Cognitive state evaluation method based on electroencephalogram analysis and visual feedback system

Abstract

The invention discloses a cognitive state evaluation method based on electroencephalogram analysis and a visual feedback system, and the method comprises the steps: collecting an electroencephalogram signal, and carrying out the preprocessing of the electroencephalogram signal; the preprocessed electroencephalogram signals are analyzed in a sliding window mode, the frequency band total energy of various waves in different frequency bands in the electroencephalogram signals is extracted, the average value is taken, and the frequency band energy average value of the various waves is obtained; extracting a blink electromyographic signal from the preprocessed electroencephalogram signal according to a frequency band energy mean value of each wave, and calculating a normalized blink frequency; calculating the concentration degree and the cognitive load; and respectively normalizing the concentration degree and the cognitive load, mapping the concentration degree and the cognitive load to a two-dimensional cognitive state matrix, and carrying out interactive visual display so as to complete the assessment of the cognitive state. According to the method, the concentration degree and the cognitive load of the user can be calculated in real time and mapped to the two-dimensional cognitive state matrix, so that visual feedback is provided, and the user is helped to understand and adjust the cognitive state of the user.

Description

Technical Field

[0001] The present invention relates to the technical field of cognitive state assessment, and in particular to a cognitive state assessment method and a visual feedback system based on electroencephalogram (EEG) analysis. Background Art

[0002] With the development of brain-computer interface technology, the use of electroencephalogram (EEG) signals to assess the user's cognitive state has become a research hotspot. However, existing technologies mostly rely on complex multi-channel EEG acquisition equipment, which is not only expensive but also has poor portability and is difficult to promote in daily application scenarios. At the same time, traditional cognitive state assessment methods often only focus on a single dimension (such as concentration or fatigue) and lack a comprehensive characterization of the cognitive state. In addition, existing methods usually require laboratory-level signal pre-processing, with high algorithm complexity, making it difficult to achieve real-time feedback and interactive applications. Therefore, the current technology has the following major problems: 1. Complex equipment, a large number of channels, and poor user experience; 2. Single cognitive state assessment and lack of multi-dimensional integration; 3. High method complexity and difficulty in real-time operation on lightweight devices; 4. Lack of intuitive visual feedback mechanism, making it difficult for users to understand and adjust their own cognitive state. Summary of the Invention

[0003] The present invention provides a cognitive state assessment method and a visual feedback system based on electroencephalogram analysis to solve the technical problems mentioned in the background technology.

[0004] To achieve the above object, the technical solution of the present invention is achieved as follows:

[0005] The present invention provides a method for cognitive status assessment based on electroencephalogram analysis, comprising the following steps:

[0006] S1, collecting EEG signals and preprocessing the EEG signals to obtain preprocessed EEG signals;

[0007] S2. Analyze the preprocessed EEG signal using a sliding window method, extract the total energy of various waves in the EEG signal in different frequency bands, and then take the average value to obtain the frequency band energy mean of various waves;

[0008] S3, extracting blink electromyographic signals from the preprocessed EEG signals based on the frequency band energy averages of various waves, and then calculating the normalized blink frequency based on the blink electromyographic signals;

[0009] S4. Calculate concentration and cognitive load based on the frequency band energy averages of various waves and the normalized blink frequency;

[0010] S5. Normalize the concentration and cognitive load respectively and map them onto a two-dimensional cognitive state matrix, and interactively visualize the two-dimensional cognitive state matrix to complete the cognitive state assessment.

[0011] On the other hand, the present invention also provides a visual feedback system, including a device end, which evaluates the user's cognitive state according to the cognitive state evaluation method.

[0012] Beneficial effects of the present invention:

[0013] 1. The present invention discloses a cognitive state assessment method based on EEG analysis. This method can calculate the user's concentration and cognitive load in real time by analyzing dual-channel EEG signals of the prefrontal lobe and combining them with myoelectric signals such as blinking. After mapping the concentration and cognitive load into a two-dimensional cognitive state matrix, the method can provide intuitive visual feedback to the user, converting abstract neurophysiological indicators into an intuitive cognitive state matrix visualization representation, thereby realizing the assessment of cognitive state and helping users understand and adjust their own cognitive state.

[0014] In addition, when evaluating cognitive status, the present invention combines electromyographic signals such as blinking to construct a two-dimensional evaluation system for concentration and cognitive load, achieving multi-dimensional fusion. The cognitive status obtained by the evaluation is more accurate and comprehensive.

[0015] 2. When collecting EEG signals, the present invention only needs to collect EEG signals from two locations, that is, only the EEG signals of two channels at the dual-channel positions AF7 and AF8 of the frontal lobe need to be collected. Compared with the traditional method of evaluating cognitive status, the present invention requires better channels, is streamlined and efficient, greatly simplifies the complexity and cost of the equipment, and improves portability and user experience.

[0016] 3. The method used in the present invention is simpler than traditional cognitive status assessment methods, is suitable for real-time operation on mobile devices and wearable devices, and has a wide range of applications.

[0017] At the same time, through individualized baseline calibration and dynamic threshold adjustment, it can be adapted to different users. In addition, through individualized baseline calibration and dynamic threshold adjustment, the accuracy of cognitive status assessment is further improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 is a flowchart of the cognitive status assessment method of the present invention;

[0019] Figure 2 Schematic diagram of a two-dimensional cognitive state matrix for visualization;

[0020] Figure 3 This is a schematic diagram of the web application interface. DETAILED DESCRIPTION

[0021] To facilitate understanding of the present invention, the present invention will be described more fully below with reference to the accompanying drawings. The accompanying drawings illustrate preferred embodiments of the present invention. However, the present invention may be implemented in many other forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to provide a more thorough and comprehensive understanding of the present disclosure.

[0022] Reference Figure 1 The present invention provides a method for evaluating cognitive status based on EEG analysis, comprising the following steps:

[0023] S1, collecting EEG signals and preprocessing the EEG signals to obtain preprocessed EEG signals;

[0024] S2. Analyze the preprocessed EEG signal using a sliding window method, extract the total energy of various waves in the EEG signal in different frequency bands, and then take the average value to obtain the frequency band energy mean of various waves;

[0025] S3, extracting blink electromyographic signals from the preprocessed EEG signals based on the frequency band energy averages of various waves, and then calculating the normalized blink frequency based on the blink electromyographic signals;

[0026] S4. Calculate concentration and cognitive load based on the frequency band energy averages of various waves and the normalized blink frequency;

[0027] S5. Normalize the concentration and cognitive load respectively and map them onto a two-dimensional cognitive state matrix, and interactively visualize the two-dimensional cognitive state matrix to complete the cognitive state assessment.

[0028] The present invention discloses a cognitive state assessment method based on EEG analysis. The method can calculate the user's concentration and cognitive load in real time by analyzing dual-channel EEG signals of the prefrontal lobe and combining them with myoelectric signals such as blinking. After mapping the concentration and cognitive load into a two-dimensional cognitive state matrix, the method can provide intuitive visual feedback to the user, converting abstract neurophysiological indicators into an intuitive cognitive state matrix visualization representation, thereby realizing the assessment of cognitive state and helping users understand and adjust their own cognitive state.

[0029] In addition, when evaluating cognitive status, the present invention combines electromyographic signals such as blinking to construct a two-dimensional evaluation system for concentration and cognitive load, achieving multi-dimensional fusion. The cognitive status obtained by the evaluation is more accurate and comprehensive.

[0030] In some embodiments, the step S1 specifically includes the following steps:

[0031] S11. Use an EEG signal acquisition device to collect dual-channel EEG signals from positions AF7 and AF8 on both sides of the frontal lobe at a set sampling frequency (e.g., 250 Hz). In addition, the EEG signals collected by the present invention are not limited to positions AF7 and AF8 on both sides of the frontal lobe, but can also be collected from two other different positions in the left and right hemispheres of the brain. In addition, the number of channels is not limited to two, but can also be three, four, or more; however, the premise is that at least two channels are set at positions in the left and right hemispheres of the brain respectively.

[0032] S12. Use a (50 Hz) notch filter to eliminate power supply interference within the dual-channel EEG signals. Then, use a (0.5-45 Hz) bandpass filter to eliminate baseline drift and noise above a set frequency within the dual-channel EEG signals. The notch filter is implemented using the Python MNE library, and the bandpass filter is designed using the butter() function in the SciPy library.

[0033] S13: Use the adaptive threshold method (the threshold is set to the signal mean ± 3 times the standard deviation) to mark the large-scale artifacts of the dual-channel EEG signals obtained in S12, such as signal anomalies caused by head movement and poor electrode contact. The marked artifact segments are repaired using the local linear interpolation method implemented in the Pandas library to ensure signal continuity, and the preprocessed dual-channel EEG signals are obtained. The adaptive threshold method is implemented based on the NumPy library;

[0034] When collecting EEG signals, the present invention only needs to collect EEG signals from two positions, that is, only needs to collect EEG signals from two channels at the dual-channel positions AF7 and AF8 of the frontal lobe. Compared with traditional cognitive state assessment methods, the present invention requires better channels, is streamlined and efficient, greatly simplifies equipment complexity and cost, and improves portability and user experience.

[0035] In some embodiments, the step S2 specifically includes the following steps:

[0036] S21, using a sliding window method to analyze the preprocessed dual-channel EEG signals for a set time (e.g., 2 seconds), with the window overlap rate set to 50%;

[0037] S22. Apply fast Fourier transform (FFT) to the data in each window and extract the total energy of the 0.5Hz to 4Hz frequency band of the two channels of delta waves at positions AF7 and AF8 on both sides of the frontal lobe, respectively, in μV. 2 ;

[0038] S23. Apply fast Fourier transform (FFT) to the data in each window and extract the total energy of the 4Hz to 8Hz frequency band of the two channels of theta waves at positions AF7 and AF8 on both sides of the frontal lobe, respectively, in μV. 2 ;

[0039] S24. Apply fast Fourier transform (FFT) to the data in each window and extract the total energy of the 8Hz to 13Hz frequency band of the two channels of α waves at positions AF7 and AF8 on both sides of the frontal lobe, in μV. 2 ;

[0040] S25. Apply fast Fourier transform (FFT) to the data in each window and extract the total energy of the 13Hz to 30Hz frequency band of the beta wave in the two channels AF7 and AF8 on both sides of the frontal lobe, in μV. 2 ;

[0041] S26. Apply fast Fourier transform (FFT) to the data in each window and extract the total energy of the Y wave from 30 Hz to 45 Hz in the two channels at positions AF7 and AF8 on both sides of the frontal lobe, in μV. 2 ;

[0042] S27. For each wave, the average total energy of the two channels AF7 and AF8 at the bilateral frontal lobe was calculated to obtain the average energy of the delta, theta, alpha, beta, and y waves, respectively. These average energy values ​​reflect different types of neural activity: delta waves are associated with deep sleep and attention deficits, theta waves are associated with working memory load and attention switching, alpha waves are associated with alertness and relaxation, beta waves are associated with active thinking and focus, and y waves are associated with higher-level cognitive processing and multisensory integration.

[0043] In some embodiments, the total energy of the 0.5 Hz to 4 Hz frequency band of the delta waves of the two channels at the positions AF7 and AF8 on both sides of the frontal lobe in S22 is calculated as follows:

[0044]

[0045] Among them, P δ,ch Indicates the total energy of the 0.5Hz to 4Hz frequency band of the delta wave on the channel on AF7 or AF8; ch indicates the channel on AF7 or AF8; f1 indicates the current frequency extracted from the delta wave; X ch Indicates the power spectrum density of channel ch at the current frequency, in μV 2 / Hz;

[0046] The total energy of the 4 Hz to 8 Hz frequency band of the two channels of theta waves at the positions AF7 and AF8 on both sides of the frontal lobe in S23 is calculated as follows:

[0047]

[0048] Among them, P θ,ch Indicates the total energy of the 4Hz to 8Hz frequency band of the theta wave of the channel on AF7 or AF8; f2 indicates the current frequency extracted by the theta wave;

[0049] The total energy of the 8 Hz to 13 Hz frequency band of the two channels of α waves at the positions AF7 or AF8 on both sides of the frontal lobe in S24 is calculated as follows:

[0050]

[0051] Among them, P α，ch Indicates the total energy of the 8Hz to 13Hz frequency band of the a wave of the channel on AF7 or AF8; f3 indicates the current frequency extracted by the α wave;

[0052] The total energy of the 13 Hz to 30 Hz frequency band of the beta waves of the two channels at the positions AF7 or AF8 on both sides of the frontal lobe in S25 is calculated as follows:

[0053]

[0054] Among them, P β，ch Indicates the total energy of the 13Hz to 30Hz frequency band of the beta wave of the channel on AF7 or AF8; f4 indicates the current frequency extracted by the beta wave;

[0055] The total energy of the 30 Hz to 45 Hz frequency band of the two channels of gamma waves at the positions AF7 or AF8 on both sides of the frontal lobe in S26 is calculated as follows:

[0056]

[0057] Among them, P γ,ch Indicates the total energy of the 30Hz to 45Hz frequency band of the gamma wave of the channel on AF7 or AF8; f5 indicates the current frequency extracted by the Y wave.

[0058] The calculation formula for the energy of the δ wave, θ wave, α wave, β wave and γ wave frequency band in S27 is:

[0059]

[0060] Among them, P δ 、P θ 、P ɑ 、P β and P γ represents the frequency band energy mean of δ wave, θ wave, α wave, β wave and γ wave respectively; P δ，AF7 、P δ，AF8 represents the total energy of the delta wave frequency band of the channel on AF7 or AF8; P θ，AF7 、P θ,AF8 represents the total energy of the frequency band of the θ wave on the channel of AF7 or AF8; P a，AF7 、P a，AF8represents the total energy of the α wave frequency band of the channel on AF7 or AF8; P β，AF7 、P β,AF8 represents the total energy of the beta wave frequency band of the channel on AF7 or AF8; P γ，AF7 、P γ，AF8 Represents the total energy of the gamma wave frequency band of the channel on AF7 or AF8 respectively.

[0061] In some embodiments, S3 specifically includes the following steps:

[0062] S31. Extract blink electromyographic signals from the preprocessed dual-channel EEG signals. The extraction formula used is as follows:

[0063]

[0064] Among them, E blink represents the electromyographic signal of eye blink; x max 、x min are the maximum and minimum values ​​of the signal amplitude in the current window (unit: μV); std(x) represents the standard deviation of the signal amplitude in the window (unit: μV); τ blink Indicates the preset blink electromyographic signal threshold (typical value is 5);

[0065] S32. Obtain the number of blinks per unit time based on the blink electromyographic signal, then preset a minimum and maximum blink frequency, and calculate a normalized blink frequency based on the number of blinks per unit time, the preset minimum and maximum blink frequency, as follows:

[0066]

[0067] Among them, Blink norm represents the normalized blink frequency, ranging from 0 to 1, representing the relative blink activity level; f blink Indicates the number of blinks per unit time (unit: times / minute); f max 、f min These are the preset minimum and maximum blink rates, with typical values ​​being 5 times / minute and 30 times / minute respectively.

[0068] In some embodiments, the S4 specifically includes the following steps:

[0069] S41. Calculate the degree of α-wave suppression based on the frequency band energy mean of the α-wave, and then calculate the degree of concentration based on the ratio of the frequency band energy mean of the β-wave and the θ-wave and the degree of α-wave suppression;

[0070] S42. Calculate the difference between the left and right hemispheres, and calculate the cognitive load based on the ratio of the frequency band energy means of the gamma and alpha waves, the frequency band energy mean of the theta wave, and the difference between the left and right hemispheres.

[0071] In some embodiments, the calculation formula for the concentration in S41 is:

[0072]

[0073] Among them, Attention means concentration; The ratio of the energy averages of the two frequency bands of beta and theta waves reflects the activity of cognitive activities; It indicates the degree of alpha wave suppression, corresponding to the enhancement of visual attention; Indicates the alpha wave energy baseline in the resting state, unit: μV 2 ; Indicates the normalized inverse of the blink frequency, unitless, range 0-1, through 1-Blink norm Calculate the normalized inverse of blink frequency The physiological phenomenon that a decrease in blink frequency is usually accompanied by an increase in concentration is taken into account; w1, w2, and w3 represent the weight coefficients of each item, with typical values ​​of 0.5, 0.3, and 0.2, respectively.

[0074] In some embodiments, the calculation formula for the cognitive load in S42 is:

[0075]

[0076] Among them, CognitiveLoad represents cognitive load; It represents the ratio of the frequency band energy averages of gamma waves and alpha waves; Indicates the difference between the left and right hemispheres.

[0077] P θ The item reflects the positive correlation between prefrontal theta wave activity and working memory load. A large number of studies have shown that an increase in theta wave energy usually corresponds to an increase in cognitive load. The term represents the ratio of high-frequency to low-frequency activity. Enhanced Y-waves are usually associated with complex cognitive processing, while weakened a-waves correspond to increased allocation of attentional resources. The β-wave asymmetry measure measures left-right hemispheric beta wave asymmetry, reflecting the complexity of the cognitive task and the impact of emotional state on cognitive load. Furthermore, these three indicators are independent and complementary, comprehensively capturing different aspects of cognitive load. The weighting coefficients are set based on their relative importance in predicting cognitive load.

[0078] In some embodiments, reference Figure 2 and Figure 3 , the S5 specifically includes the following steps:

[0079] S51. Normalize the calculated concentration and cognitive load and map them to a two-dimensional cognitive state matrix. Use the Matplotlib library and the Plotly library to interactively visualize the two-dimensional cognitive state matrix. The visualized two-dimensional cognitive state matrix includes four quadrants, each representing a different cognitive state. The first quadrant (high concentration, high cognitive load) corresponds to "cognitive striving"; the second quadrant (high concentration, low cognitive load) corresponds to "flow state"; the third quadrant (low concentration, low cognitive load) corresponds to "cognitive wandering"; and the fourth quadrant (low concentration, high cognitive load) corresponds to "cognitive fatigue." See the schematic diagram of the visualized two-dimensional cognitive state matrix for details. Figure 2 As shown;

[0080] S52. Use the Flask framework to build a web application interface. In the web application interface, a two-dimensional cognitive state matrix is ​​updated in real time according to the user's cognitive state. Color coding is used to indicate the dwell time and provide dynamic visual feedback.

[0081] S53. The visual feedback system uses the Pandas library to record the changing trajectory of the user's cognitive state. Then, it uses the Seaborn library to generate a time series heat map based on the real-time transformation data of the visually displayed two-dimensional cognitive state matrix, making it easier for users to understand the dynamic changes in their own cognitive state.

[0082] The following is a specific example of evaluating a user's cognitive state in a computer programming task, illustrating the implementation process of the method of the present invention and the data transformation in each step.

[0083] Step S1: Collect the frontal lobe dual-channel (AF7 and AF8) EEG signals of a user during the programming task, with a sampling rate of 250Hz, using the notch in the MNE library _ The EEG signal was preprocessed using a 50Hz notch filter to eliminate power supply interference and a 0.5-45Hz bandpass filter to filter the EEG signal. The preprocessed EEG signal was then obtained using numpy.mean()±3*numpy.std() to set an adaptive threshold for artifact detection, and pandas.DataFrame.interpolate() was used for local repair.

[0084] Step S2: Perform a 2-second sliding window analysis on the preprocessed signal (implemented using the numpy.lib.stride_tricks.sliding_window_view() function). Calculate the power spectra of the AF7 and AF8 channels using scipy.signal.periodogram() and extract the energy of each frequency band. During the difficult phase of the programming task, a significant increase in beta wave energy and a significant suppression of alpha waves were observed, indicating that the user was in a state of high concentration. The frequency band energy distribution is as follows:

[0085] AF7 channel: P δ，AF7 =11.8μV 2 , P θ,AF7 =19.2μV 2 , P α，AF7 =15.1μV 2 , P β，AF7 =37.8μV 2 , P γ，AF7 =10.2μV 2 ;

[0086] AF8 channel: P δ，AF8 =12.8μV 2 , P θ,AF8 =18.2μV 2 , P α，AF8 =13.3μV 2 , P β，AF8 =33.4μV 2 , P γ，AF8 =9.4μV 2 ;

[0087] Average: P δ =12.3μV 2 , P θ =18.7μV 2 , P α =14.2μV 2 , Pβ=35.6μV 2 , P γ =9.8μV 2 ;

[0088] Step S3: Use the threshold judgment algorithm to identify that the number of blinks has decreased significantly (only 3 times in 2 minutes, far lower than the normal frequency of 6-15 times / minute). Calculate the number of blinks per unit time (or blink frequency) f blink = 1.5 times / minute, using the minimum blink rate f min =5 times / minute and the blink frequency is the maximum value f max =30 times / minute for normalization to obtain the normalized blink frequency Blinknorm =(1.5-5) / (30-5)=0 (since the result is less than 0, take 0).

[0089] Step S4: Calculate the concentration and cognitive load based on the above features.

[0090] Step S4-1: Set the β / θ ratio (1.90), α wave suppression (current P α =14.2μV 2 , baseline inhibited by 35%) and is inversely proportional to blink frequency Enter the concentration formula and get:

[0091]

[0092] After normalization, the concentration value is 79%.

[0093] Step S4-2: Theta wave energy (P θ =18.7μV 2 ), γ / α ratio and hemispheric beta wave asymmetry Entering the cognitive load formula, we get:

[0094] CDgnitiveLoad=0.4×18.7+0.4×0.69+0.2×0.06=7.48+0.28+0.01

[0095] =7.77

[0096] After normalization, the cognitive load value was 65%.

[0097] Step S5: Use the Matplotlib library to build a cognitive state matrix visualization interface, map the concentration (78%) and cognitive load (65%) to the cognitive state matrix, and locate the user in the "cognitive striving" quadrant, indicating that the user is investing a lot of cognitive resources to solve programming problems. Figure 3 As shown, the system transmits the visualization results to the web interface through the Flask framework, marks this state with a red area, and uses HTML+CSS+JavaScript to achieve dynamic updates, prompting "high-intensity thinking state, it is recommended to take a short break of 5 minutes every 25 minutes to prevent cognitive fatigue."

[0098] As the programming task continued, the user gradually adapted to the task's difficulty, with cognitive load decreasing from 65% to 42%, while focus remained around 75%. The state point in the matrix gradually moved from the "Cognitive Striving" quadrant to the "Flow State" quadrant. The system, using the Pandas library to track state changes, identified this change and updated the prompt to read, "Flow State has been achieved. Maintain the current task rhythm to maintain optimal efficiency."

[0099] The method used in the present invention is simpler than traditional cognitive status assessment methods, is suitable for real-time operation on mobile devices and wearable devices, and has a wide range of applications.

[0100] At the same time, through individualized baseline calibration and dynamic threshold adjustment, it can be adapted to different users. In addition, through individualized baseline calibration and dynamic threshold adjustment, the accuracy of cognitive status assessment is further improved.

[0101] A second aspect of the present invention further provides a visual feedback system, including a device end, which evaluates the user's cognitive state according to a cognitive state evaluation method.

[0102] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art who is familiar with the technical field can easily think of changes or replacements within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. In addition, the technical solutions between the various embodiments of the present invention can be combined with each other, but it must be based on the ability of ordinary technicians in this field to implement. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the protection scope required by the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A method for cognitive status assessment based on EEG analysis, characterized in that: The steps include: S1, collecting EEG signals and preprocessing the EEG signals to obtain preprocessed EEG signals; S2. Analyze the preprocessed EEG signal using a sliding window method, extract the total energy of various waves in the EEG signal in different frequency bands, and then take the average value to obtain the frequency band energy mean of various waves; S3, extracting blink electromyographic signals from the preprocessed EEG signals based on the frequency band energy averages of various waves, and then calculating the normalized blink frequency based on the blink electromyographic signals; S4. Calculate concentration and cognitive load based on the frequency band energy averages of various waves and the normalized blink frequency; S5. Normalize the concentration and cognitive load respectively and map them onto a two-dimensional cognitive state matrix, and interactively visualize the two-dimensional cognitive state matrix to complete the cognitive state assessment.

2. The method for cognitive status assessment based on EEG analysis according to claim 1, characterized in that: The S1 specifically includes the following steps: S11, using an EEG signal acquisition device to collect dual-channel EEG signals from positions AF7 and AF8 on both sides of the frontal lobe at a set sampling frequency; S12, using a notch filter to eliminate power supply interference in the dual-channel EEG telecom; then using a bandpass filter to eliminate baseline drift and noise above a set frequency in the dual-channel EEG telecom; S13. Adopt the adaptive threshold method to mark the artifacts of the dual-channel EEG signals finally obtained in S12, and use the local linear interpolation method to repair the marked artifact segments to ensure signal continuity, thereby obtaining the preprocessed dual-channel EEG signals.

3. The method for cognitive status assessment based on EEG analysis according to claim 2, characterized in that: The S2 specifically includes the following steps: S21, using the sliding window method to analyze the preprocessed dual-channel EEG signals, with the window overlap rate set to 50%; S22, applying fast Fourier transform (FFT) to the data in each window, extracting the total energy of the 0.5 Hz to 4 Hz frequency band of the delta wave of two channels at positions AF7 and AF8 on both sides of the frontal lobe; S23, applying fast Fourier transform (FFT) to the data in each window, extracting the total energy of the 4 Hz to 8 Hz frequency band of the theta waves of two channels at positions AF7 and AF8 on both sides of the frontal lobe; S24, applying fast Fourier transform (FFT) to the data in each window, extracting the total energy of the 8 Hz to 13 Hz frequency band of the alpha waves of the two channels at positions AF7 and AF8 on both sides of the frontal lobe; S25. Apply fast Fourier transform (FFT) to the data in each window to extract the total energy of the 13 Hz to 30 Hz frequency band of the beta waves in the two channels AF7 and AF8 at both sides of the frontal lobe; S26, applying fast Fourier transform (FFT) to the data in each window, extracting the total energy of the 30 Hz to 45 Hz frequency band of the two channels of gamma waves at positions AF7 and AF8 on both sides of the frontal lobe; S27. For various waves, calculate the average value of the total frequency band energy of the two channels AF7 and AF8 at both sides of the frontal lobe, and obtain the frequency band energy mean values ​​of the delta wave, theta wave, alpha wave, beta wave and gamma wave respectively.

4. The method for cognitive status assessment based on EEG analysis according to claim 3, characterized in that: The total energy of the 0.5 Hz to 4 Hz frequency band of the delta wave of the two channels at the positions AF7 and AF8 on both sides of the frontal lobe in S22 is calculated as follows: Among them, P δ,ch Indicates the total energy of the 0.5Hz to 4Hz frequency band of the delta wave on the channel on AF7 or AF8; ch indicates the channel on AF7 or AF8; f1 indicates the current frequency extracted from the delta wave; X ch Represents the power spectral density of channel ch at the current frequency; The total energy of the 4 Hz to 8 Hz frequency band of the two channels of theta waves at the positions AF7 and AF8 on both sides of the frontal lobe in S23 is calculated as follows: Among them, P θ,ch Indicates the total energy of the 4Hz to 8Hz frequency band of the theta wave of the channel on AF7 or AF8; f2 indicates the current frequency extracted by the theta wave; The total energy of the 8 Hz to 13 Hz frequency band of the two channels of α waves at the positions AF7 or AF8 on both sides of the frontal lobe in S24 is calculated as follows: Among them, P α,ch Indicates the total energy of the 8Hz to 13Hz frequency band of the alpha wave of the channel on AF7 or AF8; f3 indicates the current frequency extracted by the alpha wave; The total energy of the 13 Hz to 30 Hz frequency band of the beta waves of the two channels at the positions AF7 or AF8 on both sides of the frontal lobe in S25 is calculated as follows: Among them, P β,ch Indicates the total energy of the 13Hz to 30Hz frequency band of the beta wave of the channel on AF7 or AF8; f4 indicates the current frequency extracted by the beta wave; The total energy of the 30 Hz to 45 Hz frequency band of the two channels of gamma waves at the positions AF7 or AF8 on both sides of the frontal lobe in S26 is calculated as follows: Among them, P γ,ch Indicates the total energy of the 30Hz to 45Hz frequency band of the gamma wave of the channel on AF7 or AF8; f5 indicates the current frequency extracted by the gamma wave. The calculation formula for the energy of the δ wave, θ wave, α wave, β wave and γ wave frequency band in S27 is: Among them, P δ 、P θ 、P ɑ 、P β and P γ represents the frequency band energy mean of δ wave, θ wave, α wave, β wave and γ wave respectively; P δ，AF7 、P δ，AF8 represents the total energy of the delta wave frequency band of the channel on AF7 or AF8; P θ，AF7 、P θ,AF8 represents the total energy of the frequency band of the θ wave on the channel of AF7 or AF8; P a，AF7 、P α,AF8 represents the total energy of the α wave frequency band of the channel on AF7 or AF8; P β,AF7 、P β，AF8 represents the total energy of the beta wave frequency band of the channel on AF7 or AF8; P γ，AF7 、P γ，AF8 Represents the total energy of the gamma wave frequency band of the channel on AF7 or AF8 respectively.

5. The method for cognitive status assessment based on EEG analysis according to claim 4, characterized in that: The S3 specifically includes the following steps: S31. Extract blink electromyographic signals from the preprocessed dual-channel EEG signals. The extraction formula used is as follows: Among them, E blink represents the electromyographic signal of eye blink; x max 、x min are the maximum and minimum values ​​of the signal amplitude in the current window respectively; std(x) represents the standard deviation of the signal amplitude in the window; τ blink Indicates the preset blink electromyographic signal threshold; S32. Obtain the number of blinks per unit time based on the blink electromyographic signal, then preset a minimum and maximum blink frequency, and calculate a normalized blink frequency based on the number of blinks per unit time, the preset minimum and maximum blink frequency, as follows: Among them, Blink norm represents the normalized blink frequency; f blink Indicates the number of blinks per unit time; f max 、f min They are the preset minimum and maximum values ​​of blink frequency respectively.

6. The method for cognitive status assessment based on EEG analysis according to claim 5, characterized in that: The S4 specifically includes the following steps: S41. Calculate the degree of α-wave suppression based on the frequency band energy mean of the α-wave, and then calculate the degree of concentration based on the ratio of the frequency band energy mean of the β-wave and the θ-wave and the degree of α-wave suppression; S42. Calculate the difference between the left and right hemispheres, and calculate the cognitive load based on the ratio of the frequency band energy means of the gamma and alpha waves, the frequency band energy mean of the theta wave, and the difference between the left and right hemispheres.

7. The method for cognitive status assessment based on EEG analysis according to claim 6, characterized in that: The calculation formula of the concentration in S41 is: Among them, Attention means concentration; The ratio of the energy averages of the two frequency bands of beta and theta waves reflects the activity of cognitive activities; It indicates the degree of alpha wave suppression, corresponding to the enhancement of visual attention; It represents the a-wave energy baseline in the resting state; Represents the normalized inverse of blink frequency, expressed as 1-Blink norm Calculation; w1, w2, w3 represent the weight coefficients of each item respectively.

8. The method for cognitive status assessment based on EEG analysis according to claim 7, characterized in that: The calculation formula of cognitive load in S42 is: Among them, CognitiveLoad represents cognitive load; It represents the ratio of the frequency band energy averages of gamma waves and alpha waves; It represents the difference between the left and right hemispheres and is used to measure the asymmetry of beta waves between the left and right hemispheres.

9. The method for cognitive status assessment based on EEG analysis according to claim 8, characterized in that: The S5 specifically includes the following steps: S51. Normalize the calculated concentration and cognitive load and map them onto a two-dimensional cognitive state matrix. Use the Matplotlib and Plotly libraries to interactively visualize the two-dimensional cognitive state matrix. The two-dimensional cognitive state matrix presented visually includes four quadrants, each representing a different cognitive state; S52. Build a web application interface, update a two-dimensional cognitive state matrix in real time according to the user's cognitive state in the web application interface, and use color coding to indicate the dwell time to provide dynamic visual feedback; S53. Use the Seabom library and generate a time series heat map based on the real-time transformation data of the visually displayed two-dimensional cognitive state matrix to help users understand the dynamic changes in their own cognitive state.

10. A visual feedback system, characterized in that: The device comprises a device end, which evaluates the user's cognitive state according to the cognitive state evaluation method according to any one of claims 1 to 9.

Images