JPH09173320A - Voice inspecting device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To easily examine everybody and to especially speedily find out the disease of respiratory organs such as the pharynx or the bronchus. SOLUTION: An A/D converter 5 converts the voice signal of a person to be measured inputted from a microphone 1 to a digital signal and time sequentially stores these signals in a voice data memory 6 later. Next, an FFT operation part 7 performs high-speed Fourier transformation to the voice signal data so that these data can be converted to frequency time sequential voice spectrum data and stored in a voice spectrum data memory 8 frequency- sequentially. A statistical operation part 9 operates standard deviation showing the degree or spread of dispersion and distribution in the voice spectrum data and calculates it as an index value (x). A discriminative evaluation part 10 compares the index value (x) with boundary values (a) and (b) decided based on conventional disease example data or statistical data so that the degree of doubt or possibility of disease at the site of the pharynx/the bronchus can be discriminated and displayed on an LCD display part 13.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a voice inspection apparatus for inspecting a body abnormality, particularly a respiratory disease, based on a voice spectrum distribution.

[0002]

2. Description of the Related Art In recent years, with the increase in so-called adult diseases such as cancer, heart disease, arteriosclerosis, and diabetes, the need for medical treatment, which emphasizes prevention and early detection, rather than treatment, has been emphasized. It has become. In particular, adult diseases are often too late when subjective symptoms occur, so early detection is desired. In particular, even for diseases such as cancer that account for a very high proportion of death and for which the overall picture of the mechanism has not yet been elucidated, it is said that early detection is a major factor in improving treatment results. There is data that the treatment rate for early detection of cancer is almost 100%.

Therefore, conventionally, as a test for early detection and early treatment of diseases, "examination" such as inquiry, visual inspection, auscultation, and palpation by a doctor is started, and periodical inspections are carried out to periodically check the physical condition. There are "regular checkups" and "group checkups" to find hidden signs of illness by collecting and tracking changes in the body's condition. In addition, as a method of examination, "screening examination" (sieving examination for finding signs of a specific disease), which is also used in regular examinations and some medical checkups.
Alternatively, there are “a biopsy” that examines the body itself, such as an X-ray test, and a “specimen test” that examines the excrement of the examinee, such as a blood test and excrement, as an object of examination. In health checkups and medical checkups, more precise inspections of dozens of items are performed by making full use of inspection technology including the latest inspection machines according to the inspection method.

Furthermore, in order to ensure the early detection of diseases and the examination for early treatment, in order to make a detailed examination of each part of the body, particularly the respiratory system, a medical diagnosis or medical examination of the respiratory system is performed. As a testing device for this purpose, it will be used for screening tests and group screenings on physical checkups, starting with the "chest X-ray testing device" that is usually used for regular physical examinations, and for examining the ventilation function of the respiratory system such as vital capacity. Function tester (spirometer, etc.) "or X-ray by injecting a contrast agent containing iodine that does not pass X-rays into the bronchus with a catheter.
There is a "bronchial contrast inspection device" that takes radiographs.

[0005] In addition, a "bronchoscope" for visually observing the details of the bronchus, the function of the nasal cavity as an airway and the lesions of the middle and lower concha, the middle and lower nasal passages, and the nasopharynx are examined. "Posterior laryngoscope", "Indirect laryngoscope" and "Laryngodirective mirror" to put a reflector in the throat and inspect the inside with the naked eye,
There is a "sputum test" in which sputum secreted from the mucous membrane of the bronchus is collected and observed with the naked eye or a microscope for color, stickiness, pathogenic bacteria, pathological cells, and abnormal substances.

[0006]

However, the above-mentioned conventional examination method and examination apparatus have the following problems. (B) In the “medical examination” at hospitals, etc., there were cases where it was too late for early detection because it was performed at the stage when subjective symptoms were already present. (B) Further, in the “regular screening”, it is difficult to perform all the early detection of diseases such as cancer that are difficult to detect, even though major screening tests are performed. (C) In addition, in "Ningen Dock" and "Detailed inspection", it is time-consuming to go out and medical expenses must be paid in full. (D) Since the examination device is expensive and has a large housing, it is difficult to install it in a general home, and further, since specialized knowledge for operation is required, it is not possible to easily perform an examination at home.

[0007] Therefore, an object of the present invention is to provide a voice inspection apparatus which can be easily examined by anyone, and in particular, a respiratory system disease such as the larynx and bronchus can be detected early.

[0008]

In order to achieve the above object, a voice inspection apparatus according to the present invention according to a first aspect of the invention has a voice input means for capturing an input voice as a voice signal of an electric signal and a voice input means for capturing the voice. Frequency spectrum extraction means for extracting the frequency spectrum data of the voice signal, based on the frequency spectrum data extracted by the frequency spectrum extraction means, a judgment evaluation means for judging the presence or absence of disease in the pharyngeal region and bronchial region. It is characterized by having.

In a preferred embodiment, the frequency spectrum extraction means converts the discrete time series data group from the audio signal by a fast Fourier transform for converting the discrete time series data group into the discrete frequency series data group. You may make it extract frequency spectrum data. Further, as a preferable aspect, for example, as described in claim 3, a statistical calculation unit for calculating a deviation indicating a variation in distribution of the frequency spectrum of the frequency spectrum data extracted by the frequency spectrum extraction unit may be provided. Good.

Further, for example, the judging and evaluating means may judge the presence or absence of a disease in the pharyngeal part and the bronchial part based on the deviation calculated by the statistical calculating means. Good. Further, for example, the determination and evaluation means determines “no disease” when the deviation is smaller than the first boundary value, and the second boundary where the deviation is larger than the first boundary value. It may be determined that “disease is present” when the value is equal to or more than the value, and “high possibility of disease” when the deviation is between the first boundary value and the second boundary value. Further, for example, as described in claim 6, at least display means for displaying the judgment result by the judgment evaluation means may be provided.

According to the present invention, after the frequency spectrum data of the voice signal taken in by the voice input means is extracted by the frequency spectrum extraction means, the judgment and evaluation means calculates the disease in the pharyngeal region and the bronchial region based on the extracted frequency spectrum data. The presence or absence of is determined. Therefore, anyone can be examined easily, and in particular, signs of respiratory disease can be detected early.

[0012]

BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, an embodiment of the present invention will be described with reference to the drawings as an embodiment applied to a voice inspection device.

A. Principles of voice inspection Generally, hoarseness or lack of voice is often "larynxitis", i.e. inflammation of the vocal cords,
It is suspected to be the cause of swelling and normal vibration. There are various causes of "larynxitis" such as infection and some kind of irritation, most of which are mild, and are often cured easily by home treatment, but when the hoarseness or lack of vocalization becomes chronic or occurs repeatedly, There may be some serious cause and it is necessary to see a doctor.

For example, in "acute laryngitis", most of the causes are inflammation of the vocal cords and the mucous membrane of the larynx due to infection of the throat due to virus infection.
For example, after a bacterial infection or inhalation of irritating gas, after a loud voice. Most of them are accompanied by subjective symptoms such as cold, cough, and sore throat (occipital region), along with hoarse voice. Usually, the symptoms disappear as the cold disappears, but when a cold occurs, "acute rhinitis" or "acute pharyngitis (throat In many cases, the mucous membrane in the back of the "pharynx" becomes complicated, and in children, the trachea and bronchus become inflamed and breathing becomes difficult.

The causes of "chronic sore throat" include excessive use of voice, smoking, and inhalation of irritating gas. , This may last for months or longer, causing discomfort or mild pain in the larynx. In particular, since symptoms similar to those of chronic laryngitis are also found in cases of malignant tumors of the larynx (laryngeal cancer, etc.), it is necessary to consult a specialist early to clarify the main body of the disease.

Here, FIGS. 1 (a) and 1 (b) are waveform diagrams of a voice spectrum for explaining the principle of voice inspection by the voice inspection apparatus according to the present embodiment. It is known that one of the signs of the above-mentioned person who is strongly suspected of having a malignant tumor of the larynx is that a groaning and throat-growing voice is heard.
When these voices are recorded and the voice spectrum is analyzed, in the case of a normal person, as shown in FIG. 1 (a), as compared with the voices concentrated near the frequency of the generated main voice, the above-mentioned diseases In the case of a person having, as shown in FIG. 1 (b), since the voice spectrum is scattered and distributed to other frequencies, the voices of those frequencies are synthesized,
It is known that the sound is often groaning.

For example, when the vocal cords are permanently deformed due to excessive use of professional voices such as a rogue master, a lecturer, a rakugo performer, etc., the voice quality of the voice is not always the same as that of a malignant tumor, but a cold or If the distribution of the voice spectrum is remarkably large compared to the voice spectrum of a normal person at normal times without any acute illness such as acute laryngitis, some kind of chronic disease or malignancy of the vocal cords or the larynx. Suspected tumor. In particular, for malignant tumors such as cancer, early detection and early treatment are important, and if a highly correlated symptom is found, it is important to promptly see a doctor.

In this embodiment, such a human voice is input and recorded, digital signal processing is performed by D / A conversion and FFT calculation, and a voice spectrum component is analyzed.
In order to determine the distribution variation and spread in the frequency sequence, the deviation (standard deviation) of the voice spectrum is statistically calculated, and the above-mentioned signs of mainly respiratory diseases are found early. Screening (screening)
It realizes a portable medical inspector with a compact size.

B. Configuration of Voice Inspection Device FIG. 2 is a block diagram showing the configuration of the voice inspection device according to one embodiment of the present invention. In the figure, a small microphone 1 is built in the housing or provided so as to protrude outside the housing, converts an input sound into an electric sound signal (analog signal), and supplies the sound signal to a bandpass filter 2. To do. The bandpass filter 2 outputs several tens of K from an audio signal.
Only the voice in the voice band of Hz to several thousand KHz is taken out and supplied to the low frequency amplifier 3. The low frequency amplifier 3 amplifies the filtered audio signal and supplies it to the sample & hold circuit 4. The sample & hold circuit 4 samples the audio signal at a predetermined timing, holds it each time, and supplies it to the A / D converter 5. A / D
The converter 5 converts the sampled audio signal into a digital signal (hereinafter referred to as audio signal data) and supplies it to the audio data memory 6.

The voice signal data memory 6 is shown in FIG.
As shown in, the supplied audio signal data is converted into a time-series audio data group [generally 2 ^M (2
Of M) = N × number of data bits (8 to 16 bits or more)]. The audio signal data is appropriately read out under the control of the control circuit 15 and supplied to the fast Fourier transform calculation processing section (hereinafter referred to as FFT calculation section) 7 and the display control section 11 described later. The FFT calculation unit 7
By FFT calculation, 2 ^M = N time series data groups are
A voice spectrum data memory for converting to frequency time-series voice spectrum data composed of 2 ^M = N real parts (XR (I)) and 2 ^M = N imaginary parts (XI (I)). Supply to 8. Voice spectrum data memory 8
Stores the voice spectrum data in time series as shown in FIG. The audio spectrum data is appropriately read under the control of the control circuit 15 and supplied to the statistical calculation unit 9 and the display control unit 11 described later.

The statistical calculation unit 9 calculates a standard deviation indicating the degree and spread of the distribution variation of the voice spectrum data by a statistical method, and the index value x thereof is judged and evaluated by the evaluation unit 1.
0 and the display control unit 11. Judgment evaluation unit 10
Is, for example, “normal” when the index value x is smaller than a certain boundary value a (x <a), or another boundary value b (b>
When it is larger than a) (x ≧ b), “abnormal” or “suspected of XX”, and when the index value x is between the boundary value a and the boundary value b (a ≦ x <b ), “Boundary area” or “there is a possibility of XX”,
Based on the boundary values a and b determined based on the conventional case data and statistical data, the degree of suspicion or possibility of disease is determined and supplied to the display control unit 11 as the evaluation determination result.

The display control unit 11 displays the above-mentioned index value x, the evaluation judgment result, the spectrum data in the voice spectrum data memory 8 and the time-series voice signal data in the voice signal data memory 6, which is the original data, It is supplied to the LCD drive unit 12 under the control of the control circuit 15. The LCD drive unit 12 drives the LCD display unit 13 under the control of the display control unit 11 to display the index value x by a numerical value, the evaluation determination result by characters or symbols, and the audio signal data in time series. As a waveform graph of, voice spectrum data can be displayed as a frequency series spectrum graph,
The spectrum intensity for each frequency division is displayed on the horizontal axis of the time series as a spectrogram that is superimposed and displayed on the vertical axis.
The LCD display unit 13 is composed of, for example, a dot matrix liquid crystal display device, and displays graphics, characters, numbers and the like by the dot matrix as described above.

Further, the operation input section 14 includes a power switch 14a for turning on / off the power source, and command commands and inputs from an operator for setting the display screen mode, the range of display data, the display magnification and the like. It is composed of various operation switches and input keys that receive data, and supplies the input command command and input data to the control circuit 15. The control circuit 15 is based on the command command and the input data, and the band pass filter 2, the low frequency amplifier 3, the sample and hold circuit 4, the A / D conversion unit 5, the audio signal data memory 6, the FFT 7, the audio, which are described above. Spectral data memory 8, statistical operation unit 9, judgment evaluation unit 1
0, the display control unit 11 and the like are controlled.

Further, in addition to the display output by the LCD display section described above, or in addition to the LCD display section, as shown in the figure, for example, a plurality of LEDs 16a, 16b, which emit different colors such as green, yellow, and red, 16c and the LED 16a,
An LED drive unit 17 for driving 16b and 16c is provided, for example, the green LED 16a is turned on in "normal", the yellow LED 16b is turned on in "boundary", and the red LED 16c is turned on in "abnormal". The inspection determination result may be displayed by a display method that is more easily understood by anyone.

Next, the oscillator 18 generates a clock signal of a predetermined frequency and supplies it to the timing clock generator 19. The timing clock generation unit 19 divides the clock signal to generate a timing clock that defines the operation timing of each unit including the control circuit 15 described above,
Supply to each part. In addition, the power supply circuit (voltage stabilization circuit) 2
0 outputs a stabilized predetermined driving voltage from a small battery 21, such as an AA battery, an AAA battery, or a button type lithium battery, and supplies it to each of the above-mentioned parts. That is, the control circuit 15 and the FFT operation unit 7 store a control program and software for an FFT operation program in a large-capacity ROM, and store a stored program control type CPU (so-called CP) that sequentially processes instructions according to the instructions of the program.
It is configured as a 1-chip microcomputer with a built-in U), is mounted in an LSI integrated circuit that operates with low power, and is mounted, and operates with minute power from the power supply circuit. The interface circuit 22 outputs various data (index value x, evaluation determination result, voice spectrum data and voice signal data) supplied from the control circuit 15 to an external device (not shown) via the external input / output terminals 23a and 23b.

The above-described voice inspection apparatus is used to perform a more detailed voice analysis such as voice pitch (fundamental wave of voice), its temporal change, and time conversion of tracheal transfer function (LPC). It is also possible to add a function of processing the data and outputting it for display.

C. Appearance of Voice Inspection Device FIG. 4 is a schematic diagram showing an example of the appearance of the voice inspection device described above. A handle 30 for holding the device with one hand is provided on one side surface of the housing of the voice inspection device, and various kinds of LEDs 16a to 16c, the LCD display unit 13, and the operation input unit 14 are provided on the front surface of the housing. Operation switches and input keys are provided. Further, on the side surface opposite to the side where the handle 30 is provided, the power on / off switch 14a of the operation input unit 14 and the external input / output terminals 23a, 23a, 23
b is provided. The microphone 1 is provided so as to project to the upper side surface of the housing so that voice can be easily input while the device is held by the handle 30.

D. Next, an operation of the above-described embodiment will be described. Here, FIG. 5 is a flowchart for explaining the operation of the above-described voice inspection device. First, the power switch 14a
When is operated and the power is turned on, in step S10, the band pass filter 2, the low frequency amplifier 3, the sample & hold circuit 4, and the A / D converter 5 are initialized. Next, in step S12, a voice is captured from the microphone 1. At this time, the LCD display unit 13 or LE
You may make it display that voice capture is started by D16a-16c. The sound signal is tens of KHz by the band pass filter 2 and the low frequency amplifier 3.
~ Amplified to a predetermined level and taken out as an audio signal in the audio band of several thousand KHz. The audio signal is sampled by the sample & hold circuit 4 at a predetermined timing and held each time, and by the A / D converter 5,
Converted to digital audio signal data.

The voice signal data is sent to step S14.
The data is stored in the audio data memory 6 in time series. The
The audio data is appropriately read according to the control of the control circuit 15.
Then, in step S16, the FFT operation unit 7
Fast Fourier transform M = N real parts (XR
2 as in (I))^M= N imaginary parts (XI (I)) and
Is converted into time-series speech spectrum data consisting of
It is. The voice spectrum data is stored in step S18.
Stored in the voice spectrum data memory 8 in chronological order
Is done. The voice spectrum data is controlled by the control circuit 15.
It is read out appropriately according to your request, and in step S20,
The statistical calculation unit 9 allows the degree and spread of distribution variation
Is calculated by a statistical method and the
The calculation result is calculated as an index value x.

Next, in the steps shown in steps S22 to S30, the judgment evaluation unit 10 compares the index value x with preset boundary values a and b,
Determine the presence or absence of disease. First, the determination evaluation unit 10 determines whether or not the index value x is smaller than a predetermined boundary value a in step S22. When the index value x is smaller than the predetermined boundary value a (x <a), “normal It is determined that the evaluation determination result is “normal” in step S24.
Display on the LCD display 13, or (further)
The LED 16a (green) is turned on.

On the other hand, when the index value x is equal to or greater than the predetermined boundary value a (x ≧ a), the process proceeds to step S26, and it is determined whether the index value x is smaller than the predetermined boundary value b (b> a). However, when the index value x is smaller than the boundary value b, that is, when the index value x is between the boundary value a and the boundary value b (a ≦ x <
b), it is determined that the "boundary area" or "there is a possibility of XX", and in step S28, the evaluation determination result is "the boundary area" or "there is a possibility of XX".
Display on the LCD display 13, or (further)
The LED 16b (yellow) is turned on.

On the other hand, when the index value x is equal to or greater than the boundary value a and is not between the boundary value a and the boundary value b, that is, the index value x
Is larger than the boundary value b (x ≧ b), it is determined to be “abnormal” or “suspected of XX”, and step S3
0, "abnormal" or "suspected of XX"
Is displayed on the LCD display unit 13, or (further) the LED 16c (red) is turned on.

As a result, the person to be inspected has the LCD display unit 13
Alternatively, the presence or absence of a disease can be confirmed by looking at the LEDs 16a to 16c. In addition, when there is a possibility of illness or when it is determined that there is an abnormality, it is possible to act so as to promptly see a doctor.

E. Modified Example of Appearance Next, FIGS. 6 and 7A and 7B are schematic views showing another example of the appearance configuration of the above-described voice inspection device. The parts corresponding to those in FIG. 3 are designated by the same reference numerals. In FIG. 6, the microphone 1 may be built in the housing in order to further reduce the size of the device, and various operation switches and input keys of the operation input unit 14 provided on the front surface of the housing may be replaced by a numeric keypad. May be Further, the external input / output terminals 23a and 23b may be built in the housing and connected to an external device (not shown) through the slot 24. Further, as shown in FIG. 7A, the LCD display unit 1
3 may be divided into a dot matrix portion for displaying a voice spectrum and a segment portion for displaying character data such as numbers in order to further improve the visibility. FIG.
As shown in (b), as in a microphone having a normal rod-shaped grip portion, in order to be able to use the microphone without a sense of discomfort, the casing has a rod-like shape, and the microphone 1 projects from the tip of the casing. You may make it provide like this.

[0035]

As described above, according to the present invention,
The following effects can be obtained. (1) The device itself can be made inexpensive. (2) Since it can be integrated, the housing can be made small, it has excellent portability, and it is possible to perform a medical examination at home, etc. without going to a hospital. (3) Since the sound is only sampled, no specialized knowledge is required for operation, and it is possible to easily and easily perform an examination by yourself. (4) Therefore, diseases in the pharynx and bronchi can be detected early.

[Brief description of the drawings]

FIG. 1 is a waveform diagram of a voice spectrum for explaining a voice inspection principle of the present invention.

FIG. 2 is a block diagram showing a configuration of a voice inspection device according to an embodiment of the present invention.

FIG. 3 is a conceptual diagram showing a data structure of an audio signal data memory and an audio spectrum data memory of the audio inspection device.

FIG. 4 is a schematic diagram showing an example of an external appearance of the audio inspection device.

FIG. 5 is a flowchart for explaining the operation of the audio inspection device.

FIG. 6 is a schematic view showing another example of the appearance of the audio inspection device.

FIG. 7 is a schematic diagram showing another example of the appearance of the audio inspection device.

[Explanation of symbols]

 1 Microphone (Voice Input Means) 2 Band Pass Filter 3 Low Frequency Amplifier 4 Sample & Hold Circuit 5 A / D Converter 6 Voice Signal Data Memory 7 FFT Operation Unit (Frequency Spectrum Extraction Means) 8 Voice Spectrum Data Memory 9 Statistical Operation Unit (Statistical calculation unit) 10 Judgment evaluation unit (Judgment evaluation unit) 11 Display control unit 12 LCD drive unit 13 LCD display unit (Display unit) 14 Operation input unit 15 Control circuit 16a, 16b LED (Display unit) 17 LED drive unit

Claims (6)

[Claims] 1. A voice input unit for capturing an input voice as a voice signal of an electric signal, a frequency spectrum extracting unit for extracting frequency spectrum data of the voice signal captured by the voice input unit, and the frequency spectrum extracting unit. A voice inspection apparatus comprising: a determination and evaluation unit that determines the presence or absence of a disease in a pharyngeal region and a bronchial region based on the extracted frequency spectrum data. 2. The frequency spectrum extracting means extracts frequency spectrum data from the audio signal by a fast Fourier transform for converting a discrete time series data group into a discrete frequency series data group. 1
The voice inspection device described. 3. The voice inspection apparatus according to claim 1, further comprising statistical calculation means for calculating a deviation indicating a variation in distribution of the frequency spectrum of the frequency spectrum data extracted by the frequency spectrum extraction means. 4. The voice inspection apparatus according to claim 3, wherein the determination evaluation unit determines the presence or absence of a disease in the pharyngeal region and the bronchial region based on the deviation calculated by the statistical calculation unit. . 5. The determination and evaluation means determines “no disease” when the deviation is smaller than a first boundary value, and “disease” when the deviation is a second boundary value larger than the first boundary value or more. 5. The voice according to claim 3, wherein "there is a possibility of disease" is determined when the deviation is between the first boundary value and the second boundary value. Inspection device. 6. The voice inspection apparatus according to claim 1, further comprising a display unit that displays a determination result of the determination evaluation unit.