KR100751257B1 - Brain waves - Google Patents

Abstract

A method for extracting an index related to a high-class recognition function based on brain waves and a method for testing the high-class recognition function are provided to improve user convenience by enabling a user to easily measure the brain wave by simply contacting a band or headset type electrode on a user forehead. A classification task is provided to a user, who wears a brain wave measuring device, such that the brain wave of the user is measured. A gamma wave is extracted by applying a BPF(Band Pass Filter) on the measured brain wave. An absolute value of the extracted gamma wave is obtained and a smoothing process is performed to generate a gamma wave magnitude graph. The gamma magnitude graphs are averaged such that a simulation-induced gamma response graph is obtained. The simulation-induced gamma response graph is differentiated and a lowermost point of the graph is set to zero. Then, uppermost and lowermost peak values are obtained from the differentiated simulation-induced gamma response graph. Biggest positive/negative peaks, which occur in a positive or negative order between 70 and 400 ms after an application of the stimulation, are extracted and a difference between the positive and negative peaks is set as an index of recognition intensity. A cross point between a lowermost point and the positive and negative peaks are set as an index of a recognition speed.

Description

Indicator extraction method related to EEG-based high cognitive function and high cognitive function test method {Brain waves}

The present invention relates to a method for extracting indicators related to high-order cognitive function (cognitive ability, concentration, workload, left / right brain balance) from the brain waves of a subject.

In general, higher-order cognitive function is the brain's cognitive process, which includes perception, attention, memory, reasoning, and determination. Significantly developed higher brain functions that are in charge of the cerebral cortex.

The higher the higher cognitive function, the more efficient the brain's learning process, which accepts information from the outside and processes it appropriately. In this case, the neural efficiency is sometimes expressed as a high state.

Higher cognitive function in humans usually deteriorates naturally as aging progresses, but is rapidly deteriorated by various causes such as dementia, stroke, trauma to the brain, and chronic fatigue.

The lowering of high cognitive function impairs perceptual ability, attention, memory, or judgment, so that normal learning efficiency is lowered and the quality of life is drastically reduced in daily life. In addition, the high-cognitive function that has been degraded once is rarely recovered to its original state, so it is important to identify it early and prevent any further degradation.

Therefore, in order to detect such degradation of high-order cognition early, more accurate and objective test tools are urgently needed. However, the existing high-order cognitive function test methods are mostly test papers or psychoneurological tests by computerized problem solving, and test subjects need to solve many problems. Even if you do not know the answer, you can take a lot of subjective effects, and if the problem is exposed to the subject, the test result is meaningless, there was a disadvantage that it is impossible to repeat the test afterwards.

In order to overcome various drawbacks of the existing psychological neurological examination methods, brain function measuring devices such as fMRI (functional magnetic resonance device), PET (positron emission imaging device), MEG (brain magnetic measuring device) and EEG (brain potential measuring device) Various objective neurophysiological tests have been tried. However, fMRI, PET, and MEG are hundreds of millions to billions of expensive and fixed equipment in fixed spaces, so they can usually be used only as a built-in type in large hospitals. On the other hand, EEG equipment, also known as electroencephalography, has the advantage that it is not only cheap but can be miniaturized to a size that is easy to carry.

EEG analysis methods are largely divided into spontaneous EEG and Evoked or Event-Related Potential. If spontaneous EEG is a signal reflecting a combination of the electrical activity of the cranial nerve cell group by various factors at the moment, the induced EEG can be seen as an electrical activity due to a given specific stimulus only. In general, induced EEG can be obtained only by repeatedly presenting a specific stimulus. However, deviations due to the complex factors experienced by spontaneous EEG can be avoided, which has the advantage of high reproducibility when repeated.

Representative methods of cognitive abilities based on induced EEG have been mainly performed by event-related potential analysis (ERP). ERP can be extracted from brain waves repeatedly presented to the subject according to task rules in which stimuli corresponding to sight or hearing are defined. That is, the brain waves are extracted only by the common stimulus by averaging the brain waves based on the time points at which each stimulus is presented. The extracted ERP is composed of several large negative / bidirectional peaks, and peaks appearing between 100 and 900 ms after stimulus presentation have been mainly used for cognitive evaluation. Usually, the larger the amplitude of these peaks, the higher the recognition intensity, and the faster the time of appearance of the peak corresponds to a faster recognition rate. Therefore, the cognitive ability determined by cognitive strength and cognitive speed is evaluated to be superior as the cognitive strength is higher and the cognitive speed is faster.

However, the existing ERP method has some problems as follows, and it is still difficult to be introduced for popular test.

First, since ERP shows peaks most clearly near the head near Fz, Cz, and Pz, it is inconvenient to attach the electrode after showing the skin through the hair to measure brain waves in this area. have.

Second, since the ERP extraction process averages the EEG as it is, it is sensitively affected by the mixed wave by only one blink or movement. Therefore, the frontal lobe (frontal forehead), which is frequently affected by the blinking of the subject, is often excluded, and there is an inconvenience in that the subject should be strictly controlled so that no wave is mixed during the measurement.

Third, since ERP is a signal included in the measured EEG with very small amplitude, it is possible to offset parts due to other factors only by averaging by at least several dozen repeated stimuli, so that the response due to a given stimulus is well extracted. Can be. Therefore, the stimulus under the same conditions should be presented more than a few dozen times, so the test time is long and uncomfortable.

Therefore, the present invention has been devised to solve the above problems, and an object of the present invention is to newly present a method for extracting cognitive indicators, and to further increase higher concentrations such as concentration, workload, and left and right brain activity balance. In order to extract cognitive function-related indicators, we provide an index extraction method related to high-order cognitive function based on EEG for a more comprehensive and efficient high-order cognitive function test.

An object of the present invention as described above comprises the steps of measuring the EEG by presenting a classification task to distinguish a plurality of stimulation to a subject equipped with an EEG; Extracting a gamma waveform by applying a gamma band band filter to the measured brain waves, taking an absolute value of the extracted gamma waveform, and performing a smoothing operation to convert the gamma wave amplitude graph into a gamma wave amplitude graph; Averaging the graph on the basis of the time points at which each stimulus is presented in the gamma wave amplitude graph and converting the graph into a stimulus-induced gamma response graph in response to the presented stimulus; Differentiate the converted stimulus-induced gamma response graph, and adjust the baseline of the graph to 0, to differentiate the amplitude of the stimulus-induced gamma response to the largest positive value and the lowest negative value. Converting the stimulus-induced gamma response graph; In the differential stimulus-induced gamma response graph, after extracting the largest continuous positive / negative peak in positive and negative order between 70 and 400 ms after the stimulus presentation, the height difference of the positive / negative peak is used as an index of cognitive intensity. And using a point crossing the baseline 0 at the positive / negative peak as an index of cognitive velocity; and an index extraction method related to EEG-based high differential cognitive function, comprising:

Measuring brain waves by presenting a classification task to distinguish a plurality of stimuli to a subject equipped with an electroencephalogram; The measured brain waves are subjected to power spectrum by FFT (Fast Fourier Transform) to extract gamma wave power values, and the gamma wave power value of the left brain is defined as the left brain activity, and the gamma wave power value of the right brain is defined as the right brain activity. Calculating an index of left and right brain activity balance by using Equation>; and an index extraction method related to a high degree of cognitive function based on brain waves, comprising:

Equation

Measuring brain waves by presenting a classification task to distinguish a plurality of stimuli to a subject equipped with an electroencephalogram; The measured brain waves are subjected to power spectrum by FFT (Fast Fourier Transform) to extract power values of theta (Thaeta) wave, power values of the slow beta (SMR) wave, and power values of the mid beta (Mid-Beta) wave. Calculating a concentration index (BTR) index by <Equation> of the index extraction method associated with the higher order cognitive function based on EEG, comprising the;

Equation

Measuring brain waves by presenting a classification task to distinguish a plurality of stimuli to a subject equipped with an electroencephalogram; In order to check whether the measured power wave is a power spectrum with a fast fourier transform (FFT) to determine whether the power spectrum distribution is a relatively high frequency or low frequency biased distribution, the highest frequency at the lowest frequency value in the specified frequency region of the power spectrum. Converting the cumulative power spectrum graph by adding values sequentially up to a value; And scaling a final cumulative value of the cumulative power spectrum to 100% and using a frequency value corresponding to any value in a range of 0 to 100% as an indicator of workload. It is achieved by the index extraction method related to the high order cognitive function.

The index extraction method related to EEG-based high cognitive function according to the present invention uses the frontal lobe, which is easy to attach the electrode, and uses the gamma subband to minimize the influence of the miscellaneous waves that are easily incorporated into the EEG, and at the same time performs a specific task with high reproducibility By using the active EEG, the general public can easily and quickly carry out the neurophysiological-based accurate high-order cognitive tests repeatedly, it is easy to track and manage whether the high-order cognitive decline by regular tests.

In addition, since the present invention is a technique for objectively quantifying the response information related to the high-order cognitive activity of the subject with respect to a specific presented stimulus, the subject's high degree of the stimulus related to the crime and the stimulus not related to the crime at the time of scientific investigation It can be widely used in cognitive response test.

In addition, since the present invention can extract and display the high cognitive ability scores for each presented individual stimulus in real time, the user will make a voluntary effort to increase this value by monitoring his high cognitive ability score displayed on the screen in real time. It can also be used for EEG-based biofeedback training for higher cognitive functioning.

Hereinafter, with reference to the accompanying drawings showing a preferred embodiment of the present invention will be described in detail, in adding reference numerals to the components of each drawing, the same components are possible even if displayed on different drawings It should be noted that the same reference numerals are used. In describing the present invention, when it is determined that the detailed description of the related known functions or configurations may unnecessarily obscure the subject matter of the present invention, the detailed description thereof will be omitted.

EEG is composed of various rhythms, delta wave (0-3.99Hz), theta wave (4-7.99Hz), alpha wave (8-12.99Hz), beta wave (13-29.99) Hz) and gamma wave (30-50 Hz). In the past, mainly delta, theta, alpha, and beta waves were used, but since the late 1990s, the possibility of gamma wave band rhythm and cognitive activity has been mentioned through various experimental results. Although it has been rapidly attracting attention, the development of high cognitive function test method that can be evaluated clinically accurately has not progressed yet.

First, the present invention introduces differential pattern indicators of event-related induced gamma response, which is a newly devised gamma wave analysis method, and solves all the problems of the aforementioned cognitive evaluation methods as follows. In this paper, we propose an index extraction method and a high cognitive test method.

First, the event-related induced gamma response is more convenient than the conventional method by measuring EEG by simply contacting with a band or headset type electrode by using the advantage that the prefrontal lobe is most predominant.

Second, since EEG uses only the high frequency signals included in the gamma band, it is almost unaffected by crosstalk caused by eye movement or body movement of the subject, which contaminates the slow frequency band. Therefore, accurate test results can be obtained without subject control. Will be.

Third, gamma wave is one of the main components of EEG during cognitive activity, and it is included in a relatively large proportion, so it is possible to clearly detect event-related induction gamma response only by averaging the small amount of stimulus several times. The time can be drastically reduced.

To this end, the present invention proposes a relatively simple type of discrimination task for distinguishing a plurality of stimuli from a subject. These divisional tasks should be designed to accompany the higher-order cognitive processes of the brain, which are basically perception, attention, memory, reasoning, and determination.

The classification task can be presented in several sensory modes such as visual and auditory, which are usually required to respond to the subject's response. It may be formal or both stimuli may be presented at the same time. In this case, if it is determined that the two stimuli match each other according to a predetermined criterion, the reaction key is pressed as long as it is designated.

In addition, the form of the stimulus presented to the subject may be a simple pattern such as a figure or a picture, a stimulus with a meaning such as a number or a word, or a stimulus such as a pure sound or a sound, or a word character (visual). And word sounds (hearing) may be mixed to suggest whether two stimuli coincide. A relatively simple and low difficulty classification task is known as the oddball task, where one stimulus (standard stimulus) is presented at a common frequency, the other stimulus (target stimulus) is presented at a rare frequency, and the subject responds only to rare stimuli. This can be done by pressing or by counting the incidence of rare stimuli.

As shown in FIG. 1, the brain waves measured by the subjects were presented to the subject, and the detailed process of analyzing the stimulus-induced gamma response from the brain waves of the subjects measured when performing the subject tasks was as follows. And cognitive velocity) can be extracted.

Cognitive ability (cognitive strength and speed) index extraction method

Gamma wave filtering process,

FIG. 2 is a diagram illustrating a gamma waveform obtained through gamma wave filtering from the EEG of FIG. 1. The gamma band band filter is applied to an EEG of a subject. In this case, the gamma band used may use 30-50 Hz, which is the entire gamma band, or may select and use only the subbands within the gamma band. Only the 34-40 Hz band, which is least affected by EMG, may be used, and the subject may use an area within 3 Hz before and after the peak gamma frequency most active when performing the task. The overall test results are similar for any of the above.

Gamma amplitude graph extraction process,

3 is a diagram illustrating a graph obtained after taking an absolute value of the gamma waveform of FIG. 2, and FIG. 4 is a diagram of a gamma wave amplitude graph formed by performing a smoothing operation on the graph of FIG. 3. The absolute value (Absolute) is taken to the gamma waveform extracted by the above process, and then smoothed to extract only the amplitude graph of the gamma wave. The smoothing process may be a moving window averaging technique, a low pass filter, or other processing techniques for extracting only amplitude lines.

Stimulus-induced gamma response graph extraction process,

FIG. 5 is a graph of an event-related induction gamma amplitude graph that responds only to a presentation stimulus by simply averaging the graph based on a time point at which each stimulus is presented in the gamma amplitude graph. Averaging can also be performed. If the task stimulus is presented only once, the stimulus-induced gamma response is detected to some extent, but the analysis is possible even if this stimulus averaging process is omitted.

On the other hand, if EEG is measured at various sites, the graphs of the spatially averaged stimulus-induced gamma response are extracted by averaging all event-related induced gamma amplitude graphs for each site. In particular, the spatial averaging of the frontal and right frontal lobes is preferred when evaluating high-order cognitive function. If the measurement site is one site, this process can be omitted, such a graph is shown in FIG.

Extraction process of cognitive strength and cognitive velocity index by differentiation

7 is differentiating the extracted stimulus-induced gamma response graph (FIG. 5, or FIG. 6), while adjusting the overall baseline of the graph to 0, while increasing the induced gamma amplitude to the greatest increase and the greatest decrease, respectively. Differentiated stimulus-induced gamma response graphs can be easily identified from peak and negative peak values.

In the differential stimulus-induced gamma response graph, find the largest continuous positive / negative peak in the order of 70-400 ms after the event stimulus presentation, and then extract the height difference between the two peaks corresponding to the response intensity. It is used as an index of cognitive strength. The large cognitive strength indicators indicate that the cognitive processes corresponding to the perception, attention, memory, reasoning, and judgment processes performed efficiently in the brain. On the other hand, the response time that crosses the value of zero between positive and negative peaks is used as the point of cognitive activity, which gives information about cognitive velocity. That is, the faster the recognition time appears, the faster the recognition speed.

Concentration Index Extraction Method

In general, the concentration of theta, which reflects the level of consciousness, decreases in the concentration state for performing the task, while the slow beta (SMR) rhythm (12-15 Hz), which means unfocused attention, Mid-Beta rhythm (16-20Hz), which means focused attention, is immersed in one subject. Therefore, the total concentration index can be quantified by the Beta-Theta Ratio (BTR) index, which is a ratio of the power sum of the slow beta and the medium beta wave to theta wave as shown in Equation 1 below.

In order for this index to have a normal distribution feature that can be statistically interpreted, it is desirable to log the BTR or use the square root value as the concentration index.

Workload Index Extraction Method

FIG. 8 is a diagram illustrating a power spectrum graph obtained by power spectrumizing brain waves measured from a subject through a Fourier transform, and a cumulative spectrum graph adjusted to a maximum value of the power spectrum graph to 100. FIG. .

Workload is the degree of mental stress that a subject feels when performing a task, and means the level of mental stress at work or the level of brain arousal induced by the work. When the workload is high, the slow rhythms (low frequency components) of the brain waves reflecting sleep or consciousness decrease, while the relatively fast rhythms (high frequency components), which mean attention, awakening or complex cognitive activity, increase.

Therefore, the workload index can be quantified using SEF-A% (Spectral Edge Frequency) index, MEF (Median Frequency), and MF (Mean Frequency) index, which quantify whether the EEG spectrum distribution is relatively biased toward the high frequency. Can be. These indicators are known as general mathematical indicators that sensitively reflect whether the value distribution is biased toward a relatively higher or lower frequency in the specified frequency region of the spectrum. Meanwhile, the indicators known as MEF and MF are included in the SEF-A% index as SEF-50% and SEF-Mean%, respectively. Therefore, only the method of calculating the SEF index will be described.

A cumulative spectrum graph can be obtained by first adding the values from the lowest frequency value to the highest frequency value in the designated frequency region of the spectrum. In this graph, the maximum value means the total cumulative value, and if the scale is adjusted to correspond to the value of 100% for convenient calculation, the graph form is as follows. The "SEF-A%" index is defined as the frequency value at the point where the value corresponds to A% in this cumulative spectrum. For example, in the case of SEF-50%, it is determined by the frequency value when the value is 50% in the cumulative spectrum, which is called the center frequency or median frequency (MEF) because it means the center value in the spectral distribution. Also called. The SEF-A% index can be extracted using the same procedure as long as the time series data is transformed into frequency space, such as amplitude spectrum, amplitude density spectrum, power density spectrum, and wavelet spectra, as well as the power spectrum of EEG. It has a similar meaning.

Left / Right Brain Activity Index Extraction Method

The higher the cognitive activity, such as the classification task, the higher the gamma wave, so the activity of left and right brain is reflected in the gamma wave power of EEG measured from the left and right brain, respectively. In this case, it is preferable to use the logarithmic or square root values of gamma wave power as left and right brain activity indicators so that the gamma wave power value has a normal distribution feature that can be statistically interpreted.

At this time, the left and right brain activity balance indicator can be quantified by the relative ratio of the right brain activity indicator to the left brain activity indicator as shown in the following equation.

1 is a view showing an example of an EEG measured from a subject,

FIG. 2 is a diagram illustrating a gamma waveform obtained through gamma wave filtering from the brain waves of FIG. 1.

FIG. 3 is a diagram illustrating a graph obtained after taking an absolute value of the gamma waveform of FIG. 2.

FIG. 4 is a diagram illustrating a gamma wave amplitude graph obtained by performing a smoothing operation on the graph of FIG. 3.

FIG. 5 is a graph illustrating a stimulus-induced gamma response graph that simply averages a graph based on a time point at which each stimulus is presented in the gamma wave amplitude graph of FIG.

FIG. 6 is a graph showing an average of all the stimulus-induced gamma response graphs for EEG measured at various parts of FIG. 4.

7 is a graph showing a graph obtained by differentiating the stimulus-induced gamma response graph of FIG.

FIG. 8 is a diagram illustrating a power spectrum graph obtained by power spectrumizing brain waves measured from a subject through a Fourier transform, and a cumulative spectrum graph obtained by adjusting a maximum value of the power spectrum graph to 100. FIG.

Claims (7)

Measuring brain waves by presenting a classification task to distinguish a plurality of stimuli to an examinee equipped with an electroencephalogram (S10); Extracting a gamma waveform by applying a gamma band band filter to the measured brain waves, taking an absolute value of the extracted gamma waveform, and performing a smoothing operation to convert the gamma waveform into a gamma wave amplitude graph (S20); Averaging the graph on the basis of the time points at which each stimulus is presented in the gamma wave amplitude graph and converting the graph into a stimulus-induced gamma response graph in response to the presented stimulus; Differentiate the converted stimulus-induced gamma response graph, and adjust the baseline of the graph to 0, to differentiate the amplitude of the stimulus-induced gamma response to the largest positive value and the lowest negative value. Converting into a stimulus-induced gamma response graph (S40); From the differential stimulus-induced gamma response graph, after extracting the largest continuous positive / negative peak in positive and negative order between 70 and 400 ms after the stimulus presentation, the height difference of the positive / negative peak is designated as an index of cognitive intensity. (S50) designating a point that intersects the baseline 0 in the positive / negative peak as an index of recognition speed; Indicator extraction method and high cognitive function test method associated with EEG-based high cognitive function, characterized in that comprises a. Measuring brain waves by presenting a classification task to distinguish a plurality of stimuli to an examinee equipped with an electroencephalogram (S10); The measured brain waves are subjected to power spectrum by FFT (Fast Fourier Transform) to extract gamma wave power values, and the gamma wave power value of the left brain is defined as the left brain activity, and the gamma wave power value of the right brain is defined as the right brain activity. Calculating left and right brain activity balance indicators by using Equation &gt; Indicator extraction method and high cognitive function test method associated with EEG-based high cognitive function, characterized in that comprises a. Equation

Measuring brain waves by presenting a classification task to distinguish a plurality of stimuli to an examinee equipped with an electroencephalogram (S10); Power spectrum of the measured brain waves by FFT (Fast Fourier Transform) to extract the power value of the Thaeta wave, the power value of the slow beta (SMR) wave, and the power value of the mid-beta wave, and then Calculating a concentration (BTR) index by using Equation (S70); Indicator extraction method and high cognitive function test method associated with EEG-based high cognitive function, characterized in that comprises a. Equation

Measuring brain waves by presenting a classification task to distinguish a plurality of stimuli to an examinee equipped with an electroencephalogram (S10); In order to check whether the measured power wave is a power spectrum with a fast fourier transform (FFT) to determine whether the power spectrum distribution is a relatively high frequency or low frequency biased distribution, the highest frequency at the lowest frequency value in the specified frequency region of the power spectrum. Step S80 of sequentially adding the values up to the values and converting the values into cumulative power spectrum graphs; Scaling a final cumulative value of the cumulative power spectrum to 100% and using a frequency value corresponding to an arbitrary value in a range of 0 to 100% as an indicator of workload; Indicator extraction method and high cognitive function test method associated with EEG-based high cognitive function, characterized in that comprises a. The method according to any one of claims 1 to 4, EEG measurement of the subject is the frontal lobe, or the index extraction method associated with the high-order cognitive function and high-cognitive function test method characterized in that the frontal lobe region. The method according to claim 1 or 2, The gamma band is any one of gamma bands selected from within 3 Hz, 30 to 50 Hz, 34 to 40 Hz, or 36 to 41 Hz before and after the frequency after the subject extracts the most active gamma frequency. Index extraction method and high cognitive function test method related to high cognitive function based on brain waves. The method of claim 4, wherein The power spectrum is an index extraction method and a high differential cognitive function test method related to the brain wave-based high-order cognitive function, characterized in that consisting of more than 4 Hz measured brain waves with less influence of distortion due to cross-border (blink and body movement).

Images