KR102238602B1 - Body Scanner for Tracing sound from the Internal Organs of a Patient and Detecting its Location of the Sound and Scanning Method thereof - Google Patents

Abstract

Disclosed are a human body scanning device capable of detecting and tracking a location of a sound generated in a patient's body, and a human body scanning method. The scanning device of the present invention includes a plurality of microphones that are densely attached to a certain area of the patient's body, detects sound in the patient's body, and determines the location of the sound in the patient's body through frequency analysis of the audio signal. Recognize it with artificial intelligence. Some of the sound inside the patient can be an important data for diagnosing a patient's illness. With artificial intelligence, it is possible to diagnose a disease and track its location through learning about the frequency characteristics of sound, which are related to the diagnosis of a specific disease, and the pattern in which internal sounds are acquired from a plurality of microphones.

Description

Body Scanner for Tracing sound from the Internal Organs of a Patient and Detecting its Location of the Sound and Scanning Method thereof

The present invention is a device that scans internal organs of the human body such as the chest or digestive tract by detecting the sound in the audible or inaudible band emitted from the patient's body, with artificial intelligence (AI) on the frequency spectrum of the sound generated from the patient's body. The present invention relates to a human body scanning device capable of extracting a sound having a medical problem by analyzing it with an algorithm and confirming its location.

The large intestine, which is a part of the human digestive system, is about 1.5m long, but the small intestine is a fairly long organ of about 6m to 7m. Therefore, in the small intestine, intestinal strictures that narrow the tubes or intestinal adhesions that stick together may occur. Intestinal stricture or adhesion can cause other serious problems while causing intestinal perforation or intestinal obstruction, so it can be an emergency situation requiring surgery or other treatment if necessary.

Even a patient undergoing intestinal stenosis may not feel pain continuously, so it is difficult to recognize the location with a stethoscope and diagnose intestinal stenosis or adhesion. Computed tomography (CT) or magnetic resonance imaging (MRI) is taken with an improved method. Since it is not easy to find, it may cause over-treatment problems due to occasional repetitive imaging.

Meanwhile, there are several types of surgical treatments performed in hospitals, one of which is open surgery performed by opening the patient's abdomen. For example, when a part of the stomach is excised to treat gastric cancer, resection surgery is performed after opening the stomach. After the gastric resection is finished, the internal organs are reorganized, and the stomach is sewn again to complete the operation. Since the small intestine is 6-7m long, even if the doctor has arranged it well, it may happen that a part of the operation is finished in a twisted or bent state and the belly is sewn again. In some patients with very severe bowel kinks or bends, food cannot go down further from the area, and if this condition is maintained for a certain period of time or longer, perforation may occur in the area. If a perforation occurs, you have no choice but to operate again. Intestinal adhesions may already be accompanied when perforation occurs, and if the operation is repeated, intestinal adhesions are more likely to occur.

If, after surgery, it is possible to confirm in advance that there is such a problem in the patient's small intestine, laparoscopic surgery or other relatively low-risk methods can be used without performing laparotomy again. However, as in diagnosing intestinal stenosis or intestinal adhesions, there is currently no easy way to determine the location of a twisted or bent small intestine. As described above, it is often difficult to find intestinal strictures and perforations even with computed tomography or magnetic resonance imaging. Even with an endoscope, you can check from the neck to the stomach and from the rectum to the large intestine, but you cannot see the small intestine.

In addition, sounds from the human body provide various information. For example, a certain audible or inaudible band of sound from the alveoli is related to the condition of the lungs, and some specific sounds may be considered in the diagnosis of lung cancer in some cases as a sign of lung cancer. Even in the case of bronchial asthma, the human body makes special sounds.

An object of the present invention is a device that scans the chest or digestive tract by detecting sound in the audible or inaudible band in the patient's body, by analyzing the frequency spectrum of the sound in the patient's body with an artificial intelligence algorithm. It is to provide a human body scanning device and a human body scanning method capable of extracting medically problematic sounds and confirming their location.

The present invention proposes a method of collecting internal sounds through a plurality of microphones, noting that some of the internal sounds of a patient can be an important means for clinically diagnosing a patient's disease. In particular, using an artificial intelligence engine or algorithm that learns the frequency characteristics of the sound (hereinafter referred to as the sound of interest) that occurs when the patient's internal organs are in a medically problematic state and the pattern of obtaining internal sounds from a plurality of microphones. Thus, we propose a method to track the location by classifying and analyzing the sounds of interest among the sounds of the patient's body.

The human body scanning apparatus of the present invention includes a sensor module, a data processing unit, and an analysis unit to recognize the location of internal sounds generated in the patient's body. The sensor module is attached to the patient's body, and has a plurality of microphones arranged spaced apart from each other to detect sound from the body. The data processing unit converts audio signals generated by the plurality of microphones into digital signals and performs fast Fourier transform to obtain frequency information of the audio signals. The analysis unit identifies and classifies the internal sound generated in the patient's body among the audio signals generated by the plurality of microphones using the frequency information provided by the data processing unit, thereby determining the location of the microphone that detects the internal sound of the human body. It can be recognized as a location where the sound inside the human body is generated.

According to an embodiment, the analysis unit learns, with artificial intelligence, a frequency characteristic of a sound generated from a preset state of a patient's internal organs, and detects an audio signal having the frequency characteristic among audio signals classified as internal sounds of the human body. can do. Here, the analysis unit may acquire the preset frequency characteristics by learning a frequency characteristic of a sound generated inside the human body based on frequency information obtained by fast Fourier transforming a plurality of audio signals as learning data.

According to another embodiment, the analysis unit may recognize a location where the internal sound of the human body is generated by analyzing a pattern in which the same internal sound is collected between the plurality of microphones using artificial intelligence.

According to an embodiment, the sensor module includes: the plurality of microphones of the unidirectional; The plurality of microphones are accommodated in a state that is spaced apart from each other, and includes an adhesive pad portion in a form that can be attached to the body of the patient. Furthermore. It is preferable that the plurality of microphones are spaced apart from each other at intervals in which at least two or more microphones detect the same internal sound inside the human body.

According to an embodiment, the human body scanning apparatus of the present invention may further include a display unit displaying a two-dimensional planar grid map. Here, the plurality of microphones are disposed at positions corresponding to the intersection points of the grid map, and the analysis unit may visually display the positions of the microphones that detect the internal sound of the human body on the two-dimensional plane grid map.

According to another embodiment, the human body scanning apparatus of the present invention includes a plurality of indicator lights attached to each of the plurality of microphones; According to the control of the analysis unit, it may further include a driving unit for turning on the indicator attached to the microphone that has detected the internal sound of the human body among the plurality of indicators.

According to another embodiment, the human body scanning apparatus of the present invention may further include a speaker that outputs the detected internal sound of the human body under the control of the analysis unit.

The classification of sounds inside the human body of the analysis unit can be performed by various algorithms. According to an exemplary embodiment, the generation positions of the plurality of audio signals are compared with each other, and an audio signal whose generation position is outside a predetermined distance from the generation position of other audio signals may be treated as noise. According to another embodiment, the analysis unit includes typical characteristic data of the internal sound of the human body, and converts an audio signal having the same characteristics as the characteristic data as internal sound through spectrum analysis of each of the plurality of audio signals. Can be classified. Here, the characteristic data may be a combination of frequencies peculiar to sound inside the human body.

The present invention also extends to a human body scanning method of a human body scanning device. How to scan the human body,

Attaching a plurality of microphones spaced apart from each other to a scan area to be scanned from the patient's body, and then collecting sound from the patient's body; Converting, by a data processing unit, the audio signals collected by the plurality of microphones into digital signals and performing fast Fourier transform to obtain frequency information of the audio signals; And classifying and classifying the internal sound generated in the patient's body among the audio signals generated by the plurality of microphones using the frequency information, and the location of the microphone that detected the internal sound of the human body It is recognized as the location where the sound inside the human body is generated.

The human body scanning apparatus of the present invention not only detects and presents sounds in audible and inaudible bands generated in the patient's body (abdomen, lungs, bronchi, etc.), but also occurs when the patient's internal organs are in a medically problematic state. Special tracking of sounds (sounds of interest) can help clinically diagnose a patient's illness.

In the human body scanning apparatus of the present invention, in tracking the sound inside the body by attaching a plurality of microphones to the body at the same time, it is possible to track while minimizing errors with the artificial intelligence learning the sound of interest. Through artificial intelligence learning the collected patterns, it is possible to more accurately track the sound of interest and recognize its location.

Since the present invention is a method of collecting and analyzing sound in audible and inaudible bands, a doctor directly performs an image presented by a conventional human body scanning device such as computed tomography (CT) or magnetic resonance imaging (MRI). It is possible to easily trace the disease in its initial state, which is difficult to detect when viewed and analyzed.

1 is a block diagram of a human body scanning apparatus according to the present invention;
2 is a diagram showing an example of a human body scanning apparatus according to the present invention;
3 is a block diagram of a human body scanning apparatus according to another embodiment of the present invention;
4 is a diagram showing an example of a human body internal sound detection map displayed on a display unit;
5 is a flowchart provided to explain the operation of the human body scanning apparatus of the present invention, and
6 is a block diagram of an apparatus for scanning a human body according to another embodiment of the present invention.

Hereinafter, the present invention will be described in more detail with reference to the drawings.

1 and 2, the human body scanning apparatus 100 of the present invention includes a sensor module 110 attached to a patient's body (chest, abdomen, etc.), a data processing module 130, and an analysis module 150. Including, the sound generated from the patient's body (lung, small intestine, bronchi, etc.) (hereinafter referred to as'internal sound') and the'location of the sound' of the sound are recognized. The data processing module 130 may be implemented as a separate configuration from the sensor module 110 attached to the patient's body (chest, abdomen, etc.), or may be implemented integrally as shown in FIG. 2. According to another embodiment, the data processing module 130 and the analysis module 150 may be implemented as a single device, such as a kiosk, and the sensor module 110 may be connected to a kiosk.

As another embodiment, a server that can be accessed through a separate network may be connected to the analysis module 150. At this time, the server (not shown) may perform a function of receiving, storing, and managing the analysis result from the analysis unit 151 described below, and the digital audio signal from the analysis module 150 instead of the analysis unit 151 It is also possible to perform the function of the analysis unit 151 by analyzing the and frequency information.

The sensor module 110 is attached to the patient's body (chest, abdomen, etc.), and has a plurality of microphones 111 for detecting sounds generated from the patient's body (lung, small intestine, bronchi, etc.), and a plurality of It is provided with an adhesive pad unit 113 for accommodating the microphone 111. The sensor module 110 receives operation power from the data processing module 130, and analog audio signals acquired by the plurality of microphones 111 are provided to the data processing module 130.

First, the adhesive pad unit 113 must be in a shape capable of accommodating a plurality of microphones 111 in a state that is spaced apart from each other and attaching to a patient's body. Therefore, it is preferable that the material of the adhesive pad part 113 can be changed in shape such as synthetic resin, fiber, silicone, or the like. As shown in FIG. 2, the adhesive pad unit 113 may be implemented as a single pad accommodating all of the microphones 111, or may be implemented in the form of three or four pads in which the entire microphones 111 are divided and disposed.

Since the present invention uses a method of recognizing the location of the microphone that recognizes the internal sound of the human body as the location of the internal sound of the human body, the microphone 111 is a unidirectional microphone having good sensitivity to sound in one direction. It is better to use it, and the narrower the receiving area in the space is, the better. Considering mounting the sensor module 110 on the patient's body as shown in FIG. 2, it is preferable to be arranged to detect sound coming from a direction perpendicular to the adhesive pad unit 113. Furthermore, in order to prevent external noise from being collected by the microphone 111 as much as possible, even if the microphone 111 is accommodated in the adhesive pad unit 113, the sound collecting surface of the microphone 111 is attached to the patient's body. I like the form that I can do.

On the other hand, the sound collected by the microphone 111 can include both audible and non-audible sounds, so it can process audio signals in the audible band 2㎐ ~ 2㎑ and non-audible band 2㎑ ~ 80㎑. It is good to have a bandwidth that is available.

Since the present invention uses a method of recognizing the location of the microphone 111 that has collected the internal sound of the human body as the location of the internal sound of the human body, the sensor module 113 can cover the scan area or the target range of the human body. It is necessary to have as many microphones 111 as there are. Therefore, hereinafter, for convenience of description, the individual microphone 111 is referred to as a “scan cell”.

For example, in the case of diagnosing intestinal stricture or intestinal adhesion, the microphone 111 should be attached to a plurality of points in the abdominal region of the patient or in a narrower small intestine region, and the thickness of the small intestine and the general small intestine in the abdomen. It is placed in consideration of the location and the like. In addition, in the case of lung diagnosis, the arrangement and position of the microphone 111 must be changed accordingly. Accordingly, the total size of the sensor module 110 or the number of microphones 111 may vary according to the attachment position.

Since the small intestine is a digestive tract that is relatively narrow and quite long, but is not unfolded and is kept in a clustered state, the more densely arranged the scan cells are, the better the more densely arranged the microphone 111 is. It is preferable that the internal sound generated at one point is disposed at intervals such that at least two or more microphones 111 can detect the sound inside the human body. Considering these circumstances, the sensor module 110 may include dozens of microphones 111. In FIGS. 1 and 3, a plurality of microphones 111 are spaced at equal intervals and disposed in a two-dimensional matrix form to cover the abdomen (or small intestine area) of a patient. Of course, the plurality of microphones 111 may be spaced apart from each other at different intervals according to the shape and arrangement of human organs obtained through experiments or statistics.

The sensor module 110 detects sound from the body and provides it to the data processing module 130. If necessary, it may be a good idea to collect sound in a condition that requires an artificial motion from the patient. For example, in the case of diagnosing intestinal stricture, it is good for diagnosing intestinal stricture to detect the internal sound of the human body when the food (or digested food) passes through the digestive tract after allowing the patient to swallow. Although the sensor module 110 is attached to the patient's body and each microphone 111 is unidirectional, external noise other than the sound inside the body can be collected, and the sound from the organ to be collected among the sounds inside the body Sounds made by organs other than those can be collected.

The plurality of microphones 111 accommodated in the adhesive pad unit 113 must be attached to or in close contact with a plurality of spaced apart points on the patient's body.

The data processing module 130 supplies operating power to the sensor module 110, converts internal sound detected by the sensor module 110 into a digital signal, extracts frequency information, and provides it to the analysis module 150. To this end, the data processing module 130 includes a connection unit 131, a communication interface 133, a power supply unit 135, a data processing unit 137, and a control unit 139. In the process of implementing in hardware, the data processing unit 137 and the control unit 139 may be implemented as separate dedicated chips, but may be implemented by combining an audio IC, a DSP chip, and an MCU.

The power supply unit 135 supplies operating power to the sensor module 110 as well as the data processing module 130. Power supply to the plurality of microphones 111 directly connects the output terminal of the power supply unit 135 to each microphone 111 through serial and/or parallel connection. In the embodiment of FIG. 3 described below, the power supply unit 135 also supplies power necessary for lighting the indicator 311.

The connection unit 131 provides the analog audio signals acquired by the plurality of microphones 111 to the data processing unit 137. The analog audio signals collected by the plurality of microphones 111 are provided to the data processing unit 137 through individual channels to be distinguished from each other. However, since it is difficult for the general data processing unit 137 to have dozens of input channels (or input terminals), the connection unit 131 is a data processing unit by muxing the analog audio signals provided by the plurality of microphones 111. Provided to 137, the data processing unit 137 demuxes the muxed signal and classifies it as an audio signal of an individual channel. If the data processing unit 137 has dozens of channels (or input terminals), the connection unit 131 is not required.

The communication interface 133 is a communication means for connecting the data processing unit 137 and the analysis unit 151, and the digital audio signals of a plurality of channels obtained by the data processing unit 130 and frequency information of the digital audio signals (FFT The execution result) is provided to the analysis module 150. The communication interface 133 is used when the data processing module 130 and the analysis unit 151 are implemented as separate devices. Therefore, the communication interface 133 is not an essential component of the present invention, and may be unnecessary when the data processing module 130 and the analysis unit 151 are integrally implemented. The communication interface 133 may use a wired interface or a wireless interface as shown in FIG. 2.

The data processing unit 137 converts an analog audio signal for each channel input through the connection unit 131 into a digital signal and provides it to the analysis unit 151. The data processing unit 137 may perform basic noise filtering to remove noise such as white noise or sound less than a preset size.

The data processing unit 137 includes an ADC 141, a noise filter unit 143, and an FFT processing unit 145.

The ADC 141 samples and quantizes the analog audio signal provided for each microphone 111 through the connector 131 and converts it into a digital signal. When the band of the microphone 111 is set to a maximum of 80 kHz, the sampling frequency should be approximately twice as high as 160 kHz according to the Nyquist theorem. The noise filter unit 143 removes primary noise such as white noise from the converted digital signal.

The FFT processor 145 extracts frequency information of the audio signal for each scan cell by performing Fast Fourier Transform (FFT) to analyze the frequency spectrum of the digital audio signal for each channel.

The control unit 139 controls the overall operation of the data processing module 130 and may be generally implemented as an MCU. According to the request of the analysis module 150, the controller 139 collects analog signals for scanning, converts them to digital signals, and provides data to the analysis module 150 as a whole.

The analysis module 150 shown in FIGS. 1 and 3 is separated from the data processing module 130 and connected through a communication channel. However, as described above, the analysis module 150 and the data processing module 130 may be integrally implemented.

The analysis module 150 may recognize a location where sound inside the human body is generated using digital audio signals and frequency information provided by the data processing module 130 and display the location. To this end, the analysis module 150 includes an analysis unit 151 and a display unit 153. Hereinafter, the operation of the analysis unit 151 of the present invention will be described with reference to FIGS. 1 to 5.

<Collection of sound: S501>

The sensor module 110 collects sound from the patient's abdomen using a plurality of microphones 111, and the data processing unit 137 converts the audio signal provided through the plurality of channels into a digital signal, and performs fast Fourier transformation. (FFT) is performed to obtain frequency information. The digital audio signal for each channel acquired by the data processing unit 137 and its frequency information are provided to the analysis unit 151. This process is performed by the control of the controller 139, and the controller 139 may receive a control command directly from the user or may receive a control command from the analysis module 150.

<Internal sound classification and location recognition: S503, S505>

The analysis unit 151 classifies and classifies the internal sound of the human body among digital audio signals of a plurality of channels (ie, a plurality of microphones) provided by the data processing unit 137. The sound collected by the sensor module 110 may include various noises other than the sound inside the human body, and in some cases, there may be no sound inside the human body and only other noises may be collected.

A specific method of classifying internal human sounds by the analysis unit 151 may be implemented with various algorithms. For example, (1) the analysis unit 151 may classify all sounds equal to or larger than a preset size as internal human sounds. The human body scanning apparatus 100 of the present invention is an auxiliary device for diagnosing intestinal stricture or lung cancer, and the final diagnosis is in the area of a doctor or medical professional, so it may be necessary to recognize all sounds of a certain size or more and display the location thereof. .

(2) As another example, the analysis unit 151 may analyze frequency information provided by the data processing module 130 using artificial intelligence. To this end, the analysis unit 151 is equipped with an artificial intelligence algorithm or engine, and learns the frequency characteristics of the internal sound of the human body tracked by the scanning device 100 of the present invention. Among the internal sounds of the human body, it is good to learn the frequency characteristics of sounds that occur when the patient's internal organs are in a medically problematic state (hereinafter, referred to as'the sound of interest'). For example, learn the characteristics of sounds related to intestinal stenosis or intestinal adhesions, sounds from the alveoli related to lung cancer, or sounds from the bronchi related to asthma.

The analysis unit 151 analyzes the frequency information provided by the data processing module 130 using an artificial intelligence algorithm to classify audio signals having similar or identical characteristics as sounds of interest, and audio signals without similar or identical characteristics. Can be classified as noise.

If there is sound inside the human body among the digital audio signals of a plurality of channels provided by the data processing unit 137, the analysis unit 151 determines the location (that is, the location of the microphone) at which the signal classified as the sound inside the body is collected. It can be recognized as the location of the sound inside the human body.

<Indicating the location of sound inside the human body: S507>

The analysis unit 151 may display a location at which the audio signal classified as internal sound inside the human body is collected to a manager or a patient. Here, the analysis module 150 can be displayed in various ways. First of all, the purpose of the display of the analysis unit 151 is to display the location of the sound inside the human body.

For example, if a patient traces the internal sound of the body after ingesting food, it tracks where the food is passing through the digestive tract. If there is an intestinal stenosis or intestinal adhesion, it is indicated that the movement of the food is no longer progressing. Therefore, doctors may suspect that there is a bowel twist, bend, bowel stricture, or bowel adhesion in the part if the sound inside the human body stops while progressing. Even in the case of lung cancer, if an abnormality occurs in a specific part of the lung, a minute sound may be generated in the part during the breathing process, and the location of the sound is recognized.

Therefore, the analysis module 150 must display a location (ie, a location where a human body internal sound is generated) at which signals classified as internal human sounds are collected. As one of the methods, as in FIGS. 1 to 3, the analysis module 150 may include a display unit 153. The display unit 153 is a general display device such as an LCD, an LED, an OLED, and a picture tube. The analysis unit 151 may display a matrix map or a grid map as shown in FIG. 4 on the display unit 153. 4 is an example of a screen displayed on the display unit 153 in the case of a digestive tract scan. Marks (including a1, a2, a3, a4, a5) corresponding to the plurality of microphones 111 of the sensor module 110 are displayed at the intersections of the matrix. In other words, each mark (including a1, a2, a3, a4, a5) corresponds to a scan cell. The analysis unit 151 displays the location of the microphone 111 that has detected the sound inside the human body on a two-dimensional plane grid map. Referring to FIG. 4, 12 marks (a1, a2) are displayed in a lit state, and the 13th mark (a3) is displayed as blinking, indicating that the last internal sound of the human body has been acquired at the current 13th mark (a3) position. I'm doing it. Thus, the doctor can confirm that the food is moving in the direction of the arrow. In the case of a gastrointestinal tract scan, sound inside the human body captured at a location not included in a series of connected flows (eg, a4, a5) can be classified as noise.

3 illustrates a human body scanning apparatus 300 according to another embodiment of the present invention. The scanning device 300 of FIG. 3 is the same device as the scanning device 100 of FIG. 1, except that the sensor module 310 further includes a plurality of indicator lights 311 along with a plurality of microphones 111, The data processing module 330 further includes a driving unit 331 for driving the indicator 311. The power supply unit 135 also supplies power necessary for lighting the indicator 311.

The indicator 311 is attached to each of the plurality of microphones 111 mounted on the scan module 310. The driving unit 331 turns on the indicator 311 attached to the microphone 111 where the internal sound of the human body is detected under the control of the analysis unit 151, so that the position where the current internal sound is detected is directly on the patient. Can be displayed. Since the position displayed on the display unit 153 does not indicate a specific position in the patient's body, the position displayed on the display unit 153 may be displayed on the patient through the indicator 311. The indicator 311 may use an LED or other similar type of lamp.

The driving unit 331 is a switch block connecting between the power supply unit 135 and the plurality of indicator lights 311, and individually regulates the supply of operating power to the plurality of indicator lights 311. In this case, the driving unit 331 may open a specific switch according to a separate control signal so that only the specific indicator 311 is turned on. When the analysis unit 151 recognizes the position where the sound inside the human body is detected, the corresponding position is transmitted to the driving unit 331 through the data processing unit 137.

According to another embodiment, the human body scanning apparatus 300 of the present invention may further include a speaker 155 that outputs the detected internal sound of the human body under the control of the analysis unit 151. An experienced doctor may use the sound output through the speaker 155 to distinguish whether the sound is internal to the human body.

In the above manner, the human body scanning apparatus of the present invention operates.

Example 1

In FIGS. 1 and 3, one data processing module 130 and a sensor module 110 are connected to the analysis module 150. However, according to another embodiment, a plurality of data processing modules may be connected to the analysis module 150. In the scanning device 600 of FIG. 6, a plurality of data processing units 531, 533, and 535 are connected to the analysis module 150, and sensor modules 511 and 513 are individually connected to each of the data processing units 531, 533, and 535. , 515) is connected. Although the plurality of data processing units 531, 533, and 535 and the sensor modules 511, 513, and 515 are not premised on any connection or combination with each other, they are usually used to scan one patient.

The sensor modules 511, 513, and 515 are examples in which one sensor module 110 is implemented as a plurality of pads as described in the embodiments of FIGS. 1 and 3. Therefore, the configuration or operation is the same as the sensor module 110. However, the number of microphones 111 provided in the individual sensor modules 511, 513, and 515 is only smaller than that of the sensor modules 110 of FIGS. 1 and 3.

The data processing units 531, 533, and 535 distribute the roles of the data processing unit 130 of FIGS. 1 and 3. The internal configuration of each of the data processing units 531, 533, and 535 is the same as that of the data processing unit 130, but the number of processed audio channels is small. Accordingly, the data processing units 531, 533, and 535 may not need the connection unit 131.

The analysis unit 551 is connected to a plurality of data processing units 531, 533, and 535, and a description of a method for classifying sounds inside the human body and displaying a sound generation location is provided with the analysis unit 551 of FIGS. 1 and 3 same.

Example 2

Previously, the recognition by artificial intelligence in the sound location recognition process in step S503 was described. Analysis (or learning) by artificial intelligence includes (1) first, a process of learning the characteristics of internal sounds of the human body, and (2) patterns in which internal sounds of the human body are collected between a plurality of scan cells of the sensor module 113 May include learning about it. Furthermore, the learning of (1) and (2) above may be limited to the sounds of interest that can help in clinical diagnosis of the patient's disease, among the internal sounds of the human body. Since the scanning apparatuses 100, 300, and 600 of the present invention use sound in an audible and inaudible band, a conventional visual human body scanning apparatus such as computed tomography (CT) or magnetic resonance imaging (MRI) is used. It is especially helpful in diagnosing diseases that are not easy to detect.

<Learning the characteristics of sounds inside the human body>

Among the various sounds produced by the human body, it is to learn the frequency characteristics of the sound inside the human body. Since such learning is a learning about the frequency characteristics of the sound, it is premised that an artificial intelligence algorithm or an artificial intelligence engine obtains frequency information by performing Fast Fourier Transform (FFT) on the audio signal, which is the training data.

<Collection pattern of internal human sounds between scan cells>

The sensor module 113 includes a plurality of microphones 111, and each microphone 111 becomes a scan cell. In other words, the location of each scan cell is a very important factor in the operation of the scanning apparatuses 100, 300, and 600 of the present invention. The internal sound of the human body can be transmitted in various directions depending on the size of the tissue that is making the sound, the volume of the sound, and the type and shape of the medium that transmits the sound, and the range and collection time of the sound can be changed. Although it will vary depending on the layout and complexity of the scan cells, several scan cells can collect the same sound at the same time or at a time difference. In summary, the pattern in which the internal sound of the human body is collected between the plurality of scan cells can also be an important factor to distinguish the internal sound of the human body from the noise. I can.

For example, no matter what shape the small intestine is arranged in the stomach of the patient, since it is one connected tube, it is captured as a single flow from a plurality of microphones 111 arranged in the form of a matrix or array. Therefore, sounds inside the human body that are not included in a series of connected flows can be classified as noise. For example, when comparing generation positions of a plurality of audio signals with each other, an audio signal whose generation position is outside a predetermined distance from the generation position of other audio signals may be treated as noise.

The analysis units 151 and 551 analyze frequency information provided by the data processing modules 130, 330, 531, 533, 535 based on (1) and (2) learning, and provide audio signals having similar or identical characteristics. Is classified as internal sound, and audio signals that do not have similar or identical characteristics can be classified as noise.

In the above, preferred embodiments of the present invention have been illustrated and described, but the present invention is not limited to the specific embodiments described above, and is generally used in the technical field to which the present invention pertains without departing from the gist of the present invention claimed in the claims. Of course, various modifications may be made by those skilled in the art, and these modifications should not be understood individually from the technical idea or perspective of the present invention.

Claims (14)

As a human body scanning device,
A sensor module attached to the patient's body and having a plurality of microphones disposed to be spaced apart from each other to collect sound from the body;
A data processing unit that converts the audio signals collected by the plurality of microphones into digital signals and performs fast Fourier transform to obtain frequency information of the audio signals; And
A microphone that detects the internal sound of the human body, including an analysis unit for classifying and classifying the internal sound generated in the patient's body among audio signals generated by the plurality of microphones using frequency information provided by the data processing unit Recognize the location of the body as the location where the sound inside the human body is generated,
The analysis unit,
Extracting the generation position of the plurality of audio signals, and processing an audio signal whose generation position among the plurality of audio signals is outside a predetermined distance from the generation position of other audio signals and is not included in a series of connected flows as noise A human body scanning device, characterized in that.
The method of claim 1,
The analysis unit,
A human body scanning apparatus, characterized in that: learning a frequency characteristic of sound generated from a preset state of a patient's internal organs by artificial intelligence, and detecting an audio signal having the frequency characteristic among audio signals classified as internal human sounds.
The method of claim 2,
The analysis unit,
A human body scanning apparatus, characterized in that, based on frequency information obtained by fast Fourier transforming a plurality of audio signals, which are training data, a frequency characteristic of a sound generated inside a human body is learned to obtain the preset frequency characteristic.
The method according to any one of claims 1 to 3,
The analysis unit,
A human body scanning apparatus, characterized in that the human body scan apparatus recognizes a location where the internal body sound is generated by analyzing a pattern in which the same internal sound is collected between the plurality of microphones with artificial intelligence.
The method of claim 1,
The sensor module,
The plurality of microphones of the cardiogram; And
A human body scanning apparatus comprising: an adhesive pad unit that accommodates the plurality of microphones in a state that is spaced apart from each other and is attached to the body of a patient.
The method of claim 5,
Wherein the adhesive pad unit accommodates the plurality of microphones by being spaced apart from each other at intervals in which at least two or more microphones detect the same internal sound inside the human body.
The method of claim 1,
Further comprising a display unit for displaying a two-dimensional plane grid map, wherein the plurality of microphones are disposed at a location corresponding to an intersection point of the grid map,
And the analysis unit displays the location of the microphone that has detected the sound inside the human body on the two-dimensional plane grid map.
The method of claim 1,
A plurality of indicator lights attached to each of the plurality of microphones; And
And a driving unit for lighting an indicator attached to a microphone that has detected the sound inside the human body from among the plurality of indicators under the control of the analysis unit.
The method of claim 1,
And a speaker for outputting the detected internal sound of the human body under the control of the analysis unit.
delete As a human body scanning method of a human body scanning device,
Attaching a plurality of microphones spaced apart from each other to a scan area to be scanned from the patient's body, and then collecting sound from the patient's body;
Converting, by a data processing unit, the audio signals collected by the plurality of microphones into digital signals and performing fast Fourier transform to obtain frequency information of the audio signals; And
Including the step of classifying and classifying the internal sound generated in the body of the patient among the audio signals generated by the plurality of microphones using the frequency information, the location of the microphone that detected the internal sound of the human body Recognized as the location of the sound inside the human body,
The analysis unit,
Extracting the generation position of the plurality of audio signals, and processing an audio signal whose generation position among the plurality of audio signals is outside a predetermined distance from the generation position of other audio signals and is not included in a series of connected flows as noise Human body scanning method, characterized in that.
The method of claim 11,
The classifying step,
Learning the frequency characteristics of sound generated from a preset state of the patient's internal organs with an artificial intelligence algorithm, and the artificial intelligence algorithm detecting an audio signal having the frequency characteristic among audio signals classified as internal sounds of the human body How to scan the human body.
The method of claim 12,
The classifying step,
A human body scanning method, characterized in that, based on frequency information obtained by fast Fourier transforming a plurality of audio signals, which are training data, the frequency characteristics of sounds generated inside the human body are learned to obtain the preset frequency characteristics.
The method according to any one of claims 11 to 13,
The classifying step,
The human body scanning method, characterized in that by analyzing a pattern in which the same internal sound inside the human body is collected between the plurality of microphones with an artificial intelligence algorithm, the artificial intelligence algorithm recognizes a location where the internal body sound is generated.

Images