JP2012205693A - Device, method, and program for analyzing breath sound - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a device, method, and program for analyzing breath sound that has enhanced accuracy for detecting abnormal sound such as aspiration.SOLUTION: The device 11 for analyzing breath sound includes: a breath sound information acquiring part 31 for acquiring breath sound information presenting continuous breath sounds made by a subject for a predetermined period; a breath sound extracting part 35 for extracting each breath sound from the continuous breath sounds based on the breath sound information acquired by the breath sound acquisition part 31; a suspicious breath sound detecting part 36 for extracting one breath sound or a plurality of breath sounds as a reference from the breath sounds extracted by the breath sound extracting part 35, and detecting, based on the reference breath sound, a suspicious breath sound suspected of an abnormal breath sound made by the subject when not in a normal condition.

Description

    The present invention relates to a respiratory sound analysis device for detecting at least one abnormality among aspiration, laryngeal invasion, pharyngeal residue, and pharyngeal retention that occurs in a sleeping subject based on a respiratory sound emitted by the subject, The present invention relates to a sound analysis method and a respiratory sound analysis program.

  FIG. 7 is a side sectional view of a human body showing an organ through which food, saliva, and the like pass.

  When food or saliva is swallowed from the oral cavity 91 into the pharynx 92, an operation in the pharyngeal phase of swallowing is attempted to send the food or saliva or the like to the esophagus 93 via the pharynx 92. At this time, an elderly person whose swallowing function is lowered tends to cause aspiration in which food, saliva, and the like enter the trachea 95 beyond the glottis 94. In addition, food or saliva enters the pharynx 92 despite the laryngeal invasion and swallowing operation in the pharyngeal phase where food and saliva enter even before the glottis 94 even without aspiration. The pharynx remains temporarily stagnant, and the pharyngeal pool in which food, saliva and the like temporarily stagnate without performing the pharyngeal movement of swallowing may occur. Laryngeal intrusion, pharyngeal residue, and pharyngeal retention can lead to aspiration. Since repeated aspiration may cause aspiration pneumonia, it is necessary to examine the presence or absence of aspiration and to take appropriate measures if aspiration is likely to occur. People who experience laryngeal invasion, pharyngeal residue, and pharyngeal retention also need to be appropriately addressed early.

  Conventionally, as a method for examining the presence or absence of at least one abnormality (hereinafter referred to as abnormality such as aspiration) among aspiration, laryngeal invasion, pharyngeal residue, or pharyngeal retention, swallowing contrast examination, swallowing endoscopy, neck Department of auscultation is known. The swallowing contrast examination is an examination in which the subject actually eats the suspected food mixed with barium while the X-ray is applied near the pharynx 92 and observes the flow of the suspected food eaten by the subject. Swallowing endoscopy is a test in which the flow of food eaten by a subject is observed using a fiberscope. These examinations are effective as a method for examining abnormalities such as aspiration during meals. Further, cervical auscultation is a method of auscultating a swallowing sound generated during swallowing and breathing sounds before and after swallowing in the cervix to determine abnormalities such as aspiration. Although this neck auscultation is slightly less accurate than the above two methods, it can be used to check not only food but also abnormalities such as aspiration caused by saliva, and can be easily tested. Is the method.

  However, none of these inspection methods can be adopted as a method for inspecting aspiration during sleep. In other words, abnormalities such as aspiration during sleep occur mainly when saliva enters the trachea 95, but saliva cannot be imaged by swallowing contrast examination. In addition, swallowing contrast examinations must be performed in a facility with X-ray equipment, and subjects may be exposed to exposure if performed for a long time. It cannot be adopted as an inspection method. The swallowing endoscopy is an inspection performed by inserting a fiberscope from the subject's nose, and is accompanied by the pain of the subject, so that it is difficult to inspect while the subject is sleeping. Although the neck auscultation is less burdensome on the subject, the doctor will follow the subject for a long time while the subject is sleeping. For this reason, a doctor's burden becomes too large and implementation is difficult.

  Further, a diagnostic system described in Patent Document 1 is known as a diagnostic system that automatically detects aspiration during sleep. This diagnostic system acquires a swallowing sound of a subject who is sleeping, and determines that aspiration has occurred when the number of times that the swallowing sound is generated is smaller than that of a healthy person. If this diagnostic system is used, a doctor's burden can be reduced.

WO2007 / 108322

  However, even if a swallowing operation that generates a swallowing sound is performed, aspiration may occur as a result. Conversely, even if no swallowing sound is generated, saliva may normally pass through the esophagus and reach the stomach. For this reason, it can not be concluded that aspiration is generally occurring just because the number of occurrences of swallowing sounds is small, and the diagnosis system of Patent Document 1 cannot detect aspiration or mistakenly mistakenly. Detects swallowing more often. Moreover, since the presence or absence of laryngeal invasion, pharyngeal residue, or pharyngeal accumulation may not coincide with the presence or absence of swallowing sound, the diagnosis system of Patent Document 1 may not detect them.

  In view of the above circumstances, an object of the present invention is to provide a respiratory sound analysis device, a respiratory sound analysis method, and a respiratory sound analysis program that improve the detection accuracy of abnormalities such as aspiration.

The respiratory sound analysis apparatus of the present invention that achieves the above-described object is characterized in that at least one abnormality among aspiration, laryngeal invasion, pharyngeal residue, and pharyngeal storage that occurs in a sleeping subject is detected by the respiratory sound generated by the subject. In the respiratory sound analysis device that detects based on
A respiratory sound information acquisition unit that acquires respiratory sound information representing continuous respiratory sounds emitted by the subject during a predetermined period;
Based on the breathing sound information acquired by the breathing sound information acquisition unit, a breathing sound extraction unit that extracts each breathing sound of each continuous breathing sound;
One or a plurality of reference exhalation sounds serving as a reference are extracted from the exhalation sounds extracted by the exhalation sound extraction unit, and an abnormal exhalation sound and a suspected exhalation sound generated in the event of the abnormality are based on the reference exhalation sounds. And a suspicious breath detection unit for detection.

  According to the respiratory sound analysis apparatus of the present invention, the expiration sound that is likely to change in the case of the abnormality is extracted from the respiratory sound, and the expiration sound is compared with the reference expiration sound, thereby causing an abnormality such as aspiration. Since an expiratory sound suspected of changing is detected, the abnormality can be detected with high accuracy.

  The suspected breath sound detection unit may update the reference breath sound and detect the suspected breath sound based on the updated reference breath sound.

  The suspected breath sound detecting unit adds a breath sound determined to be not the suspected breath sound from the breath sounds extracted by the breath sound extracting unit to the reference breath sound and generates a predetermined breath from the reference breath sound. The sound may be removed and updated as a new reference expiratory sound, and the suspected expiratory sound may be detected based on the updated reference expiratory sound.

  The reference exhalation sound includes, for example, a predetermined exhalation sound that is not accompanied by an operation in the pharyngeal phase of swallowing in addition to an exhalation sound that is generated when it is assumed that the operation in the pharyngeal phase of swallowing is performed well. included. On the other hand, the suspicious exhalation sound includes an exhalation sound when an abnormality accompanying the movement in the pharyngeal phase of swallowing and an expiratory sound when an abnormality occurs without an operation in the pharyngeal period of swallowing.

  In the respiratory sound analysis apparatus of the present invention, the suspected breath sound detecting unit detects the suspected breath sound based on a ratio of a frequency component in a predetermined frequency range including a predetermined frequency occupied in the exhaled sound. It is characterized by that.

  The respiratory sound analysis apparatus of the present invention detects the suspicious breath sound by using the characteristic found by the present inventor that the ratio of the frequency components in the predetermined frequency range changes when the abnormality occurs. Sound can be detected with high accuracy.

  In the respiratory sound analysis apparatus of the present invention, the suspected breath sound detecting unit detects the suspected breath sound based on a ratio of a value representing the sound pressure of the expiratory sound to a value representing the sound pressure of the reference expiratory sound. It is characterized by being.

  The respiratory sound analysis apparatus according to the present invention detects the suspicious breath sound with high accuracy because the ratio has been detected by the inventor using the characteristic that changes when the abnormality occurs. can do.

  The value representing the sound pressure of the reference expiratory sound and the value representing the sound pressure of the expiratory sound may be a root mean square value of a value representing the sound pressure per predetermined time in a predetermined period, or an arithmetic average It may be a value.

  In the respiratory sound analysis device of the present invention, the suspected breath sound detection unit includes a value representing a sound pressure of the breath sound and a frequency in a second frequency range occupying a frequency component of the first frequency range in the breath sound. The suspicious breath sound is detected by using a ratio of a component and a ratio of a frequency component in a third frequency range different from the second frequency range in a frequency component in the first frequency range. It is characterized by being.

  The respiratory sound analysis apparatus of the present invention uses a characteristic that the inventor has found that the ratio of the frequency components in the second frequency range and the ratio of the frequency components in the third frequency range change when abnormal. Since the suspected breath sound is detected, the suspected breath sound can be detected with higher accuracy.

  The third frequency range may be a higher frequency range than the second frequency range. Further, the third frequency range may be a range including a higher frequency than the highest frequency included in the second frequency range, and may be higher than the lowest frequency included in the second frequency range. It may be a range where the frequency is the lowest frequency.

  In the respiratory sound analysis apparatus of the present invention, the suspected breath sound detecting unit is

However, Lk is the ratio of the frequency component of the second frequency range in the frequency component of the first frequency range in the kth expiratory sound in the expiratory sound to be examined, and Hk is the examination target. The ratio of the frequency component of the third frequency range to the frequency component of the first frequency range in the kth expiratory sound in the expiratory sound, Sk is k in the expiratory sound to be examined A value representing the sound pressure in the th exhalation sound, Ln is a ratio of the frequency component in the second frequency range in the frequency component in the first frequency range in the n th reference expiratory sound, and Hn is n The ratio of the frequency component of the third frequency range to the frequency component of the first frequency range of the first reference expiratory sound, Sn is a value representing the sound pressure of the nth reference expiratory sound, and k and n are positive Integer, m is the reference call Is the number of sound.
A suspicious sound is detected by comparing the calculated value obtained by the above formula (1) with a predetermined reference value.

  The inventor has a ratio of frequency components in the second frequency range that occupy the breath sound increases in the abnormality, a ratio of frequency components in the third frequency range decreases in the abnormality, A characteristic has been found in which the value representing the sound pressure increases when the abnormality occurs. The respiratory sound analysis apparatus of the present invention uses these characteristics, multiplies the increasing rate and value, and divides by the decreasing rate to make the characteristic at the time of abnormality more obvious and detect the suspicious breath sound. Therefore, the suspicious sound can be detected with higher accuracy.

  In the respiratory sound analyzing apparatus of the present invention, the suspected expiratory sound detecting unit is a predetermined part in the first half of the expiratory period from the start of the expiratory sound to the end of the expiratory sound among the expiratory sounds extracted by the expiratory sound extracting unit. A suspicious breath sound is detected using a sound generated during a period.

  The respiratory sound analysis apparatus of the present invention detects the suspicious breath sound with high accuracy because it detects using the sound generated in the first half of the expiratory sound, which changes most remarkably at the time of the abnormality. .

In the respiratory sound analysis device of the present invention, a change information acquisition unit that acquires change information representing a change based on the breathing motion of the sleeping subject,
A storage unit that stores respiratory sound information and change information in association with the respiratory sound information acquired by the respiratory sound information acquisition unit and the change information acquired by the change information acquisition unit;
The exhalation sound extraction unit extracts an exhalation sound based on respiratory sound information and change information stored in the storage unit.

  Since the breathing sound analysis apparatus of the present invention extracts the breathing sound based on the two information of the breathing sound information and the change information, the breathing sound can be accurately extracted from the breathing sound.

The respiratory sound analysis method of the present invention that achieves the above-described object is characterized in that at least one abnormality among aspiration, laryngeal invasion, pharyngeal residue, and pharyngeal retention that occurs in a sleeping subject is detected in the respiratory sound that the subject has emitted. In the respiratory sound analysis method to detect based on
A breathing sound information acquisition step for acquiring breathing sound information representing a continuous breathing sound emitted by the subject during a predetermined period;
Based on the breathing sound information acquired by the breathing sound information acquisition unit, a breathing sound extraction step of extracting each breathing sound of each continuous breathing sound;
One or a plurality of reference exhalation sounds serving as a reference are extracted from the exhalation sounds extracted by the exhalation sound extraction unit, and an abnormal exhalation sound and a suspected exhalation sound generated in the event of the abnormality are based on the reference exhalation sounds. A suspicious breath detection step for detecting.

  By using the respiratory sound analysis method of the present invention in the respiratory sound analysis apparatus, it is possible to realize a respiratory sound analysis apparatus that detects the abnormality with high accuracy.

The respiratory sound analysis program of the present invention that achieves the above object is characterized in that at least one abnormality among aspiration, laryngeal invasion, pharyngeal residue, and pharyngeal retention that occurs in a sleeping subject is detected in the respiratory sound generated by the subject. A respiratory sound analysis program used in a computer that detects based on
A breathing sound information acquisition unit for acquiring breathing sound information representing a continuous breathing sound emitted by the subject during a predetermined period;
Based on the respiratory sound information acquired by the respiratory sound information acquisition unit, the exhalation sound extraction unit that extracts each exhalation sound of each continuous respiratory sound, and
One or a plurality of reference exhalation sounds serving as a reference are extracted from the exhalation sounds extracted by the exhalation sound extraction unit, and an abnormal exhalation sound and a suspected exhalation sound generated in the event of the abnormality are based on the reference exhalation sounds. Suspicious breath detection unit to detect,
Is a respiratory sound analysis program for causing a computer to function.

  By installing and executing the respiratory sound analysis program of the present invention on a computer, the computer can be operated as the respiratory sound analysis apparatus of the present invention.

  According to the present invention, it is possible to provide a breathing sound analysis apparatus, a breathing sound analysis method, and a breathing sound analysis program with improved detection accuracy of abnormalities such as aspiration.

It is a hardware block diagram which shows an example of the respiratory sound analysis system 1 containing one Embodiment of the respiratory sound analyzer of this invention. It is a functional block diagram of the respiratory sound analysis apparatus 11 as an example of the present invention realized by installing and executing a respiratory sound analysis program in the computer 10. It is a conceptual diagram for demonstrating the extraction operation | movement of the exhalation sound by the exhalation sound extraction part. (A) is an example of the initial screen of the respiratory sound analyzer displayed on the display device, and (b) is an example of a display screen displayed after the analysis is completed. It is an example of the display screen showing the waveform of the breathing sound displayed when the sequential playback button is pressed. It is a flowchart showing an example of operation | movement of a respiratory sound analyzer. It is side sectional drawing of the human body which shows the organ through which food, saliva, etc. pass.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

  FIG. 1 is a hardware configuration diagram showing an example of a respiratory sound analysis system 1 including an embodiment of the respiratory sound analysis apparatus of the present invention.

  The respiratory sound analysis system 1 shown in FIG. 1 includes a computer 10, a data logger 51, a microphone 52, and a respiratory band sensor 53 that operate as an embodiment of the respiratory sound analysis apparatus of the present invention. The data logger 51 is connected to each of the microphone 52 and the breathing band sensor 53 via a communication cable. The microphone 52 converts sound into an electric signal and transmits the electric signal to the data logger 51. The breathing band sensor 53 is a sensor that converts band elongation into an electrical signal. The data logger 51 is connected to the computer 10 by a communication cable. Each electrical signal transmitted from the microphone 52 and the breathing band sensor 53 to the data logger 51 is stored in a memory provided in the data logger 51 and is transmitted from the data logger 51 to the computer 10.

  Inside the computer 10 are a CPU (Central Processing Unit) 12 for executing various programs, a main memory 13 comprising a RAM (Random Access Memory), a hard disk device 14 in which various programs and information are stored, and instructions from the CPU 12 Accordingly, an external device such as a display device 15 for displaying images and characters on the display screen, an input device 16 such as a keyboard and a mouse, an optical drive device 17 for reading a program recorded on the optical disc 61, and a data logger 51 are connected. An external I / F (interface) 18, a sound output device 19 such as a speaker or headphones, and a timer 20 are provided. These various elements are connected to each other via a bus.

  The optical disc 61 stores a respiratory sound analysis program that causes the computer 10 to operate as a respiratory sound analyzer. By loading this optical disc 61 into the optical drive device 17, the respiratory sound analysis program stored in the optical disc 61 is installed in the computer 10. Note that the medium for storing the breathing sound analysis program may not be the optical disc 61 but may be another recording medium such as a memory card. When the respiratory sound analysis program is stored in another recording medium such as a memory card, the respiratory sound analysis program is installed in the computer 10 by connecting a reading device that reads the recording medium to the external I / F 18. .

  FIG. 2 is a functional block diagram of a respiratory sound analysis apparatus 11 as an example of the present invention, which is realized by installing and executing a respiratory sound analysis program in the computer 10.

  In the respiratory sound analysis device 11, when the CPU 12 reads and executes the installed respiratory sound analysis program, at least one abnormality among aspiration, laryngeal invasion, pharyngeal residue, and pharyngeal retention that has occurred in a sleeping subject is detected. (Hereinafter, referred to as an abnormality such as aspiration) is a device that detects a breath sound generated by a subject. The respiratory sound analysis device 11 includes a display device 15, an input device 16, an audio output device 19, a respiratory sound information acquisition unit 31, a change information acquisition unit 32, an information management unit 33, a storage unit 34, an exhalation sound extraction unit 35, and The suspicious breath detection unit 36 is configured. Of these, the respiratory sound information acquisition unit 31, the change information acquisition unit 32, the information management unit 33, the exhalation sound extraction unit 35, and the suspected exhalation sound detection unit 36 are functional configurations realized by the CPU 12. In addition, a data logger 51 is connected to the respiratory sound analyzer 11, and a microphone 52 and a respiratory band sensor 53 are connected to the data logger 51.

  The microphone 52 is affixed to the outer skin of the trachea directly below the cricoid cartilage of the subject, thereby capturing the respiratory sound when the subject is sleeping and converting the respiratory sound into an electrical signal. The electrical signal converted by the microphone 52 is transmitted to the data logger 51 and further transmitted from the data logger 51 to the respiratory sound analyzer 11. The breathing band sensor 53 is attached, for example, so as to wrap the subject's chest, and measures changes in expansion and contraction of the chest associated with the breathing motion, and converts the changes into electrical signals. Changes in the expansion and contraction of the chest in the present embodiment correspond to an example of changes based on the breathing motion of the present invention. The electrical signal converted by the breathing band sensor 53 is transmitted to the data logger 51, and the electrical signal is further amplified by the data logger 51 and transmitted to the respiratory sound analyzer 11. In this embodiment, the breathing band sensor 53 is used. However, as long as it is a sensor that measures a change based on the breathing motion of the subject, for example, a sensor that is attached to the nostril and detects a change in the air flow due to breathing, etc. These sensors may be used.

  The respiratory sound information acquisition unit 31 receives the electrical signal transmitted from the microphone 52 via the data logger 51, and acquires respiratory sound information representing the respiratory sound based on the received electrical signal. When acquiring respiratory sound information, the respiratory sound information acquiring unit 31 stores the respiratory sound information in the main memory 13 while adding the time information obtained from the timer 20 shown in FIG. 1 to the respiratory sound information. The change information acquisition unit 32 shown in FIG. 2 receives the electrical signal transmitted from the breathing band sensor 53 via the data logger 51, and based on the received electrical signal, the change indicating the change in the expansion and contraction of the chest. Get information. When acquiring the change information, the change information acquisition unit 32 stores the change information in the main memory 13 while adding the time information obtained from the timer 20 illustrated in FIG. 1 to the change information. After the acquisition of the respiratory sound information and the acquisition of the change information are completed, when the operator of the respiratory sound analysis device 11 performs a predetermined input using the input device 16, the information management unit 33 changes the acquired respiratory sound information and the change. The storage date and time are written in each piece of information and stored in the storage unit 34 as corresponding information. Note that the breathing sound information and the change information may be merged by the information management unit 33 and stored in the storage unit 34 as one piece of information.

  The information management unit 33 manages information transmitted from each component of the respiratory sound analyzer 11, appropriately retrieves necessary information from the storage unit 34, and transmits the necessary information to each component constituting the respiratory sound analyzer 11, In addition, information that needs to be stored is stored in the storage unit 34. The storage unit 34 stores various types of information based on instructions from the information management unit 33. The specific hardware corresponds to the hard disk device 14 shown in FIG. Note that the main memory 13 is a temporary memory that is used as necessary by each component constituting the respiratory sound analyzer 11 when temporarily storing information or storing a calculation process. The expiratory sound extraction unit 35 shown in FIG. 2 extracts an expiratory sound from the respiratory sound based on the respiratory sound information and the change information. As a result of researches by the present inventor, it has been found that when an abnormality such as aspiration occurs, the exhalation sound changes in particular. Therefore, in the respiratory sound analysis apparatus 11 of the present embodiment, an exhalation sound is extracted from the respiratory sound, and an abnormality such as aspiration is detected by analyzing the exhalation sound.

  FIG. 3 is a conceptual diagram for explaining the extraction operation of the expiration sound by the expiration sound extraction unit 35.

  The exhalation sound extraction unit 35 extracts an exhalation sound by an extraction operation described with reference to FIG. The upper graph shown in FIG. 3 shows a waveform W1 representing a respiratory sound and a waveform W2 representing a value based on the sound pressure of the respiratory sound obtained by processing the waveform W1 with a bandpass filter that passes 300 Hz to 500 Hz. ing. The reason why the waveform W1 is processed by the band-pass filter is that it is easy to extract the expiration sound in the frequency range of 300 Hz to 500 Hz. The lower graph shown in FIG. 3 shows a waveform W3 representing changes in the expansion and contraction of the chest. The vertical axis of the waveform W3 represents the amount of expansion of the chest, and becomes a high value when the chest is inflated and becomes a low value when the chest is contracted. Note that the waveform W3 and the waveform W2 are waveforms that have been subjected to smoothing processing by taking a section average. The horizontal axis of the graph shown in FIG. 3 is a time axis representing the time flowing from left to right. The exhalation sound extraction unit 35 matches the time axis of each waveform from the time information written in each of the respiration sound information and the change information, and performs the exhalation sound extraction process in order from the respiration sound with the earlier generation timing. Extract.

  The exhalation sound extraction unit 35 first detects the peak time P of the waveform W3 for each breath. Here, the detection operation of the peak time P will be described. The exhalation sound extraction unit 35 searches for a temporary peak time during which the chest changes from expansion to contraction in the waveform W3. When the temporary peak time is found, a search is made for another temporary peak time during which the chest changes from expansion to contraction for a period of 1.5 seconds before and after that, and if not, the temporary peak time is peaked. Determine as time P. In addition, even if another provisional peak time exists, if the time when the chest is inflated is the provisional peak time among the provisional peak time and another provisional peak time, the provisional peak time is set to the peak time P. Determine as. If another tentative peak time expands the chest more than the tentative peak time, another tentative peak time is used as a new tentative peak time, and a 1.5 second search is performed again. Repeat this process until it is determined.

  Next, the expiratory sound extraction unit 35 searches for a time B (a time when the value based on the sound pressure is the smallest value) in which the minimum value of the waveform W2 exists in a period of 0.2 seconds before and after the peak time P. Then, the time B is determined as the start time of the exhalation sound. Since the switching between the inspiratory sound and the expiratory sound occurs immediately before the chest starts to contract, the start time of the expiratory sound can be found by searching for the time when the sound pressure of the respiratory sound is weakest immediately before and after the peak time P. it can. Finally, the search is performed in the direction of time passage of the waveform W2 starting from the time B, and a time E having the same value as the time B (the value based on the sound pressure is the same value as the time B) is found and the expiration is performed. Determined as the end time of the sound. However, if there is no time that is the same value within 3 seconds after time B, the time when the value based on the sound pressure is the smallest value within 3 seconds is determined as time E. Since the sound pressure is almost the same at the start and end of expiration, the expiration time ends by searching for the first time after which the value based on the sound pressure becomes the same value as time B after time B. Can find out. However, since it is normally not considered that exhalation lasts longer than 3 seconds, if there is no time within 3 seconds that is the same as the value based on the sound pressure of time B, the sound pressure is adjusted within that 3 seconds. The time when the base value becomes the smallest is determined as the expiration time of the expiration sound. By repeating the above extraction operation, the exhalation sound extraction unit 35 extracts an exhalation sound from all breathing sounds. When extraction of exhalation sounds for all breath sounds is completed, the exhalation sound extraction unit 35 assigns numbers sequentially from the exhalation sounds that the subject first emitted, and numbers and start times (time B) of all the exhalation sounds extracted. The end time (time E) and the time from time B to time E (expiration period T) are transmitted to the information management unit 33. The information management unit 33 causes the storage unit 34 to store information regarding the number, start time, end time, and expiration period T transmitted from the expiration sound extraction unit 35.

  The suspicious expiratory sound detection unit 36 receives information on the number, start time, end time, and expiratory period T of the expiratory sound from the information management unit 33, extracts a reference expiratory sound as a reference from the expiratory sound, An abnormal exhalation sound and a suspicious suspected exhalation sound generated in the case of an abnormality such as swallowing are detected based on the reference exhalation sound. As a result of research conducted by the present inventor, it has been found that an abnormal exhalation sound tends to have a stronger sound pressure than an exhalation sound in which an abnormality such as aspiration has not occurred (hereinafter referred to as a normal exhalation sound). did. The abnormal expiratory sound is a normal expiratory sound in which the ratio of frequency components in the frequency range greater than 0 Hz and less than or equal to 100 Hz occupying the frequency components in the frequency range greater than 0 Hz and less than or equal to 1000 Hz (hereinafter referred to as low frequency component ratio). It turned out that there is a tendency to increase. Furthermore, the abnormal expiratory sound is a ratio of frequency components in a frequency range of 450 Hz to 800 Hz (hereinafter referred to as a high frequency component ratio) in a frequency component in a frequency range of greater than 0 Hz and equal to or less than 1000 Hz (hereinafter referred to as a reference frequency range). It has been found that there is a tendency to decrease with respect to normal breath sounds. It has also been found that these trends are particularly prominent in women. The frequency range greater than 0 Hz and less than or equal to 1000 Hz in the present embodiment corresponds to an example of the first frequency range of the present invention, and the frequency range greater than 0 Hz and less than or equal to 100 Hz in the present embodiment is the second frequency range of the present invention. The frequency range of 450 Hz to 800 Hz in this embodiment corresponds to an example of the third frequency range of the present invention. The suspected breath sound detection unit 36 of the present embodiment detects abnormal breath sounds and suspected breath sounds based on these tendencies. Note that the reference frequency range may be set as appropriate according to the performance of the measuring device or the like, and may be a frequency range greater than 0 Hz and less than or equal to 2000 Hz, for example. Further, the low frequency component ratio may be a frequency component ratio in a frequency range including a frequency higher than 100 Hz, for example, as long as it is a ratio of a frequency component including a part of a frequency range higher than 0 Hz and not higher than 100 Hz. Further, the high frequency component ratio may be a ratio of frequency components in a frequency range including a frequency less than 450 Hz or higher than 800 Hz as long as it is a ratio of frequency components including a part of a frequency range of 450 Hz to 800 Hz. . In addition, when applied to a subject having a specific disease, the low frequency component ratio may be a ratio of frequency components in a frequency range other than a frequency range greater than 0 Hz and not greater than 100 Hz, and a high frequency component ratio is also 450 Hz or more. It is good also as a ratio of the frequency component of frequency ranges other than the frequency range of 800 Hz or less. For example, the low frequency component ratio may be a frequency component ratio in a frequency range of 300 Hz to 400 Hz, and the high frequency component ratio may be a frequency component ratio in a frequency range of 250 Hz to 450 Hz. For example, the low frequency component ratio may be a ratio of frequency components in a frequency range of 250 Hz to 400 Hz, and the high frequency component ratio may be a ratio of frequency components in a frequency range of 300 Hz to 450 Hz.

  Further, as a result of research by the present inventor, it has been found that the above-mentioned tendency appears most significantly in the sound generated during the first half of the expiration period T from the start of the expiration sound to the end of the expiration sound. Therefore, the suspicious breath detection unit 36 of the respiratory sound analysis apparatus 11 of the present embodiment detects the sigh breath generated during the period from the start of the expiration period T to 25% of the expiration period T when detecting the suspicious expiration sound. Various calculations are performed. However, if the sound is generated during the first half of the exhalation sound, it may not be 25% from the beginning. For example, various calculations are performed using the exhalation sound generated during the period from 25% to 50%. You may go. Further, although the detection accuracy of the suspicious exhalation sound is lowered, various calculations may be performed using the entire exhalation sound.

Here, the extraction operation of the reference expiration sound performed by the suspicious expiration sound detection unit 36 will be described. The suspicious exhalation sound detection unit 36 is a predetermined number of times from the exhalation sound extracted by the exhalation sound extraction unit 35 to the exhalation sound that the subject first emitted to the exhalation sound (here, the seventh time). Then, 7) exhalation sounds are extracted as temporary reference exhalation sounds. In the present embodiment, seven expiratory sounds are extracted as temporary reference expiratory sounds and reference expiratory sounds to be described later, but any number of temporary expiratory sounds can be extracted as temporary reference expiratory sounds and reference expiratory sounds. . However, in order to increase detection accuracy, it is preferable to use five or more exhalation sounds. The suspicious expiratory sound detection unit 36 calculates a low frequency component ratio L _n ′, a high frequency component ratio H _n ′, and a value S _n ′ representing sound pressure in each of the seven temporary reference expiratory sounds. In addition, each frequency component ratio can be calculated | required by using a FFT (Fast Fourier Transform) analyzer. In this embodiment, the ratio of the frequency component in the frequency range for every 50 Hz that occupies the frequency component in the frequency range of greater than 0 Hz and 1000 Hz or less is obtained using an FFT analyzer. And the ratio of the frequency component of the frequency range for every 50 Hz is added, and the low frequency component ratio and the high frequency component ratio are calculated. In addition, the value representing the sound pressure is an arithmetic average obtained by squaring the value representing the sound pressure every 1/16000 sec in the period from the start of the expiration period T to 25% of the expiration period T, and the square average taking the square root. Square root value. The value representing the sound pressure of the reference exhalation sound may be a simple arithmetic mean value, but using the root mean square value has the advantage that the difference between the exhalation sound and the suspected exhalation sound becomes clearer. .

  The suspicious expiratory sound detection unit 36 obtains, for each of the seven provisional reference expiratory sounds, the provisional calculation value obtained by the following equation (a) in order from the expiratory sound that the subject first emitted.

However, L _n ′ is a low frequency component ratio in the n th exhalation sound of the seven temporary reference exhalation sounds, and H _n ′ is a high frequency in the n th exhalation sound of the seven temporary reference exhalation sounds. The frequency component ratio, S _n ′, is a value representing the sound pressure in the nth expiratory sound of the seven temporary reference expiratory sounds, n is a positive integer, and m is the number of temporary reference expiratory sounds. 7 in form.

If all the temporary calculated values obtained by the formula (a) are less than a predetermined value (2.08 in this case), the seven exhalation sounds from the first exhalation sound that the subject emitted to the seventh exhalation sound are obtained. Set as reference breath sound. If there is an expiratory sound in which the temporary calculated value obtained by the equation (a) is 2.08 or higher in the temporary reference expiratory sound, the expiratory sound that is 2.08 or higher is set as the temporary reference expiratory sound. In addition, the seven exhalation sounds that are newly added to the exhalation sound that is next to the exhalation sound that the test subject uttered most recently among the exhalation sounds used as the temporary reference exhalation sound are defined as new temporary reference exhalation sounds. To do. Then, a low frequency component ratio L _n ′, a high frequency component ratio H _n ′, and a value S _n ′ representing the sound pressure in the new temporary reference exhalation sound are newly calculated, and using the formula (a), the reference The temporary calculated value is continuously obtained by the equation (a) from the exhaled sound that exists after the exhaled sound that is excluded from the exhaled sound (the temporary calculated value becomes 2.08 or more). If an exhalation sound in which all the provisional calculated values of the seven temporary reference exhalation sounds are less than 2.08 can be extracted, the seven exhalation sounds are determined as the reference exhalation sounds.

  When the reference expiratory sound is determined, the suspicious expiratory sound detection unit 36 performs the suspicious expiratory sound detection operation in order from the expiratory sound (k = 1) first emitted by the subject. The suspected expiratory sound detection unit 36 obtains a calculated value obtained by the following expression (b) for each expiratory sound for each expiratory sound.

However, Lk is the low frequency component ratio in the exhalation sound emitted by the subject k times, Hk is the high frequency component ratio in the exhalation sound emitted by the subject k times, Sk is the exhalation sound emitted by the subject k times Ln is a value representing the sound pressure at L, n is the low frequency component ratio of the nth expiratory sound of the seven reference expiratory sounds, and Hn is the high frequency component ratio of the nth expiratory sound of the seven reference expiratory sounds , Sn is a value representing the sound pressure in the nth expiratory sound among the seven reference expiratory sounds, k and n are positive integers, m is the number of reference expiratory sounds, and is 7 in this embodiment.

  The suspected expiratory sound detection unit 36 is less than a reference value (here, 2.08) where the calculated value of the expiratory sound obtained by the equation (b) is the same value as the predetermined value used when obtaining the temporary reference expiratory sound. The exhalation sound which becomes becomes the normal exhalation sound. On the other hand, an expiratory sound having a calculated value equal to or greater than the reference value of 2.08 is determined as a suspected expiratory sound. If the calculated value is less than 2.08 from the first expiratory sound emitted by the subject to the expiratory sound emitted most recently by the subject among the reference expiratory sounds, the calculated values are sequentially obtained for the next expiratory sound. . If there is an expiratory sound determined to be a suspected expiratory sound in the reference expiratory sound, the expiratory sound determined to be the suspected expiratory sound is removed from the reference expiratory sound, and the subject emitted the latest among the reference expiratory sounds Seven expiratory sounds obtained by newly adding the expiratory sound that exists next to the expiratory sound are set as new reference expiratory sounds. Then, using the new reference expiratory sound, the calculated value is continuously obtained by the equation (b) from the expiratory sound that exists after the exhaled sound that is removed from the reference expiratory sound (the calculated value becomes 2.08 or more). Go.

  From the expiratory sound next to the expiratory sound that the subject most recently uttered among the reference expiratory sounds initially determined using the formula (a), it is determined that the expiratory sound after the next expiratory sound is a normal expiratory sound In this case, the expiratory sound is newly added to the reference expiratory sound, and the seven new expiratory sounds obtained by removing the earliest (old) expiratory sound from the reference expiratory sound among the reference expiratory sounds from the reference expiratory sound. The calculated value is obtained as a sound by the equation (b). In other words, the reference expiratory sound is composed of normal expiratory sounds that have been generated in the near future by sequentially adding the expiratory sound determined to be a normal expiratory sound to the reference expiratory sound and removing the old expiratory sound from the reference expiratory sound to update the reference expiratory sound in sequence. A suspected expiratory sound is detected by a calculated value using. By making the normal expiratory sound that has occurred in the near future a reference expiratory sound, the change in the expiratory sound appears more clearly as a calculated value, so that the detection accuracy of the suspicious expiratory sound can be increased. For example, even when the exhalation sound gradually changes with the passage of time since going to bed, since the reference exhalation sound is sequentially updated, an abnormality such as aspiration can be accurately detected. However, although the detection accuracy is lowered, all the expiratory sounds may be analyzed based on the reference expiratory sound once determined without adding the expiratory sound determined to be the normal expiratory sound to the reference expiratory sound.

  In addition, when it is determined that the expiratory sound is the suspicious expiratory sound from the expiratory sound next to the expiratory sound that the subject has most recently emitted among the reference expiratory sounds first determined using the formula (a) (suspected expiratory sound) If a sound is detected), the reference expiration sound is not updated. When the suspected expiratory sound is detected for all the expiratory sounds, the suspected expiratory sound detecting unit 36 transmits to the information managing unit 33 how many times the expiratory sound is the suspected expiratory sound. The information management unit 33 stores the information on the suspected breath sound transmitted from the suspected breath sound detecting unit 36 in the storage unit 34. In addition, the information management unit 33 causes the display device 15 to display information on the suspicious sound.

  FIG. 4A is an example of an initial screen of the respiratory sound analysis device 11 displayed on the display device 15.

  This initial screen is a screen that is displayed when the respiratory sound analyzer 11 is turned on. After the calculation by the suspicious breath sound calculation unit 33 is completed, a screen of FIG. 4B described later is displayed. Is done. The initial screen includes a file read button 71, a breath analysis button 72, a sequential playback button 73, an enlarged display button 74, a range specification display button 75, a comment display button 76, an exclusion specification button 77, a help button 78, and a respiratory sound playback order. And a reproduction condition designation unit 79 for designating the reproduction length, a suspected breath position display unit 81, an enlarged display unit 82, and a result display unit 83 representing the analysis result.

  The file read button 71 is an operation button for instructing to call a screen for designating respiratory sound information stored in the storage unit 34. When the operator of the respiratory sound analyzer 11 presses the file read button 71, an information selection screen (not shown) that displays the respiratory sound information stored in the storage unit 34 is called. The operator of the breathing sound analyzer 11 specifies one piece of breathing sound information on the information selection screen, and is stored in the storage unit 34 corresponding to the breathing sound information to be analyzed and the breathing sound information. Change information is specified. When the breathing sound information and the change information are not stored in the storage unit 34, the breathing sound information and the change information are acquired by the breathing sound information acquisition unit 31 and the change information acquisition unit 32, and then the storage unit 34. It is necessary to memorize respiratory sound information and change information. The breath analysis button 72 is an operation button for instructing the breathing sound analyzer 11 to start analysis of selected breathing sound information and change information. When the exhalation analysis button 72 is pressed, the exhalation sound extraction unit 35 and the suspected exhalation sound detection unit 36 shown in FIG. When the detection of the suspected breath sound is completed, the analysis result is displayed on the suspected breath position display unit 81 and the result display unit 83.

  FIG. 4B is an example of a display screen displayed after the analysis is completed.

  The horizontal axis of the suspected expiratory position display unit 81 is a time axis representing the time flowing from the left to the right. The suspected expiratory position display unit 81 displays a black vertical line in a portion where the suspected expiratory sound is present in the entire analyzed breathing sound, and displays a portion where a normal expiratory sound is present in white. In addition, the result display unit 83 displays the number of respiratory sounds, the number of detected suspicious sounds, and the average value (seconds) of the expiration time in the entire analyzed respiratory sounds.

  The sequential playback button 73 is an operation button for instructing the respiratory sound analyzer 11 to display the respiratory sound that has been analyzed. When the sequential playback button 73 is pressed, the breathing sound specified by the playback condition specifying unit 79 is displayed on the screen. The reproduction condition designation unit 79 includes a sort condition designation unit 791, a rank input unit 792, a number input unit 793, a reproduction respiratory rate designation unit 794, and a performance designation box 795.

  The sort condition designating portion 791 displays five radio buttons and one check box. The five radio buttons are for designating the display order of the waveform of the respiratory sound displayed when the playback button is sequentially pressed. The “None” radio button specifies that the waveform of the breathing sound is displayed in time order each time the sequential playback button is pressed. When “None” is designated, the waveform of the respiratory sound is displayed in order from the oldest respiratory sound every time the playback button is pressed. The radio button of “index” designates that the waveform of the respiratory sound is displayed in descending order of the calculated value calculated by the above formula (b) each time the playback button is sequentially pressed. The “strength” radio button designates that the waveform of the respiratory sound is displayed in order from the respiratory sound having the highest value Sk representing the sound pressure each time the playback button is sequentially pressed. Each time the “Frequency” radio button is sequentially pressed, the value obtained by dividing the low frequency component ratio Lk in the expiratory sound by the high frequency component ratio Hk in the expiratory sound, that is, the value of Lk ÷ Hk in each expiratory sound is calculated. This specifies that the waveform of the respiratory sound is displayed in order from the highest respiratory sound. The radio button for “strong cycling” divides the value obtained by multiplying the low-frequency component ratio Lk in the expiratory sound by the value Sk representing the sound pressure in the expiratory sound by the high-frequency component ratio Hk in the expiratory sound each time the play button is sequentially pressed. That is, it is designated that the waveform of the respiratory sound is displayed in order from the respiratory sound having the highest value of Lk × Sk ÷ Hk in each exhalation sound. The check box designates inversion of the order of sequential reproduction. When this check box is checked, every time the playback button is sequentially pressed, the waveform of the respiratory sound is displayed in the reverse order to the order specified by the radio button.

  The rank input unit 792 is a part for inputting the number of breathing sounds to be displayed when the breathing sounds are arranged in the order specified by the sort condition specifying unit 791. When a numerical value is input to this portion and the playback button is pressed sequentially, the waveforms of respiratory sounds that are in the input order are displayed in the order specified by the sort condition specifying unit 791. The number input unit 793 is a part that designates which number of breathing sounds to be displayed when the breathing sounds are arranged in time order in order from the exhalation sound that the subject first emitted. When a numerical value is input to this portion and the playback button is pressed sequentially, the waveform of the respiratory sound having the input order is displayed in time order regardless of the order specified in the sort condition specifying unit 791. The respiration rate designation section 794 is a portion that designates the respiration rate displayed when the replay button is sequentially pressed. The performance designation box 795 is a check box for designating output of the breathing sound from the sound output device 19 together with the display of the display screen showing the waveform of the breathing sound when the playback button 73 is sequentially pressed. .

  FIG. 5 is an example of a display screen showing the waveform of the breathing sound displayed when the playback button 73 is sequentially pressed.

  In the display screen shown in FIG. 5, the waveform of the suspicious exhalation sound is surrounded by a rectangle 731, and a hatched band mark 732 is displayed above the waveform of the normal exhalation sound. The numerical value described above the waveform of each breath sound is a numerical value indicating the ranking when the upper part is arranged in time order, and the lower part is a calculated value calculated by the above-described equation (b). For example, the waveform surrounded by the rectangle 731 shown in the center of FIG. 5 is the 1143th waveform, and the calculated value of the expiratory sound indicated by the waveform is 2.6.

  The enlarged display button 74 shown in FIGS. 4A and 4B is an operation button for designating that the expiratory sound selected in the suspected expiratory position display unit 81 is enlarged and displayed on the enlarged display unit 82. is there. The range designation display button 75 is an operation button for designating that the range selected in the suspicious breath position display unit 81 is displayed on the enlarged display unit 82. The comment display button 76 is an operation button for designating that a comment is given to the exhalation sound selected in the suspected expiratory position display unit 81 or that a comment already given is displayed. When the comment display button 76 is pressed, another screen for comment input (not shown) is displayed. The operator of the breathing sound analyzer 11 can enter a comment on the separate screen or edit the comment if there is a comment already given. The exclusion designation button 77 is an operation button for designating that even if the exhalation sound selected in the suspected exhalation position display unit 81 is an exhalation sound determined as the suspected exhalation sound, it is processed as a normal exhalation sound. For example, when the selected exhalation sound is snoring or sleeping, the operator of the breathing sound analyzer 11 can cancel the determination of the suspected exhalation sound for the exhalation sound by pressing the exclusion designation button 77. The help button 78 is an operation button for designating display of a help screen.

  Next, the operation of the respiratory sound analyzer 11 as an example of the present invention will be described with reference to FIG.

  FIG. 6 is a flowchart showing an example of the operation of the respiratory sound analyzer 11.

  When the breathing sound analyzer 11 is turned on, the CPU 12 reads and executes a breathing sound analysis program, and each component of the breathing sound analyzer 11 operates based on the breathing sound analysis program. When the breathing sound analysis program is executed, the information management unit 33 first displays the initial screen shown in FIG.

  When the electrical signals from the microphone 52 and the respiratory band sensor 53 are transmitted to the respiratory sound analyzer 11, the respiratory sound information acquisition unit 31 and the change information acquisition unit 32 are configured to transmit the respiratory sound information and the respiratory information based on the transmitted electrical signals. Change information is acquired (step S1). Then, in response to a predetermined input by the operator of the respiratory sound analysis device 11, the respiratory sound information acquisition unit 31 and the respiratory sound information acquisition unit 31 transmit the respiratory sound information and the change information to the information management unit 33, respectively. The information management unit 33 stores the respiratory sound information and the change information in the storage unit 34 in association with each other.

  In step S2, after the breathing sound information is designated by the operator of the breathing sound analysis apparatus 11, it is repeatedly executed until the breath analysis button 72 is pressed. When the exhalation analysis button 72 is pressed (Yes in step S2), the information management unit 33 reads the designated breathing sound information and change information from the storage unit 34 and transmits them to the exhalation sound extraction unit 35. The exhalation sound extraction unit 35 extracts an exhalation sound by the above-described extraction operation using the transmitted breathing sound information and change information (step S3). The exhalation sound extraction unit 35 transmits the extracted exhalation sound information to the information management unit 33. The information management unit 33 transmits the exhalation sound information to the suspicious exhalation sound detection unit 36 and causes the storage unit 34 to store the information. Next, the suspected expiratory sound detection unit 36 extracts the reference expiratory sound from the expiratory sound extracted by the expiratory sound extracting unit 35, and detects the suspected expiratory sound by the above-described detection operation (step S4). When the detection of the suspicious sound is completed, the suspicious sound detection unit 36 transmits the detected information to the information management unit 33. The information management unit 33 stores the information detected by the suspicious sound detection unit 36 in the storage unit 34 and displays the analysis result on the display device 15 based on the detected information (step S5). If the playback button 73 is not sequentially pressed on the display screen (the display screen shown in FIG. 4B) representing the analysis result (No in step S6), the process proceeds to step S10. When the sequential playback button 73 is pressed (Yes in step S6), whether or not the performance designation box 795 is checked is checked (step S7). If there is a check, the breathing sound is determined according to the condition designated by the reproduction condition designation unit 79. Is displayed on the display device 15 and the breathing sound is output from the sound output device 19 (step S8). If the performance designation box 795 is not checked, the breathing sound waveform is displayed on the display device 15 in accordance with the condition designated by the reproduction condition designation unit 79 (step S9). Note that the waveform of the breathing sound displayed in step S8 and step S9 is sequentially switched every time the playback button 73 is pressed in accordance with the order specified in the sort condition specifying unit 791. In step S10, if there is no end instruction, the process returns to step S6, and if there is an end instruction, the operation of the respiratory sound analyzer 11 is ended.

  According to the respiratory sound analysis apparatus 11 of the present embodiment, it is possible to detect an abnormal exhalation sound and a suspicious exhalation sound that are generated when an abnormality such as aspiration occurs. Further, even if the doctor does not accompany the sleeping subject, the doctor can diagnose the presence or absence of abnormalities such as aspiration occurring during sleeping in a short time by listening to the suspicious breath sound detected by the respiratory sound analyzer 11. can do. Further, since the reproduction order of the suspicious breath sounds can be designated by the reproduction condition designating unit 79, the doctor can easily select the respiratory sounds necessary for the diagnosis. Moreover, according to the respiratory sound analyzer 11 of this embodiment, not only aspiration but also laryngeal invasion, pharyngeal residue, and pharyngeal accumulation can be detected. That is, when at least one of these abnormalities occurs, food or saliva is present in a part of the airway, so that a part of the airway becomes narrow and the breathing sound changes. Since the respiratory sound analysis device 11 of the present embodiment analyzes this respiratory sound, it can detect not only aspiration but also laryngeal invasion, pharyngeal residue, and pharyngeal accumulation. Moreover, not only abnormalities such as aspiration accompanied by movement in the pharyngeal phase of swallowing but also abnormalities such as aspiration not accompanied by movement in the pharyngeal phase of swallowing can be detected. In addition, the respiratory sound analysis apparatus 11 of the present embodiment extracts an expiratory sound that is likely to change when an abnormality such as aspiration occurs from the respiratory sound, and compares the expiratory sound with a reference expiratory sound to detect an erroneous sound. Since an expiratory sound that is suspected of being changed due to an abnormality such as swallowing is detected, the detection accuracy of an abnormality such as aspiration can be increased. In addition, two elements that tend to show high values in abnormal exhalation sounds (value S representing the low-frequency component ratio L of expiratory sounds and sound pressure of expiratory sounds) are multiplied, and further elements that tend to show low values in abnormal exhalation sounds ( Since the suspected expiratory sound is detected using the value divided by the high frequency component ratio H) of the expiratory sound, an abnormality such as aspiration can be detected with high accuracy. In addition, according to the respiratory sound analysis apparatus 11 of the present embodiment, it is possible to know at what time during sleep the abnormality such as aspiration occurred, and therefore appropriate measures are taken for a person who is likely to cause an abnormality such as aspiration. Can do.

The present invention is not limited to the above-described embodiment, and various modifications can be made within the scope described in the claims. For example, in the above-described embodiment, the suspected breath sound detecting unit 36 detects the suspected breath sound with high detection accuracy by using the formula (b), but the average value of the values representing the sound pressure of the reference breath sound By comparing the ratio of the value Sk representing the sound pressure of the expiratory sound with respect to {Σ (S ₁ + S ₂ +... S _m ) ÷ m} with a predetermined reference value, the suspected expiratory sound may be detected easily. . Further, the ratio of the high frequency component ratio Hk of the expiratory sound to the average value {Σ (H ₁ + H ₂ +... H _m ) ÷ m} of the high frequency component ratio of the reference expiratory sound is compared with a predetermined reference value. A simple expiratory sound may be detected. Further, the ratio of the low frequency component ratio Lk of the expiratory sound to the low frequency component ratio {Σ (L ₁ + L ₂ +... L _m ) / m} of the reference expiratory sound is simply compared with a predetermined reference value. A suspected breath sound may be detected. Alternatively, the suspected expiratory sound may be detected using two of the value Sk representing the sound pressure of the expiratory sound, the high frequency component ratio Hk of the expiratory sound, and the low frequency component ratio Lk of the expiratory sound. Further, in the present embodiment, change information is used to improve the accuracy of extracting exhalation sounds from breath sounds, but exhalation sounds may be detected only from breath sounds.

11 Respiratory Sound Analysis Device 15 Display Device 16 Input Device 18 External I / F
DESCRIPTION OF SYMBOLS 19 Audio | voice output device 31 Respiratory sound information acquisition part 32 Change information acquisition part 33 Information management part 34 Storage part 35 Expiratory sound extraction part 36 Suspicious expiratory sound calculation part 51 Data logger 52 Microphone 53 Respiration band sensor

Claims (9)

In a respiratory sound analysis device that detects at least one abnormality among aspiration, laryngeal invasion, pharyngeal residue, and pharyngeal storage that occurred in a sleeping subject, based on the respiratory sound emitted by the subject,
A respiratory sound information acquisition unit that acquires respiratory sound information representing continuous respiratory sounds emitted by the subject during a predetermined period;
Based on the breathing sound information acquired by the breathing sound information acquisition unit, a breathing sound extraction unit that extracts each breathing sound of each continuous breathing sound;
One or a plurality of reference exhalation sounds serving as a reference are extracted from the exhalation sounds extracted by the exhalation sound extraction unit, and an abnormal exhalation sound and a suspected exhalation sound generated in the event of the abnormality are based on the reference exhalation sounds. A respiratory sound analysis device comprising a suspicious breath detection unit for detection.   The suspected breath sound detecting unit is configured to detect the suspected breath sound based on a ratio of a frequency component in a predetermined frequency range including a predetermined frequency occupied in the expiratory sound. The respiratory sound analysis device described.   The suspected expiratory sound detecting unit detects the suspected expiratory sound based on a ratio of a value representing a sound pressure of the expiratory sound to a value representing a sound pressure of the reference expiratory sound. Item 3. The respiratory sound analysis device according to item 1 or 2.   The suspicious expiratory sound detection unit includes a value representing a sound pressure of the expiratory sound, a ratio of a frequency component in a second frequency range to a frequency component in the first frequency range in the expiratory sound, and the first 2. The suspicious sound is detected using a ratio of a frequency component in a third frequency range different from the second frequency range in a frequency component in the frequency range. Respiratory sound analyzer described in 1. The suspicious breath detection unit is
However, Lk is the ratio of the frequency component of the second frequency range in the frequency component of the first frequency range in the kth expiratory sound in the expiratory sound to be examined, and Hk is the examination target. The ratio of the frequency component of the third frequency range to the frequency component of the first frequency range in the kth expiratory sound in the expiratory sound, Sk is k in the expiratory sound to be examined A value representing the sound pressure in the th exhalation sound, Ln is a ratio of the frequency component in the second frequency range in the frequency component in the first frequency range in the n th reference expiratory sound, and Hn is n The ratio of the frequency component of the third frequency range to the frequency component of the first frequency range of the first reference expiratory sound, Sn is a value representing the sound pressure of the nth reference expiratory sound, and k and n are positive Integer, m is the reference call Is the number of sound.
The respiratory sound analysis device according to claim 4, wherein a suspected expiratory sound is detected by comparing the calculated value obtained by the formula (1) with a predetermined reference value.   The suspicious expiratory sound detection unit uses a sound generated in a predetermined period in the first half of the expiratory period from the start of the expiratory sound to the end of the expiratory sound among the expiratory sounds extracted by the expiratory sound extracting unit, The respiratory sound analysis device according to any one of claims 1 to 5, wherein the respiratory sound analysis device detects a suspicious breath sound. A change information acquisition unit for acquiring change information representing a change based on the breathing motion of the sleeping subject;
A storage unit that stores the respiratory sound information and the change information in a state in which the respiratory sound information acquired by the respiratory sound information acquisition unit is associated with the change information acquired by the change information acquisition unit;
The expiratory sound extraction unit is configured to extract an expiratory sound based on respiratory sound information and change information stored in the storage unit, according to any one of claims 1 to 6. Respiratory sound analyzer. In a respiratory sound analysis method for detecting at least one abnormality of aspiration, laryngeal invasion, pharyngeal residue, and pharyngeal retention occurring in a sleeping subject based on the respiratory sound emitted by the subject,
A breathing sound information acquisition step for acquiring breathing sound information representing a continuous breathing sound emitted by the subject during a predetermined period;
Based on the breathing sound information acquired by the breathing sound information acquisition unit, a breathing sound extraction step of extracting each breathing sound of each continuous breathing sound;
One or a plurality of reference exhalation sounds serving as a reference are extracted from the exhalation sounds extracted by the exhalation sound extraction unit, and an abnormal exhalation sound and a suspected exhalation sound generated in the event of the abnormality are based on the reference exhalation sounds. A breathing sound analysis method comprising: a suspicious breath detection step for detecting. A respiratory sound analysis program used in a computer that detects at least one abnormality of aspiration, laryngeal invasion, pharyngeal residue, and pharyngeal retention that occurs in a sleeping subject based on the respiratory sound produced by the subject. And
A breathing sound information acquisition unit for acquiring breathing sound information representing a continuous breathing sound emitted by the subject during a predetermined period;
Based on the respiratory sound information acquired by the respiratory sound information acquisition unit, the exhalation sound extraction unit that extracts each exhalation sound of each continuous respiratory sound, and
One or a plurality of reference exhalation sounds serving as a reference are extracted from the exhalation sounds extracted by the exhalation sound extraction unit, and an abnormal exhalation sound and a suspected exhalation sound generated in the event of the abnormality are based on the reference exhalation sounds. Suspicious breath detection unit to detect,
Breathing sound analysis program to make the computer function as.