Snoring determination system, snoring determination method and program

Abstract

A snore determination system capable of appropriately determining whether or not a subject is snoring from a voice signal from the subject. [Solution] The snoring determination system includes a measurement means (S201) for measuring the change in noise level or sound pressure level of breathing sounds emitted by a sleeping person, a frequency spectrum acquisition means (S202) for acquiring a frequency spectrum of the change in the noise level or sound pressure level, a peak identification means (S203) for identifying a plurality of frequency peaks based on the frequency spectrum, and a determination means (S205) for determining whether or not the person is snoring based on the intensity ratio between a first peak having the greatest intensity and a second peak having the next greatest intensity among the plurality of frequency peaks.

Description

[Technical field]

[0001] The present invention relates to a snore detection system, a snore detection method, and a program. [Background technology]

[0002] The following Patent Document 1 discloses a sleep aid that can suppress snoring of a person being measured while sleeping and provide a comfortable sleep, and this sleep aid is equipped with a detection device that obtains an audio signal using a microphone installed near the person being measured and detects snoring from the audio signal. Specifically, the detection device extracts the envelope of the audio signal, and if the periodicity of the envelope continues for a certain period of time, it is determined that the sleeping person being measured is snoring. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2005-13754 A (FIG. 12) Summary of the Invention [Problem to be solved by the invention]

[0004] A voice signal acquired by a microphone from a sleeping subject contains various frequency components, and when the subject snores, the ideal periodicity of the envelope does not necessarily continue for a certain period of time.

[0005] The present invention has been made in consideration of the above problems, and its purpose is to provide a snore determination system, a snore determination method, and a program that can appropriately determine whether or not a subject is snoring from an audio signal from the subject. [Means for solving the problem]

[0006] (1) In order to solve the above problem, the snoring determination system of the present invention includes a measuring means for measuring a change in the noise level or sound pressure level of respiratory sounds emitted by a sleeping person, a frequency spectrum acquiring means for acquiring a frequency spectrum of the change in the noise level or sound pressure level, a peak identifying means for identifying a plurality of frequency peaks based on the frequency spectrum, and a determining means for determining whether or not the person is snoring based on the intensity ratio between a first peak having the greatest intensity and a second peak having the second greatest intensity among the plurality of frequency peaks.

[0007] (2) In the snoring determination system described in (1), the determination means may determine whether or not the subject is snoring based on the magnitude of the intensity of the first peak.

[0008] (3) In the snoring determination system described in (1) or (2), the measuring means may be a sound level meter.

[0009] (4) The snoring determination system according to any one of (1) to (3) may further include a Doppler sensor for acquiring Doppler data indicating a movement of the chest of the subject. The determination means may determine whether or not a waveform of the Doppler data is synchronized with a waveform of the noise level or sound pressure level based on the Doppler data and a transition of the noise level or sound pressure level. Furthermore, the determination means may determine whether or not the subject is snoring based on whether or not they are synchronized.

[0010] (5) In the snoring detection system described in (4), the detection means may determine whether the waveforms of the Doppler data and the noise level or sound pressure level are synchronized based on the difference between the timing of the maximum point of the waveform of the Doppler data and the timing of the maximum point of the waveform of the noise level or sound pressure level.

[0011] (6) In the snoring determination system described in (4), the determination means may determine that the subject is snoring when the difference between the respiratory rate calculated based on the frequency spectrum of the Doppler data and the respiratory rate calculated based on the frequency spectrum of the noise level or sound pressure level change is less than a predetermined threshold value.

[0012] (7) In the snoring determination system described in any one of (1) to (6), the system may further include a sleep depth acquisition means for acquiring a sleep depth of the person being measured, and the determination means may determine whether or not the person being measured is snoring based further on the sleep depth of the person being measured.

[0013] (8) The snoring determination method of the present invention includes a measurement step of measuring a change in noise level or sound pressure level of respiratory sounds emitted by a sleeping person, a frequency spectrum acquisition step of acquiring a frequency spectrum of the change in noise level or sound pressure level, a peak identification means for identifying a plurality of frequency peaks based on the frequency spectrum, and a determination step of determining whether or not the person is snoring based on the intensity ratio between a first peak having the greatest intensity and a second peak having the next greatest intensity among the plurality of frequency peaks.

[0014] (9) A program according to the present invention is a program for causing a computer to function as a means for acquiring a transition in noise level or sound pressure level of breathing sounds emitted from a sleeping subject, a means for acquiring a frequency spectrum of the transition in the noise level or sound pressure level, a means for identifying a plurality of frequency peaks based on the frequency spectrum, and a means for determining whether or not the subject is snoring based on an intensity ratio between a first peak having the greatest intensity and a second peak having the second greatest intensity among the plurality of frequency peaks. This program may be stored in a computer-readable information storage medium such as a semiconductor memory or a magneto-optical disk. Effect of the Invention

[0015] According to the present invention, it is possible to suitably determine whether or not a subject is snoring from a voice signal from the subject. [Brief description of the drawings]

[0016] [Figure 1] 1 is a configuration diagram of a snore determination system according to a first embodiment of the present invention. [Diagram 2] 4 is a flowchart showing the operation of the signal processing device according to the first embodiment. [Diagram 3] FIG. 13 is a diagram showing a frequency spectrum of sound data when the subject is not snoring. [Figure 4] FIG. 13 is a diagram showing a frequency spectrum of sound data when a subject is snoring. [Diagram 5] FIG. 11 is a configuration diagram of a snore determination system according to a second embodiment. [Figure 6] FIG. 13 is a diagram showing sound data and Doppler data of a subject while sleeping. [Figure 7] 10 is a flowchart showing the operation of a signal processing device according to the second embodiment. [Figure 8] FIG. 11 is a flowchart showing the operation of a signal processing device according to the third embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0017] Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings.

[0018] (Embodiment 1) Fig. 1 is a configuration diagram of a snore determination system according to a first embodiment of the present invention. The snore determination system 1 shown in the figure is arranged, for example, beside a bed in a bedroom, and determines whether or not a person to be measured sleeping on the bed is snoring based on breathing sounds made by the person.

[0019] As shown in the figure, the snoring determination system 1 includes a sound collection unit 2 and a signal processing device 3. The sound collection unit 2 is for acquiring sound data of breathing sounds emitted by a sleeping subject, and is configured using, for example, a microphone or a sound level meter. When a microphone is used, time series data showing the transition of sound pressure level (sound volume) is generated from the acquired raw data of breathing sounds, and this is used as sound data. The sound data may be generated from the raw data of breathing sounds acquired by the microphone, for example, by the sound collection unit 2 itself, or may be generated by the signal processing device 3.

[0020] In contrast, when a sound level meter is used, the sound data acquired by the sound level meter indicates the transition of the noise level of the breathing sounds emitted by the subject (time series data). From the viewpoint of protecting privacy, it is preferable to adopt a sound level meter rather than a microphone as the sound collection unit 2. In the following, the explanation will be continued assuming that the sound collection unit 2 is a sound level meter. Note that sound level meters are legal measuring instruments as defined by the Measurement Act, and there are two types: ordinary sound level meters and precision sound level meters. Either sound level meter may be adopted as the sound collection unit 2.

[0021] The signal processing device 3 acquires sound data indicating a transition of the noise level from the sound collection unit 2, and judges whether the subject is snoring or not based on the sound data. At this time, the signal processing device 3 calculates a frequency spectrum of the transition of the noise level indicated by the sound data for a certain period, and judges whether the subject is snoring or not based on the frequency spectrum. The signal processing device 3 may be configured using a known computer including, for example, a CPU and a memory.

[0022] 2 is a flow diagram showing a specific operation of the signal processing device 3. The operation shown in the figure is realized by executing a signal processing program according to this embodiment in the signal processing device 3, which is a computer. The signal processing program is executed periodically, for example, once a day, when the subject wakes up.

[0023] In the signal processing device 3, first, sound data is acquired from the sound collection unit 2 (S201). For example, this sound data indicates a change in the noise level of the breathing sound of the subject from the time of falling asleep to the time of waking up. Next, sound data of a predetermined time window (for example, one minute) is cut out from the sound data thus acquired, and FFT (Fast Fourier Transform) is performed on the cut out sound data. This obtains a frequency spectrum for the sound data of the time window. This process is performed from the beginning to the end of the sound data, that is, for all time windows set from the time of falling asleep to the time of waking up (S202). Then, peaks included in the frequency spectrum of each time window are identified. Specifically, the frequency and intensity of each peak included in the frequency spectrum are identified. Then, the frequency and intensity of the peak with the largest intensity (first peak) and the frequency and intensity of the peak with the second largest intensity (second peak) are identified (S203).

[0024] Thereafter, the processes of S204 to S206 are repeated for each time window. That is, it is determined whether the intensity of the first peak is equal to or greater than a threshold (S204). If the intensity of the first peak is less than the threshold, it is considered that the subject was not snoring during the time period related to that time window, and the process proceeds to the next time window. On the other hand, if the intensity of the first peak is equal to or greater than the threshold, it is next determined whether the intensity of the first peak is α times or greater than the intensity of the second peak (S205). Here, α is a value greater than 1, and may be, for example, a value of about 1.8 to 2.0.

[0025] Fig. 3 is a diagram showing a frequency spectrum of sound data when the subject is not snoring, and Fig. 4 is a diagram showing a frequency spectrum of sound data when the subject is snoring. In Fig. 3, the difference between the intensity of the first peak 20 and the intensity of the second peak 21 is small, and in this embodiment, it is determined that the subject is not snoring. On the other hand, in Fig. 4, the difference between the intensity of the first peak 22 and the second peak 23 is sufficiently large, and in this embodiment, it is determined that the subject is snoring.

[0026] 2, if the intensity of the first peak is not α times or more the intensity of the second peak, the process proceeds to the next time window (S205).On the other hand, if the intensity of the first peak is α times or more the intensity of the second peak, the time period of the time window is recorded in a file as a period during which the subject was snoring (S206), and the process proceeds to the next time window.

[0027] According to the snore determination system 1 described above, since the snoring determination is performed using a sound level meter, privacy protection can be realized. In addition, since the frequency spectrum of the sound data is acquired and the presence or absence of snoring by the subject is determined based on the intensity ratio between the first peak and the second peak, the accuracy of the determination can be improved. Furthermore, if the intensity of the first peak is less than the threshold, it is determined that no snoring is occurring, thereby further improving the accuracy of the snore determination.

[0028] (Embodiment 2) FIG. 5 is a configuration diagram of a snore determination system according to a second embodiment. The snore determination system 10 shown in the figure is also arranged, for example, beside a bed in a bedroom, and determines whether or not a subject is snoring from breathing sounds emitted by the subject while sleeping on the bed. The snore determination system 10 includes a sound collection unit 2, a Doppler sensor 40, and a signal processing device 30. The sound collection unit 2 has a configuration similar to that of the first embodiment. The Doppler sensor 40 is installed, for example, near the bed, facing the chest of the subject. Microwaves are emitted from the Doppler sensor 40, and the reflected waves from the chest of the subject are received by the Doppler sensor 40. The reflected waves are frequency-shifted due to the Doppler effect, and the respiratory rate and heart rate of the subject can be obtained by observing this. The reflected waves are detected as Doppler data including an I signal, which is an in-phase component of the transmitted wave, and a Q signal, which is an orthogonal component, and are output in digital form to the signal processing device 30. The Doppler data indicates the movement of the chest of the subject.

[0029] The signal processing device 30 according to the second embodiment may be configured using a known computer including, for example, a CPU and a memory, similar to the signal processing device 3 according to the first embodiment. In particular, the signal processing device 30 uses Doppler data acquired from the Doppler sensor 40 in addition to sound data acquired from the sound collection unit 2 to more accurately determine whether or not the subject is snoring.

[0030] In addition to the component derived from breathing, the Doppler data also contains components derived from body movement and heart rate. Therefore, the signal processing device 3 applies a predetermined band-limiting filter (a band-pass filter that extracts the frequency band of breathing) to the raw Doppler data acquired from the Doppler sensor 40 before executing the snoring determination process.

[0031] Fig. 6 is a diagram showing sound data and Doppler data (after application of a filter) of a subject sleeping. The upper part of the figure shows sound data obtained from the subject, with the vertical axis showing the amplitude of the sound data and the horizontal axis showing time. The lower part of the figure shows Doppler data obtained from the subject at the same time, with the vertical axis showing the amplitude of the Doppler data and the horizontal axis showing time. In addition, many two-dot chain vertical lines pass through the maximum points of the sound data. In addition, one-dot chain circles drawn on the Doppler data indicate the maximum points of the Doppler data.

[0032] As shown in the figure, the sound data of the subject during sleep and the Doppler data showing the movement of the subject's chest are usually synchronized. That is, the maximum point of the Doppler data is present near the vertical line of the two-dot chain line. On the other hand, if the sound data and the Doppler data are not synchronized, for example, it is considered that the sound data indicates a sound that is not originating from the subject, and in such a case, the accuracy of snoring determination decreases. Therefore, in the second embodiment, it is determined whether the sound data of the subject during sleep and the Doppler data showing the movement of the subject's chest are synchronized, and the presence or absence of snoring by the subject is determined based on the result.

[0033] Specifically, whether or not the two data are synchronized is determined based on the difference between the timing of the maximum point in the sound data and the timing of the maximum point in the Doppler data. For example, the magnitude of the difference between the timing (time) of each maximum point in the sound data and the timing of the maximum point in the Doppler data closest to that timing is calculated. Then, if the average value of the difference is equal to or less than a predetermined threshold value (e.g., 0.5 seconds), it may be determined that the sound data and the Doppler data are synchronized.

[0034] Here, it may be determined whether the two data are synchronized based on the difference between the timing of the minimum points of the sound data and the timing of the minimum points of the Doppler data, or based on the difference between the timing of the maximum points and minimum points (extreme value points) of the sound data and the timing of the maximum points and minimum points of the Doppler data.

[0035] Fig. 7 is a flow diagram showing the operation of the signal processing device 30. The operation shown in the figure is realized by executing a signal processing program according to this embodiment in the signal processing device 30, which is a computer. The signal processing program is also executed periodically, for example, once a day, when the subject wakes up. In the figure, the processes of S701 to S705 and S710 are the same as the processes of S201 to S205 and S206 in Fig. 2, respectively, and therefore will not be described here.

[0036] As shown in the figure, when it is determined in S705 that the intensity of the first peak is α times or more the intensity of the second peak, a maximum point in the sound data of the currently processed time window is identified (S706) in this embodiment 2. In addition, Doppler data of the same time window as the currently processed time window is acquired (S707), and a maximum point in the acquired Doppler data is identified (S708).

[0037] Then, the deviation between the maximum point identified in S706 and the maximum point identified in S708 is evaluated, and it is determined whether the deviation is equal to or smaller than a reference value (S709). For example, as described above, it is determined whether the average value of the magnitude of the deviation is equal to or smaller than a threshold value. If the deviation is greater than the reference value, the process moves to the next time window. On the other hand, if the deviation is equal to or smaller than the reference value, the time period of the time window is recorded as a transition section (S710), and the process moves to the next time window.

[0038] According to the above-described second embodiment, the presence or absence of snoring is judged by using Doppler data in combination, so that the accuracy of the judgment can be improved. In particular, when the sound data includes sounds other than the breathing sounds of the subject, it is possible to prevent the subject from being erroneously judged as snoring.

[0039] (Embodiment 3) In the second embodiment, whether the sound data and the Doppler data are synchronized is determined by comparing the extreme points of both data. Alternatively, whether the sound data and the Doppler data are synchronized may be determined by comparing the respiration rate calculated from the sound data and the respiration rate calculated from the Doppler data.

[0040] Fig. 8 is a flow diagram showing a modified example of the operation of the signal processing device 30. The operation shown in the figure is realized by executing a signal processing program according to this embodiment in the signal processing device 30, which is a computer. The signal processing program is also executed periodically, for example, once a day, when the subject wakes up. In the figure, the processes of S801 to S805 and S810 are the same as the processes of S201 to S205 and S206 in Fig. 2, respectively, and therefore will not be described here.

[0041] As shown in the figure, when it is determined in S805 that the intensity of the first peak is α times or more the intensity of the second peak, in this embodiment 3, the respiration rate of the subject is calculated based on the frequency of the first peak contained in the frequency spectrum of the sound data of the currently processed time window (S806). Next, Doppler data of the currently processed time window is acquired (S807), and the respiration rate of the subject is calculated based on the acquired Doppler data (S808). Specifically, the Doppler data is subjected to FFT processing to obtain a frequency spectrum. Then, the respiration rate is acquired from the frequency of the strongest peak contained in the frequency spectrum.

[0042] Thereafter, it is determined whether the respiration rate based on the sound data and the respiration rate based on the Doppler data are similar to each other (S809). For example, if the difference between the two respiration rates is less than a threshold, it is determined that the two respiration rates are similar to each other. If the two respiration rates are not similar to each other, the process moves to the next time window. On the other hand, if the two respiration rates are similar to each other, the time period of the time window is recorded as a breathing interval (S810), and the process moves to the next time window.

[0043] In the third embodiment described above, the presence or absence of snoring is judged by using Doppler data in combination, so that the accuracy of the judgment can be improved. In particular, when the sound data includes sounds other than the breathing sounds of the subject, it is possible to prevent the subject from being erroneously judged as snoring.

[0044] The present invention is not limited to the above-mentioned embodiments, and various modifications are possible, and such modifications are also included in the scope of the present invention. For example, since snoring occurs during so-called REM sleep, the sleep depth of the subject may be acquired, and the presence or absence of snoring by the subject may be determined based on the acquired sleep depth. In other words, if the subject is not in REM sleep, the determination that the subject is snoring may be suppressed or stopped. For example, the presence or absence of the subject's body movement may be obtained from a moving image of the subject or a load sensor installed on the bed, and the sleep depth may be determined from the information. In addition, the heart rate of the subject may be calculated from Doppler data, and the sleep depth may be determined based on the heart rate. [Explanation of symbols]

[0045] 1,10 snoring detection system, 2 sound collection unit, 3,30 signal processing device, 40 Doppler sensor.

Claims

1. A measuring means for measuring a transition of a noise level or a sound pressure level of a breathing sound emitted from a subject during sleep; a frequency spectrum acquisition means for acquiring a frequency spectrum of the transition of the noise level or the sound pressure level; a peak identification means for identifying a plurality of frequency peaks based on the frequency spectrum; a determination means for determining whether the subject is snoring based on an intensity ratio between a first peak having the highest intensity and a second peak having the second highest intensity among the plurality of frequency peaks; A snoring detection system comprising:

2. The snore detection system according to claim 1, The snoring determination system, wherein the determination means determines whether or not the subject is snoring based on the intensity of the first peak.

3. The snore detection system according to claim 1, The snoring determination system, wherein the measuring means is a sound level meter.

4. The snore detection system according to claim 1, a Doppler sensor for acquiring Doppler data indicative of chest movement of the subject; The determination means determines whether or not the waveform of the Doppler data and the waveform of the noise level or sound pressure level are synchronized based on the Doppler data and the changes in the noise level or sound pressure level, and determines whether or not the person being measured is snoring based on whether or not they are synchronized.

5. The snore determination system according to claim 4, The determination means determines whether or not the waveforms of the Doppler data and the noise level or sound pressure level are synchronized based on the difference between the timing of the extreme points of the waveform of the Doppler data and the timing of the extreme points of the waveform of the noise level or sound pressure level.

6. The snore determination system according to claim 4, The snoring determination system, wherein the determination means determines that the subject is snoring when the difference between the respiratory rate calculated based on the frequency spectrum of the Doppler data and the respiratory rate calculated based on the frequency spectrum of the change in the noise level or sound pressure level is less than a predetermined threshold value.

7. The snore detection system according to claim 1, The method further includes a sleep depth acquisition means for acquiring a sleep depth of the subject, The snoring determination system, wherein the determination means determines whether or not the subject is snoring based further on the depth of sleep of the subject.

8. A measurement step of measuring a transition of a noise level or a sound pressure level of a breathing sound emitted from a subject during sleep; a frequency spectrum acquisition step of acquiring a frequency spectrum of the transition of the noise level or the sound pressure level; a peak identification means for identifying a plurality of frequency peaks based on the frequency spectrum; a determining step of determining whether or not the subject is snoring based on an intensity ratio between a first peak having the highest intensity and a second peak having the second highest intensity among the plurality of frequency peaks; A method for determining snoring comprising:

9. A means for acquiring a transition of a noise level or a sound pressure level of a breathing sound emitted from a subject during sleep; A means for acquiring a frequency spectrum of the change in the noise level or the sound pressure level; means for identifying a plurality of frequency peaks based on the frequency spectrum; and a means for determining whether the subject is snoring based on an intensity ratio between a first peak having the highest intensity and a second peak having the second highest intensity among the plurality of frequency peaks; A program that makes a computer function as a

Images