JP6565401B2 - Sleeping habit analysis device, sleeping habit analysis method and program thereof - Google Patents

Description

  Embodiments of the present invention relate to a sleep state analyzing apparatus, a sleep state analyzing method, and a program thereof.

  Sleep apnea syndrome awakens the brain when resuming breathing from an apnea state and reduces the depth of sleep, leading to increased daytime fatigue and reduced concentration. It is a disease that increases the possibility of causing etc. In recent years, sleep apnea syndrome has been known as a disease that causes a combination of various lifestyle-related diseases such as cardiovascular disease and diabetes, and is recommended for early treatment and improvement of lifestyle and sleep environment. is there.

  However, sleep apnea syndrome has the disadvantage that there are few subjective symptoms and patients are not sufficiently manifested. For this reason, development of the new technique which can fully reveal the patient of sleep apnea syndrome is desired.

JP2013-236925A Japanese National Patent Publication No. 11-505146

  The problem to be solved by the present invention is to provide a sleep state analyzing apparatus, a sleep state analyzing method, and a program thereof that can realize the manifestation of a patient with sleep apnea syndrome.

  According to the embodiment, the sleep state analyzing apparatus can be worn on a human body, and the sleep state analyzing apparatus includes a measurement unit that measures a wearer's acceleration, and a first threshold value of the measured acceleration. Based on the first generation means for generating the first time-series data indicating the intensity of the wave in the above frequency band, and the generated first time-series data, the tendency of the wearer's wrinkles is determined. Determining means; and calculating means for calculating a risk of developing sleep apnea syndrome based on the determined tendency of wrinkles.

The figure which shows the schematic structural example of the sleep state analyzing system which concerns on each embodiment. The figure which shows the hardware structural example of the biosensor apparatus which concerns on each embodiment. The figure which shows the circuit structural example of the biosensor apparatus which concerns on each embodiment. The block diagram which shows an example of a function structure of the sleep state analysis application program which concerns on 1st Embodiment. 8 is a flowchart showing an example of a procedure of feature amount analysis section determination processing according to the embodiment. The figure which shows an example of the DCz waveform of the acceleration shown by the acceleration signal which concerns on the same embodiment. The figure which shows an example of the ACz waveform of the acceleration shown by the acceleration signal which concerns on the same embodiment. The figure which shows the various waveforms obtained by performing a predetermined | prescribed process with respect to the healthy person's ACz waveform in the feature-value analysis area which concerns on the embodiment, and the said ACz waveform. The figure which shows the various various waveforms obtained by performing a predetermined | prescribed process with respect to the ACz waveform of the test subject of sleep apnea syndrome in the feature-value analysis area which concerns on the embodiment, and the said ACz waveform. 8 is a flowchart showing an example of a procedure of first feature point calculation processing according to the embodiment. 8 is a flowchart showing an example of a procedure of second feature point calculation processing according to the embodiment. The flowchart which shows an example of the procedure of the 1st wrinkle tendency determination process which concerns on the same embodiment. The flowchart which shows an example of the procedure of the 2nd wrinkle tendency determination process which concerns on the same embodiment. The flowchart which shows an example of the procedure of the 3rd habit tendency determination process which concerns on the same embodiment. The figure which shows an example of the onset risk calculation table which concerns on the same embodiment. The figure which shows another example of the onset risk calculation table which concerns on the same embodiment. The block diagram which shows an example of a function structure of the sleep state analysis application program which concerns on 2nd Embodiment. The flowchart which shows an example of the procedure of the sleeping posture estimation process which concerns on the embodiment. The figure which shows an example of the data structure of the matching result information which concerns on the same embodiment. The block diagram which shows an example of a function structure of the sleep state analysis application program which concerns on 3rd Embodiment.

Hereinafter, embodiments will be described with reference to the drawings.
<First Embodiment>
FIG. 1 is a diagram illustrating a schematic configuration example of a sleep state analyzing system according to each embodiment. As shown in FIG. 1, the sleep state analyzing system 1 includes a biosensor device 10 and an electronic device 11, and the biosensor device 10 and the electronic device 11 are connected to be communicable. Note that some of the various functions of the biosensor device 10 may be realized by the electronic device 11. The biosensor device 10 is small, light, and thin, and is driven by a battery (for example, a built-in secondary battery). The biosensor device 10 is affixed to the human body (the chest surface) with, for example, an adhesive tape so that biometric information can be constantly measured. In addition to attachment by attachment, the attachment method to the human body may be attachment by a band, attachment by a belt, implantation by clothes, or the like. The biometric sensor device 10 has a function of simultaneously measuring a plurality of pieces of biological information such as pulse waves, electrocardiograms, temperatures, and accelerations, and sending the measured biological information to the electronic device 11 wirelessly. The biosensor device 10 also has a function of analyzing the measured biometric information and wirelessly sending analysis result information obtained as a result of the analysis to the electronic device 11. Furthermore, the biosensor device 10 also has a function of receiving a control signal or the like from the electronic device 11 wirelessly.

  The electronic device 11 is a module capable of displaying biological information and analysis result information sent from the biological sensor device 10, and corresponds to, for example, a smartphone or a tablet computer. An application for displaying biological information and analysis result information transmitted from the biological sensor device 10 is installed in the electronic device 11 in advance.

  Note that the biosensor device 10 and the electronic device 11 may be connected to each other so that they can communicate with each other by wire instead of wirelessly.

Here, the hardware configuration of the biosensor device 10 will be described with reference to FIG.
Although the biosensor device 10 includes a plurality of sensors, the analog front end of each sensor has different specifications for each sensor, so that both flexibility and high performance are required, and the size may increase. However, here, a sensor module of several millimeters square is realized by integrating a plurality of analog front ends, a CPU, and the like on a single chip using the pseudo SoC technology. The pseudo SoC technology is a technology that achieves both a size reduction equivalent to SoC and a design freedom equivalent to SiP by integrating components on a wafer. By connecting a small number of peripheral components such as an antenna and a battery to this module, the biosensor device 10 that is small, light (about 10 gram), and thin (about several mm) can be realized. It is also possible to reduce the size by using a component built-in board technology or a configuration using a dedicated LSI.

  The biosensor device 10 has, for example, an elliptical shape with a major axis of about several centimeters. As shown in FIG. 2, an electrocardiogram electrode (R) 20a, an electrocardiogram electrode (L) 20b, a photoelectric sensor are provided on the adhesion surface to the human body. A unit 22, a temperature sensor 24, and a charging terminal 26 are arranged. Since the electrocardiogram electrodes 20a and 20b are required to be positioned on the left and right sides of the heart, they are arranged at intervals along the long axis. The photoelectric unit 22 optically detects a pulse wave, and a window portion made of a transparent material that transmits light is provided on the front surface thereof.

  FIG. 3 is a block diagram illustrating a circuit configuration of the biosensor device 10. In addition to the electrocardiogram electrodes 20a and 20b, the photoelectric unit 22, the temperature sensor 24, and the charging terminal 26, the biosensor device 10 includes an electrocardiograph 30, an acceleration sensor 32, a pulse wave meter 34, and Bluetooth (registered trademark). A module 36, a system controller 38, an embedded controller (EC) 40, a lithium secondary battery 42, a CPU 44, a main memory 46, a BIOS-ROM 48, a flash memory 50, and the like are included.

  The electrocardiogram electrode (R) 20a and the electrocardiogram electrode (L) 20b are connected to an electrocardiograph 30 which is an analog front end for electrocardiogram. The electrocardiograph 30 obtains an electrocardiogram by analyzing a time-series signal obtained by sampling a potential difference between the electrocardiogram electrode (R) 20a and the electrocardiogram electrode (L) 20b.

  The photoelectric unit 22 is for detecting a volume pulse wave, and includes a light emitting element (for example, a blue LED) 22a as a light source and a photodiode (PD) 22b as a light receiving unit. A transparent window portion is provided on the front surface of the photoelectric unit 22, light from the blue LED 22a is irradiated on the skin surface through the window portion, and reflected light is incident on the PD 22b through the window portion. The blue LED 22a and the PD 22b are connected to a pulse wave meter 34 that is an analog front end for pulse waves. The sphygmograph 34 detects a change in reflected light that changes due to a change in blood flow in the capillary blood vessel, and obtains a pulse wave by analyzing the detection signal.

  An electrocardiograph 30, an acceleration sensor 32, a pulse wave meter 34, and a temperature sensor 24 are connected to the system controller 38. The temperature sensor 24 measures the temperature of the human body surface, and the acceleration sensor 32 measures the body movement of the human body. Here, it is assumed that the acceleration sensor 32 is a three-axis acceleration sensor with a sampling frequency of 1 KHz.

  The CPU 44 is a processor that controls the operation of each module and each component of the biosensor device 10. As described above, the biological sensor device 10 analyzes various biological information (eg, body temperature, skin temperature, pulse rate, heart rate, autonomic nerve activity) by analyzing the output of each sensor or the combination of outputs of a plurality of sensors. Index, blood pressure, sleep time, etc.) can be measured continuously.

  The system controller 38 is a bridge device that connects the CPU 44 to each module and each component. Also connected to the system controller 38 are a Bluetooth module 36, an embedded controller (EC) 40, a CPU 44, a main memory 46, a BIOS-ROM 48, and a flash memory 50.

  The embedded controller 40 is a power management controller for executing power management of the biosensor device 10 and controls charging of a built-in secondary battery, for example, a lithium secondary battery 42. When the biosensor device 10 is attached to the charger 52, the charging terminal 26 comes into contact with the terminal of the charger 52, and the charging current from the charger 52 is supplied to the biosensor device 10 via the charging terminal 26. The secondary battery 42 is charged. The embedded controller 40 supplies operating power to each module and each component based on the power from the lithium secondary battery 42.

  The CPU 44 is a processor that controls the operation of each module and each component in the biosensor device 10. The CPU 44 executes various software loaded from the flash memory 50 to the main memory 46. These software include an operating system (OS) and various application programs. The application program includes a sleeping habit analysis application program 100. This sleeping heel analysis application program 100 analyzes the heel based on the acceleration signal acquired by the acceleration sensor 32, and the risk that the user wearing the biosensor device 10 will develop sleep apnea syndrome (risk) ) Is high.

  FIG. 4 is a block diagram illustrating an example of a functional configuration of the sleep state analyzing application program 100. As shown in FIG. 4, the sleeping habit analysis application program 100 includes a feature amount analysis section determination unit 101, a feature point calculation unit 102, a habit tendency determination unit 103, an onset risk calculation unit 104, and the like. In this embodiment, the sleep state analyzing application program 100 is stored in the biosensor device 10 and executed by the CPU 44 in the biosensor device 10, but for example, the sleep state analyzing application program 100 may be stored in the electronic device 11 and executed by a CPU in the electronic device 11. In this case, the acceleration signal acquired by the acceleration sensor 32 is output to the electronic device 11. In addition, a part of each functional unit included in the sleep state analyzing application program 100 may be mounted on the electronic device 11. For example, the functions of the feature amount analysis section determination unit 101 and the feature point calculation unit 102 are implemented in the biosensor device 10, and the functions of the wrinkle tendency determination unit 103 and the onset risk calculation unit 104 are implemented in the electronic device 11. Good.

Hereinafter, the details of the functions of the units 101 to 104 will be described.
When receiving the input of the acceleration signal output from the acceleration sensor 32, the feature amount analysis section determination unit 101 executes a feature amount analysis section determination process using the acceleration signal. The feature amount analysis section determination process is a process executed to determine a section suitable for executing a feature point calculation process (to be described later) (hereinafter referred to as a feature amount analysis section). The feature amount analysis section is a vibration caused by posture changes such as turning over, vibration caused by body movements during awakening, vibration caused by coughing, sleeping, etc. This section shows a lot of vibrations caused by.

Here, an example of the procedure of the feature amount analysis section determination process will be described with reference to the flowchart of FIG.
First, the feature amount analysis section determination unit 101 receives an input of an acceleration signal output from the acceleration sensor 32 (step S1). In the processing of steps S2 to S4 described later, the feature amount analysis section determination unit 101 is the surface of the chest of the user wearing the biosensor device 10 among the DC (direct current) components of the acceleration indicated by the acceleration signal that has received the input. DCz that is a DC component in the vertical direction (z-axis direction) is used. Here, the case where the waveform W _DC shown in FIG. 6A is obtained from the acceleration signal as the DCz waveform (hereinafter referred to as a DCz waveform) of the DC component of the acceleration indicated by the acceleration signal. Suppose.

Subsequently, the feature amount analysis section determination unit 101 mounts a low-pass filter for each predetermined section I of the DCz waveform W _DC indicated by the acceleration signal that has received the input (step S2). In addition, the feature amount analysis section determination unit 101 performs a differentiation process on the DCz waveform W _DC after mounting the above-described low-pass filter (step S3). According to the processes in steps S2 and S3, the feature amount analysis section determination unit 101 can extract vibrations caused by posture fluctuations (vibrations caused by other than wrinkles). Note that the vibration due to the posture variation may be extracted by performing a band pass filter process or a high pass filter process.

Here, it is assumed that the DCz waveform W _{DC ′} shown in FIG. 6B is obtained by repeatedly executing the processes of steps S2 and S3 described above for each predetermined section I. Here, it is assumed that the predetermined section I is 30 [sec] (1 epoch), but the predetermined section I may be 1 [min] (2 epochs), for example.

  In the present embodiment, DCz of the DC component of the acceleration indicated by the acceleration signal is used for the feature amount analysis section determination process. For example, of the DC component of the acceleration indicated by the acceleration signal, DCx that is a DC component in a direction (x-axis direction) that is horizontal with respect to the chest surface of the user wearing the biosensor device 10 and that is perpendicular to the body axis (DCx that is a DC component of acceleration) , DCy, DCz square average), etc. may be used.

Next, the feature amount analysis section determination unit has all of the absolute values of a plurality of values constituting the waveform W _{DC ′} of the predetermined section I obtained as a result of the differentiation process being equal to or less than a first threshold value set in advance. Is determined (step S4). If it is determined as NO as a result of the determination process in step S4 (NO in step S4), the process returns to the process in step S1 described above, and the various processes described above are executed for the next predetermined section I.

On the other hand, when it is determined as a result of the determination process in step S4 that all of the absolute values are equal to or less than the first threshold value set in advance (YES in step S4), the feature amount analysis interval determination unit 101 determines that the predetermined interval I is determined to be a section where there is little vibration due to the posture variation of the user wearing the biosensor device 10. In the case of FIG. 6B, the period T1 including the above-described predetermined section I and the period T2 correspond to a section where there is little vibration due to posture variation. Here, in the process of step S4 described above, the feature amount analysis section determination unit 101 is set in advance for all of the absolute values of a plurality of values constituting the waveform W _{DC ′} of the predetermined section I. Although it is determined whether or not it is equal to or less than the first threshold value, for example, the above-described determination is performed for an average value, a median value, a variance value, or the like of a plurality of values constituting the waveform W _{DC ′} of the predetermined section I. Processing may be executed.

In the processing of steps S5 to S8, which is processing after it is determined that the vibration due to the posture variation is small, the feature amount analysis section determination unit 101 determines the AC of the acceleration indicated by the acceleration signal that has received the input. Of the (alternating current) components, ACz that is an AC component in the z-axis direction of the user wearing the biosensor device 10 is used to determine the feature amount analysis section. Here, ACZ waveform of the AC component of the acceleration represented by the acceleration signal (hereinafter, referred to as ACZ waveform) as the case where the waveform W _AC shown in FIG. 7 (a), is obtained from the acceleration signal Suppose.

Analysis of features section determining unit 101 calculates the envelope of ACz waveform W _AC of the predetermined section I shown by the acceleration signal, the input of which is accepted, to implement a low-pass filter to the ACz waveform W _AC (step S5 ). According to the processing in step S5, the feature amount analysis section determination unit 101 can generate a waveform in which vibration caused by wrinkles is emphasized. Incidentally, enhancement of vibration due to snoring performs time-frequency analysis of ACz waveform W _AC predetermined interval I, the maximum value or the frequency band of the power spectral density of at least audible range in the predetermined section I (or 20 Hz) This may be done by calculating the sum and implementing a low pass filter on the result. Also, it may only implement a low-pass filter to ACz waveform W _AC predetermined interval I.

  Next, the feature quantity analysis section determination unit 101 calculates the autocorrelation function of the waveform W1 shown in FIG. 7B obtained as a result of the process of step S5 (step S6). According to the process of step S6, the feature amount analysis section determination unit 101 can evaluate the periodicity of the waveform obtained as a result of the process of step S5. Here, it is assumed that the waveform W2 shown in FIG. 7C is obtained as a result of the process of step S6.

  Subsequently, the feature amount analysis interval determination unit 101 is the second maximum value that is the second maximum value of the waveform W2 of the autocorrelation function obtained as a result of the process of step S6 and the peak value that is the maximum value is set in advance. It is determined whether or not it is equal to or greater than the threshold value (step S7). In addition, when it determines with it being no as a result of the determination process of step S7 (NO of step S7), it returns to the process of above-described step S1.

  On the other hand, when it is determined as a result of the determination process in step S7 that the peak value is greater than or equal to the second threshold value set in advance (YES in step S7), the feature amount analysis interval determination unit 101 determines that the predetermined interval I is , It is determined that there is little vibration due to wrinkles and vibration due to wrinkles is present, the predetermined section I is determined as a feature amount analysis section (step S8), and the processing here is terminated. .

The feature amount analyzing section determining unit 101 determines a feature quantity analysis section transmits waveform information indicating the ACz waveform W _AC of the determined feature amount analysis section to the feature point calculating unit 102.

  Here, the feature amount analysis section determination unit 101 determines the feature amount analysis section by the above-described method. However, for example, the feature amount analysis section determination unit 101 obtains as a result of the process of step S5 described above. Alternatively, a period in which a plurality of values constituting the waveform W1 shown in FIG. 7B is equal to or greater than a preset third threshold may be extracted, and the extracted period may be determined as a feature amount analysis section.

  Here, with reference to FIG.8 and FIG.9, the difference in acupuncture with a healthy subject and the test subject (patient) of sleep apnea syndrome is demonstrated. FIG. 8 shows various waveforms obtained by executing predetermined processing on the ACz waveform of the healthy person and the ACz waveform in the feature amount analysis section. FIG. 9 shows ACz waveforms of subjects with sleep apnea syndrome in a feature amount analysis section and various waveforms obtained by executing predetermined processing on the ACz waveforms.

  FIG. 8A shows an ACz waveform of a healthy person, and FIG. 9A shows an ACz waveform of a subject with sleep apnea syndrome. According to FIG. 8A and FIG. 9A, it can be seen that vibrations caused by wrinkles are periodically superimposed.

  FIG. 8B is obtained as a result of performing short-time Fourier transform processing on the ACz waveform of the healthy person shown in FIG. Similarly, FIG. 9B is obtained as a result of performing short-time Fourier transform processing and performing time-frequency analysis on the ACz waveform of the subject with sleep apnea syndrome shown in FIG. 9A. It is a thing. The color intensity indicates the magnitude of the power spectrum density value of each frequency (vertical axis) at each time (horizontal axis). According to FIG.8 (b) and FIG.9 (b), the test subject of sleep apnea syndrome has a large power spectrum density in a high frequency band compared with a healthy person, and contains many high frequency in a bag ( I can see that it emits high frequency as a kite).

  FIG. 8C shows a waveform obtained by calculating the sum of power spectral densities of frequencies equal to or higher than a preset threshold TH, using the result of the time frequency analysis shown in FIG. FIG. 8D shows a waveform obtained by calculating the sum of power spectral densities of frequencies less than a preset threshold TH using the result of the time frequency analysis shown in FIG. Show. FIG. 8E shows a waveform obtained by calculating the sum of power spectral densities of all frequencies using the result of the time frequency analysis shown in FIG. 8B.

  Similarly, FIG. 9C shows a waveform obtained by calculating the sum of power spectral densities of frequencies equal to or higher than a preset threshold TH using the result of the time frequency analysis shown in FIG. 9B. Indicates. FIG. 9D shows a waveform obtained by calculating the sum of power spectral densities of frequencies less than a preset threshold TH using the result of the time frequency analysis shown in FIG. 9B. Show. Further, FIG. 9 (e) shows a waveform obtained by calculating the sum of power spectral densities of all frequencies using the result of the time frequency analysis shown in FIG. 9 (b).

  The above-mentioned threshold value TH is generally set within a range of 200 Hz to less than 300 Hz because the fundamental frequency of voice is generally 100 Hz to 200 Hz and the lowest resonance frequency is 300 Hz to 1000 Hz. .

  When comparing each waveform shown in FIGS. 8 (c) to (e) related to healthy subjects with each waveform shown in FIGS. 9 (c) to (e) related to subjects with sleep apnea syndrome, It can be seen that the difference is significant. This is because when the sleep apnea syndrome develops, the upper respiratory tract is obstructed by the position and shape of the tonsils and tongue base, so waves with fundamental frequencies emitted from the vocal cords and soft tissues are complicated. It is thought that the high frequency is generated due to reflection and resonance.

  As described above, it can be seen that there is a great difference in vibration caused by hemorrhoids between healthy subjects and subjects with sleep apnea syndrome. Therefore, the sleep heel analysis application program 100 according to the present embodiment calculates a feature point of vibration related to the heel in the feature point calculation unit 102, and determines the tendency of the heel using this calculation result. The user wearing the biosensor device 10 has a function capable of determining whether or not the risk (risk) of developing sleep apnea syndrome is high.

  As described above, the feature point calculation unit 102 executes a first feature point calculation process, which will be described later, and a second feature point calculation process, which will be described later, as a process for calculating the feature point of vibration caused by wrinkles. .

Here, an example of the procedure of the first feature point calculation process executed by the feature point calculation unit 102 will be described with reference to the flowchart of FIG.
First, the feature point calculation unit 102 receives input of waveform information indicating the ACz waveform of the feature amount analysis section determined by the feature amount analysis section determination unit 101 (step S11). Here, it is assumed that the feature point calculation unit 102 receives input of waveform information indicating the ACz waveform shown in FIG.

  Subsequently, the feature point calculation unit 102 performs a time-frequency analysis by executing a short-time Fourier transform process on the ACz waveform in the feature amount analysis section indicated by the waveform information that has received the input (step S12). According to the processing in step S12, the feature point calculation unit 102 can obtain the time frequency analysis result shown in FIG. 8B as a result of performing the time frequency analysis. Note that here, the feature point calculation unit 102 performs short-time Fourier transform processing and performs time-frequency analysis, and calculates the power spectrum density as the intensity. Frequency analysis may be performed to calculate the power spectral density as the feature amount. Further, the feature point calculation unit 102 may extract a high-frequency band wave by mounting a high-pass filter with the threshold value TH as a cutoff frequency without performing a time-frequency analysis.

  Next, the feature point calculation unit 102 calculates the sum of power spectral densities of frequencies equal to or higher than a preset threshold TH for each time based on the result of the time frequency analysis, and calculates the frequency band of the frequency band equal to or higher than the threshold TH. Time-series data indicating the wave intensity is calculated (step S13). For example, at the time t1 shown in FIG. 8B, the feature point calculation unit 102 adds the power spectral density values expressed in magnitude by the color density over the frequency equal to or higher than the threshold value TH. The sum of power spectral densities at t1 is calculated. By executing the same processing for all times, the feature point calculation unit 102 can obtain time-series data indicating the waveform shown in FIG.

  Thereafter, the feature point calculation unit 102 determines the first value of the vibration caused by wrinkles as the maximum value of the waveform indicated by the calculated time-series data and the maximum value at a predetermined interval before and after the maximum value. Are extracted as feature points (step S14), and the process is terminated. In this case, the point marked with x shown in FIG. 8C is extracted as the first feature point of vibration caused by wrinkles. The first feature point information indicating the coordinates of the extracted first feature points is sent to the wrinkle tendency determining unit 103.

Next, the second feature point calculation process executed by the feature point calculation unit 102 will be described with reference to the flowchart of FIG. In addition, since the process of step S21 and S22 is a process similar to the process of already demonstrated step S11 and S12, the detailed description is abbreviate | omitted here. Hereinafter, various processes will be described in order from the process of step S23.
After the process of step S22, the feature point calculation unit 102 calculates the sum of the power spectrum density of all frequencies for each time based on the result of the time frequency analysis, and shows the time series indicating the intensity of the waves in all frequency bands. Data is calculated (step S23). For example, at time t1 shown in FIG. 8 (b), the feature point calculation unit 102 adds the power spectral density values expressed in magnitude by the color density over all frequencies, so that the feature point calculation unit 102 has the power at time t1. Calculate the sum of spectral densities. By executing the same processing for all times, the feature point calculation unit 102 can obtain time-series data indicating the waveform shown in FIG.

  Subsequently, the feature point calculation unit 102 determines a point that is the maximum value of the waveform indicated by the calculated time-series data and that is the maximum value at a predetermined interval before and after the maximum value as the second value of vibration caused by wrinkles. Are extracted as feature points (step S24). In this case, the point marked with x shown in FIG. 8E is extracted as the second feature point of vibration caused by wrinkles. The second feature point information indicating the coordinates of the extracted second feature points is sent to the wrinkle tendency determining unit 103.

  Thereafter, the feature point calculation unit 102 calculates a time difference (interval) between the extracted second feature points as a period of vibration caused by wrinkles (step S25), and ends the processing here. In this case, C1, C2,... Shown in FIG. 8 (e) are calculated as the periods of vibration caused by wrinkles. Note that the cycle information indicating the calculated cycle is sent to the wrinkle tendency determination unit 103.

  Note that the feature point calculation unit 102 calculates the sum of power spectral densities of frequencies less than a preset threshold TH for each time based on the result of the time frequency analysis, similarly to the process of step S13 described above. By calculating the time series data indicating the intensity of the wave in the frequency band below the threshold TH, the time series data indicating the waveform shown in FIG. 8D can be obtained. Although not described in detail below, the wrinkle tendency determination unit 103, which will be described later, performs time-series data in FIG. 8D with respect to wave intensities in the frequency band equal to or higher than the threshold TH shown in the time-series data in FIG. It is also possible to calculate the ratio of the wave intensity in the frequency band less than the threshold value TH shown in FIG. 5 to determine the tendency of wrinkles of the user wearing the biosensor device 10.

  Returning again to the description of FIG. When the wrinkle tendency determination unit 103 receives the input of the first feature point information, the second feature point information, and the period information output from the feature point calculation unit 102, the biological tendency determination unit 103 The tendency of the wrinkle of the user wearing the sensor device 10 is determined.

Here, with reference to the flowchart of FIG. 12, the first wrinkle tendency determination process for determining the tendency of the wrinkle of the user wearing the biosensor device 10 based on the first feature point information described above will be described.
First, when the wrinkle tendency determining unit 103 receives the input of the first feature point information calculated by the process of step S14 by the feature point calculating unit 102 and output from the feature point calculating unit 102, the first feature Based on the coordinates of the plurality of first feature points indicated by the point information, the sum of the power spectrum densities of all feature points in the feature amount analysis section (in other words, the first marked with a cross as shown in FIG. 8C). The average value (hereinafter, referred to as the first average value) of the feature points is calculated (step S31).

  Subsequently, the wrinkle tendency determining unit 103 determines whether or not the calculated first average value is greater than or equal to a preset first classification threshold (step S32).

  As a result of the determination process in step S32, when it is determined that the calculated first average value is equal to or greater than a preset first classification threshold value (YES in step S32), the wrinkle tendency determination unit 103 determines that the biosensor The tendency of the user who wears the device 10 to wrinkle is determined as “having a high-pitched tendency” (step S33), and the process proceeds to step S35 described later.

  On the other hand, when it is determined as NO as a result of the determination process in step S32 (NO in step S32), the wrinkle tendency determining unit 103 sets the tendency of the wrinkle of the user wearing the biosensor device 10 to “no treble tendency”. (Step S34).

  Next, the wrinkle tendency determination unit 103 determines the first based on each value of the individual power spectrum density sums of all feature points and the first average value that is the average value of the power spectrum density sums of all feature points. Is calculated (step S35). Here, although the case where the first average value is used to calculate the first variance value has been described, for example, the median value of the first feature points may be used. Alternatively, an average value, a median value, or the like in the feature amount analysis section of the waveform indicated by the time series data illustrated in FIG. 8C may be used.

  Thereafter, the wrinkle tendency determining unit 103 determines whether or not the calculated first variance value is greater than or equal to a preset second classification threshold value (step S36).

  As a result of the determination process in step S36, if it is determined that the first variance value is greater than or equal to the preset second classification threshold value (YES in step S36), the wrinkle tendency determination unit 103 causes the biosensor device 10 to The tendency of the user who wears the eyelid is determined to be “with irregular tendency” (step S37), and the process here is terminated.

  On the other hand, when it is determined as NO as a result of the determination process in step S36 (NO in step S36), the wrinkle tendency determining unit 103 determines the wrinkle tendency of the user wearing the biosensor device 10 as “no irregular tendency”. "(Step S38), and the process here is terminated.

  The wrinkle tendency determining unit 103 also has a function of sending wrinkle tendency information indicating the determined wrinkle tendency to the onset risk calculating unit 104 as a result of the first wrinkle tendency determining process. For example, when the result of the determination process in step S32 is “having a high-pitched tendency” and the result of the determination process in step S36 is “having an irregular tendency”, information indicating this is the onset risk calculation unit as wrinkle tendency information 104. Here, the case where the result of the determination process in step S32 is “having a high-pitched tendency” and the result of the determination process in step S36 is “there is an irregular tendency” has been described as an example. However, since the wrinkle tendency information is sent to the onset risk calculation unit 104, detailed description thereof is omitted here.

  In addition, here, the wrinkle tendency determining unit 103 simply determines the presence / absence of “high pitch tendency” as the tendency of wrinkles of the user wearing the biosensor device 10 in the processing of step S33 and step S34. By preparing a plurality of threshold values corresponding to one classification threshold, for example, “high pitch tendency (light)”, “high pitch tendency (medium)”, “high pitch tendency (heavy)”, etc. It is also possible to determine the tendency of wrinkles including the above. Similarly, the wrinkle tendency determination unit 103 simply determines the presence or absence of “irregular tendency” as the tendency of wrinkles of the user wearing the biosensor device 10 in the processing of step S37 and step S38. By preparing multiple thresholds corresponding to the classification thresholds, irregularities such as “with irregular tendency (light)”, “with irregular tendency (medium)”, “with irregular tendency (heavy)” etc. The tendency of epilepsy may be determined including the degree of the tendency (severity).

  Referring to the flowchart of FIG. 13, the biometric sensor that is calculated by the process of step S <b> 24 by the above-described feature point calculation unit 102 and executed when the input of the second feature point information is received from the feature point calculation unit 102. The second wrinkle tendency determination process for determining the tendency of the wrinkle of the user wearing the apparatus 10 will be described.

  When the wrinkle tendency determining unit 103 receives the input of the second feature point information output from the feature point calculating unit 102, the wrinkle tendency determining unit 103 is based on the coordinates of the plurality of second feature points indicated by the second feature point information. , The sum of power spectral densities of all feature points (in other words, the value of the vertical axis of the second feature point marked with x shown in FIG. 8E) (hereinafter referred to as the second average value) ) Is calculated (step S41).

  Subsequently, the wrinkle tendency determining unit 103 determines whether or not the calculated second average value is greater than or equal to a preset third classification threshold (step S42).

  As a result of the determination process in step S42, if it is determined that the calculated second average value is greater than or equal to a preset third classification threshold (YES in step S42), the wrinkle tendency determining unit 103 may The tendency of the wrinkle of the user wearing the apparatus 10 is determined as “prone to loud sound” (step S43), and the process proceeds to step S45 described later.

  On the other hand, when it is determined as NO as a result of the determination process in step S42 (NO in step S42), the wrinkle tendency determination unit 103 determines the tendency of the wrinkle of the user wearing the biosensor device 10 as “no loud sound tendency”. (Step S44).

  Next, the wrinkle tendency determination unit 103 performs the second calculation based on each value of the sum of power spectrum densities of all feature points and a second average value that is an average value of the sum of power spectrum densities of all feature points. Is calculated (step S45). Although the case where the second average value is used to calculate the second variance value has been described here, for example, the median value of the second feature points may be used. Alternatively, an average value, a median value, or the like in the feature amount analysis section of the waveform indicated by the time series data shown in FIG.

  Thereafter, the wrinkle tendency determining unit 103 determines whether or not the calculated second variance value is equal to or greater than a preset fourth classification threshold (step S46).

  As a result of the determination process in step S46, if it is determined that the second variance value is greater than or equal to a preset fourth classification threshold value (YES in step S46), the wrinkle tendency determination unit 103 causes the biosensor device 10 to The tendency of the wrinkle of the user who wears it is determined as “with irregular tendency” (step S47), and the process here is terminated.

  On the other hand, when it is determined as NO as a result of the determination process in step S46 (NO in step S46), the wrinkle tendency determining unit 103 determines the wrinkle tendency of the user wearing the biosensor device 10 as “no irregular tendency”. "(Step S48), and the process here is terminated.

  The wrinkle tendency determination unit 103 also has a function of sending wrinkle tendency information indicating the determined wrinkle tendency to the onset risk calculation unit 104 as a result of the second wrinkle tendency determination process. For example, if the result of the determination process in step S42 is “There is a loud tendency” and the result of the determination process in step S46 is “There is an irregular tendency”, information indicating this is calculated as a tendency tendency information to calculate the onset risk. Sent to the unit 104. Here, the case where the result of the determination process in step S42 is “prone to loud sound” and the result of the determination process in step S46 is “has an irregular tendency” has been described as an example. Even if it exists, since the wrinkle tendency information is sent to the onset risk calculation unit 104, the detailed description thereof is omitted here.

  In addition, here, the wrinkle tendency determining unit 103 simply determines the presence or absence of “loud tendency” as the tendency of wrinkles of the user wearing the biosensor device 10 in the processing of step S43 and step S44. By preparing a plurality of thresholds corresponding to the third classification threshold, for example, “with loud sound tendency (light)”, “with loud sound tendency (medium)”, “with loud sound tendency (heavy)”, etc. It may be possible to determine the tendency of the habit including the degree of loud sound tendency. Similarly, the wrinkle tendency determination unit 103 simply determines whether or not there is an “irregular tendency” as the tendency of wrinkles of the user wearing the biosensor device 10 in the processing of step S47 and step S48. By preparing multiple thresholds corresponding to the classification thresholds, irregularities such as “with irregular tendency (light)”, “with irregular tendency (medium)”, “with irregular tendency (heavy)” etc. It is also possible to determine the tendency of wrinkles, including the degree of tendency.

Referring to the flowchart of FIG. 14, the biosensor device is executed when the input of the period information output from the feature point calculation unit 102 is received by the process of step S25 by the feature point calculation unit 102 described above. A third wrinkle tendency determination process for determining the wrinkle tendency of the user wearing 10 will be described.
First, when the wrinkle tendency determining unit 103 receives the input of the periodic information output from the feature point calculating unit 102, the wrinkle tendency determining unit 103 starts from the predetermined wrinkle to the next based on the period of vibration caused by the wrinkle indicated by the periodic information. An average value up to 鼾 (hereinafter referred to as a third average value) is calculated (step S51). Specifically, wrinkle tendency determining unit 103 calculates the average value of cycles C1, C2,... Shown in FIG.

  Subsequently, the wrinkle tendency determining unit 103 determines the first based on each cycle (interval) from the predetermined wrinkle to the next wrinkle indicated by the period information that has received the input, and the calculated third average value. 3 is calculated (step S52).

  Next, the wrinkle tendency determining unit 103 determines whether or not the third variance value is greater than or equal to a preset fifth classification threshold (step S53). If it is determined as NO as a result of the determination process in step S53 (NO in step S53), the wrinkle tendency determining unit 103 determines the wrinkle tendency of the user wearing the biosensor device 10 as “no irregular tendency”. "(Step S54), and the process here is terminated.

  On the other hand, when it is determined as a result of the determination process in step S53 that the third variance value is greater than or equal to a preset fifth classification threshold value (YES in step S53), the wrinkle tendency determining unit 103 determines that the biological sensor device The wrinkle tendency of the user wearing 10 is determined to be “with irregular tendency” (step S55), and the process here is terminated.

  The wrinkle tendency determination unit 103 also has a function of sending wrinkle tendency information indicating the determined wrinkle tendency to the onset risk calculation unit 104 as a result of the third wrinkle tendency determination process. For example, when the result of the determination process in step S53 is “There is an irregular tendency”, information indicating this is transmitted to the onset risk calculation unit 104 as wrinkle tendency information. If the result of the determination process in step S53 is “no irregular tendency”, information indicating this is sent to the onset risk calculation unit 104 as wrinkle tendency information.

  In addition, the wrinkle tendency determination unit 103 simply determines whether or not “irregular tendency” exists as the tendency of wrinkles of the user wearing the biosensor device 10 in the processes of step S54 and step S55. By preparing multiple thresholds that correspond to the classification thresholds, for example, irregular trends such as “with irregular tendency (light)”, “with irregular tendency (medium)”, “with irregular tendency (heavy)”, etc. It is also possible to determine the tendency of wrinkles, including to the extent of.

  Note that the result of the determination process in step S32 is “no treble tendency”, the result of the determination process in step S42 is “no loud tendency”, and the result of the determination process in steps S36, S46 and S53 is “no irregular tendency”. If the result is obtained, the wrinkle tendency information collectively indicating “healthy” may be sent to the onset risk calculation unit 104.

  In the embodiment, each classification threshold is set, and the variance value is calculated after the threshold is set. However, each may be calculated simultaneously.

  Returning again to the description of FIG. When the onset risk calculation unit 104 receives input of a plurality of wrinkle tendency information output from the wrinkle tendency determination unit 103, the onset risk calculation unit 104 calculates the risk of developing sleep apnea syndrome based on the received wrinkle tendency information. It has a function to do. Specifically, the onset risk calculation unit 104 refers to an onset risk calculation table shown in FIGS. 15 and 16 to be described later, and calculates a total value of points corresponding to the tendency of wrinkles indicated by each wrinkle tendency information. It is determined that the higher the calculated total value, the higher the risk of developing sleep apnea syndrome.

Here, the onset risk calculation table used when calculating the risk of developing sleep apnea syndrome will be described with reference to FIGS. 15 and 16 respectively.
FIG. 15 shows an onset risk calculation table used when the tendency of wrinkles is determined by the wrinkle tendency determination unit 103 without considering the degree (severity) of various wrinkles. According to this onset risk calculation table, if the wrinkle tendency indicated by the wrinkle tendency information is one of “high-pitch tendency”, “high-pitch tendency” and “irregular tendency”, “1 point” is added. However, it is defined that “0 points” is added (that is, points are not added) if the tendency of wrinkles indicated by the wrinkle tendency information is “no tendency”. In addition, for “irregular tendency”, points are added according to which of the “irregular tendency” determined in any of the first to third habit tendency determination processes by the habit tendency determining unit 103. It is stipulated that

  FIG. 16 shows an onset risk calculation table used when the tendency of wrinkles is determined by the wrinkle tendency determining unit 103 in consideration of the degree of various wrinkles. According to this onset risk calculation table, the tendency of wrinkles indicated by wrinkle tendency information is either “high-pitch tendency (light)”, “high-pitch tendency (light)”, or “regular tendency (light)” If so, it is specified that “1 point” is added. Also, if the habit tendency indicated by the habit tendency information is one of “high pitch tendency (medium)”, “loud tendency (medium)” and “irregular tendency (medium)”, “2 points” Is added. Furthermore, “3 points” if the tendency of wrinkle indicated by the wrinkle tendency information is any of “high-pitch tendency (heavy)”, “high-pitch tendency (heavy)”, and “irregular tendency (heavy)”. Is added. Further, it is defined that “0 points” is added (that is, no points are added) if the tendency of wrinkles indicated by the wrinkle tendency information is “no tendency”. In addition, for “irregular tendency”, points are added according to which of the “irregular tendency” determined in any of the first to third habit tendency determination processes by the habit tendency determining unit 103. It is stipulated that

  Here, the first to third wrinkle tendency determination processes are executed without considering the degree of wrinkle by the wrinkle tendency determination unit 103, and the wrinkle tendency determined by the first wrinkle tendency determination process is “high tone”. “There is a tendency and there is an irregular tendency”, and the tendency of the habit determined by the second habit tendency determination process is “with a loud sound tendency and no irregular tendency” and is determined by the third habit tendency determination process. A specific process executed by the onset risk calculation unit 104 will be described on the assumption that the tendency of wrinkles is “no irregular tendency”.

  In this case, when the onset risk calculation unit 104 receives input of wrinkle tendency information indicating the tendency of wrinkles determined by the first to third wrinkle tendency determination processes, the onset risk calculation unit 104 refers to the onset risk calculation table illustrated in FIG. , “1 point” due to “having a high-pitched sound tendency”, “1 point” due to “having a loud sound tendency”, and the habit tendency determined by the first habit tendency determining process “1 point” due to the fact that “has an irregular tendency” and “0 point” due to the fact that the habit tendency determined by the second and third habit tendency determination processes is “no irregular tendency” Add together. That is, the onset risk calculation unit 104 calculates the total value “3 points” as an index of the risk of developing sleep apnea syndrome. For example, when the total value is “3 points” or more and the risk of developing sleep apnea syndrome is preset to be high, the onset risk calculation unit 104 has a total value of “3 points”. Therefore, it is determined that the risk of developing sleep apnea syndrome is high. In addition, when the total value is “4 points” or more and the risk of developing sleep apnea syndrome is set to be high in advance, the onset risk calculation unit 104 is “3 points”. The risk of developing sleep apnea syndrome is determined to be low.

  When the onset risk calculation unit 104 determines that the risk of developing sleep apnea syndrome is high, the onset risk calculation unit 104 has a function of notifying the electronic device 11 of this via the Bluetooth module 36 in the biosensor device 10 or the like. You may have. According to this function, the user wearing the biosensor device 10 can know whether the risk of developing sleep apnea syndrome is high through the electronic device 11. Moreover, the onset risk calculation unit 104 uses the high risk of developing sleep apnea syndrome as analysis result information regardless of the risk of developing sleep apnea syndrome, and the Bluetooth module in the biosensor device 10. You may send to the electronic device 11 via 36 grade | etc.,.

  According to the first embodiment described above, the vibration caused by the heel is extracted from the various vibrations indicated by the acceleration signal acquired by the acceleration sensor 32, and based on the intensity of the vibration caused by the heel, Since it is possible to determine whether or not the risk of developing an apnea group is high, whether or not the user wearing the biosensor device 10 is at high risk of developing sleep apnea syndrome As a result, it is possible to realize the manifestation of a patient with sleep apnea syndrome.

  In addition, according to the present embodiment, not only the intensity of vibration caused by wrinkles but also the period of vibration caused by wrinkles and the like are provided, so that the tendency of wrinkles can be determined in a multifaceted manner. Thus, the risk of developing sleep apnea syndrome can be calculated.

<Second Embodiment>
Next, a second embodiment will be described. In the present embodiment, a case where the sleep state analysis application program 100 further includes a sleeping posture estimation unit 105 and an association unit 106 as shown in FIG. 17 in addition to the above-described units 101 to 104 will be described. To do. In the present embodiment, the same reference numerals are given to the components having the same functions as those in the first embodiment, and detailed description thereof will be omitted.

  When the sleeping posture estimation unit 105 receives input of the waveform information indicating the DCx waveform of the feature amount analysis section determined by the feature amount analysis section determination unit 101 and the waveform information indicating the DCz waveform of the feature amount analysis section, Using these waveform information, the user has a function of estimating the sleeping posture of the user wearing the biosensor device 10.

Here, an example of the procedure of the sleeping posture estimation process executed by the sleeping posture estimation unit 105 will be described with reference to the flowchart of FIG.
First, the sleeping posture estimation unit 105 receives input of waveform information indicating the DCx waveform of the feature amount analysis section determined by the feature amount analysis section determination unit 101 and waveform information indicating the DCz waveform of the feature amount analysis section. (Step S61). In addition to the above-described DCx waveform and DCz waveform, the sleeping posture estimation unit 105 is horizontal with respect to the chest surface of the user wearing the biosensor device 10 and parallel to the body axis (y direction). The input of waveform information indicating the waveform of DCy, which is the DC component, may be further received.

  Here, it is assumed that the feature amount analysis section is 30 [sec] (1 epoch), as in the first embodiment.

  Subsequently, the sleeping posture estimation unit 105 represents, for each predetermined section in the feature amount analysis section, a DCx waveform indicated by the waveform information that has received the input and a DCz waveform that is also indicated by the waveform information that has received the input. Each value is acquired (step S62). The predetermined section described above may be the same as the sampling period, for example.

  Next, the sleeping posture estimation unit 105 calculates the median value or average value of the acquired representative values as the feature amount of the feature amount analysis section (step S63).

  Thereafter, the sleeping posture estimation unit 105 sets the feature amount calculated from the DCx waveform to DCx_F, sets the feature amount calculated from the DCz waveform to DCz_F, and then sets the sleeping posture based on these feature amounts DCx_F and DCz_F. The process of determining whether the position corresponds to the position, the left-side prone position, the right-side prone position, or the prone position is performed, the process for estimating the sleeping posture is performed (step S64), and the process is terminated Let Note that sleeping posture information indicating the sleeping posture obtained as a result of the sleeping posture estimation process is sent to the associating unit 106.

  Note that when determining whether the sleeping posture corresponds to the supine position, the left-side prone position, the right-side prone position, or the prone position based on the feature values DCx_F and DCz_F, for example, an error that is one of neural networks. A back propagation method or the like may be used. It is assumed that the weights used when the error back propagation method is used are previously learned and recorded in the flash memory 50 or the like. Further, when determining whether the sleeping posture corresponds to any of the above-mentioned positions, for example, another neural network such as a support vector machine may be used, or used in the above-described wrinkle tendency determination process. Like the classification threshold values, threshold values recorded in advance corresponding to the feature amounts DCx_F and DCz_F may be used.

  In addition, the sleeping posture determined in step S64 is calculated as a discrete value here, but for example, an arctangent value of the ratio of the feature amounts DCx_F and DCz_F is calculated, and the biosensor device 10 (on the device surface) is calculated. The slope may be calculated as a continuous value.

  Returning again to the description of FIG. When the associating unit 106 receives an input of the sleeping posture information sent from the sleeping posture estimating unit 105, the associating unit 106 extracts the sleeping posture indicated by the sleeping posture information received from the input by the feature point calculating unit 102. Or the second feature point, the period from a predetermined wrinkle to the next wrinkle, the wrinkle tendency determined by the wrinkle tendency determining unit 103, or the onset risk calculated by the onset risk calculating unit 104 A process of associating with one is executed.

  In the associating unit 106, the sleeping posture based on the sleeping posture information, the first feature point and the second feature point extracted by the feature point calculating unit 102, the cycle from a predetermined heel to the next heel, and the heel tendency The association result information obtained by associating at least one of the tendency of wrinkles determined by the determination unit 103 and the onset risk calculated by the onset risk calculation unit 104 is the Bluetooth module 36 in the biosensor device 10 or the like. To be sent to the electronic device 11 as needed.

Here, an example of a screen displayed when the above-described association result information is displayed on a display unit (not shown) of the electronic device 11 will be described with reference to FIG.
In FIG. 19, the display unit of the electronic device 11 has various items indicating time, section, height of the heel, heel size, heel cycle / dispersion, heel tendency, apnea syndrome risk, sleeping posture, and the like. The case where it displays by matching is illustrated. In the item of time, the time when a series of processes by the sleep state analyzing application program 100 is executed is shown. In the section item, the feature amount analysis section determined by the feature amount analysis section determination unit 101 is shown. In the item of the height of the eyelid, an average value of the wave intensity in the high frequency band in the feature amount analysis section including the first feature point extracted by the feature point calculation unit 102 is shown. In the item of the size of the wrinkles, an average value of the wave intensity over the entire frequency band in the feature amount analysis section including the second feature point extracted by the feature point calculation unit 102 is shown. In the item of the wrinkle period / dispersion, a variance value of a period from a predetermined wrinkle to a next wrinkle in the feature amount analysis section calculated by the feature point calculation unit 102 is shown. In addition, the item of the height of the eyelids, the item of the size of the eyelids, and the item of the period / dispersion of the eyelids may indicate time series data used for calculating various average values and variance values. In the item of heel tendency, the tendency of heel determined by the heel tendency determining unit 103 is shown. Here, a case is shown in which the severity of “high-tone tendency”, “high-tone tendency”, and “irregular tendency” is displayed in a form that can be understood. In the item of apnea syndrome onset risk, the total value calculated by the onset risk calculation unit 104 is shown as an onset risk, and the higher the total value, the higher the risk of developing sleep apnea syndrome. In the sleeping posture item, the sleeping posture estimated by the sleeping posture estimation unit 105 is shown.

  According to the second embodiment described above, the sleeping posture is estimated, and the estimation result is associated with the risk of developing sleep apnea syndrome, etc., and can be presented to the user. So, when the user is in a sleeping posture, the risk of developing sleep apnea syndrome is high, and conversely, in what sleeping posture, the risk of developing sleep apnea syndrome is low Therefore, in order to reduce the risk of developing sleep apnea syndrome, it is possible to take measures such as what sleeping posture should be taken.

<Third Embodiment>
Next, a third embodiment will be described with reference to FIG. In this embodiment, a case where the sleep state analyzing application program 100 includes an obstructive heel extracting unit 107 instead of the above-described units 103 to 106 will be described. In the present embodiment, the same reference numerals are given to components having the same functions as those in the first and second embodiments described above, and detailed description thereof will be omitted.

The obstructive fistula extraction unit 107 includes time-series data (specifically, as shown in FIG. 8C) indicating the intensity of waves in the frequency band equal to or higher than the threshold value TH of the feature amount analysis section calculated by the feature point calculation unit 102. When the input of the time series data) is received, a section having a value equal to or greater than a preset occlusion determination threshold value is searched from the waveform indicated by the time series data received, and the section is determined to be occlusive. The process of extracting as a section with a high habit is executed. The occlusion determination threshold is preferably set in the range of 5 [G ² / Hz] to 20 [G ² / Hz].

  According to the third embodiment described above, since it is possible to extract wrinkles related to obstructive sleep apnea syndrome, which is said to have the largest number of patients among sleep apnea syndrome, the user himself It is possible to grasp whether it is obstructive sleep apnea syndrome or a sleep apnea syndrome other than obstructive sleep apnea.

  According to at least one embodiment described above, the manifestation of a patient with sleep apnea syndrome can be realized.

  Note that the processing of the present embodiment can be realized by a computer program. Therefore, the computer program can be installed and executed on a computer through a computer-readable storage medium storing the computer program. Similar effects can be easily realized.

  In addition, although some embodiment of this invention was described, these embodiment is shown as an example and is not intending limiting the range of invention. These novel embodiments can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the scope of the invention. These embodiments and modifications thereof are included in the scope and gist of the invention, and are included in the invention described in the claims and the equivalents thereof.

  DESCRIPTION OF SYMBOLS 1 ... Sleeping habit analysis system, 10 ... Biosensor apparatus, 11 ... Electronic device, 100 ... Sleeping habit analysis application program, 101 ... Feature-value analysis area determination part, 102 ... Feature point calculation part, 103 ... Acupuncture tendency determination part 104 ... Onset risk calculation unit, 105 ... Sleeping posture estimation unit, 106 ... Association unit, 107 ... Obstructive wrinkle extraction unit.

Claims (13)

A sleep trap analysis device that can be worn on a human body,
A measuring means for measuring the acceleration of the wearer;
First generation means for generating first time-series data indicating the intensity of waves in a frequency band equal to or higher than a first threshold value of the measured acceleration;
Determining means for determining a tendency of wrinkles of the wearer based on the generated first time-series data;
A sleep wrinkle analysis device comprising: calculating means for calculating a risk of developing sleep apnea syndrome based on the determined wrinkle tendency. The first generation means includes:
Further generating second time-series data indicating the intensity of waves in a frequency band of the measured acceleration less than the first threshold;
The determining means includes
The ratio of the wave intensity indicated by the second time series data to the wave intensity indicated by the first time series data is calculated, and the wrinkle tendency of the wearer is calculated based on the calculated ratio. The sleep state analyzing apparatus according to claim 1, wherein the sleep state analyzing apparatus is determined. The first generation means includes:
Further generating third time-series data indicating the intensity of the wave of the measured acceleration in all frequency bands,
The determining means includes
2. The sleep state analyzing apparatus according to claim 1, further determining a tendency of the wearer's eyelid based on the generated third time-series data. The first generation means includes:
Based on the generated third time-series data, further generating period information indicating the period of the wearer's eyelids,
The determining means includes
The sleep state analyzing apparatus according to claim 3, further comprising determining a tendency of the wearer's wrinkle based on the generated period information. The determining means includes
The sleep time according to claim 1, wherein a tendency of wrinkles of the wearer is determined based on at least one of an average value and a variance value of wave intensities indicated by the generated first time-series data.鼾 Analysis device. The determining means includes
The sleep time according to claim 3, wherein the tendency of the wearer's wrinkle is determined based on at least one of an average value and a variance value of wave intensities indicated by the generated third time-series data.鼾 Analysis device. The determining means includes
The sleep time according to claim 4, wherein a tendency of the wearer's wrinkle is determined based on at least one of an average value and a variance value of the wearer's wrinkle cycle indicated by the generated cycle information.鼾 Analysis device. Based on the measured acceleration, further comprising an extracting means for extracting a section where the wearer's posture fluctuation amount is small and periodic vibration is generated;
The first generation means includes:
The sleep time lapse analysis apparatus according to claim 1, wherein the first time series data is generated based on the acceleration of the extracted section. The extraction means includes
Based on the DC component of the measured acceleration, extract a section with a small amount of posture variation of the wearer,
Among the measured accelerations, an autocorrelation function is calculated based on an alternating current component of the extracted section, and a peak value of the calculated autocorrelation function is equal to or greater than a second threshold value The sleep state analyzing apparatus according to claim 8, wherein the extracted section is extracted as a section suitable for generating the first time-series data.   10. The apparatus according to claim 9, further comprising: a second extraction unit that extracts a section having a value equal to or greater than a third threshold among sections extracted by the extraction unit as a section having a high obstruction property. 10. Sleeping habit analysis device. Estimating means for estimating the sleeping posture of the wearer based on the measured acceleration;
At least one of the sleeping posture information indicating the estimated sleeping posture, the generated first time-series data, the determined tendency of wrinkles, and the calculated risk of developing sleep apnea syndrome The sleep state analyzing device according to claim 1, further comprising: a second generation unit configured to generate association result information associated with each other. It is a method applied to a sleep heel analyzer that can be worn on a human body,
Measuring the wearer's acceleration;
Generating first time-series data indicating the intensity of waves in a frequency band equal to or higher than the first threshold of the measured acceleration, and based on the generated first time-series data, Determining trends in
Calculating a risk of developing sleep apnea syndrome based on the determined tendency of hemorrhoids. A program for a sleeping habit analyzer that can be worn on a human body,
The sleep bud analyzer
A measuring means for measuring the acceleration of the wearer;
Generating means for generating first time-series data indicating a wave intensity in a frequency band equal to or higher than a first threshold value of the measured acceleration;
Determining means for determining a tendency of wrinkles of the wearer based on the generated first time-series data;
A program for functioning as a calculation means for calculating a risk of developing sleep apnea syndrome based on the determined tendency of wrinkles.

Images