JP7432233B2 - Sleep analysis method and device - Google Patents

Description

The present invention can be easily attached to a test subject, does not require special hospitalization and testing, and can be easily used by the test subject in their daily life. It is a simple, accurate and reliable way to analyze sleep (non-REM sleep, shallow sleep, deep sleep, etc.), and can be used not only for adults but also for children and infants who cannot be tested using conventional polysomnography. Related to analysis method and device.

Conventional sleep analysis methods are painful tests in which the subject is hospitalized for one day, wearing a large polysomnography device, and restrained to a bed. There were problems that made it difficult to apply to small children.

Japanese Patent Application Publication No. 2002-291710 Japanese Patent Application Publication No. 2010-99173

For example, the device shown in Patent Document 1 has been proposed as a device that solves the problem of polysomnography that requires one-day hospitalization, but since it analyzes only pulse data, there are problems with accuracy. remains. Further, Patent Document 2 discloses a sleep management system for a large number of people, but there is a problem in that the detection device itself is large-scale. Furthermore, it does not accurately determine the sleep onset time when the brain falls asleep after bed-in.

The present invention was made in view of the above-mentioned circumstances, and an object of the present invention is to easily attach it to the chest of a subject, process sleep detection information detected by three types of built-in sensors, and detect sleep. It is an object of the present invention to provide a sleep analysis method and device that performs accurate sleep analysis for various phenomena that occur at times (for example, sleep onset latency, etc.), and that allows the analysis to be performed painlessly even for children.

The present invention relates to a sleep analysis method with a small number of sensors, and the above object of the present invention is to provide temperature data obtained by a sleep information detection device equipped with a temperature sensor, an acceleration sensor, and an electrocardiograph attached to the body of a subject ; An operating method of a sleep analysis device, wherein the sleep analysis unit analyzes the sleep state of the subject based on acceleration data and electrocardiogram data, wherein the temperature data is based on the skin temperature of the subject obtained from the temperature sensor. The acceleration data is acceleration data with the subject's height direction as the Y-axis, the left-right direction as the X-axis, and the front-back direction as the Z-axis, and the acceleration data is acceleration data with the subject's height direction as the Y-axis, the left-right direction as the X-axis, and the front-back direction as the Z-axis, a heart rate, a sympathetic nerve index and a parasympathetic nerve index obtained by processing the electrocardiogram data of , a change in the intersection of the sympathetic nerve index and the parasympathetic nerve index, and the acceleration data of the subject from the acceleration sensor. The posture of the subject, the energy consumption, and the acceleration data in the Y-axis direction of the first predetermined time obtained when detecting sleep information among the acceleration data obtained by processing data and the acceleration data in the X-axis direction, and based on the acceleration data adopting the larger one and the respiration rate calculated based on the parasympathetic nerve index , the sleep analysis unit The sleep states include entering bed, falling asleep (latency time), falling asleep after waking up from sleep, and waking up, and the sleep analysis unit analyzes the sleep state when entering bed based on the posture based on the acceleration data or determining whether the patient has woken up, and determining that the heart rate has decreased, the skin temperature has increased, the breathing rate has decreased, the energy consumption has decreased, and the sympathetic nerve index has decreased and the parasympathetic index has increased due to the , the sleep analysis unit determines the sleep onset (latency time) based on the increase in the heart rate, the decrease in the skin temperature, the increase in the respiratory rate, the increase in the energy consumption, and the increase in the sympathetic This is achieved by the sleep analysis unit determining sleep deprivation caused by awakening from sleep, based on the intersection between the increase in the nervous index and the decrease in the parasympathetic index.

Further, the present invention relates to a sleep analysis device with a small number of sensors, and the above object of the present invention is to provide a sleep information detection device that can be attached to a subject's body and is equipped with a temperature sensor, an acceleration sensor, and an electrocardiograph; Acquires and processes temperature data, acceleration data, and electrocardiogram data obtained by the information detection device, and detects entering bed, falling asleep (latency time), falling asleep, waking up, REM sleep, non-REM sleep, shallow sleep, and deep sleep. a sleep analysis unit that analyzes the sleep state of sleep , the temperature data is the skin temperature of the subject obtained from the temperature sensor, and the acceleration data is oriented in the height direction of the subject on the Y axis The sleep analysis unit determines the posture of the subject (standing position, supine position, lateral position, etc.) based on a posture signal based on the acceleration data. a posture determination unit that outputs a determination result based on the electrocardiogram data; a temperature change detection unit that detects a change (increase, decrease) in the skin temperature data and outputs the detection result; a heart rate detection unit that detects a heart rate detection unit; an autonomic nerve calculation unit that calculates a sympathetic nerve index and a parasympathetic nerve index based on time/frequency analysis of the peak interval; and a change in the chiasm of the sympathetic nerve index and the parasympathetic nerve index ( a cross-determination unit that determines (increase, decrease) and outputs a determination result, and among the acceleration data, acceleration data in the Z-axis direction and Y-axis direction of the first predetermined time obtained when detecting sleep information. a respiration detection unit that compares the acceleration data and the acceleration data in the X-axis direction and detects the respiration rate based on the acceleration data and the parasympathetic index that adopts the larger one; a motion detection unit that detects energy, and determines whether to go to bed or wake up based on the posture based on the acceleration data, and detects a decrease in the heart rate, an increase in the skin temperature, and a decrease in the respiration rate. , determining sleep onset (latency time) based on the decrease in the energy consumption and the crossover caused by the decrease in the sympathetic nerve index and the increase in the parasympathetic nerve index, and determining the increase in the heart rate and the decrease in the skin temperature. Achieved by: determining arousal from sleep based on the increase in the breathing rate, the increase in the energy consumption, and the intersection caused by the increase in the sympathetic nerve index and the decrease in the parasympathetic nerve index. be done.

According to the sleep analysis method and device according to the present invention, the sleep state of the subject is determined based on the sleep detection information obtained by the sleep information detection device attached to the subject's chest, using many analysis elements (skin temperature data, heart rate data, etc.). , respiration rate, acceleration data, electrocardiogram data, etc.), it can detect symptoms such as falling asleep, falling asleep, light sleep, deep sleep, sleep apnea syndrome, and morning surge where blood pressure spikes in the morning. It is possible to accurately and reliably analyze sleep states.

In addition, many analysis elements can be obtained with just three sensors: a temperature sensor, an acceleration sensor, and an electrocardiograph, resulting in a simple hardware configuration and easy data transmission (wireless, terminal connection, etc.). be.

1 is a perspective view showing an example of the appearance of a sleep information detection device used in the present invention. FIG. 3 is a bottom view of the main body of the sleep information detection device used in the present invention. FIG. 3 is a side view and a plan view showing an example of the structure of an electrode piece. FIG. 3 is a bottom view of the sleep information detection device used in the present invention. FIG. 2 is a schematic diagram showing a state in which the sleep information detection device is attached to the chest of a subject. FIG. 1 is a schematic diagram showing an example of a usage pattern of the sleep information detection device according to the present invention. It is a block diagram showing an example of the internal configuration of a sleep information detection device. FIG. 2 is a schematic diagram for explaining an example of coordinates (xyz) of the sleep information detection device when the subject is in a standing position and in a supine position. FIG. 2 is a block diagram showing a configuration example of an electrocardiograph. FIG. 2 is a block diagram showing a configuration example of a sleep analysis section. FIG. 2 is a block diagram showing a configuration example of a respiration detection section. FIG. 2 is a waveform diagram showing an example of electrocardiographic data. FIG. 3 is a waveform diagram showing normal and abnormal examples of electrocardiographic data waveforms. FIG. 2 is a characteristic diagram showing the relationship between indicators of sympathetic nerve activity and sleep. FIG. 3 is a characteristic diagram showing the relationship between breathing frequency fluctuation width and sleep. It is a characteristic diagram showing the relationship between sympathetic nerves, parasympathetic nerves, and sleep. FIG. 3 is a characteristic diagram showing an example of the relationship between exercise intensity and exercise acceleration. FIG. 2 is a block diagram showing a configuration example of an analysis processing section. 3 is a flowchart showing an example of the operation of the present invention. 3 is a flowchart illustrating an example of the operation of the analysis processing section. 3 is a timing chart showing an example of the operation of the present invention. It is a time chart showing an example of change during sleep.

One of the best ways to analyze sleep is with electroencephalography. There are various types of brain waves. Alpha waves are waves that appear when you close your eyes and are in a resting state, and have an irregular waveform with a frequency of 8 to 13 Hz and an amplitude of about 30 to 60 μV, and beta waves are concentrated. It is a wave that often appears and is known to have an irregular waveform with a frequency of 14 to 30 Hz and an amplitude of 30 μV or less. In addition, θ waves are waves that appear when you are drowsy or in a light sleep state, and have a frequency of 4 to 7 Hz and an amplitude of about 10 to 50 μV, and δ waves appear during deep sleep or when you are under anesthesia. It is known that the waveform appears when the situation is such that the frequency is 0.5 to 4Hz and the amplitude is about 20 to 200μV. Looking at brain waves during sleep, there are two types of sleep: REM sleep and NREM sleep. When a person begins to fall asleep, they enter NREM sleep, and gradually enter deep sleep, gradually changing their heart rate, blood pressure, and body temperature. ), but this situation is called non-REM sleep, where the brain is in a resting state and the brain waves have more delta waves. Next, the brain waves rapidly return to the same state as at the beginning of sleep, with theta waves seen during light sleep increasing, and this state is called REM sleep.

In addition, NREM sleep is classified into categories 1 to 4 from the lightest level depending on the depth of sleep (sleep degree), with categories 1 and 2 being light sleep, and categories 3 and 4. is defined as deep sleep.

In the present invention, sleep onset (latency) when the brain enters sleep is detected based on sleep detection information using as few sensors as possible (temperature sensor, acceleration sensor, electrocardiograph), rather than a polygraph device using a large-scale electroencephalogram. It is possible to reliably and accurately detect the state of sleep deprivation caused by waking up from sleep and non-REM sleep, as well as the apnea syndrome that occurs during sleep, the number of times of tossing and turning, and whether the sleep is good or bad. It consists of a sleep information detection device that is worn on the subject, and a sleep analysis section that performs sleep analysis based on the sleep detection information from the sleep information detection device.

FIG. 1 shows an example of the external configuration of a sleep information detection device 100 used in the present invention. Long electrode pieces 101A and 101B are attached to the front panel, and a USB (Universal Serial Bus) terminal 102 for charging a built-in battery 130 is provided on the front surface. In this example, there are two electrodes for acquiring electrocardiographic data, and when there are three electrodes, three electrode pieces are attached. Further, although the USB terminal 102 protrudes from the front, it may have a movable structure that is buried and protrudes during use, and the installation location can be changed as appropriate.

FIG. 2 shows the bottom surface of the main body of the sleep information detection device 100, and two circular conductive materials 103A and 103B are buried on both sides of the bottom surface of the main body. Recesses 104A and 104B, which are depressions, are provided for attaching and detaching 101B.

Further, the electrode pieces 101A and 101B have the same configuration, and an example of the configuration of the electrode piece 101A is shown in FIG. FIG. 3(A) is a side view, and FIG. 3(B) is a plan view. The electrode piece 101A is made of a flexible elongated synthetic resin or the like, and a disc-shaped conductive engagement member 105A is provided at one end of one surface to contact and engage with the conductive material 103A. A cylindrical convex portion 106A that fits into a concave portion 104A of the conductive material 103A is provided on the upper surface of the engaging member 105A. The concave portion 104A and the convex portion 106A can be easily attached and detached. The other end of the other surface of the electrode piece 101A is provided with a disk-shaped electrode material 107A that contacts or presses the subject's skin, and the engaging member 105A and the electrode material 107A are connected to a conductive lead. It is electrically connected by a wire 108A, and the potential measured by the electrode material 107A is transmitted inside the sleep information detection device 100 via the lead wire 108A, the engaging member 105A (convex portion 106A), and the conductive material 103A (concave portion 104A). control unit 140, and then input to a sleep analysis unit 300, which will be described later. The area around the electrode material 107A has a known suction cup structure, and can be easily attached to the skin (chest) of the subject.

Sleep information can be obtained by fitting the two electrode pieces 101A and 101B having such a configuration into the concave portions 104A and 104B of the conductive materials 103A and 103B and the convex portions 106A and 106B of the engaging members 105A and 105B. It can be attached to the detection device 100, and when the electrode pieces 101A and 101B are attached, the structure is as shown in the perspective view of FIG. 1 and the bottom view of FIG. 4. In this state, the suction cups at both ends of the electrode pieces 101A and 101B are applied to the subject's chest, and the sleep information detection device 100 is attached to the subject as shown in FIG. It can be attached simply by placing the suction cup on the subject's chest or by pressing lightly (pressing), so it is easy to attach to children, infants, and the like.

Note that the electrode materials 107A and 107B are electrodes of an electrocardiograph, and in this example, the potential at two locations on the subject is measured, so the interval d between them must be 80 [mm] or more. is desirable. Further, the mounting member is not limited to the suction cup structure, and other known means can be used.

FIG. 6 shows various usage forms of the sleep analysis unit 300 consisting of the sleep information detection device 100 worn on a subject and a PC (personal computer), etc. The system in FIG. The detected sleep detection information is wirelessly transmitted, the sleep analysis unit 300 directly receives the sleep detection information, and the sleep analysis result, which is processed and analyzed by the sleep analysis unit 300, is output as print, display, information, etc. It looks like this. Furthermore, in the system of FIG. 6(B), sleep detection information detected by the sleep information detection device 100 is input to the sleep analysis unit 300 via the network 1. The network 1 can include a mobile communication network, a public network such as the Internet, or a fixed telephone network, and can be realized by a WAN (Wide Area Network) or a LAN (Local Area Network), and can be wired or wireless. No question. Further, the system in FIG. 6C shows an example in which the system is connected to the network 1 via the terminal device 2, and the sleep analysis unit 300 is connected to the network 1.

Although FIG. 6 shows an example in which the sleep detection information is output wirelessly from the sleep information detection device 100, wired or optical communication may also be used. The sleep detection information stored in the memory in the sleep analysis unit 300 may be read and processed by connecting to the sleep analysis unit 300 (direct transmission/reception method).

FIG. 7 shows an example of the internal configuration of the sleep information detection device 100. The sleep information detection device 100 has a sensor section 120 for detecting sleep information, and the sensor section 120 is connected to the subject's skin (epidermis) temperature. A temperature sensor 121 that detects the temperature and outputs temperature data Th, an acceleration sensor 122 that measures the acceleration of the subject's movement (exercise) and outputs acceleration data α, and two electrodes 107A and 107B that are in contact with the subject. and an electrocardiograph 123 that outputs electrocardiographic data EC based on the electrocardiographic data. The temperature sensor 121 measures the skin temperature of the subject, and the measured temperature data Th is input to the control unit 140. The acceleration sensor 122 measures acceleration according to the movement (exercise) of the subject on three orthogonal axes (x-axis, y-axis, and z-axis), and the measured acceleration data α of the x, y, and z axes is input to the control unit 140. Ru. The electrocardiograph 122 measures the potential with two electrodes 107A and 107B, and inputs electrocardiographic data EC of the subject calculated from the potential difference to the control unit 140. The measured potential is weak and is amplified by an amplifier inside the electrocardiograph 123, so it is easily affected by noise. Therefore, in order to reduce the influence of noise and improve the S/N ratio, the electrodes 107A, 107B, amplifiers, etc. are arranged close to each other.

As described above, the temperature data Th, acceleration data α, and electrocardiogram data EC output from the sensor unit 120 are input to the control unit 140, and the control unit 140 receives the input temperature data Th, acceleration data α, and electrocardiogram data EC. The data EC is stored in areas in the memory 141 that are set in advance for each data, and is transmitted to the external sleep analysis unit 300 as sleep detection information RS via the transmission unit 142. The sleep information detection device 100 also includes a rechargeable battery 130 that supplies power to each element, and the battery 130 is charged by inserting the USB terminal 102 into a USB terminal of a PC or the like. Furthermore, in the above-described direct transmission/reception method, by inserting the USB terminal 102 into a predetermined terminal section of the sleep analysis section 300, data stored in the memory 141 can be taken in.

When the simultaneous transmission mode is set by a mode changeover switch (not shown), the control unit 140 transmits the detected sleep detection information to the transmitting unit 142 without storing it in the memory 141, and the transmitting unit 142 Send sleep detection information RS to the outside. When the storage transmission mode is set by the mode changeover switch, the sleep detection information is temporarily stored in the memory 141, and thereafter the information is read out at any time and transmitted to the transmitter 142, and the transmitter 142 transmits the sleep detection information RS to an external device. Send to. In this case, the detected sleep detection information can be stored in the memory 141 and simultaneously transmitted wirelessly to the outside, or can be simply stored in the memory 141. If the information is only stored in the memory 141, it is imported into the sleep analysis unit 300 using the above-mentioned direct transmission/reception method. The transmitting unit 142 converts the input sleep detection information RSA into a format that can be received by the external sleep analysis unit 300, and wirelessly transmits it as sleep detection information RS. As a wireless transmission method, a Wi-Fi method, a Bluetooth (registered trademark) method, etc. are used.The sleep analysis section 300 is constructed from software such as a personal computer, and is based on the received sleep detection information RS. Scientifically analyze the sleep state of the subject.

The sleep information detection device 100 used for sleep analysis of the present invention is attached to the chest of the subject as shown in FIGS. 5 and 8, and the vertical direction is the y-axis (acceleration α _y ) in the standing state of FIG. 8(A). , the left-right direction is the x-axis (acceleration α _x ), and the front-back direction is the z-axis (acceleration α _z ), and in the supine position of FIG. 8(B), the left-right direction is the x-axis (acceleration α _{x ), and the vertical direction is the x-axis (acceleration α x} ). is the y-axis (acceleration α _y ), and the direction perpendicular to the page is the z-axis (acceleration α _z ), but the axial relationship can be changed as appropriate. The acceleration sensor 122 measures acceleration related to the motion (movement) of the subject's entire chest, heart, airway, diaphragm, ribs, etc., and all measured acceleration data α is input to the control unit 140. Further, the electrocardiograph 123 includes a potential difference calculation unit 123A that calculates the difference between the potential e1 of the electrode 107A and the potential e2 of the electrode 107B using equation 1 and outputs electrocardiographic data EC, as shown in FIG. Electrocardiographic data EC is also input to the control unit 140.
(Number 1)
EC=e1-e2
In addition, when the number of electrodes is three (the third electrode potential is e3), the potential difference is calculated according to Equation 2 or Equation 3 below, and electrocardiographic data EC calculated based on the potential difference is output. .
(Number 2)
EC={(e1-e2)+(e1-e3)}/2
(Number 3)
EC=e1-(e2+e3)/2

FIG. 10 shows a configuration example of the sleep analysis unit 300, and the sleep detection information RS of the temperature data Th, acceleration data α, and electrocardiogram data EC transmitted from the sleep information detection device 100 is sent to the interface ( The electrocardiographic data EC is input to the input section 301 of the I/F), and the electrocardiographic data EC is input to the peak detection section 302 via the input section 301. The temperature data Th is input as is to the analysis processing section 320, and the acceleration data α is input to the respiration detection section 310, the posture detection section 360, and the movement detection section 370. Posture data ST indicating standing, supine, etc., detected by the posture detection section 360 is input to the analysis processing section 320 and calculated from the motion acceleration α _m , and the energy consumption EN detected by the motion detection section 370 is analyzed. It is input to the processing unit 320.

Peak detection section 302 detects peak R of electrocardiographic data EC, and peak signal PS indicating detected peak R is input to peak interval detection section 303 . The peak interval detection unit 303 detects the peak interval RRI of each data based on the peak signal PS, and the peak interval signal IPS indicating the peak interval RRI is detected by the time/frequency analysis unit 350, the heart rate detection unit 340, and the CVRR (Coefficient of Variation of RR Interval) is input to the detection unit 341. The time/frequency analysis unit 350 analyzes the frequency of the peak interval RRI with respect to time, and the wavelength analysis unit 351 inputs the analyzed frequency signal, which analyzes frequencies of 0.15 Hz or more as high wavelength HF, and analyzes frequencies of 0.15 Hz or more as high wavelength HF. A frequency smaller than 15 Hz and greater than or equal to 0.04 Hz is analyzed as low wavelength LF. The analyzed high wavelength HF and low wavelength LF are input to the autonomic nerve calculation unit 352, and the sympathetic nerve (Sympathetic Nervous System) index SNS and parasympathetic nerve (Parasympathetic Nervous System) index PSNS indicating the degree of autonomic nerve activity are calculated as follows. 4.
(Number 4)
HF=parasympathetic nerve index PSNS
LF/HF=Sympathetic nerve index SNS

That is, the parasympathetic nerve index PSNS, which is an activity index of parasympathetic nerves, is high wavelength HF, and the sympathetic nerve index SNS, which is an activity index of sympathetic nerves, is LF/HF, and these parasympathetic nerve index PSNS and sympathetic nerve index SNS are analyzed. At the same time as being input to the processing unit 320, the parasympathetic nerve index PSNS is input to the respiration detection unit 310 and the apnea syndrome detection unit 342, and is subjected to Fourier analysis. Further, the peak interval signal IPS is input to the heart rate detection section 340 and the CVRR detection section 341, and the detected heart rate HN is input to the analysis processing section 320. The peak interval signal IPS is based on the electrocardiographic data EC, and since the pulsation of the heart is related to the heart rate, the heart rate HN can be easily detected.

Furthermore, the acceleration data α from the input unit 301 is input to the respiration detection unit 310, the posture detection unit 360, and the movement detection unit 370, and the respiration rate BR detected by the respiration detection unit 310 is input to the apnea syndrome detection unit 342. At the same time, it is input to the analysis processing section 320. The energy consumption EN calculated based on the posture data ST such as standing and supine from the posture detection section 360 and the motion acceleration (α _xm , α _ym , α _zm ) from the motion detection section 370 is calculated by the analysis processing section. 320. The sleep analysis information SYA analyzed by the analysis processing unit 320 is output as sleep analysis information SY via the output unit 304 of an interface (I/F) for matching the format.

The sleep information detection device 100 according to the present invention has a built-in acceleration sensor 122, and as explained in FIGS. ₈ (A) and (B), _the The acceleration _{αz of} is detected and output. Then, among the accelerations α _x , α _y , α _z of each axis, there are gravitational acceleration α _g , motion acceleration α _m , and minute body motion acceleration α _s , and for example, the x-axis acceleration α _x is the following number 5, and during exercise, there is generally a relationship as shown in the following number 6. The same applies to the y-axis acceleration α _y and the z-axis acceleration α _z .
(Number 5)
α _x = α _xg + α _xm + α _xs
(Number 6)
α _m > α _g > α _s

The output of the minute body dynamic acceleration α _s is very small, less than 0.01 [G]. Further, the motion acceleration α _m is several [G] up to a light motion position. To detect respiration, minute body motion acceleration α _s is used for each of the x, y, and z axes, but it is necessary to detect body motion acceleration α _s in a small range of about 0.01 [G] or less. Therefore, two types of bandpass filters (BPS) are used to process the acceleration data α according to the present invention in order to increase the sampling period and increase the output gain in order to increase the output by a small amount. That is, since the respiration rate is about 10 to 40 times per minute and the frequency is about 0.17 [Hz] to 0.67 [Hz], a BPF that passes this frequency is used. Further, a low-pass filter (LPF) with a gravity of 0.01 [G] or less is used. If the noise is still large, the respiration rate (respiration frequency) BR is calculated using fast Fourier transform (FFT).

What also needs to be considered is the output of the acceleration sensor 122 when the sleep information detection device 100 is attached to the subject's chest. People can be classified into people with large airway movements (z-axis direction), diaphragm movements (y-axis directions), and rib cage movements (x-axis and z-axis directions). During an initial predetermined period of time (for example, 5 seconds) when detecting information, the acceleration data α _zs in the z-axis direction, the acceleration data α _ys in the y-axis direction, and the acceleration data α _xs in the x-axis direction are automatically compared. , the two larger acceleration data are adopted. This has a significant impact on respiratory rate detection. Experiments have shown that in general, the movement of the diaphragm (in the y-axis direction) is greater in many people, but each case is different, so the movement of the airway (in the z-axis direction), the movement of the diaphragm (in the y-axis direction), and The configuration is such that the movements of the ribs (in the X-axis direction and the Z-axis direction) are compared and the larger value is automatically selected.

An example of the configuration of the respiration detection section 310 configured based on such a premise is shown in FIG. 11 and will be described. Acceleration α (α _x , α _y , α _z ) from the acceleration sensor 122 of the sleep information detection device 100 is input to the selection unit 311 , and the two larger accelerations α _{y ,} α _z , α _x are selected for high-speed sampling. 312. The high-speed sampling unit 312 performs high-speed sampling higher than the general breathing frequency in order to obtain the minute body motion acceleration α _s of 0.01 [G] or less, and collects the high-speed sampled accelerations α _{y1 ,} α _z1 , α _x1 is input to the BPF 313. The BPF313 passes a frequency band width of 0.17 to 0.67 [Hz] corresponding to the number of breaths per minute of a normal person, and the acceleration α _y2, α that has been band-pass filtered by the BPF313. _z2 and α _x2 are input to the LPF 314. The LPF 314 passes gravity of 0.01 [G] or less, and inputs the low-pass filtered accelerations α _y3 , α _z3 , α _x3 to the fast Fourier transform unit (FTT) 315 . The FTT 315 removes noise components and outputs frequency signals α _ys , α _zs , α _xs , and the frequency signals α _ys , α _zs , α _zs are input to the comparison section 316 . The comparison section 316 automatically selects the larger frequency signal and inputs it to the respiration calculation section 317. The respiration rate BR calculated from the frequency signal by the respiration calculation unit 317 is input to the analysis processing unit 320 and is also input to the apnea syndrome detection unit 342.

The apnea syndrome detection unit 342 inputs the respiration rate BR and the parasympathetic nerve index PSNS, and generates an apnea signal when there is no respiration within a predetermined period of time (for example, 10 seconds) and the respiration rate BR becomes “0”. Output NB. Furthermore, since apnea syndrome can be determined even when the parasympathetic nerve index PSNS has no output for a predetermined period of time (for example, 10 seconds), both determinations may be performed in parallel, and any Either one may be implemented. The apnea signal NB is input to the analysis processing section 320.

FIG. 12 shows a waveform example of electrocardiographic data EC, and the peak detection unit 302 detects peaks R1, R2, . . . Rn of electrocardiographic data EC, and detects peaks R1, R2, . The signal PS is input to the peak interval detection section 303. The peak interval detection unit 303 sequentially detects the intervals between peaks R, that is, the peak interval RRI1 between peak R1 and peak R2, the peak interval RRI2 between peak R2 and peak R3, the peak interval RRI3 between peak R3 and peak R4, etc., A peak interval signal IPS indicating these peak intervals RRI is input to the time/frequency analysis section 350, the heart rate detection section 340, and the CVRR detection section 341. The rate of change in the peak interval RRIV was determined based on the peak signal PS and the peak interval signal IPS of the electrocardiogram data EC, and the activity of the autonomic nerve, called CVRR (Coefficient of Variation of RR Interval), was normalized from the rate of change RRIV. Calculate coefficient CVRR. The calculation method for the coefficient CVRR is well known, and when the variance value of the peak interval is σ and the average value of the peak interval is M, it is expressed by the following equation 7. The coefficient CVRR is input to the analysis processing unit 320 and used for determining the activity of the autonomic nerve.

図１３の波形図において、正常な被験者は、図１３（Ａ）に示すようにピーク間隔ＲＲＩ１～ＲＲＩ３に多少のバラツキがあり、異常な被験者は、図１３（Ｂ）に示すようにピーク間隔ＲＲＩ１～ＲＲＩ３がほぼ均一でバラツキがない。

In the waveform diagram of FIG. 13, a normal subject has some variation in peak intervals RRI1 to RRI3 as shown in FIG. 13(A), and an abnormal subject has a peak interval RRI1 as shown in FIG. 13(B). ~RRI3 is almost uniform and there is no variation.

It is known that the relationship between the sympathetic nerve index SNS, which is an index of sympathetic nerve activity, and sleep (non-REM sleep, light sleep and deep sleep) is shown in Figure 14. From this relationship, based on the distribution of the sympathetic nerve index SNS, Can distinguish between deep sleep and light sleep. Further, FIG. 15 shows the relationship between deep sleep and light sleep with respect to the variation of respiratory frequency VRFRE, and from this relationship, deep sleep and light sleep are distinguished based on the variation of respiratory frequency VRFRE. be able to.

Further, the influence of the sympathetic nerve index SNS and the parasympathetic nerve index PSNS on sleep has characteristics as shown in FIG. 16. That is, FIG. 16(B), in which the sympathetic nerve index SNS and parasympathetic nerve index PSNS are antagonistic, shows the characteristics of a normal sleep state, and FIG. 16(A), in which the parasympathetic nerve index PSNS is widespread. indicates good sleep. On the contrary, FIG. 16C, in which the sympathetic nerve index SNS is wide-ranging and the parasympathetic nerve index PSNS is narrow, indicates bad sleep.

Next, the operation of the motion detection unit 370 to detect the consumed energy EN based on the acceleration data α will be described.

The amount of oxygen taken into the body per kilogram of body weight is used as an index for exercise intensity (strength of exercise), but since the amount of oxygen is difficult to quantify, the unit METs (Metabolic Equivalent) is used. The unit METs is a unit that indicates exercise intensity by how many times more energy can be consumed during exercise when the oxygen intake at rest is 3.5 ml/kg/min as "1". The relationship between the motion acceleration α _m and the unit METs has a characteristic as shown in FIG. 17. Therefore, by detecting the motion acceleration α _m , the corresponding METs can be determined. In the example of FIG. 17, when the motion acceleration α _ma is 9.8 METs (161 m/min running), when the motion acceleration α _mc is 3.0 METs (normal walking). In this way, the value of the unit METs is obtained from the detected motion acceleration α _m , and the consumed energy EN is calculated according to Equation 8 below. The calculated energy consumption EN is input to the analysis processing section 320.
(Number 8)
EN = METs x time x body weight x 1.05

FIG. 18 shows a configuration example of the analysis processing unit 320, which includes a temperature change detection unit 321 that detects changes (increase, decrease) in temperature data Th and outputs a detection result DT1, and a change (increase, decrease) in heart rate HN. a heart rate change detection unit 322 that detects a change (increase, decrease) in the respiration rate BR and outputs a detection result DT2; a respiration rate change detection unit 323 that detects a change (increase, decrease) in the respiration rate BR and outputs a detection result DT3; A sleep degree determination unit 324 inputs the index SNS and the parasympathetic nerve index PSNS, determines the sleep degree based on the correlation between the two, and outputs the determination result DT4, and a change in the intersection of the sympathetic nerve index SNS and the parasympathetic nerve index PSNS. (rise, fall) and outputs a determination result DT5; and a cross determination unit 325 that determines the posture of the subject (standing, supine, lateral, etc.) based on the posture signal ST and outputs a determination result DT6. an energy consumption change detection unit 327 that detects a change (increase, decrease) in energy consumption EN and outputs a detection result DT7; The sleep determining section 328 inputs the breathing signal NB and outputs the sleep analysis information SYA.

In addition, in the detection of each of the above-mentioned changes, a threshold value for determining an increase and a threshold value for determining a decrease are respectively determined, and they may be the same value or different values.

In such a configuration, an example of its operation will be described with reference to the flowchart of FIG. 19.

First, sleep information such as temperature data Th, acceleration data α, and electrocardiogram data EC is detected by the sleep information detection device 100 (step S1), and the sleep detection information is transmitted as is from the sleep information detection device 100 or is temporarily stored in the memory 141. It is stored (step S2). Thereafter, the sleep detection information (temperature data Th, acceleration data α, and electrocardiogram data EC) RS is input to the sleep analysis unit 300 via the input unit 301 (step S10), and the peak detection unit 302 detects the peak of the electrocardiogram data EC. The peak interval detection unit 303 detects the peak interval RRI based on the peak signal PS from the peak detection unit 302 (Step S12). The peak interval signal IPS from the peak interval detection section 303 is input to the heart rate detection section 340 and the CVRR detection section 341, and the heart rate detection section 340 detects the heart rate HN (step S13), and the CVRR detection section 341 detects the heart rate HN. Coefficient CVRR is detected (step S14).

Further, the peak interval signal IPS is input to the time/frequency analysis section 350 where time/frequency analysis is performed (step S20), and wavelength analysis based on the frequency analysis is performed at the wavelength analysis section 351 (step S21). The analyzed high wavelength HF and low wavelength LF are input to the autonomic nerve calculation unit 352, and a parasympathetic nerve index PSNS and a sympathetic nerve index SNS related to the autonomic nerves are calculated (step S22). The parasympathetic nerve index PSNS and the sympathetic nerve index SNS are input to the analysis processing section 320, and the parasympathetic nerve index PSNS is input to the respiration detection section 310 and the apnea syndrome detection section 342.

The respiration detection unit 310 detects the respiration rate BR based on the acceleration data α and the parasympathetic index PSNS (step S23), and the respiration rate BR is input to the analysis processing unit 320 and the apnea syndrome detection unit 342. Ru. Posture detection section 360 detects the subject's posture based on acceleration data α (step S24), and posture data ST is input to analysis processing section 320. Furthermore, the motion detection unit 370 detects the energy consumption EN based on the acceleration data α (step S25), and the apnea syndrome detection unit 342 detects the apnea syndrome based on the respiration rate BR and the parasympathetic index PSNS (step S25). S26), the apnea signal NB indicating an apnea syndrome is input to the analysis processing section 320. The analysis processing unit 320 detects a sleep state based on each input data (step S100), and sleep analysis information SYA, which is the analyzed result, is outputted via the output unit 304 (step S200).

Note that the input of each data and the order of each process described above can be changed as appropriate.

Next, an example of the operation of the analysis processing unit 320 (step S100 in FIG. 19) will be described with reference to the flowchart in FIG. 20.

The analysis processing unit 320 first determines whether or not there is a change in temperature data Th (increase from threshold Th1, decrease from threshold Th2) (step S101), and if there is a change, changes heart rate HN (threshold HN1). If there is a change, it is determined whether there is a change in the respiration rate BR (decreasing from the threshold BR1 or rising from the threshold BR2) (step S102). Determination is made (step S103). If there is a change in the breathing rate BR, the degree of sleep (light sleep, deep sleep) is determined based on the parasympathetic index PSNS and the sympathetic index SNS (step S104), and the intersection of the parasympathetic index PSNS and the sympathetic index SNS is determined. It is determined whether there is a change (decrease from threshold EC1, increase from threshold EC2) (step S105), and if there is a change, posture detection unit 360 determines the posture of the subject (step S106). Next, it is determined whether or not there is a change in the energy consumption EN from the exercise detection unit 370 (decreased from the threshold value EN1, increased from the threshold value EN2) (step S107), and if there is a change, sleep is determined (step S110), and the determination result SY is output from the output unit 304 (step S120).

In step S101, step S102, step S103, step S105, and step S107, if there is no change, the determination process of step S110 is performed.

FIG. 21 is a timing chart showing changes in sleep information analyzed by the present invention and an example of the operation, and the operation of the present invention will be described with reference to FIG. 21.

For sleep analysis, the subject wears the sleep information detection device 100 on his chest as shown in FIG. According to the present invention, it is also easy to attach the sleep information detection device 100 to the chest. In addition, from an environmental perspective, it may be connected to the sleep analysis unit 300 in the form shown in FIG. You may do so.

Figure 21 (A) shows on the time axis the time t1 when the subject went to bed for sleep, the time t2 when the subject entered sleep, the time t3 when the subject woke up from sleep, and the time t4 when he woke up. T indicates that the apnea syndrome detection unit 342 has detected an apnea syndrome.

Before entering the bed before time t1, the subject is usually in a standing position as shown in FIG. 21(B), and the xyz axes are as described in FIG. 8(A). When entering the bed from a standing position, the body is rotated by an angle π/2 with respect to the vertical y-axis, and the body is placed in a supine position (time t1 (Bed- in)). When in bed, there is generally no movement in the vertical direction (z-axis), and movement is considered to be limited to horizontal directions (x-axis and y-axis). However, the subject's heart, airways, diaphragm, ribs, etc. are moved by autonomic nerves (sympathetic nerves, parasympathetic nerves). Also, when you fall asleep from a supine sleep state (time t3 (Bed-out)) and wake up, you rotate your body by an angle π/2 and return to a standing position (time t4 (Bed-out)) ). Such a state of the subject is detected by the posture detecting section 360 based on the posture data ST, and a determination result DT6 determined by the posture determining section 326 is input to the sleep determining section 328.

Temperature data Th, which is the subject's skin temperature, is detected by the temperature sensor 121 and input to the temperature change detection unit 321 in the analysis processing unit 320 via the control unit 140 and transmission unit 142, and the rise or fall of the temperature data Th is determined as a threshold value. Detected based on. The temperature data Th maintains a constant temperature (normal temperature) before going to bed (before time t1) and before falling asleep (before time t2), as shown in FIG. 21(D). Then, the temperature data Th gradually increases when falling asleep (after time t2), and gradually decreases to return to a constant temperature (normal temperature) when waking up from sleep (after time t3). Therefore, when the temperature change detection section 321 monitors the change in the temperature data Th and inputs the detection result DT1 to the sleep determination section 328, falling asleep and falling asleep can be detected as the first factor.

Further, the electrocardiogram data EC is inputted to the heart rate detection unit 340 as a peak interval signal IPS via the peak detection unit 302 and the peak interval detection unit 303, and the heart rate HN detected by the heart rate detection unit 340 is input to the analysis processing unit. 320, and a change (increase, decrease) in heart rate HN is detected based on a threshold value. The heart rate HN maintains a constant heart rate before going to bed (before time t1) and before falling asleep (before time t2), as shown in FIG. After time t2), the heart rate gradually decreases, and when you wake up from sleep (after time t3), it gradually increases and returns to a constant heart rate HN. Therefore, when the heart rate change detection section 322 monitors the change in the heart rate HN and inputs the detection result DT2 to the sleep determination section 328, falling asleep and falling asleep can be detected as the second factor.

The respiration rate BR is input to the respiration rate change detection unit 323 in the analysis processing unit 320, and a change (increase, decrease) in the respiration rate BR is detected based on a threshold value. The breathing rate BR maintains a constant breathing rate before going to bed (before time t1) and before falling asleep (before time t2), as shown in FIG. After time t2), the breathing rate gradually decreases, and when you wake up from sleep (after time t3), it gradually increases and returns to a constant respiration rate. Therefore, when the breathing rate change detection section 323 monitors the change in the breathing rate BR and inputs the detection result DT3 to the sleep determination section 328, it is possible to detect sleep onset and sleep deprivation as the third factor.

The sympathetic nerve index SNS and the parasympathetic nerve index PSNS are input to the sleep degree determination section 324 and the crossover determination section 325, and the sleep degree determination section 324 determines whether deep sleep is achieved based on the correlation between the sympathetic nerve index SNS and the parasympathetic nerve index PSNS. The determination result DT4 is input to the sleep determination unit 328. In addition, as shown in Figure 21 (G), before falling asleep, the sympathetic nerve index SNS is high and the parasympathetic nerve index PSNS is low, but when falling asleep, the sympathetic nerve index PSNS gradually decreases and the parasympathetic Since the neural index PSNS gradually increases and intersects, falling asleep can be detected by detecting this intersect. Similarly, when a person falls asleep, the sympathetic nerve index SNS gradually increases and the parasympathetic nerve index PSNS gradually decreases and crosses each other, so that sleep loss can be detected by detecting this crossover or change. Therefore, the crossover determination unit 325 monitors the crossover or change of the sympathetic nerve index SNS and the parasympathetic nerve index PSNS, and inputs the determination result DT5 to the sleep determination unit 328, thereby detecting sleep onset and sleep deprivation as the fourth factor. can do.

As shown in FIG. 21(B), before entering the bed (before time t1), the user is in a standing position, and as shown in FIG. 21(F), the acceleration data α indicates that the acceleration α _y with respect to gravity is 1[ G], but when the patient enters the bed, the patient's posture becomes the supine position (after time t1), so the acceleration α _y becomes "0". Then, when the person wakes up (after time t4), the acceleration α _y with respect to gravity becomes 1 [G] again. Before entering the bed (before time t1), you are in a standing position, so the acceleration α _z parallel to gravity is “0”, but when you enter the bed and are in a supine position (after time t2), the acceleration α _z is equal to the gravity Since it acts in the direction, the acceleration α _z becomes 1 [G], and when you wake up (after time t4), the acceleration α _z with respect to gravity becomes “0” again. Such a posture of the subject is determined by the posture detecting section 360 and the posture determining section 326, and becomes an important element of sleep analysis. Since the posture of the subject can be determined from the accelerations α _x , α _y , and α _z of the xyz axes, and the lateral position can also be determined, it is also possible to detect the number of times the subject turns over during sleep.

The energy consumption EN detected by the exercise detection section 370 is input to the energy consumption change detection section 327, and a change in the energy consumption EN is detected based on a threshold value. The consumed energy EN maintains a constant level before going to bed (before time t1) and before falling asleep (before time t2), as shown in FIG. (after time t2), it decreases, and when you wake up from sleep (after time t3), it increases and returns to a constant level. Therefore, when the energy consumption change detection section 327 monitors the change in the energy consumption EN and inputs the detection result DT7 to the sleep determination section 328, falling asleep and falling asleep can be detected as the fifth factor. The sleep determination unit 328 outputs sleep analysis information SYA based on the detection results and determination results. When the apnea signal NB is input to the sleep determination unit 328, sleep analysis information SYA indicating an apnea syndrome is output.

According to the sleep analysis of the present invention, when all five factors of heart rate HN, skin temperature data Th, respiration rate BR, energy consumption EN, sympathetic nerve index SNS, and parasympathetic nerve index PSNS are established, Since it determines whether someone is falling asleep or falling asleep, the analysis results are accurate. It is also possible to determine apnea syndrome, the number of times the patient turns over in bed, etc.

FIG. 22 shows an example of actual detected data, and shows an apnea attack occurring during sleep and a state of taking a shower.

1 Network 2 Terminal device 100 Sleep information detection device 101A, 101B Electrode piece 102 Charging terminal 103A, 103B Conductive material 104A, 104B Recessed part 107A, 107B Electrode material 120 Sensor section 121 Temperature sensor 122 Acceleration sensor 123 Electrocardiograph 130 Battery 140 Control section 141 Memory 300 Sleep analysis section 301 Input section 302 Peak detection section 303 Peak interval detection section 304 Output section 310 Breathing detection section 311 Selection section 312 High speed sampling section 313 BPF
314 LPF
315 Fast Fourier Transform (FFT)
316 Comparison unit 317 Breathing calculation unit 320 Analysis processing unit 321 Temperature change detection unit 322 Heart rate change detection unit 323 Breathing rate change detection unit 324 Sleep degree determination unit 325 Crossover determination unit 326 Posture determination unit 327 Energy consumption change detection unit 328 Sleep determination Section 330 Sympathetic nerve detection section 331 Fourier transform section 340 Heart rate detection section 341 CVRR detection section 342 Apnea syndrome detection section 350 Time/frequency analysis section 351 Wavelength analysis section 352 Autonomic nerve calculation section 360 Posture detection section 370 Movement detection section

Claims (7)

The sleep analysis unit determines the sleep state of the subject based on temperature data, acceleration data, and electrocardiogram data obtained by a sleep information detection device equipped with a temperature sensor, an acceleration sensor, and an electrocardiograph attached to the subject's body. A method of operating a sleep analysis device that analyzes
The temperature data is the subject's skin temperature obtained from the temperature sensor,
The acceleration data is acceleration data with the height direction of the subject as the Y axis, the left and right direction as the X axis, and the front and back direction as the Z axis,
the subject's skin temperature;
Heart rate, sympathetic nerve index and parasympathetic nerve index obtained by processing the electrocardiographic data of the subject from the electrocardiograph, and changes in the intersection of the sympathetic nerve index and the parasympathetic nerve index;
a posture of the subject and energy consumption obtained by processing the acceleration data of the subject from the acceleration sensor;
Among the acceleration data, the acceleration data in the Z-axis direction, the acceleration data in the Y-axis direction, and the acceleration data in the X-axis direction for the first predetermined time obtained when detecting sleep information are compared, and the larger Acceleration data employing two methods, and a respiration rate calculated based on the parasympathetic nerve index,
The sleep analysis unit analyzes the sleep state of the subject based on
The sleep state is going to bed, falling asleep (latency time), falling asleep after waking up from sleep, and waking up,
The sleep analysis unit determines whether to go to bed or wake up based on the posture based on the acceleration data,
Based on the decrease in the heart rate, the increase in the skin temperature, the decrease in the respiration rate, the decrease in the energy consumption, and the crossover caused by the decrease in the sympathetic nerve index and the increase in the parasympathetic nerve index , The sleep analysis unit determines the sleep onset (latency time),
Based on the increase in the heart rate, the decrease in the skin temperature, the increase in the respiration rate, the increase in the energy consumption, the increase in the sympathetic nerve index and the decrease in the parasympathetic index , The sleep analysis department determines sleep deprivation caused by awakening from sleep.
A sleep analysis method characterized by: The sleep state is REM sleep, non-REM sleep, shallow sleep, or deep sleep,
The sleep analysis unit determines the degree of sleep of the non-REM sleep (light sleep and deep sleep) and the REM sleep based on the sympathetic nerve index and/or the data of the breathing frequency fluctuation range,
The sleep analysis method according to claim 1. Based on the respiration rate, if the respiration rate becomes 0 for a predetermined period of time, and/or if the parasympathetic nerve index has no output for a predetermined period of time, the sleep analysis unit detects apnea. The sleep analysis method according to claim 1 or 2, wherein the sleep analysis method is determined to be a syndrome. A sleep information detection device that can be attached to a subject's body and includes a temperature sensor, an acceleration sensor, and an electrocardiograph;
The temperature data, acceleration data, and electrocardiogram data obtained by the sleep information detection device are acquired and processed, and are used to detect entering bed, falling asleep (latency time), falling asleep after waking up from sleep, waking up, REM sleep, non-REM sleep, and shallow sleep. , a sleep analysis unit that analyzes the sleep state of deep sleep,
The temperature data is the subject's skin temperature obtained from the temperature sensor,
The acceleration data is acceleration data with the height direction of the subject as the Y axis, the left and right direction as the X axis, and the front and back direction as the Z axis,
The sleep analysis section includes:
a posture determination unit that determines the posture of the subject (standing position, supine position, lateral position, etc.) based on a posture signal based on the acceleration data and outputs a determination result;
a temperature change detection unit that detects a change (increase, decrease) in the skin temperature data and outputs a detection result;
a heart rate detection unit that detects the heart rate based on the peak interval of the electrocardiogram data;
an autonomic nerve calculation unit that calculates a sympathetic nerve index and a parasympathetic nerve index based on the time/frequency analysis of the peak interval;
a crossover determination unit that determines a change in crossover (increase, decrease) of the sympathetic nerve index and the parasympathetic index and outputs a determination result;
Among the acceleration data, the acceleration data in the Z-axis direction, the acceleration data in the Y-axis direction, and the acceleration data in the X-axis direction for the first predetermined time obtained when detecting sleep information are compared, and the larger a respiration detection unit that detects the respiration rate based on the acceleration data and the parasympathetic nerve index employing the two;
a motion detection unit that detects energy consumption based on the acceleration data;
It consists of
Determining whether to go to bed or get up based on the posture based on the acceleration data,
Falling asleep based on the decrease in the heart rate, the increase in the skin temperature, the decrease in the respiratory rate, the decrease in the energy consumption, and the crossover caused by the decrease in the sympathetic nerve index and the increase in the parasympathetic nerve index. (latency) and
sleep based on the increase in the heart rate, the decrease in the skin temperature, the increase in the respiration rate, the increase in the energy consumption, the increase in the sympathetic nerve index and the decrease in the parasympathetic nerve index, A sleep analysis device is characterized in that it determines sleep deprivation caused by awakening. The temperature data, the acceleration data, and the electrocardiogram data from the sleep information detection device are transmitted to the sleep analysis unit by terminal connection via memory, wirelessly, wired, or optical communication. The sleep analysis device according to claim 4. Determining the degree of sleep of the non-REM sleep (light sleep and deep sleep) and the REM sleep based on the sympathetic nerve index and/or data on the breathing frequency fluctuation range;
The sleep analysis device according to claim 4 or 5. Based on the respiration rate, if the respiration rate becomes 0 for a predetermined time, or/and if the parasympathetic nerve index has no output for a predetermined time, determining an apnea syndrome. The sleep analysis device according to claim 4.

Images