline working memory capacity of an individual is accurately quantified, providing a benchmark for subsequent adjustments to memory capacity. Specifically, this includes: S1.1 Preparations before measurement Equipment Selection: Dry electrode EEG equipment is used, and the equipment error meets industry standards (such as signal-to-noise ratio ≥20dB, common-mode rejection ratio ≥80dB), avoiding the problems of wet electrode equipment requiring the application of conductive paste and complicated operation, thus lowering the threshold for use; Subject preparation: Subjects were required to avoid consuming neuroactive substances (such as coffee, strong tea, and alcohol) within one hour before the measurement to prevent affecting the stability of EEG signals; rest for more than 30 minutes in advance to eliminate brain fatigue caused by previous activities; and wear clothing without metal decorations (such as without metal zippers or buttons) to avoid electromagnetic interference of metal on EEG signals.

[0038] S1.2 Basic Capacity Test Task paradigm: The n-back task paradigm is adopted (which can be replaced with a number breadth task or a letter breadth task according to the scenario to suit participants of different age groups). Stimuli (such as letters "A, B, C" or numbers "1, 2, 3") are presented to the participants at a frequency of 1 time / second, with each presentation lasting 500ms. Difficulty Adjustment and Data Recording: Starting with 1-back, gradually increase the value of n (1-back, 2-back, 3-back…), repeat the test 3 times for each n value, and take the average value to reduce random errors; record the three types of data for each test: 1. EEG signals: Focus on monitoring characteristic waves closely related to working memory, namely P300 wave (amplitude range 1-10μV, latency 300-600ms, decreased amplitude and prolonged latency indicate excessive working memory load) and theta wave (frequency 4-7Hz, increased power indicates brain fatigue). 2. Task performance data: Task accuracy (the percentage of subjects who correctly judge whether the current stimulus is consistent with n previous stimuli), reaction time (the time from stimulus presentation to the subject making a judgment, ranging from 300 to 1500 ms). 3. Capacity determination rule: Set the task accuracy threshold to 70% (the industry-recognized effective standard for judging working memory capacity). When the accuracy corresponding to a certain n value is lower than 70%, the capacity corresponding to the previous n value is the basic working memory capacity of that individual (e.g., when n=3 the accuracy is 65% and when n=2 the accuracy is 78%, then the basic capacity is 2).

[0039] Step S2: Data Preprocessing and Model Building S2.1 Preprocessing tools and operations: Use EEGLAB software or BrainVision Analyzer software to process the collected raw EEG data: S2.2 Artifact Removal: Independent Component Analysis (ICA) is used to separate and remove electrooculography artifacts (such as potentials generated by blinking) and electromyography artifacts (such as signals generated by facial muscle activity), and then the signal is further purified by 5-30Hz bandpass filtering. S2.3 Data Segmentation and Baseline Correction: Using 100ms before stimulus presentation as the baseline, the EEG signal is segmented into segments from "100ms before stimulus presentation to 1000ms after presentation" to eliminate the influence of baseline drift on the signal. S2.4 Regression Model Establishment: Based on the preprocessed EEG indicators (P300 amplitude / latency, theta wave power), task accuracy and n value, establish a linear regression model of "EEG indicator ~ accuracy ~ n value", quantify the basic working memory capacity into a specific value (unit: number of units). For example, when the model outputs "n=2.3", round it to 2 or round it to the nearest 2.

[0040] Step S3: Real-time monitoring of individual state fluctuations + periodic monitoring By employing a dual monitoring approach of "regular monitoring to identify patterns and real-time monitoring to determine the current situation," fluctuations in an individual's working memory state can be accurately captured, providing real-time data for dynamic adjustments. S3.1 Regular monitoring (pattern analysis) Monitoring frequency and time points: 3-4 fixed time points every day (such as 9:00, 12:00, 15:00, 19:00), covering four typical time periods: morning, noon, afternoon, and evening, to avoid the deviation of patterns caused by too concentrated time points; Monitoring tasks and duration: Data is collected at each time point through a 5-10 minute 2-back task (with fixed difficulty to avoid affecting the status judgment due to changes in difficulty), including EEG signals (focusing on monitoring P300 wave and theta wave) and task performance (accuracy rate, reaction time). Pattern analysis: After 7 consecutive days of monitoring, the data was analyzed using statistical software (such as SPSS) to obtain the individual state fluctuation patterns, such as "excellent state from 9:00 to 11:00 (accuracy ≥ 85%), medium state from 15:00 to 17:00 (accuracy 70%-85%), and poor state after 19:00 (accuracy < 70%)", which provides a basis for recommending time periods for subsequent memory tasks.

[0041] S3.2 Real-time monitoring (current status assessment) Monitoring timing: Perform within 5 minutes before each word memorization task begins to ensure that the monitoring results are highly consistent with the current memorization status; Data collection, such as Figure 1 As shown: Resting-state EEG signal: Collect 1-2 minutes of closed-eye resting-state EEG signal, focusing on analyzing alpha waves (frequency 8-13Hz, increased power indicates inattention and brain relaxation); Self-Rating Status Scale: Subjects are asked to rate their current "fatigue level" and "attention" on a scale of 1 to 10 (1 point = extremely fatigued / very poor attention, 10 points = no fatigue / excellent attention). The average of the two scores is taken as the self-rating score. State level classification: Combining "resting-state EEG features + self-assessment score", the individual state is divided into 4 levels, each level corresponding to "the degree of deviation from the basic working memory state" (the basic state is the stable state measured in step S1).

[0042] Step S4: Dynamic adjustment of the number of words memorized Using the logic of "initial setting based on basic capacity, adjustment based on state level, and fine-tuning based on real-time signals," the system achieves precise dynamic adaptation of the number of words memorized. S4.1 Determining the Number of Basic Memorization Points Based on the baseline working memory capacity measured in step S1, establish a clear table corresponding to the number of baseline memories, such as... Figure 2 As shown, this avoids the subjectivity of setting the initial quantity.

[0043] Note: "Maximum tolerable quantity" is the base capacity + 1 to avoid overloading due to excessive quantity.

[0044] S4.2 Adjustment based on state level Based on the status levels determined in step S3, adjust the number of words to memorize for each group according to the following rules: Excellent condition (self-assessment score ≥ 8): Increase the number of units by 1 to 2 per group based on the initial number, and the adjusted number shall not exceed the "maximum tolerable number" (e.g., basic capacity of 4 to 5 units, initial 5 units per group, can be adjusted to 6 units per group when the condition is excellent, but not exceeding the maximum of 6 units). Good condition (self-assessment score of 6-7): Maintain the initial quantity (e.g., if the initial number was 5 per group, continue to do so at 5 per group). In progress (self-assessment score 4-5): Reduce the number of items per group by 1 from the initial quantity (e.g., if the initial quantity was 5 per group, adjust to 4 per group). Poor condition (self-rating ≤3 points): Immediately pause the current memory task and allow the subject to rest for 15-30 minutes (light activity such as walking is recommended during this period; avoid prolonged sitting). After resting, repeat the real-time monitoring in step S2. If the condition is still "poor" after re-monitoring, adjust the task for the day: reduce the total number of groups by 20%-30% (e.g., from 10 groups to 7-8 groups), and reduce the number of items per group to 50%-70% of the initial number (e.g., from 5 items per group initially to 2-3 items per group).

[0045] S4.3 Real-time fine-tuning during the memory process To further adapt to changes in the memory process (such as fatigue after memorizing 3 sets), a real-time fine-tuning mechanism has been added: Fine-tune the trigger timing: After completing 2-3 sets of word memorization tasks, pause for 5 minutes and collect EEG signals for 1 minute; Fine-tuning is triggered based on the analysis of the acquired EEG signal. Fine-tuning is activated if any of the following conditions are met: Abnormally elevated theta waves (power increased by ≥20% compared to baseline, indicating brain fatigue). A decrease in P300 amplitude (a decrease of ≥15% compared to the baseline state indicates decreased attention). Fine-tuning measures: Reduce the number of repetitions per set by 1, and extend the rest time between sets from 5 minutes to 8-10 minutes to give the brain sufficient recovery time.

[0046] Step S5: Effect Evaluation and Solution Optimization By conducting short-term and long-term evaluations and continuously optimizing the solution based on user feedback, we ensure the stability and adaptability of the technical effects. S5.1 Short-term assessment (daily assessment) Evaluation indicators and recording methods: Immediate memory accuracy: 10 minutes after the memory task ends, the accuracy is calculated by "randomly selecting 50% of the memorized words for testing" (number of correctly recalled / total number of tests × 100%). 24-hour forgetting rate: 24 hours after the memory task ends, the same words are tested again, and the forgetting rate is calculated as (accuracy rate after 1-24 hours / immediate accuracy rate × 100%). Fatigue rating: After the memory task, participants rated their fatigue level on a scale of 1 to 10. Comparative analysis: Daily comparison of the differences in indicators between "after implementation of this plan" and "before implementation (such as a fixed number plan)", such as the immediate accuracy rate increasing from 75% to 88% after implementation, the 24-hour forgetting rate decreasing from 45% to 17%, and the fatigue score decreasing from 6 points to 3 points.

[0047] S5.2 Long-term assessment (monthly assessment) Evaluation indicators: Total cumulative memorization: The total number of words actually memorized by the participants in the current month (excluding the forgotten ones); Memory efficiency per unit time: Calculate "total cumulative memorization amount / total training time" (e.g., if you memorize 300 words in a month and the total training time is 30 hours, the efficiency is 10 words / hour). Comparison: Compare the evaluation data with the "control group using a fixed number of memories" (e.g., subjects of the same age and with the same basic level) to verify the advantages of this scheme (e.g., the total cumulative memory volume of this scheme is 15% higher than that of the control group, and the efficiency per unit time is 12% higher).

[0048] S5.3 Feedback Optimization (optimized every 2 weeks) Feedback collection: Collect feedback from participants through questionnaires or interviews, focusing on issues such as "whether the amount of memory is appropriate", "whether the rest time is sufficient", and "whether the task difficulty is suitable". Parameter adjustment: Optimize key parameters based on feedback and evaluation data. State level threshold: If most subjects report "self-rating 7 points but still feel tired", the self-rating threshold for "good state" can be adjusted from 6-7 points to 7-8 points; Increase or decrease in the number of items to be memorized: If subjects with a basic capacity of 2 to 3 items report "great pressure after adding 2 items", the increase in "excellent state" can be adjusted from 1 to 2 items to 1 item; Rest duration: If most participants reported that "8 minutes of rest is not enough", it can be extended to 10-12 minutes.

[0049] An EEG-based dynamic adjustment system for individual word memorization count To achieve full automation of the above methods, this invention develops an integrated management system comprising four core modules. These modules work collaboratively to ensure the feasibility and security of the technical solution. Module 1: EEG Data Acquisition Module Core function: Serving as the interface between the system and EEG devices, enabling real-time data reception and preprocessing; Specific functional details: Device Interoperability: Supports data communication with different models of dry electrode EEG devices (such as NeuroSky, Emotiv, etc.) via Bluetooth or USB, automatically identifies the device model and adapts the data format; Raw signal reception: Real-time reception of raw EEG signals transmitted by EEG devices (sampling rate 256Hz or 512Hz, in line with industry standards), stored in EDF or BDF format (common EEG data format). Automated preprocessing: Built-in core algorithms of EEGLAB software / BrainVision Analyzer software automatically complete artifact removal, data segmentation, baseline correction, and output preprocessed EEG feature data (such as P300 amplitude and theta power) without manual intervention.

[0050] Module 2: Status Analysis Module Core function: Based on EEG data and self-assessment information, automatically classify status levels and generate analysis reports; Specific functional details: Data input: Receive preprocessed data output from the EEG data acquisition module, and simultaneously obtain self-assessment status scale data filled in by the subjects through the system interface; Status level calculation: Built-in weighted calculation model of "EEG features + self-assessment score" (e.g., EEG features account for 60% and self-assessment score accounts for 40%), automatically outputs the status level (excellent / good / average / poor) and the specific percentage of deviation from the basic status; Analysis report generation: Automatically generates daily / weekly status analysis reports, including "status fluctuation curve", "recommendation of optimal memory period", and "speculation of the cause of abnormal status (e.g., poor status at a certain time may be due to insufficient sleep)", and displays them in the form of charts (such as line charts and bar charts) for easy viewing by subjects or administrators.

[0051] Module 3: Memory Adjustment Module Core functions: Generate adjustment suggestions based on status level, control the output of the vocabulary database, and enable real-time fine-tuning during the memorization process; Specific functional details: Initial quantity determination: The basic working memory capacity data stored in step S1 is retrieved, and the initial number of words for each group is automatically determined by combining the "basic capacity ~ initial quantity correspondence table"; Adjustment instruction generation: Receive the status level output by the status analysis module and generate memory quantity adjustment instructions according to preset rules (e.g., "initial 5 / group, status optimization to 6 / group"). Word database control: It works in conjunction with the system's built-in word database (which supports multiple languages ​​and difficulty levels, such as English CET-4, CET-6, TOEFL, etc.) and automatically outputs the corresponding number of words (e.g., 6 words per group) according to the adjustment command. It also supports reviewing words according to the "Ebbinghaus Revision Curve". Real-time fine-tuning trigger: During the execution of the memory task, the EEG signal transmitted by the EEG data acquisition module is received in real time. If an abnormal increase in theta wave or a decrease in P300 wave amplitude is detected, a fine-tuning instruction (reducing the number of words and extending the rest time) is automatically triggered, and the subject is prompted through the system interface that "the number of words in each group has been adjusted and the rest time has been extended".

[0052] Module 4: Data Management Module Core functions: Stores end-to-end data, ensuring data security and privacy, and supports data querying and export; Specific functional details: Data storage: Four categories of data are stored separately: EEG data: raw signal, preprocessed data, EEG feature data; Memory process data: number of words per group, memorization time, accuracy, reaction time; Evaluation data: daily instant accuracy rate, 24-hour forgetting rate, fatigue score, monthly cumulative total memory volume, and efficiency per unit time; User information: Basic information of the subjects (anonymized storage, such as using ID instead of name), basic working memory capacity, and pattern of state fluctuations; Data security: Encrypted storage: All data is encrypted using the AES-256 symmetric encryption algorithm to prevent unauthorized data theft. Access Control: Set three levels of permissions (administrator / subject / researcher): Administrator privileges: can view all participant data and configure system parameters (such as adjusting status level thresholds); Subject permissions: Subjects can only view their own memory data and assessment reports; they cannot view the data of others. Researcher permissions: Can only view anonymized summary data (such as average volume for a certain age group), cannot view personal information; Data services: Supports querying data by time (e.g., "October 2025") and indicators (e.g., "24-hour forgetting rate"), and exporting data to Excel or PDF formats for easy subsequent analysis or report generation. Example 3

[0053] To verify the effectiveness of this invention, a comparative experiment was conducted, and the detailed process is as follows: I. Experimental Subjects One hundred and eighty college students aged 18 to 25 (90 male and 90 female) were selected. All of them had basic English proficiency (College English Test Band 4 or above) and were excluded from having a history of neurological diseases (such as epilepsy, depression), taking drugs that affect nerve function recently (within 1 month), or having contraindications for EEG measurement (such as having metal implants in the head).

[0054] 180 participants were randomly divided into two groups: an experimental group (using the scheme of this invention) of 90 participants and a control group (using a fixed number scheme, 5 words per group) of 90 participants. There were no significant differences between the two groups in terms of age, gender, and English proficiency (P>0.05), ensuring the fairness of the experiment.

[0055] II. Experimental Equipment and Environment EEG equipment: NeuroSky MindWave Mobile 2 dry electrode EEG equipment was selected (error meets industry standards, sampling rate is 512Hz). System and Software: Install the supporting system of this invention (containing 4 core modules), while the control group uses the same system but disables the "dynamic adjustment function" and fixes each group to memorize 5 words; Experimental environment: a quiet laboratory with a light intensity of 300-500 lux and a temperature of 22-25℃, with each subject in an independent cubicle to avoid mutual interference; Vocabulary bank: Selected from the core vocabulary of College English Test Band 6 (500 words in total), both groups use the same vocabulary bank to ensure consistent memorization materials.

[0056] III. Experimental Period The total cycle is 30 days, divided into three phases: 1. Basic capacity measurement phase (Day 1): Only the experimental group performs step S1 to measure the basic working memory capacity; 2. Memory training phase (days 2-29): Both groups will conduct one word memorization training session per day, each lasting 30 minutes. 3. Effectiveness Evaluation Phase (Day 30): Conduct the final long-term effectiveness evaluation.

[0057] IV. Experimental Procedure (Taking the experimental group as an example) Step 1: Basic Capacity Measurement (Day 1) After entering the laboratory, the subjects rested for 30 minutes, confirmed that they had not ingested any neuroactive substances within one hour, and wore clothing without metal decorations. Wear a dry electrode EEG device and adjust the device until the signal is stable (signal-to-noise ratio ≥20dB). Perform the n-back task (letter stimulus), starting from 1-back, test 3 times for each n value, and record the EEG signal, accuracy, and reaction time; The software automatically preprocesses the data, establishes a regression model, and outputs the baseline working memory capacity (baseline capacity distribution of 90 subjects in the experimental group: 2-3 for 28 people, 4-5 for 42 people, and 6-7 for 20 people). Step 2: Memory Training (Days 2-29) Daily Schedule: Regular monitoring (9:00, 12:00, 15:00, 19:00): Participants complete a 5-minute 2-back task at the corresponding time points, and the system automatically collects data and analyzes the state patterns; Real-time monitoring (before memorization): Before starting the memorization task at 9:30, the subjects completed a 1-minute resting-state EEG collection and filled out a self-assessment questionnaire. The system output the status level. Memory quantity adjustment: The system determines the number of words per group based on the basic capacity and status level (e.g., 4-5 words per group for basic capacity, 6 words per group for excellent status), and controls the output of the word database. Fine-tuning of the memory process: After every 3 sets of memory are completed, the system collects EEG signals for 1 minute. If an increase in theta wave is detected, the system will automatically reduce 1 set and extend the rest time to 10 minutes. Short-term assessment: 10 minutes after memorization, the system automatically pops up test questions (randomly selected 50% of the words) and calculates the immediate accuracy rate; 24 hours later (before 9:30 the next day), the subject completes the 24-hour forgetting rate test and fatigue score.

[0058] Step 3: Long-term assessment (Day 30) The experimental group and the control group were compared in terms of "total cumulative memory in 30 days" (excluding words forgotten after 24 hours) and "memory efficiency per unit time" (total cumulative total / total training time, with total training time being 30 days × 30 minutes = 900 minutes). Feedback questionnaires (10 questions, 5-point scale) were collected from two groups of participants to evaluate user experience (e.g., "fatigue during the memorization process" and "satisfaction with the solution").

[0059] V. Experimental Results and Analysis 1. By Figure 3It can be seen that the experimental group was significantly better than the control group in all core indicators (P<0.05), proving that the present invention can effectively improve memory efficiency and effect and reduce the forgetting rate.

[0060] 2. By Figure 4 It can be seen that the fatigue score of the experimental group was significantly lower than that of the control group, and the user satisfaction and retention rate were significantly higher than those of the control group (P<0.05), proving that the solution of the present invention can effectively improve the user experience and enhance the willingness to use it for a long time.

[0061] 3. Verification of the accuracy of status monitoring The accuracy of state monitoring was verified by comparing the "systematically classified state levels" with the "subjects' actual memory performance (accuracy rate)". When the status is "Excellent", the proportion of actual accuracy rate ≥ 85% is 91.3%; When the status is "Good", the actual accuracy rate of 70% to 85% is 88.7%; When the status is "medium", the actual accuracy rate of 60% to 70% accounts for 85.2% of the total. When the status is "poor", the proportion of actual accuracy rates < 60% is 89.5%; This indicates that the system's state level classification is highly consistent with actual memory performance, demonstrating high monitoring accuracy and providing a reliable basis for dynamic adjustments.

[0062] VI. Implementation Conclusion As can be seen from the above specific implementation, in the "EEG-based method and system for dynamically adjusting the number of words memorized by an individual", the following applies: 1. It can accurately quantify an individual's basic working memory capacity through EEG and capture state fluctuations in real time; 2. The dynamic adjustment logic is scientifically effective and can significantly improve the efficiency, effectiveness, and stability of word memorization; 3. Significantly reduces user fatigue, improves user experience and long-term retention rate; 4. The system modules are fully functional, have low hardware requirements, can be directly integrated into existing educational products, and have good industrialization and promotion value.

[0063] The core innovations of this invention are as follows: 1. A basic working memory capacity measurement method based on n-back task and EEG indicators: Specifically, it is a technical solution that “induces EEG activity through an n-back task with progressively increasing n values, determines the basic capacity with an accuracy of 70% as the threshold, and establishes a quantitative calculation model by combining P300 and theta wave indicators”, which is different from existing capacity measurement methods that only rely on behavioral data (such as digit span).

[0064] 2. Individual state fluctuation monitoring mechanism of "regular EEG monitoring + real-time EEG + self-assessment scale": In particular, the technical details of "combining resting-state EEG characteristics (such as alpha wave power) with self-assessment scores, classifying 4 levels of state and quantifying the degree of deviation from the baseline state" solve the problem of monitoring lag caused by existing technologies that rely solely on subjective feelings or fixed periods.

[0065] 3. Dynamic adjustment rules for the number of words memorized based on status level: including "the increase or decrease in the number of words corresponding to the status of excellent / good / medium / poor (+1 to 2 words / maintain / -1 word / decrease to 50% to 70%)" and "the temporary fine-tuning mechanism triggered by abnormal EEG signals during the memorization process". This is the core logic for realizing personalized memorization.

[0066] 4. Integrate a closed-loop system for EEG data acquisition, status analysis, memory adjustment, and data management: especially the "linkage control between EEG data and word database output" and the "integrated design of data encryption and access control modules" to ensure the integrity and feasibility of the technical solution.

[0067] The beneficial effects of this invention are reflected in the following aspects: 1. By accurately capturing an individual's basic capacity and real-time status through EEG, the amount of memory can be dynamically adjusted to avoid "capacity waste" and "overload," thereby achieving a dual improvement in memory efficiency and effectiveness; 2. Based on the state monitoring of "objective EEG indicators + subjective self-evaluation", fatigue signals are identified in advance and parameters are adjusted to avoid users forcing themselves to memorize when they are fatigued, thereby reducing fatigue and improving user experience; 3. The implementation of the present invention is achieved through the combination of hardware, software and operating procedures, and it can be adapted to multiple scenarios; 4. We use dry electrode EEG devices that meet international safety standards, eliminating the risk of electric shock. They are biocompatible (the skin contact part uses medical-grade silicone) and cause no irritation even with long-term wear. 5. By using encrypted storage and hierarchical access control, the leakage of subjects' EEG data and personal information is prevented, which complies with the requirements of the Personal Information Protection Law and the Data Security Law and avoids privacy risks.

[0068] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, it is intended that all variations falling within the meaning and scope of equivalents of the claims be included within the present invention.

Claims

1. A method for dynamically adjusting the number of individual words memorized based on EEG, characterized in that, It comprises the following steps: Step S1, select a dry electrode EEG device, use n-back task paradigm in the device, present stimuli to the subjects, gradually increase the n value of the stimuli, record EEG signals, task accuracy and reaction time under different n values; Step S2, use EEGLAB software or BrainVision Analyzer software to preprocess EEG data, establish a regression model, and obtain individual-based working memory capacity based on EEG measurement; Step S3, set real-time monitoring and regular monitoring of individual state fluctuation; Regular monitoring is fixed at 3-4 time points per day, through 5-10 minutes, the device uses 2-back task paradigm to collect EEG data and task performance, and analyzes the state fluctuation rule; Real-time monitoring is to collect 1-2 minutes of resting state EEG signal before word memory, combined with the subject's self-evaluation state scale, to divide the state level, and the state level corresponds to the degree of deviation from the basic state.

2. The EEG-based individual word memory quantity dynamic adjustment method according to claim 1, wherein, The real-time monitoring in step S3 divides the subject's self-evaluation state scale into four state levels: excellent, good, medium and poor; When the state is excellent, the deviation degree is ≤10% compared with the individual-based working memory capacity; When the state is good, the deviation degree is 10-20% compared with the individual-based working memory capacity; When the state is medium, the deviation degree is 20-30% compared with the individual-based working memory capacity; When the state is poor, the deviation degree is > 30% compared with the individual-based working memory capacity.

3. The EEG-based individual word memory quantity dynamic adjustment method according to claim 2, characterized in that, According to the individual-based working memory capacity, a basic memory quantity corresponding table is established, and the memory quantity is adjusted according to the state level; When the state is excellent, the initial quantity is increased by 1-2 per group, and the adjusted quantity does not exceed the maximum tolerable quantity corresponding to the individual-based working memory capacity; When the state is good, the initial quantity is maintained; When the state is medium, the initial quantity is reduced by 1 per group; When the state is poor, the task is paused, and after resting for 15-30 minutes, the measurement is re-measured, if it is still poor, the total number of groups is reduced by 20-30% per day, and the number of groups is reduced to 50-70% of the initial number.

4. The EEG-based individual word memory quantity dynamic adjustment method according to claim 3, characterized in that, It also includes step S4, after completing 2-3 groups of memory, collect 1 minute of EEG signal, if theta wave is abnormally elevated or P300 amplitude is reduced, reduce 1 per group, and extend the inter-group rest to 8-10 minutes, the initial inter-group rest time before extension is 5 minutes.

5. The EEG-based individual word memory quantity dynamic adjustment method according to claim 1, wherein, It also includes step S5, effect evaluation and scheme optimization: short-term evaluation is to record the memory accuracy rate after 10 minutes, the 24-hour forgetting rate and the fatigue score; long-term evaluation is to statistically analyze the cumulative memory total amount and the memory efficiency per unit time every month; every 2 weeks, collect the feedback of the subjects, and adjust the state level threshold and the memory quantity increase and decrease amplitude according to the evaluation data.

6. The EEG-based individual word memory quantity dynamic adjustment method according to claim 1, wherein, Before step S1, the subjects are required to avoid taking neuroactive substances 1 hour before measurement, rest for more than 30 minutes, and wear clothes without metal decorations, and the error of the dry electrode EEG device needs to meet the industry standard.

7. The EEG-based individual word memory quantity dynamic adjustment method according to claim 1, wherein, In step S1, the EEG signal focuses on monitoring P300 amplitude, P300 latency and theta wave power, the task accuracy threshold is set to 70%, and the n-back task paradigm can be replaced by a number or letter span task.

8. The EEG-based individual word memory quantity dynamic adjustment method according to claim 7, characterized in that, When the accuracy is less than 70%, the capacity corresponding to the previous n value is the basic working memory capacity, and each n value is repeated for 3 times to take the average value; The basic memory quantity corresponding table is: the basic working memory capacity of 2-3 corresponds to the initial memory quantity of 3 / group, the basic working memory capacity of 4-5 corresponds to the initial memory quantity of 5 / group, and the basic working memory capacity of 6-7 corresponds to the initial memory quantity of 7 / group.

9. A system for dynamically adjusting the number of individual words memorized based on EEG, characterized in that, The system comprises an EEG data acquisition module, a state analysis module, a memory adjustment module and a data management module. The EEG data acquisition module is connected with an EEG device, receives original signals in real time and completes preprocessing, the preprocessing process comprises removing eye and muscle electrical artifact through independent component analysis (ICA) combined with filtering, completing data segmentation and baseline correction, and performing preprocessing operation by using EEGLAB software or BrainVision Analyzer software; The state analysis module automatically divides state grades and outputs analysis reports based on preprocessed data and self-evaluation forms, the state grade division combines resting-state EEG characteristics and self-evaluation scores, and the resting-state EEG characteristics comprise alpha wave power; The memory adjustment module generates memory quantity adjustment suggestions according to the state grades, synchronously controls the output of the number of words in each group from the word library, receives signals transmitted in real time by the EEG data acquisition module in the memory process, and triggers temporary adjustment instructions; The data management module stores EEG data, memory process data and evaluation data, supports data encryption and permission management, and the permission management comprises administrator permission and subject viewing permission, which correspond to different data operation ranges, respectively.

10. The EEG-based individual word memory quantity dynamic adjustment system according to claim 9, wherein, The EEG data acquisition module further comprises a device adaptation unit, which supports data communication with different types of dry electrode EEG devices.

Images