JP2018167062A - Respiratory sound analysis apparatus - Google Patents

Abstract

To sort and output a plurality of sound kinds included in a respiratory sound.SOLUTION: A respiratory sound analysis apparatus includes: sorting means (210-250) for sorting a respiratory sound into a normal sound and an abnormal sound; and output means (125,260,270) for outputting an arbitrary respiratory sound from the sorted respiratory sound. The respiratory sound analysis apparatus can sort a respiratory sound and output an arbitrary respiratory sound so as to suitably analyze the respiratory sound including a plurality of sound kinds. The respiratory sound analysis apparatus may further include change means (125,260,270) capable of changing an output state of an arbitrary respiratory sound for every sound kind.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a technical field of a respiratory sound analyzer that analyzes respiratory sounds including a plurality of sound types.

  As this type of device, there is known a device that discriminates between normal respiratory sound and abnormal respiratory sound with respect to a respiratory sound detected by an electronic stethoscope or the like. For example, Patent Document 1 proposes a technique of outputting an abnormal sound analysis result with a three-dimensional display device. Patent Document 2 proposes a technique for detecting a rale as an abnormal sound included in a respiratory sound. Japanese Patent Application Laid-Open No. 2004-228688 proposes a technique of dividing a sound waveform into a plurality of sections and determining a sound type.

JP 2002-538921 A JP 2009-106574 A JP2013-123494A

  However, the techniques described in Patent Documents 1 to 3 described above have a technical problem that a plurality of sound types included in the respiratory sound cannot be arbitrarily output.

  Examples of problems to be solved by the present invention include the above. It is an object of the present invention to provide a respiratory sound analysis apparatus that can separately output a plurality of sound types included in a respiratory sound.

  A breathing sound analyzer for solving the above-described problems includes a sorting unit that sorts a breathing sound into a normal sound and an abnormal sound, and an output unit that outputs an arbitrary breathing sound among the sorted breathing sounds.

  The respiratory sound analysis method for solving the above-described problems includes a classification process for classifying respiratory sounds into normal sounds and abnormal sounds, and an output process for outputting an arbitrary respiratory sound among the classified respiratory sounds.

  A computer program for solving the above-described problems causes a computer to execute a classification process for classifying respiratory sounds into normal sounds and abnormal sounds, and an output process for outputting an arbitrary respiratory sound among the classified respiratory sounds.

  The above-described computer program is recorded on a recording medium for solving the above problems.

It is a block diagram which shows the whole structure of the respiratory sound analyzer which concerns on a present Example. It is a flowchart which shows operation | movement of the respiratory sound analyzer which concerns on a present Example. It is a spectrogram figure which shows the frequency analysis result of the breathing sound containing a hair hair sound. It is a spectrogram figure which shows the frequency analysis result of the breathing sound containing a whistle voice sound. It is a graph which shows the spectrum in the predetermined timing of the breathing sound containing a hair hair sound. It is a conceptual diagram which shows the approximation method of the spectrum of the breathing sound containing a hairpin sound. It is a graph which shows the spectrum in the predetermined timing of the breathing sound containing a whistle voice sound. It is a conceptual diagram which shows the approximation method of the spectrum of the breathing sound containing a whistle voice sound. It is a graph which shows an example of a frequency analysis method. It is a figure which shows an example of a frequency analysis result. It is a conceptual diagram which shows the peak detection result of a spectrum. It is a graph which shows a normal alveolar respiratory sound base. It is a graph which shows a twist hair sound base. It is a graph which shows a continuous ra sound base. It is a graph which shows a white noise base. It is a graph which shows the continuous ra sound base shifted in frequency. It is a figure which shows the relationship between a spectrum, a base, and a coupling coefficient. It is a figure which shows an example of the base used for the observed spectrum and approximation. It is a figure which shows each base which shows a spectrum, and a coupling coefficient. It is a conceptual diagram (the 1) which shows the classification method of the continuous rarity concerning a 1st modification. It is a conceptual diagram (the 2) which shows the classification method of the continuous rarity concerning a 1st modification. It is a graph which shows the threshold value used for classification of a whistle voice sound and analogy sound concerning the 2nd modification. It is a graph which shows the initial value of the threshold value used for classification of a whistle voice sound and analogy sound concerning a 3rd modification. It is a graph (the 1) which shows the value after the adjustment of the threshold value used for classification | category with a whistle voice sound and analogy sound which concerns on a 3rd modification. It is a graph (the 2) which shows the value after the adjustment of the threshold value used for the classification | category of a whistle voice sound and analogy sound which concerns on a 3rd modification. It is a spectrogram figure of the breathing sound containing a whistle voice sound. It is a graph which shows the peak frequency and peak number of a whistle voice sound. It is a spectrogram figure of the breathing sound containing a similar sound. It is a graph which shows the peak frequency and number of peaks of an analogy sound. It is a top view which shows the example of a display in a display part. It is a spectrogram figure which shows the extraction result for every sound type. It is the conceptual diagram which shows an example of the audio | voice output for every classified sound type (the 1). It is a conceptual diagram (the 2) which shows an example of the audio | voice output for every classified sound type. It is a conceptual diagram which shows the volume adjustment method for every classified sound type. It is a conceptual diagram which shows the volume adjustment method for every frequency band. It is a conceptual diagram which shows an example of the image process performed for every sound type. It is a top view which shows an example of the image produced | generated by superimposing the image processed for every sound type. It is a conceptual diagram which shows the display color adjustment method for every classified sound type.

<1>
The respiratory sound analysis apparatus according to the present embodiment includes a classification unit that classifies a respiratory sound into a normal sound and an abnormal sound, and an output unit that outputs an arbitrary respiratory sound among the sorted respiratory sounds.

  According to the respiratory sound analysis apparatus according to the present embodiment, during the operation, the respiratory sound is first classified into a normal sound and an abnormal sound. The breathing sound separated by the sorting means may not be the normal sound and the abnormal sound, and each of the normal sound and the abnormal sound may be further classified into a plurality of sound types. For example, the abnormal sound may be classified into a whistle sound, an analogy sound, a haircut sound, and the like. In addition, it does not specifically limit about a specific classification method, What is necessary is just to be able to classify in the state which can output the arbitrary respiratory sounds mentioned later.

  When the breathing sound is separated, any breathing sound among the separated breathing sounds is output. That is, a plurality of sorted respiratory sounds are selectively output. Thereby, for example, it is possible to output only the breathing sound desired by the user. More specifically, for example, it is possible to output only the whistle voice included in the breathing sound, or output only the whistle voice and analogy sound. In addition, it does not specifically limit about an output aspect, You may output as an audio | voice or an image, and you may output in another aspect.

  As described above, by separating and outputting respiratory sounds, for example, diagnosis of a health condition can be easily performed. Specifically, for example, when it is heard in a state where a plurality of breathing sounds are mixed, it is difficult for a skilled doctor to distinguish and listen to each sound type. It becomes possible to easily discriminate the included sound types. As described above, if it is possible to easily listen to only a specific sound type included in the breathing sound, it can be utilized in doctor education and research.

  As described above, according to the respiratory sound analysis apparatus according to the present embodiment, since the respiratory sound can be sorted and any respiratory sound can be output, the respiratory sound including a plurality of sound types can be suitably analyzed. Is possible.

<2>
In one aspect of the respiratory sound analyzing apparatus according to the present embodiment, the output means outputs a plurality of respiratory sounds simultaneously among the sorted respiratory sounds.

  According to this aspect, since a plurality of respiratory sounds can be output simultaneously (in other words, in a superimposed state) as arbitrary respiratory sounds, desired respiratory sounds can be appropriately combined and output.

<3>
In another aspect of the respiratory sound analyzing apparatus according to the present embodiment, the output means outputs the arbitrary respiratory sound as a voice or a spectrum image.

  According to this aspect, arbitrary respiratory sounds among the sorted respiratory sounds are output as sound using, for example, a speaker or headphones. Alternatively, an arbitrary breathing sound is output as a spectrum image using a display such as a liquid crystal monitor. Therefore, any output breathing sound can be suitably used.

<4>
In another aspect of the respiratory sound analyzing apparatus according to the present embodiment, the respiratory sound analyzing apparatus further includes changing means capable of changing the output state of the arbitrary respiratory sound for each sound type.

  According to this aspect, since the output state of any respiratory sound (for example, output volume, image display mode, etc.) can be changed for each sound type, the output arbitrary respiratory sound can be used more suitably. For example, by increasing the volume of a specific abnormal sound included in the respiratory sound and outputting it, the specific abnormal sound can be easily heard even when a plurality of respiratory sounds are mixed. Moreover, even if it is made easy to listen once and then returned to the normal volume again, it is considered that it becomes easy to listen thereafter. Therefore, it can be effectively used for training of inexperienced doctors.

<5>
In the aspect further including the changing means described above, the changing means may be capable of changing the output volume of the arbitrary respiratory sound for each sound type.

  In this case, since the output volume can be changed for each sound type, the convenience can be improved by making it easy to hear only the desired breathing sound.

<6>
In the above aspect in which the output volume can be changed for each sound type, the changing means may be able to change the output volume of the arbitrary breathing sound for each predetermined frequency band.

  In this case, since the output volume for each sound type can be changed in more detail, for example, only the output volume in a predetermined frequency band can be increased to make it easier to hear a desired breathing sound. The predetermined frequency band may be set according to the characteristics of the sound types to be sorted.

<7>
Alternatively, in an aspect further including a changing unit, the changing unit may be capable of executing predetermined image processing for each sound type on the image indicating the arbitrary breathing sound.

  In this case, a predetermined image processing is performed on an image showing an arbitrary breathing sound (for example, a spectrogram obtained by extracting only one kind of breathing sound), and a state where it can be easily recognized visually can be realized. Examples of the predetermined image processing include color change (for example, RGB adjustment), binarization, edge detection, and the like.

<8>
In the aspect in which the above-described image processing can be performed for each sound type, the changing unit may be able to output an image obtained by performing the predetermined image processing for each sound type by superimposing the plurality of sound types. .

  In this case, since a plurality of images for each sound type separately subjected to image processing can be displayed in a superimposed manner, a plurality of sound types displayed in an appropriate manner by image processing can be recognized together in one image, for example, Comparison between a plurality of sound types can be suitably performed.

<9>
In another aspect of the respiratory sound analysis apparatus according to the present embodiment, the classification unit is an acquisition unit that acquires information about a frequency corresponding to a predetermined characteristic of the respiratory sound spectrum, and is a reference for classifying the respiratory sound. Shift means for shifting a plurality of reference spectra in accordance with information on the frequency to obtain a frequency shift reference spectrum, and the plurality of the plurality of reference spectra included in the respiratory sound based on the respiratory sound and the frequency shift reference spectrum And a ratio output means for outputting a ratio of the reference spectrum.

  According to this aspect, in the classification means, first, information regarding the frequency corresponding to the predetermined characteristic of the spectrum of the respiratory sound is acquired. Here, the “predetermined feature” means a feature that occurs at a specific frequency according to the sound type included in the spectrum of the body sound, and is, for example, a peak that appears in a frequency-analyzed signal. . Furthermore, the “information about the frequency” is not limited to the information that directly indicates the frequency, but includes information that can indirectly derive the frequency.

  When the information about the frequency is acquired, a plurality of reference spectra serving as a reference for classifying the respiratory sounds are shifted according to the information about the frequency, and the frequency shift reference spectrum is acquired. Note that the “reference spectrum” here refers to each sound type in order to classify a plurality of sound types included in the respiratory sound (for example, normal respiratory sound, continuous rarity sound, haircut sound, etc.). This is a preset spectrum. The reference spectrum is frequency-shifted according to, for example, a peak position, which is a predetermined feature acquired from a respiratory sound, and becomes a frequency-shifted reference spectrum.

  When the frequency shift reference spectrum is acquired, a ratio of a plurality of reference spectra included in the respiratory sound is output based on the respiratory sound and the frequency shift reference spectrum. Specifically, the proportion of sound types corresponding to a plurality of reference spectra is calculated in the respiratory sound to be analyzed, and the result is output. More specifically, the ratio of the reference spectrum is calculated as a coupling coefficient by executing a calculation based on a plurality of reference spectra, for example, for the spectrum of the respiratory sound.

  As a result, according to the classification means according to the present embodiment, respiratory sounds including a plurality of sound types can be suitably classified. Particularly in the present embodiment, even when a plurality of breathing sounds are mixed on the same frequency axis, the ratio of each sound type can be suitably classified.

<10>
In the aspect using the frequency shift reference spectrum described above, the predetermined feature may be a local maximum value.

  In this case, for example, a frequency analysis by a fast Fourier transform (FFT) or the like is performed on a signal indicating a respiratory sound, and information on a frequency corresponding to a maximum value (that is, a peak) of the analysis result is acquired. The Note that the information about the frequency is acquired as corresponding to the position of the maximum value, but even if the frequency is not completely coincident with the position of the maximum value, it is acquired as information about the frequency corresponding to the position near the maximum value. Also good.

  As described above, by using the maximum value as a predetermined characteristic of the spectrum of the respiratory sound, information about the frequency can be acquired more easily and accurately.

<11>
The respiratory sound analysis method according to the present embodiment includes a classification process for classifying respiratory sounds into normal sounds and abnormal sounds, and an output process for outputting an arbitrary respiratory sound among the classified respiratory sounds.

  According to the breathing sound analysis method according to the present embodiment, breathing sounds including a plurality of sound types can be suitably analyzed in the same manner as the breathing sound analysis apparatus according to the present embodiment described above.

  In the respiratory sound analysis method according to the present embodiment, various aspects similar to the various aspects of the respiratory sound analysis apparatus according to the present embodiment described above can be employed.

<12>
The computer program according to the present embodiment causes the computer to execute a classification process of classifying respiratory sounds into normal sounds and abnormal sounds, and an output process of outputting arbitrary respiratory sounds among the classified respiratory sounds.

  According to the computer program concerning this embodiment, since the same processing as the breathing sound analysis method concerning this embodiment mentioned above can be performed, the breathing sound containing a plurality of tone types can be analyzed suitably.

  Note that the computer program according to the present embodiment can also adopt various aspects similar to the various aspects of the respiratory sound analyzer according to the present embodiment described above.

<13>
The recording medium according to the present embodiment records the above-described computer program.

  According to the recording medium according to the present embodiment, it is possible to suitably analyze a respiratory sound including a plurality of sound types by causing the computer program described above to be executed by a computer.

  The breathing sound analysis apparatus and breathing sound analysis method according to the present embodiment, the operation of the computer program and the recording medium, and other gains will be described in more detail in the following examples.

  Hereinafter, embodiments of a respiratory sound analysis device, a respiratory sound analysis method, a computer program, and a recording medium will be described in detail with reference to the drawings.

<Overall configuration>
First, the overall configuration of the respiratory sound analyzer according to the present embodiment will be described with reference to FIG. FIG. 1 is a block diagram showing the overall configuration of the respiratory sound analysis apparatus according to this embodiment.

  In FIG. 1, the respiratory sound analysis apparatus according to the present embodiment includes, as main components, a body sound sensor 110, a signal storage unit 120, a signal processing unit 125, a sound output unit 130, and a base holding unit 140. , A display unit 150, an input unit 160, and a processing unit 200.

  The biological sound sensor 110 is a sensor configured to be able to detect a respiratory sound of a biological body. The biological sound sensor 110 includes, for example, an ECM (Electret Condenser Microphone), a microphone using a piezo, a vibration sensor, and the like.

  The signal storage unit 120 is configured as a buffer such as a RAM (Random Access Memory), for example, and temporarily stores a signal indicating a breathing sound detected by the biological sound sensor 110 (hereinafter referred to as “breathing sound signal” as appropriate). To remember. The signal storage unit 120 is configured to be able to output the stored signal to the audio output unit 130 and the processing unit 200, respectively.

  The signal processing unit 125 processes the sound acquired by the biological sound sensor 110 and outputs the processed sound to the audio output unit 130. The signal processing unit 125 functions as, for example, an equalizer or a filter, and processes the acquired sound so that it can be easily heard by a person.

  The audio output unit 130 is configured as, for example, a speaker or a headphone, and outputs a respiratory sound detected by the biological sound sensor 110 and processed by the signal processing unit 125.

  The base holding unit 140 is configured, for example, as a ROM (Read Only Memory) or the like, and stores a base corresponding to a predetermined sound type that can be included in the respiratory sound. The basis according to the present embodiment is an example of the “reference spectrum” in the present invention.

  The display unit 150 is configured as a display such as a liquid crystal monitor, for example, and displays image data output from the processing unit 200.

  The input unit 160 is a device that accepts input from the user, and is configured as, for example, a keyboard, a mouse, a touch panel, various switches, and the like. The input unit 160 is configured to be capable of performing an input operation for selecting at least a breathing sound to be output.

  The processing unit 200 includes a plurality of arithmetic circuits, memories, and the like. The processing unit 200 includes a frequency analysis unit 210, a frequency peak detection unit 220, a basis set generation unit 230, a coupling coefficient calculation unit 240, a signal intensity calculation unit 250, an image generation unit 260, and a breathing sound selection unit 270.

  The operation of each unit of the processing unit 200 will be described in detail later.

<Description of operation>
Next, the operation of the respiratory sound analyzer according to the present embodiment will be described with reference to FIG. FIG. 2 is a flowchart showing the operation of the respiratory sound analysis apparatus according to this embodiment. Here, a simple description for grasping the overall flow of processing executed by the respiratory sound analysis apparatus according to the present embodiment will be given. Details of each process will be described later.

  In FIG. 2, during the operation of the respiratory sound analysis apparatus according to the present embodiment, first, the biological sound sensor 110 detects the respiratory sound, and the processing unit 200 acquires the respiratory sound signal (step S101).

  When the respiratory sound signal is acquired, frequency analysis (for example, fast Fourier transform) is performed in the frequency analysis unit 210 (step S102). Further, the frequency peak detection unit 220 detects a peak (maximum value) using the frequency analysis result.

  Subsequently, a base set is generated in the base set generation unit 230 (step S103). Specifically, the base set generation unit 230 generates a base set using the base stored in the base holding unit 140. At this time, the basis set generation unit 230 shifts the basis based on the peak position (that is, the corresponding frequency) obtained from the frequency analysis result.

  When the basis set is generated, the coupling coefficient calculation unit 240 calculates the coupling coefficient based on the frequency analysis result and the basis set (step S104).

  When the coupling coefficient is calculated, the signal intensity calculation unit 250 calculates the signal intensity corresponding to the coupling coefficient (step S105). In other words, the ratio of each sound type included in the respiratory sound signal is calculated.

  When the signal strength is calculated, the image generation unit 260 generates image data indicating the signal strength. The generated image data is displayed as an analysis result on the display unit 150 (step S106).

  After the analysis result is displayed, when the sound type to be output is input by the user (step S107: YES), the respiratory sound to be output is selected by the respiratory sound selection unit 270, and the selected sound type is the audio output unit 130. Or it outputs to the display part 150 (step S108).

  Thereafter, it is determined whether or not to continue the analysis process (step S109). If it is determined to continue the analysis process (step S109: YES), the process from step S101 is executed again. If it is determined not to continue the analysis process (step S109: NO), the series of processes ends.

<Specific example of respiratory sound signal>
Next, a specific example of the respiratory sound signal analyzed by the respiratory sound analyzer according to the present embodiment will be described with reference to FIGS. FIG. 3 is a spectrogram showing the frequency analysis result of the breathing sound including the haircut sound, and FIG. 4 is a spectrogram showing the frequency analysis result of the breathing sound including the whistle sound.

  In the example shown in FIG. 3, in addition to the spectrogram pattern corresponding to the normal breathing sound, the spectrogram pattern corresponding to the haircut sound that is one of the abnormal breathing sounds is observed. The spectrogram pattern corresponding to the haircut sound has a shape close to a rhombus, as shown in the enlarged portion in the figure.

  In the example shown in FIG. 4, in addition to the spectrogram pattern corresponding to the normal breathing sound, the spectrogram pattern corresponding to the whistle voice which is one of the abnormal breathing sounds is observed. The spectrogram pattern corresponding to the whistle voice is shaped like a swan's neck as shown in the enlarged portion of the figure.

  Thus, a plurality of sound types exist in the abnormal breathing sound and are observed as spectrogram patterns having different shapes depending on the sound types. However, as can be seen from the figure, the normal breathing sound and the abnormal breathing sound are detected in a mixed state. The respiratory sound analysis apparatus according to the present embodiment executes an analysis for separating a plurality of mixed sound types in this way.

<Approximation method of respiratory sound signal>
Next, an analysis method by the respiratory sound analysis apparatus according to the present embodiment will be briefly described with reference to FIGS. FIG. 5 is a graph showing a spectrum at a predetermined timing of the breathing sound including the haircut sound, and FIG. 6 is a conceptual diagram showing an approximation method of the spectrum of the breathing sound including the haircut sound. FIG. 7 is a graph showing a spectrum of a breathing sound including a whistle sound at a predetermined timing, and FIG. 8 is a conceptual diagram showing an approximation method of a spectrum of a breathing sound including a whistle sound.

  In FIG. 5, when a spectrum is extracted at a timing at which a spectrogram pattern corresponding to the haircut sound appears with respect to a respiratory sound signal including the haircut sound (see FIG. 3), the result shown in the figure is obtained. This spectrum is considered to include normal breath sounds and haircut sounds.

  In FIG. 6, the spectrum corresponding to the normal breathing sound and the spectrum corresponding to the haircut sound can be estimated in advance by experiments or the like. For this reason, if a pre-estimated pattern is used, it is possible to know in what ratio the component corresponding to the normal breathing sound and the component corresponding to the haircut sound are included in the above-described spectrum.

  In FIG. 7, when a spectrum is extracted at a timing when a spectrogram pattern corresponding to a whistle voice appears strongly for a respiratory sound signal including a whistle voice (see FIG. 4), the result shown in the figure is obtained. This spectrum is considered to include normal breath sounds and whistle sounds.

  In FIG. 8, the spectrum corresponding to the whistle voice sound can be estimated in advance by experiments or the like in the same manner as the normal breathing sound and the haircut sound described above. For this reason, if a pre-estimated pattern is used, it is possible to know in what ratio the component corresponding to the normal breathing sound and the component corresponding to the whistle sound are included in the above-described spectrum.

  Below, each process for implement | achieving such an analysis is demonstrated more concretely.

<Frequency analysis>
The frequency analysis of the respiratory sound signal and the detection of the peak in the analysis result will be described in detail with reference to FIGS. FIG. 9 is a graph illustrating an example of the frequency analysis method, and FIG. 10 is a diagram illustrating an example of the frequency analysis result. FIG. 11 is a conceptual diagram showing a spectrum peak detection result.

  In FIG. 9, frequency analysis is first performed on the acquired respiratory sound signal. The frequency can be obtained using an existing technique such as fast Fourier transform. In the present embodiment, an amplitude value for each frequency (that is, an amplitude spectrum) is used as a frequency analysis result. In addition, what is necessary is just to determine suitably about the sampling frequency at the time of data acquisition, window size, and a window function (for example, Hanning window etc.).

  As shown in FIG. 10, the frequency analysis result is obtained as being composed of n values. “N” is a value determined by a window size or the like in frequency analysis.

  In FIG. 11, peak detection is performed on the spectrum obtained by frequency analysis. In the example shown in the figure, peaks p1 to p4 are detected at positions of 100 Hz, 130 Hz, 180 Hz, and 320 Hz. The peak detection process may be a simple process because it is only necessary to know at which frequency the peak exists. However, it is preferable that the peak detection parameters are set so that even a small peak is not missed.

  In the present embodiment, a point having a maximum value is obtained, and N points (N is a predetermined value) are detected in order from the smallest second-order differential value (that is, the absolute value is large). . The maximum value is obtained from the point where the sign of the difference switches from positive to negative. The second derivative is approximated by the difference. N values having a value smaller than a predetermined threshold (negative value) are selected in order from the smallest, and the positions are stored.

<Generation of base set>
Next, generation of a base set will be described in detail with reference to FIGS. FIG. 12 is a graph showing the normal alveolar respiratory sound base. FIG. 13 is a graph showing the haircut sound base, FIG. 14 is a graph showing the continuous ra sound base, and FIG. 15 is a graph showing the white noise base. FIG. 16 is a graph showing frequency-shifted continuous ra sound bases.

  As shown in FIGS. 12 to 15, the base corresponding to each sound type has a specific shape. Each base is composed of n numerical values (that is, amplitude values for each frequency) that are the same as the frequency analysis result. Each base is normalized so that an area surrounded by a line indicating the amplitude value for each frequency and the frequency axis becomes a predetermined value (for example, 1).

  Incidentally, here, four bases of a normal alveolar respiratory sound base, a haircut sound base, a continuous ra sound base, and a white noise base are shown, but the analysis can be executed even when there is only one base. In addition, bases other than those listed here can be used. Note that if a base corresponding to a heartbeat sound or an intestinal sound is used instead of the base corresponding to the respiratory sound mentioned here, for example, an analysis of the heartbeat sound or the bowel sound can be performed.

  In FIG. 16, the base corresponding to the continuous rale among the above-mentioned bases is frequency-shifted according to the peak position detected from the result of frequency analysis. Here, an example is shown in which the continuous ra sound base is frequency-shifted in accordance with each of the peaks p1 to p4 shown in FIG. In addition, you may frequency-shift bases other than the base corresponding to a continuous rales.

  As a result, the base set is generated as a set of normal alveolar respiratory sound bases, haircut sound bases, continuous ra sound bases corresponding to the number of detected peaks, and white noise bases.

<Calculation of coupling coefficient>
Next, calculation of the coupling coefficient will be described in detail with reference to FIGS. FIG. 17 is a diagram showing the relationship between the spectrum, the basis, and the coupling coefficient, and FIG. 18 is a diagram showing an example of the observed spectrum and the basis used for approximation. FIG. 19 is a diagram showing an approximation result by non-negative matrix factorization.

  The relationship between the spectrum y, the basis h (f), and the coupling coefficient u, which is the analysis target, can be expressed by the following mathematical formula (1).

図１７に示すように、スペクトルｙ及び各基底ｈ（ｆ）は、ｎ個の値を有している。他方、結合係数は、ｍ個の値を有している。なお、「ｍ」は、基底集合に含まれる基底の数である。

As shown in FIG. 17, the spectrum y and each base h (f) have n values. On the other hand, the coupling coefficient has m values. “M” is the number of bases included in the base set.

  In the respiratory sound analysis apparatus according to the present embodiment, the coupling coefficient of each base included in the base set is calculated using non-negative matrix factorization. Specifically, u that minimizes the optimization criterion function D expressed by the following formula (2) (however, each component value of u is not negative) may be obtained.

なお、一般的な非負値行列因子分解は、基底スペクトルの集合を表す基底行列と、結合係数を表すアクティベーション行列を共に算出する手法であるが、本実施例においては、基底行列を固定して結合係数のみを算出している。

Note that general non-negative matrix factorization is a method for calculating both a base matrix representing a set of base spectra and an activation matrix representing a coupling coefficient. In this embodiment, the base matrix is fixed. Only the coupling coefficient is calculated.

Incidentally, an approximation method other than non-negative matrix factorization may be used as means for calculating the coupling coefficient. However, even in this case, the condition that it is non-negative is desired. In the following, the reason for using the non-negative approximation method will be described with a specific example. As shown in FIG. 18, when the coupling coefficient is calculated by approximating the observed spectrum with four bases A to D think of. Note that the expected coupling coefficient u under the condition of being non-negative is 1 for the base A, 1 for the base B, 0 for the base C, and 0 for the base D. It is 0 to do. That is, on the condition that it is non-negative, the observed spectrum is approximated as a spectrum obtained by adding the base A multiplied by 1 and the base B multiplied by 1.

  On the other hand, the expected coupling coefficient u when the condition is not negative is 0 for the base A, 0 for the base B, 1 for the base C, and 1 for the base D. What to do is -0.5. That is, when the condition is not negative, the observed spectrum is approximated as a spectrum obtained by adding the base C multiplied by 1 and the base D multiplied by -0.5.

  When the above two examples are compared, higher approximation accuracy may be obtained when the non-negative condition is not used than when the non-negative condition is used. However, since the coupling coefficient u here represents the component amount for each spectrum, it must be obtained as a non-negative value. In other words, when the coupling coefficient u is obtained as a negative value, it cannot be interpreted as a component amount. On the other hand, if the approximation is performed under a non-negative condition, the coupling coefficient u corresponding to the component amount can be calculated.

In FIG. 19, in the biological analysis apparatus according to the present embodiment, as described above, a basis set including a normal alveolar respiratory sound base, a hair hair sound base, four continuous rar sound bases, and a white noise base is used. In order to calculate the coupling coefficient u, the coupling coefficient u is calculated as having seven values from u ₁ to u ₇ .

Here, it can be said that the coupling coefficient u ₁ corresponding to the normal alveolar respiratory sound base is a value indicating the ratio of the normal alveolar respiratory sound to the respiratory sound. Similarly, the coupling coefficient u ₂ corresponding to the hair hair base, the coupling coefficient u ₃ corresponding to the white noise base, the continuous coefficient shifted to 100 Hz, the coupling coefficient u ₄ corresponding to the sound base, and the continuous coefficient shifted to 130 Hz. For each of the coupling coefficient u ₅ corresponding to the sound base, the coupling coefficient u ₆ corresponding to the continuous ra sound base shifted to 180 Hz, and the coupling coefficient u ₇ corresponding to the continuous ra sound base shifted to 320 Hz, breathing is also performed. It can be said that the value indicates the ratio of each sound type to the sound. Therefore, the signal intensity of each sound type can be calculated from the coupling coefficient u.

  As described above, in this embodiment, a plurality of sound types included in the respiratory sound are classified using a plurality of bases corresponding to each sound type. However, the classification method described above is merely an example, and a plurality of sound types may be classified using other classification methods.

<Modification of sorting method>
Hereinafter, some examples of the sorting method other than the sorting method using the plurality of bases already described will be described.

<First Modification>
First, the classification method according to the first modification will be described with reference to FIGS. FIGS. 20 and 21 are conceptual diagrams showing a method for separating continuous rales according to the first modification.

  In the classification method according to the first modification, the breathing sound is classified into a continuous ra sound and other sounds. Specifically, when the peak frequency detected from the frequency analysis result of the respiratory sound signal fluctuates within a predetermined range, it is determined that the sound is a continuous rale.

  As shown in FIG. 20, the whistle sound and the like sound that are continuous rales fluctuate so that the peak positions continuously detected on the time axis fall within a predetermined range. In other words, the peak frequency changes so as to have continuity in time. Therefore, when the continuous peak position is within the predetermined range, it can be determined that the sound is a continuous ra-tone.

  On the other hand, as shown in FIG. 21, the sounds other than the continuous rarity fluctuate so that the peak positions continuously detected on the time axis do not fall within a predetermined range. In other words, the peak frequency does not have temporal continuity and changes discretely. Therefore, when the continuous peak position is not within the predetermined range, it can be determined that the sound is not a continuous rarity.

  In addition, the determination result of multiple times can also be used for determination of a continuous rarity. Specifically, when the number of times the peak position continuously detected on the time axis has fluctuated so as to be within a predetermined range has continued for a predetermined number of times, the sound is a continuous rar sound. May be determined.

<Second Modification>
Next, the classification method according to the second modification will be described with reference to FIG. FIG. 22 is a graph showing threshold values used for classification of whistle sounds and similar sounds according to the second modification.

  In the classification method according to the second modification, the continuous rales are classified into whistle sounds and analog sounds. Here, whistle voice sounds and analogy sounds can be distinguished by their pitch (ie, frequency), as whistle voice sounds are called high-pitched continuous rales and analogy sounds are called low-pitched continuous rales. Is possible. However, the peak frequency of the whistle sound and the like sound changes with time. For this reason, when trying to determine a whistle voice and similar sounds using a single threshold for the peak frequency (that is, one threshold whose value does not vary), the determination result changes over time. There is. For example, if the peak frequency changes so as to cross the determination threshold, what has been accurately determined until then will be determined as an incorrect sound type. For this reason, in the second modification, the determination threshold value is varied according to the peak frequency.

  As shown in FIG. 22, in the classification method according to the second modified example, the threshold value varies so that the ratio for determining the whistle sound and the ratio for determining the analogy sound smoothly change according to the peak frequency. For example, when the peak frequency is 200 Hz, it is determined that 7% of whistle sounds are included and 93% of similar sounds are included. When the peak frequency is 250 Hz, it is determined that 50% of whistle sounds are included and 50% of similar sounds are included. When the peak frequency is 280 Hz, it is determined that 78% of whistle sounds are included and 22% of similar sounds are included. The specific numerical values here are merely examples, and different values may be set. Moreover, you may make it have a variation characteristic which changes with sex, age, height, weight, etc. of the biological body which is a measuring object.

  By using the fluctuating threshold described above, it is possible to suitably prevent erroneous determination caused by fluctuations in peak frequency. That is, in the classification method according to the second modified example, the threshold value for determining the whistle voice sound and the analogy sound changes so as to become an appropriate value according to the peak frequency, and thus, for example, a single threshold value that does not change is used. Compared to the case, more accurate separation can be performed.

<Third Modification>
Next, a sorting method according to the third modification will be described with reference to FIGS. FIG. 23 is a graph showing the initial value of the threshold value used for classification of the whistle sound and the analogy sound according to the third modification. FIG. 24 and FIG. 25 are graphs showing the adjusted values of the threshold values used for the distinction between the whistle vocal sound and the analogy sound according to the third modification, respectively.

  The classification method according to the third modified example is also a method of classifying the continuous rar sound into a whistle voice sound and an analogy sound like the second modified example already described. Moreover, it is the same as that of the 2nd modification also about the point determined using the threshold value with respect to the peak frequency obtained from the frequency analysis result.

  As shown in FIG. 23, in the classification method according to the third modified example, the determination result is set to change with a threshold of 250 Hz as a boundary. Specifically, when the peak frequency is 250 Hz or more, it is determined that the continuous rale includes 100% of the whistle sound component and does not include the analogy sound. On the other hand, when the peak frequency is less than 250 Hz, it is determined that the continuous rale includes 100% of the similar sound component and does not include the whistle sound.

  As shown in FIG. 24, in the classification method according to the third modification, when it is determined that the whistle voice component is 100% in the immediately preceding determination, the threshold value is lowered from 250 Hz to 220 Hz. Therefore, it is easy to determine that the whistle voice component is 100%. Specifically, considering the case where the peak frequency is 230 Hz, according to the initial threshold value (see FIG. 23), it is determined as analogy, but according to the adjusted threshold value (see FIG. 24). Judged as a whistle sound.

  As shown in FIG. 25, in the classification method according to the third modification, the threshold value is increased from 250 Hz to 280 Hz when it is determined in the previous determination that the analog sound component is 100%. Therefore, it is easy to determine that the analog sound component is 100%. Specifically, considering the case where the peak frequency is 270 Hz, the whistle sound is determined according to the initial threshold value (see FIG. 23), but according to the adjusted threshold value (see FIG. 25). Judged as roaring.

  By adjusting the threshold value as described above, it is possible to suitably prevent erroneous determination due to fluctuations in peak frequency. That is, in the classification method according to the third modified example, the threshold value for determining the whistle sound and the analogy sound is adjusted to an appropriate value based on the past determination result. Compared with the case of using, more accurate determination can be performed.

  Note that the adjustment of the threshold value may be performed based on a plurality of past determination results, not just the previous determination result. Further, when a plurality of past determination results are used, each determination result may be weighted. For example, weighting may be performed so that the influence becomes smaller as the past determination results. In addition, the smooth threshold value of the second modification may be used as the initial value of the threshold value to be adjusted (see FIG. 22).

<Fourth Modification>
Next, a sorting method according to the third modification will be described with reference to FIGS. FIG. 26 is a spectrogram diagram of a breathing sound including a whistle voice, and FIG. 27 is a graph showing the peak frequency and the number of peaks of the whistle sound. FIG. 28 is a spectrogram diagram of a respiratory sound including a similar sound, and FIG. 29 is a graph showing the peak frequency and the number of peaks of the similar sound.

  The classification method according to the fourth modified example is also a method of separating continuous rales into whistle voices and similar sounds, as in the second and third modified examples already described.

  In FIG. 26, a breathing sound including a whistle voice sound is detected as a spectrum waveform having a predetermined peak. In order to detect the peak frequency F and the peak number N from here, first, a frequency-amplitude graph corresponding to a single time of the spectrum waveform (that is, a region surrounded by a white frame in the figure) is created.

  From the graph shown in FIG. 27, the peak frequency F1 and the peak number N1 of the whistle sound can be detected. It is known that the distribution of the peak frequency of the whistle voice is about 180 to 900 Hz. As can be seen from the figure, the peak number N1 of the whistle voice sound is one.

  In FIG. 28, the breathing sound including the analogy sound is detected as a spectrum waveform having a predetermined peak different from the whistle voice sound. In order to detect the peak frequency F and the number N of peaks from here, a frequency-amplitude graph corresponding to a single time of the spectrum waveform is similarly created.

  From the graph shown in FIG. 29, the peak frequency F2 and the peak number N2 of analogy sounds can be detected. It is known that the distribution of the peak frequency of the analog sound is about 100 to 260 Hz. That is, the peak frequency F2 of the analogy sound is distributed in a region lower than the peak frequency F1 of the whistle sound. Further, as can be seen from the figure, the number N2 of analogy sounds is, for example, three. That is, the peak number N2 of analogy sounds is not one, but a plurality, like the peak number N1 of whistle sounds.

  In the classification method according to the fourth modified example, the determination is performed using the difference in the characteristics of the whistle voice sound and the analogy sound described above. Specifically, based on each of the peak frequency F and the peak number N, the whistle sound and the analogy sound are separated. In this way, for example, more accurate classification can be performed as compared with the case where only the peak frequency F is used to separate the whistle sound and the analogy sound.

<Display of analysis results>
Next, display of analysis results will be described in detail with reference to FIG. FIG. 30 is a plan view showing a display example on the display unit.

  As shown in FIG. 30, the analysis results are displayed as a plurality of images in the display area 155 of the display unit 150. Specifically, the waveform of the acquired respiratory sound is displayed in the area 155a. In the region 155b, the spectrum of the acquired respiratory sound is displayed. In the region 155c, a spectrogram of the acquired respiratory sound is displayed. In the region 155d, a graph representing the time series change of the component amount of each classified sound type (here, five sound types of normal breathing sound, analogy sound, whistle sound, haircut sound, and water bubble sound) is displayed. ing. In the area 155e, the ratio of each classified sound type is displayed as a radar chart.

  Note that such an analysis result display mode is merely an example, and the analysis result may be displayed in another display mode. For example, the ratio of each classified sound type may be displayed as a bar graph or a pie chart, or may be displayed as a numerical value.

<Selection and output of sound types>
Next, the selection of the sound type by the user and the output for each selected sound type will be described with reference to FIGS. FIG. 31 is a spectrogram showing the extraction results for each sound type. FIG. 32 and FIG. 33 are conceptual diagrams showing examples of audio output for each classified sound type.

  As shown in FIG. 31, the spectrogram displayed in the area 155c of the display unit 150 may be displayed for each sound type selected by the user. That is, instead of the spectrogram of the original (acquired original breathing sound) shown in FIG. 31 (a), the spectrogram of normal breathing sound shown in FIG. 31 (b) and the spectrogram of analogy sound shown in FIG. 31 (c). The spectrogram of the whistle voice shown in FIG. 31 (d), the spectrogram of the twisting sound shown in FIG. 31 (e), and the spectrogram of the water bubble sound shown in FIG. 31 (f) may be displayed. Also, a plurality of spectrograms for each sound type may be displayed side by side.

  As shown in FIGS. 32 and 33, a graph for each sound type displayed in the region 155d of the display unit 150 may be selected, and only the selected sound type may be output as a volume. For example, in the example of FIG. 32, only the normal breathing sound is selected, and other analogy sounds, whistle sounds, haircut sounds, and water bubble sounds are not selected. For this reason, only normal breathing sounds are output from the audio output unit 130. In the example of FIG. 33, only normal breathing sound is not selected, and other analogy sounds, whistle voice sounds, haircut sounds, and water bubble sounds are all selected. For this reason, the sound output unit 130 outputs a sound obtained by synthesizing the analogy sound, the whistle sound, the haircut sound, and the water bubble sound.

<Change of output mode>
Next, a method for changing the output mode for each sound type will be specifically described with reference to FIGS. FIG. 34 is a conceptual diagram showing a volume adjustment method for each classified sound type, and FIG. 35 is a conceptual diagram showing a volume adjustment method for each frequency band. FIG. 36 is a conceptual diagram illustrating an example of image processing executed for each sound type, and FIG. 37 is a plan view illustrating an example of an image generated by superimposing images processed for each sound type. . FIG. 38 is a conceptual diagram showing a display color adjustment method for each classified sound type.

  As shown in FIG. 34, the output volume of each sorted sound type may be adjustable for each sound type. In the operation screen shown in the figure, ON / OFF for each sound type can be switched by a check box, and the volume for each sound type can be adjusted by a slider. In the example shown in the figure, a similar sound and a whistle sound are output, and the whistle sound is output at a volume higher than the similar sound.

  As shown in FIG. 35, in addition to the adjustment of the output volume for each sound type, the output volume for each frequency band may be adjusted. In the example shown in the figure, the gain can be adjusted in each frequency band of 125 Hz, 250 Hz, 500 Hz, 1 kHz, and 2 kHz.

  As shown in FIG. 36, image processing (for example, binarization, edge detection, etc.) may be performed on the spectrogram extracted for each sound type. In this way, what is difficult to recognize in the state of being extracted can be displayed in a more visually understandable state. Note that the image processing may be a combination of a plurality of processes. Also, different image processing may be performed depending on the sound type.

  As shown in FIG. 37, an image for each sound type that has undergone image processing (see FIG. 36) may be displayed superimposed. In this way, the spectrogram for each sound type can be collectively recognized with one image, so that visual grasp can be suitably performed. In addition, although the example which displayed the image of normal breath sound and the whistle voice sound superimposed is shown here, the sound type to be displayed can be selected, and only the desired sound type can be appropriately selected and displayed. Can do.

  As shown in FIG. 38, the color of the image may be adjustable for each sound type. In the example shown in the figure, the RGB values can be adjusted for each sound type by adjusting the sliders corresponding to R (red), G (green), and B (blue). In this way, it is possible to display a plurality of sound types in different colors, and it is possible to realize display in a state that is easier to grasp visually.

  As described above, according to the respiratory sound analysis apparatus according to the present embodiment, after the respiratory sounds are separated, they can be appropriately selected and output. Moreover, since the output mode can be changed for each sound type, the data for each sorted sound type can be suitably used.

  The present invention is not limited to the above-described embodiment, and can be appropriately changed without departing from the spirit or idea of the invention that can be read from the claims and the entire specification, and respiratory sound analysis accompanying such changes The apparatus, the respiratory sound analysis method, the computer program, and the recording medium are also included in the technical scope of the present invention.

DESCRIPTION OF SYMBOLS 110 Body sound sensor 120 Signal memory | storage part 125 Signal processing part 130 Audio | voice output part 140 Base holding part 150 Display part 155 Display area 160 Input part 200 Processing part 210 Frequency analysis part 220 Frequency peak detection part 230 Basis set production | generation part 240 Coupling coefficient calculation 240 Unit 250 signal intensity calculation unit 260 image generation unit 270 breath sound selection unit y spectrum h (f) basis u coupling coefficient

Claims (1)

A separation means for separating respiratory sounds into normal sounds and abnormal sounds;
An output means for outputting an arbitrary breathing sound among the sorted breathing sounds.

Images