JP4956137B2 - Sleep evaluation method, sleep evaluation device, sleep evaluation system - Google Patents

Description

  The present invention particularly relates to a sleep evaluation method, a sleep evaluation apparatus, and a sleep evaluation system that estimate a sleep state based on biological information during sleep and display the sleep state.

Due to recent advances in brain science, the importance of sleep has become widely recognized. Good quality sleep is indispensable for maintaining not only brain functions but also various body functions, and is important for improving the quality of life. “Sleeping better” leads to “living better”.
Therefore, there has been a need to evaluate the quality of daily sleep in ordinary households and to have good quality sleep.

Here, human sleep will be described.
Since the human cerebrum is so developed, REM sleep and non-REM sleep are differentiated, and each plays a different role.
REM sleep stems from the term Rapid Eye Movement, which moves the eyeballs under a closed eyelid. The brain at the time of REM sleep is in a state close to awakening and often has a dream, but the movement of the body is suppressed. In addition, since the autonomic nervous functions such as pulse, respiration, and blood pressure change irregularly, the body is active in a manner different from that at awakening.
Non-REM sleep is sleep that is not REM sleep, and is so-called restful sleep. During this non-REM sleep, body muscle tension is relatively maintained, and autonomic nerve functions such as pulse, respiration and blood pressure are stable.
A closer examination reveals that non-REM sleep is divided into four stages based on brain waves, each of which corresponds to a state of deep sleep from shallow slumber.

Currently, a method of determining a sleep state that is medically accepted and clinically applied is a polysomnography (PSG) method. In the polysomnography method, electroencephalogram, eye movement and mental electromyogram can be measured, and the sleep state can be determined from those waveforms.
However, methods for measuring electroencephalograms, myoelectricity, etc. require electrodes to be attached to the body of the subject, and the burden on the subject is very large. For this reason, it is impossible to measure an electroencephalogram, myoelectricity, etc. in a general household.

Therefore, several methods have been devised for estimating the sleep state by a simple method instead of the polysomnograph.
As such a method, a method of estimating from a trend calculated from a pulse rate, a method of analyzing a heartbeat fluctuation, and determining a sleep state based on autonomic nerve activity of sympathetic nerve activity and parasympathetic nerve activity (see Patent Document 1) ) Etc.
JP 2006-192152 A

However, in the method of Patent Document 1 (hereinafter referred to as Conventional Method 1), various improvements have been made, but there is a problem that it is greatly inferior in accuracy compared to the polysomnographic method. It was.
Furthermore, there has been no method for displaying the estimated sleep state in an easy-to-understand manner for the subject.

  This invention is made | formed in view of the above, Comprising: It aims at providing the technique which solves the said problem.

The invention according to claim 1 extracts pulse wave pulsation component data from the original signal obtained from the detection means, performs chaos analysis from the pulse wave pulsation component data, and determines the Lyapunov exponent or the The degree of sleep in the brain is measured from the change pattern of the Lyapunov exponent, the heart rate variability HRV analysis of the heart rate time-series data from the pulse wave pulsation component data, and the body sleep from the LF / HF component based on the result of the HRV analysis The degree of sleep of the body and the sleep degree of the brain are determined to be a sleep state corresponding to REM sleep when the sleep degree of the body is high and the sleep degree of the brain is low, In this case, the sleep evaluation method determines that the sleep state corresponds to non-REM sleep .
The invention according to claim 2, wherein the determination of sleep state, according to claim 1, characterized in that for a given time in real time, or stored the pulse wave beat component data based on It is a sleep evaluation method.
According to a third aspect of the invention, the attractor display or recurrence plot displayed by the chaos analysis is the sleep evaluation method according to claim 1 or 2, characterized in that the image drawing as the quality or sleep state of the sleep It is characterized by that.
The invention according to claim 4 is a pulse wave pulsation component data extraction unit that extracts pulse wave pulsation component data from an original signal obtained from the detection unit, a chaos analysis unit that performs chaos analysis, a Lyapunov exponent, Brain sleep level measuring means for evaluating brain sleep based on Lyapunov exponent change pattern, heart rate variability HRV analysis means for performing heart rate variability HRV analysis of heart rate time-series data, and body sleep level measuring body Sleep level measuring means, display means for displaying the sleep level of the body and sleep level of the brain, sleep level of the body, and sleep level of the brain, the sleep level of the body is high, The sleep evaluation device includes a sleep state evaluation unit that determines a sleep state corresponding to REM sleep when the sleep degree is low, and otherwise determines a sleep state corresponding to non-REM sleep .
Invention of claim 5, sleep claim 4, characterized in that the attractor display or recurrence plot displayed by the chaos analysis, further comprising an image drawing means for image drawing as a quality or sleep state of the sleep It is an evaluation device.
A sixth aspect of the present invention is a sleep evaluation system including the sleep evaluation device according to the fourth or fifth aspect.
The invention according to claim 7 is the sleep evaluation system according to claim 6 , further comprising network means and server means.

According to the sleep evaluation method, the sleep evaluation apparatus, and the sleep evaluation system of the present invention, it is possible to make a determination regarding REM sleep and non-REM sleep by determining the sleepiness level of the body and the sleepiness level of the brain representing the sleep state.
In addition, it is possible to visually express the quality of sleep with an image displayed by attractor display and recurrence plot display by chaos theory.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
First, as shown in FIG. 1, a bedding 2 such as a mattress, a bed, a futon, a pillow, a cushion, a cushion, and a sofa is equipped with a piezoelectric sensor 1, and an output of an original signal of the piezoelectric sensor is measured by a measuring device. 3 is input. The measuring device 3 performs various processes from the pulse wave signal, evaluates the sleep of the subject, and displays the evaluation on the personal computer 4 that is a display unit.

Next, as shown in FIG. 2, the piezoelectric sensor 1 is a piezoelectric sensor that can acquire the pulse wave of the subject. In the analog circuit, power noise and the like are cut, and components due to the pulsation of the living heart That is, after amplification through a high-pass filter and a low-pass filter that pass around 1 Hz, A / D (analog / digital) conversion is performed at 500 Hz and 12 bits.
The measurement device 3 includes a pulse wave pulsation component data extraction unit 11, a chaos analysis unit 12, a brain sleepiness measurement unit 13, an HRV analysis unit 14, a body sleepiness measurement unit 15, and a sleep state evaluation unit. 16 is composed. In addition, data of each unit is controlled by a control unit or a storage unit (not shown) and is temporarily accumulated. The pulse wave pulsation component data extraction unit 11 extracts pulse wave pulsation component data from the original signal obtained from the piezoelectric sensor 1. The chaos analysis unit 12 performs various chaos analyses. The brain sleepiness measuring unit 13 evaluates the brain sleepiness from the Lyapunov exponent or the change pattern of the Lyapunov exponent, which is the result of the chaos analysis. The HRV analysis unit 14 performs heart rate variability (HRV) analysis of heart rate time series data. The body sleep level measuring unit 15 quantitatively evaluates the body sleep level based on the result of the HRV analysis. The sleep state evaluation unit 16 determines the sleep state from the sleep level of the brain and the sleep level of the body.
The display unit 4 that displays sleep evaluation is a device having an intelligent display function such as a PC, and displays a sleep evaluation that is graphical and has a high visual effect.

Next, the operation | movement at the time of the data analysis of the sleep evaluation apparatus in embodiment of this invention is demonstrated in detail with reference to the flowchart of FIG.
First, in step S1, the pulse wave pulsation component data extraction unit 11 performs a filtering process on the original signal of the piezoelectric sensor. In this process, the measurement device 3 extracts pulse wave pulsation component data due to the pulsation of the heart near 1 Hz through an FIR bandpass filter that passes 0.7 Hz to 1.4 Hz with respect to the digitized data. .

  From here, the flowchart is broadly divided into chaos analysis embedded in a multidimensional phase space according to the Turkens theorem and peak analysis for peak detection. The chaos analysis that is the flow on the right side of the page will be described first.

In step S <b> 2, the chaos analysis unit 12 performs chaos analysis in which pulse wave pulsation component data is embedded in a multidimensional phase space according to the Turkens theorem.
The Turkens theorem is a method commonly used in chaos analysis. FIG. 4 shows an outline of turnens embedding. When the time series data is x (k) (k = 0,1,2,3...) And m state variables are to be restored, the vector x (i) = {x ( i), x (i + τ), x (i + 2τ), ... x (i + mτ)}. For example, when the number of state variables is 3, x (i) = {x (i), x (i + τ), x (i + 2τ)}.
This τ is a parameter and is called an embedding delay time. When this vector x (i) is sequentially plotted in the three-dimensional state space (coordinate axes are x (i), x (i + τ) and x (i + 2τ)) (i = 0,1,2) ,..., N), the trajectory as shown on the right side of FIG. The shape of this orbit is called an attractor.
The calculated trajectory can be used to display “sleeping art” in another embodiment of the present invention described later.

Next, in step S3, the chaos analyzer 12 calculates the Lyapunov exponent λ.
The Lyapunov exponent is one of the chaos indices, and is a numerical value indicating the degree of dependence of the chaos on the initial value. The distance between two adjacent trajectories drawn by the attractor increases with time. It is an amount indicating the degree of separation.

For the calculation of the Lyapunov exponent λ, there are a Wolf method, a Kantz method, and the like. From the viewpoint of calculation speed and accuracy, an approximate calculation method by the Sano-Sawada (Sano-Sawada) method is used.
FIG. 5 shows a conceptual diagram of the three-dimensional case of the Lyapunov exponent calculation algorithm of the Sano-Sawada (Sano-Sawada) method. Assuming that a microsphere (hypersphere) with a radius ε is given as an initial value to a three-dimensional chaotic dynamical system, what was initially a sphere is stretched in the e1 direction by being mapped once, and the e3 direction The result is an ellipsoid. If the logarithms of the enlargement rates per unit time in the e1, e2, and e3 directions are λ1, λ2, and λ3, these λ1, λ2, and λ3 are called Lyapunov exponents, and these sets are called Lyapunov spectra. In practice, the operation of calculating the hypersphere is repeated for all points of the attractor, and the Lyapunov spectrum is calculated from the average of the sum of these.
The Lyapunov exponent usually means the first component of the Lyapunov spectrum, and is called the “first Lyapunov exponent” or the “maximum Lyapunov exponent” because it is the first component of the Lyapunov spectrum. ".

In the sleep evaluation device according to the embodiment of the present invention, the Lyapunov exponent λ is calculated by sliding calculation.
Referring to the example of the sliding calculation in FIG. 6, the sliding calculation is performed by using a window width W as the time width and the number S of sliding points, starting from the first point of the data and using all data at the W point. Next, it is a process of repeating the same process using all the data of the W point from the point slid by the S point from the first point of the data. Thereby, the calculation result can be displayed with a time delay of about half of the window width W.

Next, in step S <b> 4, the brain sleep level measuring unit 13 calculates “brain sleep level” from the Lyapunov exponent λ.
Recent research results of fingertip pulse wave chaos analysis suggest that the Lyapunov exponent λ value by chaos analysis of pulse wave pulsation component data reflects the central brain emotion, excitement, and suppression. ing.
That is, it is considered that the greater the Lyapunov exponent λ value of the pulse wave pulsation component data, the more excited the brain, and the smaller the Lyapunov exponent λ value, the more suppressed the brain activity.
Therefore, it is assumed that the smaller the Lyapunov exponent λ, the deeper the “sleeping degree of the brain”, and the higher the Lyapunov exponent λ, the shallower the “sleeping degree of the brain”.
In the apparatus according to the embodiment of the present invention, the above-mentioned Lyapunov exponent λ is calculated in four steps with respect to the “degree of sleep of the brain” by providing three threshold values to obtain a measurement result.
FIG. 7 shows a calculation example of the “brain sleep level” expressed in these four stages. In the display unit described later, it is also possible to display “brain sleep level” as in this calculation example.

Next, peak analysis, which is the flow on the left side of the page, in the flowchart of FIG. 3 will be described.
First, in step S5, the HRV analysis unit 14 detects a peak from the pulse wave pulsation component data and creates PP interval data corresponding to the RR interval of the electrocardiogram.
The relationship between the RR interval and the PP interval will be described with reference to FIG. 8. The RR interval and PP interval, which are one beat on the electrocardiogram, correspond to each other although their peak positions are off. is doing.
Using this P-P interval data, sliding calculation of HF and LF analysis of heart rate variability HRV analysis is performed. The sliding calculation is performed in the same manner as the sliding calculation of the Lyapunov exponent λ described above.

Next, in step S6, the HRV analysis unit 14 performs HRV analysis of peak analysis on the above-described pulse wave pulsation component data. The HRV analysis is a heartbeat variability analysis.
First, the discrete values of the P-P interval are interpolated by linear interpolation, FFT (Fast Fourier Transform) processing is performed on the interpolated curve, spectrum analysis is performed, and the heartbeat fluctuation component is extracted as a peak on the frequency axis.
Referring to FIG. 9A, there are shown a fluctuation waveform at the PP interval and a waveform of each fluctuation component when the fluctuation waveform is decomposed into the above three frequency components from the RR interval. Similarly to the fluctuation waveform of the RR interval, each frequency component can be extracted from the PP interval.

Here, the HRV analysis unit 14 extracts “LF / HF” from the peak.
First, when the spectrum analysis of the heartbeat variability described above is performed, it can be seen that the variability is composed of waves of a plurality of frequencies. These are classified into the following three types of fluctuation components.
1. HF (High Frequency) component, which is a fluctuation that coincides with respiration 2. LF (Low Frequency) component that fluctuates in a cycle of around 10 seconds Trend component that fluctuates at a frequency lower than the measurement limit FIG. 9B shows the result of spectrum analysis for the fluctuation waveform of the PP interval shown in FIG.
As can be seen from this figure, peaks are observed at two frequencies near 0.07 Hz and 0.25 Hz. The former is the LF component and the latter is the HF component. Note that the trend component cannot be read from the figure because it is below the measurement limit.

Next, in step S <b> 7, the body sleep level measuring unit 15 calculates the “body sleep level” from the LF component and the HF component.
First, the above-described LF component represents the degree of sympathetic nerve tension, and the greater the amplitude of this component, the greater the tension.
On the other hand, the HF component represents the degree of tension of the parasympathetic nerve, and the larger the amplitude of this component, the more relaxed it is.
Since there are individual differences in the amplitude values of the LF component and the HF component, LF / HF, which is the amplitude ratio of the LF component and the HF component, is useful for estimating the tension of the subject.
That is, it can be said that the greater the LF / HF value, the higher the degree of tension, and the smaller the LF / HF value, the lower the degree of tension and the more relaxed.

Here, the smaller the LF / HF value is, the more relaxed it is, so the “body sleepiness” is deeper, and the higher the LF / HF value, the more nervous, so “body sleepiness” is shallower.
In the apparatus according to the embodiment of the present invention, as for the “sleeping degree of the body”, LF / HF is calculated in four stages by setting three thresholds as the “sleeping degree of the brain” to obtain a measurement result. .
FIG. 10 shows an example in which the “degree of sleep” is displayed at four levels.

Finally, in step S8, the sleep state evaluation unit 16 evaluates the sleep state corresponding to the REM sleep and the non-REM sleep using the “body sleep” and “brain sleep”.
In general, non-REM sleep is a state where the muscles that maintain the body posture (anti-gravity muscles, posture muscles) are maintained, the body is awake, but the brain is sleeping, and almost no dreams are seen State.
On the other hand, REM sleep is said to be a dreaming state in which the brain is awake while the body is asleep with almost no muscle tension maintaining the body posture.
Therefore, it can be inferred that non-REM sleep is when “sleepiness” is small and “sleepiness” is high, and REM sleep is when “sleepiness” is large and “brain sleep” is low. .
The sleep state evaluation unit 16 outputs the sleep state evaluation corresponding to the non-REM sleep and the REM sleep to the display unit 4.
Thereby, a series of sequences of sleep state evaluation is completed.

  The above steps can be performed in real time or for a predetermined time based on the stored pulse wave component data.

By configuring as described above, the following remarkable effects occur.
First, in the conventional method, the sleep depth, sleep state, REM sleep, and non-REM sleep are estimated only from the function of the autonomic nerve.
However, in the sleep evaluation device according to the embodiment of the present invention, the “sleeping degree of the body” is calculated from the tension / relaxation of the autonomic nerve, and the “sleeping degree of the brain” is numerically expressed from the emotion / excitement / suppression of the central brain. From the relationship between the two, REM sleep and non-REM sleep are estimated.
That is, by introducing an index of “brain sleep level” as well as “body sleep level”, it is possible to determine the sleep state with higher accuracy than before without imposing a burden on the subject.

In addition, the Lyapunov exponent calculation of the chaos theory of the prior art is for the time series of heart rate, and in order to calculate one index, data of a length of about 15 minutes was required, so in real time The result could not be calculated.
Actually, a sufficient number of measurement points are required to calculate the Lyapunov exponent, and about 15 minutes of data are required to obtain one Lyapunov exponent λ. For this reason, it was difficult to calculate a result in real time.
However, in the method according to the embodiment of the present invention, it is possible to evaluate sleep for every arbitrary unit time in real time by performing sliding calculation.

In addition, the chaos analysis of the biological data can be calculated not only on the P-P interval time series data but also directly using the time series data of the pulse wave pulsation component or the time series data of the electrocardiogram. know.
Therefore, if the Lyapunov exponent λ of the pulse wave pulsation component data is calculated, it is possible to calculate one exponent with time series data of about 1 minute to 2 minutes. That is, the time delay due to the sliding calculation can be suppressed to about 1 minute.
In this sliding calculation, it is desirable to perform calculation by discarding data of 5 minutes or less for data with a lot of noise. In addition, data of 5 minutes or more can be supplemented from previous and subsequent data. Furthermore, since data with a lot of noise is often recorded when turning over, the number of turning over can be measured by counting the number of times of noise.

Furthermore, since “the sleep level of the body” and “the sleep level of the brain” can be calculated at the same time, the result can be drawn as a graph to determine the sleep state in more detail.
FIG. 11 is an example of this graph. The horizontal axis represents the sleep level of the body as four levels, and the vertical axis represents the sleep level of the brain as four levels.
The quadrant in the lower right of the graph (the brain sleep level is 1-2 and the body sleep level is 3-4) is REM sleep, and other values are non-REM sleep.

Here, in the apparatus according to the embodiment of the present invention, as described above, it is possible to measure the "body sleep" and "brain sleep" in real time, so when the sleep state changes, You can create a reporting system.
Furthermore, since the non-REM sleep of the above-mentioned graph is spread over three quadrants, there is a possibility that new knowledge about sleep can be obtained by evaluating these as separate sleep states.
Also, by looking at this graph, the observer can check what sleep state the subject is in. Thereby, for example, the wife can check the sleeping state of the husband, or the husband can check the sleeping state of the wife.
You can wake up at a comfortable waking time or check the quality of your sleep.
Furthermore, it has been understood that the Lyapunov exponent based on the chaotic theory of pulse wave pulsation component data reflects the activity related to the brain emotion, which is the upper center (Ken Seino, Yoshiharu Yamamoto, Fluctuation Science and Technology- Introducing Chuomatics-”Tohoku University Press 2004, P71-P90, see), and you can get clues about the subject's psychological state.
That is, it is possible to create a system that gives a stimulus such as an odor during REM sleep so as to have a good dream.

In addition, by measuring the amount of sleep equivalent to non-REM sleep and REM sleep from the average of "sleepiness of the body" and "sleepiness of the brain" overnight, a clue to know whether the subject's sleep was comfortable can get.
In addition, it is possible to draw a state in which the above-mentioned graph changes in real time in a layered manner, and it is conceivable to construct a system for evaluating changes in sleep quality from this pattern.
For example, it is generally said that REM sleep and non-REM sleep are healthy four times a night in a 90 minute cycle, but data can be easily taken to determine if this is correct.

<Other embodiments>
As for the output of the sleep evaluation apparatus according to the above-described embodiment of the present invention, it is sufficient if a graph can be displayed as shown in FIGS. 7 and 10, but this can be output in a more graphical and varied manner. it can. An example of this display method will be described with reference to FIG. 12 as an example according to another embodiment of the present invention.
The upper three in FIG. 12 are examples of attractor display. Multi-dimensional attractor trajectory data is projected to three dimensions, and the value of the third dimension is graphically displayed in color as warm color to cold color information.
The bottom three in FIG. 12 are called recurrence plot displays, which are techniques for visualizing the structure of attractors. The recurrence plot is a method for visualizing the correlation between each point on the attractor, and is an excellent method for detecting non-stationarity of a time series signal.
In order to create a recurrence plot, first, a two-dimensional image in which the length of one side is the total number N of points on the attractor is prepared. Next, the distance D (i, j) = | v (i) −v (j) | between the two points on the attractor is calculated. An image of N × N pixels created based on D (i, j) is a recurrence plot. This recurrence plot is displayed graphically in warm to cold colors.
The above-mentioned attractor display and recurrence plot display also perform sliding calculation and display information at an arbitrary time in real time.

By configuring as described above, the following remarkable effects occur.
First, by showing the attractor display image and recurrence plot display as images rather than numerical values, it is possible to appeal to the human sense as an image. In other words, it can express sleep art, "sleep art".
For example, in the case of FIG. 12, the awakening is close to “body sleep” and “brain sleep” are shallow “shallow sleep”, “deep sleep” corresponding to non-REM sleep, and “dream sleep” corresponding to REM sleep. Attractor and recurrence plots were displayed for each state. It can be seen that the three sleep states are displayed in completely different patterns.
Such interesting display changes can be expected to attract the subject's interest in sleep and provide an opportunity to think about sleep and health.
Furthermore, this “sleep art” has the practicality that if you get used to it, you can intuitively understand the state of sleep. That is, when “sleep art” is observed on a daily basis, if the sleep state is abnormal, it is immediately noticed, so an effect on preventive medicine can be expected.
In addition, the “sleep art” can be displayed with the graph of the temporal change in sleep evaluation described above.

Note that the configuration and operation of the above-described embodiment are examples, and it is needless to say that the configuration and operation can be appropriately changed and executed without departing from the gist of the present invention.
For example, in the embodiment of the present invention, the sensor provided in the bedding is a piezoelectric sensor, but another type of pressure sensor, pulse wave sensor, or electrocardiogram sensor may be used. Also, various materials such as blankets and duvets can be used for the bedding.

As described above, the sleep evaluation method, the sleep evaluation device, and the sleep evaluation system according to the present invention can be easily used in daily life, are useful for high-precision sleep state estimation, and bedding manufacturers and the like develop and improve bedding that is comfortable to sleep. To help.
In addition, since the health level can be estimated from the data at the time of daily sleep, the health check can be performed without taking the trouble of the measurer. In addition, the quality of daily sleep is evaluated in a general home, and it helps to perform good quality sleep.
Moreover, there is a possibility that it can be used as a new medical treatment method not only by checking the health level of healthy persons but also by measuring the data of inpatients in hospital facilities and administering the data while analyzing the data.
Furthermore, it is possible to visually appeal the state of sleep by “sleep art”.

It is a figure explaining the sleep evaluation apparatus which concerns on embodiment of this invention. It is a block diagram of the sleep evaluation apparatus which concerns on embodiment of this invention. It is a flowchart which shows the operation | movement at the time of the analysis which concerns on embodiment of this invention. It is a figure explaining attractor embedding by the Turkens theorem concerning an embodiment of the invention. It is a figure explaining the Lyapunov exponent calculation algorithm which concerns on embodiment of this invention. It is a figure explaining the sliding calculation which concerns on embodiment of this invention. It is a figure which shows the example of the "brain sleep degree" result graph which concerns on embodiment of this invention. It is a figure explaining the RR interval and PP interval which concern on embodiment of this invention. It is a figure which shows the relationship between the (a) RR space | interval fluctuation | variation which concerns on embodiment of this invention, and the frequency component which comprises this fluctuation. (B) It is a figure which shows the result of having performed the spectrum analysis of RR space | interval fluctuation | variation. It is a figure which shows the example of the "body sleepiness" result graph based on embodiment of this invention. It is a figure which shows the example of the graph of the "degree of sleep of a brain" and the "degree of sleep of a body" concerning embodiment of this invention. It is a figure which shows the example of a display of "sleeping art" concerning other embodiment of this invention.

Explanation of symbols

1 Piezoelectric sensor (detection means)
2 Bedding 3 Measuring device 4 Display unit (Display means, image drawing means)
11 Pulse wave pulsation component data extraction unit (Pulse wave pulsation component data extraction means)
12 Chaos analysis part (chaos analysis means)
13 Brain sleep level measuring unit (Brain sleep level measuring means)
14 HRV analysis unit (Heart rate variability HRV analysis means)
15 Body sleep level measuring unit (Body sleep level measuring means)
16 Sleep state evaluation part (sleep state evaluation means)

Claims (7)

Extract pulse wave pulsation component data from the original signal obtained from the detection means,
Chaos analysis from the pulse wave pulsation component data,
Measure the sleepiness of the brain from the Lyapunov exponent or the change pattern of the Lyapunov exponent from the result of the chaos analysis,
Heart rate variability HRV analysis of heart rate time-series data from the pulse wave pulsation component data,
From the result of the HRV analysis, the sleepiness of the body is measured from the LF / HF component ,
Regarding the sleep level of the body and the sleep level of the brain, when the sleep level of the body is high and the sleep level of the brain is low, it is determined as a sleep state corresponding to REM sleep, and in other cases, it is determined as non-REM sleep. A sleep evaluation method characterized by determining a corresponding sleep state . The sleep evaluation method according to claim 1 , wherein the determination of the sleep state is performed in real time or for a predetermined time based on the stored pulse wave pulsation component data. The sleep evaluation method according to claim 1 or 2 , wherein an image of the attractor display or recurrence plot display by the chaos analysis is drawn as sleep quality or sleep state. Pulse wave pulsation component data extraction means for extracting pulse wave pulsation component data from the original signal obtained from the detection means;
Chaos analysis means for chaos analysis,
A means of measuring the sleepiness of the brain for evaluating the sleepiness of the brain from the Lyapunov exponent or a change pattern of the Lyapunov exponent;
Heart rate variability HRV analysis means for performing heart rate variability HRV analysis of heart rate time series data;
A means for measuring sleepiness of the body for measuring sleepiness of the body;
Display means for displaying the sleepiness of the body and sleepiness of the brain ;
Regarding the sleep level of the body and the sleep level of the brain, it is determined that the sleep state corresponding to REM sleep when the sleep level of the body is high and the sleep level of the brain is low, otherwise it corresponds to non-REM sleep The sleep evaluation apparatus characterized by including the sleep state evaluation means to determine with the sleep state to do. The sleep evaluation apparatus according to claim 4 , further comprising an image drawing unit that draws an image of the attractor display or recurrence plot display by the chaos analysis as sleep quality or sleep state. A sleep evaluation system comprising the sleep evaluation device according to claim 4 . The sleep evaluation system according to claim 6 , further comprising network means and server means.

Images