JP2005304941A - Sleep state detector - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a sleep state detector surely detecting a normal/abnormal state by detecting daily sleep environment without letting a user go to a specific inspection institute such as hospital. <P>SOLUTION: This sleep state detector detects sounds around a bed body by a microphone 3, detects the sounding of snore by a snore detecting part 31 based on obtained sound signals, and estimates whether the detected snore is steady or not by a snore state estimation part 32. This sleep state detector estimates whether the user falls into an apnea state or not by an apnea state estimation part 33 based on the estimated snoring state and the detected snore sound signals, determines whether the estimated apnea state is an abnormal state or not by an abnormality determination part 44 and, when the user is in the abnormal state, notifies it to a notification destination by a notification part 50. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a sleep state detection apparatus that detects a user's sleep state during sleep and reports the health state of the user to a predetermined report destination.

  In recent years, attention has been paid to the presence of a disease that occurs during sleep such as so-called apnea syndrome, in which breathing stops during sleep. Along with this, a system for grasping the health state of a user during sleep has been proposed. ing. However, since many of these systems, for example, are equipped with electrodes for measuring an electrocardiogram, sensors for detecting respiration, etc. on the user's body, the burden on the user is heavy, and during measurement (during sleep) ) Has a problem that the sensor attached to it is detached. On the other hand, an attempt has been made to measure a user's health condition during sleep in a non-contact manner (see, for example, Patent Document 1 and Patent Document 2).

  Specifically, Patent Document 1 discloses a medical biological information detection apparatus that detects biological information of animals including humans. In particular, the medical biological information detection apparatus includes a first time point that is updated with respect to means for inputting moving image information related to biological information of a subject and time-series information on the moving image input at a predetermined frame rate. Primary processing means for detecting a primary information change amount in a corresponding area of each of the image at the second time point and the image at the second time point that is updated in a predetermined relationship with respect to the first time point; Then, a time-series signalized primary information change amount which is arbitrary time-series information is generated by newly arranging this primary information change amount as an element, and this time-series signalized primary is generated as necessary. Secondary processing means for extracting up to time-series signalized secondary information change amount having physical significance different from the information change amount, and high-order processing for extracting higher-order feature amounts as necessary Means, and information and features after processing Means for outputting, in which they were and means for internal control. As a result, this medical biological information detecting apparatus can extract biological information as a moving image by optical remote observation without contact or restraining such as electrode fixation to a subject.

  Further, Patent Document 2 discloses an apnea detection unit that images a monitored living body in a lying state and detects the occurrence of an apnea state based on a change in the captured image, and apnea detection by the apnea detection unit. A biological monitoring apparatus is disclosed that has a warning output and a notification output means for outputting predetermined information about a monitored living body when a state is detected. In particular, in this living body monitoring apparatus, the apnea detection means determines that an apnea state has occurred when apneas exceeding a preset time have occurred continuously a predetermined number of times. Thereby, in this living body monitoring apparatus, it can be reliably detected that the monitored living body has entered an apnea state, and appropriate monitoring of the monitored living body can be performed.

JP-A-7-124126 JP 2002-306455 A

  However, the conventional techniques including Patent Document 1 and Patent Document 2 are used exclusively in specific inspection institutions capable of investigating the presence or absence of abnormal states during sleep such as apnea syndrome. It is targeted. Accordingly, since diseases that occur during sleep, such as apnea syndrome, often have no subjective symptoms, it is desirable to be able to detect the presence or absence of abnormal conditions during daily sleep.

  Then, this invention is proposed in view of the situation mentioned above, and it aims at providing the sleep state detection apparatus which can detect the presence or absence of an abnormal state reliably by detecting a daily sleep state. To do.

  In order to solve the above-described problem, the sleep state detection device according to the present invention detects the ambient sound of the bed body lying when the user goes to bed, and detects snoring based on the obtained audio signal. Estimate whether the snoring detected is stationary. Then, the sleep state detection device according to the present invention estimates whether or not the user has entered an apnea state based on the estimated snoring state, and whether or not the estimated apnea state is an abnormal state. In addition, if it is determined that the user is in an abnormal state, a report is sent to the report destination.

  According to the sleep state detection device of the present invention, an abnormal state caused by apnea that occurs during sleep can be reliably detected without going to a specific laboratory such as a hospital, and early notification can be made by reporting that fact. Correspondence becomes possible.

  Hereinafter, specific embodiments to which the present invention is applied will be described in detail with reference to the drawings.

  The present invention is applied to a sleep state detection device that detects a user's sleep state during sleep and reports the health state of the user to a third party.

  For example, as shown in FIG. 1, this sleep state detection device is installed in a bedroom or the like where a user US sleeps and is configured to have a bed-like appearance. Specifically, the sleep state detection apparatus is configured by providing various sensors for detecting various types of information during sleep of the user US on the bed body 1 on which the user US lies at bedtime.

  As this sensor, a pressure mat sensor 2 provided in the form of a sheet on the bed body 1, a plurality of microphones 3 and a plurality of side human body detection sensors 4 provided on the sides of the bed body 1, and a ceiling with respect to the bed body 1 And a plurality of upper human body detection sensors 5 and a single camera 6 provided above. Here, the side human body detection sensor 4 is provided at a height when the user US is sleeping, and at a height when the user US is not sleeping.

  Sensor signals indicating the detection values of these various sensors are analyzed by the signal processing module 7 attached to or incorporated in the bed main body 1 to estimate the sleep state of the user US. Based on the estimation result of the sleep state of the user US, the sleep state detection device uses a notification module 8 built in the signal processing module 7 to, for example, a predetermined notification destination RP such as a hospital or a close relative outside the home, Notification is made via the gateway GW and the network NT such as the Internet. The reporting module 8 reports via a non-IP (Internet Protocol) network (not shown) when reporting to the in-house reporting destination RP.

[Configuration of Sleep State Detection Device]
The various sensors and the signal processing module 7 and the notification module 8 in such a sleep state detection device have a circuit configuration as shown in FIG. Specifically, the sleep state detection device includes a sensor unit 10 corresponding to the above-described sensor group, a pre-processing unit 20 corresponding to the above-described signal processing module 7, a sleep state detection unit 30, and a health state detection unit 40, And a reporting unit 50 corresponding to the reporting module 8 described above.

  The sensor unit 10 includes the pressure mat sensor 2, the microphone 3, the side human body detection sensor 4, the upper human body detection sensor 5, and the camera 6 described above.

  The microphone 3 is composed of, for example, a capacitor microphone, and always detects a snoring sound or the like uttered by the user US. The audio signal detected by the microphone 3 is supplied to an A / D (Analog to Digital) conversion unit 21 of the preprocessing unit 20.

  The camera 6 is composed of, for example, a CCD (Charge Coupled Device) camera, and always images the upper body or the whole body of the user US in a state lying on the bed body 1 from above. Note that the camera 6 may be provided with a light that emits infrared light when imaging using visible light is difficult, such as when the room in which the sleep state detection device is installed is dark. The video signal imaged by the camera 6 is supplied as a sensor signal to the A / D conversion unit 22 of the preprocessing unit 20.

  The pressure mat sensor 2 is a pneumatic or piezoelectric sensor laid on the floor surface of the bed body 1 and actively detects minute body movements and vibrations of the lying user US. The pneumatic pressure mat sensor 2 is a mat body in which air is sealed. For example, as shown in FIG. 3, a minute body movement or vibration of the user US is regarded as a change in the air pressure, and the change in the air pressure is detected. Is transmitted to the microphone 4b via the air tube 4a, thereby converting minute body movements and vibrations of the user US into electric signals. Further, when the pressure mat sensor 2 is a piezoelectric type, although not particularly illustrated, the pressure mat sensor 2 converts a minute shape change of the mat body itself generated in response to a minute body movement or vibration of the user US into an electric signal as it is.

  Since the pressure mat sensor 2 detects minute movements and vibrations of the user US in both the pneumatic type and the piezoelectric type, the user's US breathing, heartbeat, body movement, etc. during sleep are electrically It can be acquired as a signal. The biological signal detected by the pressure mat sensor 2 is supplied to the A / D conversion unit 23 of the preprocessing unit 20 as a sensor signal.

  The side human body detection sensor 4 and the upper human body detection sensor 5 are pyroelectric sensors, and passively detect a large body movement of the user US lying on the bed main body 1. The lateral human body detection sensor 4 detects the body movement of the user US in the height direction with respect to the bed surface of the bed body 1, and the upper human body detection sensor 5 detects the body movement of the user US in the lateral direction with respect to the bed surface. Thereby, the side human body detection sensor 4 and the upper human body detection sensor 5 detect the body movement in the state in which the user US lies and the body movement in the state in which the user US has woken up.

  Since the lateral human body detection sensor 4 and the upper human body detection sensor 5 have predetermined directivities within the detection range, the directing direction is set in the direction in which the user US lies from the rear or upper side of the bed body 1. By setting so as to be, it is detected that the user US has turned over or wakes up during sleep. For example, as shown in FIG. 1, the sleep state detection apparatus is provided with the side human body detection sensor 4 at a height for detecting the user US in the state lying on the bed body 1 and rising from the head. When the height direction from the floor surface of the bed body 1 is changed in three stages so as to be provided at a height for detecting the user US in the state, the direction of the side human body detection sensor 4 is set to the horizontal direction. Is set to, based on the second-stage lateral human body detection sensor 4 that detects the user US in the state where the head is lifted up, the fact that the user US has temporarily woken up is detected, and the number of wake-ups is determined. Can be counted.

  As described above, the side human body detection sensor 4 and the upper human body detection sensor 5 accurately detect the presence / absence of a wake-up even in a situation where the user US has a poor sleeping phase and information cannot be detected by the pressure mat sensor 2. It is possible to do. The sensor signals detected by the side human body detection sensor 4 and the upper human body detection sensor 5 are supplied to the A / D conversion unit 24 of the preprocessing unit 20.

  Each of the various sensors constituting the sensor unit 10 is not directly attached to the body of the user US, and can detect biological information as it is, so that it is removed by turning over the user US during sleep or the like. In addition, the burden imposed on the user can be reduced.

  The preprocessing unit 20 is converted by the four A / D conversion units 21, 22, 23, 24 that convert the sensor signals of analog signals into digital signals, and these A / D conversion units 21, 22, 23, 24 convert And a filter unit 25 for performing frequency separation processing on the digital signal.

  The filter unit 25 performs frequency separation processing on the digital signals converted by the A / D conversion units 21, 22, 23, and 24 and separates them into a plurality of frequency components. Here, the filter unit 25 performs different frequency separation processes set in advance according to each sensor signal.

  Specifically, the filter unit 25 separates a low-frequency component near 0.2 Hz from the digital biological signal based on the biological signal detected by the pressure mat sensor 2, and extracts this as a respiratory signal. Further, the filter unit 25 separates a high-frequency component near 1 Hz from the digital biological signal based on the biological signal detected by the pressure mat sensor 2, and extracts this as a heartbeat signal.

  The sleep state detection device determines whether or not the user US has shifted to the sleep state based on, for example, the period and amplitude stability within a predetermined time for the respiratory signal and the heartbeat signal, and estimates the sleep depth. be able to. Although details will be described later, the filter unit 25 extracts the main component signal of the snoring sound from the digital audio signal based on the audio signal detected by the microphone 3 in order to detect the presence or absence of apnea. The filter unit 25 supplies the plurality of digital signals subjected to frequency separation in this way to the sleep state detection unit 30.

  As described above, the preprocessing unit 20 performs processing necessary for detecting the sleep state of the user US based on the signal supplied from the sensor unit 10.

  The sleep state detection unit 30 determines the sleep state from the snoring utterance, the snoring detection unit 31 and the snoring state estimation unit 32, the apnea state estimation unit 33 for determining the apnea state, and the heart rate state. A heart rate state estimation unit 34 for determining the number of wake-ups, a wake-up number estimation unit 35 for determining the number of wake-ups, and a body motion state estimation unit 36 for determining the body motion state. Of these, the snoring detection unit 31, the snoring state estimation unit 32, and the apnea state estimation unit 33 are provided for detecting apnea syndrome, as will be described in detail later.

  The snoring detection unit 31, among the digital signals supplied from the filter unit 25 in the preprocessing unit 20, when the main component signal of the snoring sound based on the audio signal detected by the microphone 3 is at a predetermined level or higher, It is determined that the user US is screaming. The snoring detection unit 31 supplies a signal indicating that the snoring utterance has been detected to the snoring state estimation unit 32.

  When snoring is detected by the snoring detection unit 31, the snoring state estimation unit 32 estimates whether or not the snoring sound is steady. Although the details will be described later, the snoring state estimation unit 32 can estimate the snoring continuity by obtaining the correlation of the temporal change of the main component signal of the snoring sound. The snoring state estimation unit 32 supplies a signal indicating the estimated snoring state to the apnea state estimation unit 33.

  The apnea state estimation unit 33 estimates whether or not the user US has entered an apnea state based on the temporal change of the snoring state estimated by the snoring state estimation unit 32. Although the details will be described later, the apnea state estimation unit 33 can estimate the presence or absence of the apnea state based on the correlation of the principal component signals of the snoring sound obtained by the snoring state estimation unit 32.

  The heartbeat state estimation unit 34 estimates a heartbeat state based on a heartbeat signal based on a biological signal detected by the pressure mat sensor 2 among the digital signals supplied from the filter unit 25 in the preprocessing unit 20. At this time, the heart rate state estimation unit 34 compares the information stored in the sleep state storage unit 37 that stores the heart rate and respiration rate of the user US in a normal state with the heart rate based on the detected heart rate signal. Then, based on the difference, the heart rate state of the user US is estimated. The cycle of the heartbeat signal frequency-separated from the digital signal based on the biological signal detected by the pressure mat sensor 2 by the filter unit 25 in the preprocessing unit 20 is generally constant. Therefore, the heartbeat state estimation unit 34 can appropriately detect abnormalities such as palpitation, arrhythmia, and cardiac arrest during sleep by detecting the unsteady state and the cardiac arrest state of the heartbeat signal.

  Of the digital signals supplied from the filter unit 25 of the preprocessing unit 20, the wake-up number estimation unit 35 is a video signal imaged by the camera 6, a biological signal detected by the pressure mat sensor 2, and a side human body detection sensor. 4 estimates the number of times (frequency) that the user US has woken up during sleep based on the digital biological signal based on the biological signal detected by No. 4, and estimates the sleep depth and the like by combining with the peak frequency and volume of the snoring sound. To do. For example, the wake-up number estimation unit 35 detects a signal indicating that the user US has disappeared from the field of view of the camera 6, a signal indicating that the signal from the pressure mat sensor 2 has been interrupted, and a side where the sleeping position of the user US is detected. Based on the signals of the human body detection sensor 4 and the side human body detection sensor 4 that detects the wake-up position, it can be determined that a temporary wake-up has occurred.

  The wake-up number estimating unit 35 can also count the number of wake-ups based on the signal detected by the upper human body detection sensor 5. At this time, the wake-up number estimating unit 35 may use the video signal captured by the camera 6 in order to identify the wake-up before and after falling asleep. For example, when the brightness of the entire image captured by the camera 6 is high, the wake-up number estimation unit 35 determines that the room in which the sleep state detection device is installed is bright and before sleep, while the brightness is low. In this case, it can be determined that it is a time zone for sleeping.

  The body movement state estimation unit 36 is based on the biological signal detected by the pressure mat sensor 2 and the biological signal detected by the upper human body detection sensor 5 among the digital signals supplied from the filter unit 25 of the preprocessing unit 20. Based on the digital biological signal, the body movement state of the user US is estimated. Specifically, the body movement state estimation unit 36 estimates the body movement state of the user US based on the magnitude of the amplitude fluctuation of the digital biological signal and the length of time that the amplitude fluctuation continues. Thereby, the body movement state estimation part 36 can detect the frequency etc. of the user US turning over.

  Thus, the sleep state detection unit 30 detects the sleep state of the user US based on the signal supplied from the preprocessing unit 20.

  The health state detection unit 40 sets an score as a determination condition for determining that the health state of the user US is at an abnormal level, and whether or not the health state of the user US is at an abnormal level. An abnormality level estimation unit 42 that estimates the above, a timer unit 43 that counts time information, and an abnormality determination unit 44 that determines an abnormal state.

  The abnormal level determination setting unit 41 is abnormal if the various health states such as time and frequency (frequency) of falling into an apnea state, heartbeat instability, frequency of waking up (frequency), number of times of turning (frequency) are abnormal. Is determined as a score by the user US in advance. For example, in the abnormal level determination setting unit 41, “score 0 (none)” is set as an apnea setting score when the user US does not normally fall into an apnea state. At this time, the abnormal level determination setting unit 41 sets an abnormal level for each user US, for example, by setting by a switch operation of the user US. The score set in the abnormality level determination setting unit 41 is read by the abnormality determination unit 44.

  The abnormal level estimation unit 42 individually scores the above-described various health states based on the sleep state detected by the sleep state detection unit 30 and the time information counted by the timer unit 43, and the health state of the user US It is estimated whether or not is an abnormal level. The abnormality level estimation unit 42 supplies the obtained score to the abnormality determination unit 44.

  The timer unit 43 counts time information used when the abnormal level estimation unit 42 estimates whether or not the health state is an abnormal level.

  The abnormality determination unit 44 determines whether or not the user US is in an abnormal state based on the score read from the abnormality level determination setting unit 41 and the score supplied from the abnormality level estimation unit 42. For example, the abnormality determination unit 44 supplies from the abnormality level estimation unit 42 when “score 0” is set as the apnea setting score in the abnormality level determination setting unit 41, assuming that the patient does not normally fall into an apnea state. If the measured apnea score has any value, that is, if the apnea condition is recognized even a little, it is determined that the condition is abnormal. In addition, the abnormality determination unit 44, for example, frequently goes to the toilet of the user US based on the information obtained by scoring the number of wake-ups estimated by the wake-up number estimation unit 35 in the sleep state detection unit 30 by the abnormality level estimation unit 42. The operation is judged and it is judged whether or not it is an abnormal state.

  As described above, the health state detection unit 40 determines the health state (degree of abnormality) of the user US based on the sleep state detected by the sleep state detection unit 30.

  The reporting unit 50 includes a reporting destination setting unit 51 that sets a reporting destination RP, a reporting information setting unit 52 that sets information to be reported to the reporting destination RP, a reporting method selection unit 53 that selects a reporting method, A VOIP calling unit 54 as one of reporting methods for performing VOIP (Voice Over IP) calls, a mail transfer unit 55 as one of reporting methods for performing mail transfer, and a reporting method for converting a digital signal into an analog signal And a D / A (Digital to Analog) converter 56 as one.

  In the report destination setting unit 51, a report destination RP to be reported is arbitrarily registered in advance by the user US. For example, the telephone number or e-mail address of the report destination RP is registered in the report destination setting unit 51. Note that not only one notification destination RP but also a plurality of notification destinations RP can be registered.

  In the report information setting unit 52, the type of information to be reported to the report destination RP is arbitrarily set in advance by the user US. For example, the report information setting unit 52 is set to transmit a heartbeat signal when an abnormality occurs in the heartbeat. The notification information setting unit 52 also includes information based on various biological signals detected by the sensor unit 10 such as snoring detected by the microphone 3, respiration detected by the pressure mat sensor 2, images captured by the camera 6, and the like. It is also possible to set to transmit a plurality of information, such as Furthermore, in the notification information setting unit 52, a predetermined alarm or message indicating an abnormal state can be set instead of information based on various biological signals detected by the sensor unit 10.

  Based on the information on the report destination RP set in the report destination setting unit 51 and the information to be reported set in the report information setting unit 52, the report method selection unit 53 selects an appropriate report method (VOIP, mail, analog). Select the line). For example, the notification method selection unit 53 is abnormal in heart rate when a plurality of e-mail addresses are set in the notification destination setting unit 51 and the notification information setting unit 52 is set to transmit a heart rate signal. In the case where the notification occurs, the mail transfer unit 55 is selected as a notification method, and control is performed so as to transmit a heartbeat signal to any or all of the registered plurality of notification destinations.

  The VOIP call unit 54 is provided as one of reporting methods, and performs a VOIP call with the report destination RP. In the sleep state detection device, by providing the VOIP call unit 54, it is possible to perform a two-way call with the report destination RP, and not only transmits information indicating the health state to the report destination RP. It becomes possible to report the current status of the user US in real time.

  The mail transfer unit 55 is provided as one of reporting methods, and transfers electronic mail to the reporting destination RP. In the sleep state detection device, by providing the mail transfer unit 55, it is possible to appropriately report even if the user US falls into a situation where it cannot be verbally reported.

  The D / A converter 56 is provided as one of the reporting methods, and analogizes a digital audio signal detected by the microphone 3 and digitized, or a digital video signal imaged by the camera 6 and digitized. Re-convert to audio and video signals. By providing the D / A conversion unit 56, it is possible to report the current state of the user US as visual and / or auditory information, and the reporting destination RP can accurately grasp the situation.

  As described above, when the abnormality determination unit 44 in the health state detection unit 40 determines that the user US is in an abnormal state, the notification unit 50 notifies the predetermined notification destination RP in an appropriate notification form. Report to that effect.

[Apnea state detection operation by sleep state detection device]
In such a sleep state detection device, as described above, it is possible to detect whether or not the user US is in an apnea state during sleep. Here, the apnea syndrome has a characteristic that it does not suddenly fall into an apneic state but falls into an apneic state after uttering a non-steady snoring from a state of making a steady snoring. Therefore, in order to automatically detect an apnea state, it is important to detect snoring and grasp the state of snoring. This apnea detection method will be described with reference to FIGS. 4 (a) and 4 (b).

  Ordinary snoring is known to have a unique peak frequency around 1 kHz, and its period is relatively constant. Therefore, in the sleep state detection apparatus, the main component of the snoring sound is extracted by the filter unit 25 of the preprocessing unit 20 from the digital audio signal based on the audio signal detected by the microphone 3 in the sensor unit 10. Then, the snoring detection unit 31 determines that the user US is snoring when the principal component signal is at a predetermined level or higher.

  Here, since the period of snoring during normal sleep is substantially constant, the principal component signal has an amplitude peak at a constant time interval as shown in, for example, time region A in FIG. Appear. On the other hand, the apnea state of the user US who has an apnea syndrome, the main component signal of the snoring sound, as shown in the time region B in FIG. The amplitude becomes random as compared with the period and amplitude in the time domain A in FIG. 4A, and becomes unsteady and unstable.

  In the sleep state detection device, in order to quantify this situation, when snoring is detected by the snoring detection unit 31, the snoring state estimation unit 32 estimates whether or not the snoring uttered is steady. To do. At this time, the snoring state estimation unit 32 obtains the correlation coefficient r of the main component signal of the snoring sound as a parameter.

The correlation coefficient is expressed as shown in the following equation, and takes a value of 0 to 1 as shown in FIG.

  When the main component signal of the snoring sound is stationary, as shown in the time domain A in FIG. 4 (b), it takes a value near 1.0, and when it is unsteady, FIG. As shown in the time region B in (b), the value changes at a lower value than that in the stationary case. Therefore, the snoring state estimation unit 32 sets a predetermined threshold value TH for the correlation coefficient r, and estimates that it is a sign of an apnea state when the correlation coefficient falls below the threshold value TH.

  Further, when the snoring is changed from the unstable state to the apnea state, the main component signal of the snoring sound has an amplitude of 0 as shown in a time region C in FIG. Therefore, after the snoring state estimation unit 32 estimates the signs of an apnea state, when the amplitude of the main component signal of the snoring sound is lower than a predetermined level, the apnea state estimation unit 33 causes the user US to Presumed that the patient had apnea.

  In this way, it is possible to appropriately detect whether or not the user US is in an apnea state during sleep. In the sleep state detection device, the apnea state is not detected based only on the audio signal detected by the microphone 3, but based on the respiratory signal based on the biological signal detected by the pressure mat sensor 2. Thus, the determination may be made using a mathematical expression for calculating the correlation coefficient r in a composite manner.

[Effect of the embodiment]
As described above in detail, according to the sleep state detection device, the ambient sound of the bed main body 1 is detected by the microphone 3 and the snoring detection unit 31 detects the snoring utterance based on the obtained audio signal. The snoring state estimation unit 32 estimates whether the detected snoring is steady. Based on the estimated snoring state, whether or not the user US has entered an apnea state is estimated by the apnea state estimation unit 33, and whether the estimated duration of the apnea state is an abnormal state or not. The abnormality determining unit 44 determines whether or not the user US is in an abnormal state, and the reporting unit 50 reports to the reporting destination RP.

  Thus, an abnormal state caused by apnea occurring during sleep can be reliably detected without going to a specific inspection organization such as a hospital, and an early response can be made by reporting that fact.

  Moreover, the snoring state estimation unit 32 can estimate the sleep depth of the user US by detecting a frequency specific to snoring, for example, around 1 kHz, by performing frequency analysis on the signal obtained by the microphone 3.

  Furthermore, since the abnormality level determination setting unit 41 that sets a determination condition for determining that the health state of the user US is at an abnormal level is set as a reporting condition, the condition for determining the abnormality is different depending on the user US. However, it is possible to appropriately report the abnormal state.

  Furthermore, since the report destination setting unit 51 for setting the report destination RP and the report information setting unit 52 for setting information to be reported to the report destination RP are provided, the abnormal state of the user US can be determined appropriately. It is possible to report only information necessary for the user US and the report destination RP. Furthermore, since the notification method selection unit 53 for selecting an appropriate notification method is provided, it is possible to perform appropriate notification according to the communication environment of the notification destination RP, the contents of the abnormal state, and the like.

  The above-described embodiment is an example of the present invention. For this reason, the present invention is not limited to the above-described embodiment, and various modifications can be made depending on the design and the like as long as the technical idea according to the present invention is not deviated from this embodiment. Of course, it is possible to change.

It is a figure which shows the structure of the sleep state detection apparatus shown as embodiment of this invention. It is a block diagram which shows the circuit structure of the sleep state detection apparatus. It is a figure which shows the structure of a pneumatic mat sensor. (a) is a figure which shows the example of a waveform of the principal component signal of a snoring sound, (b) is a figure which shows the example of a waveform of the correlation coefficient of the principal component signal of a snoring sound.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Bed main body 2 Pressure mat sensor 2a Air tube 2b Microphone 3 Microphone 4 Side human body detection sensor 5 Upper human body detection sensor 6 Camera 7 Signal processing module 8 Notification module 10 Sensor part 20 Pre-processing part 21, 22, 23, 24 A / D conversion unit 25 filter unit 30 sleep state detection unit 31 snoring detection unit 32 snoring state estimation unit 33 apnea state estimation unit 34 heart rate state estimation unit 35 wake-up number estimation unit 36 body movement state estimation unit 37 sleep state storage unit 40 health state Detection unit 41 Abnormal level determination setting unit 42 Abnormal level estimation unit 43 Timer unit 44 Abnormality determination unit 50 Notification unit 51 Notification destination setting unit 52 Notification information setting unit 53 Notification method selection unit 54 VOIP call unit 55 Mail transfer unit 56 D / A Conversion unit GW gateway NT network RP report destination US Over The

Claims (5)

In the sleep state detection device for detecting the sleep state of the user at the time of sleep and reporting the health state of the user to a predetermined report destination,
Ambient sound detecting means for detecting ambient sounds of the bed body lying when the user goes to bed;
Based on the audio signal detected by the ambient sound detection means, snoring detection means for detecting the utterance of snoring,
Snoring state estimation means for estimating whether snoring detected by the snoring detection means is stationary;
An apnea state estimating means for estimating whether the user has entered an apnea state from the state of snoring estimated by the snoring state estimating means;
An abnormality determining means for determining whether the apnea condition estimated by the apnea condition estimating means is an abnormal condition;
A sleep state detection apparatus comprising: notification means for reporting to the report destination when the abnormality determination means determines that the user is in an abnormal state. The snoring detection means detects a snore-specific frequency among the audio signals detected by the ambient sound detection means,
The sleep state detection device according to claim 1, wherein the snoring state estimation unit estimates the depth of sleep of the user based on the frequency detected by the snoring detection unit. The abnormality determining means is
An abnormal level estimator for estimating an abnormal level of the apnea state estimated by the apnea state estimating means;
An abnormal level determination setting unit for inputting a determination condition for determining that the health state of the user is an abnormal level;
The sleep state according to claim 1, wherein when the abnormal level estimated by the abnormal level estimation unit satisfies a determination condition set in the abnormal level determination setting unit, the sleep state is determined as an abnormal state. Detection device. The reporting means is
Report destination setting means for setting any one of a plurality of registered report destinations;
The sleep state detection apparatus according to claim 1, further comprising: a notification method selection unit that selects a notification method to the notification destination set by the notification destination setting unit. When the user is determined to be in an abnormal state, the notification unit notifies the notification destination of an audio signal detected by the ambient sound detection unit and a predetermined alarm or message indicating that the user is in an abnormal state. The sleep state detection apparatus according to claim 1, further comprising: a report information setting unit that determines the sleep state.

Images