KR101921582B1 - Medical diagnosis system, server, and method thereof - Google Patents

Abstract

A medical diagnostic system is disclosed. A medical diagnosis system according to an embodiment of the present invention includes at least one endoscope system and a server. The endoscope system includes a probe unit and a sensor unit. The probe unit includes an image sensor for generating first image data with respect to a long term mucous membrane of a subject and a sub-probe having a light transmitting unit for transmitting laser light to the organ's long term mucosa and receiving scattered light reflected from the long mucous membrane. The sensor unit generates at least one of Raman spectroscopic data and laser scanning optical image data using the scattered light collected by the sub-probe. The server comprises: a collection unit for collecting, from the endoscope system, endoscopic measurement data including at least one of the second image data and Raman spectroscopic data, and lesion data including lesion judgment information determined by a medical professional regarding the subject; A learning unit for learning lesion data and generating lesion judgment reference information for each lesion; And a determination unit for determining a long-term state of the subject using the lesion judgment reference information learned by the learning unit. According to the embodiment of the present invention, information of the lesion can be obtained more objectively, and the accuracy of the lesion diagnosis can be enhanced.

Description

[0001] MEDICAL DIAGNOSIS SYSTEM, SERVER, AND METHOD THEREOF [0002]

The present invention relates to a medical diagnostic system, and more particularly, to a medical diagnostic system based on endoscopic measurement data learning, a server thereof, and a method for generating reference information for lesion judgment.

Optical endoscopes have been used by medical professionals such as stomach, bronchial, esophagus, duodenum, and rectum to observe organs inside the human body that can not directly see lesions and to use them for cancer diagnosis or surgery. The diagnosis of lesions using optical endoscopy is based on the visual observation of the difference between the abnormal tissue of the organ's mucous membrane and the surrounding normal tissue, and further diagnosis of the long-term condition based on histopathological findings such as histological examination. In the case of the diagnostic method based on the glance observation of the optical endoscopic image data, the accuracy of the diagnosis of the lesion depends on the experience and skill of the diagnostic practitioner.

The present invention has been devised on the basis of the above-described technical background, and it is an object of the present invention to provide a medical diagnostic system, a server, and a reference information for determining a lesion, which can acquire information of a lesion objectively more objectively, .

A medical diagnostic system according to an aspect of the present invention includes at least one endoscopic system; And a server.

Wherein the at least one endoscope system comprises an image sensor for generating first image data with respect to a long term mucous membrane of a subject and a light transmission unit for transmitting laser light to the long term mucous membrane of the subject and receiving scattered light reflected from the long term mucous membrane A probe unit including a sub-probe detachably arranged; And a sensor unit for generating at least one of Raman spectroscopic data and second image data which is laser scanning optical image data using the scattered light collected by the sub-probe.

Wherein the server comprises: endoscope measurement data including at least one of the second image data and the Raman spectroscopic data from each of the one or more endoscope systems, and lesion data including lesion judgment information determined by a medical professional regarding the subject A collecting unit for collecting the collected objects; A learning unit for learning the lesion data to generate lesion judgment reference information for each lesion; And a determination unit for determining a long-term state of the subject using the lesion judgment reference information learned by the learning unit.

Wherein the sensor unit comprises: a high magnification image sensor that generates the second image data having a magnification higher than that of the first image data using the scattered light collected by the sub-probe; And a spectroscope for generating the Raman spectroscopic data using the scattered light collected by the sub-probe, wherein the lesion data includes at least one of the first image data, the second image data, the Raman spectroscopic data, Information, and the lesion judgment information.

The learning unit may determine the lesion judgment reference information by learning the relationship between the first image data, the second image data, the Raman spectroscopic data, the body information of the subject, and the lesion judgment information.

The determination unit may determine the long-term state of the subject using the body information of the subject, the endoscope measurement data generated for the organ mucosa of the subject by the endoscope system, and the lesion judgment reference information.

The learning unit learns the relationship between the image pattern of the lesion tissue of each of the first image data and the second image data, the waveform distribution pattern of the Raman spectroscopic data, the body information of the subject, and the lesion judgment information, The reference information can be generated.

Wherein the server generates feedback data when the determination result of the determination unit regarding the long-term state of the subject coincides with the judgment of the medical professional, and when the long-term condition of the subject does not match the judgment of the medical professional, When the number of times of generation of the data exceeds the set reference value, the lesion determination reference information may be updated by reflecting the weight to the lesion data corresponding to the feedback data.

A medical diagnosis server according to another aspect of the present invention includes at least one of Raman spectroscopic data and laser scanning optical image data relating to a long term mucous membrane of a subject and lesion data including lesion judgment information determined by a medical professional regarding the subject A storage unit for storing the data; A learning unit for learning the lesion data to generate lesion judgment reference information for each lesion; And a determination unit for determining a long-term state of the subject using the lesion judgment reference information learned by the learning unit.

Wherein the learning unit learns the relationship between the low magnification image data on the organ mucosa of the subject, the laser scanning optical image data, the Raman spectroscopic data, the body information of the subject, and the lesion judgment information to determine the lesion judgment criterion information have.

The determination unit may determine the long-term state of the subject using the body information of the subject, endoscopic measurement data generated with respect to the organ's mucous membrane, and the lesion judgment reference information.

The learning unit learns the relationship between the image pattern of the lesion tissue of each of the low magnification image data and the laser scanning optical image data, the waveform distribution pattern of the Raman spectroscopic data, the body information of the subject and the lesion judgment information, The judgment criterion information can be generated.

Wherein the medical diagnosis server judges whether or not the judgment result of the judging unit regarding the long term state of the subject coincides with the judgment of the medical professional and generates feedback data when the long term condition of the subject does not agree with the judgment of the medical professional, When the number of times of generation of the feedback data exceeds the set reference value, the lesion determination reference information may be updated by reflecting the weight to the lesion data corresponding to the feedback data.

According to still another aspect of the present invention, there is provided a method for acquiring lesion data including at least one of laser-scanned optical image and Raman spectroscopy data acquired for a long term mucous membrane of a subject and lesion judgment information determined by a medical professional regarding the subject To do; And generating lesion judgment criterion information for judging the long-term condition of the subject by learning lesion data.

The generation of the lesion criterion information comprises learning the relationship between the low magnification image data, the laser scanning optical image data, the Raman spectroscopy data, the body information of the subject, and the lesion judgment information for the organ mucosa of the subject And determining the lesion criterion information.

The generation of the lesion criterion information may comprise generating an image pattern of lesion tissue of each of the low magnification image data and the laser scanning optical image data, a waveform distribution pattern of the Raman spectroscopic data, a body information of the subject, And learning the at least two relationships to generate the lesion criterion information.

The method comprising: determining whether a determination result of the determination unit regarding the long-term status of the subject coincides with a judgment of a medical professional; Generating feedback data when the long-term condition of the subject does not match the judgment of the medical professional; And updating the lesion judgment reference information by reflecting a weight on lesion data corresponding to the feedback data when the number of times of generation of the feedback data exceeds a set reference value; .

According to the embodiment of the present invention, information of the lesion can be obtained more objectively, and the accuracy of the lesion diagnosis can be enhanced.

1 is a configuration diagram of a medical diagnosis system according to an embodiment of the present invention.
2 is a configuration diagram of an endoscope system constituting a medical diagnosis system according to an embodiment of the present invention.
3 is a perspective view of a probe unit constituting an endoscope system according to an embodiment of the present invention.
4 is an illustration of lesion data according to an embodiment of the present invention.
5 is a flowchart of a method for generating reference information for lesion judgment according to an embodiment of the present invention.
FIG. 6 is an exemplary diagram showing learning of lesion data according to an embodiment of the present invention. FIG.
7 is an exemplary view of second image data obtained according to an embodiment of the present invention.
8 is an exemplary diagram of Raman spectroscopic data obtained according to an embodiment of the present invention.
9 is another example of lesion data learned in a server according to an embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily carry out the present invention.

The present invention may be embodied in many different forms and is not limited to the embodiments described herein. In order to clearly illustrate the present invention, parts not related to the description are omitted, and the same or similar components are denoted by the same reference numerals throughout the specification. In addition, since the sizes and thicknesses of the respective components shown in the drawings are arbitrarily shown for convenience of explanation, the present invention is not necessarily limited to those shown in the drawings.

In this specification, when a part is referred to as "including " an element, it is to be understood that it may include other elements as well, without departing from the other elements unless specifically stated otherwise.

As used herein, 'to' is a unit for processing at least one function or operation, and may be, for example, software, FPGA or hardware component. The functions provided in the " part " may be performed separately by a plurality of components, or may be integrated with other additional components. It is to be understood that the term " portion " herein is not limited to software or hardware, but may be embodied in an addressable storage medium and configured to reproduce one or more processors.

The medical diagnosis system according to the embodiment of the present invention collects endoscopic measurement data such as CCD image data, laser scanning optical image data, and Raman spectroscopic data acquired for the organ mucosa of the subject by the endoscope system, The lesion judgment criterion information can be accurately diagnosed using the lesion judgment criterion information by learning the relation between the lesion judgment criterion information and the lesion judgment criterion information.

1 is a configuration diagram of a medical diagnosis system according to an embodiment of the present invention. Referring to FIG. 1, the medical diagnosis system according to the present embodiment includes a plurality of endoscopic systems 100, 200, 300, and 400, a server 500, and a storage unit 600.

The plurality of endoscope systems 100, 200, 300, and 400 and the server 500 may be connected to each other through a wired and / or wireless communication network, and may transmit data to each other. In an embodiment, a plurality of endoscopic systems may include endoscopic systems 100, 200, 300, and a remote endoscopic system 400.

Illustratively, a plurality of endoscopic systems 100, 200, 300, and 400 may be provided at the hospital side for performing diagnosis and surgery for organs, respectively, using an endoscope apparatus. The remote endoscope system 400 includes an operation unit for performing an endoscope examination remotely and a display for displaying data such as images obtained by the endoscope apparatus.

2 is a configuration diagram of an endoscope system constituting a medical diagnosis system according to an embodiment of the present invention. Referring to FIG. 1, an endoscope system 100 includes a terminal 110, a laser light source 112, a probe unit 120, and a sensor unit 180.

3 is a perspective view of a probe unit 120 constituting an endoscope system according to an embodiment of the present invention. 1 to 3, the probe unit 120 is inserted into the body of the examinee or the operation subject, and is provided for observing organs inside the subject.

In the present embodiment, the probe unit 120 is not necessarily limited to being inserted into the organs of the examinee's body to observe the inside of the organ, and the configuration for examining the probe unit of the tissue examination specimen collected from the organ of the examinee is also the embodiment May be included in the example.

The probe unit 120 may include an image sensor 130 for generating a low magnification image and a sub-probe 140 for performing a laser beam scanning and a scattering light collecting function to generate a high magnification image.

The image sensor 130 may be illustratively provided as a charge coupled device (CCD) or the like. The image sensor 130 may be provided at a distal end of the probe unit 120. In order to adjust the brightness in the organ, light sources 131 and 132 may be provided at the distal end of the probe unit 120.

Meanwhile, in the present embodiment, the image sensor 130 is a CCD device. However, the image sensor 130 may be one of various sensors for sensing an image, for example, CMOS, and the like.

The sub-probe 140 includes a light transmitting portion 150 and a stabilizing portion 160. The light transmitting portion 150 is illustratively provided to be inserted into an instrument channel of the probe portion 120 and is detachably disposed with respect to the probe portion 120. [

The optical transmission unit 150 may be provided as a confocal probe or a Raman probe composed of an optical fiber bundle. Illustratively, some of the optical fibers of the light transmitting portion 150 may be provided as light receiving portions for receiving the scattered light of the laser light, and others may be provided as light emitting portions for scanning the laser light.

The tip portion of the light transmitting portion 150 constituting the sub-probe 140 may protrude from the distal end portion of the probe portion 120 and may be elongated in the forward direction Z. [ In one embodiment, the light transmitting portion 140 may be provided to have flexibility that can be bent in the up, down, left, and right directions.

The stabilizer 160 compensates for the shake of the sub-probe 140 and performs a stabilization function of controlling the attitude of the sub-probe 140 to adjust the attitude of the sub-probe 140 to a high magnification image (laser scanning optical image) generated by the sensor unit 180 It is possible to prevent the occurrence of shaking.

In order to scan the organ mucosa inside the human body, the laser light source 112 generates a predetermined single wavelength or multi-wavelength laser light. A bandpass filter 114 passes a laser beam of a specific wavelength to be scanned to a long-term mucous membrane among the laser beams generated by the laser beam source 112. The laser light source 112 and the band pass filter 114 may be illustratively provided to generate laser light having at least one of 415 nm to 540 nm.

The laser light having passed through the band pass filter 114 is provided to the light transmitting portion 150 through a coupling 116 and is transmitted to the light transmitting portion 150 through at least a part of the optical fiber bundles 151 (Mucosa) of the < / RTI > At this time, coupling 116 may be a beam splitter, which is a reflecting mirror or other optical device, which illustratively reflects part of the light beam and transmits the other part.

In one embodiment, a plurality of laser beams having different wavelengths can be irradiated to the mucosa of the organ through the light transmitting portion 150. [ Illustratively, a blue laser beam with a wavelength of 415 nm and a green laser beam with a wavelength of 540 nm can be irradiated to the long-term mucosa.

The blue laser light has a wavelength shorter than that of the green laser light and can be reflected by superficial capillary networks and received by the light transmitting part 150. Since the green laser light has a longer wavelength than the blue laser light, it can be reflected by the subepithelial tissue passing through the skin and received by the light transmitting part 150.

The scattered light collected through the optical transmission unit 150 is transmitted to the sensor unit 180 via the coupling 116 via the long pass filter 170. The sensor unit 180 may include a sensor unit 182 and a spectrograph 184. [

The spectroscope 184 spectrums the laser light including the Raman scattered light collected by the probe unit 120 to generate spectral data. Here, the spectrum data may include measurable Raman spectroscopic data, and the Raman spectroscopic data includes Raman variation waveforms, wherein the abscissa represents the Raman variation and the ordinate represents the intensity (brightness).

The sensor unit 182 generates a high magnification image (laser scanning optical image) using the information of the scattered light collected by the probe unit 120 or extracts the high magnification image or the Raman spectroscopy The spectrum can be sensed. The sensed high magnification image or the Raman spectroscopic spectrum may be generated as high magnification image information or Raman spectroscopic spectral information in the sensor unit 182 and transmitted to the terminal 110.

That is, the sensor unit 182 according to the present embodiment may be a high sensitivity sensor capable of sensing the high magnification image and the Raman spectrum.

The terminal 110 is electrically coupled to the laser light source 112, the probe unit 120 and the sensor unit 180 to control the operation of the laser light source 112, the probe unit 120, and the sensor unit 180 Simultaneously, data such as CCD image data (first image data), laser scanning optical image data (second image data), and Raman spectroscopic data generated by the probe unit 120 and the sensor unit 180 can be displayed have.

At this time, the first image data, which is CCD image data, is formed as low-magnification image data than the second image data, which is laser scanning optical image data, so that the first image data can be referred to as low-magnification image data.

1, the lesion data generated by each of the endoscopic systems 100, 200, 300, and 400 is provided to the server 500 so as to accurately examine lesions on an objective basis, / RTI >

The server 500 learns lesion data provided from the endoscope systems 100, 200, 300, and 400 to generate lesion judgment reference information for each lesion. The server 500 may include a determination unit 510, a learning unit 520, a remote control unit 530, and a collection unit 540.

4 is an illustration of lesion data according to an embodiment of the present invention. Referring to FIGS. 1 and 4, in one embodiment, the lesion data may include endoscopic measurement data obtained for an organ of a subject, lesion judgment information, and body information of the subject.

The endoscopic measurement data may include first image data (CCD image data), second image data (laser scanning optical image data), and Raman spectroscopic data acquired for the organ's mucous membranes.

The lesion judgment information is a diagnosis result of the lesion determined by the medical professional regarding the subject. The subject's body information may include various information such as sex, age, height, weight, race, nationality, smoking amount, drinking amount, family history, etc. of the subject.

The remote control unit 530 and the collecting unit 540 collect the lesion data from the remote endoscope system 400 and the endoscope systems 100, 200, 300, and 400, respectively. The remote control unit 530 and the collecting unit 540 can collect the lesion data from the endoscope systems 100, 200, 300, and 400 through the wired / wireless communication network.

The learning unit 520 may learn the lesion data collected from the endoscope systems 100, 200, 300, and 400 in a deep data-based deep learning method to generate lesion judgment reference information for each lesion. Illustratively, the learning unit 520 may generate lesion criterion information by learning lesion data using a Recurrent Neural Network algorithm.

The learning unit 520 learns the relationship between the first image data (CCD image data), the second image data (laser scanning optical image data), Raman spectroscopic data, the subject's body information, and the lesion judgment information, Information can be determined.

In an embodiment, the learning unit 520 may include at least two of the image pattern of the lesion tissue of each of the first image data and the second image data, the waveform distribution pattern of Raman spectroscopic data, the subject's body information, So as to generate the lesion judgment reference information.

The determination unit 510 determines the long-term state of the subject using the lesion judgment reference information learned by the learning unit 520. [ Diagnostic information can be generated in the form of probabilistic information about a specific disease or probabilistic information about a number of diseases.

In the embodiment, the judging unit 510 judges, based on the body information of the subject, the endoscope measurement data generated for the organ mucosa of the subject by the endoscope system, and the lesion judgment reference information generated by the learning unit 520, Can be determined.

5 is a flowchart of a method for generating reference information for lesion judgment according to an embodiment of the present invention. Referring to FIGS. 1 and 5, lesion data is collected from a plurality of endoscope systems 100, 200, 300, and 400 by the server 500 (S11). The lesion data may include first image data (CCD image data), second image data (laser scanning optical image data), Raman spectroscopic data, body information of a subject, and lesion judgment information of a medical professional.

When the lesion data is collected, the learning unit 520 learns lesion data collected from the endoscope systems 100, 200, 300, and 400, and generates lesion judgment reference information for lesions (S12).

FIG. 6 is an exemplary diagram showing learning of lesion data according to an embodiment of the present invention. FIG. Referring to FIG. 6, the learning unit 520 calculates the relationship between the first image data (CCD image data), the second image data (laser scanning optical image data), Raman spectroscopic data, the subject's body information, The lesion judgment criterion information can be determined.

FIG. 7 is an exemplary view of second image data obtained according to an embodiment of the present invention, and FIG. 8 is an exemplary view of Raman spectroscopic data obtained according to an embodiment of the present invention. Referring to FIGS. 1 and 5 to 8, the learning unit 520 is configured to classify the lesion tissue 12 of each of the first image data (CCD image data), the second image data 10 (laser scanning optical image data) (RS1, RS2, RS3), peak values (P1, P2, P3) and peak widths (RW1, RW2, RW3), various physical information of the subject, The lesion judgment reference information can be generated by learning the relation of the lesion judgment information by the lesion judgment reference information.

The determination unit 510 determines the long-term state of the subject using the lesion judgment reference information learned by the learning unit 520 (S13). In the embodiment, the determination unit 510 receives endoscopic measurement data (CCD image, laser scanning optical image, Raman spectroscopic data) generated for the subject's body information, the endoscope mucosa of the subject by the endoscope system, To the lesion judgment criterion information generated by the lesion judgment criterion information.

In addition, the server 500 determines whether the determination result of the determination unit 510 regarding the long-term status of the subject matches the determination of the medical expert (S14). If the long-term condition of the subject does not match the determination of the medical professional, And generates feedback data (S15).

If the number of generation of the feedback data exceeds the set reference value N, the lesion determination reference information may be updated by reflecting the weight to the lesion data corresponding to the feedback data (S16, S17). The reference value may be set to an integer equal to or greater than zero. In the case where the judgment of the doctor according to the gross observation of the lesion data and the judgment result of the server 500 obtained by applying the lesion data to the lesion judgment criterion information are different, a high weight is given to the lesion data to improve the learning performance.

9 is another example of lesion data learned in a server according to an embodiment of the present invention. Referring to FIG. 9, the server 500 includes first image data as low magnification CCD image data, second image data as a high magnification laser scanning optical image, Raman spectroscopy data, first and second lesion judgment information and final lesion judgment information, The lesion judgment criterion information for each lesion can be generated by learning the relationship of the subject's body information.

The primary lesion judgment information is a result of the lesion judgment based on the gross observation of the endoscopic measurement data (CCD image data, laser scanning type optical image, Raman spectroscopic data) of the subject by the doctor, Means the result of ultimately judging the lesion based on the results obtained from the examination.

In this embodiment, the weights of the lesion data may be determined differently according to the first lesion judgment data of the doctor, the final lesion judgment data, and the lesion judgment result of the server 500. Illustratively, the weights of lesion data can be set according to Table 1 below.

Judgment result of doctor's lesion Lesion judgment result
(server) weight
(weight) Results of primary lesion judgment Final lesion judgment result positivity positivity positivity W1 positivity positivity voice W2 positivity voice positivity W3 positivity voice voice W4 voice positivity positivity W5 voice positivity voice W6 voice voice positivity W7 voice voice voice W8

For example, when the first lesion judgment result on the lesion of stomach cancer is 'negative', the server lesion judgment result is 'positive', and the final lesion judgment result is 'positive', a high weight value W5), more reliable lesion judgment criterion information can be secured and used for lesion diagnosis.

The lesion judgment reference information generated by learning using the lesion data can be provided with different reliability (examination accuracy) of CCD image data, laser scanning optical image data, and Raman spectroscopy data for various lesions. For example, it can be determined that the reliability of the CCD image data is high for the first lesion, the reliability of the laser scanning optical image data is high for the second lesion, and the reliability of the Raman spectroscopic data is high for the third lesion. Therefore, by diagnosing the long-term condition by raising the rate of reflecting highly reliable data on a specific lesion, it is possible to improve the accuracy of the lesion detection.

The medical diagnosis system and the lesion judgment criterion generation method according to the present embodiment are different from the conventional endoscope system and diagnosis method in which only the visual observation of the organ of the examinee is performed and the examinee makes the judgment about the examinee based on the visual observation, Based on the optical observation of the subject's organs, the lesion data is collected and objective lesion criterion information is generated by learning the collected lesion data to diagnose the long-term condition such as precancerous lesion and cancer of the examinee.

According to the present embodiment, it is possible to provide more reliable diagnostic information based on objective data (lesion judgment criterion information), instead of performing diagnoses on gastritis, stomach cancer and the like based on the experience of a diagnostician.

The medical diagnosis system according to an embodiment of the present invention can be utilized for diagnosing conditions of mucous membranes of various organs such as stomach, bronchus, esophagus, duodenum, rectum and the like by way of example.

According to this embodiment, objective lesion judgment criteria information can be established for each lesion, and diagnosis of the histopathological information can be made by subdividing the histopathological information into Gastric adenocarcinoma, Gastritis, Helicobacter pylori infection, chronic atrophic gastritis and alcoholic gastric ulcer .

In addition, during the screening endoscopy examination, it is possible to accurately determine the screening by dividing the entire normal part of the mucosa of the digestive tract. In addition, a more accurate differential diagnosis is possible in the case of colorectal disease. For example, it can be accurately distinguished by various lesions such as Crohn's disease, ulcerative colitis, and tuberculous enteritis.

According to the present embodiment, accurate diagnosis results can be obtained by using a high magnification image (laser scanning optical image) without injecting contrast agents such as fluorescein sodium into the organs of a subject to be examined And the advantage of being able to eliminate side effects by the contrast agent is also provided.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments, and that various changes and modifications may be made without departing from the scope of the invention, It is natural to belong to the scope.

100 to 400: endoscope system 110: terminal
112: laser light source 114: bandpass filter
116: Coupling 120: Probe portion
130: image sensor 131, 132: light source
140: Sub-probe 150:
151: Optical fiber 160:
180: Sensor unit 182: High magnification image sensor
184: spectroscope 500: server
510: Judgment section 520:
530: remote control unit 540:
600:

Claims (13)

A sub-probe having an image sensor for generating first image data for a long term mucous membrane of the subject, and a light transmitting portion for transmitting the laser light to the long term mucous membrane and receiving scattered light reflected from the long mucous membrane, ; And a sensor unit for generating at least one of Raman spectroscopic data and second image data that is laser-scanned optical image data using the scattered light collected by the sub-probe. And
Acquiring, from each of the at least one endoscope system, endoscopic measurement data including at least one of the second image data and the Raman spectroscopic data, and a collection for collecting lesion data including lesion judgment information determined by a medical professional regarding the subject part; A learning unit for learning the lesion data to generate lesion judgment reference information for each lesion; And a determination unit for determining a long-term state of the subject using the lesion judgment criterion information learned by the learning unit.
Lt; / RTI >
A part of the sub-probe is protrudable from the distal end of the probe part,
The sensor unit includes:
A high magnification image sensor for generating the second image data having a magnification higher than that of the first image data using the scattered light collected by the sub-probe; And
And a spectroscope for generating the Raman spectroscopic data using the scattered light collected by the sub-probe,
Wherein the lesion data includes the first image data, the second image data, the Raman spectroscopic data, the body information of the subject, and the lesion judgment information,
Wherein the learning unit obtains a relationship between a lesion tissue image pattern of the first image data, a lesion tissue image pattern of the second image data having a higher magnification than the first image data, a waveform distribution pattern of the Raman spectroscopy data, And generates the lesion judgment reference information. delete The method according to claim 1,
Wherein the learning unit determines the lesion judgment criterion information by further considering the body information of the subject,
Wherein the determination unit determines the long-term state of the subject using the body information of the subject, the endoscopic measurement data generated for the organ mucosa of the subject by the endoscope system, and the lesion judgment reference information. delete The method according to claim 1,
The server comprises:
Wherein the feedback data generation unit generates feedback data when the long-term state of the subject does not agree with the judgment of the medical professional, The lesion determination criterion information is updated by reflecting the weight to the lesion data corresponding to the feedback data,
Wherein the medical diagnosis system assigns a higher weight to a case in which the judgment of the medical professional and the judgment result of the judgment unit are not coincident with each other. A storage unit for storing lesion data including at least one of low magnification image data, Raman spectroscopic data and laser scanning optical image data relating to a long term mucous membrane of the subject, and lesion judgment information determined by a medical professional regarding the subject;
A learning unit for learning the lesion data to generate lesion judgment reference information for each lesion; And
And a determination unit for determining a long-term state of the subject using the lesion judgment reference information learned by the learning unit,
Wherein the learning unit obtains a relationship between a lesion tissue image pattern of the low magnification image data, a lesion tissue image pattern of the laser scanning optical image data having a higher magnification than the low magnification image data, a waveform distribution pattern of the Raman spectroscopic data, And generates the lesion judgment criterion information by learning. The method according to claim 6,
Wherein the learning unit determines the lesion judgment criterion information by further considering the body information of the subject,
Wherein the determination unit determines the long term state of the subject using the body information of the subject, endoscopic measurement data generated with respect to the organ mucosa of the subject, and the lesion judgment reference information. delete The method according to claim 6,
Wherein the feedback data generation unit generates feedback data when the long-term state of the subject does not agree with the judgment of the medical professional, If the judgment result of the healthcare professional and the judgment result of the judgment unit are not coincident with each other, the medical care professional can update the lesion judgment criterion information by reflecting the weight to the lesion data corresponding to the feedback data, And gives a higher weight than when the determination result of the determination unit matches the determination result of the determination unit. The diagnosis method according to claim 6, wherein the server comprises at least one of a laser scanning optical image and Raman spectroscopy data for organ tissues collected from a human body of a subject, and a lesion judgment Collecting lesion data including information; And
Wherein the server learns the lesion data and generates lesion judgment criterion information for judging the long-term condition of the subject for each lesion. 11. The method of claim 10,
Generating the lesion criterion information may comprise:
At least two relationships among the low magnification image data, the laser scanning optical image data, the Raman spectroscopic data, the body information of the subject, and the lesion judgment information for the organ mucosa of the subject are learned and the lesion judgment reference information is determined Lt; / RTI > 12. The method of claim 11,
Generating the lesion criterion information may comprise:
Learning the relationship between the image pattern of the lesion tissue of each of the low magnification image data and the laser scanning optical image data, the waveform distribution pattern of the Raman spectroscopic data, the body information of the subject, and the lesion judgment information, ≪ / RTI > 11. The method of claim 10,
Judging whether or not the judgment result of the server regarding the long-term state of the subject coincides with the judgment of the medical professional;
Generating feedback data when the long-term condition of the subject does not match the judgment of the medical professional; And
And updating the lesion judgment reference information by reflecting a weight on lesion data corresponding to the feedback data when the number of times of generation of the feedback data exceeds a set reference value.

Images