阅读说明：本技术 一种基于幅值的自动脑电去伪迹方法 (Automatic electroencephalogram artifact removing method based on amplitude ) 是由 冯春雨 王怀瑞 黎彤亮 田书娟 马芹颖 王铭维 王彦永 方际宇 于 2020-11-10 设计创作，主要内容包括：本发明公开了一种基于幅值的自动脑电去伪迹方法,其特征在于,包括以下步骤：判断脑电数据长度；根据考虑的频域范围对原始脑电进行带通滤波,去除工频干扰并在一定程度上修正基线漂移；将滤波后的脑电数据分为分段数个等长度的小段,每段数据的时间长度不小于t_(min)；计算相对均值；计算相对标准差；计算相对均值阈值、相对标准差阈值、通道高阈值和通道低阈值；以上述划分的段为单位进行处理,判断小段上是否存在伪迹。该方法不需要额外通道和人工参与,计算量小,能够快速有效去除rsEEG中的伪迹,从而满足临床分析脑电的要求。(The invention discloses an automatic electroencephalogram artifact removing method based on amplitude, which is characterized by comprising the following steps of: judging the length of the electroencephalogram data; performing band-pass filtering on the original electroencephalogram according to the considered frequency domain range, removing power frequency interference and correcting baseline drift to a certain extent; dividing the filtered electroencephalogram data into a plurality of segments with equal length, wherein the time length of each segment of data is not less than t min (ii) a Calculating a relative mean value; calculating relative standard deviation; calculating a relative mean threshold, a relative standard deviation threshold, a channel high threshold and a channel low threshold; and processing the divided segments as a unit to judge whether the small segments have the artifacts. The method does not need additional channels and manual participation, has small calculated amount, and can quickly and effectively remove the artifacts in the rsEEG, thereby meeting the requirement of clinical analysis of the EEG.)

1. An automatic electroencephalogram artifact removing method based on amplitude is characterized by comprising the following steps:

step 1, judging the length of electroencephalogram data: for the electroencephalogram data of the channel row sample point sequence, if the sample point number/sampling rate is less than 60, namely the electroencephalogram acquisition time is less than 60 seconds, generating a data processing report, and quitting the processing;

step 2, performing band-pass filtering on the original electroencephalogram according to the considered frequency domain range, removing power frequency interference and correcting baseline drift to a certain extent;

step 3, dividing the filtered electroencephalogram data into a plurality of segments with equal lengthSmall segment of degree, the time length of each segment of data is not less than t_min；

Step 4, calculating a relative mean value;

step 5, calculating a relative standard deviation;

step 6, calculating a relative mean threshold, a relative standard deviation threshold, a channel high threshold and a channel low threshold;

step 7, processing the segments divided in the step 3 as units, and judging whether the segments have artifacts or not;

step 8, processing the data according to the bad lead condition;

and 9, repeating the steps 7 and 8 until the whole electroencephalogram data is processed, and generating a data report.

2. The amplitude-based automatic electroencephalogram artifact removing method according to claim 1, wherein the step 2 of performing band-pass filtering on the original electroencephalogram according to the considered frequency domain range, removing power frequency interference and correcting baseline drift to a certain extent specifically comprises the following steps:

determining the frequency domain maximum f from the considered frequency domain range_maxAnd frequency domain minimum f_minThe method specifically comprises the following steps: when resting state electroencephalogram rhythm information is concerned, the frequency band is generally selected from 1-40Hz, namely f_max＝40，f_min1 is ═ 1; performing band-pass filtering on the original electroencephalogram by using a digital filter (FIR or IIR); and if the frequency domain range contains 50/60Hz power frequency interference, removing the power frequency interference by using a notch filter, and filtering to correct baseline drift.

3. The amplitude-based automatic electroencephalogram artifact removing method according to claim 1, wherein the calculating relative mean value in the step 4 specifically is as follows:

4.1, calculating the mean value of each channel on each small section to obtain a mean value matrix of a channel number row sample point number column;

step 4.2, if all the values of the mean matrix are larger than the absolute mean M_a(ii) a Generating a data processing report and quitting the processing; otherwise, for each row of the mean matrix, from all less than M_aNumerical value ofCalculating the median, if all the values in a certain row are not less than M_aThen the median of the row is noted as M_aObtaining a mean vector of channel dimensions;

step 4.3, the obtained mean vector is selected from all vectors less than M_aThe median of the values of (A) is calculated and recorded as a relative mean value M_r。

4. The amplitude-based automated electroencephalogram artifact removal method of claim 3, wherein said absolute mean M_aIs 4.

5. The automatic amplitude-based electroencephalogram artifact removing method according to claim 3, wherein the calculating of the relative standard deviation in the step 5 specifically comprises:

step 5.1, calculating the standard deviation of each channel on each small section to obtain a standard deviation matrix of a channel row and segment sequence;

step 5.2, if all the values of the standard deviation matrix are larger than the absolute standard deviation threshold S_aIf yes, returning an error code 0x002, and exiting the processing; otherwise, for each row of the standard deviation matrix, from all less than S_aIf all the values in a certain row are not less than S_aThen the median of the row is marked as S_aObtaining a standard deviation vector of the nchans dimension;

step 5.3, for the obtained standard deviation vector, all are less than S_aThe median of the values of (A) is calculated and recorded as the relative standard deviation S_r。

6. The amplitude-based automatic electroencephalogram artifact removing method according to claim 5, wherein the absolute standard deviation threshold S is_aIs 25.

7. The amplitude-based automatic electroencephalogram artifact removing method according to claim 5, wherein the calculating of the relative mean threshold, the relative standard deviation threshold, the channel high threshold and the channel low threshold in the step 6 specifically comprises:

step 6.1, relative mean M_rThe product of the sum of the parameter P1 and the relative mean threshold Trm is used to determine whether the baseline wander is too large;

step 6.2, relative standard deviation S_rThe product of the sum and the parameter P2 is recorded as a relative standard deviation threshold value Trs, and the relative standard deviation threshold value Trs is used for judging whether the amplitude fluctuation of the channel is too small;

step 6.3, relative mean M_rAnd relative standard deviation S_rThe sum of the parameters P3 is recorded as the high channel threshold H _ Tc, relative mean M_rAnd relative standard deviation S_rThe difference between the parameters P3 is denoted as the low channel threshold L _ Tc, and these two thresholds are used to determine whether the amplitude of the sampling point is too large.

8. The automatic electroencephalogram artifact removing method based on the amplitude as claimed in claim 7, wherein the step 7 is to process the segments divided in the step 3 as a unit to judge whether the segments have artifacts, and specifically comprises the following steps:

step 7.1, if the average value of a certain channel is more than T_rmIf so, judging that the channel has too large drift and marking the channel as a bad conductor;

7.2, if the standard deviation of a certain channel is smaller than Trs, judging that the amplitude fluctuation of the channel is too small, and marking the channel as bad lead;

step 7.3, for any sampling point, if its amplitude and its following interval are N_pIs not in the interval [ L _ Tc H _ Tc ]]And if so, judging that the sample point has too large amplitude and marking as a pseudo trace point. If the proportion of the pseudo trace points to the whole data exceeds P5, the channel is marked as a bad leader.

9. The automatic electroencephalogram artifact removing method based on the amplitude according to claim 8, wherein the step 8 of processing data according to the bad lead condition specifically comprises the following steps:

step 8.1, if no bad guide exists, no treatment is carried out;

step 8.2, if the number of the bad leads is not more than P4; considering that the bad derivative is in an acceptable range, and performing interpolation bad derivative processing by using eeglab or other electroencephalogram processing software;

and 8.3, if the number of the bad derivatives is more than P4, the bad derivatives are considered to be beyond the acceptable range, and the whole data is deleted.

Technical Field

The invention belongs to the technical field of brain signal processing, and particularly relates to an automatic electroencephalogram artifact removing method based on amplitude.

Background

Because the electroencephalogram, especially the electroencephalogram collected at the scalp, is an extremely weak physiological signal, has a small amplitude (generally within 100 microvolts), is easily interfered by various noises, forms electroencephalogram and electroencephalogram artifacts, and affects the quality of the electroencephalogram signals. Therefore, preprocessing is needed before the electroencephalogram signals are analyzed, artifact components in the electroencephalogram signals are removed, and clean electroencephalogram signals are obtained.

Clinically, the electroencephalogram artifacts are generally removed by adopting a filtering and visual inspection mode, namely, a band-pass filter is used for filtering out the frequency components which are not concerned, and then a small section which is relatively smooth is found out from the whole electroencephalogram data by observation and is used as a representative for analysis. The most used methods for removing artifacts in academic research include regression methods, Independent Component Analysis (ICA) methods, and filtering methods based on wavelet transform. The regression method assumes that the ocular artifacts are the combination of the signals acquired by the ocular acquisition channels, calculates the ocular components through a regression model, and then deletes the ocular components from the acquired electroencephalogram signals. ICA is an electroencephalogram artifact removing method with higher acceptance in the current research field, and is a new signal analysis method developed based on a blind source separation technology. ICA aims at mutual independence and non-gaussian nature of random variables, and separates mutually independent source signals from a given multi-derivative mixed signal when neither the source signal nor the mixed system is known. In electroencephalogram signal analysis, ICA can effectively separate artifact signals such as electrooculogram and electrocardio from original signals. The basic idea of the artifact removing method based on wavelet transformation is to select a proper wavelet function family, perform wavelet decomposition on the signal with the artifact, decompose the signal to different sub-frequency bands, and obtain an approximate coefficient and a detail coefficient. Then setting a proper threshold value to process the wavelet coefficient, setting the high-frequency band and low-frequency band coefficients corresponding to the artifact to be zero, and reserving useful signal coefficients, thereby achieving the purpose of filtering the artifact.

The manual visual inspection method has high requirements on experience of readers, has strong subjectivity, consumes a large amount of time for completing one part of electroencephalogram processing, and is difficult to apply in reality. The regression method needs an extra electro-oculogram acquisition channel, and because of the bidirectional mixing of electroencephalogram and electro-oculogram, the regression method can lose part of real electroencephalogram signals while removing artifacts. Solving the unmixing matrix in the ICA artifact removing method is an optimization problem, so the computation complexity of the ICA is high. In the common ICA false removing process, the false mark component separated from the source signal needs to be visually identified, so that the method is not suitable for automatic processing of the electroencephalogram signal. Although some scholars have proposed some methods for automatically identifying artifact components, such methods still require pre-defining appropriate thresholds to identify artifact signals, depending on the experience of the operator. The wavelet-based artifact removal method also has the problems of selection of wavelet basis, threshold values and the like. In addition, the wavelet-based artifact removal method has poor effect on processing some high-amplitude artifacts with frequencies close to those of normal electroencephalogram.

Therefore, there is a need to provide a new automatic electroencephalogram artifact removal method.

Disclosure of Invention

In view of the above, the invention researches an automatic electroencephalogram artifact removing method based on amplitude for various artifacts which are present in resting state electroencephalogram (rsEEG) data and are embodied as amplitude abnormity. The method does not need additional channels and manual participation, has small calculated amount, and can quickly and effectively remove the artifacts in the rsEEG, thereby meeting the requirement of clinical analysis of the EEG.

In order to solve the technical problem, the invention discloses an automatic electroencephalogram artifact removing method based on amplitude, which comprises the following steps:

step 1, judging the length of electroencephalogram data: for the electroencephalogram data of the channel row sample point sequence, if the sample point number/sampling rate is less than 60, namely the electroencephalogram acquisition time is less than 60 seconds, generating a data processing report, and quitting the processing;

step 2, performing band-pass filtering on the original electroencephalogram according to the considered frequency domain range, removing power frequency interference and correcting baseline drift to a certain extent;

step 3, dividing the filtered electroencephalogram data into a plurality of subsection small sections with equal length, wherein the time length of each section of data is not less than t_min；

Step 4, calculating a relative mean value;

step 5, calculating a relative standard deviation;

step 6, calculating a relative mean threshold, a relative standard deviation threshold, a channel high threshold and a channel low threshold;

step 7, processing the segments divided in the step 3 as units, and judging whether the segments have artifacts or not;

step 8, processing the data according to the bad lead condition;

and 9, repeating the steps 7 and 8 until the whole electroencephalogram data is processed, and generating a data report.

Optionally, the step 2 of performing band-pass filtering on the original electroencephalogram according to the considered frequency domain range, removing power frequency interference, and correcting baseline drift to a certain extent specifically includes:

determining the frequency domain maximum f from the considered frequency domain range_maxAnd frequency domain minimum f_minThe method specifically comprises the following steps: when resting state electroencephalogram rhythm information is concerned, the frequency band is generally selected from 1-40Hz, namely f_max＝40，f_min1. The raw brain electrical is band-pass filtered using a digital filter (FIR or IIR). And if the frequency domain range contains 50/60Hz power frequency interference, removing the power frequency interference by using a notch filter, and filtering to correct baseline drift.

Optionally, the calculating the relative mean value in step 4 specifically includes:

4.1, calculating the mean value of each channel on each small section to obtain a mean value matrix of a channel number row sample point number column;

step 4.2, if all the values of the mean matrix are larger than the absolute mean M_a(ii) a Generating a data processing report and quitting the processing; otherwise, for each row of the mean matrix, from all less than M_aTo obtain a median value, e.g. ofAll values of a row of fruits are not less than M_aThen the median of the row is noted as M_aObtaining a mean vector of channel dimensions;

step 4.3, the obtained mean vector is selected from all vectors less than M_aThe median of the values of (A) is calculated and recorded as a relative mean value M_r。

Optionally, the absolute mean M_aIs 4.

Optionally, the calculating the relative standard deviation in step 5 specifically includes:

step 5.1, calculating the standard deviation of each channel on each small section to obtain a standard deviation matrix of a channel row and segment sequence;

step 5.2, if all the values of the standard deviation matrix are larger than the absolute standard deviation threshold S_aIf yes, returning an error code 0x002, and exiting the processing; otherwise, for each row of the standard deviation matrix, from all less than S_aIf all the values in a certain row are not less than S_aThen the median of the row is marked as S_aObtaining a standard deviation vector of the nchans dimension;

step 5.3, for the obtained standard deviation vector, all are less than S_aThe median of the values of (A) is calculated and recorded as the relative standard deviation S_r。

Optionally, the absolute standard deviation threshold S_aIs 25.

Optionally, the calculating of the relative mean threshold, the relative standard deviation threshold, the channel high threshold, and the channel low threshold in step 6 specifically includes:

step 6.1, relative mean M_rThe product of the sum of the parameter P1 and the relative mean threshold Trm is used to determine whether the baseline wander is too large;

step 6.2, relative standard deviation S_rThe product of the sum and the parameter P2 is recorded as a relative standard deviation threshold value Trs, and the relative standard deviation threshold value Trs is used for judging whether the amplitude fluctuation of the channel is too small;

step 6.3, relative mean M_rAnd relative standard deviation S_rThe sum of the parameters P3 is recorded as the high channel threshold H _ Tc, relative mean M_rAnd relative standard deviation S_rThe difference between the parameters P3 is denoted as the low channel threshold L _ Tc, and these two thresholds are used to determine whether the amplitude of the sampling point is too large.

Optionally, in the step 7, the segment divided in the step 3 is processed as a unit, and whether artifacts exist on the small segment is determined, specifically:

step 7.1, if the average value of a certain channel is more than T_rmIf so, judging that the channel has too large drift and marking the channel as a bad conductor;

7.2, if the standard deviation of a certain channel is smaller than Trs, judging that the amplitude fluctuation of the channel is too small, and marking the channel as bad lead;

step 7.3, for any sampling point, if its amplitude and its following interval are N_pIs not in the interval [ L _ Tc H _ Tc ]]And if so, judging that the sample point has too large amplitude and marking as a pseudo trace point. If the proportion of the pseudo trace points to the whole data exceeds P5, the channel is marked as a bad leader.

Optionally, the processing of the data according to the bad lead condition in the step 8 specifically includes:

and 8.1, if no bad conductor exists, not processing.

Step 8.2, if the number of the bad leads is not more than P4; considering that the bad derivative is in an acceptable range, and performing interpolation bad derivative processing by using eeglab or other electroencephalogram processing software;

and 8.3, if the number of the bad derivatives is more than P4, the bad derivatives are considered to be beyond the acceptable range, and the whole data is deleted.

Compared with the prior art, the invention can obtain the following technical effects:

the method of the present invention uses the main operations of averaging, standard deviation and median, with a time complexity of nlog (n). Therefore, the method has the advantages of small operand and high processing speed. In addition, the required threshold is generated through electroencephalogram signal self-adaptation, extra channels and manual participation are not needed, the requirement of acquisition equipment is lowered, the acquisition flow is simplified, and the automation level of electroencephalogram processing is improved.

Of course, it is not necessary for any one product in which the invention is practiced to achieve all of the above-described technical effects simultaneously.

Drawings

The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the invention and not to limit the invention. In the drawings:

FIG. 1 shows that the ratio of the artifact points of all the channels does not exceed P5, and no processing is performed;

FIG. 2 shows that the number of channels with the proportion of the pseudo trace points with overlarge amplitude exceeding P5 is greater than P4, and the whole data is deleted;

FIG. 3 shows that the number of channels with the proportion of the pseudo trace points with too small amplitude exceeding P5 is larger than P4, and the whole data is deleted;

FIG. 4 shows that the number of channels containing pseudo trace points with a proportion exceeding P5 is not more than P4, and interpolation is performed to obtain bad leads;

FIG. 5 is data after interpolation bad lead processing according to the present invention;

FIG. 6 is a flow chart of the brain electrical artifact removal of the present invention.

Detailed Description

The following embodiments are described in detail with reference to the accompanying drawings, so that how to implement the technical features of the present invention to solve the technical problems and achieve the technical effects can be fully understood and implemented.

The method is mainly used for removing the artifacts of the resting state electroencephalogram, compared with the task state electroencephalogram, the eye electrical components in the resting state electroencephalogram data are relatively few, the artifacts mainly comprise power frequency interference and noise introduced by the surrounding environment, and the artifacts are more expressed as amplitude which is more normal and has obvious difference (too high or too low), so the artifacts can be effectively removed through time domain analysis. Based on this situation, we propose an automatic artifact removal method for offline electroencephalogram data processing. Because the amplitude of the electroencephalogram is influenced by a plurality of factors such as the collected person, the electroencephalogram collecting equipment, the collection standard degree and the like, the amplitude of different electroencephalogram data is different, and therefore the same threshold value cannot be adopted for judgment and processing. The method automatically generates a series of thresholds according to the amplitude information of the electroencephalogram, the thresholds are used for judging whether artifact components exist in the electroencephalogram, and different processing modes are adopted according to the influence range and degree of the artifacts.

The symbolic meanings and values used in the algorithm are shown in table 1:

symbolic meanings and values used in the algorithm of Table 1

Note: note: the three parameters of nchans, npnts and srate in the table are determined by specific electroencephalogram data, and P1, P2, P3 and P5 are empirical values. In order to ensure the quality of the interpolation bad derivatives, 0.3 selected in the process of calculating P4 is used as a multiplication factor. Some adjustments to these parameters may be made if it is desired to change the degree of constraint. floor is a rounded down function and max is a rounded up function.

The invention discloses an automatic electroencephalogram artifact removing method based on amplitude, which comprises the following steps of:

step 1, judging the length of the electroencephalogram data. And for the electroencephalogram data of the nchans row and the npnts column, if the npnts/srate is less than 60, namely the electroencephalogram acquisition time is less than 60 seconds, generating a data processing report, and quitting the processing.

Step 2, determining the frequency domain maximum value f according to the considered frequency domain range_maxAnd frequency domain minimum f_minThe method specifically comprises the following steps: when resting state electroencephalogram rhythm information is concerned, the frequency band is generally selected from 1-40Hz, namely f_max＝40，f_min1. The raw brain electrical is band-pass filtered using a digital filter (FIR or IIR). If the frequency domain range contains 50/60Hz power frequency interference, the notch filter is used for removing the power frequency interference, and the filtering can correct baseline drift.

Step 3, dividing the filtered electroencephalogram data into epochs small segments with equal length, wherein the time length of each segment of data is not less than t_min。

Step 4, calculating a relative mean value:

4.1, calculating the mean value of each channel on each segment to obtain a mean value matrix of the ephchs columns of the nchans rows;

step 4.2, if all the values of the mean matrix are larger than the absolute mean M_aThen, the error code 0x001 is returned and the process is exited. Otherwise, for each row of the mean matrix, from all less than M_aIf all the values in a certain row are not less than M_aThen the median of the row is noted as M_aObtaining a mean vector of the nchans dimensions;

step 4.3, the obtained mean vector is selected from all vectors less than M_aThe median of the values of (A) is calculated and recorded as a relative mean value M_r。

Step 5, calculating the relative standard deviation:

step 5.1, calculating the standard deviation of each channel on each segment to obtain a standard deviation matrix of the ephchs columns of the nchans rows;

step 5.2, if all the values of the standard deviation matrix are larger than the absolute standard deviation threshold S_aAnd generating a data processing report and quitting the processing. Otherwise, for each row of the standard deviation matrix, from all less than S_aIf all the values in a certain row are not less than S_aThen the median of the row is marked as S_aObtaining a standard deviation vector of the nchans dimension;

step 5.3, for the obtained standard deviation vector, all are less than S_aThe median of the values of (A) is calculated and recorded as the relative standard deviation S_r。

And 6, calculating a relative mean threshold, a relative standard deviation threshold, a channel high threshold and a channel low threshold.

Step 6.1, relative mean M_rThe product of the sum of the parameter P1 and the relative mean threshold Trm is used to determine whether the baseline wander is too large;

step 6.2, relative standard deviation S_rThe product of the sum and the parameter P2 is recorded as a relative standard deviation threshold value Trs, and the relative standard deviation threshold value Trs is used for judging whether the amplitude fluctuation of the channel is too small;

step 6.3, relative mean M_rAnd relative standard deviation S_rThe sum of the parameters P3 is recorded as the high channel threshold H _ Tc, M_rAnd S_rThe difference between P3 is recorded as the low channel threshold L _ Tc, and the two thresholds, i.e., the high channel threshold H _ Tc and the low channel threshold L _ Tc, are used to determine whether the amplitude of the sampling point is too large.

And 7, processing the segments divided in the step 3 as a unit, and judging whether the segments have artifacts:

step 7.1, if the average value of a certain channel is more than T_rmIf so, judging that the channel has too large drift and marking the channel as a bad conductor;

7.2, if the standard deviation of a certain channel is smaller than Trs, judging that the amplitude fluctuation of the channel is too small, and marking the channel as bad lead;

step 7.3, for any sampling point, if its amplitude and its following interval are N_pIs not in the interval [ L _ Tc H _ Tc ]]And if so, judging that the sample point has too large amplitude and marking as a pseudo trace point. If the proportion of the pseudo trace points in the whole data exceeds P5, marking the channel as a bad guide;

and 8, processing the data according to the bad lead condition:

step 8.1, if no bad leadings exist, no processing is carried out, as shown in fig. 1, the electroencephalogram screenshot is data with more stable amplitude fluctuation, and according to judgment of the method, the artifact point proportion of all channels does not exceed P5 (in the example, the P5 takes a value of 0.2), and the data of the section does not have bad leadings and does not need to be processed;

step 8.2, if the number of the bad leads is not more than P4; considering that the bad derivative is in an acceptable range, and performing interpolation bad derivative processing by using eeglab or other electroencephalogram processing software;

as shown in fig. 4, the electroencephalogram screenshot indicates that artifact data exists in a part of channels, and according to the judgment of the invention, the proportion of artifact points of 4 channels, namely C3, P3, O1 and O2, exceeds P5, the number of the channels is not greater than P4, and according to step 8.2 in the specific embodiment, the number of bad leads of the section of data is within an acceptable range, and interpolation bad lead processing is adopted.

And 8.3, if the number of the bad derivatives is more than P4, the bad derivatives are considered to be beyond the acceptable range, and the whole data is deleted.

As shown in fig. 2, the electroencephalogram screenshot is data with large amplitude fluctuation, and according to the method of the present invention, it is determined that the number of channels with artifact point ratio exceeding P5 is 16, and the number of channels is greater than P4 (in this example, the value of P4 is 4), and according to step 8.3 in the specific implementation, the bad derivative of the segment of data exceeds the acceptable range, and the entire segment of data is deleted.

As shown in fig. 3, the electroencephalogram screenshot is data with small amplitude fluctuation, and according to the method of the present invention, the number of channels with the artifact point ratio exceeding P5 is determined to be 16, and the number of channels is greater than P4, and according to step 8.3 in the specific implementation, the bad derivative of the data exceeds the acceptable range, and the whole data is deleted.

As shown in FIG. 5, the data obtained by interpolation and bad-lead processing of the electroencephalogram data shown in FIG. 4 shows that the amplitudes of 4 channels such as C3, P3, O1, O2 and the like are relatively stable after being processed

And 9, repeating the steps 7 and 8 until the whole electroencephalogram data is processed, and generating a data report.

While the foregoing description shows and describes several preferred embodiments of the invention, it is to be understood, as noted above, that the invention is not limited to the forms disclosed herein, but is not to be construed as excluding other embodiments and is capable of use in various other combinations, modifications, and environments and is capable of changes within the scope of the inventive concept as expressed herein, commensurate with the above teachings, or the skill or knowledge of the relevant art. And that modifications and variations may be effected by those skilled in the art without departing from the spirit and scope of the invention as defined by the appended claims.