WO2010044452A1

WO2010044452A1 - Information judgment aiding method, sound information judging method, sound information judgment aiding device, sound information judging device, sound information judgment aiding system, and program

Info

Publication number: WO2010044452A1
Application number: PCT/JP2009/067879
Authority: WO
Inventors: 末治宮原; 千弥喜安; 昭一松永; 雄滝川
Original assignee: 国立大学法人長崎大学
Priority date: 2008-10-16
Filing date: 2009-10-16
Publication date: 2010-04-22
Also published as: JPWO2010044452A1; JP5093537B2

Abstract

When sound such as pulmonary sound is collected and judged, good aiding is provided for making a correct judgment (diagnosis) by making use of the experience of the judgment worker (the doctor in the case of a medical act). The result of determination whether each piece of sound information of the same kind as the objective sound to be examined is a normal or abnormal sound is previously imparted to the objective sound. The sound information is classified depending on the feature thereof and stored in a database. Sound information most similar to the inputted sound information to be judged is sought from among the sound information classified as normal sound and stored in the database, and sound information most similar to the inputted sound information to be judged is sought from among the sound information classified as abnormal sound. Upon the seek, the inputted sound information, the sought normal sound information, and the sought abnormal sound information are almost consecutively outputted.

Description

Information determination support method, sound information determination method, sound information determination support device, sound information determination device, sound information determination support system, and program

The present invention is a sound information determination support method suitable for determining a state based on various sounds such as lung sounds, a sound information determination method, a sound information determination support device, a sound information determination device, and a sound information determination support The present invention relates to a system and a program to which those methods are applied.

2. Description of the Related Art Conventionally, it has been performed to collect sound from an object (subject) to be examined, and to use the collected sound to determine or diagnose the state of the object (target).

For example, in the medical field, when a doctor diagnoses a patient's lung sound using a stethoscope at a medical institution, the stethoscope may be applied to the patient's chest, and the doctor may hear the lung sound directly and hear the lung sound The doctor was trying to determine if he was normal. Such conventional direct listening and judgment methods rely largely on the experience of the doctor, and there is a problem that the diagnosis results vary depending on the ability of the doctor making the diagnosis. On the other hand, various proposals have conventionally been made to take lung sound as data into an apparatus, analyze it, and perform diagnosis automatically.

Patent Document 1 discloses one example of a method for capturing and analyzing heart sounds and lung sounds as data.

JP 2005-296643

As mentioned above, one of the problems in the diagnosis of lung sounds using a conventional stethoscope is that the judgment of whether the lung sounds heard with a stethoscope are abnormal depends on the doctor's own experience. There is a point. That is, if the doctor has many diagnostic experience with lung sound, it can be judged relatively easily whether there is an abnormality in the patient's lung sound based on past experience. However, if the lung sound of a case of a type that the doctor has never heard in the past, or if he has heard the same type of sound in the past, he or she will see a patient with the same case and wait for a long time. If the period has passed, you may not be able to make an accurate judgment as a doctor.

Such a problem is, for example, as described in Patent Document 1, if lung sounds obtained by a stethoscope are taken as data into an information processing apparatus, data analysis is performed in the information processing apparatus, and abnormality determination is performed. Regardless of the accuracy, the device will automatically diagnose, it will not be necessary for the doctor to make a diagnosis, and it will lead to the solution of the problem in that it will not be influenced by the doctor's skill.

However, simply judging by the information processing apparatus does not lead to improvement in the ability of a doctor who makes a diagnosis and is not necessarily preferable. In addition, abnormalities due to lung sounds are assumed in various cases and can not be simply analyzed data, and there are many cases where the doctor's experience is important in order to finally determine the symptoms and disease names.

In addition, although the problem in the case of diagnosing the lung sound has been described so far, the determination result is also obtained in the case of other determination processing that is necessary to collect and determine the sound from the target (subject) to be examined. There is a similar problem that depends on the skill of the judgment worker.

The present invention has been made in view of such a point, and when collecting and judging sounds such as lung sounds, an accurate judgment (diagnosis) utilizing the experience of a judgment worker (a doctor in the case of medical treatment) The purpose is to provide good support.

According to the present invention, sound information of the same type as the inspection object is assigned identification of normal sound or abnormal sound in advance, classification is performed based on the characteristics of sound information, and is stored as a database.
Then, for the input sound information to be judged, among the sound information accumulated in the database, the sound information most similar among the sound information classified as the normal sound is searched and classified as the abnormal sound. The most similar sound information is searched among the different sound information.
When the search is performed, the input sound information of the determination target, the sound information of the searched normal sound, and the sound information of the searched abnormal sound are alternately output.
Alternatively, it is judged whether the inputted sound information of the judgment object is the normal sound or the abnormal sound from comparison between the inputted sound information of the judgment object, the sound information of the searched normal sound, and the sound information of the searched abnormal sound I do.

According to the present invention, since the input sound, the normal sound similar to the input sound, and the abnormal sound similar to the input sound are alternately output and reproduced, the person who makes the determination is Whether the input sound is close to a normal sound or close to an abnormal sound can be judged while comparing it with the sound actually made into a database, and a material for making a judgment can be provided. In addition, it is judged whether the inputted sound information of the judgment object is the normal sound or the abnormal sound by comparison between the inputted sound information of the judgment object, the sound information of the searched normal sound, and the sound information of the searched abnormal sound. By doing this, it is possible to make a highly accurate determination while comparing with the database-ized sound.
For example, by applying the process of the present invention to the diagnosis of lung sounds, a doctor or the like can hear the patient's lung sounds, compare with normal lung sounds close to it, and compare with abnormal lung sounds, and correct diagnosis You will be able to provide the materials you need to Alternatively, it can be automatically determined whether normal lung sound or abnormal lung sound.

In addition, when diagnosing this lung sound, a plurality of lung sounds with high similarity as lung sound of normal sound and / or lung sound of abnormal sound are compared between the lung sound of determination object and the sound information of the database. It is possible to search for the closest case candidate among the plurality of candidate sounds by outputting the plurality of types of lung sounds searched and searched almost continuously with the lung sound of the determination target.

Also, for each lung sound of the abnormal sound stored in the database, when the disease name or disease name corresponding to the lung sound is registered in advance and the retrieved abnormal sound is output, the lung sound of the abnormal sound is applicable. By displaying the symptom name or the disease name, it is possible to support the diagnosis of the symptom name or the disease name.

In addition, the database also registers the sound information of the past lung sound of the same judgment subject, and alternately outputs the inputted lung sound of the judgment target and the past lung sound of the same judgment subject. By doing this, it is possible to understand the change in the lung sound of the person to be judged, and to judge, for example, whether there is no change in the state of the lung, or whether the state of the lung is good or bad.

In addition, the input lung sound information of the determination target is divided into information of a plurality of sections in a predetermined time unit, and the features of the input lung sound information of each of the divided sections are the characteristics of the information accumulated in the database By comparing the abnormal sound and the section in which the highest similarity is detected and displaying the section notifying the section, it can be easily understood which part of the lung sound to be judged has an abnormality. As a doctor, etc. diagnoses, it will be possible to know which part of the lung sound to focus on.

In addition, when dividing into information of a plurality of sections, the frequency spectrum obtained from the information of the plurality of sections is superimposed, and the minimum value at each frequency is determined as a stationary noise component, and the determined stationary noise component By extracting features from the frequency spectrum of each section, stationary noise can be removed well from the input judgment target lung sound, and stationary noise is removed, and accurate search can be performed from the sound. You will be able to do it.

Furthermore, the sound information accumulated as a database includes noise sound information of noise likely to be included when collecting lung sound, and the input sound information is compared with the sound information of noise. By comparing the sound information of the normal sound and the abnormal sound with the exception of the information of the section determined as the noise, noise of a certain fixed pattern such as a sound striking a microphone as a sound input unit A search that effectively eliminates the impact is possible.

Furthermore, by configuring a system including a terminal for transmitting input sound information and a sound information processing apparatus for searching for sound information transmitted from the terminal with a database, for example, a mobile telephone terminal It becomes possible to easily perform judgment support and judgment using various communication terminals.

It is explanatory drawing which shows the example of the lung sound collection state by one embodiment of this invention. It is a block diagram which shows the example of a system configuration by one embodiment of this invention. It is a flowchart which shows the example of a registration process of the identification dictionary by one embodiment of this invention. It is a flowchart which shows the example of the lung sound determination processing by one embodiment of this invention. Lively child care It is explanatory drawing which shows the example of a display of the measurement position by one embodiment of this invention. It is explanatory drawing which shows the example of division | segmentation to the identification area by one embodiment of this invention. It is explanatory drawing which shows the example of calculation of the average spectrum by one embodiment of this invention. It is explanatory drawing which shows the example of the normalization of a spectrum by one embodiment of this invention, and a feature extraction process. It is explanatory drawing which shows the example of selection of the short area frequency spectrum characteristic by one embodiment of this invention. It is explanatory drawing which shows the example of estimation of the spectrum of stationary noise by one embodiment of this invention. It is explanatory drawing which shows the example of a subtraction process of stationary noise by one embodiment of this invention. It is explanatory drawing which shows the example of a waveform of a normal lung sound (a) and an abnormal lung sound (b). It is an explanatory view showing an example of a display by a 1 embodiment of the present invention. It is a block diagram which shows the example of a system by other embodiment of this invention. It is explanatory drawing which shows the process example by another embodiment of this invention. It is a flowchart which shows the process example by further another embodiment of this invention.

Hereinafter, an embodiment of the present invention will be described with reference to FIGS. 1 to 13.

FIG. 1 is a view showing an example of the collection state of the lung sound in the present example. In this example, the lung sound of the subject a (patient) is collected and diagnosed by the lung sound diagnosis support device 20 by the operation of the doctor b. A microphone 11 is attached to the stethoscope 10 possessed by the doctor b, and sounds (lung sounds) picked up by the microphone 11 are input to the lung sound diagnosis support apparatus 20. The stethoscope 10 has a microphone 11 attached near the tip of a stethoscope having a general configuration in which a doctor b listens to lung sounds. Even when lung sounds are collected by the lung sound diagnosis support apparatus 20 It is also possible for the doctor b to hear the lungs as in the case of the medical examination. Instead of the microphone 11, a more sensitive electronic stethoscope using a piezo element or the like may be used.

The lung sound diagnosis support device 20 is constituted by, for example, a computer device and its peripheral equipment, and displays an image indicating a measurement point on the display 21 connected to the computer device. In the computer device, software (program) for functioning as a lung sound diagnostic device is installed. The software is for executing operation processing shown in the flowcharts of FIG. 3 and FIG. 4 described later. For operation, for example, a keyboard 22 connected to a computer device is used. In this configuration, the doctor b looks at the image displayed on the display 21 and performs an input operation of the lung sound. The display 21 displays the waveform of the input lung sound, the characteristics subjected to frequency analysis, and the like.

FIG. 2 is a view showing a configuration example of the lung sound diagnosis support apparatus 20 of the present example. The output signal of the microphone 11 attached to the stethoscope 10 is supplied to the characteristic adjustment unit 24 to perform signal adjustment (correction) processing according to the type of the stethoscope 10 and individual characteristics. As the signal adjustment processing, for example, correction of frequency characteristics is performed. The lung sound signal processed by the characteristic adjustment unit 24 is supplied to the analog-to-digital converter 25 and sampled at a fixed cycle to digitize it. The sampling period is, for example, 0.1 ms.

The digital data converted by the analog-to-digital converter 25 is supplied to the fast Fourier transformer 26 for each predetermined unit of data, and the time axis is converted to the frequency axis by the fast Fourier transform (FFT) operation. The process of converting to the frequency axis is, for example, data of 50 Hz intervals in the range of 50 Hz to 5000 Hz, and data indicating how much the signal component at each frequency position is within a predetermined time. The data subjected to Fourier transform is supplied to the data processing unit 27 and data analysis processing is performed. The analyzed data is stored in the memory 28. The memory 28 also stores, as a database, data of lung sound as a reference, which is necessary to support lung sound diagnosis. Furthermore, the database may store data of lung sound measured in the past of each subject (patient). The process using the data of this past lung sound will be described later (example in FIG. 16). The memory 28 may use other storage means such as a hard disk.

The data analysis process in the data processing unit 27 is executed under the control of the control unit 30. The operation information of the keyboard 22 for performing an operation such as start and stop of lung sound capturing is also supplied to the control unit 30, and the control unit 30 controls each unit based on the operation information.

The lung sound diagnosis support apparatus 20 of the present embodiment includes the display control unit 29, and the display 21 connected to (or incorporated in) the display data generated by the display control unit 29 is the lung sound diagnosis support apparatus 20. Image, etc. can be displayed. The display on the display 21 by the display control unit 29 is also executed under the control of the control unit 30.

In addition, the lung sound diagnosis support device 20 according to the present embodiment includes the sound reproducing unit 31, and outputs the input lung sound and the lung sound searched from the database from the earphone 32 connected to the sound reproducing unit 31. is there. The earphone b is worn by the doctor b shown in FIG. 1 and listens to the reproduced sound. A speaker may be used instead of the earphone 32. The input lung sound and the retrieved lung sound are configured to be output almost continuously and reproduced. The specific output processing state of the lung sound will be described later.

Next, in the lung sound diagnosis support apparatus 20 of the present embodiment, an example of processing for providing support when the doctor b determines the lung sound of the patient a will be described below.
First, the memory 28 stores a database that is learning dictionary data of the lung sound that is the source. This database is formed by the process shown in the flowchart of FIG. 3 based on the control of the control unit 30.
First, lung sound data of various patients are collected as learning dictionary data (step S11). The more lung sound data, the better. For example, data of at least several hundred people are prepared. Then, the doctor performs classification work on each of the collected lung sounds (step S12). Specifically, the doctor listens to each of the collected lung sounds and classifies whether the lungs of a person with normal lungs, the lungs of a person with lung abnormalities or noises. At the time of this classification, whether it is abnormal or normal is judged for each section divided into relatively short predetermined time intervals. The details of an example of determination for each section will be described later.

Then, processing is performed to extract features from the frequency spectrum for each classified lung sound (step S13), classification based on the extracted features is performed, and stored as a database (step S14). As a database of sounds, data of normal lung sounds, data of abnormal lung sounds, and data of lung sounds including noise are prepared. A detailed example of classification when accumulating as a database will be described later.
In addition, when using the lung sound diagnosis support apparatus 20 of the present example for the first time, it is necessary to prepare database data to some extent, but using the lung sound diagnosis support apparatus 20, it is input for determination support. Lung sound data may be left as it is in the lung sound diagnosis support apparatus 20 and used as database data. In this case, the lung sound diagnosis support device 20 is made to input the diagnosis result (discrimination as to whether it is normal or abnormal and the disease name or the disease name) made by the doctor based on the support by the lung sound diagnosis support device 20. Good. Moreover, when using lung sound data input for determination support as database data, identification data is added to identify the subject (patient) of the data, and lung data of the same subject is also added. You may make it compare, when is input.

Next, with reference to the flowchart of FIG. 4, an example of processing to support the determination of the lung sound of the patient in the state where the database is prepared as described above will be described. The processing for performing the determination support of the lung sound of the patient is also executed based on the control of the control unit 30.
First, the lung sound of the patient is input using the stethoscope 10 as shown in FIG. 1 (step S21). As this lung sound, for example, inhalation and exhalation are input at least once, preferably a plurality of times. Here, the presence or absence of designation of the measurement position of the body receiving the input lung sound with the stethoscope 10 is judged (step S22), and if the designation of the measurement position is not given, the stethoscope 10 is operated by the doctor's operation. The information on the measurement position applied to is input and added to the lung sound information (step S23).

Then, a detection process of the breathing cycle is performed for the input lung sound, and the information is classified into the information of the inspiratory period and the information of the expiratory period (step S24). Thereafter, features are extracted from lung sound information, and stationary noise removal processing is performed (step S25).
The feature information of the lung sound from which the stationary noise has been removed is compared with the features of the normal lung sound and the features of the abnormal lung sound and the features of the noise accumulated in the database (step S26). At the time of this comparison, if data of the position at which each lung sound was measured is added, the comparison may be preferentially performed from the lung sound measured at the same position in the database.
Then, in the comparison, it is judged whether or not a component similar to noise is present in the input lung sound (step S 27), and if it is present, for a section where a component similar to the noise is present, It removes from comparison object (step S28).

Then, in the normal lung sounds in the database, if there is a normal lung sound whose characteristics are similar to the input lung sound, it is taken out, and the lung that is input in the abnormal lung sounds in the database If there is an abnormal lung sound whose characteristics are similar to the sound, it is taken out (step S29). If none of them is similar, for example, only the input lung sound is output from the sound reproduction unit 31 and reproduced from the earphone 32. Based on the reproduced lung sound, it is determined whether the doctor b is a normal sound or an abnormal sound (step S31).

When similar normal lung sounds and abnormal lung sounds are extracted from the database in step S29, the extracted normal lung sounds, abnormal lung sounds, and input lung sounds are substantially continuous. Then, the sound is output from the audio reproduction unit 31 and reproduced from the earphone 32.
The order of reproduction may be any order, for example, as an example, reproduction of input lung sounds → reproduction of normal lung sounds in the database → reproduction of abnormal lung sounds in the database, and The lung sounds are alternately reproduced. Or, the input lung sound is reproduced as the reproduction of the input lung sound → the reproduction of the normal lung sound in the database → the reproduction of the input lung sound → the reproduction of the abnormal lung sound in the database The lung sounds may be alternately reproduced. In any case, it is preferable to reproduce in the order in which the differences in the lung sounds are easy to understand, and it is also possible to repeatedly reproduce necessary lung sounds repeatedly by key operation or the like.

At the time of this lung sound output, for example, the waveform and frequency spectrum of each lung sound are displayed on the display 21, and the disease name or disease name added to the abnormal lung sound retrieved from the database and being reproduced is also displayed. May be When displaying the waveform and frequency spectrum of the lung sound, they may be displayed side by side in the same screen so that the waveform and frequency spectrum of each lung sound can be compared.
In addition, the display 21 is also used to indicate whether the lung sound currently being output (reproduced) is the lung sound of the patient who has been input, or whether the normal lung sound or the abnormal lung sound retrieved from the database can be distinguished. Furthermore, when there is an abnormal lung sound similar to the input lung sound, a mark is displayed to indicate which section of the input lung sound is similar to the abnormal lung sound.

In the example of FIG. 4, one normal lung sound having similar characteristics to the input lung sound and one abnormal lung sound having similar characteristics to the input lung sound are extracted and output. Although normal lung sounds and abnormal lung sounds whose characteristics are similar to the input lung sound may be extracted, plural types of them may be extracted and output from the most similar ones. Also in this case, each time one normal lung sound or abnormal lung sound is reproduced, the input lung sound as a comparison target is reproduced to reproduce the input lung sound and the searched lung sound. And may be performed alternately.

Next, an example of details of processing in each step of the flowchart of FIG. 4 will be described in order with reference to FIG.
First, an example of the process of specifying the position at which the lung sound has been measured, which is performed in steps S22 and S23 of the flowchart of FIG. 4, will be described with reference to FIG.
By setting the mode for designating the measurement position of the lung sound, the display 21 displays, for example, a screen for instructing the input of the measurement point shown in FIG.
That is, for example, as shown in FIG. 5A, an image of the human body m1 on the front side in the vicinity of the lungs is displayed, and in the human body m1, six numbers of measurement points P1 to P8 are circled. , And a number corresponding to the current measurement position therein is entered by operating the keyboard 22 or the like. In addition, when the measurement point on the back side is designated, for example, as shown in FIG. 5B, the human body m2 on the back side in the vicinity of the lung is displayed, and 8 measurement points are displayed in the human body m2. P11 to P18 are displayed as circled numbers, and a number corresponding to the current measurement position therein is entered by operating the keyboard 22 or the like.
By designating the measurement position in this manner, the information of the measurement position is added to the input lung sound information. The lung sound information stored as a database may also have information on the position at which the lung sound information was measured. As described above, by having the measurement position information also on the database side, it is possible to preferentially compare the lung sounds measured at the same position and search for candidate lung sounds.

Next, an example of processing for dividing the input lung sound information into identification intervals will be described with reference to FIG.
In step S24 of the flowchart of FIG. 4, the input lung sound data is divided into a period of inspiration and a period of expiration. FIG. 6 shows an example in which the waveform of the input lung sound of one breath is divided into a first half inspiration period and a second half expiration period. The process described in Japanese Patent Application Laid-Open No. 2007-190081, which is an invention by the same inventor as that of the present patent application, is applicable to the process of dividing into an inspiratory period and an expiratory period.

Then, the inside of the inspiratory period and the inside of the expiratory period are divided into a plurality of sections at relatively short time intervals, and processing is performed to calculate an average spectrum of frequency components from each of the divided sections. Here, as shown in FIG. 4, the divided sections are set in a state where they are temporally overlapped with the preceding and following sections. In the following description, the section of one unit is referred to as a discrimination section. The example shown in FIG. 4 shows an example in which it is divided into 15 identification sections (section 1 to section 15) from the period of inspiration to the period of expiration.

FIG. 7 is a diagram showing an example of calculation processing of an average spectrum of frequency components in each identification section.
In the upper waveform diagram of FIG. 7, the vertical axis represents amplitude and the horizontal axis represents time. As shown in the upper part of FIG. 7, the signal waveform in the identification unit extracted in one unit is subjected to fast Fourier transform (FFT) for each data of a predetermined sampling number (for example, 256 samplings), and results of fast Fourier transform In averaging in one identification period, as shown on the lower side of FIG. 7, an average spectrum indicating the power for each frequency is detected. In the lower side of FIG. 7, the vertical axis is power, and the horizontal axis is frequency.

Then, as shown in FIG. 8, a process of normalizing the average spectrum calculated in this way with the average power for each relatively short frequency interval is performed. That is, in the example of FIG. 8, normalization is performed with an average power of 2 to 3 kHz to obtain normalized values r ₀ , r ₁ ,..., R _N-1 every (3000 / N) Hz. N is the number of dimensions of the feature representing the spectrum.
This normalized value is represented by an N-dimensional feature vector represented by the following equation.
v = [r ₀ , r ₁ , ..., r _N-1 ]

A process of comparing similar N-dimensional feature vectors obtained in this manner with those of lung sounds stored in the database is performed. As for the data of each lung sound stored in the database, the sound wave form itself of the lung sound is stored as data, and the N-dimensional feature vector obtained by the same processing is stored as data. The comparison can be made quickly.

In addition, when storing the lung sound data of abnormal sound in the database, a process of preselecting only the identification section in which the abnormal signal component is included in the sound form of one breath consisting of inspiration and expiration is carried out. I am supposed to do it. That is, when the lung sound data taken in as a database is determined as normal / abnormal / noise by the doctor in step S12 of FIG. 3, for example, as shown on the left side of FIG. Also in the sound form of one breath, a normal sound tag is attached to the section of only normal sound, and an abnormal sound tag is attached to the section in which an abnormality appears in the waveform. The section to which the abnormal tag is assigned is a section including a normal sound and an abnormal sound.
Then, the feature vector obtained from the identified section in the section to which the abnormal sound tag is attached is registered in the database as the feature of the abnormal lung sound. The right side of FIG. 9 is a distribution of normal sound and abnormal sound in which the feature vector is represented by the first main component and the second main component. Of the distribution shown in FIG. It is distributed in a specific range.

Next, processing of classifying lung sounds of abnormal sound stored in the database by classification will be described.
Abnormal sounds of lung sounds are classified according to the nature of the sound, such as whether it is continuous or intermittent, or bass or treble. The type of abnormal lung noise that appears in the lungs differs depending on the type of disease. When classifying in the database, as typical abnormal lung sounds, Snoring sound (representative disease: pulmonary emphysema), 2. Hoarse voice (representative disease: bronchial asthma), 3. Torsion (representative disease: pulmonary fibrosis), 4. Blister sound (representative disease: bronchiectasis), 5. Five abnormal lung sounds of frictional noise (representative disease: pneumonitis) are registered in the database. Examples of representative diseases of abnormal lung sounds of the five types are shown below.

1. Snoring sound (bass continuity la sound)
・ Bronchial asthma ・ Occlusive pulmonary disease (lung emphysema, chronic bronchitis)
・ Bronchoectasis ・ Brows retention ・ tracheal ・ Bronchial stenosis

2. Hoarse voice (treble continuity la tone)
・ Bronchial asthma

3. Torsional sound (fine intermittent sound)
・ Pulmonary fibrosis / Idiopathic interstitial pneumonia · Pneumonitis (hypersensitivity, drug property, radiation)
-Mild heart failure-Lung edema initial stage-Pneumonia early stage-Obese panbronchiolitis

4. Water bubble sound (rough intermittent noise)
-Bronchiectasis-Pneumonia-Chronic bronchitis-Emphysema (when infected)
· Heart failure · Advanced pulmonary edema

5. Pleural friction noise · pneumonitis

For example, when searching for abnormal lung sounds whose characteristics are similar to those of the input lung sounds by classifying the abnormal lung sounds into these five types of classification, for example, the abnormal lung sounds of which classification are input lung sounds. It is possible to efficiently search for similar abnormal lung sounds by judging whether it is close to and selecting the most similar abnormal lung sound in the determined classification. These five classifications are one example, and other classifications may be performed.

Next, noise removal processing will be described.
In the present embodiment, it is assumed that stationary noise is removed from both the input waveform data of the lung sound to be diagnosed and the waveform data of the lung sound to be accumulated as a database. Stationary noise here is background noise that appears over the entire section of lung sound data.
10 and 11 show stationary noise removal processing.
FIG. 10 shows the frequency spectrum detected (estimated) from each discrimination interval in one breath, superimposed on one table. When frequency spectra are superimposed as shown in FIG. 10, the frequency spectra of the respective identification sections become similar. Here, in the present example, the minimum value at each frequency position in the superimposed frequency spectrum is set as the spectrum of stationary noise.

After the stationary noise spectrum is set in this way, the stationary noise spectrum is subtracted from the frequency spectrum of each discrimination interval, as shown in FIG. In this way, from the stationary noise spectrum removed, the feature detection processing of each identification section is performed.

Stationary noise is thus eliminated. With regard to noise appearing at a specific location, when a similarity is detected by comparison with data registered as noise in the database, the corresponding identified section is removed from the lung sound data to be compared. This removal process is executed in steps S27 and S28 of the flowchart of FIG.
Also, noise appearing periodically in lung sounds may be determined and removed.

By comparing the feature obtained by performing such noise removal with the database, a normal lung sound similar to the input lung sound in the database and an abnormal lung sound similar to the input lung sound are extracted. FIG. 12 shows the frequency spectrum (a) of the normal lung sound and the frequency spectrum (b) of the abnormal lung sound, which are extracted in the feature comparison.
Thus, the frequency spectrum similar to the input lung sound is detected, and the waveform data of the lung sound having the identification section of the frequency spectrum of FIG. 12A is extracted from the database as the waveform data of normal lung sound, The waveform data of the lung sound having the identification section of the frequency spectrum of FIG. 12B is extracted from the database as the waveform data of abnormal lung sound. The waveform data of the normal lung sound and the waveform data of the abnormal lung sound thus taken out are supplied to the sound reproduction unit 31 (FIG. 2) together with the waveform data of the input lung sound, and are outputted almost continuously from an earphone or the like.

FIG. 13 shows a display example on the display 21 in the case of outputting the normal lung sound, the abnormal lung sound and the input lung sound.
In this example, the upper part of the display screen shows the waveform of the input lung sound and the representative value of the frequency spectrum analyzed from its lung sound waveform.
The middle part of the display screen shows the frequency spectrum of similar normal lung sounds retrieved.
The lower part of the display screen shows the frequency spectrum of the similar abnormal lung sound retrieved. At the lower display position of the abnormal lung sound, the symptom name or disease name added to the lung sound is simultaneously displayed. In addition, as to the normal lung sound and the abnormal lung sound, a lung sound waveform may be displayed as in the case of the input lung sound.
Then, the lung sound currently output from the sound reproduction unit 31 and reproduced from the earphone 32 is displayed so as to indicate which lung sound it is. In the example of FIG. 13, it is indicated that the normal lung sound in the database is being reproduced, and the characters “reproducing” are displayed in the middle.
When a plurality of abnormal lung sound candidates are detected, the plurality of candidates may be simultaneously displayed on the display screen and each may be sequentially reproduced. Even when a plurality of normal lung sound candidates are detected, the plurality of candidates may be simultaneously displayed on the display screen and each may be sequentially reproduced.

When the waveform of the input lung sound is displayed, it is displayed so as to indicate which section in the input waveform the section (identification section) in which the similarity to the abnormal lung sound is detected is identified. That is, in the example of FIG. 13, a section in which the similarity to the abnormal lung sound in the waveform is detected is indicated by a dashed circle x. The frequency spectrum of the section is also shown for the frequency spectrum. Further, although not shown in FIG. 13, if there is data about the position at which each lung sound was measured (that is, the measurement position described in FIG. 5), that position may be displayed.
The order in which the lung sounds are output is, as described above, output in a substantially continuous manner in the order in which the differences in the lung sounds are easily understood.

Thus, while displaying information about each lung sound, the normal lung sound and the abnormal lung sound in which the similarity to the input lung sound and the input lung sound are detected are sequentially output, as shown in FIG. The doctor b who is operating in such a state can compare the lung sound measured from the patient a directly with the similar normal lung sound and abnormal lung sound stored in the database, and based on the comparison, the doctor Can make an accurate diagnosis. That is, even if it is an abnormal lung sound that the doctor has not heard directly from the patient in the past, it can be diagnosed whether it is abnormal (suspected) by contrast with the abnormal sound reproduced from the database, It will be able to properly support the doctor's diagnosis. The case name and disease name are also displayed, which helps the doctor diagnose.
Further, as shown in FIG. 13, a display is made to show the section in which the coincidence with the abnormal lung sound is detected, so it is possible to know which part of the patient's lung sound should be noted and heard. It will be able to properly support the doctor's diagnosis.
In the example described in FIG. 13 and the like, although one normal lung sound closest to the input lung sound and one abnormal lung sound closest to the input lung sound are respectively reproduced, both normal lung sound and abnormal lung sound, or With regard to either one, a plurality of candidate lung sounds such as two or three may be extracted in order from the one closest to the input lung sound, and may be reproduced while displaying the respective information. By increasing the number of candidates in this manner, the material for the doctor to make a diagnosis is increased, which is preferable.

In addition, the lung sound diagnosis support device of the present embodiment can be applied to the training of a doctor. That is, the lung sound of the person corresponding to the patient is input as training, the normal sound and the abnormal sound similar to the lung sound are output, and the doctor is trained so that the doctor can make an accurate diagnosis. It will also be possible.

In the embodiment described above, the lung sound diagnosis support apparatus is configured using a computer device. However, the lung sound diagnosis support apparatus may be configured as a dedicated lung sound diagnosis support apparatus. The software (program) necessary for configuring the lung sound diagnosis support apparatus using an information processing apparatus such as a computer apparatus is stored in various storage media such as a disk and distributed, and the Internet etc. It may be distributed via the Internet.

In the embodiment described above, the lung sound diagnosis support apparatus is prepared as a database. However, for example, an information storage apparatus as one database is provided at a predetermined location where data can be transferred via the Internet or the like. The lung sound diagnosis support device may be prepared to communicate with the information storage device to acquire a lung sound at normal time or a lung sound at abnormal time.

As described above, in the case of a system configuration in which the device for acquiring lung sound and the device on the database side are divided, the device for acquiring lung sound may be a communication terminal such as a mobile phone terminal.
FIG. 14 is a diagram showing an example of a system configuration in which the mobile telephone terminal 40 and the sound information determination support center 50 are divided. The lung sound of the subject c (patient) is input from the microphone 42 attached to the voice input terminal 41 of the mobile phone terminal 40. Then, the communication circuit provided in the mobile phone terminal 40 is connected to the sound information determination support center 50 side, and transmits the input lung sound data to the sound information determination support center 50. The connection between the mobile phone terminal 40 and the sound information determination support center 50 is made by, for example, dialing the telephone line to which the sound information determination support center 50 is connected, and transmitting it over the connected telephone line. Alternatively, a site on the Internet prepared by the sound information determination support center 50 may be accessed by the mobile phone terminal 40 and transmitted via the connected site.

In the sound information determination support center 50, the transmission processing unit 51 for receiving the transmitted lung sound data, the database unit 52 in which the lung sound data is stored as a database, and the data similar to the received lung sound data A search processing unit 53 for searching from a database is provided. The data stored in the database unit 52 is the same as the database stored in the memory 28 by the lung sound diagnosis support apparatus 20 shown in FIG. 2 and includes data of lung sound of normal sound and data of lung sound of abnormal sound. Are classified and stored. The search process performed by the search processing unit 53 is also the same as the search process of similar normal sound and abnormal sound lung sound performed by the lung sound diagnosis support device 20 described above.

When the sound information determination support center 50 searches for lung sounds of normal sound and abnormal sound similar to the transmitted lung sound, the data of lung sound of the searched normal sound and abnormal sound is transmitted to the mobile phone terminal. Send to 40 The cellular phone terminal 40 alternately and continuously reproduces the transmitted lung sound of the normal sound, the lung sound of the abnormal sound, and the input lung sound. This reproduction is output from, for example, the speaker 44 provided in the mobile phone terminal 40 shown in FIG. Further, the display unit 43 of the mobile phone terminal 40 may display information about the lung sound of the normal sound or the abnormal sound that has been output. For example, based on the classification of the detected abnormal sound, the name of a possible disease or the like in the classification may be displayed.

FIG. 15 is a diagram showing the process of the mobile telephone terminal 40 and the sound information determination support center 50 and the transmission state of data in time series. As described with reference to FIG. 15, first, lung sound input processing is performed at the mobile phone terminal 40 (step S41). Then, it is connected to the sound information determination support center 50 by a telephone line or the like, and the lung sound is transmitted (step S42).
The sound information determination support center 50 searches the database using the transmitted lung sound to search for the nearest normal sound and the nearest abnormal sound (step S43). The data of the searched normal sound and abnormal sound are transmitted to the mobile phone terminal 40 (step S44), and the respective sounds are alternately reproduced by the mobile phone terminal 40 (step S45). At the time of transmission in step S44, data of input lung sound may also be transmitted. Alternatively, the input lung sound stored in the mobile phone terminal 40 may be reproduced.

By thus separating the terminal for inputting the lung sound from the device for searching from the database, the device shown in FIG. 1 can be obtained without preparing a dedicated lung sound determination support device. Similar processing is possible.

In the example of FIG. 15, the lung sounds are searched to output similar normal sounds and abnormal sounds, but the sound information determination support center 50 receives the input lung sounds transmitted from the mobile phone terminal 40, Information of the subject of the lung sound may be added and stored in the database. Then, when the input lung sound is transmitted from the mobile phone terminal 40 to the sound information determination support center 50, the stored data of the lung sound of the same person to be measured in the past is sent to the mobile phone terminal 40. The lung sound that is currently input and the past lung sound may be alternately and continuously reproduced. The alternate reproduction of the currently input lung sound and the past lung sound may be performed, for example, by switching between the alternate reproduction of the similar normal sound and the abnormal sound described above and the reproduction mode. Alternatively, after alternately reproducing the currently input lung sound and the past lung sound, it is also possible to alternately reproduce similar normal sound and abnormal sound so that they can be compared with each other.

FIG. 16 is a flowchart showing an example of processing in the case of reproducing an input lung sound and a past lung sound. The sound information determination support center 50 determines whether the input lung sound includes identification information for specifying a subject (step S51). Then, when there is identification information for specifying the subject, it is judged whether or not there is a memory of the past lung sound of the same subject in the database (step S52). Here, when there is a memory of the past lung sound, the past lung sound and the present input lung sound are alternately outputted, and are alternately reproduced by the mobile phone terminal 40 (step S53). In the case where a plurality of lung sounds in the past are registered in a plurality of databases, various settings can be assumed, such as outputting the latest sound or outputting a sound before a predetermined date. If there is no storage of lung sound in the past in step S51, and if there is no identification information for specifying the subject in step S52, the processing here ends.

In the configuration shown in FIG. 14, the sound information determination support center 50 holds and searches the database, but the mobile phone 40 may access and search the data of the database. Further, although the example of FIG. 14 is an example using a mobile telephone terminal, any other communication terminal may be used as long as it is a communication terminal that can be connected to a communication line or the Internet. Alternatively, software for performing the process of the present embodiment may be installed in a personal computer connectable to the Internet to perform the same process as the mobile phone terminal 40.

Also, in each of the processes described so far, normal lung sound and abnormal lung sound most similar to the input lung sound are selected in the device, and it is alternately reproduced with the input lung sound to support diagnosis by a doctor etc. However, based on the detection of the lung sound most similar to the input lung sound, the apparatus may automatically determine whether the input lung sound is a normal lung sound or an abnormal lung sound. When determining whether this input lung sound is normal lung sound or abnormal lung sound, it is determined whether the normal lung sound from which the input lung sound is detected or normal lung sound is closer, and the lung sound determined to be closer to that. If the lung sound is normal, it is diagnosed as normal, and if the lung sound judged to be close is abnormal lung sound, it is diagnosed as abnormal. At that time, it is possible to display the assumed disease name and the like based on the classification of abnormal lung sound and the assumed disease information.

Further, although the embodiment described above is an example configured as an apparatus for supporting a doctor's diagnosis of lung sounds, it may be configured as an apparatus for supporting diagnosis of sounds other than lung sounds. For example, when an abnormality in a structure such as a concrete structure is diagnosed with a sound by tapping the structure, a sound when the internal state is normal and a sound when the internal state is abnormal are displayed. The normal sound and the abnormal sound similar to the sound under diagnosis may be selected from the database, and the selected normal sound and the abnormal sound may be continuously output to the input sound.

In addition to the case where each apparatus described above is configured as a dedicated apparatus, software (program) for performing each process is installed in a general-purpose information processing apparatus such as various computer apparatuses. , And may be configured as an apparatus that performs the same processing. The software is configured as a program that performs, for example, the process illustrated in the flowchart of FIG. 4.

DESCRIPTION OF SYMBOLS 10 ... Stethoscope, 11 ... Microphone, 20 ... Lung sound diagnostic assistance apparatus, 21 ... Display, 22 ... Keyboard, 24 ... Characteristic adjustment part, 25 ... Analog-digital converter, 26 ... High-speed Fourier transformer, 27 ... Data processing Unit 28 28 Memory 29 Display control unit 30 Control unit 31 Audio playback unit 32 Earphone 40 Mobile phone terminal 41 Voice input terminal 42 Microphone 43 Display unit 44 Loudspeaker, 50: sound information determination support center, 51: transmission processing unit, 52: database unit, 53: search processing unit, a: examinee, b: doctor, c: examinee, m1, m2: human body, P1 ~ P8, P11 ~ P18 ... Measurement point

Claims

The sound information of the same type as the judgment object is given in advance a discrimination as to whether it is a normal sound or an abnormal sound, classified based on the characteristics of the sound information and accumulated as a database,
Among the sound information stored in the database for the input sound information to be judged, sound information most similar among sound information classified as normal sound is searched and classified as abnormal sound Search the most similar sound information in the sound information,
A sound information determination support method, which alternately outputs the sound information of the determination target, the sound information of the searched normal sound, and the sound information of the searched abnormal sound.
In the sound information determination support method according to claim 1,
Sound information accumulated in the database is lung sound information,
A sound information determination support method, which alternately outputs the lung sound of the determination target, the lung sound of the retrieved normal sound, and the lung sound of the retrieved abnormal sound.
In the sound information determination support method according to claim 2,
A plurality of lung sounds with high degree of similarity are retrieved as lung sounds of normal sound and / or lung sounds of abnormal sound by comparison between the lung sound of the judgment object and the sound information of the database;
A sound information determination support method comprising: outputting a plurality of types of retrieved lung sounds alternately with the lung sound to be determined.
In the sound information determination support method according to claim 2 or 3,
For each lung sound of abnormal sound accumulated in the database, a disease name or disease name corresponding to the lung sound is registered in advance,
A sound information determination support method, characterized in that when outputting the searched abnormal sound, a disease name or a disease name corresponding to a lung sound of the abnormal sound is displayed.
In the sound information determination support method according to any one of claims 2 to 4,
In the above-mentioned database, it registers also about the sound information of the past lung sound of the same judgment subject,
A sound information determination support method, which alternately outputs the input lung sound of the determination target and the past lung sound of the same determination target person.
In the sound information determination support method according to any one of claims 2 to 5,
Dividing the input lung sound information of the determination target into information of a plurality of sections in a predetermined time unit;
The features of the input lung sound information of each of the divided sections are compared with the features of the information stored in the database, the section in which the highest similarity to the abnormal sound is detected, and the section is notified A sound information determination support method characterized by performing display.
In the sound information determination support method according to claim 6,
The frequency spectrums obtained from the information of the plurality of sections are superimposed, the minimum value at each frequency is determined to be a stationary noise component, and the determined stationary noise component is removed from the frequency spectrum of each section to obtain features. A sound information determination support method, which is extracted and compared with lung sounds stored in the database.
In the sound information determination support method according to any one of claims 2 to 7,
The sound information accumulated as the database includes noise sound information that is likely to be included when collecting lung sounds,
The input sound information is compared with the sound information of the normal sound and the abnormal sound except the information of the section determined as the noise by the comparison with the sound information of the noise. Information judgment support method.
A database in which classification of sound information of the same type as the judgment object is classified in advance based on the characteristics of the sound information, with identification of whether the sound is normal sound or abnormal sound, respectively,
Among the sound information stored in the database for the input sound information to be judged, sound information most similar among sound information classified as normal sound is searched and classified as abnormal sound Search processing means for searching for the most similar sound information among the sound information;
And sound output means for alternately outputting the sound information of the normal sound, the sound information of the abnormal sound, and the input sound information of the determination object, which are obtained by the search processing means. Sound information judgment support device.
A terminal for inputting sound information and transmitting the input sound information;
A sound information determination support system comprising: a sound information processing apparatus that compares and searches sound information transmitted from the terminal and sound information stored in a database and transmits a search result to the terminal;
The sound information processing apparatus is
As the database, sound information of the same type as the determination target is given identification of normal sound or abnormal sound in advance, and classification is performed based on the characteristics of the sound information and accumulated.
For sound information to be judged transmitted from the terminal, sound information most similar among sound information classified as normal sound is searched among sound information accumulated in the database, and abnormal sound Searching for the most similar sound information among the sound information classified as c, and transmitting the search result to the terminal,
The sound information determination support system, wherein the terminal outputs or displays a sound based on the transmitted search result.
In a program installed in an information processing apparatus to execute processing,
An accumulation process in which sound information of the same type as the determination target is previously identified as a normal sound or an abnormal sound, classified according to the characteristics of the sound information, and accumulated as a database.
Among the sound information stored in the database for the input sound information to be judged, sound information most similar among sound information classified as normal sound is searched and classified as abnormal sound Search processing for searching for the most similar sound information among the sound information;
A program executing output processing for alternately outputting the sound information of the determination target, the sound information of the searched normal sound, and the sound information of the searched abnormal sound.
The various sound information of the same type as the sound information of the judgment target is classified in advance based on the respective sound information, and identification information of normal sound or abnormal sound is added to the various sound information subjected to the classification. After, it accumulates as a database,
Of the normal sound and the abnormal sound accumulated in the database, a normal sound and an abnormal sound most similar to the sound information of the judgment target are searched,
A sound information judging method comprising judging whether the sound information of the judgment object is a normal sound or an abnormal sound by comparing the sound information of the judgment object with the most similar normal sound and the abnormal sound.
In the sound information determination method according to claim 12,
The sound information of the judgment target is sound information divided into predetermined sections, and the most similar normal sound and abnormal sound among the normal sound and the abnormal sound accumulated as the database for each of the divided sound information A sound information judging method characterized by searching for and.
The various sound information of the same type as the sound information of the judgment target is classified in advance based on the respective sound information, and identification information of normal sound or abnormal sound is added to the various sound information subjected to the classification. Database to store
Search processing means for searching for the normal sound and the abnormal sound most similar to the sound information to be judged among the normal sound and the abnormal sound accumulated in the database;
A sound determining unit that determines whether the sound information to be judged is the normal sound or the abnormal sound by comparing the sound information to be judged with the most similar normal sound and the abnormal sound; Information judgment device.
In a program installed in an information processing apparatus to execute processing,
The various sound information of the same type as the sound information of the judgment target is classified in advance based on the respective sound information, and identification information of normal sound or abnormal sound is added to the various sound information subjected to the classification. And accumulation processing to be accumulated as a database,
Search processing for searching for the normal sound and the abnormal sound most similar to the sound information to be judged among the normal sound and the abnormal sound accumulated in the database;
A determination process is performed to determine whether the sound information of the determination target is the normal sound or the abnormal sound by comparing the sound information of the determination target with the most similar normal sound and the abnormal sound. program.