JP2009254611A - Cough detector - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a cough detector which can precisely detect cough. <P>SOLUTION: A cough detector 1 characteristically comprises a myogenic potential-measuring means 210 which detects information of myogenic potential which is caused in association with muscular contraction in a subject, a posture-detecting means 220 which detects information of posture of the subject, and a cough-detecting means 3 which detects coughing on the basis of the information of myogenic potential and the information of posture. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a cough detection device.

  Cough is a common symptom of respiratory diseases (particularly asthma, chronic obstructive pulmonary disease, bronchitis, etc.). The current situation is that the diagnosis of cough depends on an interview, but the patient does not always have cough at the time of the examination, and the doctor can only hear the subjective symptoms from the patient. Moreover, even if the patient remembers the symptoms at the time of awakening during the day, the cough during sleep is limited to expressions such as severe cough and inability to sleep. Therefore, there has been a problem that objective evaluation cannot be performed and effective treatment cannot be performed.

  Therefore, in Patent Document 1, in order to objectively evaluate cough, a signal obtained from an acceleration sensor attached to the subject's throat is collated with a pre-registered cough pattern. There is described a cough detection device that determines whether a signal is based on cough.

In Non-Patent Document 1, in order to monitor the occurrence frequency of cough for a plurality of subjects for 24 hours, signals are acquired from an electromyograph (EMG) and a macrophone attached to the subjects, and the obtained signals are received by a personal computer. By analyzing, cough is counted.
Japanese Patent Laid-Open No. 9-98964 Coughing Frequency in patients with persistent cough: assessment using a 24hour ambulatory recorder (European Respiratory Journal 1994, 7, 1246-1253)

  However, in the cough detection device described in Patent Document 1, since the acceleration sensor is attached to the subject's throat, for example, the movement of the throat when swallowing an object in the mouth is erroneously detected as cough. there is a possibility. Further, the motion of shaking the head is a routine operation, which also affects the throat and may cause false detection. Thus, when an acceleration sensor is attached to the throat, there are many factors that cause false detection.

  In the method described in Non-Patent Document 1, cough is detected based on the measured myoelectric potential and audio signal. Since the myoelectric potential is also detected due to muscle contraction caused by an operation other than cough, an error occurs. Detection will occur, and the audio signal will also cause erroneous detection due to the voice of the subject or external noise.

  This invention is made | formed in view of the above problems, and it aims at providing the cough detection apparatus which can detect a cough accurately.

1. A myoelectric potential measuring means for detecting myoelectric potential information generated as the subject's muscle contracts;
Posture detection means for detecting posture information of the subject;
Cough detection means for detecting cough based on the myoelectric potential information and the posture information;
A cough detection device characterized by comprising:

  2. 2. The cough detection device according to 1 above, wherein the myoelectric potential measuring means detects myoelectric potential information of a respiratory muscle of a subject.

  3. 3. The cough detection device according to 2 above, wherein the respiratory muscle is an external oblique muscle.

  4). The cough detecting means detects candidate cough from myoelectric potential information, and determines the candidate cough whose posture information changes within a predetermined range before and after detecting the candidate cough as a cough. The cough detection device according to any one of the above.

5. Having a microphone to detect the sound signal;
4. The cough detection device according to any one of 1 to 3, wherein the cough detection unit detects cough based on a sound signal detected by the microphone, the myoelectric potential information, and the posture information.

  6). The cough detection means detects a candidate cough from myoelectric potential information, a change in posture information before and after the candidate cough is within a predetermined range, and a sound pressure value of a sound signal detected by the microphone is a predetermined value or more 6. The cough detection device according to 5 above, wherein the candidate cough is determined as cough.

  According to the present invention, it is possible to provide a cough detection device capable of detecting cough with high accuracy by including cough detection means for detecting cough based on myoelectric potential information and posture information.

  Although the present invention will be described based on an embodiment, the present invention is not limited to the embodiment.

[Device configuration]
FIG. 1 is a configuration diagram of a cough detection device 1 according to the present embodiment. The cough detection device 1 includes a sensor unit 2 attached to a subject and a cough detection unit 3 that detects cough by analyzing a signal from the sensor unit 2. The sensor unit 2 includes a myoelectric potential measurement unit 210 and a posture detection unit 220.

  The myoelectric potential measurement unit 210 functions as a “myoelectric potential measurement unit”, the posture detection unit 220 functions as a “posture detection unit”, and the cough detection unit 3 functions as a “cough detection unit”.

  The myoelectric potential measurement unit 210 and the posture detection unit 220 are connected to the cough detection unit 3 via the communication medium L, respectively. This communication medium L may be wired or wireless. When connected wirelessly, the chance of giving unnecessary vibration and noise to the sensor used in the sensor unit 2 via the line as in the case of wired connection can be reduced, and the behavior of the subject is restricted. Can be reduced.

  The “myoelectric potential measurement unit 210” includes a detection electrode and an electrode unit 211 including a grounding electrode used by being attached to the surface of the subject's skin, an amplifier 212 that amplifies a myoelectric potential signal detected by the electrode unit 211, and an amplifier 212. Filter circuit 213 for removing low frequency components below a predetermined frequency and high frequency components above a predetermined frequency from the myoelectric signal amplified by the above, and an A / D converter for converting the myoelectric signal processed by the filter circuit 213 into digital data 214, and an I / F 215 that transmits digital data converted by the A / D converter 214 to the cough detection unit 3.

  A detection electrode and a ground electrode of the electrode unit 211 are attached to the skin surface near the respiratory muscles of the subject. The myoelectric potential measurement unit 210 measures the myoelectric potential generated with the contraction of the respiratory muscle of the subject. Here, the respiratory muscle is a muscle that controls respiratory motion, and includes an intercostal muscle, a diaphragm, an external oblique muscle, an internal oblique muscle, a rectus abdominis muscle, and the like. The coughing action is an action of raising the intrathoracic pressure and then immediately releasing the vocal cords to exhale at once. In order to increase the intrathoracic pressure, after breathing in the lungs and closing the vocal cords of the throat, contraction of the external oblique muscle, the internal oblique muscle, and the rectus abdominis muscle is performed.

  In addition, as an attachment position of the electrode part 211, in order to acquire the myoelectric potential information of an external oblique muscle among respiratory muscles, attaching to the skin surface near the external oblique muscle is preferable. This is because the influence of the myocardium can be reduced and the cough associated with the movement of the respiratory muscle can be accurately detected when the measurement location is far from the heart.

  The “posture detection unit 220” includes an accelerometer 221 that converts an input change in posture into an electric signal, an amplifier 222 that amplifies the electric signal converted by the accelerometer 221, and a predetermined frequency from the electric signal amplified by the amplifier 222. Filter circuit 223 for removing the above high-frequency components, A / D converter 224 for converting the myoelectric signal processed by filter circuit 223 into digital data, and cough detection for the digital data converted by A / D converter 224 The I / F 225 is transmitted to the unit 3.

  In the posture detection unit 220, the posture information of the subject is detected by the accelerometer 221. Posture information is a concept that represents a posture state and a change in posture. The accelerometer 221 is attached to the subject's abdomen or the like. The accelerometer 221 uses a three-axis acceleration sensor, and detects changes in the movement of the subject's body in the left-right direction (X-axis), the vertical direction (Y-axis), the front-rear direction (Z-axis), and the direction with respect to gravity. Thus, the subject can detect the posture state such as the body posture and the change in posture.

  The electrode part 211 and the accelerometer 221 are preferably attached directly or indirectly to the human body with a double-sided tape or the like applied with an adhesive having a low sensitivity to the skin. Moreover, you may make it provide the cover material and cushion material for noise prevention around these.

  In the present embodiment, the filter circuits 213 and 223 for removing the low frequency component below the predetermined frequency and the high frequency component above the predetermined frequency are provided, so that the signal of the high frequency component due to sudden electrical noise is removed in advance. .

  In the present embodiment, the posture detection unit 220 describes an example using a triaxial acceleration sensor. However, the present invention is not limited to this, and the subject is detected by detecting a change in angular velocity using a triaxial gyro sensor. A change in posture may be detected.

  Further, the myoelectric potential measurement unit 210 and the posture detection unit 220 may be integrated. In this case, it is desirable that the position where the sensor unit 2 is attached to the subject is on the rectus abdominis muscle, from the viewpoint of improving both the action of the respiratory muscles and the detection accuracy of the posture change of the subject.

  The “cough detection unit 3” is an I / F 31 that receives data from the sensor unit 2. A CPU 32 that processes data received by the I / F 31 according to a program, a ROM 33 that stores programs and data necessary for processing by the CPU 32, and a RAM 34 that temporarily stores programs and data necessary for processing by the CPU 32 Has. Furthermore, a storage device 35 that stores the processing results in the CPU 32 in a semiconductor memory, hard disk, DVD-R, CD-R, or the like inserted in the device. The cough detection unit 3 includes an input unit 36 for inputting data, and a display unit 37 for displaying a processing result in the CPU 32.

  The cough detection unit 3 may be configured with a dedicated information processing device or a general-purpose personal computer. If it is a personal computer, it is preferably a personal digital assistant (PDA) that can be easily carried. In the former case, it may be integrated with the sensor unit 2 and attached to the waist of the subject by a belt or the like. In this way, troublesome wiring becomes unnecessary. Further, in the case of such a form, the A / D converters 214 and 224 are provided in the sensor unit 2, but an A / D converter may be provided in the cough detection unit 3.

[Control flow]
FIG. 2 is an explanatory diagram of a control flow performed by the cough detection device 1 according to the present embodiment. The control flow is executed by the CPU 32. Prior to this control flow, the data detected by the sensor unit 2 is stored in the storage device 35 together with time-lapse information such as time.

  In step S <b> 11, the CPU 32 acquires myoelectric potential information and posture information from the data stored in the storage device 35.

  In step S12, candidate cough is extracted from the myoelectric potential information. For example, if the variance of the myoelectric potential data in the unit section is equal to or greater than a predetermined value, the candidate cough is extracted as a candidate cough. For the purpose of improving detection accuracy, data of a predetermined frequency band may be extracted from the myoelectric potential data, and the candidate cough may be determined from the extracted band data.

  In step S13, it is determined from the posture information data corresponding to the section (timing) in which the candidate cough is detected in step S12 whether or not the subject's posture has changed before and after the section. The change in posture means, for example, a change from a standing position to a sitting position, a sitting position to a supine position, or a mutual change. The actual waveform example 1 of the posture change will be described later.

  When it is determined that the posture has not changed (step S13: No), in step S21, the candidate cough extracted in step S12 is determined to be cough, and the cough is recorded in the storage device 35.

  On the other hand, if it is determined that there has been a change in posture (step S13: Yes), the CPU 32 determines that the candidate cough is a false cough and determines that there is no cough and ends.

[Detection waveform example]
FIG. 3 shows waveform data in the myoelectric potential measuring unit 210 and the posture detecting unit 220 when coughing occurs. FIG. 4 shows waveform data in the myoelectric potential measurement unit 210 and posture detection unit 220 when the subject changes posture.

  The column of reference numeral 1 in the figure shows the waveform data of myoelectric potential from the myoelectric potential measuring unit 210, and the columns of reference numerals 2x, 2y, and 2z indicate the x-axis direction, y-axis direction from the accelerometer 221, Waveform data in the z-axis direction is shown. As described above, the acceleration sensor is attached so that the x axis, the y axis, and the z axis correspond to the left and right direction (X axis), the vertical direction (Y axis), and the front and rear direction (Z axis) of the subject's body. . The horizontal axis represents time, and the vertical axis represents potential (column 1) and acceleration (column 2), respectively.

  In the column 1 of FIG. 3, reference numerals c1, c2, c3, and c4 indicate cough occurrence timings. As shown in the figure, it can be seen that when cough occurs, the variance value in the unit interval of the waveform data of the myoelectric potential is large. Thus, the occurrence of cough can be detected from the myoelectric potential information. On the other hand, however, when cough is detected only by the myoelectric potential information, a problem of false detection (false cough) that causes an event other than cough to be determined as cough occurs. This will be described below.

  In FIG. 4, the state of the subject is standing at the timing of S1. In S2, the timing of changing from the standing position to the sitting position is shown, and in S3, the sitting position is stable. S4 is the timing at which the sitting position changes again from the standing position, and S5 is the standing stable state.

  As shown in the figure, at the timings S2 and S4, contraction of the external oblique muscles accompanying the posture change of the subject occurs. Therefore, as shown by α1 and α2 in column 1 of FIG. As can be seen from FIG. And it is difficult to distinguish the waveform data at the time of the posture change from the waveform data at the time of the occurrence of cough shown in FIG. 3, and an erroneous detection that determines the posture change as cough occurs.

  In the present embodiment for such a problem, as described in the control flow of FIG. 2, waveform data (for example, a large variance value) having features at the time of cough detected from waveform data of myoelectric potential is immediately obtained. Judgment is suspended as a candidate cough without being judged as cough. The final determination is made by performing a NOT operation to determine whether or not the subject's posture has changed (step S13).

  An example of posture change detection will be described. The posture change is detected by comparing the output of the accelerometer 221 in S1 (standing position) and S3 (sitting position) in the section before and after the section S2 in which the candidate cough α1 is detected in step S12. When the posture is changed, the output of the acceleration sensor of the posture detection unit 220 is greatly changed as shown in columns 2x, 2y, and 2z in FIG.

  In the example shown in FIG. 4, in particular, in the change from the standing position (S1) to the sitting position (S3) or from the sitting position (S3) to the standing position (S5), the three-axis acceleration sensor shown in the columns 2x and 2z in FIG. It can be seen that the change in waveform data in the x and z directions is large. Then, the cough detection unit 3 calculates a difference between the preceding and following sections, and determines that the posture has changed when the difference is equal to or greater than a predetermined value. Since the change in the waveform data of the low frequency component is large when the posture is changed, the high frequency component is removed by averaging processing or the like, and the change in posture is detected from the waveform data after the removal. May be.

  If it is determined that the subject's posture has changed from the waveform data of the posture detector 220 in the sections (S1 and S3) before and after the section S2 in which the candidate cough α1 is detected, The candidate cough is determined to be a pseudo-cough by judging that it is due to muscle contraction and not accompanying the occurrence of cough. On the other hand, if it is determined that the posture has not changed, it is determined that the candidate cough is a true cough.

  According to the present embodiment, it is possible to detect cough with high accuracy by detecting cough based on myoelectric potential information and posture information.

[Second Embodiment]
A cough detection device according to the second embodiment will be described with reference to FIGS. 5 and 6. FIG. 5 is a configuration diagram of the cough detection device 1 according to the second embodiment. In the second embodiment, a sound detection unit 230 is further added to the embodiment shown in FIG. The configuration other than the sound detection unit 230 is the same as that of the embodiment shown in FIG.

  In the drawing, a “sound detection unit 230” is a so-called condenser microphone type sound collection unit. The sound detection unit 230 attached to the subject converts the sound around the subject into a sound signal by the microphone 231, and the microphone 231 An amplifier 232 that amplifies the detected sound signal, a filter circuit 233 that removes low frequency components below a predetermined frequency and high frequency components above a predetermined frequency from the sound signal amplified by the amplifier 232, and the sound signal processed by the filter circuit 233 Is converted to digital data, and an I / F 235 that transmits the digital data converted by the A / D converter 234 to the cough detection unit 3.

  Also in the second embodiment, the sound detection unit 230 may be integrated with the myoelectric potential measurement unit 210 and the posture detection unit 220.

  FIG. 6 is an explanatory diagram of a control flow performed by the cough detection device 1 according to the second embodiment. The control flow is a flow executed by the CPU 32 of the cough detection unit 3. The steps other than step S14 are the same as the control flow shown in FIG.

  In the second embodiment, even if it is waveform data (step S12: Yes) having cough characteristics detected from waveform data of myoelectric potential, and there is no change in the posture of the subject (step S13: No). Suspend judgment without judging cough. In step S14, if the sound pressure detected by the sound detection unit 230 corresponding to the section (timing) in which the candidate cough is detected in step S12 is equal to or greater than a predetermined value (step S14: Yes), the CPU 32 Determines that the candidate cough is cough (step S21). On the other hand, when it is determined that the sound pressure has fallen below the predetermined value (step S14: No), it is determined that the candidate cough is a false cough and the cough is not present, and the process ends (step S22).

  According to this embodiment, the cough can be detected with high accuracy by determining whether the candidate cough detected from the myoelectric potential information is a pseudo cough or a true cough based on the posture information and the sound signal. It becomes possible.

It is a lineblock diagram of cough detection device 1 concerning this embodiment. It is explanatory drawing of the control flow which the cough detection apparatus 1 in this embodiment performs. The waveform data in the myoelectric potential measuring unit 210 and the posture detecting unit 220 at the time of occurrence of cough are shown. The waveform data in the myoelectric potential measurement unit 210 and the posture detection unit 220 when the subject changes his / her posture is shown. It is a block diagram of the cough detection apparatus 1 which concerns on 2nd Embodiment. It is explanatory drawing of the control flow which the cough detection apparatus 1 in 2nd Embodiment performs.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Cough detection apparatus 2 Sensor part 210 Myoelectric potential measurement part 211 Electrode part 220 Posture detection part 221 Accelerometer 230 Sound detection part 231 Microphone 3 Cough detection part 32 CPU
33 ROM
34 RAM

Claims (6)

A myoelectric potential measuring means for detecting myoelectric potential information generated as the subject's muscle contracts;
Posture detection means for detecting posture information of the subject;
Cough detection means for detecting cough based on the myoelectric potential information and the posture information;
A cough detection device characterized by comprising: The cough detection device according to claim 1, wherein the myoelectric potential measuring unit detects myoelectric potential information of a respiratory muscle of a subject. The cough detection device according to claim 2, wherein the respiratory muscle is an external oblique muscle. The cough detecting means detects a candidate cough from myoelectric potential information, and determines a candidate cough whose change in posture information is within a predetermined range before and after the detection of the candidate cough as a cough. 4. The cough detection device according to any one of 3. Having a microphone to detect the sound signal;
The cough detection device according to claim 1, wherein the cough detection unit detects cough based on a sound signal detected by the microphone, the myoelectric potential information, and the posture information. . The cough detection means detects a candidate cough from myoelectric potential information, a change in posture information before and after the candidate cough is within a predetermined range, and a sound pressure value of a sound signal detected by the microphone is a predetermined value or more The cough detection device according to claim 5, wherein the candidate cough is determined as cough.

Images