JP4571709B1 - Vibration resistance electroencephalograph - Google Patents

Abstract

An object of the present invention is to realize a vibration-resistant electroencephalograph capable of measuring without mixing vibration noise even when a user is in an operating state, and capable of being easily worn. A plurality of electrodes 2, a vibration sensor 3 disposed on the frame 1, a filter unit 4 for removing electrical noise components from signals detected by the electrodes 2, and an amplifying unit 5 for amplifying a signal that has passed through the filter unit 4. A digital converter 6 that converts the analog signal amplified by the amplifier 5 into a digital signal, a filter 7 that removes electrical noise components from the signal detected by the vibration sensor 3, and a signal that has passed through the filter 7 An amplifying unit 8 for amplifying the signal, a digital converting unit 9 for converting an analog signal amplified by the amplifying unit 8 into a digital signal, the digital converting unit 9 and the digital converting unit It solved the problem by anti-vibration electroencephalograph consisting calculating unit 10 for removing operation of the vibration noise component from the output value of.
[Selection] Figure 1

Description

  The present invention relates to a vibration-resistant electroencephalograph.

  In order to correctly detect an electroencephalogram, which is a weak electric signal of about 50 microvolts, various noise removal techniques have been conventionally used.

  First, a technique called differential amplification has been used for external noise that enters from surrounding electrical equipment such as fluorescent lamps. This is a technique for removing noise mixed in both of the head by simultaneously measuring a portion where a brain wave is generated and a portion where the brain wave is not generated and subtracting both measured values.

  Further, when a commercial power source is used as a power source for an electroencephalograph, a hum noise of 50 Hz or 60 Hz is mixed. Therefore, an analog filter or a digital filter has been used to remove this noise.

  In addition, since the electrode to be mounted on the head is made of metal, a technique called an unpolarized electrode using silver chloride and an electrolytic solution has been used to suppress noise generated when the electrode itself generates power. .

  However, even with an electroencephalograph that combines the above technologies, it is difficult to remove vibration noise generated by the vibration of the electrode, and the user must always remain stationary so that the electrode does not vibrate during measurement. I had to beat.

  As a method for removing this vibration noise, there is a method using an active electrode. Normally, electrodes used for measuring biosignals have a structure in which a long conducting wire extends from the electrode plate and can be connected to the amplifier of the measuring instrument body. This active electrode is an integrated electrode plate and amplifier. As a result, it is possible to reduce noise mixed in a long conductor. This technology is used in pulse oximeters that are used for the purpose of continuously measuring the pulse and the like, and can also measure the user during exercise. However, since the electroencephalogram signal is a voltage signal that is one hundredth of the pulse signal, vibration noise cannot be removed by this method.

  In addition, there is a method of fixing specially shaped electrodes such as three legs to the scalp with conductive paste, which is a method that EEG researchers have used in the past, especially for measuring the EEG of subjects who are moving It is very effective to do. However, with this method, it is impossible for the person himself to wear the electrode alone, and it takes much time to install the electrode even if the specialist assists the wearing.

  Japanese Patent Laid-Open No. 2001-231767 mentions a method for statistically calculating and reducing the amount of noise, which is mainly intended to instantaneously measure the evoked waveform in an ARB test (auditory brainstem response). It is not intended for use in normal waveform measurements. In addition, the calculation is complicated because it is not a method of identifying the cause of noise and taking measures.

JP 2001-231767 A

  The present invention has been made in view of the above circumstances, and the problem to be solved is that measurement is possible without mixing vibration noise even when the user is in an operating state, and simple mounting is possible. This is the realization of a vibration-resistant electroencephalograph.

The present invention has been made to solve the above-described problems, and the invention according to claim 1 includes a frame 1, a plurality of electrodes 2 disposed on the frame 1, and a vibration sensor 3 disposed on the frame 1. A filter unit 4 that removes an electrical noise component from a signal detected by the electrode 2, an amplifier unit 5 that amplifies a signal that has passed through the filter unit 4, and an analog signal that is amplified by the amplifier unit 5 is converted into a digital signal. A filter unit 7 that removes an electrical noise component from signals detected by the digital conversion unit 6 and the vibration sensor 3, an amplification unit 8 that amplifies the signal that has passed through the filter unit 7, and an analog signal that is amplified by the amplification unit 8 Is converted into a digital signal, and a calculation unit 10 that performs a removal calculation of vibration noise components from the output values of the digital conversion unit 9 and the digital conversion unit 6 A vibration-resistant electroencephalograph which al made.

According to the second aspect of the present invention, there is provided a plurality of annular frames 1 detachably attached to the head and at least one projectingly disposed on the inner wall of the frame 1 and at least one disposed on the frame 1 via a conductive wire that can be bent. An electrode 2, a vibration sensor 3 disposed close to the electrode 2 or fixed to the frame 1, and a filter unit 4 that removes signals other than approximately 4 Hz to approximately 30 Hz among signals detected by the electrode 2, An amplification unit 5 that amplifies the signal that has passed through the filter unit 4, a digital conversion unit 6 that converts an analog signal amplified by the amplification unit 5 into a digital signal, and approximately 4 Hz among signals detected by the vibration sensor 3. A filter unit 7 that removes signals other than approximately 30 Hz from the filter unit 7, an amplifier unit 8 that amplifies the signal that has passed through the filter unit 7, and an analog signal amplified by the amplifier unit 8 It comprises a digital conversion unit 9 for converting to a digital signal, and a calculation unit 10 for performing an operation for removing vibration noise components from the output values of the digital conversion unit 9 and the digital conversion unit 6, and the calculation unit 10 includes the digital conversion unit 6 and the digital conversion. The ratio coefficient calculating function 11 for calculating the difference between the maximum value and the minimum value for each fixed time from the output data of the unit 9 and calculating the ratio between them, and the intensity of the data of the digital conversion unit 6 and the digital conversion unit 9 for each frequency band Intensity calculation function 12 for each frequency band to be calculated, intensity for each frequency band calculated by the intensity calculation function 12 for each frequency band, and intensity for each frequency band from which vibration noise components are removed by the ratio coefficient calculated by the coefficient calculation function 11 The frequency band-specific intensity correction function 13 to be calculated and the frequency corrected by the frequency band-specific intensity correction function 13 It is vibration resistant electroencephalograph, characterized in consisting of inverse transform calculation function 14 for returning the band-by-band intensity on the waveform data.

The invention described in claim 3 is the vibration-resistant electroencephalograph according to claim 1 having an impedance check function.

  In the electroencephalogram measurement, a vibration-resistant electroencephalograph that can be measured without mixing vibration noise even when the user is in an operating state and can be easily worn has been realized.

  Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram showing the appearance and configuration of the present invention. First, the user attaches the annular frame 1 to the head. At that time, an electrode projecting and arranged on the inner wall of the frame 1 is applied to the back of the head, and two electrodes arranged on the frame 1 via a bendable conductor are attached to both ears. In this description, the number of electrodes is three in total, one occipital electrode and two earlobe electrodes, and one of the earlobe electrodes is a reference electrode. However, the electrode protrudes from the inner wall of frame 1 or is folded into frame 1. The number of electrodes arranged via a bendable conductive wire may be larger than this.

  The electroencephalograph of the present invention first detects the potential difference between the occipital electrode and the reference electrode and the potential difference between the earlobe electrode and the reference electrode, and subtracts them. This is a general technique called differential amplification for the purpose of removing noise by subtracting both when noise is mixed in the same amount.

  Next, the electroencephalograph of the present invention passes the filter unit 4 so as to further remove signals other than approximately 30 Hz from approximately 4 Hz of the electroencephalogram signal after differential amplification. This is because the electroencephalogram is an AC wave of 4 to 30 hertz, so that signals in other frequency bands are removed as noise.

  Next, the electroencephalograph of the present invention amplifies the electroencephalogram signal that has passed through the filter 4 from about 5,000 times to about 10,000 times by the amplifying unit 5. This is because the electroencephalogram is a weak signal of several microvolts, and it is impossible to perform later calculations as it is.

  Next, the electroencephalograph of the present invention digitally converts the analog signal amplified by the amplifier 5 by the digital converter 6.

  On the other hand, the vibration signal detected by the vibration sensor 3 which is arranged close to the earlobe electrode which is not the reference electrode or which is fixed to the frame 1 is removed by the filter unit 7 from about 4 Hz to other than about 30 Hz. The signal that has passed through the filter unit 7 is amplified by the amplification unit 8 and then converted into a digital signal by the digital conversion unit 9.

  The electroencephalogram data output from the digital conversion unit 6 and the vibration data output from the digital conversion unit 9 are respectively determined by a ratio coefficient calculation function 11 to obtain a difference between a maximum value and a minimum value for a predetermined time, for example, 1 second, and to determine a ratio between the two. calculate. This is because it is used as a ratio coefficient for aligning the intensity difference between the detected electroencephalogram signal and the vibration signal.

  Further, the data of the digital conversion unit 6 and the digital conversion unit 9 are converted into the intensity for each frequency band by the intensity calculation function 12 for each frequency band. This may be a fast Fourier transform operation or a fast wavelet transform.

  Next, the electroencephalogram data and the vibration data converted into the intensity for each frequency band by the intensity calculation function 12 for each frequency band are subjected to the following calculation by the intensity correction function 13 for each frequency band together with the ratio coefficient calculated by the coefficient calculation function 11. Is made. “(Intensity by frequency band in brain wave data−Intensity by frequency band in vibration data × ratio coefficient) / Intensity by frequency band in brain wave data” That is, for each frequency band, the product of vibration intensity multiplied by a proportional coefficient Subtract from the EEG intensity and divide the result by EEG intensity. Therefore, the calculation result is calculated for each frequency band.

  The frequency band intensity correction function 13 multiplies the value calculated in 0024 for each frequency band corresponding to the real part and the imaginary part calculated in the process of intensity calculation for each frequency band in the electroencephalogram data.

  Next, the inverse transformation calculation function 14 inversely transforms the value calculated in 0025 into waveform data.

  FIG. 2 is a diagram illustrating the transmission directions of the electroencephalogram signal and the vibration signal, and FIG. 3 is a diagram illustrating the internal configuration of the calculation unit.

  Embodiments of the present invention will be described below in detail with reference to the drawings. Usually, in electroencephalogram measurement, a plate-shaped electrode covered with a silver chloride film with a diameter of about 1 cm is attached to the head with a conductive paste, and the electrical signal detected by the electrode is differentially amplified by a technique called differential amplification. After the processing, a method of amplifying the signal from several thousand to 10,000 times and extracting the electroencephalogram signal by removing the noise through the noise filter is adopted.

  The arrangement method of the electrode in that case is as follows. First, an electrode is affixed as a reference electrode to one part of the body where no brain waves such as earlobe are generated. Next, attach electrodes to one part of the head that does not generate brain waves, such as the other earlobe, and to multiple parts of the head that generate brain waves in the back of the head, the top of the head, the temporal region, and the front of the head. The potential difference between the electrode and the reference electrode is measured.

  A difference processing method called differential amplification performed as the next step will be described. Difference processing in electroencephalogram measurement refers to noise reduction by subtracting the potential difference between the earlobe from the potential difference between the reference electrode and the occipital region, between the reference electrode and the top of the head, between the reference electrode and the temporal region, or between the reference electrode and the frontal region. It is to remove. The reason why noise removal is possible with this method is that the potential difference between the earlobe includes only noise, and the other potential differences include noise and brain waves. Only remains. Including the process of amplifying the differentially processed signal, this is called differential amplification, which is a method used for almost all electroencephalogram measurements. (Hereinafter, an example in which a reference electrode is disposed on one earlobe and electrodes are disposed on the other earlobe and the back of the head)

  After this differential amplification is performed, the measured signal is applied to an analog filter or a digital filter to extract the brain wave to be measured in order to remove frequency components other than the frequency band of 4 Hz to 30 Hz, which is usually the brain wave frequency band.

  However, the above-described differential amplification method is effective only when the impedance between the binaural ridges and the other parts is the same, and noise remains if there is an impedance difference. That is, a calculation formula is established: (measured value by differential amplification) = (impedance between reference electrode and occipital region × reference potential between reference electrode and occipital region) − (impedance between both ears × potential difference between both ears).

  Therefore, in the medical field and the like, both impedances are lowered by rubbing the scalp by rubbing the scalp and the earlobe with gauze before attaching the electrodes. When this process is performed, the impedance drops to about 5 kilohms, but when this is not performed, it is not uncommon for the impedance to be 1 megaohm or more. However, this process is very time-consuming.

  Then, what happens if this keratin is not removed, for example, if the impedance between the reference electrode and the back of the head is 1 megaohm and the impedance between the earlobe is 5 kiloohms, both are 5 kiloohms? Compared with the noise, the noise is 200 times larger. However, in general electroencephalogram measurement, since a noise filter is usually used, there is not much problem when the user is stationary. This is because in the stationary state, it is rare that noise is mixed from 4 Hz to 30 Hz which is the frequency band of the electroencephalogram. (Although hum noise of 50 Hz to 60 Hz mixed from commercial power supply is maximized, it can be removed by a filter because it is out of the frequency band of brain waves.)

  However, when the user is in an operating state and the electrode is vibrating, vibration noise is mixed in the frequency band of the electroencephalogram. FIG. 4 shows a waveform when vibration noise is mixed by having the subject shake his / her head during EEG measurement. Since the intensity of this vibration noise increases in proportion to the impedance difference in differential amplification, it was a general perception that the electroencephalograph electrode cannot be easily worn and cannot be measured by a moving person. . .

  In order to remove this vibration noise, first, two differential amplifier circuits were prepared, and a verification experiment was conducted to determine whether noise removal was possible by subtracting each of them. That is, the calculation formula is as follows. (Reference electrode and earlobe potential difference-Binaural potential difference)-(Reference electrode and parietal potential difference-Reference electrode and temporal potential difference) If two differential amplifiers contain vibration noise of the same strength For example, vibration noise in the output value should be reduced.

  During this verification experiment, an electroencephalograph with the following configuration was created. First, in order to measure a weak signal such as an electroencephalogram, it is most preferable to use a circuit called an instrumentation amplifier as a differential amplifier circuit. In this experiment, an IC in which this instrumentation amplifier is made into one chip is used. did. In addition, an 8th-order Butterworth low-pass filter and a CR high-pass filter were created and combined using an operational amplifier, a resistor and a capacitor as the filter unit. As the amplifying unit, a non-inverting amplifier including an operational amplifier, a resistor, and a variable resistor was used. As the digital converter, one of a plurality of analog / digital converters of the microcomputer was used.

  For output, data was transmitted to a personal computer via an RS232C cable by serial communication. The personal computer can display the received data as an image with dedicated software.

  The order in which the signals pass was the order of the differential amplifier circuit, the amplifier unit, the filter unit, the digital conversion unit, and the personal computer. This is because a commercial power supply is used as a circuit power supply, so that it is inevitable that hum noise of 50 Hz to 60 Hz is mixed, and it is effective to apply a filter after amplification.

  However, as a result of the verification experiment of 0036, vibration noise was not reduced. It was found that this was because the vibration intensity was different depending on the part of the head. That is, when moving the head, it is estimated that the vibration noise is small when the head is close to the axis of rotation, and the vibration noise increases as the head is moved away from the axis.

  Therefore, a single differential amplifier circuit was used, and a brain electrode was measured when vertical and horizontal swings of the neck were applied to 7 subjects, with a plate electrode attached to the back of the head and the top of the head near the rotation axis of the head. . That is, the calculation formula is as follows (reference electrode and occipital potential difference−reference electrode and parietal potential difference).

  As a result of this experiment, an effect of reducing vibration noise appeared in all seven subjects. (However, the lateral noise was larger than the vertical swing of the neck.) This proved that the vibration noise increased in proportion to the distance from the rotation axis of the head.

  Since the rotation axis of the head varies depending on the movement of the head (vertical swing of the neck, lateral swing, jump movement, etc.), vibration noise cancellation is impossible with the simple differential amplification method as described above, and it matches the movement. It was found that it is necessary to perform differential processing after multiplying one potential difference value by a coefficient that fluctuates.

  However, for example, if one of the two differential amplification values of the 0036 is multiplied by a coefficient, the electroencephalogram signal is distorted, so that one coefficient in the one differential amplification circuit of the 0041 has a coefficient. If it is multiplied, noise will also be multiplied.

  Therefore, a vibration sensor was placed close to the electrode, and the vibration sensor signal was multiplied by a coefficient to subtract from the electroencephalogram signal. This is because since the signal from the vibration sensor does not include an electroencephalogram, the electroencephalogram signal should not be distorted even if this signal is multiplied by a coefficient and subtracted from the electroencephalogram signal.

  In order to perform the verification experiment of 0045, first, an annular frame 1 for fixing electrodes and vibration sensors is created, and an electrode having a hook shape so that the electrode hits the scalp of the back of the head is formed as shown in FIG. Protrusively arranged on the inner wall. This is because the alpha wave, which is an important component of the electroencephalogram, appears the largest from the back of the head, so that the electrode is simply grounded to the scalp of the back of the head without being blocked by the hair.

  Next, two electrodes having electrodes attached to the clip opening are prepared, and one end of a bendable conductor is connected to each of the two electrodes, and the other end of the conductor is connected to FIGS. As shown in FIG. One of the electrodes was used as a reference electrode, and the other electrode and the occipital electrode were used as normal electrodes.

  In order to facilitate the vibration experiment in this experiment, the electroencephalograph was improved to a portable type, and a battery was used as a power source. The configuration of the electronic circuit is the same as that of the above-mentioned 0037, but the order in which the signals pass is the order of the differential amplifier circuit, the filter unit, the amplification unit, the digital conversion unit, and the personal computer by replacing the amplification unit and the filter unit. This is because the use of a battery as a power source reduces the amount of hum noise, and the allowable range for the magnitude of the measurement signal increases when the filter unit is placed in front of the amplifier unit. (If the order is amplifying section and filter, the allowable range of the filter may be exceeded if the noise is large.)

  The output was serial communication as in the case of 0038, but a weak radio was used for data transmission to the personal computer. This is in order to be able to carry this portable electroencephalograph. In addition, the personal computer side can receive data with a weak wireless device, and the image can be displayed with dedicated software as described above.

  Next, a triaxial acceleration sensor was attached near the occipital electrode of the annular frame 1 and the three output terminals were combined and connected to the input terminal of the electroencephalograph. Since the signal from the acceleration sensor is much larger than the electroencephalogram, the gain was adjusted by the amplifier. As a result, a signal with the same shape as the vibration noise waveform in the electroencephalograph could be measured.

  Since the result of 0050 was confirmed, the electroencephalograph was improved so that both an electroencephalogram signal and a vibration signal could be measured simultaneously. In other words, an electroencephalograph differential amplifier circuit, a filter unit, and an amplifier unit are added, and the signals from these signals enter another one of the multiple analog / digital converters of the microcomputer. Improved to be able to measure. In addition, the variable amplifier can be used to manually change the amplification factor by the amplifier.

  Using these devices, both EEG signals and vibration signals were measured simultaneously. That is, the above-mentioned annular frame 1 was attached to the head, the occipital electrode was brought into contact with the scalp of the occipital region, two earlobe clips were sandwiched between both ears, and the electroencephalogram signal and vibration signal were measured simultaneously when the head was moved.

  As a result of this experiment, since the output timing and shape of the noise signal of both the brain wave signal and the vibration signal were similar when moving the head, it is possible to create a program that simply subtracts the noise signal and remove the noise. I verified it. However, as a result of this verification, it was found that noise removal cannot be achieved by simply subtracting both. This proved to be due to a slight shift in the timing and shape of both signals.

  Therefore, we considered that both the EEG signal and the vibration signal were subjected to fast Fourier transform processing, and the result obtained by subtracting the intensity for each frequency band was returned to the waveform signal. This was programmed, and a verification experiment was performed again. As a result, even if the intensity for each frequency band was subtracted, it was not possible to return to a real-time waveform. This is because the phase information of the waveform due to time is lost in the calculation result of the fast Fourier transform.

  Therefore, the next calculation result was multiplied for each frequency band corresponding to the real part and the imaginary part calculated by the fast Fourier transform process in the electroencephalogram data. (Intensity by frequency band in EEG data-Intensity by frequency band in vibration data x Ratio coefficient) / Intensity by frequency band in EEG data

  However, the value of (intensity by frequency band in brain wave data-intensity by frequency band in vibration data x ratio coefficient) may be negative, and noise cannot be removed successfully even if this data is converted back to a waveform. Therefore, we decided to convert negative values to zero.

  Also, in the experiments so far, instead of changing the ratio coefficient, the amplifier on the vibration sensor side was manually moved to adjust the gain. However, as a result of analyzing the measurement data, the maximum of the electroencephalogram signal and the vibration signal in a certain time was analyzed. It was found that it was possible to calculate from the ratio of the difference between the value and the minimum value. That is, it was found that the difference between the intensity of the electroencephalogram signal and the vibration signal is proportional to the difference between the maximum value and the minimum value of those signals. Therefore, “(maximum value of electroencephalogram signal−minimum value of electroencephalogram signal) / (maximum value of vibration signal−minimum value of vibration signal) at a fixed time” is used as the ratio coefficient.

  As a result of these improvements and reverse conversion to waveforms, it was possible to output a signal from which vibration noise was removed in almost real time. (1) and (2) in FIG. 5 are diagrams in which the ratio coefficient is calculated from the difference between the maximum value and the minimum value of the electroencephalogram signal and the vibration signal in one second, and the maximum value of the electroencephalogram signal in this case is 74. The minimum value is 48 and the difference is 26. The maximum value of the vibration signal is 104, the minimum value is 75, and the difference is 29. Therefore, the ratio coefficient is 0.897 obtained by dividing 26 by 29. (The unit of the vertical axis in FIGS. 5 (1) and (2) is a voltage value proportional to volts, but not volts.) (3) and (4) in FIG. 5 are the fast Fourier transform (FFT) of the electroencephalogram signal and the vibration signal. FIG. 6 is a diagram in which a fast Fourier transform result of a vibration signal is multiplied by a ratio coefficient.

  (5) in FIG. 6 represents the result of subtracting (4) from (3) in FIG. 5 and converted to zero when the result is negative, and (6) in FIG. 6 represents (3) in FIG. FIG. 6 is a diagram illustrating a result obtained by multiplying a real number and an imaginary number calculated in the process of obtaining (3) / (5) and returning the waveform by inverse Fourier transform.

  Further, (7) in FIG. 6 is an electroencephalogram signal measured with the head stationary, and (8) in FIG. 6 is a diagram showing a result of fast Fourier transform of the signal in (7). Comparing (6) and (7) in Fig. 6, it can be seen that the EEG waveform with the vibration noise removed and the EEG waveform without vibration are almost the same shape. Compare (5) and (8). Then, it can be seen that the intensity for each frequency of the electroencephalogram from which vibration noise has been removed and the intensity for each frequency of the electroencephalogram measured without applying vibration are substantially the same for the intensity for each frequency.

  In the electroencephalogram measurement, the peak value of the intensity of 8 to 13 Hz, which is the frequency band of the alpha wave, is particularly important. When comparing (6) and (7) in Fig. 6, the peak value of the intensity of the alpha wave Can be seen to be almost the same amount. From this, it can be understood that the vibration noise mixed in from 4 Hz to 30 Hz, which is the frequency band of the electroencephalogram, can be accurately removed by the above calculation method.

  In FIG. 7, the upper waveform is an electroencephalogram signal including vibration noise detected by an electroencephalograph electrode, the middle waveform is a vibration signal detected by the vibration sensor, and the lower waveform after the vibration noise is removed by the present invention. It is brain wave vibration.

  In the experiment related to these vibrations, the phenomenon that the occipital electrode is detached from the scalp frequently occurs due to the vibration of the frame 1 attached to the head. Therefore, a function of checking the impedance of the electrode at a constant cycle should be added. Thought. That is, when the electrode is grounded to the scalp, the impedance is about 300 kilohms, but when it is removed from the scalp, it exceeds 2 megohms, so whether or not the electrode is removed by measuring this impedance. This is a warning to the user.

  Therefore, first, a sine wave of about 10 Hz is weakened to about 10 microamperes, and is passed between the reference electrode and the back of the head and between the reference electrode and the earlobe electrode, and the reaction is detected by the automatic impedance bridge circuit, thereby causing the reference electrode to the back of the head. And the impedance between the reference electrode and the earlobe electrode can be measured.

  The signal detected by the automatic impedance bridge circuit is amplified, filtered, rectified and inputted to the analog / digital converter of the microcomputer, and the microcomputer calculates and outputs this data. In addition, the microcomputer controls the power supply of these impedance measuring units to check the impedance at regular intervals. As the filter processing, a digital filter may be used after digital conversion, or the impedance may be calculated by calculation without rectifying the signal.

  As a result of creating a vibration-resistant electroencephalograph with such a configuration and conducting a measurement experiment, an electroencephalograph that can perform stable measurement with little vibration noise mixing even when the head is shaken and that can be easily worn has been realized.

It is a figure which shows the external appearance and structure of the vibration-proof electroencephalograph which concerns on embodiment of this invention. It is a figure which shows the transmission direction of each electroencephalogram signal and vibration signal of the vibration-proof electroencephalograph which concerns on embodiment of this invention. It is a figure showing the internal structure of the calculating part of the vibration-resistant electroencephalograph which concerns on embodiment of this invention. It is a figure explaining the waveform when vibration noise mixes at the time of electroencephalogram measurement. It is a figure explaining the computing equation for removing the vibration noise of the vibration-resistant electroencephalograph which concerns on embodiment of this invention. It is the figure explaining the computing equation for removing the vibration noise of the vibration-proof electroencephalograph which concerns on embodiment of this invention, and its effect. It is a figure which shows the effect of the vibration-proof electroencephalograph which concerns on embodiment of this invention.

DESCRIPTION OF SYMBOLS 1 Frame 2 Electrode 3 Vibration sensor 4 Filter part 5 Amplification part 6 Digital conversion part 7 Filter part 8 Amplification part 9 Digital conversion part 10 Calculation part 11 Ratio coefficient calculation function 12 Strength calculation function according to frequency band 13 Strength correction function according to frequency band 14 Inverse transformation operation function

Claims (2)

A filter unit 4 for removing an electrical noise component from a signal detected by a frame 1, a plurality of electrodes 2 arranged in the frame 1, a vibration sensor 3 arranged in the frame 1, a signal detected by the electrode 2, and the filter unit 4 An amplifying unit 5 for amplifying the passed signal, a digital converting unit 6 for converting the analog signal amplified by the amplifying unit 5 into a digital signal, and a filter unit 7 for removing an electrical noise component from the signal detected by the vibration sensor 3 And an amplifier 8 that amplifies the signal that has passed through the filter unit 7, a digital converter 9 that converts an analog signal amplified by the amplifier 8 into a digital signal, and outputs of the digital converter 9 and the digital converter 6 The calculation unit 10 performs an operation for removing the vibration noise component from the value. The calculation unit 10 outputs the output data of the digital conversion unit 6 and the digital conversion unit 9 to each other. Calculated by the ratio coefficient calculation function 11 for calculating the comparison condition, the frequency conversion intensity calculation function 12 for calculating the intensity of the data of the digital converter 6 and the digital conversion section 9 for each frequency band, and the intensity calculation function 12 for each frequency band. The frequency band strength correction function 13 for calculating the frequency band strength obtained by removing the vibration noise component by the frequency band strength and the ratio coefficient calculated by the coefficient calculation function 11 and the frequency band strength correction function 13 An anti-vibration electroencephalograph comprising an inverse transformation calculation function 14 for returning the intensity for each frequency band to waveform data. The ratio coefficient calculating function (11) is a function for calculating the ratio between the maximum value and the minimum value at a predetermined time from the output data of the digital conversion unit (6) and the digital conversion unit (9). Vibration resistance electroencephalograph.

Images