Colonoscopy Examination Site Display System

Abstract

This invention provides a system and method for indicating the site of colonoscopy. [Solution] The method for indicating a colonoscopy examination site according to the present invention includes the steps of: a model loading / condition setting unit loading a colonoscopy image analysis model and setting the analysis conditions of the analysis model; a control unit initializing the analysis screen and displaying an image of a normal colon; an image preprocessing unit preprocessing the colonoscopy image received via the image receiving unit so that subsequent image analysis can be performed smoothly; an image analysis unit analyzing the preprocessed colonoscopy image using an AI-based image analysis model; an image analysis unit detecting and indicating at least one of the examination site, diverticulum site, and lesion site in the colonoscopy image based on the analysis results; and a control unit providing the analysis results analyzed by the image analysis unit.

Description

Technical Field

[0001] The present invention relates to a colonoscopy site indication system and method. More specifically, when applying image recognition technology to indicate the main examination site during the colonoscopy process, the present invention relates to a colonoscopy site indication system and method for notifying an examiner of situations such as a diverticulum where the colon wall protrudes, or indicating it on the examination screen so that it is different from the normal colon.

[0002] The national research and development project supported by the present invention is as follows.

[0003] Problem specific number not assigned Problem number RS-2024-00510314 Department name Small and Medium-sized Venture Business Department Problem management agency name Small and Medium-sized Enterprise Technology Information Promotion Agency Research project name Development of Startup Growth Technology (R&D) Research problem name Development of technology for improving the quality of colonoscopy based on artificial intelligence - Cecum detection, endoscopy speed, examination time measurement technology Problem execution agency name WaySen Co., Ltd. Research period October 1, 2024 to September 30, 2025

Background Art

[0004] Today, due to Westernized eating habits and the development of diagnostic techniques, the incidence of colorectal cancer is increasing rapidly. 80-90% of colorectal cancers start from polyps (adenomas), which are small tumors that occur in the colon. Early detection and removal of the polyps by colonoscopy can significantly reduce the mortality rate due to colorectal cancer.

[0005] The purpose of colorectal cancer screening is to detect colorectal cancer early and reduce colorectal cancer-related mortality. Previous research has shown that the effect of colorectal cancer screening on reducing colorectal cancer mortality varies depending on the screening method. Cancer screening methods must have high sensitivity and specificity, be free from risks or complications, and be inexpensive. Methods currently proposed for colorectal cancer screening include fecal occult blood testing, sigmoidoscopy, colonoscopy, and double-contrast colonoscopy. Fecal occult blood testing has been reported to reduce colorectal cancer mortality by 15-33% in large-scale randomized clinical trials conducted in Western Europe. While fecal occult blood testing has advantages such as being free from complications, inexpensive, and relatively easy to perform, it has been pointed out that it has problems such as low sensitivity and predictive value for benign cases in a single test, and a high false-positive rate leading to additional testing.

[0006] Therefore, although colonoscopy screening has recently been recommended, it is applied restrictively to national health checkups targeting a large number of ordinary people (for example, as an additional test for those with abnormal findings in differential occult blood tests) due to its relatively high cost, rare but serious complications (such as colon perforation), discomfort and inconvenience to examinees due to pre-examination procedures, and a shortage of skilled endoscopists.

[0007] On the other hand, Korean Published Patent No. 10-2022-0140924 discloses a "deep learning-based colonoscopy image analysis method and image analysis system using the same," which is characterized by including an endoscope computer that receives images taken by a colonoscopy device, a server that acquires the images transmitted to the endoscope computer via an application downloaded to the endoscope computer using a hooking method, and is equipped with a diagnostic algorithm that performs correction of the acquired images and examination based on the corrected images based on deep learning, and transmits the corrected images and diagnostic results derived via the diagnostic algorithm to the endoscope computer, and a display device that receives and outputs the corrected images and diagnostic results from the endoscope computer.

[0008] In the case of the above-mentioned patent documents, while it is possible to operate without the manufacturer's API (Application Programming Interface) by reading and processing images using a window hooking method, and has the advantage of improving diagnostic reliability by correcting images due to light reflection and accurately measuring the size of polyps, there is no separate means to inform the examiner of conditions such as diverticula during the colonoscopy process or to instruct (display) on the examination screen that it is different from a normal colon, and thus it carries the risk that the examiner may cause complications such as perforation. [Overview of the project] [Problems that the invention aims to solve]

[0009] The present invention was created by comprehensively considering the above-mentioned matters, and its purpose is to provide a colonoscopy site indication system and method that allows examiners to perform the examination while paying attention to complications such as perforation, helps in setting the probe movement path, and enables examiners to perform a more thorough colonoscopy by applying image recognition technology to the colonoscopy process to indicate the main examination site, informing the examiner of situations such as diverticula where the colon wall protrudes, or indicating (displaying) on ​​the examination screen that it is different from a normal colon.

[0010] Another objective of the present invention is to provide a colonoscopy site indication system and method that allows the examiner to confirm the areas examined to date by indicating the main colon examination sites, to use as an indicator of colonoscopy quality by indicating whether or not a cecum examination has been performed, and to record whether or not a diverticulum is present, thereby enabling the examiner to proceed with the next colonoscopy with prior knowledge of the risk of perforation. [Means for solving the problem]

[0011] To achieve the above objective, a colonoscopy site indication system according to one embodiment of the present invention is A model loading / condition setting unit loads a colonoscopy image analysis model and sets the analysis conditions for the analysis model, An image receiving unit that receives colonoscopy image frames, An image preprocessing unit that preprocesses colonoscopy images received via the image receiving unit to facilitate subsequent image analysis, An image analysis unit analyzes the colonoscopy image preprocessed by the image preprocessing unit using an AI (artificial intelligence) based image analysis model, and detects and indicates at least one of the examination site, diverticulum site, and lesion site in the colonoscopy image based on the analysis results. The system is characterized by including a control unit that controls the status check and operation of the model loading / condition setting unit, image receiving unit, image preprocessing unit, and image analysis unit, and when the model loading / condition setting unit completes loading of the colonoscopy image analysis model and setting of analysis conditions for the analysis model, initializes the analysis screen to display an image of a normal colon and provides the analysis results analyzed by the image analysis unit.

[0012] Here, the preprocessing of the colonoscopy image by the image preprocessing unit may include segmenting the analysis area and adjusting the input size.

[0013] Furthermore, the image analysis model of the image analysis unit can consist of a single image analysis model that detects the examination site and the diverticulum site.

[0014] Furthermore, the image analysis model of the image analysis unit can consist of an inspection site detection model for detecting the inspection site and a diverticulum detection model for detecting the diverticulum site.

[0015] Furthermore, the image analysis model of the image analysis unit can consist of an examination site detection model for detecting the examination site, a diverticulum detection model for detecting the diverticulum site, and a lesion detection model for detecting the lesion site.

[0016] In this case, the lesion detection model may include a lesion attribute classification function that distinguishes whether a lesion is benign or malignant.

[0017] Furthermore, the image analysis model of the image analysis unit can consist of a cecum / diverticulum detection model that detects the cecum and diverticula, and a lesion detection model that detects lesion sites.

[0018] In this case, the lesion detection model may include a lesion attribute classification function that distinguishes whether a lesion is benign or malignant.

[0019] In addition, when the image analysis unit detects and indicates at least one of the examination site, diverticulum site, and lesion site in the colon endoscope image, the examination site may include the appendix projection, cecum, ascending colon, transverse colon, descending colon, sigmoid colon, and rectum.

[0020] In addition, when the image analysis unit detects and indicates at least one of the examination site, diverticulum site, and lesion site in the colon endoscope image, if it is the appendix projection, the control unit can send an instruction condition change command to the image analysis unit to indicate the cecum and make the diverticulum indicated differently in the picture of the normal colon.

[0021] In addition, when the image analysis unit detects and indicates at least one of the examination site, diverticulum site, and lesion site in the colon endoscope image, the control unit can send an instruction condition change command to the image analysis unit to prevent the diverticulum from being indicated in the picture of the normal colon, or to notify the colon diverticulum and to indicate the examination time and the collection time.

[0022] In addition, in order to achieve the above object, a method for indicating a colon endoscope examination site according to an embodiment of the present invention includes: a) A step in which a model loading / condition setting unit loads a colon endoscope image analysis model and sets analysis conditions of the analysis model; b) A step in which a control unit initializes an analysis screen and displays a picture of the normal colon; c) A step in which an image preprocessing unit preprocesses the colon endoscope image received via an image receiving unit so that subsequent image analysis can be performed smoothly; d) A step in which an image analysis unit analyzes the preprocessed colon endoscope image using an AI-based image analysis model; e) A step in which the image analysis unit detects and indicates at least one of the examination site, diverticulum site, and lesion site in the colon endoscope image based on the analyzed result; f) A step in which a control unit provides the analysis result analyzed by the image analysis unit, and it is characterized in this point.

[0023] Here, the preprocessing of the colonoscopy image by the image preprocessing unit in step c) can include cutting the analysis region and adjusting the input size.

[0024] Also, in step d), the image analysis model can be composed of a single image analysis model that detects the examination site and the diverticulum site.

[0025] Also, in step d), the image analysis model can be composed of an examination site detection model that detects the examination site and a diverticulum detection model that detects the diverticulum site.

[0026] Also, in step d), the image analysis model can be composed of an examination site detection model that detects the examination site, a diverticulum detection model that detects the diverticulum site, and a lesion detection model that detects the lesion site.

[0027] In this case, the lesion detection model can be equipped with a lesion attribute classification function for classifying whether the lesion is benign or malignant.

[0028] Also, in step d), the image analysis model can be composed of a cecum / diverticulum detection model that detects the cecum and the diverticulum and a lesion detection model that detects the lesion site.

[0029] In this case, the lesion detection model can be equipped with a lesion attribute classification function for classifying whether the lesion is benign or malignant.

[0030] Also, in step e), when the image analysis unit detects and indicates at least one of the examination site, the diverticulum site, and the lesion site in the colonoscopy image, the examination site can include the appendix projection, the cecum, the ascending colon, the transverse colon, the descending colon, the sigmoid colon, and the rectum.

[0031] Furthermore, in step e), when the image analysis unit detects and indicates at least one of the examination site, diverticulum site, and lesion site in the colonoscopy image, the control unit can send an instruction condition change command to the image analysis unit to indicate the cecum if the appendix is ​​detected, and to indicate a different diverticulum in a normal colon image.

[0032] Furthermore, in step e), when the image analysis unit detects and indicates at least one of the examination site, diverticulum site, and lesion site in the colonoscopy image, the control unit may send an instruction condition command to the image analysis unit to either not indicate a diverticulum in the image of a normal colon, or to indicate a colonic diverticulum, and to indicate the examination time and retrieval time.

[0033] Furthermore, in order to achieve the above objectives, the colonoscopy site indication system according to another embodiment of the present invention is A model loading / condition setting unit loads a colonoscopy analysis model and a voice keyword recognition model, and sets the analysis conditions for the analysis model. An image receiving unit that receives colonoscopy image frames, An image preprocessing unit that preprocesses colonoscopy images received via the image receiving unit to facilitate subsequent image analysis, An image analysis unit analyzes the colonoscopy image preprocessed by the image preprocessing unit using an AI (artificial intelligence) based image analysis model, and detects and indicates at least one of the examination site, diverticulum site, or lesion site in the colonoscopy image based on the analysis results. While the image analysis unit performs image analysis, the speech recognition unit reads audio from a buffer that stores audio, analyzes it using an AI-based speech keyword recognition model, recognizes speech keywords based on the analysis results, and transmits them to the image analysis unit. The system controls the status checks and operations of the model loading / condition setting unit, image receiving unit, image preprocessing unit, image analysis unit, and voice recognition unit. Once the model loading / condition setting unit completes loading of the colonoscopy image analysis model and setting of analysis conditions for the analysis model, it initializes the analysis screen, displays an image of a normal colon, and provides the analysis results analyzed by the image analysis unit. The system is characterized by including a control unit that provides the analysis results by linking the analysis target detected by the image analysis model with the voice commands (keywords) related to the analysis target spoken by the examiner.

[0034] Here, the preprocessing of the colonoscopy image by the image preprocessing unit may include segmenting the analysis area and adjusting the input size.

[0035] Furthermore, the image analysis model of the image analysis unit can consist of a single image analysis model that detects the examination site and the diverticulum site.

[0036] Furthermore, the image analysis model of the image analysis unit can consist of an inspection site detection model for detecting the inspection site and a diverticulum detection model for detecting the diverticulum site.

[0037] Furthermore, the image analysis model of the image analysis unit can consist of an examination site detection model for detecting the examination site, a diverticulum detection model for detecting the diverticulum site, and a lesion detection model for detecting the lesion site.

[0038] In this case, the lesion detection model may include a lesion attribute classification function that distinguishes whether a lesion is benign or malignant.

[0039] Furthermore, the image analysis model of the image analysis unit can consist of a cecum / diverticulum detection model that detects the cecum and diverticula, and a lesion detection model that detects lesion sites.

[0040] In this case, the lesion detection model may include a lesion attribute classification function that distinguishes whether a lesion is benign or malignant.

[0041] Furthermore, when the image analysis unit detects and indicates at least one of the examination site, diverticulum site, and lesion site in the colonoscopy image, the examination site may include the appendix, cecum, ascending colon, transverse colon, descending colon, sigmoid colon, and rectum.

[0042] Furthermore, when the image analysis unit detects and indicates at least one of the examination site, diverticulum site, and lesion site in the colonoscopy image, the control unit can send an instruction condition change command to the image analysis unit to indicate the cecum if the appendix is ​​detected, and to indicate a different diverticulum in an image of a normal colon.

[0043] Furthermore, the control unit can send a command to the image analysis unit to change the instruction conditions when the image analysis unit detects and instructs at least one of the examination site, diverticulum site, and lesion site in the colonoscopy image. This command will either prevent the diverticulum from being indicated in the image of a normal colon, or indicate the presence of a colonic diverticulum, and will also instruct the unit to specify the examination time and retrieval time.

[0044] Furthermore, in order to achieve the above objectives, the method for indicating the colonoscopy site according to another embodiment of the present invention is: p) The model loading / condition setting unit loads the colonoscopy analysis model and the voice keyword recognition model, and sets the analysis conditions for the analysis models. q) The control unit initializes the analysis screen and displays an image of a normal colon, r) The image preprocessing unit preprocesses the colonoscopy images received via the image receiving unit so that subsequent image analysis can be performed smoothly. s) The image analysis unit analyzes the colonoscopy images preprocessed by the image preprocessing unit using an AI-based image analysis model, t) The image analysis unit detects and indicates at least one of the examination site, diverticulum site, and lesion site in the colonoscopy image based on the results of the analysis, u) The speech recognition unit reads audio from a buffer that stores audio while the image analysis unit is performing image analysis, analyzes it using an AI-based speech keyword recognition model, recognizes speech keywords based on the analysis results, and transmits them to the image analysis unit. v) The control unit is characterized by including the step of providing an analysis result by linking the object to be analyzed detected by the image analysis model of the image analysis unit with the voice command (keyword) related to the object to be analyzed spoken by the examiner.

[0045] Here, in step r), the preprocessing of the colonoscopy image by the image preprocessing unit may include cutting the analysis region and adjusting the input size.

[0046] Furthermore, in step s), the image analysis model may consist of a single image analysis model that detects the examination site and the diverticulum site.

[0047] Furthermore, in step s), the image analysis model may consist of an inspection site detection model for detecting the inspection site and a diverticulum detection model for detecting the diverticulum site.

[0048] Furthermore, in step s), the image analysis model may consist of an examination site detection model for detecting the examination site, a diverticulum detection model for detecting the diverticulum site, and a lesion detection model for detecting the lesion site.

[0049] In this case, the lesion detection model may include a lesion attribute classification function that distinguishes whether a lesion is benign or malignant.

[0050] Furthermore, in step s), the image analysis model may consist of a cecum / diverticulum detection model for detecting the cecum and diverticula, and a lesion detection model for detecting lesion sites.

[0051] In this case, the lesion detection model may include a lesion attribute classification function that distinguishes whether a lesion is benign or malignant.

[0052] Furthermore, in step t), when the image analysis unit detects and indicates at least one of the examination site, diverticulum site, and lesion site in the colonoscopy image, the examination site may include the appendix, cecum, ascending colon, transverse colon, descending colon, sigmoid colon, and rectum.

[0053] Furthermore, in step t), when the image analysis unit detects and indicates at least one of the examination site, diverticulum site, and lesion site in the colonoscopy image, the control unit can send an instruction condition change command to the image analysis unit to indicate the cecum if the appendix is ​​present, and to indicate a different diverticulum in a normal colon image.

[0054] Furthermore, in step t), when the image analysis unit detects and indicates at least one of the examination site, diverticulum site, and lesion site in the colonoscopy image, the control unit may send an instruction condition change command to the image analysis unit to either not indicate a diverticulum in the image of a normal colon, or to indicate a colonic diverticulum, and to indicate the examination time and retrieval time. [Effects of the Invention]

[0055] According to the present invention, when applying image recognition technology to the colonoscopy process to indicate the main examination area, the present invention provides the advantage of allowing the examiner to perform the examination while paying attention to complications such as perforation, which helps in setting the probe movement path, and enables the examiner to perform a more thorough colonoscopy.

[0056] Furthermore, by specifying the main colonoscopy sites, it is possible to confirm the sites that have been examined to date. By indicating whether or not a cecum examination has been performed, this can be used as an indicator of the quality of the colonoscopy. Additionally, because it is possible to record whether or not diverticula have been present, the examiner can perform the next colonoscopy with prior knowledge of the risk of perforation. [Brief explanation of the drawing]

[0057] [Figure 1] This diagram schematically shows the configuration of a colonoscopy site indication system according to one embodiment of the present invention. [Figure 2A] This figure shows an example of an image analysis model configuration. [Figure 2B] This figure shows an example of an image analysis model configuration. [Figure 2C] This figure shows an example of an image analysis model configuration. [Figure 2D] This figure shows an example of an image analysis model configuration. [Figure 3] This flowchart shows the execution process of a method for indicating the colonoscopy site according to one embodiment of the present invention. [Figure 4] This flowchart shows the execution process of the first modified example of the colonoscopy site indication method shown in Figure 3. [Figure 5] This flowchart shows the execution process of the second modified example of the colonoscopy site indication method shown in Figure 3. [Figure 6] This flowchart shows the execution process of the third modified example of the colonoscopy site indication method shown in Figure 3. [Figure 7] This figure schematically shows the configuration of a colonoscopy site indication system according to another embodiment of the present invention. [Figure 8] This flowchart shows the execution process of a method for indicating the colonoscopy site according to another embodiment of the present invention. [Figure 9] This flowchart shows the execution process of a modified version of the method for indicating the colonoscopy site according to another embodiment of the present invention. [Figure 10A] This diagram shows the main parts of the large intestine. [Figure 10B] This diagram shows the main parts of the large intestine. [Figure 10C] This diagram shows the main parts of the large intestine. [Figure 11] This figure shows an example of a representation of the large intestine. [Figure 12] This diagram shows the areas to be examined in the large intestine (in the case of a normal large intestine). [Figure 13] This diagram shows the location of the colon during examination (in the case of a left colonic diverticulum). [Figure 14] This diagram shows the location of the colon during examination (for left and right colonic diverticula). [Figure 15] This figure shows another example of how to indicate the examination site of the large intestine (in the case of a left colonic diverticulum). [Figure 16] This diagram shows the location of the colon during examination (for right and left colonic diverticula). [Figure 17A] This diagram shows the location of the colon examination (indicating whether or not diverticula were found, distinguishing between the left and right sides). [Figure 17B] This diagram shows the location of the colon examination (indicating whether or not diverticula were found, distinguishing between the left and right sides). [Figure 18A] This figure shows an example of analysis results provided (an example of a normal colon examination). [Figure 18B] This figure shows an example of analysis results provided (an example of a normal colon examination). [Figure 18C] This figure shows an example of analysis results provided (an example of a normal colon examination). [Figure 19A] This figure shows an example of analysis results (an example of finding a diverticulum in the cecum). [Figure 19B] This figure shows an example of analysis results (an example of finding a diverticulum in the cecum). [Figure 19C] This figure shows an example of analysis results (an example of finding a diverticulum in the cecum). [Figure 20A] This figure shows an example of analysis results provided (an example of finding diverticula in the sigmoid colon). [Figure 20B] This figure shows an example of analysis results provided (an example of finding a diverticulum in the sigmoid colon). [Figure 20C]This figure shows an example of analysis results provided (an example of finding a diverticulum in the sigmoid colon). [Figure 21A] This diagram shows an example of how analysis results are provided (an example of a results report). [Figure 21B] This diagram shows an example of how analysis results are provided (an example of a results report). [Figure 21C] This diagram shows an example of how analysis results are provided (an example of a results report). [Figure 21D] This diagram shows an example of how analysis results are provided (an example of a results report). [Modes for carrying out the invention]

[0058] Embodiments of the present invention will be described in detail below with reference to the attached drawings.

[0059] Figure 1 is a schematic diagram showing the configuration of a colonoscopy site indication system according to one embodiment of the present invention.

[0060] Referring to Figure 1, a colonoscopy site indication system 100 according to one embodiment of the present invention can be configured to include a model loading / condition setting unit 110, an image receiving unit 120, an image preprocessing unit 130, an image analysis unit 140, and a control unit 150.

[0061] The model loading / condition setting unit 110 loads the colonoscopy image analysis model and sets the analysis conditions for the analysis model. Here, the analysis conditions for the analysis model can be set to, for example, a predicted probability value of 0.85 or higher.

[0062] The image receiving unit 120 receives colonoscopy image frames.

[0063] The image preprocessing unit 130 preprocesses the colonoscopy images received via the image receiving unit 120 so that subsequent image analysis can be performed smoothly. Here, the preprocessing of the colonoscopy images by the image preprocessing unit 130 described above may include cropping of the analysis area and adjustment of the input size.

[0064] The image analysis unit 140 analyzes the colonoscopy image preprocessed by the image preprocessing unit 130 using an AI (artificial intelligence) based image analysis model, and detects and indicates at least one of the examination site, diverticulum site, and lesion site in the colonoscopy image based on the analysis results. Here, the image analysis model of such an image analysis unit 140 can be composed of a single image analysis model that detects the examination site and diverticulum site, as shown in Figure 2A. In Figure 2A, (a) shows an example of an internal indication from colon entry, and (b) shows an example of a cecum-centered indication.

[0065] Furthermore, the image analysis model of the image analysis unit 140 can be composed of an examination site detection model (Model A) for detecting the examination site and a diverticulum detection model (Model B) for detecting the diverticulum site, as shown in Figure 2B. In Figure 2B, (a) shows an example of internal indication from entry into the large intestine, and (b) shows an example of cecum-centered indication.

[0066] Furthermore, as shown in Figure 2C, the image analysis model of the image analysis unit 140 can consist of an examination site detection model (Model A) for detecting the examination site, a diverticulum detection model (Model B) for detecting the diverticulum site, and a lesion detection model (Model C) for detecting the lesion site. In this case, the lesion detection model (Model C) can be equipped with a lesion attribute classification function that distinguishes whether the lesion is benign or malignant.

[0067] Furthermore, the image analysis model of the image analysis unit 140 can consist of a cecum / diverticulum detection model (Model A) for detecting the cecum and diverticula, and a lesion detection model (Model B) for detecting lesion sites, as shown in Figure 2D. In this case, the lesion detection model (Model B) may be equipped with a lesion attribute classification function that distinguishes whether the lesion is benign or malignant.

[0068] Furthermore, when the image analysis unit 140 detects and indicates at least one of the examination site, diverticulum site, and lesion site in the colonoscopy image, the examination site may include the appendix, cecum, ascending colon, transverse colon, descending colon, sigmoid colon, and rectum, as shown in Figures 2A to 2C.

[0069] The control unit 150 controls the status checks and operations of the model loading / condition setting unit 110, the image receiving unit 120, the image preprocessing unit 130, and the image analysis unit 140. Once the model loading / condition setting unit 110 has completed loading the colonoscopy image analysis model and setting the analysis conditions for the analysis model, the control unit 150 initializes the analysis screen, displays an image of a normal colon, and provides the analysis results analyzed by the image analysis unit 140. In this case, when the image analysis unit 140 detects and indicates at least one of the examination site, diverticulum site, and lesion site in the colonoscopy image, the control unit 150 can send an instruction condition change command to the image analysis unit 140, which indicates the cecum if the appendix is ​​detected, and indicates a different diverticulum in the image of a normal colon.

[0070] Furthermore, the control unit 150 can send a command to the image analysis unit 140 to change the instruction conditions, which will either prevent the image of a normal colon from indicating a diverticulum, or indicate the presence of a colonic diverticulum (for example, an alarm and area indication), and to indicate the examination time and retrieval time, when the image analysis unit 140 detects and indicates at least one of the examination area, diverticulum area, and lesion area in the colonoscopy image.

[0071] In Figure 1, reference number 160 indicates a database (DB), which stores and manages various software programs for system operation, data and information necessary when the model loading / condition setting unit 110, image receiving unit 120, image preprocessing unit 130, and image analysis unit 140 perform functions or processes related to model loading and condition setting, image preprocessing, and image analysis, as well as colonoscopy image analysis result data from the image analysis model.

[0072] Here, the model loading / condition setting unit 110, image receiving unit 120, image preprocessing unit 130, image analysis unit 140, control unit 150, and database (DB) 160 described above may be integrated as a whole and configured as a single computer system.

[0073] The following describes a method for indicating a colonoscopy site based on a colonoscopy site indication system according to one embodiment of the present invention having the configuration described above.

[0074] Figure 3 is a flowchart showing the execution process of a method for indicating the colonoscopy site according to one embodiment of the present invention.

[0075] Referring to Figure 3, in the method for indicating a colonoscopy site according to one embodiment of the present invention, first, the model loading / condition setting unit 110 loads a colonoscopy image analysis model and sets the analysis conditions of the analysis model (for example, a predicted probability value of 0.85 or higher) (step S301).

[0076] Then, the control unit 150 initializes the analysis screen and displays an image of a normal colon (step S302).

[0077] As described above, once the model loading and analysis condition settings are complete, and the analysis screen has been initialized and a picture of a normal colon displayed, the control unit 150 determines whether or not colonoscopy image analysis is necessary (step S303). If colonoscopy image analysis is required in this determination, the image preprocessing unit 130 reads the colonoscopy image (image frame) received via the image receiving unit 120 and preprocesses it to enable smooth subsequent image analysis (step S304). Here, the preprocessing of the colonoscopy image by the image preprocessing unit 130 may include cutting the analysis area and adjusting the input size.

[0078] Once the preprocessing of the colonoscopy images is complete, the image analysis unit 140 analyzes the preprocessed colonoscopy images using an AI-based image analysis model (step S305). Here, as described above, the image analysis model can consist of a single image analysis model that detects the examination site and the diverticulum site, as shown in Figure 2A. Alternatively, the image analysis model can consist of an examination site detection model (Model A) that detects the examination site and a diverticulum detection model (Model B) that detects the diverticulum site, as shown in Figure 2B. Furthermore, the image analysis model can consist of an examination site detection model (Model A) that detects the examination site, a diverticulum detection model (Model B) that detects the diverticulum site, and a lesion detection model (Model C) that detects the lesion site, as shown in Figure 2C. In this case, the lesion detection model (Model C) can be equipped with a lesion attribute classification function that distinguishes whether the lesion is benign or malignant. Additionally, the image analysis model can consist of a cecum / diverticulum detection model (Model A) that detects the cecum and diverticula, and a lesion detection model (Model B) that detects the lesion site, as shown in Figure 2D. In this case, the lesion detection model (Model B) may be equipped with a lesion attribute classification function that distinguishes whether a lesion is benign or malignant.

[0079] Furthermore, the image analysis unit 140 detects and indicates at least one of the examination site, diverticulum site, and lesion site in the colonoscopy image based on the results of the analysis (steps S306 to S308). Here, when the image analysis unit 140 detects and indicates at least one of the examination site, diverticulum site, and lesion site in the colonoscopy image, the examination site may include the appendix, cecum, ascending colon, transverse colon, descending colon, sigmoid colon, and rectum.

[0080] Now, let's explain steps S306 to S308 in more detail.

[0081] In step S305, once the colonoscopy image analysis is completed by the image analysis unit 140, the control unit 150 determines whether or not a diverticulum has been detected (step S306). If a diverticulum is detected, the control unit changes the image of the normal colon to an image of a colon with diverticulosis and displays it (see Figures 13 to 16) (step S307), and indicates the examination site on the image of the colon (step S308).

[0082] Subsequently, if the determination in step S303 does not result in any further requests for colonoscopy image analysis, the control unit 150 provides the analysis results that were analyzed up to the immediate prior by the image analysis unit 140 (step S309).

[0083] On the other hand, Figure 4 is a flowchart showing the execution process of the first modified example of the colonoscopy site indication method shown in Figure 3.

[0084] Referring to Figure 4, this differs from Figure 3 in that it adds the processes of saving the start time of the examination (step S404), saving the start time of the cecum instruction (step S410), displaying the examination time and retrieval time (step S411), determining the change in instruction conditions (step S412), changing the instruction conditions (step S413), and saving the end time of the examination (step S414) to the process of Figure 3 described above. The rest of the process is the same as in Figure 3. Therefore, the explanation of the parts that are the same as in Figure 3 (steps S401-S403, S405-S409, S415) will be replaced by the explanation in Figure 3 (i.e., the explanation of steps S301-S303, S304-S308, S309), and only the parts that differ from Figure 3 will be explained.

[0085] If the determination in step S403 of Figure 4 indicates that colonoscopy image analysis is required, the control unit 150 stores the examination start time in the database 160 (step S404).

[0086] Subsequently, as explained in Figure 3, the image preprocessing unit 130 reads the colonoscopy images (image frames) received via the image receiving unit 120 and preprocesses them to facilitate subsequent image analysis (step S405).

[0087] Furthermore, after the control unit 150 indicates the examination site on the colon diagram (step S409), it saves the start time of the cecum indication (step S410). Subsequently, the control unit 150 displays the examination time and retrieval time (step S411). Here, the examination time refers to the time from the start time of the examination to the current (end) time, and the retrieval time refers to the time from the start time of the cecum indication to the current (end) time.

[0088] Furthermore, the control unit 150 determines whether or not the instruction conditions have been changed (step S412), and if a change in the instruction conditions is requested, it changes the instruction conditions (step S413). At this point, the control unit 150 can send an instruction condition change command to the image analysis unit 140, which instructs the image to indicate the cecum if it is the appendix, and to indicate a diverticulum in a normal colon image.

[0089] Furthermore, the control unit 150 can send a command to the image analysis unit 140 to change the instruction conditions, such as not indicating diverticula in the image of a normal colon, or indicating the presence of colonic diverticula (for example, an alarm or area indication), and specifying the examination time and retrieval time.

[0090] On the other hand, if the determination in step S403 does not result in any further requests for colonoscopy image analysis, the control unit 150 saves the examination completion time (step S414) and provides the analysis results up to the present (examination completion) (step S415).

[0091] Figure 5 is a flowchart showing the execution process of a second modified example of the colonoscopy site indication method shown in Figure 3.

[0092] Referring to Figure 5, this differs from Figure 3 in that it adds a step to determine whether or not a lesion is detected (S509) and a step to indicate lesion information on the colon image (S510) to the process described in Figure 3 above; the rest is the same as Figure 3. Therefore, here as well, the explanation of the parts that are the same as Figure 3 will be replaced by the explanation of Figure 3, and only the parts that differ from Figure 3 will be explained.

[0093] In Figure 5, the control unit 150 indicates the examination site on the image of the large intestine (step S508), then determines whether or not a lesion has been detected (step S509), and if a lesion is detected, indicates the lesion information on the image of the large intestine (step S510).

[0094] Figure 6 is a flowchart showing the execution process of a third modified example of the colonoscopy site indication method shown in Figure 3.

[0095] Referring to Figure 6, this differs from Figure 4 in that it adds a step to determine whether or not a lesion is detected (S611) and a step to indicate lesion information on the drawing of the large intestine (S612), while the rest of the process is identical to that of Figure 4.

[0096] In other words, if colonoscopy image analysis is required in the determination in step S603 of Figure 6, the control unit 150 stores the examination start time in the database 160 (step S604).

[0097] Subsequently, as explained in Figure 3, the image preprocessing unit 130 reads the colonoscopy images (image frames) received via the image receiving unit 120 and preprocesses them to facilitate subsequent image analysis (step S605).

[0098] Furthermore, after indicating the examination site on the image of the large intestine (step S609), the control unit 150 saves the start time of the cecum indication (step S610). Subsequently, the control unit 150 determines whether or not a lesion has been detected (step S611), and if a lesion is detected, it indicates the lesion information on the image of the large intestine (step S612).

[0099] Furthermore, the control unit 150 displays the examination time and the retrieval time (step S613). Here, the examination time refers to the time from the start time of the examination to the current (end) time, and the retrieval time refers to the time from the start time of the cecal instructing to the current (end) time.

[0100] Furthermore, the control unit 150 determines whether or not it is necessary to change the instruction conditions (step S614), and if a change in the instruction conditions is requested, it changes the instruction conditions (step S615). At this point, the control unit 150 can send an instruction condition change command to the image analysis unit 140, which instructs the image to indicate the cecum if it is the appendix, and to indicate a diverticulum in a normal colon image.

[0101] Furthermore, the control unit 150 can send a command to the image analysis unit 140 to change the instruction conditions, such as not indicating diverticula in the image of a normal colon, or indicating the presence of colonic diverticula (for example, an alarm or area indication), and specifying the examination time and retrieval time.

[0102] On the other hand, if there are no further requests for colonoscopy image analysis in the determination in step S603, the control unit 150 saves the examination completion time (step S616) and provides the analysis results up to the present (end of examination) (step S617).

[0103] Figure 7 is a schematic diagram showing the configuration of a colonoscopy site indication system according to another embodiment of the present invention.

[0104] Referring to Figure 7, the colonoscopy site indication system 700 according to another embodiment of the present invention has basically the same components as the colonoscopy site indication system 100 according to one embodiment described earlier with reference to Figure 1. However, this colonoscopy site indication system 700 according to another embodiment differs in that it further includes a voice recognition unit 750.

[0105] As shown in Figure 7, the colonoscopy site indication system 700 according to another embodiment of the present invention can be configured to include a model loading / condition setting unit 710, an image receiving unit 720, an image preprocessing unit 730, an image analysis unit 740, a voice recognition unit 750, and a control unit 760.

[0106] The model loading / condition setting unit 710 loads the colonoscopy image analysis model and sets the analysis conditions for the analysis model. Here, the analysis conditions for the analysis model can be set, for example, to a predicted probability value of 0.85 or higher.

[0107] The image receiving unit 720 receives colonoscopy image frames.

[0108] The image preprocessing unit 730 preprocesses the colonoscopy images received via the image receiving unit 720 to facilitate subsequent image analysis. Here, the preprocessing of the colonoscopy images by the image preprocessing unit 730 described above may include cropping of the analysis area and adjustment of the input size.

[0109] The image analysis unit 740 analyzes the colonoscopy image preprocessed by the image preprocessing unit 730 using an AI (artificial intelligence) based image analysis model, and detects and indicates at least one of the examination site, diverticulum site, and lesion site in the colonoscopy image based on the analysis results. Here, the image analysis model of such an image analysis unit 740 can be composed of a single image analysis model that detects the examination site and diverticulum site, as shown in Figure 2A. In Figure 2A, (a) shows an example of an internal indication from the entry into the colon, and (b) shows an example of a central indication of the cecum.

[0110] Furthermore, the image analysis model of the image analysis unit 740 can be composed of an examination site detection model (Model A) for detecting the examination site and a diverticulum detection model (Model B) for detecting the diverticulum site, as shown in Figure 2B. In Figure 2B, (a) shows an example of internal indication from entry into the large intestine, and (b) shows an example of cecum-centered indication.

[0111] Furthermore, as shown in Figure 2C, the image analysis model of the image analysis unit 740 can consist of an examination site detection model (Model A) for detecting the examination site, a diverticulum detection model (Model B) for detecting the diverticulum site, and a lesion detection model (Model C) for detecting the lesion site. In this case, the lesion detection model (Model C) may be equipped with a lesion attribute classification function that distinguishes whether the lesion is benign or malignant.

[0112] Furthermore, the image analysis model of the image analysis unit 740 can be composed of a cecum / diverticulum detection model (Model A) for detecting the cecum and diverticulum, and a lesion detection model (Model B) for detecting lesion sites, as shown in Figure 2D. In this case, the lesion detection model (Model B) may be equipped with a lesion attribute classification function that distinguishes whether the lesion is benign or malignant.

[0113] Furthermore, when the image analysis unit 740 detects and indicates at least one of the examination site, diverticulum site, and lesion site in the colonoscopy image, the examination site includes the appendix, cecum, ascending colon, transverse colon, descending colon, sigmoid colon, and rectum.

[0114] The speech recognition unit 750 reads audio from a buffer (located in the internal memory of the speech recognition unit 750) that stores audio while the image analysis unit 740 is performing image analysis, analyzes it using an AI-based speech keyword recognition model, recognizes speech keywords based on the analysis results, and transmits them to the image analysis unit 740.

[0115] The control unit 760 controls the status checks and operations of the model loading / condition setting unit 710, image receiving unit 720, image preprocessing unit 730, image analysis unit 740, and voice recognition unit 750. Once the model loading / condition setting unit 710 has completed loading the colonoscopy image analysis model and setting the analysis conditions for the analysis model, the control unit 760 initializes the analysis screen, displays an image of a normal colon, and provides the analysis results analyzed by the image analysis unit 740. The analysis results are provided by linking the analysis target detected by the image analysis model with the voice commands (keywords) related to the analysis target spoken by the examiner.

[0116] Here, the control unit 760 described above can transmit a command to the image analysis unit 740 to change the instruction conditions when the image analysis unit 740 detects and instructs at least one of the examination site, diverticulum site, and lesion site in the colonoscopy image, to instruct the cecum if the appendix is ​​detected, and to instruct a different diverticulum in the image of a normal colon.

[0117] Furthermore, the control unit 760 can send a command to the image analysis unit 740 to change the instruction conditions, which will either prevent the image of a normal colon from indicating a diverticulum, or indicate the presence of a colonic diverticulum (for example, by an alarm or area indication), and to indicate the examination time and retrieval time, when the image analysis unit 740 detects and indicates at least one of the examination area, diverticulum area, and lesion area in the colonoscopy image.

[0118] In Figure 7, reference number 770 indicates a database (DB), which stores and manages various software programs for system operation, as well as data and information necessary when the model loading / condition setting unit 710, image receiving unit 720, image preprocessing unit 730, image analysis unit 740, image recognition unit 740, and speech recognition unit 750 perform functions or processes related to model loading and condition setting, image preprocessing, image analysis, and speech recognition, and colonoscopy image analysis result data from the image analysis model.

[0119] Here, the model loading / condition setting unit 710, image receiving unit 720, image preprocessing unit 730, image analysis unit 740, speech recognition unit 750, control unit 760, and database (DB) 770 described above may be integrated as a whole and configured as a single computer system.

[0120] Next, a method for indicating a colonoscopy site based on another embodiment of the present invention having the configuration described above will be explained. Figure 8 is a flowchart showing the execution process of a method for indicating the colonoscopy site according to another embodiment of the present invention.

[0121] Referring to Figure 8, in another embodiment of the present invention, the method for indicating a colonoscopy site involves the model loading / condition setting unit 710 first loading a colonoscopy image analysis model and a voice recognition model, and setting the analysis conditions for the analysis model (for example, a predicted probability value of 0.85 or higher) (step S801).

[0122] Furthermore, the control unit 760 initializes the analysis screen and displays an image of a normal large intestine (step S802).

[0123] As described above, once the loading of the colonoscopy image analysis model, the loading of the speech recognition model, and the setting of the model's analysis conditions are complete, and the analysis screen has been initialized and a picture of a normal colon has been displayed, the control unit 760 determines whether or not colonoscopy image analysis is necessary (step S803). If colonoscopy image analysis is required in this determination, the image preprocessing unit 730 reads the colonoscopy image (image frame) received via the image receiving unit 720 and preprocesses it to enable smooth subsequent image analysis (step S804). Here, the preprocessing of the colonoscopy image by the image preprocessing unit 730 may include cutting the analysis area and adjusting the input size.

[0124] Once the preprocessing of the colonoscopy images is complete, the image analysis unit 740 analyzes the preprocessed colonoscopy images using an AI-based image analysis model (step S805). Here, as described above, the image analysis model can consist of a single image analysis model that detects the examination site and the diverticulum site, as shown in Figure 2A. Alternatively, the image analysis model can consist of an examination site detection model (Model A) that detects the examination site and a diverticulum detection model (Model B) that detects the diverticulum site, as shown in Figure 2B. Furthermore, the image analysis model can consist of an examination site detection model (Model A) that detects the examination site, a diverticulum detection model (Model B) that detects the diverticulum site, and a lesion detection model (Model C) that detects the lesion site, as shown in Figure 2C. In this case, the lesion detection model (Model C) can be equipped with a lesion attribute classification function that distinguishes whether the lesion is benign or malignant. Additionally, the image analysis model can consist of a cecum / diverticulum detection model (Model A) that detects the cecum and diverticula, and a lesion detection model (Model B) that detects the lesion site, as shown in Figure 2D. In this case, the lesion detection model (Model B) may include a lesion attribute classification function that distinguishes whether a lesion is benign or malignant.

[0125] Furthermore, the image analysis unit 740 detects and indicates at least one of the examination site, diverticulum site, and lesion site in the colonoscopy image based on the results of the analysis (steps S806 to S808). Here, when the image analysis unit 740 detects and indicates at least one of the examination site, diverticulum site, and lesion site in the colonoscopy image, the examination site may include the appendix, cecum, ascending colon, transverse colon, descending colon, sigmoid colon, and rectum.

[0126] Now, let's explain steps S806 to S808 in more detail.

[0127] In step S805, once the colonoscopy image analysis is completed by the image analysis unit 740, the control unit 760 determines whether or not diverticula have been detected (step S806). If diverticula are detected, the control unit changes the image of the normal colon to an image of a colon with diverticulosis and displays it (see Figures 13 to 16) (step S807), and indicates the examination site on the image of the colon (S808).

[0128] On the other hand, if colonoscopy image analysis is required in the determination in step S803, the voice recognition unit 750 reads audio from the buffer that stores audio while the image analysis unit 740 is performing image analysis, analyzes it using an AI-based voice keyword recognition model (step S809), recognizes voice keywords based on the analysis results (step S810), and transmits them to the image analysis unit 740.

[0129] In other words, the voice recognition unit 750 recognizes a voice keyword and determines whether or not that voice keyword is a diverticulum keyword (step S811). If it is a diverticulum keyword, it transmits the information to step S807 so that it is reflected in changing the image to one of a diverticular colon.

[0130] Furthermore, if the above determination does not identify the diverticulum keyword, the voice recognition unit 750 determines whether or not it is an examination site keyword (step S812). If it is an examination site keyword, it transmits this information to step S808 so that it is reflected in indicating the examination site on the picture of the large intestine.

[0131] Thereafter, if there are no further requests for colonoscopy image analysis in the determination in step S803, the control unit 760 provides the analysis results up to the present (end of examination) by the image analysis unit 740 by linking the analysis target detected by the image analysis model of the image analysis unit 740 with the voice command (keyword) related to the analysis target spoken by the examiner (step S813).

[0132] Figure 9 is a flowchart showing the execution process of a modified version of the colonoscopy site indication method according to another embodiment of the present invention.

[0133] Referring to Figure 9, this differs from Figure 8 in that it adds the following processes to the process in Figure 8 described above: saving the start time of the examination (step S904), saving the start time of the cecum instruction (step S910), determining whether a lesion is detected or not (step S911), indicating lesion information on the colon map (step S912), displaying the examination time and retrieval time (step S913), determining whether the instruction conditions need to be changed (step S914), changing the instruction conditions (step S915), and saving the end time of the examination (step S920). The rest of the process is the same as in Figure 8. Therefore, the explanation for the parts that are the same as in Figure 8 will be replaced by the explanation in Figure 8, and only the parts that differ from Figure 8 will be explained.

[0134] If colonoscopy image analysis is required in step S903 of Figure 9, the control unit 760 stores the examination start time in the database 770 (step S904).

[0135] Subsequently, as explained in Figure 8, the image preprocessing unit 730 reads the colonoscopy images (image frames) received via the image receiving unit 720 and preprocesses them to facilitate subsequent image analysis (step S905).

[0136] Furthermore, after indicating the examination site on the image of the large intestine (step S909), the control unit 760 saves the start time of the cecum indication (step S910). Subsequently, the control unit 760 determines whether or not a lesion has been detected (step S911), and if a lesion is detected, it indicates the lesion information on the image of the large intestine (step S912).

[0137] Furthermore, the control unit 760 displays the examination time and the retrieval time (step S913). Here, the examination time refers to the time from the start time of the examination to the current (end) time, and the retrieval time refers to the time from the start time of the cecal instructing to the current (end) time.

[0138] Furthermore, the control unit 760 determines whether or not it is necessary to change the instruction conditions (step S914), and if a change in the instruction conditions is requested, it changes the instruction conditions (step S915). At this point, the control unit 760 can send an instruction condition change command to the image analysis unit 740, which instructs the image to indicate the cecum if it is the appendix, and to indicate a diverticulum in a normal colon image.

[0139] Furthermore, the control unit 760 can send a command to the image analysis unit 740 to change the instruction conditions, such as not indicating diverticula in the image of a normal colon, or indicating the presence of colonic diverticula (for example, an alarm or area indication), and specifying the examination time and retrieval time.

[0140] On the other hand, if there are no further requests for colonoscopy image analysis in the determination in step S903, the control unit 760 saves the examination completion time (step S920) and provides the analysis results up to the present (end of examination) (step S921).

[0141] The following provides further explanation regarding the colonoscopy site indication system and method according to the present invention, as described above.

[0142] Figures 10A, 10B, and 10C show the main parts of the large intestine.

[0143] Referring to Figures 10A, 10B, and 10C, Figure 10A shows the main parts of the large intestine, Figure 10B shows the examination procedure during probe insertion (direction for colonoscopy photography), and Figure 10C shows the examination procedure during probe retrieval.

[0144] Referring to Figures 10A and 10B, the colonoscope is inserted through the rectum and enters the large intestine. Once inside the large intestine, the colonoscope images the inside of the large intestine in the following order: sigmoid colon → descending colon → transverse colon → ascending colon → cecum → appendix.

[0145] Referring to Figures 10A and 10C, this is an examination of the probe retrieval process, in which the colonoscope is inserted to the end of the large intestine and then retrieved while photographing the inside of the large intestine in the order of appendix → cecum → ascending colon → transverse colon → descending colon → sigmoid colon → rectum. In Figure 10A, the ileocecal valve is the part where the cecum and small intestine join.

[0146] In the series of colonoscopy probe movements and image acquisition described above, when the probe is inserted to position 8 in the probe insertion process shown in Figure 10B, it is assumed that the cecum has been confirmed. Then, mainly after confirming the cecum, the examination is performed while retrieving the probe. A skilled examiner will also observe while inserting the probe.

[0147] Figure 11 shows an example of a representation of the large intestine.

[0148] Referring to Figure 11, (a) shows a normal colon, (b) shows a left colonic diverticulum, and (c) shows a right colonic diverticulum. Left colonic diverticulum (b) mainly occurs in the sigmoid colon and descending colon, while right colonic diverticulum (c) mainly occurs in the cecum.

[0149] Here, the diverticula mentioned above are induced by increased pressure within the large intestine. (b) The left colon-developing diverticulum (pseudodiverticulum) shows that a portion of the barrier (mucosal layer and submucosal tissue) protrudes, resulting in the development of multiple diverticula in the left large intestine. This is acquired and commonly occurs in Westerners, but recently it has been showing an increasing trend in East Asians. (c) The right colon-developing diverticulum (true diverticulum) shows that the entire barrier, including the muscular layer, protrudes, resulting in the development of one diverticulum in the right large intestine. This is congenital and commonly occurs in East Asians.

[0150] Figure 12 shows the indicators for the colon examination site (in the case of a normal colon).

[0151] Referring to Figure 12, this shows how to indicate the examination site during the colonoscopy probe insertion process. Generally, the inside of the large intestine is observed (examined) in the following order: start of examination → rectal examination → sigmoid colon examination → descending colon examination → transverse colon examination → ascending colon examination → cecum examination.

[0152] Figure 13 shows the location of the colon during examination (in the case of a left colonic diverticulum).

[0153] Referring to Figure 13, this shows how to indicate the examination site during the colonoscopy probe insertion process. Similar to the normal colon in Figure 12, the inside of the colon is observed (examined) in the following order: start of examination → rectal examination → sigmoid colon examination → descending colon examination → transverse colon examination → ascending colon examination → cecum examination. In this example, a diverticulum is found in the sigmoid colon.

[0154] Figure 14 shows the indication of the examination site in the large intestine (for left and right colonic diverticula).

[0155] Referring to Figure 14, this shows how to indicate the examination site during the colonoscopy probe insertion process. Similarly, the inside of the large intestine is observed (examined) in the following order: start of examination → rectal examination → sigmoid colon examination → descending colon examination → transverse colon examination → ascending colon examination → cecum examination. In this example, diverticula are found in the sigmoid colon, descending colon, and cecum.

[0156] Figure 15 shows another example of how to indicate the examination site in the large intestine (in the case of a left colonic diverticulum).

[0157] Referring to Figure 15, this shows how to indicate the examination site during the colonoscopy probe insertion process. Similarly, the inside of the large intestine is observed (examined) in the following order: start of examination → rectal examination → sigmoid colon examination → descending colon examination → transverse colon examination → ascending colon examination → cecum examination. In this example, only the presence or absence of cecal insertion is confirmed, and a diverticulum is found in the sigmoid colon.

[0158] Figure 16 shows the indication of the examination site in the large intestine (for right and left colonic diverticula).

[0159] Referring to Figure 16, this shows how to indicate the examination site during the colonoscopy probe retrieval process. The inside of the large intestine is observed (examined) in the following order: start of examination → cecum observation → ascending colon observation → transverse colon observation → descending colon observation → sigmoid colon observation → rectum observation (i.e., while retrieving the probe in reverse order).

[0160] Figures 17A and 17B show the indication of the examination site in the large intestine (distinguishing between left and right sides to indicate whether or not diverticula were found).

[0161] Referring to Figures 17A and 17B, this shows a simplified indication of whether or not diverticula have been found, distinguishing only between the left and right colon. As shown in the figures, if diverticula are found in the right colon, left colon, right colon, or left colon, a specific area of ​​the left, right, or left / right colon is indicated so that the examiner can visually confirm it. Figure 17A shows the main examination area indicated, and Figure 17B shows the confirmation of whether or not cecal insertion was performed.

[0162] Figures 18A, 18B, and 18C show examples of analysis results (examples of normal colon examinations).

[0163] Referring to Figures 18A, 18B, and 18C, this shows that the analysis results of a normal colonoscopy are provided, where (a) shows the colonoscopy start screen, (b) shows a screen showing an example of cecal examination during the probe insertion process, and (c) shows a screen showing an example of cecal examination during the probe retrieval process. In Figures 18A, 18B, and 18C, reference number 810 indicates the colonoscopy image analysis software screen, 820 indicates the colonoscopy image analysis area, 830 indicates the ileocecal valve, 840 indicates the appendix, and 850 indicates the cecum. Also, T indicates the examination time and W indicates the retrieval time.

[0164] Figures 19A, 19B, and 19C illustrate examples of analysis results (examples of diverticulum detection in the cecum).

[0165] Referring to Figures 19A, 19B, and 19C, this shows the results of the analysis of diverticulum detection in the cecum. Similarly, Figure 19A shows the colonoscopy start screen, Figure 19B shows an example of cecal examination during probe insertion, and Figure 19C shows an example of cecal examination during probe retrieval. In Figures 19A, 19B, and 19C, reference numbers 810 represent the colonoscopy image analysis software screen, 820 represent the colonoscopy image analysis area, 830 represent the ileocecal valve, 840 represent the appendix, 850 represent the cecum, and 860 represent the diverticulum. Also, T represents the examination time and W represents the retrieval time.

[0166] Figures 20A, 20B, and 20C show examples of analysis results (examples of diverticulum detection in the sigmoid colon).

[0167] Referring to Figures 20A, 20B, and 20C, this shows the results of the analysis of diverticulum detection in the sigmoid colon. Similarly, Figure 20A shows the start screen of the colonoscopy, Figure 20B shows an example of sigmoid colon examination during the probe insertion process, and Figure 20C shows an example of sigmoid colon examination during the probe retrieval process. In Figures 20A, 20B, and 20C, reference number 810 indicates the colonoscopy image analysis software screen, 860 indicates the diverticulum, and 870 indicates the sigmoid colon. Also, T indicates the examination time and W indicates the retrieval time.

[0168] Figures 21A, 21B, 21C, and 21D show examples of how analysis results are provided (examples of result reports).

[0169] Referring to Figures 21A, 21B, 21C, and 21D, the colonoscopy results are as follows: Figure 21A shows the probe insertion and indication of the examination site, indicating that a cecal examination was performed and there were no colonic diverticula; Figure 21B shows whether or not to perform a cecal examination, indicating that a cecal examination was performed and there were no cecal diverticula; Figure 21C shows whether or not to perform a cecal examination and indication of diverticula, indicating that a cecal examination was performed and there were colonic diverticula in the cecum; and Figure 21D shows whether or not to perform a cecal examination and a simplified indication of left colonic diverticula, indicating that a cecal examination was performed and there were colonic diverticula in the left colon.

[0170] As described above, the colonoscopy site indication system and method according to the present invention have the advantage of allowing the examiner to perform the examination while paying attention to complications such as perforation, and can be used to set the probe movement path, when indicating the main examination site by applying image recognition technology to the colonoscopy process, by informing the examiner of a situation such as a diverticulum where the colon wall protrudes or by indicating (displaying) on ​​the examination screen that it is different from a normal colon.

[0171] Furthermore, by specifying the main colonoscopy sites, it is possible to confirm the sites that have been examined to date. By indicating whether or not a cecum examination has been performed, this can be used as an indicator of the quality of the colonoscopy. Additionally, the presence or absence of diverticula can be recorded, which has the advantage of allowing the examiner to proceed with the next colonoscopy knowing the risk of perforation in advance.

[0172] Although the present invention has been described in detail above with respect to preferred embodiments, it is obvious to a person of the ordinary skill in the art that the present invention is not limited thereto and can be modified and applied in various ways without departing from the technical idea of ​​the present invention. Therefore, the true scope of protection of the present invention should be interpreted by the following claims, and all technical ideas within an equivalent scope should be interpreted as being included within the scope of the rights of the present invention.

Claims

1. A model loading / condition setting unit loads a colonoscopy image analysis model and sets the analysis conditions for the analysis model, An image receiving unit that receives colonoscopy image frames, An image preprocessing unit that preprocesses colonoscopy images received via the image receiving unit to facilitate subsequent image analysis, An image analysis unit analyzes the colonoscopy image preprocessed by the image preprocessing unit using an AI (artificial intelligence)-based image analysis model, and detects and indicates at least one of the examination site, diverticulum site, and lesion site in the colonoscopy image based on the analysis results. A colonoscopy site indication system comprising: a control unit that controls the status check and operation of the model loading / condition setting unit, image receiving unit, image preprocessing unit, and image analysis unit, and when the model loading / condition setting unit completes loading of the colonoscopy image analysis model and setting of analysis conditions for the analysis model, it initializes the analysis screen, displays an image of a normal colon, and provides the analysis results analyzed by the image analysis unit.

2. The colonoscopy site indication system according to claim 1, wherein the preprocessing of the colonoscopy image by the image preprocessing unit includes cutting the analysis area and adjusting the input size.

3. The colonoscopy site indication system according to claim 1, characterized in that the image analysis model of the image analysis unit consists of a single image analysis model for detecting the examination site and the diverticulum site.

4. The colonoscopy examination site indication system according to claim 1, characterized in that the image analysis model of the image analysis unit comprises an examination site detection model for detecting the examination site and a diverticulum detection model for detecting the diverticulum site.

5. The colonoscopy examination site indication system according to claim 1, characterized in that the image analysis model of the image analysis unit comprises an examination site detection model for detecting the examination site, a diverticulum detection model for detecting the diverticulum site, and a lesion detection model for detecting the lesion site.

6. The colonoscopy site indication system according to claim 1, characterized in that the image analysis model of the image analysis unit comprises a cecum / diverticulum detection model for detecting the cecum and diverticula, and a lesion detection model for detecting lesion sites.

7. The colonoscopy site designation system according to claim 6, characterized in that the lesion detection model includes a lesion attribute classification function that distinguishes whether a lesion is benign or malignant.

8. The colonoscopy examination site designation system according to claim 1, wherein when the image analysis unit detects and designates at least one of the examination site, diverticulum site, and lesion site in the colonoscopy image, the examination site includes the appendix, cecum, ascending colon, transverse colon, descending colon, sigmoid colon, and rectum.

9. The colonoscopy examination site indication system according to claim 1, characterized in that when the image analysis unit detects and indicates at least one of the examination site, diverticulum site, and lesion site in the colonoscopy image, the control unit transmits an instruction condition change command to the image analysis unit to indicate the cecum if the appendix is ​​detected, and to indicate a different diverticulum in an image of a normal colon.

10. The colonoscopy examination site indication system according to claim 1, characterized in that when the image analysis unit detects and indicates at least one of the examination site, diverticulum site, and lesion site in the colonoscopy image, the control unit transmits an instruction condition change command to the image analysis unit to either not indicate a diverticulum in the image of a normal colon, or to indicate a colonic diverticulum, and to indicate the examination time and retrieval time.

11. A model loading / condition setting unit loads a colonoscopy image analysis model and a voice keyword recognition model, and sets the analysis conditions for the analysis models. An image receiving unit that receives colonoscopy image frames, An image preprocessing unit that preprocesses colonoscopy images received via the image receiving unit to facilitate subsequent image analysis, An image analysis unit analyzes the colonoscopy image preprocessed by the image preprocessing unit using an AI (artificial intelligence) based image analysis model, and detects and indicates at least one of the examination site, diverticulum site, or lesion site in the colonoscopy image based on the analysis results. While the image analysis unit performs image analysis, the speech recognition unit reads audio from a buffer that stores audio, analyzes it using an AI-based speech keyword recognition model, recognizes speech keywords based on the analysis results, and transmits them to the image analysis unit. A colonoscopy site indication system comprising: a control unit that controls the status check and operation of the model loading / condition setting unit, image receiving unit, image preprocessing unit, image analysis unit, and voice recognition unit, and when the model loading / condition setting unit completes loading of the colonoscopy image analysis model and setting of analysis conditions for the analysis model, initializes the analysis screen, displays an image of a normal colon, and provides the analysis results analyzed by the image analysis unit, and provides the analysis results by linking the analysis target detected by the image analysis model with the voice command (keyword) related to the analysis target spoken by the examiner.

12. The colonoscopy site indication system according to claim 11, wherein the preprocessing of the colonoscopy image by the image preprocessing unit includes cutting the analysis region and adjusting the input size.

13. The colonoscopy site indication system according to claim 11, characterized in that the image analysis model of the image analysis unit is composed of a single image analysis model for detecting the examination site and the diverticulum site.

14. The colonoscopy examination site indication system according to claim 11, characterized in that the image analysis model of the image analysis unit comprises an examination site detection model for detecting the examination site and a diverticulum detection model for detecting the diverticulum site.

15. The colonoscopy examination site indication system according to claim 11, characterized in that the image analysis model of the image analysis unit comprises an examination site detection model for detecting the examination site, a diverticulum detection model for detecting the diverticulum site, and a lesion detection model for detecting the lesion site.

16. The colonoscopy site indication system according to claim 11, characterized in that the image analysis model of the image analysis unit comprises a cecum / diverticulum detection model for detecting the cecum and diverticula, and a lesion detection model for detecting lesion sites.

17. The colonoscopy site indication system according to claim 16, characterized in that the lesion detection model includes a lesion attribute classification function that distinguishes whether a lesion is benign or malignant.

18. The colonoscopy examination site indicating system according to claim 11, wherein when the image analysis unit detects and indicates at least one of the examination site, diverticulum site, and lesion site in the colonoscopy image, the examination site includes the appendix, cecum, ascending colon, transverse colon, descending colon, sigmoid colon, and rectum.

19. The colonoscopy examination site indication system according to claim 11, characterized in that when the image analysis unit detects and indicates at least one of the examination site, diverticulum site, and lesion site in the colonoscopy image, the control unit transmits an instruction condition change command to the image analysis unit to indicate the cecum if the appendix is ​​detected, and to indicate a different diverticulum in the image of a normal colon.

20. The colonoscopy examination site indication system according to claim 11, characterized in that when the image analysis unit detects and indicates at least one of the examination site, diverticulum site, and lesion site in the colonoscopy image, the control unit transmits an instruction condition change command to the image analysis unit to either not indicate a diverticulum in the image of a normal colon, or to indicate a colonic diverticulum, and to indicate the examination time and retrieval time.

Images