JP2959376B2 - Monitoring device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a monitoring device for determining whether a person is present on a bed, for example.

[0002]

2. Description of the Related Art Conventionally, this type of monitoring device is, for example, a 17th monitoring device.
As shown in FIG. 18 and FIG. 18 , a recording mat 5 composed of an insulating plate 1, metal plate antenna members 2a and 2b, an insulating layer 3, and a shield 4 is used to detect a change in electrostatic charge due to body movement of a sleeping person. The presence / absence of body movement is determined by detecting by 2a and 2b (Japanese Patent Laid-Open No. 55-160539).

Further, an infrared sensor or an ultrasonic sensor is provided on a bed or a wall surface around the bed to detect a change in infrared or ultrasonic generated by movement of a person. It was turned on and off depending on the weight of the person.

Further, a film-shaped piezoelectric element having a sharp detection sensitivity and a small reaction time constant is used, and if the output voltage from the piezoelectric element is equal to or higher than a predetermined value, it is determined that the piezoelectric element is present (Japanese Patent Application Laid-Open No. -258246).

[0005]

However, in the above-mentioned conventional configuration, the configuration of the recording mat is complicated, the production cost is high, and there is no moisture permeability or moisture absorption.

[0006] Further, in the configuration for detecting infrared rays or ultrasonic waves, when a person falls asleep and no body movement occurs, there is no change in infrared rays or ultrasonic waves. In the configuration using a pressure switch, heavy luggage or the like is placed on a mattress. If left unattended, the switch would turn on and malfunction.

[0007] In the configuration detection sensitivity by sensitive piezoelectric element, receiving bed near the others walking or bogie, noise due to the vibration propagation in accordance with the movement of the wheelchair, or the influence of electromagnetic induction noise from such transceiver and a motor Easy
was there.

The present invention has been made to solve the above-mentioned conventional problems, and it is intended to determine the presence / absence of a person without imposing a burden on a person at all and without giving a sense of incongruity, and to immediately notify a monitoring person of the result of the determination. Aim.

[0009]

In order to solve the above-mentioned problems, a monitoring apparatus according to the present invention comprises a body movement detecting means for detecting a body movement of a person, and an output signal from the body movement detecting means converted into time-series data. A data collection unit for converting, a statistic calculation unit for calculating a statistic such as a variance of the time-series data collected by the data collection unit, and a human based on the statistic calculated by the statistic calculation unit Determining means for determining presence or absence, wherein the determining means determines that the statistic is equal to a first threshold.
If the value is smaller than the value, it is determined that the user is absent.
Is greater than a second threshold greater than the first threshold
Is determined, and once the presence determination is made, the statistic is
If it is not less than the first threshold, the presence determination is continued.

[0010] Further, a body movement detecting means for detecting body movement of a person, a data collecting means for converting an output signal from the body movement detecting means into time series data, and the time series data collected by the data collecting means. A variance value calculating unit that calculates a statistic such as a variance value of data; a frequency component calculating unit that calculates a frequency component of the time-series data collected by the data collecting unit; and the statistic calculated by the variance value calculating unit And determining means for determining the presence or absence of a person based on the power spectrum calculated by the frequency component calculating means, wherein the determining means determines that the person is absent if the statistic is smaller than a first threshold value. Sometimes, when the statistic is larger than a second threshold larger than the first threshold, it is determined that the statistic is present. When the statistic is equal to or larger than the first threshold and equal to or smaller than the second threshold, the peripheral is determined. Power spectrum calculated by the number component calculating means determines that standing if larger than a predetermined threshold, the power spectrum is to determine the absence if said predetermined threshold value or less.

Further, the body movement detecting means comprises a flexible coaxial cable-like piezoelectric element, and is disposed in a device such as bedding, a bed, a cushion, a chair, and a toilet seat which comes into contact with the human body.

[0012]

According to the present invention, the presence / absence of a person is determined non-invasively with no burden on the user. The output signal from the body motion detecting means is converted into time-series data, and the presence or absence of a person is separated by calculating statistics such as the variance of the time-series data. In addition, the distinction between a person and an object or a person's life or death is made in the same way. This is because as long as a person is present, even if the subject is in a resting state, minute vibrations due to heartbeat, breathing, etc. continue to occur on the body surface, and the variation statistics based on the no-signal state where there is no body movement It is equivalent to treating quantity as an index of body movement. In particular, since the output signal from the body movement detecting means is converted into digitized time-series data, the determination is made with good reproducibility without being affected by temperature fluctuation or the like.

At this time, the determination means has at least two thresholds, and a dead zone is provided between the first threshold and the second threshold to determine the presence or absence of a person, thereby further improving the determination accuracy. This is because a large body movement signal is always output from the body movement detection means when a person transitions from the absence state to the presence state or from the presence state to the absence state, and there is no body movement in the absence state. This is based on the fact that a signal state always appears. For example, when the first threshold is set to a value larger than the second threshold, the state is changed from absent to present only by a body motion signal larger than the first threshold, and is first detected by a body motion signal smaller than the second threshold. From the state to the absence. This enhances the resistance to physical and electrical noise.

By calculating the frequency components of the time series data, the presence and absence of a person can be more clearly separated. This is because, when a person is present, at least a periodic micro-vibration accompanying a heartbeat or breathing can be detected as a body motion signal as long as the person is present, even if there is no large body movement such as stretching out or turning over. Therefore, even if the S / N ratio of the signal obtained from the body motion detecting means cannot be obtained, if there is a sharp power spectrum associated with heartbeat or respiration in a specific frequency band,
The existence of a person can be confirmed.

In addition, since the body movement detecting means is constituted by a flexible coaxial cable-shaped piezoelectric element, the piezoelectric element itself does not need to be shielded, is resistant to electromagnetic wave noise, and is inexpensive and easy to manufacture.

[0016]

EXAMPLES Hereinafter the first embodiment of the present invention with reference to FIGS. 1-7 will be described. FIG. 1 is a perspective view when the monitoring device is provided on a bed. In FIG. 1, reference numeral 6 denotes a body movement detecting means for detecting a body movement of a human body on a bed. Here, a polymer piezoelectric material such as polyvinylidene fluoride (PVDF) is formed into a thin film, and flexible electrode films are attached to both surfaces. Is made of a flexible piezoelectric element formed into a tape shape. Outside the piezoelectric element, a shield process for reducing the influence of electromagnetic wave noise and a waterproof process for protecting from moisture and dirt are performed. The body movement detecting means 6 is fixed to the surface of the mattress 7 so as to be located immediately below the chest when the human body goes to sleep as shown in FIG. A sheet, a thin bed pad, a blanket, or the like may be further placed on this. Reference numeral 8 denotes a signal processing device for determining the presence or absence of a person based on an output signal from the body movement detecting means 6 including a piezoelectric element, and is attached to a bedside or a back surface. That is, the present monitoring device is constituted by the body movement detecting means 6 and the signal processing device 8.

As shown in the block diagram of FIG. 2, the signal processing device 8 comprises a data collection means 9, a variance value calculation means 10, a determination means 11, a display means 12, and a storage means 13. The band amplifier 9a of the data collecting means 9 is an analog amplifier circuit having a band pass filter (BPF) characteristic as shown in FIG. 3 so that a human body motion signal can be detected separately from noise.

FIG. 4 shows an output waveform from the band amplification section 9a when a person actually goes to bed. FIG. 4B is an enlarged view of a portion S in FIG. 4A. As shown in FIG. 4A, a large output is obtained when a body motion such as entering or leaving the bed or turning over is performed, but in other resting states, a small output is obtained due to minute vibrations caused by heartbeat, respiration, and the like. . When there is no vibration, the reference voltage Vref is kept almost unchanged because there is no vibration.

In FIG. 2, the band-amplified output signal is a basic clock (for example, 100H) generated by the timer 9b.
According to z), the data is sampled at regular intervals by the A / D converter 9c and discretized. The data temporary storage unit 9d sequentially stores the time-series data vi output from the A / D conversion unit 9c at regular intervals (for example, 0.01 seconds), and stores a predetermined number n (for example, 500) of the total number of data. At this point, all data contents are output to the variance value calculating means 10, and the same operation is repeated thereafter. The variance σ2 can be expressed as (Equation 1).

[0020]

(Equation 1)

[0021]

(Equation 2)

Further, if it is assumed that the time-series data vi swings up and down almost uniformly as compared with the no-signal state (Equation 1)
Can be transformed as (Equation 3).

[0023]

(Equation 3)

The variance value calculating means 10 calculates the variance value σ ² per unit time (for example, 5 seconds) by statistical calculation corresponding to the above (Equation 1) or (Equation 3). This variance σ ²
Is the degree of variation of the signal as compared with the no-signal state, and can be used as an index of the amount of body movement. If the determination means 11 variance sigma ² obtained in dispersion value calculating means 10, it is greater than a predetermined threshold value T0 standing, absence and determine otherwise
However, in order to enhance the resistance to physical and electrical noise, the determination means 11 has two thresholds, and the first threshold T
A dead zone is provided between 1 and the second threshold value T2 to determine the presence or absence of a person .

First, the judging means 11 sets the variance value calculating means 10
If the variance value σ ² obtained in is larger than a predetermined threshold value T0,
Otherwise, a case where it is determined that the user is absent will be described. FIG. 5 shows the presence / absence state of a hospitalized patient for 10 hours and the transition of the variance value obtained by the variance value calculation means 10. 5A shows the presence / absence result of the correct answer, FIG. 5B shows the transition of the variance value, and FIG. 5C shows the determination result by the determination means 11 using the threshold value T0 . Here, the A / D conversion section 9c samples the output waveform from the band amplification section 9a at 100 Hz, and the variance value calculation means 10 obtains a variance value for each 500 time series data, that is, for 5 seconds. The judgment was made each time. As a result, it became clear that presence / absence can be separated at a judgment accuracy rate of about 98%.

The cause of the erroneous determination of about 2% in the above is that the vibration is transmitted to the body movement detecting means 6 due to the passage of a vibration source near the monitoring device, and the erroneous determination is made that the body is present. Conversely, it is considered that the human body temporarily floats in the air from the body motion detection unit 6 when turning over and the vibration cannot be detected, so that the user is erroneously determined to be absent. Therefore, next, determination means 1
1 has two thresholds, a first threshold T1 and a second threshold T2
When determining human standing, the absence is provided a dead zone between the Nitsu
I will describe. FIG. 6 is a flowchart showing the operation of the determination means 11 in this case . First, in step 111, the initial state is set to “absent”, and in step 112, the variance value σ ² calculated by the variance value calculating means 10 is read. If it is "absent", the process proceeds to step 115. In step 114, the read variance value σ ² is equal to the first threshold T1.
If it is smaller, the state is changed to the "absent" state in step 116, and the process proceeds to step 117. Otherwise, the process directly proceeds to step 117 to output the current state, and returns to step 112. On the other hand, in step 115, if the read variance value σ ² is larger than the second threshold T2,
After the state is changed, the process proceeds to step 117. Otherwise, the process directly proceeds to step 117 to output the current state, and similarly returns to step 112. The first threshold T1 determined here
And the second threshold value T2 is constant, and 0 <T1 <T2.

[0027] That the present embodiment, first and state transitions for the first time by the absence in the standing by the second threshold value T 2 greater body motion signal from
The logic determination condition has a Schmitt structure so that the state transition from the presence to the absence is made for the first time by a body motion signal smaller than the threshold value T1 of 1 . This is because, as is clear from FIG. 4 (a), a large body movement signal is always output from the body movement detection means when the person shifts from the absence state to the presence state or from the presence state to the absence state. In the absence state, a no-signal state without any body movement always appears.

[0028] FIG. 7 is a representation of 10 hours of standing by means of exactly the same data as the data shown in FIG. 5, the transition of the variance values calculated by the variance value calculating means 10 absent state. FIG.
7A shows the presence / absence result of the correct answer, FIG. 7B shows the transition of the variance value, and FIG. 7C shows the determination result by the determination means 11. Here, the A / D conversion section 9c samples the output waveform from the band amplification section 9a at 100 Hz, and the variance value calculation means 10 obtains a variance value for each 500 time series data, that is, for 5 seconds. The judgment was made each time. As a result, it was proved that the presence / absence could be separated with a correct answer rate of 100%, and the resistance to physical and electrical noise was enhanced.

The display means 12 displays the presence / absence determination result obtained by the determination means 11 with a lamp each time. The storage unit 13 stores a variance value as an index of the amount of body movement calculated by the variance value calculation unit 10. Here, the determination means 11, the display means 12, the storage means 13 and the like may be provided outside the signal processing device 8 so that remote monitoring can be performed from a place at a distance.

According to the above configuration, the presence or absence of a person can be determined non-invasively without putting any burden on the user. Particularly it converts the output signal from the body motion detecting means 6 in a time series data to calculate a variance value of the time series data, this amount
Based on the scattered value, the judgment means 11 determines whether the first threshold value T1 and the second threshold value
A dead zone is set between values T2 to determine the presence or absence of a person.
To enhance the resistance to physical and electrical noise and
And absence can be separated.

[0031] Incidentally, by storing the dispersion value calculated in the storage unit 13 in a distributed value calculating unit 10 as an indicator of the body movement amount per unit time, is prolonged human condition monitoring performed easily. If the raw time-series data output from the A / D converter 9c is accumulated as it is, an enormous amount of data will be stored.
3 can be realized inexpensively and data analysis can be simplified. Thus, for example, the present invention can be applied to estimating a health state such as a sleep depth and a prediction of occurrence of bed congestion based on the amount of body movement overnight.

Further, it is used here a variance value as an index indicating the body movement amount, which is not binding on the present invention. For example, to derive the amount of body movement per unit time by statistically processing time series data, standard deviation, squared averaging,
A mode, median, or the like may be used, or the waveform of the sampled time-series data may be further shaped by digital signal processing. As a pre-process for calculating a statistic such as a variance, a low-pass filter (LPF) may be further applied, an interpolation process, or an envelope process may be performed. Also, instead of making a determination every unit time, the output value from the body movement detecting means 6 is integrated, and the time required until the predetermined amount of body movement is reached is determined.
Absence may be determined. The body movement detecting means 6 may use a pressure-sensitive resistance element made of, for example, conductive rubber or conductive carbon instead of the piezoelectric element.

Also, as described above, using the threshold value T0
When judging absence, minutes are counted for every 5 seconds of time-series data.
Determine the scattered value and determine the presence / absence each time
However, when it is impossible to repeat the state inversion between the presence and absence at a high frequency of about 5 seconds , correct the determination by changing the determination output by the determination means 11 to, for example, about 1 minute instead of every 5 seconds. It is clear that the rate can be increased to almost 100%. As a matter of course, provisional determination results may be temporarily output at intervals of 5 seconds, and a plurality of the results may be collected and a majority decision may be made to make a comprehensive determination.

Next, a second embodiment of the present invention will be described with reference to FIGS.
This will be described with reference to FIG. In FIG. 8, the same functional blocks as those in the first embodiment are denoted by the same reference numerals, and description thereof will be omitted.
This embodiment is different from the first embodiment in that the data collection means 9
Frequency component calculating means 14 for calculating the frequency component of the time-series data collected by the user, and the presence of a person according to the power spectrum calculated by the frequency component calculating means 14,
It is provided with the determination means 15 for determining absence. The frequency component calculation means 14 extracts only the frequency components (for example, 0.5 to 2.0 Hz) accompanying the minute vibration of the person at rest, and outputs the total amount of the power spectrum density in the frequency band, that is, the power P.

As apparent from FIG. 4 (a), when a person is present, there is either a large body movement due to entering, leaving the bed, turning over, or a minute vibration due to heartbeat or breathing. Can be roughly classified. Large body movements are transient (ie, non-periodic) signals that have very large waveform amplitudes and only relatively high frequency components. On the other hand, although the amplitude of the waveform itself is small, the periodic vibration concentrated in a specific frequency band can be continuously recognized.
Generally, a person's heart rate at rest is 38 to 110 beats / minute (0.
63-1.8 Hz), and the respiratory rate is 10-20 times / min (0.
17 to 0.33 Hz), because the body movement detecting means 6 detects this biological signal.

That is, when the person is in a resting state, the signal obtained from the body movement detecting means 6 is a signal in which the vibration accompanying the heartbeat and respiration of the person is superimposed, and such a periodic signal is extracted from the output signal. In this case, the presence of a person can be confirmed even if the amount of body movement (that is, the amplitude of the output signal) is small.

The frequency component calculation means 14 includes a filter unit 14
a, a waveform shaping unit 14b and an FFT calculating unit 14c,
These are all realized by digital signal processing means. The filter unit 14a is a cyclic band-pass filter such as Butterworth, and passes a frequency component necessary for heartbeat signal extraction. The waveform obtained by the filter unit 14a is further transformed by a waveform shaping unit 14b by an envelope process or the like so that the fundamental period component is emphasized, and then a specific frequency band (for example, 0.5 to 2.0 Hz) by the FFT calculation unit 14c. Is calculated.

Next, the operation of the determination means 15 will be described with reference to the flowchart of FIG. First, in step 151, the variance value σ2 calculated by the variance value calculation means 10 is read. In step 152, the variance value σ2 thus read is smaller than the threshold value T3 corresponding to the first threshold value T1 described in FIG. If there is, the process proceeds to step 152, where the current state is determined to be “absent”, and the process returns to step 151. Otherwise, the process proceeds to step 154, and if the read variance value σ2 is a variance value larger than the threshold value T4 corresponding to the first threshold value T2 described in FIG. 6, the process proceeds to step 155 to change the current state to “ And the process returns to step 151. The threshold value T3 and the threshold value T4 determined here are constant, and 0 <T3 <T4. If the read variance value σ2 satisfies T3 ≦ σ2 ≦ T4, the process proceeds to step 156, where the power P calculated by the frequency component calculation means 14 is read, and if the power P is larger than the predetermined threshold T5 in step 157, the process proceeds to step 155.
Then, the process proceeds to step 153, and if not, the process proceeds to step 153, where the current state is determined to be “absent”.

By the above operation, the frequency component calculating means 14
Is the power P of the specific frequency band necessary for discrimination of a person at rest.
Is extracted and transmitted to the determination means 15 so that a person can be identified from both the magnitude of the amplitude and the frequency component. Therefore, even when the S / N ratio of the signal obtained from the body motion detecting means 6 cannot be obtained, the presence of a person can be confirmed if the power in a specific frequency band, that is, the periodic oscillation accompanying the heartbeat or respiration is recognized. Specifically, even in a situation where the distance between the human body and the body movement detecting means 6 is large due to the blanket or the amount of clothes, or in a case where the amount of pulsation is weak and the minute vibration on the body surface is difficult to be transmitted to the body movement detecting means 6 in an elderly person or the like. It is possible to determine the presence or absence of a person successfully. Further, even if high-frequency noise from a commercial power supply or a radio wave is superimposed on the output signal from the body movement detecting means 6, this can be effectively separated to extract only a necessary signal.

Since the frequency component calculation means 14 is performed by digital signal processing, the circuit configuration of the band amplification section 9a can be simplified. In particular, when it is desired to pass periodic vibrations caused by heartbeat or respiration, since the frequency is very low, a conventional analog circuit necessarily requires a signal transmission capacitor (for example, about several tens of μF) having a very large value.
In this case, the whole apparatus becomes large, and it is difficult to secure the accuracy of the value itself, and there is a problem in reproducibility against temperature fluctuations. However, these problems can be solved.

Further, a plurality to the body motion detecting means 6 as shown in FIG. 10 can independently measure the body movement of each part of the body (Fig. 1
0 using a piezoelectric element of the nine points 6a-6i), a configuration in which input as separate signals to the signal processor 8
Is also good. The signal processing device 8 has the same configuration as that of the embodiment shown in FIG. 8, and the variance value .sigma.2 calculation by the variance value calculation means 10 and the power P calculation in the specific frequency band by the frequency component calculation means 14 are performed by the output waveform of each piezoelectric element. Are determined independently from each other, and the judging means 15 judges the presence and absence by focusing only on the maximum value among them.

FIG. 11 shows the output waveforms from the body motion detecting means 6 (9 points 6a to 6i) when a person lays down on a bed in a resting state, and at the same time, the respiration and heartbeat obtained by a polygraph by another measuring device. 3 shows an output waveform (for 20 seconds). In this measurement, since the subject was located in the center of the bed, 6b,
Particularly large signals are output at portions 6e and 6f. The amplitude is the largest at 6b corresponding to the chest position, but with respect to the periodic vibration accompanying the heartbeat, 6h corresponding to the leg characteristically captures the venous flow (blood flow). 6b shows that the heartbeat signal is superimposed on the respiration signal.

In the above configuration, by using the body movement detecting means 6 as a plurality of piezoelectric elements, a signal corresponding to the body movement for each body part is directly output. When a person is sleeping, the sleeping posture may change sequentially. In the case where only one piezoelectric element is arranged as in the past, a gap is created between the body and the mattress 7 due to a slight overturn, etc. even if the body can ride on the piezoelectric element and detect the signal until a certain point. ,
Thereafter, there was a risk that the signal could not be detected at all. However, according to the present embodiment, almost all of the area on the bed became a detection area and the blind spot disappeared. Also, minute vibrations caused by heartbeat and breathing propagate through the body surface and the mattress 7 with delay. However, in a system in which one piezoelectric element is long, tension and contraction overlap, and the signal obtained as a whole is canceled. In contrast to this, in the present embodiment, only the body movement of each part can be detected, so that the periodic characteristics of the signal can be easily grasped. Furthermore, since the amount of body movement for each body part can be detected, it is possible to monitor which part of the body is moving while sleeping, and it is possible to specify a bed sore part and to prevent bed slippage by measuring time without body movement with higher accuracy.

Incidentally, the larger the number of piezoelectric elements used, the higher the spatial resolution naturally, but the signal processing speed and the storage capacity are also required accordingly.
It is practical to set several points to several tens points.

[0045] Next a third embodiment of the present invention will be described with reference to FIG. 12. This embodiment is different from the first and second embodiments in that a flexible coaxial cable-shaped piezoelectric element is used for the body movement detecting means 6. The body movement detecting means 6 composed of the coaxial cable has a conductive core wire 6 made of copper at the center.
01, and a piezo rubber 602 outside thereof. Further, the outside is covered with a stitched ground electrode 603,
The outside is covered with a coating 604. When the coaxial cable is bent or pressed, the piezo rubber 602 is spontaneously polarized, and a potential difference is generated between the core wire 601 and the ground electrode 603.

In the above structure, the body movement detecting means 6 is made of a flexible coaxial cable-shaped piezoelectric element, so that the piezoelectric element itself does not need to be shielded, is resistant to electromagnetic wave noise, and is inexpensive and easy to manufacture.

Next, a fourth embodiment of the present invention will be described with reference to FIG. When the determination result by the determination means 11 is present and the variance value calculated by the variance value calculation means 10 is smaller than a predetermined value, a duration measurement means 16 for measuring the duration is provided. If the duration time measured by the means 16 exceeds a predetermined value (for example, 2 hours), a notifying means 18 for notifying an external centralized monitoring center 17 may be provided.
Good. Further, the notifying unit 18 may have a configuration for notifying the external centralized monitoring center also when the determination result of the determining unit 11 changes from the absent state to the absent state and when the absent state changes to the absent state. .

In other words, when there is almost no continuous body movement for a certain period of time (for example, two hours) despite the presence of a person in the bed, it is considered that the danger of floor congestion has increased, and the centralized monitoring center is notified or monitored. This is a configuration in which a notification is made in the same manner when the presence or absence of the target person is reversed.
Here, the notification means 18 has a serial data communication means,
A telemeter system is constructed by connecting to a centralized monitoring center 17 via a communication line 19.

It is generally known that a healthy person causes some body movement every several tens of minutes even during bedtime. However, some elderly people and physically handicapped who have weak physical strength can do so by themselves. Easy to get bedsores. In addition, since bed sickness progresses rapidly overnight or in a few hours, caregivers have been forced to change positions at regular intervals even for elderly people who can still move, but this monitoring device If there is a danger of floor congestion, it is immediately reported to the centralized monitoring center 17, so that floor congestion can be prevented beforehand, the burden on the caregiver can be greatly reduced, and there is an effect that sleeping of the sleeping person is not disturbed. On the other hand, by reporting when the presence / absence state is reversed, the wandering of the elderly with dementia at night or the accident of falling from the bed can be immediately detected from a remote place.

The notification means 18 notifies the centralized monitoring center 17 only when there is a danger of floor congestion or when the presence / absence state is reversed, so that the communication line 19 need not be occupied. That is, it is easy to use the communication line 19 for another purpose or to connect a plurality of the monitoring devices to form a 1: n communication system.

It should be noted that the centralized monitoring center 17 may be located in the same premises as the monitoring device or in a remote location outdoors. The notifying unit 18 may be configured to drive an emergency buzzer, an emergency lamp, and the like provided in the signal processing device 8 instead of performing remote notification via the communication line 19.

Next, a fifth embodiment of the present invention will be described with reference to FIGS.
6 will be described. A biological signal extracting unit 20 for extracting a human heart rate according to a power spectrum in a specific frequency band among frequency components of the time-series data collected by the data collecting unit 9, and a heartbeat extracted by the biological signal extracting unit 20; First determining means 21 for determining whether or not the number deviates from a predetermined range; and duration measuring means 22 for measuring the time during which the heart rate extracted by the biological signal extracting means 20 continuously deviates from the predetermined range. Second determining means 23 for determining whether or not the duration has exceeded a predetermined value; display means 24 for displaying the heart rate extracted by the biological signal extracting means 20; and storage means 25 for storing the heart rate. And a fluctuation amount calculating means 26 for calculating the fluctuation amount from the heart rate data stored in the storage means 25, and it is determined whether or not the fluctuation amount calculated by the fluctuation amount calculating means 26 is within a predetermined range. Comprising a third determination unit 27, further first determination unit 21, the second determination unit 23, when the judgment condition is satisfied in either third determination means 27 may comprise a notification means 28 for notifying.

In the above configuration, the biological signal extracting means 20
Represents a heartbeat filter unit 20a, a heartbeat waveform shaping unit 20b,
It comprises an autocorrelation function calculation section 20c and a heart rate calculation section 20d, all of which are realized by digital signal processing means. The heartbeat filter unit 20a is a cyclic band-pass filter such as Butterworth, and passes a frequency component necessary for heartbeat signal extraction. The waveform obtained by the heartbeat filter unit 20a is further converted by a square processing so that the R wave in the heartbeat is emphasized in the heartbeat waveform shaping unit 20b. It is generally known that among the P, Q, R, and S waves of the heartbeat, the R wave has the highest frequency component and a large amplitude. Therefore, the basic period is derived based on the RR interval. Next, in the autocorrelation function calculation section 20c, the heartbeat waveform shaping section 20
An autocorrelation function R (τ) as shown in (Equation 4) or the autocorrelation coefficient C (τ) as shown in (Equation 5) is calculated for the waveform converted in b.

[0054]

(Equation 4)

[0055]

(Equation 5)

[0056]

(Equation 6)

Further, the heart rate calculating section 20d calculates the basic period T of the time series data from the calculation result of the autocorrelation function calculating section 20c, and outputs this as the heart rate per minute.

FIG. 15 shows an output waveform 15 (a) of the data collection means 9 based on an output waveform (30 seconds) from the body movement detection means 6 when a person lays down on a bed in a resting state, and a heartbeat filter section. An output waveform 15 (b) of 20a, an output waveform 15 (c) of the heartbeat waveform shaping unit 20b, and an output waveform 15 (d) of the autocorrelation function calculation unit 20c are shown, respectively. Fifteen
In the graph of (d), the horizontal axis is the time shift amount τ, and the vertical axis is the autocorrelation coefficient C (τ). The heart rate calculation unit 20d calculates the basic period T which is the first peak point, and when the value of τ at this peak point is 0.1 <τ <5.0 and the value of C (τ) is 0.2 or more. If so, it is determined that the correct heart rate has been detected, and the heart rate per minute is output. Otherwise, it outputs a determination result of “measurement impossible”. The detection of the basic period T may be determined by, for example, further performing an FFT operation on the data of 15 (d) and detecting the power spectrum. 15 (d)
In the example, the basic cycle T is 0.85 seconds, so that the heart rate of this person is 71 beats per minute. In the body movement detecting means 6 composed of a piezoelectric element, the heart rate is temporarily "measurable" when there is a large body movement such as rolling over, but returns to the measurable state when the state of rest is restored. The output frequency of the heart rate signal from the biological signal extraction unit 20 may be performed every 30 seconds using time-series data of 30 seconds as shown in FIG. 15 or by shifting the time-series data to be sequentially processed. It may be more frequent (for example, every 5 seconds) by going.

The first judging means 21 has a variable threshold setting section 21a, from which the heart rate is output from the biological signal extracting means 20 and within a range determined by the setting section 21a (for example, 20 to 20).
If it deviates from (200 beats / minute), an abnormality determination is output.

The duration measuring means 22 has a variable threshold setting section 22a, and the heart rate output from the biological signal extracting means 20 is within a range determined by the setting section 22a (for example, 30 to 30).
150) is continuously output.

The second judging means 23 has a setting section 23a which can change the threshold value. The duration outputted from the duration measuring means 22 is a value determined by the setting section 23a (for example, 5 minutes).
If it exceeds the limit, it outputs an abnormality judgment.

The storage means 25 is constituted by a FIFO (first-in first-out) memory and stores the heart rate output from the biological signal extraction means 20 for a long time (for example, for the latest one week). The fluctuation amount of the heart rate is calculated from the variance value from the heart rate histogram as shown in FIG. Alternatively, the change of the heart rate (time-series data) as the amount of fluctuation is calculated by frequency analysis to calculate the power of the power spectrum as a multiple of the frequency.

The third judging means 27 has a variable threshold setting section 27a. When the fluctuation amount calculated by the fluctuation amount calculating means 26 exceeds the range defined by the setting section 27a, an abnormality judgment is output. This is based on the fact that the heartbeat of a healthy person fluctuates within a certain range when viewed over a long period of time, and that there is some abnormality if there is no fluctuation. In addition, when a person falls asleep after entering the bed, the required metabolic rate is reduced, so that a slow wave is observed in the heart rate and the respiratory rate.

The notifying means 28 turns on the alarm lamp if any one of the first judging means 21, the second judging means 23 and the third judging means 27 outputs an abnormality judgment and the condition is satisfied. Drive the buzzer.

In the above configuration, the presence or absence of an abnormal state of a person can be monitored from the heart rate of the person. In particular, by providing a plurality of determination conditions, the accuracy of the abnormal state determination can be improved. In addition, since a trend of a biological signal of a specific person can be grasped for a long time, a transition of a health condition can be grasped. When the present monitoring device is operated, continuous monitoring is performed, so that necessary data is automatically accumulated without a medical professional being present. In addition, circulatory diseases can be automatically detected from the fluctuation amount of the heartbeat.

It should be noted that the biological signal calculated here may be configured to prevent sudden death due to apnea or the like by using a respiratory rate or the like. The heartbeat and the respiration may be separated from the same signal, and the conditions for abnormality determination may be combined. Further, the abnormality determination may be performed on the amount of body movement or the body movement part. Further, when an abnormality is determined from a biological signal, it may be determined from a change rate or a long-term fluctuation tendency instead of a value itself such as a heart rate per minute.
Further, the threshold value serving as the determination condition may be changed according to the time zone or the season. The notification means 28 may construct a telemeter system for notifying a centralized monitoring center or the like, as in the sixth embodiment shown in FIG. A plurality of piezoelectric elements may be provided in the body movement detecting means 6. Also, the monitoring device has been described by taking a person as an example, but may be adapted to other living things such as animals.

[0067]

As is apparent from the above description, the following effects can be obtained according to the monitoring apparatus of the present invention.

(1) The output signal from the body movement detecting means is time-based
Statistics such as variance of time series data converted to column data
And the determining means determines a first threshold based on the statistic.
A dead zone between the threshold and the second threshold to determine the presence or absence of a person
To increase immunity to physical and electrical noise
The presence and absence of a person can be separated.

(2) Calculated by statistic calculation means
The size of the statistic is calculated by the frequency component calculation means.
The presence or absence from both sides of the power spectrum
Therefore , the S / N ratio of the signal obtained from the body movement detecting means can be obtained.
Power in a particular frequency band
If periodic vibrations associated with beats and breathing are recognized, the presence of a person
You can check. Also, the commercial signal is output to the output signal from the body motion detection means.
Even if high frequency noise is superimposed from sources or radio waves,
It can separate well and extract only the necessary signals
You.

(3) A flexible coaxial cable is used as the body movement detecting means.
The piezoelectric element itself is sealed
No need to install, it is strong against electromagnetic wave noise, and is manufactured at low cost
Also becomes easier.

[Brief description of the drawings]

FIG. 1 is a perspective view of a monitoring device according to a first embodiment of the present invention.

FIG. 2 is a block diagram of the apparatus.

FIG. 3 is a filter characteristic diagram of a band amplification unit 9a in the device.

4A is a diagram showing an output waveform from a band amplification unit 9a in the device. FIG. 4B is an enlarged view of an S portion of an output waveform from the band amplification unit 9a in the device.

FIG. 5 is a diagram showing a determination result by a determination unit 11 in the apparatus.

FIG. 6 is a flowchart illustrating the operation of a determination unit 11 in the apparatus .

7A is a diagram showing a presence / absence state in the device; FIG. 7B is a diagram showing a transition of a variance value in the device; FIG. 7C is a diagram showing a determination result by a determination unit 11 in the device;

FIG. 8 is a block diagram of a monitoring device according to a second embodiment of the present invention.

FIG. 9 is a flowchart for explaining the operation of a determination unit 15 in the apparatus.

FIG. 10 is a perspective view of the monitoring device.

FIG. 11 is a diagram showing an output waveform from a band amplification unit 9a and a polygraph measurement result in the same device.

FIG. 12 is a structural diagram of a body motion detecting unit 6 of the monitoring device according to the third embodiment of the present invention.

FIG. 13 is a block diagram of a monitoring device according to a fourth embodiment of the present invention.

FIG. 14 is a block diagram of a monitoring device according to a fifth embodiment of the present invention.

15A is a diagram showing an output waveform of a data collection means 9 in the device; FIG. 15B is a diagram showing an output waveform of a heartbeat filter 20a in the device; FIG. 15C is an output of a heartbeat waveform shaping unit 20b in the device; (D) A diagram showing an output waveform of the autocorrelation function calculator 20c in the same device.

FIG. 16 is a distribution diagram of the heart rate stored in the storage means 25 in the apparatus.

FIG. 17 is a sectional view of a conventional monitoring device.

FIG. 18 is a sectional view of the apparatus taken along line AA.

[Explanation of symbols]

 Reference Signs List 6 body motion detecting means 8 signal processing device 9 data collecting means 10 variance value calculating means 11 determining means 12 display means 13 storage means

────────────────────────────────────────────────── ─── Continuation of the front page (56) References JP-A-3-258246 (JP, A) JP-A-4-22334 (JP, A) (58) Fields investigated (Int. Cl. ⁶ , DB name) A61B 5/00 102 A61B 5/11

Claims (3)

(57) [Claims] 1. A body movement detecting unit for detecting a body movement of a person, a data collecting unit for converting an output signal from the body movement detecting unit into time series data, and the time series collected by the data collecting unit. A variance value calculating unit that calculates a statistic such as a variance value of data; and a determining unit that determines presence or absence of a person based on the statistic calculated by the variance value calculating unit. Is determined to be absent if less than a first threshold, and when the absence is determined, the statistic is determined to be present if the statistic is greater than a second threshold greater than the first threshold, and once the presence determination is made, A monitoring device that continues the presence determination if the statistic is equal to or greater than the large 1 threshold. 2. A body movement detecting means for detecting a body movement of a person, a data collecting means for converting an output signal from the body movement detecting means into time series data, and the time series collected by the data collecting means. A variance value calculating unit that calculates a statistic such as a variance value of data; a frequency component calculating unit that calculates a frequency component of the time-series data collected by the data collecting unit; and the statistic calculated by the variance value calculating unit And determining means for determining the presence or absence of a person based on the power spectrum calculated by the frequency component calculating means, wherein the determining means determines that the person is absent if the statistic is smaller than a first threshold value. Sometimes, if the statistic is larger than a second threshold larger than the first threshold, it is determined that the statistic is present, and if the statistic is equal to or larger than the first threshold and equal to or smaller than the second threshold, the frequency component is determined. A monitoring device which determines that the power spectrum is calculated when the power spectrum calculated by the minute calculating means is greater than a predetermined threshold value, and determines that the power spectrum is absent when the power spectrum is equal to or less than the predetermined threshold value. 3. The body movement detecting means comprises a flexible coaxial cable-shaped piezoelectric element, and is disposed in a device that comes into contact with a human body, such as a bedding, a bed, a cushion, a chair, and a toilet seat. Monitoring equipment.