JP2013121489A - Sleep stage detection device, sleep stage calculation device, and sleep stage detection system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a sleep stage detection device, a sleep stage calculation device, and a sleep stage detection system, which reduce influences caused due to individual difference between subjects, and further precisely detect the sleep stage.SOLUTION: Brain wave data detected by three electrodes 11 to 13 are Fourier transformed to obtain spectral analysis data, calculation represented by X=δ/(δ+θ+α+β+emg) is executed based on δ, θ, α, and β waves included in the spectral analysis data and frontalis electromyogram (emg) to obtain the content rate X of the δ wave. The content rate X and first to third thresholds TH1, TH2, and TH3 are compared with each other, thereby determining which stage the subject is in, REM sleep stage, light non-REM sleep stage, or deep non-REM sleep stage. Since data of the frontalis electromyogram is used to obtain the content rate X, highly precise detection of the sleep stage is allowed to be detected reflecting the sleep stage of the subject.

Description

  The present invention relates to a sleep state detection device that detects a sleep state of a subject, a sleep state calculation device that calculates a sleep state of a subject by calculation, and a sleep state detection system.

  In general, human sleep is classified into REM (Rapid eye movement) sleep and non-REM sleep according to the depth of sleep, and non-REM sleep is classified into shallow non-REM sleep and deep non-REM sleep. Yes. Measuring the depth of such sleep is an important factor in managing the physical condition of the subject.

  As an apparatus for detecting the depth of sleep, a device described in, for example, Japanese Patent Application Laid-Open No. 2009-112402 (Patent Document 1) has been known. In Patent Document 1, an electroencephalogram detection terminal is attached in the vicinity of the subject's head, the electroencephalogram of the test subject is detected using the electroencephalogram detection terminal, and a δ wave (frequency 0.5 to 4.0 Hz) included in the detected electroencephalogram. ), Α wave (frequency 8.0 to 12.0 Hz), σ wave (frequency 12.0 to 16.0 Hz), β wave (frequency 13.0 to 40.0 Hz) content, and content rate It is disclosed to determine whether a subject's sleep state, that is, whether the subject's sleep is REM sleep, shallow non-REM sleep, or deep non-REM sleep.

JP 2009-112402 A

  However, in the conventional example disclosed in Patent Document 1 described above, the brain wave of the subject is detected, and the sleep state using the content and content of δ wave, α wave, σ wave, and β wave included in the brain wave Therefore, there is a problem that it is difficult to uniquely detect a sleep state for many subjects. In addition, since the electromyogram data of the subject is not used, there is a problem that the detection accuracy of the sleep state is lowered.

  The present invention has been made to solve such a conventional problem, and the object of the present invention is to reduce the influence of individual differences among subjects and to detect a sleep state with higher accuracy. It is in providing the possible sleep state detection apparatus, the sleep state calculation apparatus, and the sleep state detection system.

  In order to achieve the above object, the sleep state detection device according to claim 1 of the present application is a sleep state detection device that detects a sleep state of a subject, and at least three electrodes provided on the forehead portion of the subject; A passive high-pass filter that removes a desired low-frequency component from the electroencephalogram detected by the electrode, and an active type that removes a frequency component higher than a predetermined frequency corresponding to an electromyogram from the output signal of the high-pass filter A low-pass filter, spectral analysis means for spectrally analyzing a signal that has passed through the low-pass filter, and storage medium input / output means for writing spectral analysis data output from the spectral analysis means to a portable storage medium. It is characterized by that.

  The invention according to claim 2 further includes original waveform output means for outputting a signal that has passed through the low-pass filter as original waveform data, and the storage medium input / output means includes the original waveform data in addition to the spectrum analysis data. Is written in the storage medium.

  The invention according to claim 3 is that the three electrodes are adhesive conductive gel electrodes that can be attached to the forehead portion of the subject, and each electrode is attached to the forehead portion using an adhesive insulating gel sheet. It is characterized by sticking to.

  The invention according to claim 4 is further provided with a pulse wave detecting means provided on the forehead portion for detecting the pulse wave of the subject, and the storage medium input / output means is connected to the portable storage medium with the pulse wave. The pulse wave data detected by the wave detection means is written.

  The invention according to claim 5 is further provided with acceleration detection means provided on the forehead portion for detecting the orientation of the subject's head, and the storage medium input means is configured to detect the acceleration on the portable storage medium. The acceleration data detected by the means is written.

  The sleep state calculation device according to claim 6 is a sleep state calculation device that calculates the sleep state of the subject based on the spectrum analysis data obtained based on the brain wave of the subject, the memory storing the spectrum analysis data. Based on data reading means for reading storage data of the medium, and δ wave, θ wave, α wave, β wave, and forehead electromyogram data included in the spectrum analysis data, the δ wave level is changed to δ, θ wave level. Is θ, the α wave level is α, the β wave level is β, and the forehead electromyogram level is emg, the calculation of “X = δ / (δ + θ + α + β + emg)” is performed to calculate the content rate X of the δ wave. Based on the calculation means to be obtained and the content rate X obtained by the calculation means, it is detected whether the sleep state of the subject is awakening, REM sleep, shallow non-REM sleep, or deep non-REM sleep sleep Characterized in that it comprises a state detecting means.

  In the invention according to claim 7, the data reading means reads the original waveform data of the electroencephalogram stored in the storage medium in addition to the spectrum analysis data, and based on the original waveform data, An electrooculogram reading unit for reading a figure is further provided, and the sleep state detection unit is configured to determine whether the sleep state of the subject is awake based on the electrooculogram read by the electrooculogram reading unit in addition to the content rate X. , REM sleep, shallow non-REM sleep and deep non-REM sleep.

  In the invention according to claim 8, in addition to the spectrum analysis data, at least one of the pulse wave data of the subject and the acceleration data indicating the orientation of the subject's head is stored in the storage medium. In addition to the spectrum analysis data, at least one of the pulse wave data of the subject stored in the storage medium and the acceleration data indicating the orientation of the subject's head is read, and the sleep state detecting means reads the spectrum analysis data In addition, detecting whether the subject's sleep state is awakening, REM sleep, shallow non-REM sleep, or deep non-REM sleep using at least one of the pulse wave data and the acceleration data Features.

  The invention according to claim 9 further includes a moving average processing means for performing a moving average process on the spectrum analysis data read by the data reading means.

  The sleep state detection system according to claim 10, wherein the sleep state detection system detects a sleep state of a subject, and includes at least three electrodes provided in a forehead portion of the subject and an electroencephalogram detected by the electrodes. A passive high-pass filter that removes a desired low-frequency component, an active low-pass filter that removes a desired high-frequency component from the output signal of the high-pass filter, and a spectrum that performs spectrum analysis on the signal that has passed through the low-pass filter Based on the data of the δ wave, θ wave, α wave, β wave, and forehead electromyogram included in the analyzing means and the spectrum analysis data output from the spectrum analyzing means, the δ wave level is changed to δ, θ wave level. Where X is α, α wave level is α, β wave level is β, and electromyogram level is emg, “X = δ / (δ + θ + α + β + emg)” The calculation means for performing the calculation to obtain the content rate X of δ wave, and based on the content rate X obtained by the calculation means, the sleep state of the subject is awakening, REM sleep, shallow non-REM sleep, and deep non-REM And a sleep state detecting means for detecting which one of sleeps.

  According to an eleventh aspect of the present invention, there is provided an original waveform output unit that outputs a signal that has passed through the low-pass filter as original waveform data, and an electrooculogram reading unit that reads an electrooculogram based on the original waveform data. In addition to the content rate X, the sleep state detection unit may be configured such that the sleep state of the subject is awake, REM sleep, shallow non-REM sleep, based on an electrocardiogram read by the electrooculogram read unit. , Deep non-REM sleep is detected.

  The invention according to claim 12 is further provided with a pulse wave detection means provided on the forehead portion for detecting the pulse wave of the subject, wherein the sleep state detection means is a pulse detected by the pulse wave detection means. Based on the wave data, it is detected whether the sleep state of the subject is awakening, REM sleep, shallow non-REM sleep, or deep non-REM sleep.

  The invention according to claim 13 is further provided with acceleration detection means provided on the forehead portion for detecting the orientation of the subject's head, wherein the sleep state detection means is acceleration data detected by the acceleration detection means. The sleep state of the subject is detected from awakening, REM sleep, shallow non-REM sleep, and deep non-REM sleep.

  The invention described in claim 14 further includes moving average processing means for performing moving average processing on the spectrum analysis data output from the spectrum analysis means.

  In the sleep state detection apparatus according to claim 1 of the present invention, δ waves, θ waves, α waves, from an electroencephalogram data detected by three electrodes using an active low pass filter and a passive high pass filter, Since the waveform data of β wave and electromyogram Emg is acquired and spectrum analysis data is acquired by Fourier transforming this waveform data, the electroencephalogram data used for detecting the sleep state of the subject can be measured with high accuracy. It becomes possible.

  In the invention of claim 2, since the original waveform data of the electroencephalogram data processed by each filter is acquired and stored in the storage medium, a highly accurate sleep state using the original waveform data as well as the spectrum analysis data is acquired. Detection can be performed.

  In the invention of claim 3, since an adhesive conductive gel electrode is used as an electrode to be attached to the subject and an adhesive insulating gel sheet is applied to the subject's forehead, it can be easily attached and removed easily. It becomes possible.

  In the invention of claim 4, since the pulse wave data of the subject is detected by the pulse wave detection means and the detected pulse wave data is stored in the storage medium, the sleep state can be detected with high accuracy using the pulse wave data. Can do.

  In the invention of claim 5, since the direction of the head of the subject is detected by the acceleration detecting means and the detected acceleration data is stored in the storage medium, it is possible to detect the sleep state with high accuracy using the acceleration data. .

  In the sleep state computing device according to claim 6, X = δ / (δ + θ + α + β + emg) based on the data of δ wave, θ wave, α wave, β wave, and forehead electromyogram (egg) included in the spectrum analysis data. The δ-wave content rate X is obtained by performing the above calculation, and the sleep state of the subject is detected based on the content rate X. Therefore, the sleep state is more accurate than when the conventional electromyogram is not used. Detection is possible.

  According to the seventh aspect of the present invention, an electrooculogram is read based on the original waveform data, and REM sleep or non-REM sleep is detected based on the electrooculogram. Therefore, a more accurate sleep state can be detected.

  In the invention of claim 8, since at least one of pulse wave data and acceleration data is read in addition to spectrum analysis data, and the sleep state of the subject is detected using these data, the sleep state can be detected with higher accuracy. Can be detected.

  In the invention of claim 9, since the spectrum analysis data is subjected to moving average processing, the entire spectrum analysis data can be smoothed, and a highly accurate sleep state can be detected while avoiding frequent data fluctuations.

  In the sleep state detection system according to claim 10, brain wave data is detected by an electrode attached to a subject, and further, the brain wave data is Fourier transformed to obtain spectrum analysis data, and δ wave, θ included in the spectrum analysis data Based on the data of wave, α wave, β wave, and forehead electromyogram (egg), calculation of X = δ / (δ + θ + α + β + egg) is performed to obtain the content rate X of δ wave, and based on this content rate X Since the sleep state of the subject is detected, the sleep state can be detected with higher accuracy than in the case where a conventional electromyogram is not used.

  According to the eleventh aspect of the invention, an electrooculogram is read based on the original waveform data of the electroencephalogram detected by the electrode, and REM sleep or non-REM sleep is detected based on the electrooculogram. The state can be detected.

  In the twelfth aspect of the invention, the pulse wave detection means detects the pulse wave data of the subject, and the sleep state of the subject is detected based on the pulse wave data. Therefore, the sleep state can be detected with higher accuracy.

  In the invention of claim 13, since the acceleration data of the subject is detected by the acceleration detecting means and the sleep state of the subject is detected based on the acceleration data, the sleep state can be detected with higher accuracy.

  In the invention of claim 14, since the spectrum analysis data is subjected to moving average processing, the entire spectrum analysis data can be smoothed, and a highly accurate sleep state can be detected while avoiding frequent data fluctuations.

It is a block diagram which shows the structure of the sleep state detection system which concerns on 1st Embodiment of this invention. It is a characteristic view which shows the frequency characteristic of the filter provided in the sleep detection system which concerns on 1st Embodiment of this invention. It is explanatory drawing which shows a mode that the electrode provided in the sleep state detection system which concerns on 1st Embodiment of this invention is attached to a test subject's forehead part. It is a characteristic view which shows the relationship between the change of the content rate X of (delta) wave calculated | required with the sleep state detection system which concerns on 1st Embodiment of this invention, and the sleep state calculated | required based on this content rate X. It is a characteristic view which shows the change of (delta) wave measured with respect to the two test subjects P1 and P2. It is a characteristic view which shows the change of the content rate of (delta) wave of two test subjects P1, P2 measured using the sleep state detection system which concerns on 1st Embodiment of this invention. It is a characteristic view which shows the change of the sleep state of the two test subjects P1 and P2 calculated | required based on the content rate X shown in FIG. It is explanatory drawing which shows the spectrum analysis data calculated | required with the sleep state detection system which concerns on 1st Embodiment of this invention, and the spectrum analysis data of this 24 times moving average and 60 times moving average. It is explanatory drawing which shows the comparison with the content rate X of the (delta) wave calculated | required using the electromyogram emg, and the content rate Y of the (delta) wave calculated | required without using the electromyogram emg. It is explanatory drawing which shows the electromyogram of a test subject's genital muscle (muscle near jaw), and the electromyogram detected by a forehead part. It is a block diagram which shows the structure of the sleep state detection system which concerns on 2nd Embodiment of this invention. The electrocardiogram in the REM sleep period and the direct current retinal electrostatic potential generated before and after the eyeball detected by the forehead part are changed according to the movement of the eyeball and relate to the second embodiment of the present invention. It is explanatory drawing. It is explanatory drawing which concerns on 2nd Embodiment of this invention and shows the original waveform data in a non-REM sleep period, and the original waveform data in a REM sleep period. It is a block diagram which shows the structure of the sleep state detection apparatus of the sleep state detection system which concerns on 3rd Embodiment of this invention. It is a block diagram which shows the structure of the sleep state calculating apparatus of the sleep state detection system which concerns on 3rd Embodiment of this invention. It is explanatory drawing which shows a mode that the electrode provided in the sleep state detection system which concerns on 3rd Embodiment of this invention, a pulse-wave detection part, and a triaxial acceleration sensor are attached to a test subject's forehead part. It is a characteristic view which shows the sleep state detected in 1st Example of the sleep state detection system which concerns on 3rd Embodiment of this invention. It is a characteristic view which shows the pulse wave data and acceleration data which are detected by 1st Example of the sleep state detection system which concerns on 3rd Embodiment of this invention. It is a characteristic view which shows the sleep state detected in 2nd Example of the sleep state detection system which concerns on 3rd Embodiment of this invention. It is a characteristic view which shows the pulse wave data and acceleration data which are detected by 2nd Example of the sleep state detection system which concerns on 3rd Embodiment of this invention. It is a characteristic view which shows the sleep state detected in 3rd Example of the sleep state detection system which concerns on 3rd Embodiment of this invention. It is a characteristic view which shows the pulse wave data and acceleration data which are detected by 3rd Example of the sleep state detection system which concerns on 3rd Embodiment of this invention. It is a characteristic view which shows the sleep state detected in the 1st-3rd Example of the sleep state detection system which concerns on 3rd Embodiment of this invention. It is a table | surface which shows the sleep state detected in 1st Example of the sleep state detection system which concerns on 3rd Embodiment of this invention. It is a table | surface which shows the sleep state detected in 2nd Example of the sleep state detection system which concerns on 3rd Embodiment of this invention. It is a table | surface which shows the sleep state detected in 3rd Example of the sleep state detection system which concerns on 3rd Embodiment of this invention. It is explanatory drawing which shows the respiratory data of the sleep state detection system which concerns on 4th Embodiment of this invention. It is explanatory drawing which shows the fluctuation | variation of the sympathetic nerve and the parasympathetic nerve of the sleep state detection system which concerns on embodiment of this invention. It is a characteristic view which shows the change of the oxygen saturation when a respiratory stop occurs, and the change of a pulse rate of the sleep state detection system which concerns on embodiment of this invention. It is explanatory drawing which shows the acceleration data detected with the triaxial acceleration sensor used with the sleep state detection system which concerns on 3rd Embodiment of this invention.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

[Description of First Embodiment]
FIG. 1 is a block diagram showing the configuration of the sleep state detection system according to the first embodiment of the present invention, and FIG. 2 is an explanation showing frequency characteristics of the notch filter 17, high-pass filter 18, and low-pass filter 19 shown in FIG. FIGS. 3 and 3 are explanatory views showing a state in which the first to third electrodes 11, 12 and 13 are attached to the forehead portion of the subject.

  As shown in FIG. 1, the sleep state detection system 101 obtains a sleep state of the subject based on the sleep state detection device 102 that detects the sleep state of the subject and the electroencephalogram data detected by the sleep state detection device 102. The sleep state calculation device 103 is configured.

  The sleep state detection device 102 includes three electrodes: a first electrode 11 serving as a reference electrode, a second electrode 12 serving as an active electrode, and a third electrode (neutral point electrode) 13 serving as a so-called body ground. I have. And each electrode 11, 12, 13 is arrange | positioned along with the horizontal direction in the test subject's forehead part. That is, as shown in FIGS. 3A and 3B, the third electrode 13 is attached to the center of the forehead portion of the subject P, and the first electrode 11 and the second electrode 12 are attached to the left and right, respectively. And each electrode 11,12,13 can be comprised, for example with an adhesive conductive gel, By setting it as such a structure, it can affix on a test subject's forehead part easily. Further, an adhesive insulating gel sheet 25 is attached so as to cover each electrode 11, 12, 13. Thereby, it can prevent that each electrode 11, 12, 13 remove | deviates from a test subject's forehead part, and it becomes possible to remove easily after use.

  As shown in FIG. 1, the first electrode 11 is connected to the plus terminal of the buffer amplifier 14, and the output terminal of the buffer amplifier 14 is connected to the plus terminal of the differential amplifier 16. On the other hand, the second electrode 12 is connected to the plus terminal of the buffer amplifier 15, and the output terminal of the buffer amplifier 15 is connected to the minus terminal of the differential amplifier 16. Further, the third electrode 13 serving as a body ground is connected to the ground terminal of the differential amplifier 16.

  The subject's brain waves are extracted by bipolar derivation (derivation method for obtaining a difference in potential change between the electrodes) using the first electrode 11 as a reference electrode and the second electrode 12 as an active electrode.

  On the output side of the differential amplifier 16, a notch filter 17, a passive high-pass filter 18, an active low-pass filter 19, an amplification unit 20, and a spectrum analysis processing unit 21 are provided.

  The notch filter 17 is configured as a band stop filter having a high Q value, and is a filter that removes a component of a commercial frequency band (50 Hz or 60 Hz) included in a signal output from the differential amplifier 16.

  The high-pass filter 18 is a passive type, and has a characteristic that the cut-off characteristic gradually decreases with respect to a frequency lower than the frequency 1 [Hz] as shown in a band F1 in FIG. Therefore, a large amount of brain waves in a frequency band lower than 1 [Hz] can be prevented. In other words, it is possible to appropriately detect brain waves in a frequency band of less than 1 [Hz] and improve the detection accuracy of the entire brain waves.

  The low-pass filter 19 is an active type, and a cutoff frequency is set to 90 [Hz] as shown in FIG.

  Therefore, each filter 17, 18, and 19 has a cut-off characteristic as shown in FIG. Specifically, the gain gradually decreases from a frequency of 1 [Hz] to 0.5 [Hz], an attenuation region is set to a commercial frequency of 50 [Hz], and a frequency of 90 It has a characteristic in which the vicinity of [Hz] is the upper limit.

  The amplification unit 20 shown in FIG. 1 amplifies the signal that has passed through the low-pass filter 19 and outputs the amplified signal to the spectrum analysis processing unit 21. The spectrum analysis processing unit 21 performs an FFT (Fast Fourier Transform) process on the signal output from the amplification unit 20, and converts the electroencephalogram data into spectrum analysis data for each frequency.

  The storage medium input / output unit 23 has a function of writing spectrum analysis data to the memory card 24 when a memory card 24 (storage medium) that can be carried by the operator is inserted. In the present embodiment, a process of writing the spectrum analysis data output from the spectrum analysis processing unit 21 into the memory card 24 is performed.

  The sleep state calculation device 103 stores a data reading unit 31 that reads data written in the memory card 24, a calculation control unit 32 that performs overall control, and a program related to calculation processing in the calculation control unit 32. And a memory 34 that writes and reads various data when the arithmetic control unit 32 executes arithmetic processing. Furthermore, an input operation unit 35 for receiving various input operations by the operator and a display unit 36 for displaying various data on the screen are provided.

  In addition, the calculation program stored in the calculation program storage unit 33 includes an application program that calculates the sleep state of the subject based on the spectrum analysis data. The sleep state calculation device 103 can be configured as an integrated computer including a central processing unit (CPU), storage means such as a RAM, a ROM, and a hard disk.

  The calculation control unit 32 controls the sleep state calculation device 103 as a whole. In particular, based on the data of δ wave, θ wave, α wave, β wave, and forehead electromyogram (egg) included in the spectrum analysis data, a process for obtaining the δ wave content rate X by an arithmetic expression described later is executed. To do. Furthermore, the process which detects whether the sleep state of the test subject P is REM sleep, shallow non-REM sleep, or deep non-REM sleep based on the calculated content rate is executed. That is, the calculation control unit 32 has a function as a calculation unit, and further has a function as a sleep state detection unit.

  Next, the operation of the sleep state detection system 101 according to the present embodiment configured as described above will be described. As shown in FIG. 3, when the subject goes to sleep with the electrodes 11, 12, 13 attached to the forehead portion of the subject P, the brain waves and myoelectric waves of the subject P are caused by the electrodes 11, 12, 13. The waveform in the figure is detected.

  The detected waveform data passes through the notch filter 17, the high-pass filter 18, and the low-pass filter 19 shown in FIG. 1, whereby the waveform having the frequency characteristic shown in FIG. 2 is acquired. That is, the frequency is in the range of 0.5 [Hz] to 90 [Hz], and among these, in the range of 0.5 [Hz] to 1 [Hz], the gain is gradually attenuated, and further 50 [ [Hz] Waveform data in which a signal in the vicinity is cut is acquired.

  The waveform data that has passed through the low-pass filter 19 is amplified by the amplification unit 20 and then subjected to spectrum analysis processing by the spectrum analysis processing unit 21. Specifically, fast Fourier transform (FFT) processing is executed, and the spectrum of each frequency is detected.

  Then, the spectrum analysis data is written into the memory card 24 by a user operation at the storage medium input / output unit 23 shown in FIG. Thereafter, the memory card 24 to which the data has been written is inserted into the data reading unit 31 of the sleep state calculation device 103 by the user, and the spectrum analysis data written to the memory card 24 is read by the control of the calculation control unit 32. It is done.

  In the present embodiment, among the entire spectrum analysis data, waveform data of 0.5 to 4 [Hz] is δ wave, waveform data of 5 to 7 [Hz] is θ wave, and 8 to 13 [Hz]. The waveform data is an α wave, the waveform data of 14 to 30 [Hz] is a β wave, and the waveform data of 31 to 48 [Hz] and 52 to 90 [Hz] is an electromyogram (egg). When the levels of δ wave, θ wave, α wave, β wave, and electromyogram (amplitude of spectrum analysis data) are indicated by the same symbols δ, θ, α, β, and emg, respectively (1) The content rate X of the δ wave level with respect to the total of the entire waveform level is calculated by the equation.

X = δ / (δ + θ + α + β + egg) (1)
That is, under the control of the calculation control unit 32 shown in FIG. 1, the application program stored in the calculation program storage unit 33 is executed, and the calculation of (1) above is performed based on the spectrum analysis data stored in the memory card 24. Execute. Then, the ratio of the δ wave to the total of each waveform level is obtained, and this is set as the content rate X.

  By the above processing, for example, a characteristic curve as shown by reference numeral S1 in FIG. 4B can be obtained. In addition, curve S1 shown in FIG.4 (b) has shown the moving average of the measurement result of the content rate X for the past 60 times. Details of the moving average process will be described later.

  In the present embodiment, the first threshold value TH1, the second threshold value TH2, and the third threshold value TH3 are set for the content rate X (curve S1) (where TH1 <TH2 <TH3). Further, based on the respective threshold values TH1 to TH3, it is determined that the subject P is awake (in an arousal state) when the content ratio X is less than TH1, and the subject P is determined when it is between TH1 and TH2. Is determined to be in a REM sleep state, and if it is between TH2 and TH3, it is determined that the subject P is in a shallow non-REM sleep state, and if it is greater than TH3, the subject P is in a deep non-REM sleep state. to decide.

  And in the arithmetic control part 32, as shown to Fig.4 (a) by comparing the magnitude | size of each threshold value TH1-TH3 and curve S1, the sleep state (REM sleep, shallow) after the test subject P fell asleep. Non-REM sleep and deep non-REM sleep) can be detected. That is, the subject P falls asleep after about 30 minutes from the start of measurement, and then repeats REM sleep, shallow non-REM sleep, deep non-REM sleep, shallow non-REM sleep, and REM sleep with a period of about 60 minutes. I understand.

  Therefore, for example, when the subject P sleeps at home, the sleep state detecting device 102 shown in FIG. 1 is used to acquire the brain wave and electromyogram of the subject P and further store this data in the memory card 24. By bringing the memory card 24 to the hospital, the sleep state of the subject P can be analyzed by the sleep state calculation device 103 installed in the hospital. Therefore, the doctor in charge can recognize at a glance what sleep state the subject P is asleep in, and can easily diagnose the sleep of the subject P.

  Here, in the present embodiment, as shown in the above-described equation (1), the electromyogram emg is used when calculating the content rate X of the δ wave that is a guideline for determining the sleep state. The reason for using the electromyogram emg will be described.

  FIG. 5 is a characteristic diagram showing the δ wave level of each of the subjects P1 and P2 obtained by measuring the brain waves during sleep on two subjects P1 and P2 and acquired from the spectrum analysis data of the brain waves. a) shows the data of the subject P1, and FIG. 5 (b) shows the data of the subject P2. In the subject P1, the δ wave level has a large amplitude, and in the subject P2, the δ wave level has a small amplitude. That is, it can be said that there are individual differences in the δ wave level.

  Therefore, when the content rate of the δ wave with respect to the whole level of the electroencephalogram is obtained without using the electromyogram emg, the measurement accuracy of the content rate may decrease due to individual differences in the δ wave level. In other words, the content rate of the δ wave calculated without using the electromyogram emg may vary greatly depending on individual differences, and it may be impossible to detect a sleep state with high accuracy.

  On the other hand, when the content rate X of the δ wave is obtained by (1) using the electromyogram emg, since the electromyogram emg is added to the denominator of the arithmetic expression, the influence of the individual difference of the δ wave level is reduced. Then, as shown in FIGS. 6 (a) and 6 (b), the content rate X of the δ wave has substantially the same amplitude in both the subjects P1 and P2. In addition, the sleep state detection results using the first threshold TH1 to the third threshold TH3 are as shown in FIGS. 7A and 7B, and the influence of individual differences is reduced, and the sleep state is detected with high accuracy. Will be able to. That is, in this embodiment, the δ-wave content rate X with respect to the sum of the waveform levels in each frequency band including the electromyogram emg is obtained, and the sleep state is detected using this content rate X. High-precision detection with reduced influence becomes possible.

  Next, the reason for calculating the content rate X using a moving average will be described. FIG. 8A is a characteristic diagram showing spectrum analysis data obtained by Fourier transforming the original waveform data output from the amplifying unit 20, and is an integrated value for 5 seconds of the spectrum analysis data obtained every 5 seconds. Is shown. FIG. 8B shows a case where the moving average processing of the spectrum analysis data (integrated value) for 24 times is performed, that is, the average value of the spectrum analysis data obtained this time and the spectrum analysis data for the past 23 times is obtained this time. The characteristic when it is set as the spectrum analysis data of is shown. Furthermore, FIG.8 (c) has shown the characteristic at the time of implementing the moving average process of the spectrum analysis data for 60 times.

  Then, as understood from FIGS. 8A to 8C, by performing the moving average process, the spectrum analysis data directly obtained from the original waveform data can be smoothed, and FIG. As shown in the above, it is possible to smoothly perform the process when the sleep state is detected by comparison with the first threshold value TH1 to the third threshold value TH3.

  That is, in the present embodiment, since the moving average processing for a plurality of times (for example, 60 times) is added to the spectrum analysis data acquired in time series, it is possible to detect a stable sleep state with little fluctuation. This moving average process is executed by the arithmetic control unit 32 shown in FIG. That is, the arithmetic control unit 32 has a function as a moving average processing means for performing a moving average process on the spectrum analysis data.

  Next, when the ratio of the δ wave is obtained without using the content rate X calculated using the above-described equation (1) and the electromyogram emg shown in the denominator of the equation (1), that is, (2 ) The difference in the content rate Y calculated using the equation will be described.

Y = δ / (δ + θ + α + β) (2)
FIG. 9 is a characteristic diagram showing changes in the content rate X calculated using the equation (1) and the content rate Y calculated using the equation (2). As shown in FIG. 9, the content ratios X and Y show almost the same changes, but as understood from the curve “Y−X”, the difference between Y and X is not a constant value. That is, as shown in the equation (1), by calculating the content rate using the forehead electromyogram emg as the denominator of the arithmetic expression, a more accurate sleep state considering the influence of the electromyogram emg It can be said that detection becomes possible.

  Next, as shown in FIG. 3, it is explained that the electromyogram (frontal electromyogram) of the subject P is measured by the electrodes 11, 12, 13 provided on the forehead portion of the subject P. To do. In this embodiment, an electromyogram detected at the forehead portion of the subject P is used as the electromyogram emg. Generally, since the electromyogram of the face of the subject P is a change of the genital muscle (muscle near the jaw), it is desirable to measure by installing an electrode near the jaw of the subject P. However, since the electromyogram is measured when the subject P1 sleeps, it is difficult to install the electrode near the jaw, and the forehead electromyogram emg is used in this embodiment.

  Hereinafter, it will be described with reference to the characteristic diagram shown in FIG. 10 that the forehead electromyogram emg and the electromyogram of the genital muscle substantially coincide. The curve q1 shown in FIG. 10 shows the electromyogram of the subject's genital muscle, and the curve q2 shows the forehead electromyogram emg measured at the forehead part. As can be understood by comparing the curves q1 and q2, the timing of the amplitude of the waveforms of both is substantially the same, although there is a difference in amplitude. Therefore, even when the forehead electromyogram emg measured using the electrodes 11, 12, 13 provided in the forehead portion is used, the measurement accuracy of the electromyogram is measured in the vicinity of the genital muscle. It can be said that it is not much different from electromyogram.

  That is, in this embodiment, a forehead electromyogram is detected using the electrodes 11, 12, and 13 provided in the forehead portion of the subject P, and the content rate X of the δ wave is determined using this forehead electromyogram. By obtaining, it becomes possible to detect a sleep state without reducing accuracy and without attaching an electromyogram detection electrode in the vicinity of the jaw.

  Thus, in the sleep state detection system 101 according to the first embodiment of the present invention, the three electrodes 11, 12, 13 are provided on the forehead portion of the subject P, and measurement is performed at each electrode 11, 12, 13. Based on the electroencephalogram and the forehead electromyogram emg, the content rate X of the δ wave is calculated by the above-described equation (1), and the sleep state of the subject P is obtained based on the content rate X. That is, any of the REM sleep, the shallow non-REM sleep, and the deep non-REM sleep is determined by comparing the value of the δ wave content rate X with the first threshold TH1, the second threshold TH2, and the third threshold TH3 shown in FIG. It is estimated whether it is.

  And in the formula (1) for calculating the content rate X, the forehead electromyogram emg is added to the denominator, and the content rate X more suitable for the actual sleep state is compared with the case where this is not added. It becomes possible to calculate, and it becomes possible to improve the detection accuracy of a sleep state. In addition, since the influence due to individual differences can be reduced by using the forehead electromyogram emg, the same threshold values (first to third threshold values TH1, TH2, TH3) are used for many subjects. It becomes possible to detect a sleep state.

  Further, a passive filter is used as the high-pass filter 18 and has a characteristic that the filter characteristic gradually attenuates as shown in the band F1 in FIG. 2, so that a large amount of electroencephalogram data below 1 [Hz] is detected. It is possible to prevent the detection and improve the detection accuracy of the electroencephalogram.

[Description of Second Embodiment]
Next, a second embodiment of the present invention will be described. FIG. 11 is a block diagram illustrating a configuration of a sleep state detection system according to the second embodiment. As shown in FIG. 11, the sleep state detection system 101a according to the second embodiment includes a sleep state detection device 102a and a sleep state calculation device 103a.

  The sleep state detection device 102a is different from the sleep state detection device 102 of the first embodiment shown in FIG. 1 in that the sleep state detection device 102a includes an original waveform processing unit 22, and the waveform amplified by the amplification unit 20 is obtained. Output to the spectrum analysis processing unit 21 and output to the original waveform processing unit 22. Since the other configuration is the same as that of the sleep state detection device 102 shown in FIG. 1, the same reference numerals are given and the description of the configuration is omitted.

  The original waveform processing unit 22 has a function of removing noise from the original waveform data amplified by the amplification unit 20 and outputting it to the storage media input / output unit 23.

  In comparison with the sleep state calculation device 103 of the first embodiment shown in FIG. 1, the sleep state calculation device 103 a reads the original waveform data in addition to the spectrum analysis data by the data reading unit 31, and the original The difference is that an electrooculogram detection unit (electrocardiogram reading means) 37 that detects an eye movement diagram (hereinafter, referred to as “electrocardiogram”) of the subject P based on the waveform data is provided. Since the other configuration is the same as that of the sleep state calculation device 103 shown in FIG. 1, the same reference numerals are given and description of the configuration is omitted.

  The electrooculogram detection unit 37 obtains an electrooculogram waveform based on the electromyogram included in the original waveform data. Specifically, the electromyogram detected in the time zone in which the subject P is estimated to be in the REM sleep period is determined to be an electrooculogram. For example, in the REM sleep period, it is known that the subject's eyes move frequently, and the electrooculogram at this time is measured using an electrode provided in the vicinity of the subject's eyes, as shown in FIG. A waveform like this is obtained. At this time, when the forehead electromyogram is measured with the electrodes provided in the forehead portion, a waveform as shown in FIG. 12B is obtained. This forehead electromyogram shows a signal that has spread as the direct-current retinal electrostatic potential generated before and after the eyeball detected by the forehead is changed by the movement of the eyeball. Comparing both waveforms in FIGS. 12A and 12B, it can be seen that the forehead electromyogram generates a signal synchronized with the electrooculogram. Therefore, if the forehead electromyogram emg is measured only in the REM sleep period based on the electroencephalogram data detected using the electrodes 11, 12, and 13, it can be determined that this is an electrooculogram.

  In addition, since it is known that the subject P enters the REM sleep period immediately after falling asleep and after a period of time has passed since the transition from deep non-REM sleep to shallow non-REM sleep, the REM sleep period should be estimated based on this condition. Can do.

  In the second embodiment, it is assumed that the forehead electromyogram emg detected by the above process is an electrocardiogram in the period estimated to be the REM sleep period. The sleep state is detected with higher accuracy by comparison with. Specifically, before and after the transition from REM sleep to non-REM sleep, there may be no significant difference in spectrum analysis data. That is, in the example shown in FIG. 4B, it has been described that when the content rate X of the δ wave included in the electroencephalogram data exceeds the second threshold value TH2, it is determined that the sleep is shallow non-REM sleep. In the vicinity of the two-threshold TH2, it may not be possible to accurately determine whether it is REM sleep or non-REM sleep only by spectrum analysis data.

  Therefore, in the second embodiment, an electrooculogram is detected by the electrooculogram detection unit 37 shown in FIG. 11, and the detection result is adopted for the determination of the sleep state, thereby detecting the sleep state with higher accuracy. Do.

  Next, the operation of the sleep state detection system 101a according to the second embodiment will be described. In the second embodiment, the waveform data amplified by the amplification unit 20 is branched into two systems, one of which is output to the spectrum analysis processing unit 21 and the other is output to the original waveform processing unit 22. Then, the spectrum analysis processing unit 21 generates spectrum analysis data by Fourier-transforming the original waveform data as in the first embodiment described above. Further, the original waveform processing unit 22 removes noise from the waveform data amplified by the amplification unit 20 and outputs the waveform data after the noise removal to the storage media input / output unit 23 as original waveform data. Both the spectrum analysis data and the original waveform data are written to the memory card 24 by the storage media input / output unit 23.

  The sleep state computing device 103a measures changes in the electrooculogram based on the spectrum analysis data stored in the memory card 24 and the original waveform data. When it is determined that an electrooculogram is generated, it is determined that REM sleep occurs even if the content rate X of the δ wave at this time shows a large value.

  That is, as shown in FIG. 4B, when the content rate X of the δ wave increases and exceeds the second threshold value TH2, it is determined that the sleep is non-REM sleep (shallow non-REM sleep or deep non-REM sleep). Is done. However, if the electrooculogram at this time has a large amplitude, it is determined that the muscles in the vicinity of the eyes are moving, and it is determined that the sleep is not REM sleep but REM sleep.

  13A shows the original waveform data during deep non-REM sleep, and FIG. 13B shows the original waveform data during REM sleep. When both are compared, the REM sleep shown in FIG. 13B is shown. There is a region where the time is larger in amplitude. The electrooculogram detection unit 37 detects a waveform based on the electrooculogram based on the original waveform data, and outputs the detection result to the calculation control unit 32.

  Then, similarly to the first embodiment described above, the arithmetic control unit 32 calculates the δ-wave content rate X by the above-described equation (1) based on the spectrum analysis data, and sleeps based on the content rate X. It is determined whether the state is REM sleep, shallow non-REM sleep, or deep non-REM sleep. Further, when it is determined that the electrooculogram is detected by the electrooculogram detection unit 37, even when the sleep state detection result based on the content rate X is shallow non-REM sleep or deep non-REM sleep. Since the subject's eyes are moving frequently, it is determined that the subject is REM sleep.

  Thus, in the sleep state detection system 101a according to the second embodiment, in addition to the content rate X of δ wave obtained based on the spectrum analysis data, it is REM sleep or non-REM sleep based on the electrooculogram. Therefore, in addition to the effects of the first embodiment described above, it is possible to detect the sleep state with high accuracy based on the eye movements of the subject.

[Description of Third Embodiment]
Next, a third embodiment of the present invention will be described. 14 and 15 are block diagrams showing the configuration of the sleep state detection system 101b according to the third embodiment, and FIG. 14 shows the configuration of the sleep state detection device 102b for storing data in the memory card 24. FIG. 15 shows the configuration of the sleep state calculation device 103b that reads the data stored in the memory card 24 and calculates the sleep state of the subject. Moreover, FIG. 16 is explanatory drawing which shows a mode that the insulating gel sheet 25a is attached to the forehead part of the test subject P. FIG.

  As shown in FIG. 16, the insulating gel sheet 25 a used in the sleep state detection apparatus 102 b according to the third embodiment includes the first to third electrodes 11 and 12, similarly to the insulating gel sheet 25 shown in FIG. 3 described above. 13 is mounted, and a three-axis acceleration sensor (acceleration detecting means) 44 for detecting a tilt in a three-dimensional direction and a pulse wave detecting sensor (pulse wave detecting means) 40 for detecting a pulse wave at the forehead are mounted. Has been. And based on each data detected by each electrode 11, 12, 13, the pulse wave detection sensor 40, and the triaxial acceleration sensor 44, the sleep state of the test subject P is detected by the method mentioned later.

  As shown in FIG. 14, the sleep state detection apparatus 102b according to the third embodiment includes buffer amplifiers 14 and 15, a differential amplifier 16, a notch filter 17, a high-pass filter 18, a low-pass filter 19, an amplification unit 20, and a spectrum analysis process. Unit 21 and storage medium input / output unit 23. In addition, since these structures are the same as that of the sleep state detection apparatus 102 of 1st Embodiment shown in FIG. 1, description is abbreviate | omitted.

  In addition to the above components, the sleep state detection apparatus 102b according to the third embodiment further includes a pulse wave detection sensor 40, an LED control unit 45, a current / voltage conversion unit 46, a filter 47, and An amplifying unit 48 is provided.

  The pulse wave detection sensor 40 includes a red LED 41 that emits light having a wavelength of red light (for example, 660 nm), an infrared LED 42 that emits light having a wavelength of infrared light (for example, 880 nm), and a photodiode 43. The red LED 41 emits red light toward the forehead part, and the reflected light is detected by the photodiode 43. Further, infrared light is irradiated toward the forehead portion from the infrared LED 42, and the reflected light is detected by the photodiode 43.

  The LED control unit 45 acquires a detection signal from the pulse wave detection sensor 40 by controlling the red LED 41, the infrared LED 42, and the photodiode 43 provided in the pulse wave detection sensor 40. The current / voltage converter 46 converts the current signal detected by the photodiode 43 into a voltage signal.

  The filter 47 removes noise included in the signal detected by the pulse wave detection sensor 40 and outputs the noise to the amplifying unit 48. The amplifying unit 48 outputs a signal related to the pulse wave from which the noise component has been removed by the filter 47 to the storage media input / output unit 23.

  The triaxial acceleration sensor 44 detects the acceleration in each of the X axis, Y axis, and Z axis directions, and detects the direction in which the subject's forehead is facing. For example, information such as whether the subject is facing upward or the lateral direction can be acquired. The detection signal of the triaxial acceleration sensor 44 is output to the storage media input / output unit 23.

  Hereinafter, the relationship between the orientation of the subject's face and the acceleration in each of the X-axis, Y-axis, and Z-axis directions detected by the three-axis acceleration sensor 44 will be described with reference to the characteristic diagram shown in FIG. When the subject is on his back, the X-axis and Y-axis accelerations are hardly detected, and the Z-axis acceleration is detected. When the subject's head is facing right, the X-axis acceleration is detected, the smaller acceleration is detected on the Y-axis, and the Z-axis acceleration is hardly detected. Similarly, the output of the three-axis acceleration sensor 44 is determined for each of the leftward, upside down, waking up, and left / right movements. That is, the state of the subject's face can be recognized based on the X-axis, Y-axis, and Z-axis accelerations.

  The storage media input / output unit 23 illustrated in FIG. 14 stores a signal related to the pulse wave and a signal detected by the triaxial acceleration sensor 44 in the memory card 24.

  The sleep state calculation device 103b shown in FIG. 15 has the same configuration as the sleep state calculation device 103 shown in FIG.

  And in 3rd Embodiment, in addition to each brain wave waveform (delta wave etc.) demonstrated in 1st Embodiment, more highly accurate by employ | adopting a test subject's pulse-wave data and a test subject's head three-dimensional position data. The sleep state of the subject is detected. That is, the subject's sleep state is determined to be any one of awakening, REM sleep, shallow non-REM sleep, and deep non-REM sleep by the method using the electroencephalogram shown in the first embodiment, and the determination result is converted to pulse wave data and By correcting based on the three-dimensional position data, a sleep state with higher accuracy is detected.

  Specifically, based on the pulse wave of the subject, the pulse rate and the oxygen saturation level in the blood are detected, and if the pulse rate is high or the oxygen saturation level is low, there is a high possibility of being awake. The sleep state detection result obtained based on the electroencephalogram is determined based on the judgment.

  In addition, when the subject's body movement (movement to change the body orientation, etc.) is detected based on the three-dimensional position data, there is a high possibility that an artifact is generated along with this movement. The sleep state detection result obtained in this way is corrected. This makes it possible to detect a sleep state with higher accuracy.

  Next, the operation of the sleep state detection system 101b according to the third embodiment configured as described above will be described. As shown in FIG. 16, since each of the three electrodes 11, 12, 13 is provided on the insulating gel sheet 25 a, when the subject P enters a sleep state, the subject P, by the electrodes 11, 12, 13. The electroencephalogram and electromyogram waveforms in the sleep state are detected.

  As in the first embodiment described above, the detected waveform data passes through the notch filter 17, the high-pass filter 18, and the low-pass filter 19, thereby obtaining a desired frequency component. Then, the spectrum analysis processing unit 21 performs spectrum analysis processing. Then, spectrum analysis data obtained by the analysis process is written in the memory card 24.

  On the other hand, the pulse wave data of the subject P is acquired by the pulse wave detection sensor 40 mounted on the insulating gel sheet 25a, and the detected pulse wave data is converted into a voltage signal by the current / voltage conversion unit 46, and the filter 47, And output to the storage media input / output unit 23 via the amplification unit 48. Further, the X-axis, Y-axis, and Z-axis acceleration data are detected by the triaxial acceleration sensor 44, and this acceleration data is output to the storage media input / output unit 23. The storage media input / output unit 23 performs processing for writing spectrum analysis data, pulse wave data, and acceleration data to the memory card 24.

  Thereafter, the memory card 24 in which each data is written is inserted into the data reading unit 31 of the sleep state computing device 103b shown in FIG. 15 by the user, and the spectrum analysis data and pulse wave data written in the memory card 24 are written. , And acceleration data are read under the control of the arithmetic control unit 32.

  As a result, the sleep state can be acquired based on the electroencephalogram data as in the first embodiment described above. In the present embodiment, the sleep state acquired based on the electroencephalogram data is corrected based on the body motion of the subject P and the pulse wave data. First, as a first example of the third embodiment, processing for correcting a sleep state based on body movement will be described with reference to the characteristic diagrams shown in FIGS. 17 and 18 and the correspondence table shown in FIG. . Note that FIG. 17 and FIG. 18 have the same time axis, and therefore the timings A to F shown in FIG. 17 and the timings A to F shown in FIG. 18 match.

  FIG. 17A is data showing a result (inspection determination result) of detecting the sleep state of the subject P using a dedicated device (a high-precision detection device used in a hospital or the like), and the sleep state over time. Changes are shown. Specifically, it is detected by distinguishing into four stages of arousal state (indicated by “W” in the figure), REM sleep, shallow non-REM sleep, and deep non-REM sleep. FIG. 17B shows the result of detecting a sleep state using the sleep state detection system 101b according to the third embodiment, and the sleep state is also detected in four stages.

  FIG. 18 (a) is a characteristic diagram showing a change in the oxygen saturation of the subject P, and FIG. 18 (b) is a characteristic diagram showing a change in the pulse rate of the subject P. FIG. 18 (c) These are the characteristic views which show the subject's P body motion (3-axis acceleration). As described above, the oxygen saturation is detected by the photodiode 43 with the red light output from the red LED 41 reflected by the blood vessel of the subject P, and similarly, the infrared light output from the infrared LED 42 is detected by the subject P. The light reflected by the blood vessel can be detected by the photodiode 43 and obtained based on the detection results of both. Similarly, the pulse rate of the subject P can be obtained. Since the method for detecting the oxygen saturation and the pulse rate is a well-known technique, a detailed description thereof will be omitted.

  The body movement of the subject P can be obtained based on the detection result of the triaxial acceleration sensor 44, and the acceleration in each direction of the X axis, the Y axis, and the Z axis is shown in FIG. Detected. For example, when the subject P moves the head in the left-right direction, acceleration is detected in the X-axis and Y-axis directions, and when the subject moves up the head, the acceleration is detected in the Z-axis direction. The

  In the first example of the third embodiment, the sleep state detection result acquired from the electroencephalogram data is corrected based on the body motion of the subject P. Specifically, when the sleep state detected based on the electroencephalogram data (sleep state obtained by the method described in the first embodiment) is acquired as shown by the solid line in FIG. The sleep state is corrected based on the body motion data (triaxial acceleration output voltage) shown in (c).

  Hereinafter, a detailed description will be given with reference to the correspondence table shown in FIG. At time A shown in FIGS. 17 and 18, some body movement occurs in the X-axis direction as shown in FIG. 18 (c). At this time, the pulse rate does not change as shown in FIG. 18B, and the oxygen saturation hardly changes as shown in FIG. Further, since the sleep state detected based on the electroencephalogram data (data indicated by the solid line in FIG. 17B) indicates shallow non-REM sleep at time A, this detection result is not changed. That is, the total sleep stage at time A is determined to be shallow non-REM sleep.

  Further, at time B, as shown in FIG. 18 (c), a large body motion occurs on the X axis and the Y axis. Therefore, artifacts (noise) to the electroencephalogram due to body movement are assumed, and no significant change in the pulse rate is observed, so the REM sleep, which is the detection result by the electroencephalogram, is changed to a shallow non-REM sleep. That is, as shown by the dotted line portion at time B in FIG. 17B, the sleep state detection result is changed from REM sleep to shallow non-REM sleep, and the total sleep stage is set to shallow non-REM sleep. Similarly for C, D, E, and F shown in FIG. 24, the sleep state due to the electroencephalogram is corrected based on the body motion, and the total sleep stage is obtained.

  Thus, in the first example of the third embodiment, the body motion of the subject P is detected using the three-axis acceleration sensor 44, and the sleep state obtained based on the electroencephalogram data is used as the body motion data. Therefore, the sleep state can be detected with higher accuracy in consideration of the body motion data of the subject P.

  Next, as a second example of the third embodiment, a process of correcting the sleep state based on the pulse rate will be described with reference to the characteristic diagrams shown in FIGS. 19 and 20 and the correspondence table shown in FIG. . Note that FIG. 19 and FIG. 20 have the same time axis, and therefore the timings A to G shown in FIG. 19 and the timings A to G shown in FIG. 20 match.

  FIG. 19A is data showing the inspection determination result, and shows a change in the sleep state over time. Moreover, FIG.19 (b) has shown the result of having detected the sleep state using the sleep state detection system 101b which concerns on 3rd Embodiment, and the sleep state is also distinguished and detected in 4 steps.

  FIG. 20A is a characteristic diagram showing a change in the oxygen saturation of the subject P, FIG. 20B is a characteristic diagram showing a change in the pulse rate of the subject P, and FIG. These are characteristic views showing the body movement of the subject P.

  In the second example of the third embodiment, the sleep state detection result acquired from the electroencephalogram data is corrected based on the pulse rate of the subject P. Specifically, when the sleep state detected based on the electroencephalogram data is acquired as shown by the solid line in FIG. 19B, the sleep state is calculated based on the pulse rate shown in FIG. to correct. At this time, the fluctuation of the pulse rate is varied in the range of 17% (pulse rate 72) to 22% (pulse rate 75) with respect to the average pulse rate during sleep (pulse rate 61.4 in this example). When detected, the sleep state (total sleep stage) is corrected. In addition, the range of 17% to 22% shows an example, and is not limited to this range.

  Hereinafter, a detailed description will be given with reference to the correspondence table shown in FIG. At time A shown in FIGS. 19 and 20, the pulse rate exceeds 72 [times / min] as shown in FIG. At this time, almost no body movement occurs as shown in FIG. Further, as shown in FIG. 19B, since the sleep state detection result by the brain wave is awakening (W), this determination is not changed. That is, the overall sleep stage at time A is determined to be awake.

  At time B, the pulse rate exceeds 72 [times / min] as shown in FIG. Therefore, since it is determined that there is a high possibility of being awakened, it is determined that the REM sleep that is the determination result by the electroencephalogram is the electroencephalogram awakening (aW), and the REM sleep that is the determination result by the electroencephalogram is the electroencephalogram awakening ( aW). In other words, the sleep state determination result is changed from REM sleep to electroencephalogram awakening (aW) as shown by the dotted line portion at time B in FIG. 19B (in the figure, the electroencephalogram awakening is the same as the normal awakening “W”. The total sleep stage is defined as brain wave awakening (aW). Similarly, for C, D, E, F, and G shown in FIG. 25, the sleep state due to the electroencephalogram is corrected based on the pulse rate, and the total sleep stage is obtained. Here, electroencephalogram awakening (aW) generally means that a sudden change in EEG frequency lasts for 3 seconds or more, and that there is sleep for 10 seconds or more before the appearance, and in addition to the change on EEG It is said that the pulse rate is transiently increased or decreased. Since it is difficult to obtain only an electroencephalogram from the forehead, a transient increase or decrease in pulse rate was defined as an electroencephalogram awakening (aW).

  As described above, in the second example of the third embodiment, the pulse rate and oxygen saturation of the subject P are detected using the pulse wave detection sensor 40, and the sleep state obtained based on the electroencephalogram data is expressed as described above. Since correction is performed based on the pulse rate and oxygen saturation data, it is possible to detect a sleep state with higher accuracy in consideration of the pulse wave of the subject P.

  Next, as a third example of the third embodiment, with respect to the process of correcting the sleep state based on the pulse rate (a numerical value higher than the second example), the characteristic diagrams shown in FIGS. 21 and 22 and FIG. This will be described with reference to the correspondence table shown in FIG. Note that FIG. 21 and FIG. 22 have the same time axis, and therefore the timings A to F shown in FIG. 21 and the timings A to F shown in FIG. 22 match.

  FIG. 21A is data showing the inspection determination result, and shows a change in the sleep state with time. Moreover, FIG.21 (b) has shown the result of having detected the sleep state using the sleep state detection system 101b which concerns on 3rd Embodiment, and the sleep state is also distinguished and detected in 4 steps.

  22A is a characteristic diagram showing a change in the oxygen saturation of the subject P, FIG. 22B is a characteristic diagram showing a change in the pulse rate of the subject P, and FIG. These are characteristic views showing the body movement of the subject P.

  And in 3rd Example of 3rd Embodiment, the detection result of the sleep state acquired from the electroencephalogram data is correct | amended based on the pulse rate of the test subject P similarly to 2nd Example mentioned above. Specifically, when the sleep state detected based on the electroencephalogram data is acquired as shown by the solid line in FIG. 21B, the sleep state is calculated based on the pulse rate shown in FIG. to correct. At this time, the fluctuation of the pulse rate occurred when a fluctuation of 22% (pulse rate 75) or more was detected with respect to the average pulse rate during sleep (pulse rate 61.4 in this example). It is assumed that the sleep state is corrected. The pulse rate threshold is not limited to 22%.

  Hereinafter, a detailed description will be given with reference to the correspondence table shown in FIG. At time G shown in FIGS. 21 and 22, the pulse rate exceeds 75 [times / min] as shown in FIG. At this time, body movement occurs as shown in FIG. Further, as shown in FIG. 21B, since the sleep state detection result by the brain wave is awakening (W), this determination is not changed. That is, it is determined that the overall sleep stage at time G is awake.

  At time B, the pulse rate exceeds 75 [times / min] as shown in FIG. Therefore, since it is determined that there is a high possibility of being awakened, it is determined that the REM sleep that is the determination result by the electroencephalogram is the electroencephalogram awakening (aW), and the REM sleep that is the determination result by the electroencephalogram is the electroencephalogram awakening Change to That is, the sleep state detection result is changed from REM sleep to electroencephalogram awakening, and the total sleep stage is set to electroencephalogram awakening (aW), as indicated by the dotted line portion shown at time B in FIG. Similarly, for A, C, D, E, and F shown in FIG. 26, the sleep state by the electroencephalogram is corrected based on the pulse rate, and the total sleep stage is obtained.

  FIG. 23 shows data obtained by combining the sleep states changed in the first to third embodiments. As shown in the figure, it is understood that by adopting the first to third embodiments and correcting the sleep state obtained based on the electroencephalogram, the inspection determination result shown in FIG. The That is, it is possible to detect the sleep state with higher accuracy by correcting the sleep state obtained from the electroencephalogram data using the body motion data and the pulse rate data.

  In this way, in the sleep state detection system 101b according to the third embodiment, the sleep state obtained based on the electroencephalogram is corrected based on the body motion of the subject P and the pulse wave data, so sleep with higher accuracy. The state can be detected.

  In addition, when only brain wave data is used, brain wave awakening (aW) cannot be detected, but subject P is brain wave awakening (aW) by detecting pulse wave data and the pulse rate of subject P. It is possible to detect whether the sleep state is more precise and fine.

  In the third embodiment described above, the example in which the sleep state of the subject is corrected using both the pulse wave data and the acceleration data indicating the movement of the head of the subject P has been described. It is also possible to correct the sleep state using at least one of the acceleration data.

  It is also possible to determine whether the subject's autonomic nerve is sympathetic dominant or parasympathetic dominant based on the pulse wave data, and to determine the sleep state based on this. That is, it is known that the sympathetic nerve is dominant when the subject is awake, the parasympathetic nerve is dominant during shallow non-REM sleep and deep non-REM sleep, and the autonomic nerve is irregular during REM sleep. Therefore, the sleep state can be determined with higher accuracy by obtaining the subject's autonomic nerve.

  Hereinafter, it will be described in detail with reference to the explanatory diagram shown in FIG. Various indexes reflecting the autonomic nervous function can be acquired from the power spectrum (characteristic diagram on the right side of FIG. 28) of pulse interval (RR interval) fluctuation. For example, LF (0.04-0.15 Hz), HF (0.2-0.4 Hz) (LF is affected by both sympathetic and parasympathetic nerves, HF reflects parasympathetic nerves), Based on the LF / HF value, it is possible to determine whether or not the sympathetic nerve is dominant.

  And the HF power value becomes the maximum during shallow non-REM sleep, and the power value of LF takes a very high value equivalent to wakefulness W during REM sleep and takes a value different from that during non-REM sleep. It can be used as an index for shallow non-REM sleep and REM sleep, which are difficult to judge at the sleep stage.

[Description of Fourth Embodiment]
Next, a fourth embodiment of the present invention will be described. The system configuration is the same as in FIGS. 14 and 15 shown in the third embodiment. And in 4th Embodiment, based on the pulse wave detected by the pulse wave detection sensor 40, a test subject's respiration is detected, Based on the detected respiration data, the data regarding sleep quality, such as sleep apnea syndrome, are obtained. get.

  Specifically, pulse wave data detected by the pulse wave detection sensor 40 shown in FIG. 14 is written to the memory card 24, and the data written to the memory card 24 by the data reading unit 31 shown in FIG. Read. And the calculation control part 32 acquires respiration data based on this pulse wave data.

  FIG. 27A is a characteristic diagram showing how respiratory data is acquired from pulse wave data. When pulse wave data indicated by a solid line in FIG. 27A is detected, an envelope of this pulse wave data is shown. Based on the line, respiration data as indicated by the alternate long and short dash line in the figure is acquired. That is, since the amplitude of the pulse wave data indicated by the solid line in FIG. 27A fluctuates according to respiration, the respiratory data of the subject can be obtained by obtaining the envelope of the pulse wave data, and consequently the subject. The state of breathing during sleep can be examined.

  For example, when the subject has obstructive sleep apnea syndrome (OSAS), the amplitude of the respiratory data decreases during time t1, as shown by a curve r1 in FIG. Therefore, it is possible to determine whether or not the subject has obstructive sleep apnea syndrome by detecting such a change in respiratory data.

  In addition, when the subject has a central sleep apnea syndrome (CSAS), as shown by a curve r2 in FIG. 27 (b), the respiratory data has almost no amplitude during time t1. Become. Therefore, it is possible to determine whether or not the subject has a central sleep apnea syndrome by detecting such a change in respiratory data.

  Furthermore, although there is no apparent apnea, hypopnea, or hypoxemia, a false arousal reaction with increased respiratory effort associated with upper airway stenosis occurs, resulting in upper airway resistance syndrome, a condition that disrupts sleep There is a case. Even in such a case, this symptom can be specified by detecting the amplitude of the respiratory data. Specifically, as shown in FIG. 27C, when the amplitude of the respiratory data becomes small in a specific time zone, it can be determined that the upper airway resistance syndrome is present.

  As described above, in the sleep state detection system according to the fourth embodiment, the subject's respiratory data is acquired based on the pulse wave data, and further, based on the respiratory data, whether the subject has sleep apnea syndrome. Alternatively, it is possible to determine whether the patient has an upper airway resistance syndrome, and further, it is possible to determine an obstructive sleep apnea syndrome and a central sleep apnea syndrome. Therefore, it is possible to determine not only the information regarding the depth of sleep but also the respiratory state during sleep.

  As a modification of the above-described fourth embodiment, it is also possible to determine whether or not breathing has stopped based on the pulse rate detected from the pulse wave data and the oxygen saturation. FIG. 29 is a characteristic diagram showing changes in oxygen saturation and pulse rate when the subject stops breathing. As shown in FIG. 29 (a), the oxygen saturation gradually decreases when the subject stops breathing. The oxygen saturation continues to decrease for a while after resuming breathing. And when oxygen saturation falls to a certain level, oxygen saturation will rise after that and will return to a normal value. Further, as shown in FIG. 29 (b), the pulse rate gradually decreases as the subject's breathing stops, and gradually increases and returns to the normal value when the breathing is resumed.

  Then, based on the data shown in FIG. 29, it can be determined whether or not breathing has stopped based on the subject's oxygen saturation or pulse rate, which in turn can be useful for examining the respiratory state during sleep. it can. For example, if it is determined that breathing has stopped based on oxygen saturation or pulse data, it can be determined that the subject is suspected of having sleep apnea syndrome.

  As described above, the sleep state detection device, the sleep state calculation device, and the sleep state detection system of the present invention have been described based on the illustrated embodiment, but the present invention is not limited to this, and the configuration of each unit is the same It can be replaced with an arbitrary configuration having the above function.

  For example, in the above-described embodiment, a configuration in which the sleep state detection devices 102, 102a, and 102b and the sleep state calculation devices 103, 103a, and 103b are separated is taken as an example, but these may be integrated.

  The present invention can be used to detect a sleep state by detecting a brain wave of a subject.

DESCRIPTION OF SYMBOLS 11, 12, 13 Electrode 14 Buffer amplifier 15 Buffer amplifier 16 Differential amplifier 17 Notch filter 18 High pass filter 19 Low pass filter 20 Amplifying part 21 Spectrum analysis processing part 22 Original waveform processing part 23 Storage media input / output part 24 Memory card 25, 25a Insulating gel sheet 32 Calculation control unit 33 Calculation program storage unit 34 Memory 35 Input operation unit 36 Display unit 37 Electrooculogram detection unit 40 Pulse wave detection sensor 41 Red LED
42 Infrared LED
DESCRIPTION OF SYMBOLS 43 Photodiode 44 3-axis acceleration sensor 45 LED control part 46 Current / voltage conversion part 47 Filter 48 Amplification part 101,101a Sleep state detection system 102,102a Sleep state detection apparatus 103,103a Sleep state calculation apparatus

Claims (14)

In the sleep state detection device for detecting the sleep state of the subject,
At least three electrodes provided on the forehead of the subject;
A passive high-pass filter that removes a desired low-frequency component from the electroencephalogram detected by the electrode;
An active low-pass filter that removes a frequency component higher than a predetermined frequency corresponding to an electromyogram from the output signal of the high-pass filter;
Spectrum analysis means for spectrum analysis of the signal that has passed through the low-pass filter;
Storage medium input / output means for writing spectrum analysis data output from the spectrum analysis means to a portable storage medium;
A sleep state detection apparatus comprising: An original waveform output means for outputting the signal that has passed through the low-pass filter as original waveform data;
The sleep state detection apparatus according to claim 1, wherein the storage medium input / output unit writes the original waveform data to the storage medium in addition to the spectrum analysis data.   The three electrodes are adhesive conductive gel electrodes that can be attached to the forehead portion of the subject, and each electrode is attached to the forehead portion using an adhesive insulating gel sheet. The sleep state detection apparatus according to claim 1 or 2.   Pulse wave detection means provided on the forehead portion for detecting the pulse wave of the subject is further provided, and the storage medium input / output means is a pulse wave detected by the pulse wave detection means on the portable storage medium. Data is written, The sleep state detection apparatus of any one of Claims 1-3 characterized by the above-mentioned.   Provided in the forehead portion is further provided with acceleration detection means for detecting the orientation of the subject's head, and the storage medium input means writes acceleration data detected by the acceleration detection means to the portable storage medium. The sleep state detection device according to claim 1, wherein the sleep state detection device is a sleep state detection device. In the sleep state calculation device that calculates the sleep state of the subject based on the spectrum analysis data obtained based on the brain wave of the subject,
Data reading means for reading storage data of a storage medium in which the spectrum analysis data is stored;
Based on the data of δ wave, θ wave, α wave, β wave and forehead electromyogram included in the spectrum analysis data, δ wave level is δ, θ wave level is θ, α wave level is α, β wave When the level is β and the level of the forehead electromyogram is emg, calculation means for calculating the content rate X of the δ wave by executing the calculation of “X = δ / (δ + θ + α + β + egg)”;
A sleep state detection means for detecting whether the sleep state of the subject is awakening, REM sleep, shallow non-REM sleep, or deep non-REM sleep based on the content rate X obtained by the calculation means; ,
A sleep state computing device comprising: The data reading means reads the original waveform data of the electroencephalogram stored in the storage medium in addition to the spectrum analysis data,
An electrooculogram reading means for reading an electrooculogram based on the original waveform data;
The sleep state detection means is based on an electrooculogram read by the electrooculogram reading means in addition to the content rate X, and the sleep state of the subject is awakening, REM sleep, shallow non-REM sleep, and deep non-REM The sleep state calculation apparatus according to claim 6, wherein the sleep state calculation device detects which one of sleep. In addition to the spectral analysis data, the storage medium stores at least one of the pulse wave data of the subject and the acceleration data indicating the orientation of the subject's head,
The data reading means reads at least one of the pulse wave data of the subject stored in the storage medium and the acceleration data indicating the direction of the subject's head in addition to the spectrum analysis data,
The sleep state detection means uses the at least one of the pulse wave data and the acceleration data in addition to the spectrum analysis data, and the sleep state of the subject is awake, REM sleep, shallow non-REM sleep, and deep non-REM sleep, The sleep state calculation device according to claim 6, wherein the sleep state calculation device is detected.   The sleep state computing device according to any one of claims 6 to 8, further comprising moving average processing means for moving average processing the spectrum analysis data read by the data reading means. In the sleep state detection system that detects the sleep state of the subject,
At least three electrodes provided on the forehead of the subject;
A passive high-pass filter that removes a desired low-frequency component from the electroencephalogram detected by the electrode;
An active low-pass filter that removes a desired high-frequency component from the output signal of the high-pass filter;
Spectrum analysis means for spectrum analysis of the signal that has passed through the low-pass filter;
Based on the δ wave, θ wave, α wave, β wave, and forehead electromyogram data included in the spectrum analysis data output from the spectrum analysis means, the δ wave level is δ, the θ wave level is θ, α When the wave level is α, the β wave level is β, and the electromyogram level is emg, calculation means for calculating the content rate X of the δ wave by executing the calculation of “X = δ / (δ + θ + α + β + egg)”;
A sleep state detection means for detecting whether the sleep state of the subject is awakening, REM sleep, shallow non-REM sleep, or deep non-REM sleep based on the content rate X obtained by the calculation means; ,
A sleep state detection system comprising: An original waveform output means for outputting the signal that has passed through the low-pass filter as original waveform data; and an electrooculogram reading means for reading an electrocardiogram based on the original waveform data;
The sleep state detection means is based on an electrooculogram read by the electrooculogram reading means in addition to the content rate X, and the sleep state of the subject is awakening, REM sleep, shallow non-REM sleep, and deep non-REM The sleep state detection system according to claim 10, wherein one of sleep is detected.   Provided in the forehead portion, further comprising pulse wave detection means for detecting the pulse wave of the subject, wherein the sleep state detection means is based on the pulse wave data detected by the pulse wave detection means; The sleep state detection according to claim 10, wherein the sleep state is detected from awakening, REM sleep, shallow non-REM sleep, and deep non-REM sleep. system.   Provided in the forehead portion, further comprising an acceleration detection means for detecting the orientation of the subject's head, wherein the sleep state detection means is based on acceleration data detected by the acceleration detection means, the sleep state of the subject The sleep state detection system according to any one of claims 10 to 12, which detects any one of awakening, REM sleep, shallow non-REM sleep, and deep non-REM sleep.   The sleep state detection system according to any one of claims 10 to 13, further comprising moving average processing means for moving average processing the spectrum analysis data output from the spectrum analysis means.

Images