JP2009279122A - Status detector - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To easily carry out the measurement of swallowing without giving a load to a patient and to accurately carry out the measurement of coughs by using sound signals and body movement signals obtained from one sensor. <P>SOLUTION: Vibration signals of a living body are collected by one sensor, the obtained vibration signals are separated to the body movement signals and the sound signals, the status of a subject is detected from the respective signals, thereby the status of the subject such as the swallowing or the coughs can easily and accurately be detected by without giving a load to the subject. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a state detection device that detects swallowing or coughing.

  Many developed countries, including Japan, are facing an aging society, and diseases associated with aging are increasing. Medical institutions can easily detect the presence or absence of various parts of the human body due to illness in daily and initial screenings, in order to help develop a treatment policy by knowing the patient's illness, degree of illness, and cause of illness. A possible evaluation method is desired. In particular, there are cases where the patient's swallowing and coughing movements may be accompanied by illness, and therefore swallowing and cough detection means are being actively developed.

  Conventionally, the detection of swallowing has been to observe the situation where the patient swallows the contrast agent while performing X-ray projection of the pharynx. However, X-rays are used in X-ray projection, and it is not preferable to expose a human body to X-rays for a long time. In addition, the X-ray projection apparatus is expensive and requires special qualifications for its handling, and requires management of radioactive materials, etc., and cannot be easily used in a medical field. Therefore, as a simpler and safer method than X-ray projection, multiple sensors such as an acceleration sensor that captures the movement of the throat, a microphone that captures the sound of swallowing, and a surface electromyogram measurement electrode around the pharynx are attached to the human body. A method has been proposed (see Patent Document 1).

In addition, conventionally, a method for detecting a cough has been proposed, in which a microphone that captures the sound of a cough is attached to the throat, a voice emitted along with the cough is detected, and a time waveform signal is analyzed to detect the state. It has been adopted (see Patent Document 2).
JP 2005-304890 A JP-A-1996-38481

  These sensors used to measure swallowing are by no means small, and if many sensors are attached in the vicinity of the patient's throat, the burden on the patient is increased. Then, it becomes difficult to maintain a rest, and it is prevented that a patient's state is detected accurately. Furthermore, it leads to worsening of symptoms.

  Moreover, cough is measured using only a voice signal, but it is difficult to clearly separate cough from speech and screams contained in the voice signal, and it is difficult to detect cough with high accuracy. Therefore, an object of the present invention is to detect swallowing and coughing easily and accurately without burdening a patient by using a voice signal and a body motion signal obtained from one sensor.

  The above object is achieved by the invention described below.

1. A collecting means for collecting a biological vibration signal;
Separating means for separating the vibration signal into a body motion signal and an audio signal;
Detecting means for detecting the state of the subject from the body motion signal and the audio signal;
A state detection apparatus comprising:

2. The body motion signal is a low frequency component signal included in the vibration signal,
2. The state detection apparatus according to claim 1, wherein the sound signal is a signal of a high frequency component included in the vibration signal.

  3. 3. The state detection apparatus according to 1 or 2, wherein the state is swallowing or coughing.

  4). The detection means detects swallowing of a subject based on a swallowing sound detected from the audio signal and a swallowing motion of a throat detected from the body motion signal. The state detection device described in 1.

  5). The detection means detects a subject's cough based on a cough sound detected from an audio signal and an abdominal cough movement detected from the body motion signal. The state detection device described.

  6). 6. The state detection device according to any one of 2 to 5, wherein the low-frequency component signal is a signal having a frequency of less than 20 Hz, and the high-frequency component signal is a signal having a frequency of 20 Hz or more. .

  By obtaining the audio signal and body motion signal that accompany the occurrence of cough and swallowing, and detecting the patient's state using the obtained audio signal and body motion signal, the patient's state can be detected with one sensor. Therefore, the patient's condition can be easily detected without giving a burden to the patient, and the condition of the patient can be detected with high accuracy since the condition is comprehensively detected from the two signals.

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

[Device configuration]
Swallowing in the present invention is an exercise in which a food or drink is formed as a bolus in the oral cavity, and the bolus and oropharyngeal secretions are sent to the stomach through the pharynx and esophagus.

  Cough refers to movement of respiratory muscles and bronchial smooth muscle that occurs when a foreign substance is attached to the throat, trachea, bronchus, or pleura to discharge the foreign substance to the outside of the body and protect the lungs.

  When detecting swallowing, which is a movement of the throat, the microphone 10 is attached to the throat as shown in FIG. 1, and when detecting cough, the microphone 10 is attached to the abdomen as shown in FIG. This is because the throat shows a characteristic movement during swallowing, and the abdomen shows a characteristic movement during coughing. When detecting cough, it is particularly preferable to attach it around the diaphragm, since it can detect even a weak cough with high sensitivity. In addition, when detecting coughing, when it is attached to the abdomen, there are fewer factors causing erroneous detection than when it is attached to the throat, and coughing can be detected with high accuracy. The movement of the throat is caused by, for example, the daily movement of the neck or the swallowing movement that is normally performed, but the abdomen is originally a region with little movement and does not move much unless it is intentionally twisted. .

  The microphone 10 is preferably attached directly or indirectly to the human body with a double-sided tape or the like applied with an adhesive having low sensitivity to the skin. Further, it is preferable to provide a noise-preventing cover material or cushion material around the microphone 10.

  FIG. 3 is a configuration diagram of the state detection device. The state detection device 1 includes a state measurement unit 20 and a state detection unit 30, and the state measurement unit 20 is connected to the state detection unit 30 via the communication medium L. This communication medium L may be wired or wireless. If connected wirelessly, unnecessary vibrations or lead wires touch the body or clothes through the wire to the sensor used in the state measurement unit 20 as in the case of wired, and give noise similar to biological sound. Opportunities can be reduced. Therefore, there is almost no erroneous detection, and highly accurate detection can be performed. Moreover, the restrictions on the behavior of the subject can be reduced.

  Next, a method for obtaining a swallowing or cough signal when swallowing or coughing occurs will be described. When a swallowing or coughing motion occurs, an audible sound (20 Hz to 20 kHz) is generated, and the swallowing motion or coughing motion vibrates the air in contact with it because the body part moves. The frequency of the air vibration generated in this way is a low frequency because the frequency generation source is due to the operation of a part of the human body, and is a frequency in the inaudible range (˜20 Hz). Therefore, the air vibration signal detected by the microphone includes an audible frequency signal that can be heard by the human ear as a result of swallowing and coughing, and an inaudible frequency signal.

  Here, an audible signal generated by swallowing or coughing is referred to as an audio signal, and a signal generated by vibrating the air by moving part of the body accompanying swallowing or coughing is referred to as a body movement signal. Called.

  Since the body movement signal and the voice signal are detected by the single microphone 10, they are included in the same electric signal, and thus need to be separated for use separately. As a method of separating the body motion signal and the voice signal, a frequency filter that can be separated at the boundary frequency is used by utilizing the fact that each frequency is different. Specifically, a high-frequency audio signal is cut using a low-pass filter to obtain a low-frequency body motion signal, and a high-frequency filter is used to cut a low-frequency body motion signal to obtain a high-frequency audio signal. Since the body motion signal is a signal of 20 Hz or less which is an inaudible range as described above, the cut-off frequency of the low-pass filter is 20 Hz. Also, since the audio signal is a signal of 20 Hz or higher that is an audible range, the cutoff frequency of the high-pass filter is 20 Hz.

  The state measurement unit 20 amplifies the low-frequency component, a low-pass filter circuit 22L that detects only a low-frequency component that is lower than or equal to a predetermined frequency from the audio signal converted by the microphone 10, and a high-pass filter circuit 22H that detects a high-frequency component that is higher than or equal to a predetermined frequency. Amplifying circuit 23L, amplifying circuit 23H for amplifying high frequency components, A / D converting circuit 24L for A / D converting the amplified low frequency, A / D converting circuit 24H for A / D converting the amplified high frequency, A / The I / F 25 transmits the D-converted signal to the state detection unit 30. As described above, 20 Hz is preferable as the predetermined frequency.

  Since the low-frequency signal and the high-frequency signal are each processed by being transmitted using a single-channel signal line, it is preferably handled as a two-channel stereo signal.

  The state detection unit 30 stores the I / F 25 that receives data from the state measurement unit 20, the CPU 31 that processes the body movement data received by the I / F 25 according to the program, and the programs and data necessary for the processing by the CPU 31. ROM 32, RAM 33 for temporarily storing programs and data necessary for processing in CPU 31, external storage device 34 for storing processing results in CPU 31 in a hard disk, DVD-R, CD-R, etc., state detection unit 30 includes an input unit 35 on the main body of the state detection apparatus 1 for inputting data such as necessary information such as a subject ID, and a display unit 36 for displaying a processing result in the CPU 31.

  The state detection unit 30 may be configured with a dedicated information processing apparatus 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.

  As shown in FIG. 4, the state measurement unit 20 inputs the output of the microphone 10 to the A / D conversion circuit 41, and then inputs the output to the low-pass filter circuit 42L and the high-pass filter circuit 42H, and the state detection unit through the I / F 25 A configuration of outputting to 30 may be adopted. With such a configuration, processing downstream from the A / D conversion circuit 41 can be performed in the computer, so there is no need to prepare the low-pass filter circuit 22L, high-pass filter circuit 22H, amplifier circuit 23L, and amplifier circuit 23H as hardware. It can be processed easily. On the other hand, if the configuration shown in FIG. 3 is adopted, the outputs of the low-pass filter circuit 22L and the high-pass filter circuit 22H are A / D converted and taken into the computer, so that the audio signal and the body motion signal are separately A / D converted. Thus, the dynamic range of each signal can be increased, and the signal can be analyzed precisely.

  Next, an example of an electrical signal that is obtained using the microphone 10 and is separated into a voice signal and a body motion signal when swallowing or coughing occurs will be described. FIG. 5 is an electrical signal obtained from the microphone 10 when swallowing occurs, and FIG. 6 is an electrical signal obtained from the microphone 10 when coughing occurs. In FIG. 5, two waveform groups each having a high intensity waveform are measured, and each waveform group represents swallowing. It can be seen that swallowing is a phenomenon of about 1 second. In FIG. 6, two waveform groups each having a high intensity waveform are measured, and each waveform group represents a cough. The waveform group on the left side in the figure is a case where coughing occurs continuously, and the waveform group on the right side in the figure is a case where coughing occurs once. It can be seen that the cough that occurred only once has a shorter generation time than swallowing and is about 0.3 seconds.

  Next, with respect to the original signal obtained by using the microphone 10 in the case of swallowing, the low-frequency component signal obtained by applying a low-pass filter process that passes a frequency of less than 20 Hz to the original signal, and the original signal FIG. 7 shows an example of a high-frequency component signal obtained by performing a high-pass filter process for passing a frequency of 20 Hz or higher.

  As shown in the figure, the low-frequency component is a signal waveform obtained by removing a component having a large vibration frequency from the original signal, and body movements associated with swallowing can be taken out. The high frequency component is a signal waveform from which a component having a small vibration frequency is removed, and a swallowing sound that can be heard by a human ear can be extracted.

  Similarly, in the case of cough, the microphone 10 can be used to obtain an original signal, low-pass filter processing can be performed to detect body movement accompanying cough, and high-pass filter processing can be performed to obtain a cough sound.

[Swallowing and cough measurement process]
Next, description will be given of a state measurement in which a voice signal and a body motion signal are collected as data by the microphone 10 when swallowing or cough occurs. FIG. 8 is a diagram illustrating a state measurement flow according to the present embodiment. This state measurement flow is a flow executed by the CPU 31 based on the state measurement program in the ROM 32. It is assumed that an ID or the like for specifying the subject is input in advance by the input unit 35.

  First, the CPU 31 determines whether or not the start of state measurement is instructed from the input unit 35 (step S10). If it is determined that the start of state measurement has been instructed (step S10; Yes), the CPU 31 stores the audio signal and body movement signal measurement data from the state measurement unit 20 in the external storage device 34 via the RAM 33. Is started (step S11). At this time, the measurement data is stored in association with the ID and time of the subject. If it is determined that the start of state measurement is not instructed (step S10; No), the process returns to step S10 and waits until the start of state measurement is instructed.

  Next, the CPU 31 determines whether or not the end of state measurement has been instructed (step S12). If it is determined that the end of the state measurement is instructed (step S12; Yes), the CPU 31 ends the operation of storing the measurement data from the state measurement unit 20 in the external storage device 34 via the RAM 33 (step S13). If it is determined that the end of state measurement is not instructed (step S12; No), the process returns to step S12 and waits until the end of state measurement is instructed.

  Next, the concept and detection flow of detection processing for detecting cough and swallowing from the obtained audio signal and body motion signal will be described with reference to FIGS. 9, 10 and 11. FIG. FIG. 9 is a diagram illustrating a state detection flow according to the present embodiment.

  This state detection flow is a flow executed by the CPU 31 based on the state detection program in the ROM 32.

  First, the CPU 31 determines whether or not an instruction to start state detection is given from the input unit 35 (step S20). If it is determined that the start of state measurement has been instructed (step S20; Yes), the CPU 31 performs swallowing detection processing while calling and expanding the measurement data stored in the external storage device 34 to the RAM 33 (step S21).

  The concept of swallowing and cough detection processing is to calculate the variance value of the body motion signal and the voice signal in the sampling window, and determine the presence or absence of swallowing or cough based on the variance value.

  FIG. 10 is a conceptual diagram showing the concept of sampling of body motion signals for swallowing in the state detection process of the present embodiment, with the horizontal axis indicating the elapsed time t and the vertical axis indicating the data level.

  In the figure, a plurality of sampling windows W are shown. That is, the sampling window is a detection time width for collecting data for a predetermined time. In the present embodiment, data existing in each sampling window W is handled as a group of data. The sampling window W is provided so that, for example, about half of the region overlaps between adjacent sampling windows so that no detection omission occurs.

  FIG. 11 is a conceptual diagram showing the concept of the state detection process of the present embodiment. FIG. 11B shows data, where the horizontal axis indicates the elapsed time t and the vertical axis indicates the data level. FIG. 11A shows the variance value V of each sampling window in the data of FIG. 11B, and also shows the threshold value Vth for the variance value V.

  A sampling window having a variance value V larger than the threshold value Vth is detected as swallowing. In the example of FIG. 11A, four swallows A, B, C, and D are detected. For a portion where a window having a variance value V greater than the threshold value Vth is continuous, the continuous portion is counted as one swallow. For example, for the portion C in FIG. 11A, four windows C1 to C4 having a variance value V larger than the threshold value Vth are continuous. In this case, the four window portions are counted as one swallow. Further, as the strength of swallowing, for example, the value of the dispersion value V can be used. About the part of C in Fig.11 (a), the value which added four dispersion values V of continuous C1-C4 can be used as the intensity | strength of swallowing, for example. In order to improve accuracy, a specific frequency band may be further extracted before obtaining the variance.

  As described above, when the swallowing detection process is performed in step S21 and the result is received, and it is determined in step S22 that swallowing is not detected, the state detection flow ends. If it is determined in step S22 that swallowing has been detected, the process proceeds to step S23.

  In step S23, the presence / absence of swallowing is detected from the high-frequency swallowing sound signal in the same manner. However, since the signal intensity differs between the swallowing body motion signal and the voice signal, the magnitude of the threshold value Vth is changed. As described above, when the swallowing detection process is performed in step S23 and the result is received, and it is determined in step S24 that swallowing is not detected, this state detection flow ends. The case where swallowing can be detected at a low frequency and swallowing cannot be detected at a high frequency is a case where the skin moves when the subject moves and the microphone picks up the low frequency component. If it is determined in step S24 that swallowing has been detected, the process proceeds to step S25.

  In step S <b> 25, the CPU 31 performs an operation of saving the detection result in the external storage device 34 via the RAM 33.

  A similar flow can be detected when detecting cough. However, a value suitable for cough detection is adopted as the value of Vth in the body motion signal and the voice signal.

  As described above, swallowing and coughing can be detected easily and accurately by using one microphone and detecting swallowing and coughing using both body motion signals and audio signals.

It is a schematic diagram which shows the condition which attached the microphone to the throat part. It is a schematic diagram which shows the condition which attached the microphone to the abdomen. It is a block diagram of a state detection apparatus. It is a block diagram of a state measurement part. It is an example of the electrical signal obtained from the microphone when swallowing occurs. It is an example of the electrical signal obtained from the microphone when cough occurs. It is an example of the original signal of swallowing, a low frequency component, and a high frequency component. It is a flowchart showing the state measurement flow which concerns on this embodiment. It is a flowchart showing the state detection flow which concerns on this embodiment. It is a conceptual diagram which shows the concept of the sampling of the body movement signal of swallowing. It is a conceptual diagram which shows the concept of the state detection process of this embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Microphone 20 State measurement part 30 State detection part 22L, 42L Low pass filter circuit 22H, 42H High pass filter circuit 23L, 23H Amplification circuit 24L, 24H, 41 A / D conversion circuit 25 I / F
31 CPU
32 ROM
33 RAM
34 External storage device 35 Input unit 36 Display unit

Claims (6)

A collecting means for collecting a biological vibration signal;
Separating means for separating the vibration signal into a body motion signal and an audio signal;
Detecting means for detecting the state of the subject from the body motion signal and the audio signal;
A state detection apparatus comprising: The body motion signal is a low frequency component signal included in the vibration signal,
The state detection device according to claim 1, wherein the audio signal is a signal of a high-frequency component included in the vibration signal. The state detection apparatus according to claim 1, wherein the state is swallowing or coughing. The detection unit detects swallowing of a subject based on a swallowing sound detected from the audio signal and a swallowing motion of a throat detected from the body motion signal. The state detection apparatus of crab. The said detection means detects a subject's cough based on the cough sound detected from an audio | voice signal, and the cough movement of the abdominal part detected from the said body motion signal. The state detection device described in 1. 6. The state detection according to claim 2, wherein the low frequency component signal is a signal having a frequency of less than 20 Hz, and the high frequency component signal is a signal having a frequency of 20 Hz or more. apparatus.

Images