Skill proficiency evaluation method based on beta-alpha electroencephalogram index measurement

Abstract

The invention discloses a beta-alpha electroencephalogram index-based skill proficiency evaluation method, which is characterized by comprising the following steps of: setting a consciousness control processing benchmark task, an automatic processing benchmark task and a professional skill test task, and collecting electroencephalogram signals of a tested person; according to the preprocessed electroencephalogram data, calculating an average wave band power value of baseline normalization of alpha and beta wave bands within 300-450ms; constructing a beta-alpha index, calculating a beta-alpha index value TRi when the judgment is correct and a beta-alpha index value TWi when the judgment is wrong in the test task of the ith professional skill, calculating the positive familiarity score of the testee as Score + = Total familiarity + / N, and calculating the negative familiarity score of the testee as Score-= Total familiarity- / N, the larger the numerical value is, the higher the habitual error is; the method has the advantages that the beta-alpha combination index is constructed by analyzing the task specificity change of alpha and beta wave bands in the electroencephalogram signal data, and the familiarity degree of the tested person for specific professional skill problems is objectively and directly evaluated.

Description

Technical Field

[0001] The present invention relates to the technical field of electroencephalogram (EEG) nerve function measurement, and in particular to a skill proficiency evaluation method based on beta-alpha EEG index measurement. Background Art

[0002] Across all industries, experts who possess proficiency in specific professional fields are highly valued and in short supply. The key difference between them and ordinary professionals, who possess knowledge and can make accurate judgments through careful consideration, lies in their level of proficiency, which is reflected in the different brain processing modes for specific problems. Experts tend to operate in an intuitive, automated mode of brain processing, while ordinary professionals tend to operate in a conscious, controlled mode of processing. The transition from ordinary professionals to domain experts requires a long period of learning, practice, reflection, and comprehensive training, requiring continuous repetition and accumulation to achieve a transformation in processing mode. However, there is currently a lack of direct and accurate methods for evaluating proficiency based on this brain processing mode, which has, to a certain extent, hindered the accurate assessment and effective training of professionals in specialized fields.

[0003] Existing research shows that EEG activity patterns can reflect individual differences in information processing and decision-making patterns. In particular, neural oscillations in the beta and alpha bands play a crucial role in cognitive decision-making. This paper constructs a β-α index for measurement and analysis. By observing differences in this index signal under specific conditions, it proposes an evaluation method for measuring proficiency. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a skill proficiency evaluation method based on β-α EEG index measurement, which can effectively evaluate an individual's proficiency in skills in a specific professional field by analyzing the conditional difference characteristics of EEG signals.

[0005] The technical solution adopted by the present invention to solve the above technical problems is: a skill proficiency evaluation method based on β-α EEG index measurement, comprising the following specific steps:

[0006] (1) Have the person being tested wear a multi-channel EEG measurement device to record EEG signals;

[0007] (2) Setting up a conscious control processing benchmark task, an automatic processing benchmark task, and a professional skills test task, wherein the professional skills test task includes two situations where the participant makes correct and incorrect judgments, collecting the EEG signals of the subjects in the three types of tasks, and obtaining the original EEG data of the subjects;

[0008] (3) Preprocessing the raw EEG data, including signal amplification, segmentation, signal noise reduction, bandpass filtering and artifact removal;

[0009] (4) Based on the preprocessed EEG data, the baseline normalized band power averages of α and β within the time window of 300-450ms were calculated for the subjects under different tasks, and the index value of β-α was calculated, that is, the baseline normalized band power average of β minus the baseline normalized band power average of α;

[0010] (5) Compare the changes in the β-α index values ​​of the subjects under different tasks and calculate the average β-α index value S of the conscious control processing benchmark task. CP and the average β-α index value S of the automatic processing benchmark task AP There are N professional skills, each professional skill corresponds to a test task, and the average β-α index value TR when the judgment is correct in the test task of the i-th professional skill is calculated one by one i i∈N, if TR i <(S CP +2S AP ) / 3, it means that the subject's i-th professional skill has reached the positive proficiency level; calculate the average β-α index value TW when the judgment error occurs in the test task of the i-th professional skill one by one i i∈N, if TW i <(S CP +2S AP ) / 3, it means that the subject's i-th professional skill has reached negative proficiency, that is, habitual errors; count all skills that have reached positive proficiency and whose correct number is greater than 2×the number of errors. 熟悉+ , then the final positive familiarity score of the subject is: Score + = Total 熟悉+ / N, the larger the value, the higher the familiarity of the professional skills in the field; count all skills that have reached negative familiarity and the number of errors > 2 × the number of correct answers Total 熟悉- , then the final negative familiarity score of the subject is: Score-=Total 熟悉- / N, the larger the value, the higher the habitual errors in professional skills in this field.

[0011] Furthermore, in the step (2), the conscious control processing benchmark task is a mental arithmetic task of square roots within 20, the automatic processing benchmark task is a mental arithmetic task of addition and subtraction within the number 3, and the professional skills test task includes two situations in which the participants judge correctly and incorrectly, which are the main skills knowledge and decision-making of jobs in different professional fields. Each professional skills test task includes four components: background description, question, correct answer, and interference answer. The question part requires short text. The test extracts the EEG data of the test person within 2000ms from the display of the question. The conscious control processing benchmark task and the automatic processing benchmark task each include more than 20 tasks. The number of professional skills test tasks is set according to actual conditions, and each task is repeated 3 times.

[0012] Furthermore, in the step (3), during the preprocessing of the raw EEG data, the raw EEG data is band-pass filtered at 0.5 to 40 Hz, and the FastICA algorithm based on the maximum negative entropy principle is used to perform independent component analysis and data reorganization to effectively remove artifact interference; and the α and β bands of the preprocessed EEG data are retained as target data for subsequent analysis.

[0013] Furthermore, in step (4), the baseline normalized band power index value is calculated as follows:

[0014] (4-1) Perform wavelet transform on the pre-processed EEG data to obtain the band power values ​​of the α and β bands in the CPz channel in the interval 300 to 450 ms after the stimulus (question display) page appears;

[0015] (4-2) Band power calculations were performed for the conscious control processing benchmark task, the automatic processing benchmark task, the correct judgment professional skill test task, and the incorrect judgment professional skill test task. The baseline was taken from -300ms to -100ms before the stimulus, and the baseline-normalized band power (dB) was obtained. The calculation relationship is as follows: Palpha_Scp_dB=10×log10(Palpha_Scp / base_α_Scp)

[0016] Pbeta_Scp_dB=10×log 10(Pbeta_Scp / base_β_Scp)

[0017] Palpha_Sap_dB=10×log 10(Palpha_Sap / base_α_Sap)

[0018] Pbeta_Sap_dB=10×log 10(Pbeta_Sap / base_β_Sap)

[0019] Palpha_Tr_dB i =10×log 10(Palpha_Tr i / base_α_Tr i )

[0020] Pbeta_Tr_dB i =10×log 10(Pbeta_Tr i / base_β_Tr i )

[0021] Palpha_Tw_dB i =10×log 10(Palpha_Tw i / base_α_Tw i )

[0022] Pbeta_Tw_dB i =10×log 10(Pbeta_Tw i / base_β_Tw i )

[0023] Among them: the symbol Palpha represents the alpha band power, the symbol Pbeta represents the beta band power, the symbol Scp represents the conscious control processing benchmark task, the symbol Sap represents the automatic processing benchmark task, the symbol Tr represents the professional skill test task with correct judgment, and the symbol Tw represents the professional skill test task with incorrect judgment. Indicates the average of this type of task, that is: It represents the average of multiple repeated tests of the test task of the i-th professional skill, that is: k is the frequency value, which ranges from 9 to 30 Hz; the symbol dB represents the value obtained after baseline normalization processing, and the symbol base represents the average band power during the baseline period of -300ms to -100ms before stimulation. For example, Palpha_Scp represents the α-band power in the conscious control processing benchmark task, Palpha_Scp_dB represents Palpha_Scp after baseline normalization processing, and base_α_Scp represents the average band power of the α-band during the baseline period of -300ms to -100ms before stimulation in the conscious control processing benchmark task.

[0024] Furthermore, in the step (5), the S calculated based on the β-α index valueCP 、S AP , TR i 、TW i Total 熟悉+ Total 熟悉- , Score+, Score- calculation formula is:

[0025] S CP =Pbeta_Scp_dB-Palpha_Scp_dB,S AP =Pbeta_Sap_dB-Palpha_Sap_dB

[0026] TR i =Pbeta_Tr_dB i -Palpha_Tr_dB i ,TW i =Pbeta_Tw_dB i -Palpha_Tw_dB i

[0027] Score + = Total 熟悉+ / N

[0028] Score-=Total 熟悉- / N.

[0029] Compared with the prior art, the present invention has the following advantages:

[0030] (1) This method objectively and directly evaluates the brain processing mode of the subjects for specific professional skill problems by analyzing the task-specific changes in the α and β bands in the EEG signal data. It plays an important role in measuring the skill familiarity of the subjects and in rationally identifying, selecting and training professional skill experts. For professional field skill test questions, this method also proposes formatting processing specifications that are convenient for EEG testing: background description, question, correct answer, and interference answer.

[0031] (2) This method uses the baseline normalized band power values ​​of α and β to construct a β-α combination index to extract the neural activity characteristics within the critical time window of 300-450ms. It is simple, sensitive and efficient for identifying processing modes and familiarity levels. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 Flowchart of the present invention. DETAILED DESCRIPTION

[0033] The present invention will be described in further detail below with reference to the accompanying drawings and embodiments.

[0034] As shown in the figure, the skill proficiency evaluation method based on β-α EEG index measurement includes the following specific steps:

[0035] (1) The person being tested is provided with a multi-channel EEG measurement device to record EEG signals. The multi-channel EEG measurement device can use an existing EEG acquisition system, such as the EMOTIV EPOC FlexSaline Sensor Kit EEG acquisition system.

[0036] (2) Set up a benchmark task for conscious control processing, using a square root mental arithmetic task within 20: There are 20 items in total, and the stimulus presentation order is as follows: the question is displayed, and after a random delay of 600ms to 1000ms, the correct answer and distractor answer are displayed, waiting for the subject to press a button to select. The automated processing benchmark task uses mental arithmetic tasks involving addition and subtraction within the number 3, such as 1 + 1 = ?. There are 20 items in total, and the stimulus presentation order is the same as above. The professional skills test tasks involve participants making correct and incorrect judgments, and use key skills and decision-making questions from different professional fields. Each test task consists of four components: a context description, a question, a correct answer, and distractor answers. The question portion is required to be brief. For example, the decision-making question in driving skills includes the context description "You are driving a vehicle on a city road and encounter a traffic light at an intersection. Please make a driving decision based on the signal." The question is "Red light," the correct answer is "Brake," and the distractor answer is "Turn left." The number of professional skills test tasks is set based on the actual situation, and each task is repeated three times. The stimulus presentation sequence was as follows: the background description was displayed for 10 seconds, then disappeared, followed by the question. After a random interval of 600 to 1000 ms, the correct answer and distractor answer were displayed, and the subject was asked to press a key to select. EEG signals were collected from the subjects in the three tasks from the start of question presentation to 2000 ms to obtain the raw EEG data.

[0037] (3) Preprocessing the raw EEG data, including signal amplification, segmentation interception, signal noise reduction, bandpass filtering and artifact removal; wherein: the raw EEG data is bandpass filtered at 0.5-40 Hz, and the FastICA algorithm based on the maximum negative entropy principle is used for independent component analysis and data reorganization to effectively remove electrooculogram, electromyography and other artifact interference; and the preprocessed EEG data retains α (9-13 Hz) and β (14-30 Hz) as the target data for subsequent analysis, providing a clear and reliable basic signal for familiarity assessment;

[0038] (4) Based on the pre-processed EEG data, the baseline normalized band power averages of α and β within the time window of 300-450ms are calculated for the subjects under different tasks, and the index value of β-α is calculated, that is, the baseline normalized band power average of β minus the baseline normalized band power average of α; specifically:

[0039] (4-1) Perform wavelet transform on the pre-processed EEG data to obtain the band power values ​​of the α and β bands in the CPz channel in the interval of 300 to 450 ms after the problem display page appears;

[0040] (4-2) Band power calculations were performed for the conscious control processing benchmark task, the automatic processing benchmark task, the correct judgment professional skill test task, and the incorrect judgment professional skill test task. The baseline was taken from -300ms to -100ms before the stimulus, and the baseline-normalized band power (dB) was obtained. The calculation relationship is as follows: Palpha_Scp_dB=10×log10(Palpha_Scp / base_α_Scp)

[0041] Pbeta_Scp_dB=10×log 10(Pbeta_Scp / base_β_Scp)

[0042] Palpha_Sap_dB=10×log 10(Palpha_Sap / base_α_Sap)

[0043] Pbeta_Sap_dB=10×log 10(Pbeta_Sap / base_β_Sap)

[0044] Palpha_Tr_dB i =10×log 10(Palpha_Tr i / base_α_Tr i )

[0045] Pbeta_Tr_dB i =10×log 10(Pbeta_Tr i / base_β_Tr i )

[0046] Palpha_Tw_dB i =10×log 10(Palpha_Tw i / base_α_Twi )

[0047] Pbeta_Tw_dB i =10×log 10(Pbeta_Tw i / base_β_Tw i )

[0048] Among them: the symbol Palpha represents the alpha band power, the symbol Pbeta represents the beta band power, the symbol Scp represents the conscious control processing benchmark task, the symbol Sap represents the automatic processing benchmark task, the symbol Tr represents the professional skill test task with correct judgment, and the symbol Tw represents the professional skill test task with incorrect judgment. Indicates the average of this type of task, that is: It represents the average of multiple repeated tests of the test task of the i-th professional skill, that is: k is the frequency value, which ranges from 9 to 30 Hz; the symbol dB represents the value obtained after baseline normalization processing, and the symbol base represents the average band power during the baseline period of -300ms to -100ms before stimulation. For example, Palpha_Scp represents the alpha band power in the conscious control processing benchmark task, Palpha_Scp_dB represents Palpha_Scp after baseline normalization processing, and base_α_Scp represents the average band power of the alpha band during the baseline period of -300ms to -100ms before stimulation in the conscious control processing benchmark task;

[0049] (5) Compare the changes in the β-α index values ​​of the subjects under different tasks and calculate the average β-α index value S of the conscious control processing benchmark task. CP and the average β-α index value S of the automatic processing benchmark task AP There are N professional skills, each professional skill corresponds to a test task, and the average β-α index value TR when the judgment is correct in the test task of the i-th professional skill is calculated one by one i i∈N, if TR i <(S CP +2S AP ) / 3, it means that the subject's i-th professional skill has reached the positive proficiency level; calculate the average β-α index value TW when the judgment error occurs in the test task of the i-th professional skill one by one i i∈N, if TW i <(S CP +2S AP ) / 3, it means that the subject's i-th professional skill has reached negative proficiency, that is, habitual errors; count all skills that have reached positive proficiency and whose correct number is greater than 2×the number of errors.熟悉+ , then the final positive familiarity score of the subject is: Score + = Total 熟悉+ / N, the larger the value, the higher the familiarity of the professional skills in the field; count all skills that have reached negative familiarity and the number of errors > 2 × the number of correct answers Total 熟悉- , then the final negative familiarity score of the subject is: Score-=Total 熟悉- / N, the larger the value, the higher the habitual errors of professional skills in this field. CP 、S AP , TR i 、TW i Total 熟悉+ Total 熟悉- , Score+, Score- calculation formula is:

[0050] S CP =Pbeta_Scp_dB-Palpha_Scp_dB,S AP =Pbeta_Sap_dB-Palpha_Sap_dB

[0051] TR i =Pbeta_Tr_dB i -Palpha_Tr_dB i ,TW i =Pbeta_Tw_dB i -Palpha_Tw_dB i

[0052] Score + = Total 熟悉+ / N

[0053] Score-=Total 熟悉- / N.

[0054] The protection scope of the present invention includes but is not limited to the above embodiments, and its protection scope is subject to the claims. Any replacement, deformation, and improvement of this technology that can be easily thought of by those skilled in the art fall within the protection scope of the present invention.

Claims

1. A skill proficiency evaluation method based on β-α EEG index measurement is characterized by The specific steps include: (1) Have the person being tested wear a multi-channel EEG measurement device to record EEG signals; (2) Setting up a conscious control processing benchmark task, an automatic processing benchmark task, and a professional skills test task, wherein the professional skills test task includes two situations where the participant makes correct and incorrect judgments, collecting the EEG signals of the subjects in the three types of tasks, and obtaining the original EEG data of the subjects; (3) Preprocessing the raw EEG data, including signal amplification, segmentation, signal noise reduction, bandpass filtering and artifact removal; (4) Based on the preprocessed EEG data, the baseline normalized band power averages of α and β within the time window of 300-450ms were calculated for the subjects under different tasks, and the index value of β-α was calculated, that is, the baseline normalized band power average of β minus the baseline normalized band power average of α; (5) Compare the changes in the β-α index values ​​of the subjects under different tasks and calculate the average β-α index value S of the conscious control processing benchmark task. CP and the average β-α index value S of the automatic processing benchmark task AP There are N professional skills, each professional skill corresponds to a test task, and the average β-α index value TR when the judgment is correct in the test task of the i-th professional skill is calculated one by one i i∈N, if TR i <(S CP +2S AP ) / 3, it means that the subject's i-th professional skill has reached the positive proficiency level; calculate the average β-α index value TW when the judgment error occurs in the test task of the i-th professional skill one by one i i∈N, if TW i <(S CP +2S AP ) / 3, it means that the subject's i-th professional skill has reached negative proficiency, that is, habitual errors; count all skills that have reached positive proficiency and whose correct number is greater than 2×the number of errors. 熟悉+ , then the final positive familiarity score of the subject is: Score + = Total 熟悉+ / N, the larger the value, the higher the familiarity of the professional skills in the field; count all skills that have reached negative familiarity and the number of errors > 2 × the number of correct answers Total 熟悉- , then the final negative familiarity score of the subject is: Score-=Total 熟悉- / N, the larger the value, the higher the habitual errors in professional skills in this field.

2. The skill proficiency evaluation method based on β-α EEG index measurement according to claim 1, characterized in that: In the step (2), the conscious control processing benchmark task is a mental arithmetic task of square roots within 20, the automatic processing benchmark task is a mental arithmetic task of addition and subtraction within the number 3, and the professional skills test task includes two situations in which the participants judge correctly and incorrectly, which are the main skills knowledge and decision-making for jobs in different professional fields. Each professional skills test task includes four components: background description, question, correct answer, and interference answer. The test extracts the EEG data of the test person within 2000ms from the display of the question to the time when the question is displayed. The conscious control processing benchmark task and the automatic processing benchmark task each include more than 20 tasks, and each task is repeated 3 times.

3. The skill proficiency evaluation method based on β-α EEG index measurement according to claim 1, characterized in that: In the step (3), during the preprocessing of the raw EEG data, the raw EEG data is band-pass filtered at 0.5 to 40 Hz, and the FastICA algorithm based on the maximum negative entropy principle is used to perform independent component analysis and data reorganization to effectively remove artifact interference; and the α and β bands of the preprocessed EEG data are retained as target data for subsequent analysis.

4. The skill proficiency evaluation method based on β-α EEG index measurement according to claim 2, characterized in that: In the step (4), the baseline normalized band power index value is calculated as follows: (4-1) Perform wavelet transform on the pre-processed EEG data to obtain the band power values ​​of the α and β bands in the CPz channel in the interval 300 to 450 ms after the stimulus (question display) page appears; (4-2) Band power calculations were performed for the conscious control processing benchmark task, the automatic processing benchmark task, the correct judgment professional skill test task, and the incorrect judgment professional skill test task. The baseline was taken from -300ms to -100ms before the stimulus, and the baseline-normalized band power (dB) was obtained. The calculation relationship is as follows: Among them: the symbol Palpha represents the alpha band power, the symbol Pbeta represents the beta band power, the symbol Scp represents the conscious control processing benchmark task, the symbol Sap represents the automatic processing benchmark task, the symbol Tr represents the professional skill test task with correct judgment, and the symbol Tw represents the professional skill test task with incorrect judgment. Indicates the average of this type of task, that is: It represents the average of multiple repeated tests of the test task of the i-th professional skill, that is: k is the frequency value, which ranges from 9 to 30 Hz; the symbol dB represents the value obtained after baseline normalization processing, and the symbol base represents the average band power during the baseline period of -300ms to -100ms before stimulation. For example, Palpha_Scp represents the α-band power in the conscious control processing benchmark task, Palpha_Scp_dB represents Palpha_Scp after baseline normalization processing, and base_α_Scp represents the average band power of the α-band during the baseline period of -300ms to -100ms before stimulation in the conscious control processing benchmark task.

5. The skill proficiency evaluation method based on β-α EEG index measurement according to claim 4, characterized in that: In the step (5), the S calculated based on the β-α index value CP 、S AP , TR i 、TW i Total 熟悉+ Total 熟悉 The calculation formula of -, Score+, and Score- is: S CP =Pbeta_Scp_dB-Palpha_Scp_dB,S AP =Pbeta_Sap_dB-Palpha_Sap_dB TR i =Pbeta_Tr_dB i -Palpha_Tr_dB i ,TW i =Pbeta_Tw_dB i -Palpha_Tw_dB i

Images