Brainwave Analysis Device

The wearable EEG device with a single-chip processor and advanced filtering and analysis techniques addresses the inconsistency in S/N ratios, enabling accurate classification of higher brain functions by normalizing signal quality and noise removal.

JP2026181979APending Publication Date: 2026-11-19BIO SEARCH CO LTD +1
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
JP2025077842
Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
Filing Date
2025-05-08
Publication Date
2026-11-19

AI Technical Summary

Technical Problem

Existing electroencephalogram (EEG) analysis devices fail to accurately analyze and classify higher brain functions due to varying signal-to-noise ratios (S/N ratios) among different EEG devices, especially in non-shielded environments, leading to inconsistent analytical accuracy.

Method used

A wearable EEG device with a single-chip processor that includes frequency and voltage filters, an AD conversion unit, and an analysis unit that estimates S/N ratios through nonlinear analysis, compares reference and load spectra, and uses artificial intelligence to remove noise, allowing for accurate classification of higher brain functions regardless of measurement environment.

Benefits of technology

The device enables precise analysis and classification of higher brain functions by normalizing S/N ratios and filtering noise, ensuring consistent results across varying environments and EEG devices.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 2026181979000001_ABST
    Figure 2026181979000001_ABST
Patent Text Reader

Abstract

This invention provides an electroencephalogram (EEG) analyzer capable of accurately analyzing and classifying the characteristics of a subject's higher-order brain functions from their electroencephalogram, regardless of the measurement environment. [Solution] An electroencephalogram (EEG) analysis device capable of analyzing the characteristics of a subject's higher brain functions from the subject's electroencephalogram (EEG) consists of an EEG 10 and an analysis unit 20. The EEG 10 has two head electrodes 11, a reference electrode 12, and a processor 13 that outputs EEG signals, and detects EEGs from the frontal lobe, including at least the prefrontal cortex to the primary motor cortex. The analysis unit 20 estimates the signal-to-noise ratio of the EEG 10 by performing a nonlinear analysis of the subject's resting EEG signal, and calculates the difference between the reference spectrum obtained from the subject's resting EEG signal and the load spectrum obtained from the subject's EEG signal under a specific workload. The analysis unit 20 compares this difference with a quantity related to the EEG 10's signal-to-noise ratio as a threshold, and analyzes the characteristics of the subject's higher brain functions based on the frequency band in which activation occurs.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to an electroencephalogram analysis device, and in particular, to an electroencephalogram analysis device capable of analyzing characteristics of a subject's higher brain functions from the subject's electroencephalogram.

Background Art

[0002] Various studies have been conventionally conducted to link electroencephalograms with physiological mechanisms, psychological activities, and the like. For example, in Non-Patent Document 1, frequency spectrum and power values in the range of 1 Hz to 60 Hz are calculated by performing frequency analysis on rhythmic electroencephalograms, and studies have been conducted to assign meanings to each frequency band such as δ (1-3 Hz), θ (4-7 Hz), α (8-12 Hz), β (13-30 Hz), and γ (30-100 Hz).

[0003] Furthermore, a sensibility evaluation device that calculates a sensibility evaluation value from brain physiological information obtained from an electroencephalograph, as described in Patent Document 1 for example, is also known. In the technology of Patent Document 1, the subjective psychological axis of sensibility and the brain physiological index of sensibility are fused, and sensibility is visualized using feature data and weighting coefficients extracted from brain physiological information measured by the electroencephalograph.

[0004] As such electroencephalographs for measuring electroencephalograms, those used in a shielded environment in a hospital or laboratory are common. For electroencephalographs used in a shielded environment, a specialist affixes a plurality of electrodes connected with cables to the scalp, and electroencephalograms are measured under restricted movement conditions.

[0005] In recent years, wearable electroencephalographs that allow tests that may impose physical burden to be performed not in a shielded environment but, for example, outdoors, as described in Patent Document 2 for example, have also been known.

[0006] Using such wearable electroencephalographs, it becomes easy to obtain brainwave signals during activities such as playing a musical instrument or exercising. Therefore, attempts are being made to analyze the characteristics of a subject's higher-order brain functions using their brainwave signals during specific work loads. While there are said to be many individual differences in how people use their brains, some regularities have been found in rhythmic brainwaves recorded during specific work loads. [Prior art documents] [Patent Documents]

[0007] [Patent Document 1] Japanese Patent Publication No. 2019-202031 [Patent Document 2] Japanese Patent Publication No. 2024-063102 [Non-patent literature]

[0008] [Non-Patent Document 1] Takada, Muneki; Miyao, K.; Takada, Masumi; Kinoshita, Fumiya; Tahara, Hiroshi. 2019. Development and evaluation of a sports vision training system that prevents mild cognitive impairment using virtual reality. Descente Sports Science, Vol. 40, pp. 97-109. [Overview of the Initiative] [Problems that the invention aims to solve]

[0009] When attempting to analyze the characteristics of a subject's higher brain functions from their electroencephalogram (EEG), an electroencephalograph (EEG) is used to measure the EEG. Conventional EEG devices differ in electrode structure and placement, and even EEG devices with the same structure produce different signal-to-noise ratios (S / N ratios). Existing EEG analysis devices have not taken into account the influence of these S / N ratio differences between EEG devices, resulting in insufficient analytical accuracy. While this is less of an issue for EEG devices used in shielded environments such as hospitals and laboratories, where measurements can be taken under consistent conditions, the influence of the S / N ratio can be particularly significant for wearable EEG devices depending on the measurement environment. Therefore, there has been a need for the development of an analysis device that can more accurately analyze and classify the characteristics of higher brain functions regardless of the measurement environment.

[0010] In view of these circumstances, the present invention aims to provide an electroencephalogram (EEG) analysis device that can correctly analyze and classify the characteristics of a subject's higher brain functions from the subject's electroencephalogram, regardless of the measurement environment. [Means for solving the problem]

[0011] To achieve the above-described objectives of the present invention, the electroencephalogram (EEG) analysis device according to the present invention may be any device that includes: a wearable electroencephalograph (EEG) capable of detecting brain waves from the frontal lobe, including at least the prefrontal cortex to the primary motor cortex, and outputting them as EEG signals; and an analysis unit that estimates the signal-to-noise ratio (S / N ratio) of the EEG by nonlinearly analyzing the EEG signal of the subject at rest using the EEG signal from the EEG, calculates the difference between a reference spectrum obtained from the EEG signal of the subject at rest and a load spectrum obtained from the EEG signal of the subject under a specific workload, compares the difference with a quantity related to the S / N ratio of the EEG as a threshold, and analyzes the characteristics of the subject's higher brain functions based on the frequency band in which activation appears.

[0012] Here, the electroencephalograph processor should consist of a single-chip processor having a frequency filter capable of extracting analog signals within a bandwidth of 0.05 Hz to 150 Hz from the potential difference signals between each head electrode and a reference electrode, a voltage filter capable of extracting analog signals below the potential band relevant to electroencephalography, and an AD conversion unit capable of AD-converting the extracted analog signals into digital signals at a predetermined sampling frequency.

[0013] Furthermore, the system may include a database that stores the characteristics of the subject's higher-order brain functions and the frequency bands in which activation occurs, and the analysis unit may use this database to analyze the characteristics of the subject's higher-order brain functions from the electroencephalogram signals during a specific task load.

[0014] Furthermore, the analysis unit may further compare the frequency spectrum of the left brainwave signal with the frequency spectrum of the right brainwave signal to analyze the characteristics of the subject's higher-order brain functions.

[0015] Furthermore, the analysis unit may further compare a pre-reference spectrum obtained from the subject's resting electroencephalogram (EEG) signal before a specific task load with a post-reference spectrum obtained from the subject's resting EEG signal after a specific task load, in order to analyze the characteristics of the subject's higher-order brain functions.

[0016] Furthermore, the analysis unit may also include an artificial intelligence filter that removes noise from the electroencephalogram (EEG) signals.

[0017] Furthermore, the program may be configured to function as an analysis unit of an electroencephalogram (EEG) analysis device, and to function as a personal ability estimation device that estimates the individual abilities of a subject from the characteristics of higher brain functions analyzed by the analysis unit.

[0018] Furthermore, the program may be configured to cause a computer to function as the analysis unit of an electroencephalogram (EEG) analysis device, and to function as a brain function estimation device that estimates the brain function of a subject from the characteristics of higher brain functions analyzed by the analysis unit. [Effects of the Invention]

[0019] The electroencephalogram (EEG) analysis device of the present invention has the advantage of being able to accurately analyze and classify the characteristics of a subject's higher brain functions from the subject's electroencephalogram, regardless of the measurement environment. [Brief explanation of the drawing]

[0020] [Figure 1] Figure 1 is a schematic block diagram illustrating the electroencephalogram (EEG) analysis device of the present invention. [Figure 2] Figure 2 is a schematic block diagram illustrating a one-chip processor suitable for use in the electroencephalogram (EEG) analysis device of the present invention. [Figure 3] Figure 3 is an example of a graph showing the spectral power against the bandwidth of the electroencephalogram signal of a subject under workload, obtained by the electroencephalograph of the electroencephalogram analysis device of the present invention. [Modes for carrying out the invention]

[0021] The embodiments for carrying out the present invention will be described below with reference to the illustrations. The electroencephalogram (EEG) analysis device of the present invention is an EEG analysis device capable of analyzing the characteristics of a subject's higher brain functions from the subject's brainwaves. Here, higher brain functions refer to functions such as language, action, perception, cognition, memory, attention, judgment, and emotion that are performed in the cerebrum. When there is impairment in such functions, symptoms such as memory impairment, attention deficit, executive function disorder, and social behavior disorder may occur. The EEG analysis device of the present invention analyzes these characteristics of higher brain functions from the subject's brainwaves. Furthermore, the EEG analysis device of the present invention can not only determine higher brain dysfunction, but also estimate the characteristics of higher brain functions, including the subject's individuality such as brain function and personal abilities.

[0022] FIG. 1 is a schematic block diagram for explaining the electroencephalogram analysis apparatus of the present invention. As shown in FIG. 1, the electroencephalogram analysis apparatus of the present invention is mainly composed of an electroencephalograph 10 and an analysis unit 20.

[0023] The electroencephalograph 10 is capable of detecting a subject's electroencephalogram and outputting it as an electroencephalogram signal. In order to analyze the characteristics of the subject's higher brain functions, it is necessary to apply various task loads to the subject, so the electroencephalograph 10 is preferably a wearable one. The wearable electroencephalograph 10 consists of head electrodes 11, a reference electrode 12, and a processor 13. The head electrodes 11 are two head electrodes 11 arranged on the left and right of the subject's head R , 11 L . The two head electrodes 11 R , 11 L are arranged on the left and right of the subject's head, so as to be configured to detect electroencephalograms from the frontal lobe including at least the prefrontal cortex to the primary motor cortex. The two head electrodes 11 R , 11 L do not necessarily need to be accurately arranged symmetrically on the left and right of the head. The two head electrodes 11 R , 11 L do not have to be arranged symmetrically on the left and right as long as they are arranged at positions where electroencephalograms from the frontal lobe including at least the prefrontal cortex to the primary motor cortex can be detected. The reference electrode 12 is an electrode that determines a reference potential. For example, it may be one arranged on the subject's ear. The reference electrode 12 does not particularly need to be arranged on the ear as long as it is a place where the reference potential can be measured. The head electrode 11 relative to the reference electrode 12 R and / or the head electrode 11 L As long as the electrodes are configured to detect potential difference signals between the respective electrodes, the head electrodes 11 and the reference electrode 12 may be dry electrodes or wet electrodes. Furthermore, one head electrode 11 may be used as the reference electrode 12, and the configuration may be such that a potential difference signal with the other head electrode 11 can be detected. That is, the head electrode 11 may be shared as the reference electrode 12.

[0024] The processor 13 has each head electrode 11 configured in this way R ,11 L The potential difference signal from the reference electrode 12 is processed and output as an electroencephalogram (EEG) signal. The wearable EEG 10 is used in environments with significant external noise influence, rather than in shielded environments such as hospitals or laboratories. Therefore, it is preferable that the processor 13 has noise immunity. Specifically, it is preferable that the processor 13 that converts the obtained potential difference signal into an EEG signal consists of a single-chip processor. Figure 2 is a schematic block diagram illustrating a single-chip processor suitable for use in the EEG analysis device of the present invention. In the figure, parts with the same reference numerals as in Figure 1 represent the same components. As shown in the figure, it is preferable that the processor 13 consists of a single-chip processor having a frequency filter 14, a voltage filter 15, and an AD conversion unit 16.

[0025] The frequency filter 14 controls each head electrode 11 R ,11 L It is sufficient if the system can extract analog signals within a frequency range of, for example, 0.05 Hz to 150 Hz from the potential difference signal between the reference electrode 12 and the system. That is, it is sufficient if it can divide the frequency bands into, for example, δ (0.05-3 Hz), θ (4-7 Hz), α (8-12 Hz), β (13-30 Hz), and γ (30-100 Hz). Furthermore, a higher frequency band division, such as 100 Hz-150 Hz, may also be used. The frequency filter 14 can be used to divide the frequency bands into δ, θ, α, β, and γ, but this division will vary depending on the subject and will change accordingly.

[0026] Furthermore, the voltage filter 15 only needs to be capable of extracting analog signals below the potential band related to brain waves. In other words, the voltage filter 15 only needs to remove signals in the potential band that constitute external noise, such as non-biological noise, such as power supply noise or environmental noise. The voltage filter 15 may also be configured to remove non-biological noise using artificial intelligence, for example. Note that the voltage filter 15 does not necessarily need to block all signals other than brain waves; it is sufficient if it can block obvious non-biological noise. In other words, the output of the voltage filter 15 may include biological signals other than brain waves, such as some electromyography.

[0027] The AD conversion unit 16 is capable of AD conversion of the extracted analog signal into a digital signal at a predetermined sampling frequency. In other words, it is preferable that the processor 13, which consists of a single-chip processor, extracts signals related to electroencephalography (EEG) from the input analog signal in a wide bandwidth as analog signals, converts them into digital signals via a predetermined voltage filter, and outputs them. Such a processor 13 makes it possible to output EEG signals necessary for analyzing the characteristics of the subject's higher brain functions.

[0028] Thus, it is preferable that the processor 13 of the electroencephalogram (EEG) analyzer of the present invention consists of a one-chip processor in which the processing circuit that receives analog signals, namely the potential difference signals from the head electrodes and reference electrodes, and outputs digital signals, namely the EEG signals, specifically the frequency filter 14, voltage filter 15, and AD conversion unit 16, are all configured as a single semiconductor chip. The frequency filter 14, voltage filter 15, and AD conversion unit 16 are all enclosed within a single semiconductor package, there are no external terminals connecting the parts, and all analog signal processing is performed by the one-chip processor. Therefore, noise immunity in processing weak signals, namely the potential difference signals from the head electrodes and reference electrodes, is greatly improved.

[0029] The analysis unit 20 uses electroencephalogram (EEG) signals from the electroencephalograph (EEG) 10 to analyze the characteristics of the subject's higher-order brain functions based on the frequency bands in which activation occurs. First, the analysis unit 20 estimates the signal-to-noise ratio (S / N ratio) of the EEG 10 by performing a nonlinear analysis on the subject's resting EEG signals. Nonlinear analysis and the S / N ratio will be described later. To estimate the S / N ratio, it is sufficient to first measure the resting EEG signals for each subject and each EEG meter. Then, during the analysis, the difference between the reference spectrum obtained from the subject's resting EEG signal and the load spectrum obtained from the subject's EEG signal under a specific workload is calculated. This difference between the resting and workload EEG signals is compared using a threshold value related to the S / N ratio of the EEG 10 to determine the frequency band in which activation occurs. The analysis unit 20 can be composed of electronic circuits or a computer capable of such signal processing.

[0030] Furthermore, the analysis unit 20 of the electroencephalogram (EEG) analyzer of the present invention may also have an artificial intelligence filter that removes biological noise such as blinking that is unnecessary for the analysis of the characteristics of higher brain functions. That is, in addition to a noise filter for non-biological noise on the EEG meter 10 side, it is also possible to provide a noise filter for biological noise that is unnecessary for the analysis on the analysis unit 20 side. Note that the analysis unit 20 can be configured by software. Therefore, the target of removal for the noise filter by the analysis unit 20 can be easily changed by changing the parameters of the software. As described above, the output of the voltage filter 15 on the EEG meter 10 side may include biological signals other than EEG, but such signals can be easily filtered by adjusting the parameters according to the purpose and application of the analysis. Necessary EEG signals and unnecessary biological signals can be distinguished by an algorithm, but an artificial intelligence filter that can more flexibly reflect the analysis history is preferred.

[0031] With this configuration, the electroencephalogram (EEG) analysis device of the present invention can accurately analyze and classify the characteristics of a subject's higher brain functions from their electroencephalogram, regardless of the measurement environment.

[0032] The electroencephalograph 10 only needs to output brainwave signals to the analysis unit 20 described later via wired communication using a cable. However, it may also be capable of outputting signals via wireless communication. That is, the electroencephalograph 10 may have a wireless transmitter that transmits brainwave signals wirelessly. In that case, the analysis unit 20 only needs to have a wireless receiver that receives wireless signals from the wireless transmitter. For example, if the analysis unit 20 is to be operated by a program on an electronic information terminal such as a computer, the electronic information terminal and the electroencephalograph 10 can be connected using a communication method such as Bluetooth (registered trademark) to send and receive brainwave signals.

[0033] The following provides a detailed explanation of nonlinear analysis of electroencephalogram (EEG) signals and the signal-to-noise ratio (S / N ratio). The EEG device 10 has left and right head electrodes 11 R ,11 L Two series of electroencephalogram (EEG) signals are recorded. Here, the EEG signal of the subject at rest is defined as the reference EEG. The time series data are then recorded as {x r L (t)}, {x r R This is represented as (t). It is desirable to measure the electroencephalogram (EEG) signal during rest for at least 3 minutes.

[0034] Then, in order to estimate the signal-to-noise ratio, the translation error E is calculated using nonlinear analysis. trans The reciprocal of or a statistic related thereto is calculated. That is, in the electroencephalogram analyzer of the present invention, the signal-to-noise ratio is the translation error E calculated by nonlinear analysis. trans This refers to the reciprocal of or a statistic related to it.

[0035] The attractor of chaotic motion is constructed by embedding time-series data {x(t)}, as proposed by Takens, into phase space. Here, embedding is the process of defining the following vectors in N-dimensional phase space (embedding space), whose components are the data values ​​at times t, t+Δt, t+2Δt, ..., t+(N-1)Δt, and drawing the trajectory.

number

[0036] The Wayland algorithm assumes that the difference vector v(t) = x(t+τ) - x(t) in the embedding space characterizes the nonlinear variation of the trajectory, and estimates the translation error defined below in the m-dimensional embedding space. From the delayed coordinate system {x(t)}, a continuous trajectory without false intersections can be reconstructed in the higher-dimensional embedding space.

[0037] The linear correlation between adjacent vectors x(t) and x(t+τ) is removed by resampling the time series with respect to each embedding delay τ. For example, the autocorrelation function r(t) is used to estimate the smallest time difference t=τ at which the linear correlation between elements of the time series data disappears. When the time difference t is increased, t=τ is defined as the point at which l / e ≤ 1 for the first time, and the linear correlation between elements of the time series data is considered to have disappeared. This τ is called the embedding delay.

[0038] The steps (1)-(5) of the Wayland algorithm are shown below. (1) Embed the delayed coordinate system {x(t)} into each space. (2) M start times t0 are randomly selected. (3) Translation error E trans The reciprocal of or a statistic related to it can be calculated as follows:

number

number

[0039] To give a concrete example of the translation error obtained in this way, when using a two-station simple electroencephalograph from Bio Search Co., Ltd., the translation error was estimated to be 0.5. In other words, the analysis unit 20 can use the translation error, which is a quantity related to the S / N ratio of each electroencephalograph 10 obtained in this way, as a threshold. For example, when using different electroencephalographs 10, calibration can be performed by estimating the translation error, which is a quantity related to the S / N ratio. By using a quantity related to the S / N ratio of the electroencephalograph 10, it becomes possible to correctly analyze and classify the characteristics of the subject's higher brain functions from the subject's brain waves, regardless of the measurement environment of the brain wave signal. Furthermore, if an electroencephalograph 10 consisting of a single chip processor as described above is used, it is also possible to reduce the influence of external noise.

[0040] Next, the left and right head electrodes 11 of the subject at rest R ,11 L By performing frequency analysis on the time-series data of the reference electroencephalogram (EEG) obtained from the source, the frequency spectrum in the range of 0.05 Hz to 150 Hz is calculated. Here, this frequency spectrum data is defined as the reference spectrum. The time-series data is then divided into P f [x r L (t)],P f [x r R This is represented as (t).

[0041] Subsequently, electroencephalogram (EEG) signals are recorded during specific workloads of the subjects. The time-series data are then divided into {x L (t)}, {x R This is represented as (t). By performing frequency analysis on the time-series data of this electroencephalogram signal, the frequency spectrum in the range of 0.05 Hz to 150 Hz is calculated, for example. Here, this frequency spectrum data is defined as the load spectrum. The time-series data is then expressed as P f [x L (t)],P f [x R This is represented as (t).

[0042] Then, the difference between the reference spectrum and the load spectrum is calculated as follows. Note that the difference between the reference spectrum and the load spectrum is measured by the right head electrode 11 R Even just the difference from that would be fine, and the left head electrode 11 L Even just the difference from that is fine. Furthermore, both left and right head electrodes 11 R ,11 L The difference between each of them would also be acceptable.

number

[0043] Note that the left and right head electrodes 11 R ,11 L The time-series data of one electroencephalogram (EEG) may be used as a reference for the time-series data of the other EEG. That is, the frequency spectrum of the left EEG signal and the frequency spectrum of the right EEG signal may be further compared to analyze the characteristics of the subject's higher-order brain functions. Specifically, the left-right balance regarding the activation of brain activity may be captured using the following difference or similar statistical measures.

number

[0044] At this time, the left and right head electrodes 11 of the subject in a resting state R ,11 L The measurement of the reference spectrum obtained from this can be omitted.

[0045] Then, electrodes 11 on the left and right heads of the subject R ,11 LBy processing the electroencephalogram (EEG) signals as described above, the degree of brain activity activation in response to a specific task load, or the number of time steps that can be considered activated, is calculated. In this process, the integral value of the spectral power in each band should be normalized using the median and interquartile deviation of the spectral power in each band obtained during rest. Specifically, the left and right EEG signals are normalized using the following formula, and this value is called the reference value of the spectral power in each band. For example, the bands are divided into 11 sections.

number

[0046] Then, the duration of time during which the spectral power in each frequency band increases and the characteristic brain activity of each band increases drastically is calculated for each subject. In other words, the translational error, which is a quantity related to the signal-to-noise ratio, can be used as a threshold for comparison. Specifically, for example, the frequency bands in which the baseline value of spectral power exceeds the reciprocal of the translational error can be extracted, as shown below.

number

[0047] In this way, the frequency bands in which brain activity activation in response to a specific task load of the subject is observed are extracted. Figure 3 shows an example of a graph of spectral power against frequency bands of the brainwave signal of a subject during a task load, obtained by the electroencephalograph of the electroencephalogram analysis device of the present invention. In this example, the left head electrode 11 L Regarding the electroencephalogram signals from P, the resting reference spectrum P f [x r L (t) and load spectrum P during work load f [x L This graph compares (t) with (t). As shown in the figure, it can be seen that for a specific workload, there is a frequency band in which the baseline value of spectral power increases significantly by more than the reciprocal of the translational error, i.e., a frequency band in which activation occurs.

[0048] The following shows some results of analyzing the characteristics of higher brain functions using the electroencephalogram (EEG) analyzer of the present invention, after giving subjects a specific workload. Table 1 below shows the results of classifying the EEG signals by bandwidth for two groups of subjects—a normal group and a cognitive impairment group—when a specific workload related to spatial cognition was applied, using the EEG analyzer of the present invention. In the table, "left brain" refers to the left head electrode 11. L EEG signals from the brain, "right brain" is shown on the right side of the head electrode 11 R The electroencephalogram (EEG) signals were analyzed using the EEG analysis device of the present invention. A "**" was used to indicate a significant increase in the cognitive impairment group compared to the normal group, and an "*" was used to indicate a significant trend. A "-" was used if neither category applied. [Table 1] Note that the frequency band divisions in the table, such as δ, θ, α, Low-α, High-α, β, Low-β, High-β, and γ, vary depending on the subject and should be adjusted accordingly. The frequency filter 14 described above should be used for appropriate division. From Table 1, it can be seen that, as a characteristic of higher brain functions, the workload related to spatial cognition is significantly increased in the high-bandwidth area of ​​the left brain in the cognitive impairment group. That is, in the cognitive impairment group, activation is observed in the high-bandwidth area of ​​the left brain, indicating an inappropriate state. Specifically, if there is a significant difference between High-β and γ in the left brain and no change in the right brain, it can be determined that there is or is at risk of spatial cognitive impairment. Furthermore, even if there is a significant difference between High-β and γ in the left brain, if changes are also observed in the right brain, the entire brain is activated, and therefore, spatial cognitive impairment cannot be determined. Therefore, using this classification, it is possible to analyze whether a subject has spatial cognitive impairment, a higher brain function, based on their electroencephalogram (EEG) signals during a spatial cognition-related workload.

[0049] Furthermore, Table 2 below shows the results of classifying the electroencephalogram (EEG) signals of two groups of subjects—a normal group and a cognitively impaired group—by bandwidth using the EEG analysis device of the present invention when a memory-related workload was applied as a specific task load. In the table, "left brain" refers to the left head electrode 11.L EEG signals from the brain, "right brain" is shown on the right side of the head electrode 11 R The electroencephalogram (EEG) signals were analyzed using the EEG analysis device of the present invention. A "**" was used to indicate a significant increase in the cognitive impairment group compared to the normal group, and an "*" was used to indicate a significant trend. A "-" was used if neither category applied. [Table 2] Table 2 shows that, as a characteristic of higher brain functions, the workload related to memory tends to increase significantly in the high-bandwidth region in the cognitive impairment group, and this tendency is particularly pronounced in the left hemisphere. Specifically, if there are significant differences in β, High-β, and γ in the left hemisphere, and little change in the right hemisphere, it can be determined that there is memory impairment or that the individual is at risk of developing memory impairment. Furthermore, even if there are significant differences in β, High-β, and γ in the left hemisphere, if there are also significant differences in the right hemisphere, it cannot be determined that there is memory impairment because the entire brain is activated. Therefore, by using this classification, it is possible to analyze whether or not a subject has memory impairment, which is a higher brain function, based on the electroencephalogram (EEG) signals during memory-related workloads.

[0050] Furthermore, Table 3 below shows the results of classifying the electroencephalogram (EEG) signals of two groups of subjects—a normal group and a cognitively impaired group—by frequency band using the EEG analysis device of the present invention when a specific workload related to orientation was applied. In the table, "left brain" refers to the left head electrode 11. L EEG signals from the brain, "right brain" is shown on the right side of the head electrode 11 R The electroencephalogram (EEG) signals were analyzed using the EEG analysis device of the present invention. A "**" was used to indicate a significant increase in the cognitive impairment group compared to the normal group, and an "*" was used to indicate a significant trend. A "-" was used if neither category applied. [Table 3] Table 3 shows that, as a characteristic of higher brain functions, the workload related to orientation tends to increase significantly in the high-bandwidth in the cognitive impairment group, and this tendency is particularly pronounced in the left hemisphere. Specifically, if there are significant differences in β, High-β, and γ in the left hemisphere, and little change in the right hemisphere, it can be determined that the individual has or is at risk of disorientation. However, even if there are significant differences in β, High-β, and γ in the left hemisphere, if changes are also observed in the right hemisphere, the entire brain is activated, and therefore, disorientation cannot be determined. Thus, using this classification, it is possible to analyze whether or not a subject has disorientation, a higher brain function, based on their electroencephalogram (EEG) signals during an orientation-related workload.

[0051] Thus, the electroencephalogram (EEG) analysis device of the present invention makes it possible to accurately analyze and classify the characteristics of a subject's higher brain functions from their EEG by estimating the signal-to-noise ratio for each EEG device and using it as a threshold. Furthermore, by using an EEG device 10 that can measure in higher frequency bands such as 100Hz-150Hz, it becomes possible to analyze the characteristics of higher brain functions using even more information.

[0052] When analyzing the characteristics of higher brain functions of subjects using the electroencephalogram (EEG) analysis device of the present invention, no activation of brain activity corresponding to a specific workload is observed in the group showing aptitude for that workload. On the other hand, in the group showing no aptitude for that specific workload, there is a significantly increased frequency band, indicating activation. Therefore, by comparing the magnitude of the activation in the group showing activation and the group not showing activation, i.e., the group showing aptitude and the group not showing activation, it is possible to analyze whether or not a person is aptitude for a particular type of thinking.

[0053] Next, we will describe an example of using the electroencephalogram (EEG) analysis device of the present invention to estimate the individual abilities of subjects, as one of the characteristics of higher brain functions. Below, Table 4 shows the results of classifying the electroencephalogram signals by bandwidth for two groups of subjects, a science group and a humanities group, when a specific workload, for example, a mathematical task, was applied, using the EEG analysis device of the present invention. In the table, "left brain" refers to the left head electrode 11L EEG signals from the brain, "right brain" is shown on the right side of the head electrode 11 R The electroencephalogram (EEG) signals from the brain were analyzed using the EEG analysis device of the present invention. Furthermore, "left brain / right brain" refers to the left and right head electrodes 11. R ,11 L The analysis was performed using the time-series data of one brainwave as a reference to the time-series data of the other brainwave. A "-" was added to the frequency band where the group showing aptitude significantly increased. In other words, a "-" was added to the frequency band where the science-oriented group, who are thought to be aptitude for the mathematical task, significantly increased. For "left brain / right brain," a "+" was added to the frequency band where the group not showing aptitude significantly increased. In other words, a "+" was added to the frequency band where the humanities-oriented group, who are thought to be aptitude for the mathematical task, significantly increased. Furthermore, a "±" was added to the frequency band where the group showing aptitude was significantly smaller. [Table 4] Table 4 shows that, as a characteristic of higher-order brain function, the group exhibiting aptitude for mathematical tasks (the science-oriented group) tends to show a significant increase in the high-bandwidth area of ​​the right hemisphere, which is responsible for spatial cognitive processing. Specifically, if there is a significant difference in High-β, Low-γ, and Mid-γ between the left and right hemispheres, it can be determined that the individual has a science-oriented brain. Conversely, if there is no significant difference in High-β, Low-γ, and Mid-γ between the left and right hemispheres, it can be determined that the individual has a humanities-oriented brain. Therefore, using this classification, it becomes possible to estimate whether a subject has a humanities-oriented or science-oriented brain, which is one of the characteristics of higher-order brain function, based on their electroencephalogram (EEG) signals during a mathematical task.

[0054] Furthermore, as a specific workload, for example, a classical studies task, the electroencephalogram (EEG) signals of two groups of subjects—a science-oriented group and a humanities-oriented group—were classified by frequency band using the EEG analysis device of the present invention, as shown in Table 5 below. In the table, "left brain" refers to the left head electrode 11. L EEG signals from the brain, "right brain" is shown on the right side of the head electrode 11 RThe electroencephalogram (EEG) signals from the brain were analyzed using the EEG analysis device of the present invention. Furthermore, "left brain / right brain" refers to the left and right head electrodes 11. R ,11 L The analysis was performed using the time-series data of one brainwave as a reference to the time-series data of the other brainwave. A "-" was added to the frequency band in which the group showing aptitude significantly increased. In other words, a "-" was added to the frequency band in which the humanities group, who are thought to be aptitude for classical tasks, significantly increased. For "left brain / right brain," a "-" was added to the frequency band in which the group showing aptitude significantly increased. In other words, a "-" was added to the frequency band in which the humanities group, who are thought to be aptitude for classical tasks, significantly increased. Furthermore, a "±" was added to the frequency band in which the group not showing aptitude significantly decreased. [Table 5] Table 5 shows that, as a characteristic of higher-order brain function, the science-oriented group, which shows a lack of aptitude for the workload related to classical tasks, tends to show a significant increase in the right hemisphere in the high-bandwidth range. Specifically, if a significant difference is observed in High-β, Low-γ, and Mid-γ between the left and right hemispheres, it can be determined that the individual has a science-oriented brain. Conversely, if no significant difference is observed in High-β, Low-γ, and Mid-γ between the left and right hemispheres, it can be determined that the individual has a humanities-oriented brain. Therefore, using this classification, it becomes possible to estimate whether a subject has a humanities-oriented or science-oriented brain, which is one of the characteristics of higher-order brain function, based on the electroencephalogram (EEG) signals during the workload related to classical tasks.

[0055] Furthermore, Table 6 below shows the results of classifying the electroencephalogram (EEG) signals of two groups of subjects—a science-oriented group and a humanities-oriented group—by bandwidth using the EEG analysis device of the present invention, under a specific workload, such as creating an information processing program. In the table, "left brain" refers to the left head electrode 11. L EEG signals from the brain, "right brain" is shown on the right side of the head electrode 11 R The electroencephalogram (EEG) signals from the brain were analyzed using the EEG analysis device of the present invention. Furthermore, "left brain / right brain" refers to the left and right head electrodes 11. R ,11 LThe analysis was performed using the time-series data of one electroencephalogram (EEG) as a reference for the time-series data of the other EEG. A "+" was added to the frequency bands where the group showing no aptitude showed a significant increase. A "±" was added to the frequency bands where the group showing aptitude showed a significantly smaller decrease. A "-" was added to the frequency bands where the group showing aptitude showed a significant increase. [Table 6] Table 6 shows that, as a characteristic of higher brain function, the group lacking aptitude for the information processing program creation workload tends to show a significant increase in the left brain at high bandwidths, while the group showing aptitude tends to show a significant increase in the left brain at low bandwidths. Specifically, if right-brain dominance is observed in High-β, Low-γ, and Mid-γ of the left / right brain, it can be determined that the individual lacks aptitude for information processing program creation. Conversely, if left-brain dominance is observed in δ and θ of the left / right brain, it can be determined that the individual has aptitude for information processing program creation. Furthermore, this result suggests that pattern recognition is performed in the left brain. Therefore, by using this classification, it is possible to estimate whether a subject has aptitude for information processing program creation, which is an individual ability, as one of the characteristics of higher brain function, based on their electroencephalogram (EEG) signals during the information processing program creation workload.

[0056] Thus, with the electroencephalogram (EEG) analysis device of the present invention, by estimating the signal-to-noise ratio for each EEG device and using this as a threshold, it is possible to accurately analyze and classify the characteristics of a subject's higher-order brain functions from their EEG, such as whether they have a humanities-oriented or science-oriented brain, and even whether they are suited to specific tasks such as programming or playing a musical instrument. Furthermore, by using an EEG device 10 that can measure in a higher frequency band such as 100Hz-150Hz, it becomes possible to analyze the characteristics of higher-order brain functions using even more information.

[0057] The electroencephalogram (EEG) analysis device of the present invention may include a database 30 that stores the classification information described above, as shown in Figure 1. Specifically, the database 30 should store the characteristics of a subject's higher brain function and the frequency bands in which activation occurs, linked together. For example, it should store the results of classifying the characteristics of higher brain function for multiple groups of subjects using the EEG analysis device of the present invention. The analysis unit 20 can then use such a database 30 to analyze the characteristics of a subject's higher brain function from the EEG signal during a specific task load.

[0058] In the electroencephalogram (EEG) analysis device of the present invention, such analysis results may be displayed graphically on a display screen of a mobile device or computer as appropriate.

[0059] In the illustrated example above, an example was described in which the difference between a reference spectrum obtained from the subject's resting electroencephalogram (EEG) signal and a load spectrum obtained from the subject's EEG signal during a specific workload was calculated. However, the EEG analysis device of the present invention may further analyze the characteristics of the subject's higher brain functions using the subject's resting EEG signal after a specific workload. Specifically, the analysis unit 20 may further compare a pre-reference spectrum obtained from the subject's resting EEG signal before a specific workload with a post-reference spectrum obtained from the subject's resting EEG signal after a specific workload. By using the post-reference spectrum in this way, the characteristics of the subject's higher brain functions can be analyzed and classified more accurately.

[0060] Furthermore, using the electroencephalogram (EEG) analysis device of the present invention, it becomes possible to estimate brain function as, for example, the subject's higher-order brain function. For example, for the three items mentioned above—spatial cognition, orientation, and memory—a distance index representing which age group the degree of brain activity is similar to can be calculated, and this index can be added to the subject's actual age to present as brain function. Specifically, brain function can be presented using a procedure such as the following. (1) First, three groups of subjects—young people (20-39 years old), middle-aged people (40-59 years old), and elderly people (60-80 years old)—were given a load on three items: spatial cognition, orientation, and memory. The left and right head electrodes 11 of the subjects were then measured. R ,11 L The two electroencephalogram (EEG) signals are processed by the analysis unit 20 of the EEG analysis device of the present invention, and the average power spectral density is calculated for all nine bands: w-α, High-α, β, Low-β, High-β, and γ. (2) Using the mean power spectral density obtained, select the indicators for which the significance probability (P-value) is less than 0.01, by performing a Kruskal-Wallis test, for example, assuming that there is no difference in the median distributions of the three groups for all indicators. Perform a box-Cox transformation on these selected indicators and store the parameters used in this transformation in advance. (3) Create an evaluation plane from all combinations of selecting two indicators from the selected indicators. (4) Then, when a subject whose brain function we wanted to estimate was given a load on three items: spatial cognition, orientation, and memory, the left and right head electrodes 11 of the subject were measured. R ,11 L The two electroencephalogram (EEG) signals are processed by the analysis unit 20 of the EEG analysis device of the present invention, and the average power spectral density is calculated for all nine bands: w-α, High-α, β, Low-β, High-β, and γ. (5) Select the index chosen in (2) above and perform a box-cox transformation using pre-stored parameters. Using the results of this calculation, calculate the Mahalanobisno distance from the center coordinates of each of the three groups (young, middle-aged, and elderly) in each evaluation plane. (6) Multiply the distances of each of these three groups by a coefficient and add them together to calculate the difference age. Perform the same operation for all combinations in (3) above, find the average value and calculate the average difference age. (7) The result of adding the average difference age calculated in (6) above to the subject's actual age is presented as brain function.

[0061] In the illustrated example described above, the analysis unit 20 was described on the premise that it is composed of an electronic circuit or computer capable of performing predetermined signal processing. However, the present invention is not limited to this, and may be a program. That is, it may be a program that causes a computer to function as the analysis unit 20 of the electroencephalogram (EEG) analysis device of the present invention described above. The program should be configured to function as a personal ability estimation device that estimates the individual ability of a subject from the characteristics of higher brain functions analyzed by the analysis unit 20. For example, such a program can be installed on an electronic information terminal, the electronic information terminal and the electroencephalograph 10 can be connected by a communication method such as Bluetooth (registered trademark), and the personal ability estimation program that estimates the individual ability of a subject can be executed.

[0062] Furthermore, the program may be configured to function as a brain function estimation device that estimates the subject's brain function from the characteristics of higher brain functions analyzed by, for example, the analysis unit 20. For example, such a program can be installed on an electronic information terminal, the electronic information terminal and the electroencephalograph 10 can be connected via a communication method such as Bluetooth (registered trademark), and the brain function estimation program that estimates the subject's brain function can be executed.

[0063] It should be noted that the electroencephalogram (EEG) analysis device of the present invention is not limited to the illustrated example described above, and various modifications can be made without departing from the spirit of the present invention. [Explanation of symbols]

[0064] 10. Electroencephalograph (EEG) 11 R ,11 L head electrode 12 Reference electrodes 13 processors 14 Frequency Filters 15 Voltage Filter 16 AD Conversion Unit 20 Analysis Department 30 databases

Claims

1. An electroencephalogram (EEG) analyzer capable of analyzing the characteristics of a subject's higher brain functions from the subject's electroencephalogram, wherein the EEG analyzer is A wearable electroencephalograph comprising two head electrodes positioned on the left and right sides of the subject's head, a reference electrode, and a processor that outputs an electroencephalogram signal using the potential difference signal between each head electrode and the reference electrode, capable of detecting brain waves from the frontal lobe, including at least the prefrontal cortex to the primary motor cortex, and outputting it as an electroencephalogram signal, An analysis unit that uses the electroencephalogram (EEG) signal from the aforementioned EEG to estimate the signal-to-noise ratio of the EEG by performing a nonlinear analysis of the subject's resting EEG signal, calculates the difference between the reference spectrum obtained from the subject's resting EEG signal and the load spectrum obtained from the subject's EEG signal during a specific task load, compares the difference with a quantity related to the EEG's signal-to-noise ratio as a threshold, and analyzes the characteristics of the subject's higher brain functions based on the frequency band in which activation appears. An electroencephalogram (EEG) analyzer characterized by being equipped with the following:

2. In the electroencephalogram (EEG) analyzer according to claim 1, the processor of the electroencephalograph is: A frequency filter capable of extracting analog signals within a bandwidth of 0.05 Hz to 150 Hz from the potential difference signals between each head electrode and the reference electrode, A voltage filter capable of extracting analog signals below the potential band related to brain waves, An AD conversion unit capable of converting the extracted analog signal into a digital signal at a predetermined sampling frequency, An electroencephalogram (EEG) analyzer characterized by comprising a single-chip processor having the following features.

3. An electroencephalogram (EEG) analyzer according to claim 1, further comprising a database that stores the characteristics of a subject's higher brain functions and the frequency bands in which activation occurs, The aforementioned analysis unit is an electroencephalogram (EEG) analysis device characterized by using a database to analyze the characteristics of a subject's higher-order brain functions from EEG signals during a specific task load.

4. An electroencephalogram (EEG) analyzer according to claim 1, wherein the analysis unit further compares the frequency spectrum of the left EEG signal with the frequency spectrum of the right EEG signal to analyze the characteristics of the subject's higher brain functions.

5. An electroencephalogram (EEG) analyzer according to claim 1, wherein the analysis unit further compares a prereference spectrum obtained from the resting EEG signal of the subject before a specific work load with a postreference spectrum obtained from the resting EEG signal of the subject after a specific work load, and analyzes the characteristics of the subject's higher brain functions.

6. An electroencephalogram (EEG) analyzer according to claim 1, wherein the analysis unit further comprises an artificial intelligence filter for removing noise from the EEG signal.

7. A program for causing a computer to function as an analysis unit of the electroencephalogram analysis device described in claim 1, wherein the program is configured to function as a personal ability estimation device that estimates the personal abilities of a subject from the characteristics of higher brain functions analyzed by the analysis unit.

8. A program for causing a computer to function as an analysis unit of the electroencephalogram analysis device described in claim 1, wherein the program is configured to function as a brain function estimation device that estimates the brain function of a subject from the characteristics of higher brain functions analyzed by the analysis unit.

Citation Information

Patent Citations

  • Sensibility evaluation device, sensibility evaluation method, and sensibility multiaxial model construction method

    JP2019202031A

  • Biological information recording device and biological information recording program

    JP2024063102A