JP2013078543A - Autonomic nerve activity index calculation method and autonomic nerve activity index calculation device, and visual display system for mind-body balance using the same - Google Patents

Abstract

PROBLEM TO BE SOLVED: To visually recognize mind-body balance based upon an autonomic nerve activity index by easily measuring the autonomic nerve activity index in daily life without using a large-scale electrocardiograph.SOLUTION: Data are acquired from a human recorder including a three-axis acceleration sensor and a heart beat sensor, abnormal values are removed from RR values obtained from the heart beat sensor, and RR-interval time-series data are generated through equal-interval sampling. An an autonomic nerve activity index is calculated by applying fast Fourier transformation, activity determinations on actions such as a standstill state, a walking state, and a fast walking state are made based upon acceleration information, and a calorie consumption is calculated based upon a preset movement coefficient and a basal metabolic coefficient to calculate an activity quantity. The mind-body balance is displayed while being classified into four types according to a combination of whether the action determination result is large or small and whether the activity quantity calculation result is large or small.

Description

  The present invention relates to a human recorder incorporating a composite sensor and a “mental health” measurement technique (mental stress measurement technique) using a sympathetic nerve activity index and a parasympathetic nerve activity index calculated using the human recorder.

  In the autonomic nerve test, CVRR (electrocardiogram RR interval variation coefficient) using an electrocardiograph is well known as a test for cardiac parasympathetic nerve function. In CVRR, after resting for about 15 minutes, the electrocardiogram at rest and about 6 deep breaths per minute are compared. About 100 heartbeats are analyzed, the average value of the RR interval is M, the standard deviation is SD, and CVRR is defined as SD / M × 100. In the case of neurodegenerative diseases, CVRR tends to decrease.

Japanese Unexamined Patent Application Publication No. 2011-120618 (Patent Document 1) is a prior art technique for obtaining a sympathetic nerve activity index from CVRR based on such electrocardiographic data of a measurement subject. In this document,
The electrocardiogram data of the measurement subject is measured by an electrocardiogram monitor, and the high frequency component HF, the low frequency component LF, and the CVRR of the heart rate variability are calculated based on the measured electrocardiogram data. Based on these calculated values, a sympathetic nerve activity index (LF / HF) and a parasympathetic nerve activity index (CVRR × LF / HF) are calculated, and the calculated sympathetic nerve activity index and sub-exchange are calculated. The relationship with the nerve activity index was displayed two-dimensionally on the display device.

JP 2011-120618 A

  However, in the acquisition of electrocardiographic data using the electrocardiograph as described above, it is appropriate in principle to perform measurement while the subject is kept at rest for 10 minutes or more, and it is expensive and large-scale installed in medical facilities. Therefore, it was difficult to calculate the sympathetic nerve activity index and the parasympathetic nerve activity index in the state where the subject normally spends daily life at home.

  The present invention has been made in view of these points. First, acquisition of a sympathetic nerve activity index and a parasympathetic nerve activity index (hereinafter collectively referred to as “autonomic nerve activity index”) is also obtained in daily life. It is a technical problem to realize a human recorder that can be used, and secondly, to realize a visual output of “balance between heart and body” based on an autonomic nerve activity index obtained from the human recorder.

The present invention employs the following means in order to solve the above problems.
Claim 1 of the present invention relates to a balance display method for a heart and a body, and includes a heart rate sensor, a triaxial acceleration sensor, a control means, a clock means, and a storage means, and is mounted on the surface of the human body of the subject. A heart and body balance display method executed by a human recorder and an information processing apparatus that calculates and displays various information from values of each means obtained from the human recorder, wherein the heart rate sensor is controlled by the control means RR interval time series data is obtained from the three-axis acceleration sensor, and the heart rate value and acceleration information are temporarily stored in the storage means, and then synchronized with the clock means. RR
Interval time series data and acceleration information are output from the output means, and in the information processing apparatus, abnormal values are removed from the RR values obtained from the heart rate sensor, and the RR interval time series is sampled at equal intervals. Forming data, applying fast Fourier transform to calculate an autonomic nerve activity index, performing at least a stationary state, a walking state, a fast walking state, and the like based on the acceleration information, and the behavior determination From the combination of the step of calculating calorie consumption based on a preset motion coefficient and basal metabolism coefficient from the result and calculating the amount of activity, the magnitude of the action determination result, and the magnitude of the activity amount calculation result This is a balance display method for the body and the body, which consists of the steps of classifying and displaying the four types of body balance.

  Claim 2 of the present invention relates to a visual display system for the balance between the heart and the body, and includes a heart rate sensor, a three-axis acceleration sensor, a control means, a clock means, and a storage means, and a person to be measured. A visual display system for balance between the body and the body executed by a human recorder mounted on the surface of the human body and an information processing device that calculates and displays various information from the values of each means obtained from the human recorder. In the human recorder, RR interval time series data is acquired from the heart rate sensor, the control means for acquiring acceleration information of the measurement subject from the triaxial acceleration sensor, and the heart rate value and acceleration information are temporarily stored in the storage means. After that, the RR interval time-series data synchronized with the clock means and the acceleration information are output to the information processing apparatus. In the information processing apparatus, the abnormal value is removed from the RR value obtained from the heart rate sensor, the RR interval time-series data is formed by sampling at an equal interval, and an autonomic nerve activity index is applied by applying a fast Fourier transform And calculating at least calorie consumption based on the preset motion coefficient and basal metabolism coefficient from the behavior determination result. Then, the control means for calculating the amount of activity and the calculation result by the control means are classified into four types of balance between the heart and the body from the combination of the magnitude of the action determination result and the magnitude of the activity amount calculation result. This is a visual display system for the balance between the body and the body, comprising display means for displaying the image.

  A third aspect of the present invention relates to a sleep determination method, which includes a heart rate sensor, a triaxial acceleration sensor, a control unit, a clock unit, and a storage unit, and is mounted on the surface of the human body of the measurement subject. A sleep determination method that is executed by a human recorder and an information processing device that calculates and displays various types of information from the values of each means obtained from the human recorder, wherein the control means causes an RR for a predetermined time from the heart rate sensor. Interval time series data is obtained, acceleration information of the measurement subject is obtained from the three-axis acceleration sensor, and the RR interval time series data is temporarily stored in the storage means, and then synchronized with the clock means. Series data and acceleration information are output from the output means, and in the information processing apparatus, RR interval time series data obtained from the heart rate sensor, Falling asleep and a behavior determining result based on the acceleration information is determined sleep state determining how the sleep of wake or sleep rhythm.

  Claim 4 of the present invention relates to sleep determination, wherein the change in body posture is not more than a predetermined value from the posture information, and the heart rate standard deviation of the RR interval time series data from the heart rate sensor at that time is 25 to 75% or less. The sleep state determination method according to claim 3, wherein the sleep state is determined when it becomes.

  Claim 5 of the present invention relates to wakefulness determination, which is within 45 to 135 minutes before detecting the sitting or standing position from the posture information, and the RR interval time series obtained from the heart rate sensor. The standard deviation of the data exceeds a predetermined average value, and the time when the difference (slope) before and after the heartbeat standard deviation is within 0.1 to 0.3 is continuously 2 minutes before and after (total 5 minutes) or more. The sleep state determination method according to claim 3, wherein wakefulness is determined when there are two or more changes in posture obtained from posture information between 10 minutes before and after.

  Claim 6 of the present invention relates to sleep cycle rhythm determination, and generates a sleep graph by smoothing an autonomic nervous activity index obtained by applying a fast Fourier transform from the RR time series data in a predetermined time unit, The sleep state determination method according to claim 3, wherein a plurality of peaks are searched from the sleep rhythm graph and a sleep cycle rhythm is detected at the peak interval.

  According to the present invention, it is possible to realize a human recorder that can acquire an autonomic nerve activity index even in daily life. In addition, it is possible to visually grasp the balance between the heart and the body based on the autonomic nerve activity index obtained from the human recorder.

The internal block diagram of the human recorder which is embodiment of this invention 1 is an external view of a human recorder according to an embodiment of the present invention. Back view of the human recorder of the embodiment 1 is a block diagram showing the configuration of an information processing apparatus according to an embodiment Explanatory drawing which shows the abnormal value removal filter table of embodiment The graph which shows the RR interval time series data of embodiment The graph which shows the sympathetic nerve activity parameter | index (LF / HF) of embodiment. Explanatory drawing which shows the action determination of embodiment Table showing coefficient of motion for calorie consumption calculation Table showing the basal metabolism coefficient by gender and age of the embodiment Explanatory drawing which shows the state which displayed the balance of the heart and body in embodiment according to the type visually The graph which shows the sleep onset judgment (heart rate standard deviation at the time of sleep) in embodiment The graph which shows arousal determination (heartbeat standard deviation at the time of sleep) in embodiment The graph figure which shows the sleep rhythm determination in embodiment

The present invention will be described with reference to the drawings.
The system according to the present embodiment includes a human recorder and an information processing apparatus.

  FIG. 1 is a block diagram of a human recorder used in this embodiment, and FIG. 4 is a block diagram of an information processing apparatus.

  The human recorder (HR) has a user memory (UMEM) connected by a bus (BUS) with a central processing unit (CPU) and a main memory (MM) as the center, and an application system together with an operating system (OS). A program (APL) and a database (DB) are stored.

  The bus (BUS) is further equipped with a heart rate detection pad (PAD) via a body surface temperature sensor, a triaxial acceleration sensor, a heart rate sensor, and an interface (I / O).

  The human recorder (HR) is provided with a USB interface and a wireless interface (IF) as an external communication interface, and its housing is provided with a power switch (SW) on its upper surface as shown in FIG. The USB interface has two electrodes (ELT) on the back (Fig. 3). A heart rate detection pad (PAD) is detachably attached to the two electrodes (ELT) by a snap method.

The information processing device (PC) connected to the human recorder (HR) is a general-purpose personal computer, and as shown in FIG. 4, a large-scale storage device centered on a central processing unit (CPU) and a bus (BUS). (HD), display (DISP) as output device, keyboard (KBD) as input device, mouse (MOU) as auxiliary input device, USB interface (USB I / O) as external communication interface, wifi or It has a wireless interface (IF) based on Bluetooth communication. The large-scale storage device (HD) is provided with an operating system (OS), an application program (APL), a database (DB), and various tables (TBL).

  The human recorder (HR) is worn on the chest of the measurement subject and acquires heart rate data (RR interval time series data) in real time. Specifically, the information processing apparatus that has received the heartbeat data via the USB interface (USB I / O) or the wireless interface (IF) stores the heartbeat data in a memory (MM), and An autonomic nerve activity index is calculated according to the program (APL). This process is explained below.

  First, an abnormal value is removed from RR interval time series data acquired from a human recorder (HR). To remove this abnormal value, an abnormal value removal filter table stored in a large-scale storage device (HD) is used. FIG. 5 shows this abnormal value removal filter table.

From the RR interval time-series data obtained from the human recorder (HR), the instantaneous heart rate (HR) at the average beat of 8 beats is calculated. And H.H. For 8 beats of R, meanH. R (8) is calculated.

(Formula 1) | 60 / t (n) 60 / t (n1) | ≦ 15

meanH. R (8) is calculated as a moving average. Therefore, meanH. R (8) will change. This meanH. After calculating R (8), the difference from the next heartbeat to be compared is obtained, and | meanH. When R (8) 60 / t | ≧ 30, the instantaneous heart rate is removed as an error value.

  Since the RR interval time-series data from which abnormal values are excluded in this way is data sampled at unequal intervals, it cannot be applied to the fast Fourier transform. Accordingly, linear interpolation (linear interpolation) and cubic spline interpolation (curve interpolation) are applied to the data to resample at equal intervals. If the sampling interval is short, the accuracy of the calculation result is high, but there is a weak point that the performance is lowered. On the other hand, when the sampling interval is increased, the accuracy of the calculation result is lowered but the performance is improved. In this embodiment, the data range is 60 seconds, and the sampling interval is 0.25 seconds. As a result, the point data is 60 × 1 / 0.25 = 240 points. In order to make this a power of 2, 16 pieces of data having an RR interval value of 0 are supplemented to create 256 points of data (2 ^ 8 = 256240 = 16).

  FIG. 6 shows an example of RR interval time-series data obtained in this way with a data range = 60 seconds and a sampling interval = 0.25 rows.

  A fast Fourier transform (FFT) is applied to the target data by applying a window function such as a Hanning window or a mamming window. This application result is shown in FIG.

From this conversion result, a low frequency component (LF), a high frequency component (HF), and a sympathetic nerve activity index (LF / HF) are calculated. The low frequency component (LF) is an integral value of a power spectrum of 0.04 to 0.15 Hz, and the high frequency component (HF) is an integral value of a power spectrum of 0.15 to 0.4 Hz.

  Next, the information processing apparatus performs behavior determination based on data obtained from a three-axis acceleration sensor of a human recorder (HR).

  First, the amount of activity of the measurement subject is calculated from each acceleration data (three-axis composite value) in the three-dimensional direction obtained from the three-axis acceleration sensor. Specifically, as the first step, the amount of activity is determined as one of a stationary state, a walking state, a fast walking state, and a running state from the three-axis composite value of the three-axis acceleration sensor.

  Specifically, the triaxial composite value of 0 to 0.9G is stationary, 0.9 to 1.2G is stationary, 1.2 to 1.5G is walking, and 1.5 to 1.7G is early. The walking state 1.7 to 1.9G is classified as a jogging (running) state, and the 1.9G or higher is classified as a running state. Then, using one minute data of the three-axis sensor, the behavior determination is performed by assigning within which threshold the combined value of the three-axis acceleration of each record is.

Next, calorie consumption calculation is performed from the behavior determination obtained above. Specifically, it can be calculated by the following formula.

(Formula 2)
Calories burned = exercise coefficient x basal metabolism coefficient x body weight (kg) x time (min)

The coefficient of exercise is determined based on each action as shown in FIG. 9 (Japan Sports Association Sports Science Committee).

The coefficient of basal metabolism is determined by age and sex as shown in FIG.
Therefore, for example, when the person to be measured is a 56-year-old male with a weight of 60 kg and a state of about 1.6 G continues for 60 minutes, it is determined that he is in a fast-walking state. Since the exercise coefficient of fast walking (for example, 100 m / min) is 0.1083, the basal metabolism coefficient of a 56-year-old man is 0.0153, weight = 60, time 60 minutes, it is calculated as follows.

Calories burned = 0.1083 x 0.0153 x 60 x 60
= 5.97Kcal

Note that the tables shown in FIGS. 8 to 10 are tabulated and stored in the large-scale storage device (HD) of the information processing apparatus. Personal information such as the age and weight of the person to be measured is also tabulated and registered in the large-scale storage device (HD). That is, the central processing unit (CPU) of the information processing apparatus performs behavior determination with reference to each table when data from the three-axis sensor is input, and the amount of activity (calorie consumption) based on the calculation program of Equation 2 above. ) Is calculated.

  The central processing unit (CPU) of this information processing device is determined to be stationary when the calorie consumption calculated from the data from the three-axis sensor is 0.02 Kcal or less. According to the inventor's experiment, the sympathetic nerve activity index (LF / HF) calculated at this time is a sympathetic nerve activity index (10 minutes resting state using an electrocardiograph in a medical facility such as a hospital) LF / HF).

The greatest feature of this embodiment is that it is possible to measure a sympathetic nerve activity index (LF / HF) not only in such a resting state (resting state) but also in a state where the body is moving in daily life. The balance between heart and body can be displayed visually from the relationship between the value and the amount of activity. The relationship between the sympathetic nerve activity index (LF / HF) and the amount of activity may be displayed in real time every minute, or the sympathetic nerve activity index (LF / HF) may be 1 hour, several hours, 1 day. Alternatively, an average value for several days may be applied.

  From the sympathetic nerve activity index (LF / HF) and the amount of activity obtained in the above description, the “balance between the heart and the body” can be classified into four types as shown in FIG. Such a display is output as it is from the display (DISP) of the information processing apparatus.

  That is, Type I belongs to the “excitement / energetic” type in which the activity amount is 0.02 or more and the sympathetic activity index (LF / HF) is 4 or more.

  Type II belongs to the “concentrated / nervous” type having an activity amount of less than 0.02 and a sympathetic nerve activity index (LF / HF) of 4 or more.

  Type III belongs to the “natural body / behavioral” type with an activity amount of 0.02 or more and a sympathetic activity index (LF / HF) of less than 4.

Type IV belongs to the type of “relaxation / charging” with an activity amount of less than 0.02 and a sympathetic nerve activity index (LF / HF) of less than 4.
In this type classification, the quadrants of the arc portions of the type determined outwardly at predetermined radii every minute on the same axis from the central axis of the XY axes are displayed in a colored manner.

  In this way, the person being measured knows his / her type per minute, so that he / she can visually recognize his / her “balance between heart and body” and can objectively grasp himself / herself. .

  In this type classification, since the amount of activity per minute at desk work was 0.02 or less, this value was set as the minimum activity value (at rest and at rest), but it should be limited to this. Absent.

  As a person to be measured, type classification in a normal life of 24 hours for a male in his 30s who weighs 60 kg will be described. Here, it is assumed that the human recorder (HR) has been worn for 24 hours.

  It is assumed that the measurement subject took 12 hours of work and 12 hours of sleep. The daily average value of the sympathetic nerve activity index (LF / HF) was 4.5. The threshold value for the calorie consumption was provisionally 2000 Kcal.

  First, action determination is performed based on data from the triaxial acceleration sensor. Here, during work, the acceleration was always within the range of 0.9 to 1.2 G, so it is determined as “still state”.

Next, the total calorie consumption for each action determination is calculated as follows.
Calorie consumption = 0.0304 (coefficient for desk work) x 0.955 (male, coefficient for 30s) x 60 (weight) x 720 (minutes)
+0.0170 (coefficient for sleep) x 0.955 (male, coefficient for 30s) x 60 (weight) x 720 (minutes) = 1955 Kcal
Based on the above calculation results, the average of sympathetic nerve activity index (LF / HF) is 4 or more and the amount of activity (calorie consumption) is 2000 Kcal or less.

Next, sleep determination using this system will be described.
The human recorder (HR) and the information processing apparatus (PC) used in this embodiment may be the same as those described above.

  Based on the RR time-series data obtained from the heart rate sensor of the human recorder (HR) and the acceleration information obtained from the triaxial acceleration sensor, the central processing unit (CPU) sleeps based on the program of the information processing device (CPU). Determine.

  As shown in FIG. 12, the sleep detection is performed based on the heart rate (RR time-series data) at the timing when the action is determined based on information from the three-axis acceleration sensor (see FIG. 8) and is determined to be stationary (flooring). When the standard deviation becomes half (50%), it is determined that the patient falls asleep. If the sleep onset cannot be determined in this way, the threshold is raised in units of 10%, such as 60%, 70%..., Based on the 50% reference, and this determination is repeated until detection. The threshold value is set to 50% as a reference value, but may be varied in the range of 20% to 80%.

  In the awakening determination, as shown in FIG. 13, it is detected that the person to be measured has stood up from the bed based on information from the triaxial acceleration sensor. Then, (A) the heartbeat (RR time series data) within 90 minutes before the rising state is acquired, and (B) the standard deviation exceeds the average value, and the difference (slope) before and after the standard deviation Is searched for the portion of the graph in which the time within 0.2 is continuous for 2 minutes before and after (total 5 minutes) or more. The average value at this time is calculated using 50% of data centering on the time of “entry time + rise time) / 2”. At this time, data with large fluctuations at the start and end of measurement are not adopted. And (C) Arousal determination is performed when the posture of the measurement subject obtained from the triaxial acceleration sensor is changed twice or more in 10 minutes before and after.

  If the number of candidates that can determine arousal is 0, the arousal determination is repeated with the conditions (A) to (C) relaxed. Specifically, the number of detections of the posture change state in (C) is one or more. The condition (C) is excluded. (B) Increase or decrease by 0.1 unit within the standard deviation slope of 0.1 to 0.3, expand the detection time range of (A) from 90 minutes to 45 to 135 minutes, etc. What is necessary is just to perform arousal determination.

  In the sleep cycle rhythm determination, as shown in FIG. 14, a sleep graph is generated by smoothing a sympathetic nerve activity index (LF / HF) obtained by applying fast Fourier transform from the RR time series data in a predetermined time unit. Then, a plurality of peaks are searched from the sleep rhythm graph, and a sleep cycle rhythm is generated at the peak interval.

  Specifically, the sympathetic nerve activity index (LF / HF) calculated in units of 1 minute is smoothed with an average of 5 minutes to create a sleep rhythm graph. The peak is detected by searching for the peak from the top of the sleep rhythm graph and adopting the maximum peak. Then, data of 50 minutes before and after the maximum peak adopted above is removed. Search for the next largest peak. Here, the data for 30 minutes before and after are removed. When you reach the end, return to the top of the sleep rhythm graph and search for the next peak. Then, 30-minute data before and after the adopted peak is removed. The maximum number of peaks detected in this way is six.

  Thus, by measuring the sleep rhythm of the person being measured and outputting it as a graph, it is possible to visually grasp the sleep rhythm of the subject.

  As mentioned above, although this invention was demonstrated based on embodiment, this invention is not limited to this. For example, in the embodiment, a numerical value based on the sympathetic nerve activity index (LF / HF) is mainly used as the autonomic nerve activity index, but various calculations are performed using the numerical value based on the parasympathetic nerve activity index (CVRR × LF / HF). You may go.

  The present invention can be used in the health equipment industry used in daily life.

HR Human Recorder PC Personal computer (Information processing device)
CPU Central processing unit MM Main memory PAD Heart rate detection pad ELT Electrode BUS Bus UMEM User memory OS Operating system APL Application program DB Database TBL Table DISP Display device KBD Keyboard

Claims (6)

A heart rate sensor, a three-axis acceleration sensor, a control means, a clock means, a storage means, a human recorder mounted on the surface of the subject's human body, and various values from the means obtained from the human recorder A heart and body balance display method executed by an information processing device that calculates and displays information,
RR interval time series data is acquired from the heart rate sensor by the control means,
Obtaining acceleration information of the person being measured from the three-axis acceleration sensor;
Outputting the RR interval time-series data and the acceleration information synchronized by the clock means from the output means after temporarily storing the heart rate value and acceleration information in the storage means;
In the information processing apparatus,
Removing abnormal values from the RR values obtained from the heart rate sensor, sampling at equal intervals to form RR interval time-series data, and applying fast Fourier transform to calculate an autonomic nerve activity index;
Performing at least a stationary state, a walking state, a fast walking state and the like based on the acceleration information; and
Calculating the amount of activity by calculating calorie consumption based on a preset exercise coefficient and basal metabolism coefficient from the behavior determination result; and
A heart / body balance display method comprising a step of classifying and displaying four types of balance between the heart and the body based on a combination of the magnitude of the action determination result and the magnitude of the activity amount calculation result. A heart rate sensor, a three-axis acceleration sensor, a control means, a clock means, a storage means, a human recorder mounted on the surface of the subject's human body, and various values from the means obtained from the human recorder A visual display system for balance between the body and the body executed by an information processing device that calculates and displays information,
In the human recorder,
Control means for acquiring RR interval time-series data from the heart rate sensor and acquiring acceleration information of the measurement subject from the three-axis acceleration sensor;
Output means for outputting RR interval time-series data and acceleration information synchronized by the clock means to the information processing apparatus after temporarily storing the heart rate value and acceleration information in the storage means;
In the information processing apparatus,
The abnormal value is removed from the RR value obtained from the heart rate sensor, the RR interval time series data is formed by sampling at an equal interval, the autonomic nerve activity index is calculated by applying fast Fourier transform, and the acceleration information Control that performs at least a behavioral determination such as a resting state, a walking state, a fast walking state, and the like, and calculates calorie consumption based on a preset motion coefficient and a basal metabolic coefficient from the behavior determination result to calculate an activity amount Means,
From the result of calculation by the control means, a display means for displaying by classifying and displaying the four types of balance between the heart and the body from the combination of the magnitude of the action determination result and the magnitude of the activity amount calculation result. Balance visual display system. A heart rate sensor, a three-axis acceleration sensor, a control means, a clock means, a storage means, a human recorder mounted on the surface of the subject's human body, and various values from the means obtained from the human recorder A sleep determination method executed by an information processing device that calculates and displays information,
By the control means, RR interval time series data of a certain time is acquired from the heart rate sensor,
Obtain acceleration information of the person being measured from the 3-axis acceleration sensor,
After the RR interval time series data and acceleration information are temporarily stored in the storage means, the RR interval time series data and acceleration information synchronized by the clock means are output from the output means,
In the information processing apparatus,
A sleep state determination method for determining sleep state of sleep onset, awakening or sleep rhythm from RR interval time series data obtained from the heart rate sensor and an action determination result based on the acceleration information. The sleep state is falling asleep,
4. Sleep according to claim 3, wherein a change in body posture is determined to be less than a predetermined value based on the posture information, and sleep onset is determined when a heart rate standard deviation of RR interval time series data from the heart rate sensor at that time is 25 to 75% or less. State determination method. The sleep state is arousal;
Within 45-135 minutes before detecting that the posture information is sitting or standing,
The standard deviation of the RR interval time series data obtained from the heart rate sensor exceeds a predetermined average value, and the difference (slope) before and after the pulse standard deviation is within 0.1 to 0.3. Is continuous for more than 2 minutes before and after (total 5 minutes),
And the sleep state determination method of Claim 3 which determines arousal when the change of the attitude | position obtained from attitude | position information between ten minutes before and after is 2 times or more. The sleep state is a sleep cycle rhythm;
Smoothing an autonomic nervous activity index obtained by applying a fast Fourier transform from the RR time-series data in a predetermined time unit to generate a sleep graph;
The sleep state determination method according to claim 3, wherein a plurality of peaks are searched from the sleep rhythm graph and a sleep cycle rhythm is detected at the peak interval.

Images