Respiratory rate measuring device, respiratory rate measuring system, respiratory rate measuring program, and respiratory rate measuring method

Abstract

To when contactlessly measuring a respiratory rate of a human or an animal using a radar, after eliminating false detection and improving accuracy, highly accurately determine a sleep state (good sound sleep, bed rest, stertor or respiratory arrest) of a human or an animal, and determine an active state (absence, presence, activity or inactivity) of the human or the animal.SOLUTION: A respiratory rate measuring device calculates, for a difference signal between beat signals with different transmission and reception periods, a frequency integration spectrum within a prescribed integration period of a frequency spectrum so as not to split a time variation of peak intensity of the frequency integration spectrum into two peaks in one respiratory cycle. Then, the respiratory rate measuring device extracts from beat signals of a plurality of transmission / reception cycles, a plurality of pieces of time-series data of transmission signals with a plurality of discrete frequencies out of sweep frequencies and selects time-series data with a maximum amplitude out of the plurality of pieces of time-series data. The respiratory rate measuring device can thus solve the above-mentioned problem.SELECTED DRAWING: Figure 6

Description

[Technical Field]

[0001] The present disclosure relates to a technique for measuring the respiration rate of a human or animal without contact using radar. [Background technology]

[0002] By measuring the respiratory rate of a human or animal without contact using radar, it is possible to monitor sleep apnea syndrome and the like in the human or animal (see, for example, Patent Documents 1 to 5, etc.).

[0003] A first prior art method for measuring respiration rate is shown in Figure 1. In Figure 1, a Doppler signal (I component and Q component) is acquired between the transmitted and received signals of a Doppler radar. Then, based on the period of the Doppler signal (I component and Q component), the respiration rate of a human or animal [times / min] = 60 / period [sec] is measured. However, because the Doppler signal (I component and Q component) does not contain information about distance, vibrations of a bed or wall surface may be mistaken for respiration of a human or animal.

[0004] A second prior art method for measuring respiration rate is shown in Figure 2. In Figure 2, a beat signal between the transmit and receive signals of an FMCW radar is acquired. Then, a differential signal between the beat signals with different transmit and receive cycles is calculated. The differential signal has a finite amplitude for a breathing human or animal, but has almost no amplitude for a stationary bed or wall. The frequency spectrum of the differential signal is then calculated. Furthermore, the human or animal's respiration rate [times / min] = 60 / cycle [sec] is measured based on the period of time change in the peak phase or peak intensity of the frequency spectrum. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Japanese Patent Application Laid-Open No. 2014-210137 [Patent Document 2] Japanese Patent Application Laid-Open No. 2017-134040 [Patent Document 3] Japanese Patent Application Publication No. 2018-094225 [Patent Document 4] Japanese Patent Application Publication No. 2019-152441 [Patent Document 5] Special Publication No. 2018-003751 Summary of the Invention [Problem to be solved by the invention]

[0006] Here, breathing of humans or animals during sleep has the following characteristics. First, as sleep deepens, the respiratory rate decreases, but when the air temperature or body temperature is high, the respiratory rate increases. Second, when the respiratory rate is low, the displacement of the body surface does not change sinusoidally over time, but changes in a pulsed manner over time (see, for example, Figure 4 described below). Third, depending on the size of the subject, the displacement of the body surface is a short distance of about a few mm. Fourth, when sleep is deep, the position of the subject does not change much and remains in the same position.

[0007] The first problem to be solved in measuring respiration rate is shown in Figure 3. In Figure 1, the displacement of the body surface is equal to or greater than half the wavelength of the transmitted signal, so the Doppler signal (I and Q components) has a phase range of 360° or more and a large amplitude. In Figure 3, as explained as the third characteristic of breathing during human or animal sleep, the displacement of the body surface is only about or less than half the wavelength of the transmitted signal, so unlike Figure 1, the Doppler signal (I and Q components) cannot be obtained.

[0008] In the upper part of Figure 3, the I component signal has an initial phase of 90° and a narrow phase range centered around 90° depending on the distance of the human or animal, so its amplitude is small and its signal period is half the original respiratory cycle. The Q component signal has an initial phase of 0° and a narrow phase range centered around 0° depending on the distance of the human or animal, but its amplitude is large and its signal period maintains the original respiratory cycle. Therefore, as explained as the fourth characteristic of human or animal breathing during sleep, when the position of the subject does not change much and remains in the same position, it may not be possible to measure the subject's respiratory rate with high accuracy.

[0009] In the middle part of Figure 3, the I component signal has an initial phase of 45° and a narrow phase range centered around 45° depending on the distance of the human or animal, so the signal period maintains the original respiratory cycle but has a medium amplitude. The Q component signal has an initial phase of -45° and a narrow phase range centered around -45° depending on the distance of the human or animal, so the signal period maintains the original respiratory cycle but has a medium amplitude. However, as explained above as the fourth feature, the respiratory rate of the subject can be measured with a certain degree of accuracy even when the subject's position does not change much and remains in the same position.

[0010] In the lower part of Figure 3, the I component signal has an initial phase of -45° and a narrow phase range centered around -45° depending on the distance of the human or animal, so the signal period maintains the original respiratory cycle but has a medium amplitude. The Q component signal has an initial phase of -135° and a narrow phase range centered around -135° depending on the distance of the human or animal, so the signal period maintains the original respiratory cycle but has a medium amplitude. However, as explained in the fourth feature above, the respiratory rate of the subject can be measured with a certain degree of accuracy even when the subject's position does not change much and remains in the same position.

[0011] The second problem to be solved in respiration rate measurement is shown in FIG. 4. In FIG. 3, the I or Q component signal only has a frequency proportional to velocity and may result in a null signal. In FIG. 2, the beat signal or differential signal has a frequency proportional to distance and therefore does not result in a null signal. However, when measuring the respiration rate of a human or animal based on the peak frequency of the frequency spectrum, the accuracy of distance measurement and the accuracy of respiration rate measurement are low due to the limited frequency band. Furthermore, when measuring the respiration rate of a human or animal based on the period of time change in the peak phase of the frequency spectrum, if the phase range exceeds 360° and the continuity of the phase change is lost, it is impossible to respond to sudden changes in body surface displacement. Therefore, when measuring the respiration rate of a human or animal based on the period of time change in peak intensity of the frequency spectrum, a problem arises: as described as the second characteristic of human or animal breathing during sleep, when the respiration rate is low, the body surface displacement does not change sinusoidally over time but changes in a pulse-like manner over time.

[0012] That is, as shown in the upper part of FIG. 4, when the displacement of the body surface changes suddenly (for example, when the chest or abdomen expands or contracts), the amplitude of the differential signal increases. As a result, as shown in the lower part of FIG. 4, the peak intensity of the frequency spectrum increases. On the other hand, as shown in the upper part of FIG. 4, when the displacement of the body surface changes very little (for example, when the chest or abdomen expands or contracts completely), the amplitude of the differential signal decreases. As a result, as shown in the lower part of FIG. 4, the peak intensity of the frequency spectrum decreases. Therefore, as shown in the lower part of FIG. 4, the time change in the peak intensity of the frequency spectrum splits into two peaks during one actual respiratory cycle, and the respiratory rate of a human or animal is erroneously detected as approximately twice the actual respiratory rate, making it impossible to measure with high accuracy.

[0013] As described above, when measuring the respiratory rate of a human or animal without contact using radar, it was not possible to eliminate false detections and achieve high accuracy. Therefore, it was not possible to determine the sleep state of a human or animal (restful sleep, quiet, heavy breathing, or respiratory cessation) with high accuracy. In fact, it was not possible to determine the activity state of a human or animal (absence, presence, activity, or inactivity).

[0014] Therefore, in order to solve the above problems, the present disclosure aims to measure the respiratory rate of a human or animal non-contactly using radar, eliminate false detections and improve accuracy, and then determine the sleeping state of the human or animal (restful sleep, quiet, heavy breathing or respiratory cessation) with high accuracy, as well as the activity state of the human or animal (absent, present, active or inactive). [Means for solving the problem]

[0015] To solve the above problem, a frequency integrated spectrum within a predetermined integration period of the frequency spectrum is calculated for a differential signal between beat signals of different transmission and reception cycles, and the time change in peak intensity of the frequency integrated spectrum is prevented from splitting into two peaks in one respiratory cycle. Then, multiple time series data for transmission signals having multiple discrete frequencies among the sweep frequencies are extracted from the beat signals of multiple transmission and reception cycles, and the time series data with the maximum amplitude is selected from the multiple time series data. As a result, it is possible to accurately determine the sleep state (restful, quiet, heavy breathing, or respiratory cessation) of a human or animal and to determine the activity state (absence, presence, activity, or inactivity) of the human or animal based on the frequency integrated spectrum within the predetermined integration period or the time series data with the maximum amplitude, as described below.

[0016] Specifically, the present disclosure provides a respiration rate measurement device that measures the respiration rate of a measurement target using an FMCW radar, the respiration rate measurement device comprising: a beat signal acquisition unit that acquires a beat signal between transmission and reception signals of the FMCW radar; a differential signal calculation unit that calculates a differential signal between the beat signals of different transmission and reception cycles; a frequency spectrum calculation unit that calculates the frequency spectrum of the differential signal; a frequency integrated spectrum calculation unit that calculates a frequency integrated spectrum within a predetermined integration period of the frequency spectrum so that a change over time in the peak intensity of the frequency integrated spectrum does not split into two peaks in one respiratory cycle; and an absence determination unit that determines that the measurement target is absent from the irradiation range of the FMCW radar when a target activity amount measured based on the peak intensity of the frequency integrated spectrum is smaller than a predetermined activity amount threshold, or when a target distance measured based on the peak frequency of the frequency integrated spectrum is larger than a predetermined distance threshold.

[0017] With this configuration, even when the displacement of the body surface hardly changes (for example, when the chest or abdomen is completely swollen), the amplitude of the differential signal becomes small, but the peak intensity of the frequency integrated spectrum becomes large. Therefore, even when the displacement of the body surface changes over time not in a sinusoidal manner but in a pulsed manner, the activity amount and distance of a human or animal can be measured with high accuracy without false detection. Then, the activity state of a human or animal (presence or absence within the irradiation range of the FMCW radar) can be determined based on the frequency integrated spectrum within a specified integration period.

[0018] The present disclosure also provides a respiration rate measurement device characterized by further comprising an activity determination unit that determines that the measurement object is active when, after it is determined that the measurement object is present within the irradiation range of the FMCW radar, a distance deviation, which is the difference between the maximum and minimum values ​​of the target distance within a predetermined activity determination period measured based on the peak frequency of the frequency integration spectrum, is greater than a predetermined distance deviation threshold.

[0019] According to this configuration, the activity state of a human or animal (active with tossing and turning in bed or inactive without tossing and turning in bed) can be determined based on the frequency integrated spectrum within a predetermined integration period.

[0020] The present disclosure also provides a respiration rate measurement device further comprising: a time series data extraction unit that extracts, after it is determined that the measurement object is present within the irradiation range of the FMCW radar and that the measurement object is inactive, multiple time series data for a transmission signal having multiple discrete frequencies among sweep frequencies from the beat signal of multiple transmission and reception periods; a respiration rate measurement unit that measures the measurement object's respiration rate based on the fluctuation period of the time series data having the maximum amplitude among the multiple time series data; a respiration rate reliability determination unit that determines the reliability of the measurement object's respiration rate based on the standard deviation of the fluctuation period of the time series data having the maximum amplitude among the multiple time series data; and a sound sleep determination unit that determines whether the measurement object is sleeping soundly / at rest when the ratio of periods in which the reliability of the measurement object's respiration rate within a specified sound sleep determination period is higher than a specified reliability threshold is larger / smaller than a specified period ratio threshold.

[0021] With this configuration, even when the displacement of the body surface is only about half or less than half the wavelength of the transmitted signal, and even when the position of the human or animal does not change much and remains the same, the respiratory rate of the human or animal can be measured with high accuracy without false detection. Furthermore, even when the time-series data with the maximum amplitude does not contain distance information and is therefore likely to contain noise from the bed or wall, the reliability of the human or animal's respiratory rate can be determined with high accuracy without false detection. Therefore, the sleep state of the human or animal (restful sleep without convulsions, etc., or rest with convulsions, etc.) can be determined with high accuracy based on the time-series data with the maximum amplitude.

[0022] The present disclosure also provides a respiratory rate measuring device that is further characterized by including an alarm issuing unit that does not determine whether the subject is sleeping soundly and issues an alarm for the respiratory rate of the subject when, for the specified sound sleep judgment period, the ratio of periods in which the reliability of the respiratory rate of the subject is higher than the specified reliability threshold is greater than the specified period ratio threshold, but the average value of the respiratory rates of the subject that have a reliability higher than the specified reliability threshold for the respiratory rates of the subject during the specified sound sleep judgment period is greater than the specified respiratory rate threshold.

[0023] According to this configuration, the sleep state (heavy breathing or calm breathing) of a human or animal can be determined with high accuracy based on the time-series data having the maximum amplitude.

[0024] The present disclosure also provides a respiration rate measurement device characterized by further comprising a respiration cessation determination unit that determines that the measurement target has stopped breathing when it is determined that the measurement target is present in the irradiation range of the FMCW radar and that the measurement target is not active, and when a subtraction value obtained by subtracting a median value within a shorter period than the expected respiration cessation period or an instantaneous value at a certain time point from a median value within a longer period than the expected respiration cessation period, for the target activity amount measured based on the peak intensity of the frequency integrated spectrum, is greater than a predetermined subtraction value threshold.

[0025] According to this configuration, the sleep state (respiratory cessation or non-respiratory cessation) of a human or animal can be determined with high accuracy based on the frequency integrated spectrum within a predetermined integration period.

[0026] The present disclosure also provides a respiratory rate measurement device characterized in that the respiratory arrest determination unit determines that the measurement subject has stopped breathing when the subtraction value obtained by subtracting the median value within the long period of time having the same time center as the time center of the long period of time, or the instantaneous value at the same time as the time center of the long period of time, from the median value within the long period of time, for the target activity amount measured based on the peak intensity of the frequency integrated spectrum, is greater than the predetermined subtraction value threshold.

[0027] This configuration makes it possible to prevent erroneous determination of respiratory arrest as a sleeping state of a human or animal based on the frequency integrated spectrum within a predetermined integration period.

[0028] The present disclosure also provides a respiratory rate measurement device characterized in that the respiratory stop determination unit extracts the maximum value of the subtraction value within the estimated respiratory stop period while moving the estimated respiratory stop period on a time axis, and determines the time from when the maximum value of the subtraction value within the estimated respiratory stop period starts to become larger than the predetermined subtraction value threshold to when it starts to become smaller as one continuous respiratory stop period of the measurement subject.

[0029] According to this configuration, the sleep state (one continuous episode of respiratory cessation) of a human or animal can be determined with high accuracy based on the frequency integrated spectrum within a predetermined integration period.

[0030] The present disclosure also provides a respiration rate measurement system including the respiration rate measurement device described above and a radar transmitting and receiving device for the FMCW radar.

[0031] According to this configuration, it is possible to provide a system having the above-described effects.

[0032] The present disclosure also provides a respiration rate measurement program for causing a computer to execute the processing steps performed by each processing unit included in the respiration rate measurement device described above.

[0033] According to this configuration, it is possible to provide a program having the above-described effects.

[0034] The present disclosure also provides a respiration rate measurement method, characterized by executing the processing steps performed by the processing units included in the respiration rate measurement device described above.

[0035] According to this configuration, it is possible to provide a processing procedure having the above-described effects. [Effects of the Invention]

[0036] In this way, the present disclosure can measure the breathing rate of a human or animal non-contactly using radar, eliminating false detections and improving accuracy, and can also determine the sleeping state of a human or animal (restful, quiet, heavy breathing or respiratory cessation) with high accuracy, as well as the activity state of a human or animal (absent, present, active or inactive). [Brief explanation of the drawings]

[0037] [Figure 1] FIG. 1 shows a first respiratory rate measurement according to the prior art. [Figure 2] FIG. 1 illustrates a second respiratory rate measurement according to the prior art. [Figure 3] FIG. 1 is a diagram showing a first problem to be solved in measuring respiratory rate. [Figure 4] FIG. 10 is a diagram showing a second problem to be solved in measuring respiratory rate. [Figure 5] FIG. 1 is a diagram illustrating the configuration of a respiratory rate measurement system according to the present disclosure. [Figure 6] FIG. 2 is a diagram illustrating a configuration of a sleep state determination unit according to the present disclosure. [Figure 7] FIG. 1 illustrates a respiratory rate measurement procedure of the present disclosure. [Figure 8] FIG. 10 is a diagram illustrating a calculation process of a frequency integrated spectrum according to the present disclosure. [Figure 9] FIG. 10 is a diagram illustrating the effect of calculating a frequency integrated spectrum according to the present disclosure. [Figure 10] FIG. 10 is a diagram illustrating a calculation process of time-series data according to the present disclosure. [Figure 11] FIG. 1 illustrates a respiratory rate measurement process according to the present disclosure. [Figure 12] FIG. 10 is a diagram illustrating a process for determining the reliability of a respiration rate according to the present disclosure. [Figure 13] FIG. 10 is a diagram illustrating the absence / activity determination procedure of the present disclosure. [Figure 14] FIG. 10 is a diagram illustrating an absence determination process according to the present disclosure. [Figure 15] FIG. 10 is a diagram illustrating an activity determination process of the present disclosure. [Figure 16] FIG. 1 is a diagram showing a procedure for determining whether a person is asleep or at rest according to the present disclosure. [Figure 17] FIG. 10 is a diagram showing the sound sleep / rest determination process of the present disclosure. [Figure 18] FIG. 10 is a diagram illustrating an alarm issuing process according to the present disclosure. [Figure 19] FIG. 10 illustrates a respiratory arrest determination procedure of the present disclosure. [Figure 20] FIG. 10 is a diagram showing a breathing arrest determination process of the present disclosure. [Figure 21] FIG. 10 is a diagram showing a breathing arrest determination process of the present disclosure. [Figure 22] FIG. 10 is a diagram showing a breathing arrest determination process of the present disclosure. [Figure 23] FIG. 10 is a diagram showing a breathing arrest determination process of the present disclosure. [Figure 24] FIG. 10 is a diagram showing a breathing arrest determination process of the present disclosure. DETAILED DESCRIPTION OF THE INVENTION

[0038]

[0023] The following embodiments of the present disclosure will be described with reference to the accompanying drawings. The embodiments described below are examples of implementation of the present disclosure, and the present disclosure is not limited to the following embodiments.

[0039] (Configuration of the respiration rate measurement system of the present disclosure) The configuration of the respiration rate measurement system of the present disclosure is shown in Fig. 5. The respiration rate measurement system S includes an FMCW radar transceiver 1 and a respiration rate measurement device 2. The FMCW radar transceiver 1 irradiates a human or animal T with FMCW radar. The respiration rate measurement device 2 includes a beat signal acquisition unit 21, a differential signal calculation unit 22, a frequency spectrum calculation unit 23, a frequency integrated spectrum calculation unit 24, a time series data extraction unit 25, a respiration rate measurement unit 26, a respiration rate reliability determination unit 27, and a sleep state determination unit 28. The respiration rate measurement device 2 can be realized by installing the respiration rate measurement program shown in Fig. 7 on a computer.

[0040] The configuration of the sleep state determination unit of the present disclosure is shown in Fig. 6. The sleep state determination unit 28 includes an absence determination unit 281, an activity determination unit 282, a sound sleep determination unit 283, an alarm issuance unit 284, and a breathing arrest determination unit 285. The sleep state determination unit 28 can be realized by installing the sleep state determination programs shown in Figs. 13, 16, and 19 in a computer.

[0041] The respiratory rate measurement procedure of the present disclosure is executed using the units from the beat signal acquisition unit 21 to the respiratory rate reliability determination unit 27. The absence / activity determination procedure of the present disclosure is executed using the absence determination unit 281 and the activity determination unit 282. The sound sleep / rest determination procedure of the present disclosure is executed using the sound sleep determination unit 283 and the alarm issuance unit 284. The respiratory arrest determination procedure of the present disclosure is executed using the respiratory arrest determination unit 285. Each processing procedure of the present disclosure will be described below.

[0042] (Respiration Rate Measurement Procedure of the Present Disclosure) The respiratory rate measurement procedure of the present disclosure is shown in Figure 7. The calculation process of the frequency integrated spectrum of the present disclosure is shown in Figure 8. The calculation effect of the frequency integrated spectrum of the present disclosure is shown in Figure 9. The calculation process of the time series data of the present disclosure is shown in Figure 10. The measurement process of the respiratory rate of the present disclosure is shown in Figure 11. The determination process of the reliability of the respiratory rate of the present disclosure is shown in Figure 12. The beat signal acquisition unit 21 acquires and stores the beat signal between the transmitted and received signals of the FMCW radar (step S1). Below, steps S2 to S4 and steps S5 to S7 may be performed in parallel or one after the other.

[0043] The differential signal calculation unit 22 calculates the differential signal between the beat signals of different transmission and reception periods (step S2). The frequency spectrum calculation unit 23 calculates the frequency spectrum of the differential signal (step S3). The differential signal and frequency spectrum are the same as those in the second prior art shown in FIG.

[0044] The frequency integrated spectrum calculation unit 24 calculates the frequency integrated spectrum within a predetermined integration period of the frequency spectrum so that the time change in peak intensity of the frequency integrated spectrum does not split into two peaks in one respiratory cycle (step S4).

[0045] In the left column of Fig. 8, the frequency spectrum has a peak near a frequency corresponding to the distance of the human or animal T, and the peak value of the frequency spectrum changes over time according to the displacement of the body surface. In the right column of Fig. 8, the frequency integrated spectrum also has a peak near a frequency corresponding to the distance of the human or animal T, and the predetermined integration period of the frequency spectrum is set according to the policy described below.

[0046] As shown in the upper part of FIG. 9, when the displacement of the body surface changes suddenly (for example, when the chest or abdomen expands or contracts), the amplitude of the differential signal increases. As a result, the peak intensity of the frequency spectrum increases, as shown in the lower part of FIG. 9. On the other hand, when the displacement of the body surface changes very little (for example, when the chest or abdomen expands or contracts completely), as shown in the upper part of FIG. 9, the amplitude of the differential signal decreases. As a result, the peak intensity of the frequency spectrum decreases, as shown in the lower part of FIG. 9.

[0047] However, as shown in the upper and lower parts of Fig. 9, even when the displacement of the body surface reaches a maximum value (for example, when the chest or abdomen is completely expanded), the peak intensity of the frequency integrated spectrum also reaches a maximum value. Of course, as shown in the upper and lower parts of Fig. 9, when the displacement of the body surface changes suddenly (for example, when the chest or abdomen expands or contracts), the peak intensity of the frequency integrated spectrum also changes suddenly. In other words, the time change in the peak intensity of the frequency integrated spectrum does not split into two peaks during one original respiratory cycle, unlike the second conventional technology.

[0048] Here, if the predetermined integration period of the frequency spectrum is short, less than half the expected shortest respiratory cycle of the human or animal T, the time change in peak intensity of the frequency integration spectrum will accentuate the bottom between the peaks, thereby improving the accuracy of measuring the respiratory rate of the human or animal T. On the other hand, if the predetermined integration period of the frequency spectrum is long, approximately equal to the expected respiratory cessation period of the human or animal T, the time change in peak intensity of the frequency integration spectrum will raise the bottom between the peaks, thereby improving the accuracy of measuring the activity amount and distance of the human or animal T.

[0049] In this way, even when the displacement of the body surface hardly changes (for example, when the chest or abdomen is completely swollen), the amplitude of the differential signal becomes small, but the peak intensity of the frequency integrated spectrum becomes large. Therefore, even when the displacement of the body surface does not change over time in a sinusoidal manner but in a pulsed manner, the activity amount and distance of the human or animal T can be measured with high accuracy without false detection.

[0050] The time series data extracting unit 25 extracts a plurality of time series data for a transmission signal having a plurality of discrete frequencies among the sweep frequencies from the beat signals of a plurality of transmission and reception cycles (step S5). In FIG. 10, the time series data extracting unit 25 extracts a plurality of time series data (voltage v 11 , time t 11 ), (voltage v 12 , time t 12 ), .... The time series data extraction unit 25 then extracts time series data (voltage v 21 , time t 21 ), (voltage v 22 , time t 22 ), .... Furthermore, the time series data extraction unit 25 extracts time series data (voltage v 31 , time t 31 ), (voltage v 32 , time t 32), and extract...

[0051] The respiration rate measurement unit 26 measures the respiration rate of the human or animal T based on the fluctuation period of the time-series data having the maximum amplitude among the plurality of time-series data (step S6). In FIG. 11, the respiration rate measurement unit 26 acquires five pieces of time-series data each having an I component and a Q component for a transmission signal having five discrete frequencies. Then, the respiration rate measurement unit 26 measures the respiration rate [times / min]=60 / cycle [sec] of the human or animal T based on the fluctuation period of the time-series data having the maximum amplitude among the five pieces of time-series data each having an I component and a Q component (the signal period maintains the original respiration period). On the other hand, the respiration rate measurement unit 26 does not measure the respiration rate of the human or animal T based on the fluctuation period of the time-series data having the minimum amplitude among the five pieces of time-series data each having an I component and a Q component (the signal period is half the original respiration period).

[0052] In this way, even when the displacement of the body surface is only about half or less than half the wavelength of the transmitted signal, and even when the position of the human or animal T does not change much and remains in the same position, the respiratory rate of the human or animal T can be measured with high accuracy without false detection.

[0053] The respiration rate reliability determination unit 27 determines the reliability of the respiration rate of the human or animal T based on the standard deviation of the fluctuation period of the time series data having the maximum amplitude among the plurality of time series data (step S7). In Fig. 12, the respiration rate measurement unit 26 and the respiration rate reliability determination unit 27 measure intervals t1, t2, t3, and t4 between negative to positive zero crossings or positive to negative zero crossings in the time series data having the maximum amplitude and in which the midpoint between the maximum and minimum values ​​is offset to zero.

[0054] In the left column of Fig. 12, the respiration rate reliability determination unit 27 determines that the smaller the standard deviation of the intervals t1, t2, t3, and t4 between negative and positive zero crossings, the higher the reliability of the respiration rate of the human or animal T. When the respiration rate reliability determination unit 27 determines that the reliability of the respiration rate of the human or animal T is high, it can adopt and output the respiration rate of the human or animal T, as will be described later with reference to Figs. 17 and 18.

[0055] In the right column of Fig. 12, the respiratory rate reliability determination unit 27 determines that the greater the standard deviation of the intervals t1, t2, t3, and t4 between negative and positive zero crossings or between positive and negative zero crossings, the lower the reliability of the respiratory rate of the human or animal T. When the respiratory rate reliability determination unit 27 determines that the reliability of the respiratory rate of the human or animal T is low, it can reject or not output the respiratory rate of the human or animal T, as will be described later with reference to Figs.

[0056] In this way, even when the time series data with the maximum amplitude does not contain distance information and is therefore likely to contain noise from the bed or wall, etc., the reliability of the respiratory rate of a human or animal T can be determined with high accuracy without false detection based on the standard deviation of the fluctuation period of the time series data.

[0057] (Disclosure of Absence / Activity Determination Procedure) The absence / activity determination procedure of the present disclosure is shown in Figure 13. The absence determination process of the present disclosure is shown in Figure 14. The activity determination process of the present disclosure is shown in Figure 15. The activity determination process of the present disclosure is executed after it is determined that a human or animal T is present within the irradiation range of the FMCW radar.

[0058] The absence determining unit 281 measures the amount of target activity based on the peak intensity of the frequency integrated spectrum (step S11), or measures the target distance based on the peak frequency of the frequency integrated spectrum (step S12).

[0059] When the target activity amount is smaller than a predetermined activity amount threshold (step S13, YES), the absence determination unit 281 determines that a human or animal T is absent from the irradiation range of the FMCW radar (step S14). On the other hand, when the target activity amount is larger than a predetermined activity amount threshold (step S13, NO), the absence determination unit 281 determines that a human or animal T is present in the irradiation range of the FMCW radar (step S15). Here, the predetermined activity amount threshold is set to a value between the target activity amount of the human or animal T and the target activity amount of environmental noise (such as a bed or a wall).

[0060] 14, the target activity amount is larger than the predetermined activity amount threshold up to time t0 and from time t1, so it is determined that a human or animal T is present in the irradiation range of the FMCW radar. On the other hand, the target activity amount is smaller than the predetermined activity amount threshold from time t0 to time t1, so it is determined that a human or animal T is not present in the irradiation range of the FMCW radar.

[0061] When the target distance is greater than a predetermined distance threshold (step S13, YES), the absence determination unit 281 determines that a human or animal T is absent from the irradiation range of the FMCW radar (step S14). On the other hand, when the target distance is less than the predetermined distance threshold (step S13, NO), the absence determination unit 281 determines that a human or animal T is present in the irradiation range of the FMCW radar (step S15). Here, the predetermined distance threshold is set to a value between the target distance of the human or animal T and the target distance of environmental noise (such as a bed or a wall).

[0062] 14, up to time t0 and from time t1, the target distance is smaller than the predetermined distance threshold, so it is determined that a human or animal T is present in the irradiation range of the FMCW radar. On the other hand, from time t0 to time t1, the target distance is larger than the predetermined distance threshold, so it is determined that a human or animal T is not present in the irradiation range of the FMCW radar.

[0063] In this way, the activity state (presence or absence in the irradiation range of the FMCW radar) of the human or animal T can be determined based on the frequency integrated spectrum within a predetermined integration period. Here, the time change in peak intensity of the frequency integrated spectrum has a relatively large intensity even at the bottom between the peaks compared to the time change in peak intensity of the frequency spectrum (see FIG. 9), so that a target activity amount that is sufficiently large compared to a predetermined activity amount threshold can be obtained.

[0064] The activity determination unit 282 calculates a distance deviation, which is the difference between the maximum and minimum values ​​of the target distance within a predetermined activity determination period measured based on the peak frequency of the frequency integrated spectrum (step S16). Here, the predetermined activity determination period is set to a value approximately equal to the period from the start to the end of one activity of the human or animal T, such as turning over in bed.

[0065] When the distance deviation is larger than the predetermined distance deviation threshold (step S17, YES), the activity determination unit 282 determines that the human or animal T is active (step S18). On the other hand, when the distance deviation is smaller than the predetermined distance deviation threshold (step S17, NO), the activity determination unit 282 determines that the human or animal T is not active (step S19). Here, the predetermined distance deviation threshold is set to a value that is slightly smaller than the change in target distance from the start to the end of an activity such as one turn over of the human or animal T.

[0066] 15, at time t0, the target distance during the predetermined activity determination period before time t0 changes significantly, and the distance deviation during the predetermined activity determination period before time t0 is larger than the predetermined distance deviation threshold, so it is determined that the human or animal T is active. On the other hand, at time t1, the target distance during the predetermined activity determination period before time t1 remains almost unchanged, and the distance deviation during the predetermined activity determination period before time t1 is smaller than the predetermined distance deviation threshold, so it is determined that the human or animal T is not active. Similar activity determinations are also performed at times other than t0 and t1.

[0067] In this way, it is possible to determine the activity state of the human or animal T (active with tossing and turning, etc., or inactive without tossing and turning, etc.) based on the frequency integrated spectrum within a predetermined integration period. Here, the time change in peak intensity of the frequency integrated spectrum has a relatively large intensity even at the bottom between the peaks compared to the time change in peak intensity of the frequency spectrum (see FIG. 9), so it is possible to obtain the target distance and therefore the distance deviation with sufficiently high accuracy.

[0068] (Procedure for determining restful sleep and quietness disclosed herein) The procedure for determining whether a person is asleep or not according to the present disclosure is shown in Fig. 16. The process for determining whether a person is asleep or not according to the present disclosure is shown in Fig. 17. The process for issuing an alarm according to the present disclosure is shown in Fig. 18. The process for determining whether a person is asleep or not according to the present disclosure is executed after it has been determined that a human being or animal T is present within the irradiation range of the FMCW radar and that the human being or animal T is not in motion.

[0069] The sound sleep determination unit 283 calculates the ratio of periods in which the reliability of the respiratory rate of the human or animal T within a predetermined sound sleep determination period is higher than a predetermined reliability threshold (step S21). Here, the predetermined sound sleep determination period is set to a value approximately equal to the period from the start to the end of a single state such as a convulsion of the human or animal T, and the predetermined reliability threshold is set to a value slightly higher than the low reliability from the start to the end of a single state such as a convulsion of the human or animal T.

[0070] When the period ratio is larger than the predetermined period ratio threshold (step S22, YES), the sound sleep determination unit 283 determines that the human being or animal T is sleeping soundly (step S25). On the other hand, when the period ratio is smaller than the predetermined period ratio threshold (step S22, NO), the sound sleep determination unit 283 determines that the human being or animal T is resting (step S27). Here, the predetermined period ratio threshold is set to a value between the period ratio when the human being or animal T is in a state where there is a convulsion or the like and the period ratio when the human being or animal T is not in a state where there is a convulsion or the like.

[0071] 17, at time t0, the reliability of the respiration rate within the predetermined sound sleep determination period before time t0 is 8 points high and 2 points low, and the validity rate of the respiration rate within the predetermined sound sleep determination period before time t0 (8 / 10) is greater than the predetermined period ratio threshold, so it is determined that the human or animal T is sleeping soundly. On the other hand, at time t1, the reliability of the respiration rate within the predetermined sound sleep determination period before time t1 is 7 points high and 3 points low, and the validity rate of the respiration rate within the predetermined sound sleep determination period before time t1 (7 / 10) is smaller than the predetermined period ratio threshold, so it is determined that the human or animal T is resting. Similar sound sleep determinations are also performed at times other than t0 and t1.

[0072] In this way, based on the time series data having the maximum amplitude, the sleep state (restful sleep without convulsions, etc. or rest with convulsions, etc.) of the human or animal T can be determined with high accuracy. Here, the signal period of the time series data having the maximum amplitude maintains the original respiratory period (see FIG. 11), so the respiratory rate and the reliability of the respiratory rate can be obtained with sufficiently high accuracy.

[0073] However, just because the period ratio is larger than the predetermined period ratio threshold (step S22, YES), it is not possible to determine whether the breathing of the human or animal T is rough or calm. The alarm issuing unit 284 calculates an average value of the breathing rates of the human or animal T within the predetermined sound sleep determination period that has a higher reliability than the predetermined reliability threshold (step S23).

[0074] When the average value is greater than the predetermined respiration rate threshold (step S24, YES), the alarm issuing unit 284 does not determine that the human or animal T is sleeping soundly, and issues an alarm for the respiration rate of the human or animal T (step S26). On the other hand, when the average value is smaller than the predetermined respiration rate threshold (step S24, NO), the alarm issuing unit 284 does not issue an alarm for the respiration rate of the human or animal T, and determines that the human or animal T is sleeping soundly (step S25). Here, the predetermined respiration rate threshold is set to a value between the respiration rate when the human or animal T is breathing heavily and the respiration rate when the human or animal T is breathing calmly.

[0075] 18, at time t0, the reliability of the respiration rate within the predetermined sound sleep determination period before time t0 is 8 points high and 2 points low, the effectiveness rate of the respiration rate within the predetermined sound sleep determination period before time t0 (8 / 10) is greater than the predetermined period ratio threshold, and the average effective respiration rate within the predetermined sound sleep determination period before time t0 is greater than the predetermined respiration rate threshold, so an alarm is issued for the respiration rate of human or animal T. On the other hand, at time t1, the reliability of the respiration rate within the predetermined sound sleep determination period before time t1 is 10 points high, the effectiveness rate of the respiration rate within the predetermined sound sleep determination period before time t1 (10 / 10) is greater than the predetermined period ratio threshold, and the average effective respiration rate within the predetermined sound sleep determination period before time t1 is smaller than the predetermined respiration rate threshold, so a sound sleep is determined for human or animal T. Similar sound sleep determination and alarm issuance are also performed at times other than t0 and t1.

[0076] In this way, based on the time series data having the maximum amplitude, it is possible to determine with high accuracy the sleep state (heavy breathing or calm breathing) of the human or animal T. Here, since the signal period of the time series data having the maximum amplitude maintains the original breathing period (see FIG. 11), it is possible to obtain the breathing rate and the reliability of the breathing rate with sufficiently high accuracy.

[0077] (Respiration arrest determination procedure of the present disclosure) The respiratory arrest determination procedure of the present disclosure is shown in Fig. 19. The respiratory arrest determination process of the present disclosure is shown in Fig. 20 to Fig. 24. The respiratory arrest determination process of the present disclosure is executed in parallel with the sound sleep / rest determination process of the present disclosure after it is determined that a human or animal T is present in the irradiation range of the FMCW radar and that the human or animal T is not moving.

[0078] The respiratory arrest determination unit 285 calculates a subtraction value by subtracting a median value within a shorter period than the expected respiratory arrest period from a median value within a longer period than the expected respiratory arrest period, for the amount of target activity measured based on the peak intensity of the frequency integrated spectrum (step S31). Alternatively, the respiratory arrest determination unit 285 calculates a subtraction value by subtracting an instantaneous value at a single time from a median value within a longer period than the expected respiratory arrest period, for the amount of target activity measured based on the peak intensity of the frequency integrated spectrum (step S31). Here, when the amount of target activity decreases over a period longer than the expected respiratory arrest period, respiratory arrest is determined, and when the amount of target activity decreases only for a period shorter than the expected respiratory arrest period, respiratory arrest is determined.

[0079] When the subtraction value is larger than the predetermined subtraction value threshold (step S32, YES), the breathing arrest determination unit 285 determines that the human being or animal T has stopped breathing (step S33). On the other hand, when the subtraction value is smaller than the predetermined subtraction value threshold (step S32, NO), the breathing arrest determination unit 285 determines that the human being or animal T has not stopped breathing (step S34). Here, the predetermined subtraction value threshold is set to a value between the subtraction value when the human being or animal T has stopped breathing and the subtraction value when the human being or animal T has not stopped breathing.

[0080] In Figure 20, the median value within a short period of time has noise removed compared to the instantaneous value at a single time, but temporary decreases in target activity amount are not removed. On the other hand, the median value within a long period of time has noise removed compared to the instantaneous value at a single time, and temporary decreases in target activity amount are also removed. At time t0, the subtraction value is larger than the predetermined subtraction value threshold, so it is determined that human or animal T has stopped breathing. Similar determinations of respiratory cessation are also made at times other than t0.

[0081] In this way, the sleep state (respiratory arrest or non-respiratory arrest) of the human or animal T can be determined with high accuracy based on the frequency integrated spectrum within a predetermined integration period. Here, the time change of the peak intensity of the frequency integrated spectrum has a relatively large intensity even at the bottom between the peaks compared to the time change of the peak intensity of the frequency spectrum (see FIG. 9), so that the bottom between the peaks will not be used to determine respiratory arrest.

[0082] 21, the breathing arrest determination unit 285 calculates a subtraction value by subtracting the median value of a short period having the same end time as the end time of the long period from the median value of the long period. Alternatively, the breathing arrest determination unit 285 calculates a subtraction value by subtracting the instantaneous value at a single time that is the same as the end time of the long period from the median value of the long period.

[0083] As a result, the median value over a long period of time lags behind the median value over a short period of time and the instantaneous value at a single time. However, at time t0, when breathing actually stops, the subtraction value is larger than the predetermined subtraction value threshold, so it is correctly determined that breathing has stopped in human or animal T. On the other hand, at time t2, when breathing does not actually stop, the subtraction value is larger than the predetermined subtraction value threshold, so it is erroneously determined that breathing has stopped in human or animal T.

[0084] 22, the breathing arrest determination unit 285 calculates a subtraction value by subtracting the median value within a short period having the same time center as the long period from the median value within the long period. Alternatively, the breathing arrest determination unit 285 calculates a subtraction value by subtracting the instantaneous value at a time that is the same time as the time center of the long period from the median value within the long period.

[0085] Then, the median value over a long period of time does not lag behind the median value over a short period of time and the instantaneous value at a single time. Therefore, at time t0, when breathing actually stops, the subtraction value is larger than the predetermined subtraction value threshold, so it is correctly determined that breathing of the human or animal T has stopped. On the other hand, at time t2, when breathing does not actually stop, the subtraction value is smaller than the predetermined subtraction value threshold, so it is correctly determined that breathing of the human or animal T has not stopped.

[0086] Therefore, in FIG. 22, it is possible to prevent erroneous determination of cessation of breathing as the sleeping state of a human or animal T based on the frequency integrated spectrum within a predetermined integration period.

[0087] The breathing arrest determination unit 285 extracts the maximum value of the subtraction value within the estimated breathing arrest period while moving the estimated breathing arrest period on the time axis (step S35). Then, the breathing arrest determination unit 285 determines the period from the time when the maximum value of the subtraction value within the estimated breathing arrest period starts to become larger than the predetermined subtraction value threshold to the time when it starts to become smaller as one continuous breathing arrest period of the human or animal T, rather than multiple intermittent breathing arrest periods (step S36).

[0088] In Figure 23, at time t0, the maximum value of the subtraction value during the assumed breathing-hold period t before time t0 begins to become larger than the predetermined subtraction value threshold. Then, at time t+t1, the maximum value of the subtraction value during the assumed breathing-hold period t before time t+t1 begins to become smaller than the predetermined subtraction value threshold. Therefore, one continuous breathing cessation of human or animal T is determined from time t0 to time t+t1. Then, the point of breathing cessation of human or animal T is determined at the time center (t0+t1) / 2 between the time center of the assumed breathing-hold period t before time t0 and the time center of the assumed breathing-hold period t before time t+t1.

[0089] In the left column of Figure 24, at time t0, the subtraction value has a maximum value greater than the predetermined subtraction value threshold. The subtraction value remains the same throughout the period t, from the maximum subtraction value within the assumed respiratory cessation period t before time t0 to the maximum subtraction value within the assumed respiratory cessation period t before time t0+t. Thus, at time t0, the point of respiratory cessation of the human or animal T is determined.

[0090] In the middle column of Figure 24, at time t1, the subtraction value has a first maximum value greater than the predetermined subtraction value threshold, and shortly after time t1, the subtraction value has a second maximum value greater than the predetermined subtraction value threshold. The subtraction value remains the same throughout the period t, from the maximum value of the subtraction value within the assumed respiratory cessation period t before time t1 to the maximum value of the subtraction value within the assumed respiratory cessation period t before time t1+t. Thus, the point of respiratory cessation for human or animal T is determined at time t1.

[0091] In the right column of Figure 24, from time t2 to time t3, the subtraction value has a substantially flat maximum value that is larger than a predetermined subtraction value threshold. From the maximum subtraction value within the assumed breathing-hold period t before time t2 to the maximum subtraction value within the assumed breathing-hold period t before time t3+t, the subtraction value remains the same over the period t+(t3-t2). Therefore, from time t2 to time t3+t, one continuous breathing cessation of human or animal T is determined. Then, at time (t2+t3) / 2, the point of breathing cessation of human or animal T is determined.

[0092] In this way, the sleep state (one continuous episode of respiratory cessation) of the human or animal T can be determined with high accuracy based on the frequency integrated spectrum within a predetermined integration period. [Industrial Applicability]

[0093] The respiratory rate measurement device, respiratory rate measurement system, respiratory rate measurement program, and respiratory rate measurement method disclosed herein use radar to measure the respiratory rate of a human or animal in a non-contact manner, thereby reducing stress on the human or animal, continuously recording the daily sleeping state of a human or animal, monitoring the sleeping state of a human or animal during hospitalization or care, and monitoring for sleep apnea syndrome and the like in a human or animal. [Explanation of symbols]

[0094] S: Respiratory rate measurement system T: Human or animal 1: FMCW radar transmitter and receiver 2:Respiration rate measuring device 21: Beat signal acquisition unit 22: Differential signal calculation unit 23: Frequency spectrum calculation section 24: Frequency integration spectrum calculation section 25: Time series data extraction unit 26:Respiration rate measuring section 27: Respiration rate reliability judgment unit 28: Sleep state determination unit 281: Absence Judgment Department 282: Activity Judgment Department 283: Sleep judgment department 284: Alarm section 285: Breathing stop determination unit

Claims

1. A respiration rate measurement device that measures the respiration rate of a subject using an FMCW radar, a beat signal acquisition unit that acquires a beat signal between a transmission signal and a reception signal of the FMCW radar; a differential signal calculation unit that calculates a differential signal between the beat signals of different transmission and reception periods; a frequency spectrum calculation unit that calculates a frequency spectrum of the difference signal; a frequency integrated spectrum calculation unit that calculates a frequency integrated spectrum within a predetermined integration period of the frequency spectrum, so that even when a time change in peak intensity of the frequency spectrum is split into two peaks in one respiratory cycle, the time change in peak intensity of the frequency integrated spectrum does not split into two peaks in one respiratory cycle; an absence determination unit that determines that the measurement target is absent from the irradiation range of the FMCW radar when the target activity amount measured based on the peak intensity of the frequency integrated spectrum is smaller than a predetermined activity amount threshold, or when the target distance measured based on the peak frequency of the frequency integrated spectrum is larger than a predetermined distance threshold, and stops determining whether the measurement target is active; A respiratory rate measuring device comprising:

2. After determining that the measurement target is present within the irradiation range of the FMCW radar, an activity determination unit that determines that the measurement target is active when a distance deviation, which is a difference between a maximum value and a minimum value of a target distance within a predetermined activity determination period measured based on a peak frequency of the frequency integrated spectrum, is larger than a predetermined distance deviation threshold; The respiration rate measuring device of claim 1 further comprising:

3. It is determined that the measurement target is present in the irradiation range of the FMCW radar and that the measurement target is not active, a time series data extracting unit that extracts, from the beat signal of a plurality of transmission and reception periods, a plurality of time series data for a transmission signal having a plurality of discrete frequencies among the sweep frequencies; a respiration rate measurement unit that measures the respiration rate of the subject based on a fluctuation period of the time series data having the maximum amplitude among the plurality of time series data; a respiration rate reliability determination unit that determines the reliability of the respiration rate of the measurement subject based on the standard deviation of the fluctuation period of the time series data having the largest amplitude among the plurality of time series data; a sound sleep determination unit that determines whether the subject is sleeping soundly or at rest when a ratio of a period in which the reliability of the respiratory rate of the subject within a predetermined sound sleep determination period is higher than a predetermined reliability threshold is larger / smaller than a predetermined period ratio threshold; The respiration rate measuring device of claim 2 further comprising:

4. an alarm issuing unit that does not determine whether the subject is sleeping soundly and issues an alarm for the respiratory rate of the subject when, for the respiratory rate of the subject during the predetermined sound sleep determination period, a ratio of periods in which the reliability of the respiratory rate of the subject is higher than the predetermined reliability threshold is greater than the predetermined period ratio threshold, but an average value of the respiratory rates of the subject during the predetermined sound sleep determination period that have a reliability higher than the predetermined reliability threshold is greater than a predetermined respiratory rate threshold; The respiration rate measuring device of claim 3 further comprising:

5. It is determined that the measurement target is present in the irradiation range of the FMCW radar and that the measurement target is not active, a respiratory arrest determination unit that determines that the measurement object has stopped breathing when a subtraction value obtained by subtracting a median value within a shorter period than the assumed respiratory arrest period or an instantaneous value at a certain time from a median value within a longer period than the assumed respiratory arrest period is greater than a predetermined subtraction value threshold value, regarding the target activity amount measured based on the peak intensity of the frequency integrated spectrum; The respiration rate measuring device according to claim 2 , further comprising:

6. The respiratory arrest determination unit determines that the measurement object has stopped breathing when a subtraction value obtained by subtracting a median value within the short period having the same time center as a time center of the long period or an instantaneous value at the same time as the time center of the long period from a median value within the long period is greater than the predetermined subtraction value threshold value, for the target activity amount measured based on the peak intensity of the frequency integrated spectrum.

6. The respiration rate measuring device according to claim 5.

7. The breathing arrest determination unit extracts the maximum value of the subtraction value within the estimated breathing arrest period while moving the estimated breathing arrest period on a time axis, and determines the period from the time when the maximum value of the subtraction value within the estimated breathing arrest period starts to become larger than the predetermined subtraction value threshold to the time when the maximum value begins to become smaller than the predetermined subtraction value threshold as one continuous breathing arrest period of the measurement object.

7. The respiration rate measuring device according to claim 5 or 6.

8. A respiration rate measurement system comprising: the respiration rate measurement device according to any one of claims 1 to 7; and a radar transmitting and receiving device for the FMCW radar.

9. A respiration rate measurement program for causing a computer to execute each processing step performed by the beat signal acquisition unit, the differential signal calculation unit, the frequency spectrum calculation unit, the frequency integrated spectrum calculation unit, and the absence determination unit provided in the respiration rate measurement device according to any one of claims 1 to 7.

10. A respiration rate measurement method comprising: executing each processing step performed by the beat signal acquisition unit, the differential signal calculation unit, the frequency spectrum calculation unit, the frequency integrated spectrum calculation unit, and the absence determination unit provided in the respiration rate measurement device according to any one of claims 1 to 7.

Images