Dental lesion information visualization method and system

Abstract

The invention relates to a dental lesion detection method and a system using the same. A dental lesion information visualization method according to the present disclosure may comprise the steps of: acquiring data of a lesion analysis model that outputs data on a type and a position of a lesion included in a panoramic image obtained by capturing an image of a patient's oral cavity; and inputting the panoramic image to a lesion analysis model to output data about the type and location of the lesion and the lesion detection confidence score.

Description

Technical Field

[0001] The present disclosure relates to a method for visualizing dental lesion information and a system applying the method. More specifically, the present disclosure relates to a method for identifying dental lesions by analyzing an image of a patient's oral cavity and a system applying the method. Background Art

[0002] Lesions that can be diagnosed through dental imaging related to teeth include various types, such as caries, periodontal disease, fissures, and apical periodontitis. Among these different lesions, in order for the dentist to accurately determine the exact location and size of the lesion related to the tooth, computed tomography (CT) imaging may be required.

[0003] However, CT imaging often raises concerns among patients due to the high cost of the procedure and concerns about radiation exposure. In addition, detecting lesions in CT images requires dentists to spend a lot of time analyzing the images.

[0004] Therefore, when a patient goes to a dental clinic, a panoramic image is first captured, and then the dentist visually checks the lesion in the oral cavity by analyzing the panoramic image.

[0005] However, since the panoramic image is created by unfolding the predetermined arches of the maxillary and mandibular bones onto a plane, it may be difficult for a dentist to visually identify the lesion on the panoramic image if the lesion is located in an area overlapping with adjacent structures due to the anatomical structure. In order to help dentists reduce the possibility of being unable to visually identify the lesion, an artificial intelligence (AI)-based method for visualizing dental lesion information may be considered. However, as mentioned above, if the lesion is located in an overlapping area due to the anatomical structure, the accuracy of the AI-based dental lesion visualization may also be reduced.

[0006] As mentioned above, in some cases, such as when the lesion is located in an area that overlaps with adjacent structures, both visual recognition of panoramic images and AI-based visualization of dental lesions may be challenging. Despite this, there is currently no available panoramic imaging viewer software or panoramic imaging device that can clearly inform the dentist whether it is a situation where visualization of dental lesion information will be difficult. As a result, unnecessary CT scans may be performed, resulting in a waste of time and cost. Summary of the invention

[0007] [Technical issues]

[0008] An object to be achieved by some embodiments of the present disclosure is to provide a method for displaying data related to a suspected area of ​​an unclear dental lesion on a panoramic image and a computing system applying the method.

[0009] Another object to be achieved by some embodiments of the present disclosure is to provide a method for providing a panoramic image viewer and a computing system applying the method, wherein the panoramic image viewer provides a user interface for activating a computed tomography (CT) image viewer or a CT imaging module for a suspected dental lesion area on a panoramic image analyzed by an artificial intelligence (AI) model of machine learning with minimal intuitive operations.

[0010] Another object to be achieved through some embodiments of the present disclosure is to provide a method for analyzing a patient's panoramic image using a dental lesion information visualization method based on machine learning (in which multiple artificial neural network models collaborate to improve lesion identification accuracy) and a computing system applying the method.

[0011] Another object to be achieved by some embodiments of the present disclosure is to provide a method of pre-processing a panoramic image to perform visualization of dental lesion information of a patient.

[0012] Another object to be achieved by some embodiments of the present disclosure is to provide a method for identifying the severity of a patient's dental lesions by analyzing a panoramic image of the patient's oral cavity.

[0013] Another object to be achieved by some embodiments of the present disclosure is to provide a method for generating and providing a list of lesions identified in a patient's oral cavity by analyzing a panoramic image of the patient's oral cavity.

[0014] Another object to be achieved by some embodiments of the present disclosure is to provide a method using an AI model to automatically identify lesions that are located in overlapping areas between anatomical structures in a panoramic image of a patient and are therefore difficult for a dentist to visually assess.

[0015] The objects of the present disclosure are not limited to the above-mentioned objects, and other objects not stated herein should be clearly understood by a person of ordinary skill in the art to which the present disclosure belongs from the following description.

[0016] [Technical solution]

[0017] A method for visualizing dental lesion information according to one embodiment of the present disclosure may include the following steps: inputting data about a panoramic image of a patient's mouth into a machine-learned lesion analysis model, determining a region of interest of the panoramic image by using the data output by the lesion analysis model, and superimposing and displaying an indicator pointing to the region of interest on the panoramic image.

[0018] In one embodiment, the step of determining the region of interest may include the following steps: determining teeth that need to be verified and whose lesion detection confidence score is less than or equal to a reference value by using data output by a lesion analysis model; identifying the region of the teeth that need to be verified in the panoramic image; and designating a portion of the region of the teeth that need to be verified as the region of interest.

[0019] In one embodiment, the area of ​​the tooth that needs to be verified may include a first sub-area, a second sub-area, and a third sub-area obtained by sequentially dividing the area from the crown toward the root in a direction perpendicular to the tooth axis, and the step of designating a part of the area of ​​the tooth that needs to be verified as the area of ​​interest may include: the step of designating a sub-area corresponding to the lesion type of the tooth that needs to be verified as the area of ​​interest in the first to third sub-areas, wherein the lesion type of the tooth that needs to be verified is determined using data output by the lesion analysis model.

[0020] In one embodiment, the step of designating a sub-region corresponding to the lesion type of the tooth to be verified as a region of interest may include the following steps: when the lesion type of the tooth to be verified is a caries type, designating a first sub-region as a region of interest; when the lesion type of the tooth to be verified is a periodontal disease type, designating a second sub-region as a region of interest; and when the lesion type of the tooth to be verified is a periapical lesion type, designating a third sub-region as a region of interest.

[0021] In one embodiment, the area of ​​the tooth to be verified may include a first sub-area, a second sub-area, and a third sub-area obtained by sequentially dividing the area from the crown toward the root in a direction perpendicular to the tooth axis, the first sub-area may include 1-1 sub-areas to 1-N sub-areas (where N is a natural number greater than or equal to 2) obtained by dividing the first sub-area into N sub-areas in a direction parallel to the tooth axis, and the second sub-area may include 2-1 sub-areas to 2-N sub-areas (where N is a natural number greater than or equal to 2) obtained by dividing the second sub-area into N sub-areas in a direction parallel to the tooth axis. , N is a natural number greater than or equal to 2), the third sub-region may include 3-1 sub-regions to 3-N sub-regions obtained by dividing the third sub-region into N sub-regions in a direction parallel to the tooth axis direction (where N is a natural number greater than or equal to 2), and the step of designating a part of the area of ​​the tooth to be verified as the area of ​​interest may include: the step of designating one of 1-1 sub-regions to 1-N sub-regions, 2-1 sub-regions to 2-N sub-regions and 3-1 sub-regions to 3-N sub-regions as the lesion position of the tooth to be verified using data output by the lesion analysis model.

[0022] In one embodiment, the dental lesion information visualization method may further include the step of displaying a computed tomography (CT) image of the patient's oral cavity in response to a user's selection input of the indicator.

[0023] In one embodiment, the step of displaying the CT image of the oral cavity may include displaying the CT image with a field of view (FoV) initially set based on the location of the region of interest.

[0024] In one embodiment, the dental lesion information visualization method may further include the following steps: in response to a user's selection input of an indicator, searching for a CT image of the patient's mouth on a server; and in response to determining that there is no CT image of the patient's mouth based on the search results, running a CT imaging module.

[0025] According to another embodiment of the present disclosure, a method for visualizing dental lesion information may include the following steps: inputting data about a panoramic image of a patient's mouth into a machine-learned lesion analysis model, using the data output by the lesion analysis model to generate a lesion detection list, and displaying the lesion detection list together with the panoramic image.

[0026] The lesion detection list may include tooth identification numbers of teeth in which lesions were identified, types of lesions detected, and lesion detection confidence scores for each identified lesion.

[0027] In one embodiment, the dental lesion information visualization method may further include the following steps: in response to a user's selection input of a record of a first lesion included in a lesion detection list, searching for a CT image of the patient's mouth; and in response to determining that there is no CT image of the patient's mouth based on the search results, running the CT imaging module.

[0028] According to another embodiment of the present disclosure, a method for visualizing dental lesion information may include the following steps: acquiring data of a lesion analysis model, which outputs data about the type and location of lesions included in a panoramic image of a patient's mouth, the lesion analysis model including a first artificial neural network, a second artificial neural network, and a third artificial neural network; inputting the data of the panoramic image into the first artificial neural network, and performing segmentation processing on each tooth area included in the panoramic image by using the data output from the first artificial neural network; inputting the segmented panoramic image data into the second artificial neural network, and identifying the lesions included in the panoramic image of each tooth by using the data output from the second artificial neural network; and inputting an image of the tooth area with the identified lesions from the segmented panoramic image into the third artificial neural network, and outputting data about the type and location of the lesion and a lesion detection confidence score by using the data output from the third artificial neural network.

[0029] In one embodiment, the step of performing segmentation processing on each tooth region included in the panoramic image may include the following steps: cropping the panoramic image to display only the area of ​​interest; masking the tooth region, crown region, pulp region and bone level region in the cropped panoramic image; and inputting the masked panoramic image into the first artificial neural network.

[0030] In one embodiment, the step of identifying lesions included in the panoramic image of each tooth may include the following steps: dividing each tooth region included in the segmented panoramic image into first to third sub-regions by sequentially dividing the tooth region from the crown toward the root in a direction perpendicular to the tooth axis; inputting data of the first sub-region into a second artificial neural network of a first type, and identifying the presence of caries lesions in the panoramic image by using data output from the second artificial neural network of the first type; inputting data of the second sub-region into a second artificial neural network of a second type, and identifying the presence of periodontal disease lesions in the panoramic image by using data output from the second artificial neural network of the second type; and inputting data of the third sub-region into a second artificial neural network of a third type, and identifying the presence of apical periodontitis lesions in the panoramic image by using data output from the second artificial neural network of the third type.

[0031] In one embodiment, the step of identifying the presence of periodontal lesions in the panoramic image may include: identifying the presence of periodontal lesions in the panoramic image by using data about the degree of depression of the bone level area included in the second sub-area relative to the crown area, the data being output from a second artificial neural network of the second type.

[0032] In one embodiment, the step of identifying the presence of an apical periodontitis lesion in a panoramic image may include: identifying the presence of an apical periodontitis lesion in the panoramic image by using data indicating an inflammatory area in a third sub-area, the inflammatory area being adjacent to a tooth and lacking Hounsfield unit (HU) values ​​corresponding to bones and teeth, the data being output from a second artificial neural network of a third type.

[0033] In one embodiment, the step of identifying the presence of caries lesions in the panoramic image may include: identifying the presence of caries lesions in the panoramic image by using data indicating an inflammatory area having an HU value different from the HU value of the crown area included in the first sub-area, the data being output from a second artificial neural network of the first type.

[0034] According to another embodiment of the present disclosure, a dental lesion information visualization system may include: one or more processors; and a memory configured to load one or more instructions. Here, the one or more processors may perform operations by running the stored one or more instructions, the operations including: inputting data about a panoramic image of a patient's oral cavity into a machine-learned lesion analysis model; determining an area of ​​interest in the panoramic image by using data output by the lesion analysis model; superimposing and displaying an indicator pointing to the area of ​​interest on the panoramic image; generating a lesion detection list using data output by the lesion analysis model; and displaying the lesion detection list together with the panoramic image. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 An exemplary environment in which a dental lesion information visualization system according to an embodiment of the present disclosure may be applied is shown.

[0036] Figure 2 It shows reference Figure 1 Block diagram of the dental lesion information visualization system described.

[0037] Figure 3 is a flowchart illustrating a dental lesion information visualization method according to another embodiment of the present disclosure.

[0038] Figure 4 is used to describe in detail such as reference Figure 3 A flowchart describing some of the operations.

[0039] Figure 5 By way of example, it is shown that Figure 3 The described method detects input of an indicator and displays the results of an oral computed tomography (CT) image.

[0040] Figure 6 Sub-regions corresponding to the types of lesions of teeth that require verification are shown by way of example to facilitate understanding of some embodiments of the present disclosure.

[0041] Figure 7 Sub-regions of a tooth are shown that require verification in order to understand some embodiments of the present disclosure.

[0042] Figure 8 An example screen is shown in which a region of interest is superimposed and displayed on a panoramic image according to some embodiments of the present disclosure.

[0043] Fig. 9 is a flow chart illustrating a method for visualizing dental lesion information according to yet another embodiment of the present disclosure.

[0044] Fig.10 By way of example, the execution of Fig. 9 The result of the step of generating a lesion detection list is shown.

[0045] Fig.11 By way of example, the execution of Fig. 9 The results of the steps of detecting the input of lesion records and displaying an oral CT image are shown.

[0046] Fig.12 The result of performing the step of displaying data output by the lesion analysis model on the user terminal in conjunction with the patient management system as may be performed in some embodiments of the present disclosure is shown.

[0047] Fig.13 is a flow chart illustrating a method for visualizing dental lesion information according to yet another embodiment of the present disclosure.

[0048] Fig.14 and Fig.15 The results of pre-processing a panoramic image as may be performed in some embodiments of the present disclosure are shown by way of example.

[0049] Figures 16 to 18 The results of segmentation pre-processing of each anatomical structure included in a panoramic image, as may be performed in some embodiments of the present disclosure, are shown by way of example.

[0050] Figures 19 to 23 is a diagram illustrating, by way of example, a method for identifying lesions included in a panoramic image as may be performed in some embodiments of the present disclosure.

[0051] Fig.24 The hardware configuration of the dental lesion information visualization system according to an embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0052] Hereinafter, example embodiments of the present disclosure will be described in detail with reference to the accompanying drawings. The advantages and features of the present disclosure and the manner in which these advantages and features are achieved will become apparent with reference to the embodiments described in detail below in conjunction with the accompanying drawings. However, the present disclosure may be implemented in many different forms and should not be construed as being limited to the embodiments set forth herein, and embodiments are provided so that the present disclosure will be thorough and complete and will fully convey the scope of the present disclosure to those skilled in the art, and the present disclosure is limited only by the scope of the appended claims.

[0053] In the following description of the present disclosure, in order to avoid making the subject matter of the present invention unclear, a detailed description of known functions and configurations incorporated herein will be omitted.

[0054] Hereinafter, some embodiments of the present disclosure will be described with reference to the accompanying drawings.

[0055] Figure 1 FIG. 2 shows an environment in which a dental lesion information visualization system according to an embodiment of the present disclosure can be applied. Figure 1 As shown, the dental lesion information visualization system 100 according to this embodiment can implement the dental lesion information visualization method by interacting with the patient management system 200 and the user terminal 300. Figure 1 The dental lesion information visualization system 100 and the patient management system 200 are shown as separate systems, but in some embodiments, the dental lesion information visualization system 100 and the patient management system 200 may also be implemented as software modules within a single system.

[0056] In the following, it will be described in more detail Figure 1 and Figure 2 Each component shown in .

[0057] The patient management system 200 is a system including patient information and patient images. The patient management system 200 can output an electronic chart of the patient according to the input from the user terminal 300. Here, the electronic chart may include the patient's personal information, medical history, diagnosis information, and a panoramic image of the patient's affected area. Such a document may be a medical imaging file (Digital Imaging and Communications in Medicine: DICOM) containing the patient's personal information, past test information, and details of the image itself, a network document in a format such as Hypertext Markup Language (HTML) or Extensible Markup Language (XML), but is not limited thereto.

[0058] Furthermore, according to one embodiment of the present disclosure, the patient management system 200 may receive data related to the identified lesion type, the identified detailed location of the lesion, and the lesion detection confidence score from the dental lesion information visualization system 100 , and display it on the user terminal 300 .

[0059] The user terminal 300 may output the received electronic chart on a screen. The user terminal 300 may be equipped with a specific display device for presenting the electronic chart.

[0060] In addition, the user terminal 300 can also display on the display device a panoramic image having a region of interest set for the dental lesion received from the dental lesion information visualization system 100 and a CT image corresponding to the region of interest. In addition, it can receive input from the user instructing it to display the panoramic image or the CT image.

[0061] The target image storage unit 130 of the dental lesion information visualization system 100 can receive data about a panoramic image of the patient's oral cavity and can input the received panoramic image into the lesion analysis model. Here, the panoramic image can also be received from the patient management system 200 in response to a request from the user terminal 300.

[0062] According to some embodiments of the present disclosure, the dental lesion analysis unit 140 may analyze the lesion using the patient's panoramic image stored in the target image storage unit 130. The dental lesion analysis unit 140 may send the result of analyzing the panoramic image to the lesion analysis model training unit 110. The lesion analysis model training unit 110 may update the lesion analysis model by further training the lesion analysis model using the analysis result sent from the dental lesion analysis unit 140 and user feedback.

[0063] The dental lesion information visualization system 100 can use the data output by the lesion analysis model to specify a region of interest in the panoramic image. The region of interest may refer to a region in the panoramic image that may potentially contain a lesion.

[0064] The dental lesion information visualization unit 120 in the dental lesion information visualization system 100 may superimpose an indicator pointing to the region of interest on the panoramic image and display it to the user via the display device on the user terminal 300. In addition, in response to the user's selection input of the indicator, the dental lesion information visualization system 100 may display a CT image of the patient's oral cavity corresponding to the panoramic image on the user terminal 300. Therefore, the dental lesion information visualization system 100 may intuitively indicate to the user that the indicator is a selectable graphic object, for example, by automatically changing the cursor to a selectable shape in response to a mouse hover event on the indicator.

[0065] The operations of determining the region of interest, displaying an indicator to the user, and displaying the CT image may be performed by Figure 2 The dental lesion information visualization unit 120 shown in FIG. More detailed implementations of these operations will be further described below.

[0066] The dental lesion analysis unit 140 in the dental lesion information visualization system 100 can generate a lesion detection list for lesions included in the input panoramic image. According to some embodiments of the present disclosure, the lesion detection list can be displayed to the user via a display device on the user terminal 300, and can also be stored in the patient management system 200.

[0067] In addition, the dental lesion information visualization unit 120 of the dental lesion information visualization system 100 may receive a user input regarding a specific lesion record included in the lesion detection list from the user terminal 300. In response to the input, the dental lesion information visualization system 100 may display an oral CT image of a patient associated with the lesion detection list on the user terminal 300.

[0068] The dental lesion information visualization system 100 can pre-process the panoramic image, input information about the panoramic image into the lesion analysis model to identify the lesions contained in the panoramic image, and output the type and location of the lesion and the lesion detection confidence score. Here, the lesion analysis model includes the first artificial neural network to the third artificial neural network, and the dental lesion information visualization system 100 can perform the above operations based on the data output from the plurality of neural networks.

[0069] The operations of preprocessing the panoramic image, identifying the lesions contained in the panoramic image, and outputting the type, location, and detection confidence score of the lesions can be performed by Figure 2 The dental lesion analysis unit 140 shown in FIG.

[0070] So far, reference has been made to Figure 1 and Figure 2 The configuration and operation of the dental lesion information visualization system 100 and the environment containing the dental lesion information visualization system 100 are described. The dental lesion information visualization system 100 can be understood to operate with the user terminal 300 according to the client-server model. However, in some embodiments, the system can be configured in a client-independent manner without a server. In this case, the operations performed by the dental lesion information visualization system 100 described above will be understood to be performed on the user terminal 300.

[0071] According to the present embodiment, lesions included in overlapping areas between anatomical structures in a panoramic image of a patient, which are difficult for a dentist to visually assess, can be automatically identified by an AI model.

[0072] In the following, reference is made to Figures 3 to 8 , a more detailed description of a dental lesion information visualization method according to another embodiment of the present disclosure will be provided. Unless otherwise specified, the steps described in the following flowcharts may be understood to be performed by the dental lesion information visualization system 100 .

[0073] Figure 3 is a flowchart showing a dental lesion information visualization method according to the present embodiment.

[0074] In step S100, the dental lesion information visualization system 100 may input data of a panoramic image of a patient's oral cavity into a lesion analysis model. The panoramic image may be an image in which each tooth included in the image is assigned a tooth identification number.

[0075] Next, in step S200, the dental lesion information visualization system 100 may use the data output by the lesion analysis model to specify a region of interest in the panoramic image. Figures 4 to 8 A more detailed description of the steps of determining a region of interest is provided.

[0076] Reference Figure 4 In step S210, the dental lesion information visualization system 100 may determine the teeth that need to be verified. Here, the teeth that need to be verified may refer to teeth whose confidence score of the lesion analysis model for identifying the lesion in each tooth in the panoramic image is less than or equal to the reference value. In addition, the reference value may be a value preset by the user or an initial setting value of the system.

[0077] In step S220, the dental lesion information visualization system 100 may identify the area of ​​the tooth that needs to be verified. Figure 6 In FIG. 6 , an exemplary region 611 of a tooth to be verified is shown in a box form. In addition, the region 611 of a tooth to be verified can be sequentially divided into three sections from the crown toward the root in a direction perpendicular to the tooth axis. Figure 6 The three-segment area 611 of the tooth to be verified includes the uppermost area in the crown direction as the first sub-area 613, the area between the first sub-area 613 and the third sub-area 612 as the second sub-area 614, and the lowermost area in the root direction as the third sub-area 612.

[0078] In some embodiments related to step S220, the tooth region 611 to be verified may be sequentially divided into three sections along a direction perpendicular to the tooth axis from the root toward the crown. Figure 6 The tooth area 611 that needs to be verified, which is divided into three sections, can include the lowermost area in the root direction as the first sub-area 612, the area between the first sub-area 612 and the third sub-area 613 as the second sub-area 614, and the uppermost area in the crown direction as the third sub-area 613.

[0079] In addition, each of the first to third sub-regions may include further subdivided regions. For example, the first sub-region may be divided in a direction parallel to the tooth axis direction, and the first sub-region includes sub-regions from the 1-1 sub-region to the 1-N sub-region. It should be noted that the variable N is a natural number greater than or equal to 2. For example, N may be 3. In other words, Figure 7 As shown, the area of ​​the tooth to be verified can be divided into a total of 9 (3x3) sub-areas. In addition, these subdivided areas of the tooth to be verified can be sequentially referred to as the first sub-area, the second sub-area, and the third sub-area along the direction perpendicular to the tooth axis from the crown toward the root. The three areas further divided from the first sub-area can be designated as the 1-1st sub-area, the 1-2nd sub-area, and the 1-3rd sub-area from the mesial to the distal direction, respectively.

[0080] In step S230, the dental lesion information visualization system 100 may designate a portion of the tooth region to be verified as a region of interest. Figure 6 , the third sub-region 612 is shown as the determined region of interest. The portion of the tooth region to be verified that is designated as the region of interest may be a region corresponding to a lesion type identified in the lesion analysis of the tooth to be verified.

[0081] In some embodiments related to step S230, if the data output by the lesion analysis model indicates that the tooth to be verified may have a caries-type lesion, the dental lesion visualization system 100 may designate the first sub-region as the region of interest. In some embodiments related to step S230, the dental lesion information visualization system 100 may designate the second sub-region as the region of interest in response to data output by the lesion analysis model indicating that the tooth to be verified contains a periodontal disease-type lesion.

[0082] In other embodiments related to step S230 , the dental lesion information visualization system 100 may designate the third sub-region as the region of interest in response to data output by the lesion analysis model indicating that the tooth to be verified contains a lesion of the periapical lesion type.

[0083] In other embodiments related to step S230, the dental lesion information visualization system 100 may determine the detailed location of the lesion within the designated region of interest based on the data output by the lesion analysis model. Figure 8 .

[0084] Reference Figure 8 When the tooth region 811 to be verified is designated as the first sub-region, the dental lesion information visualization system 100 can designate the 1st to 3rd sub-regions 812 located at the far end of the first sub-region as the region of interest containing caries type lesions based on the data output by the lesion analysis model. In addition, the 1st to 3rd sub-regions 812 can be displayed as an overlay on the patient's panoramic image.

[0085] According to this embodiment, the dental lesion detection system can provide the user with the exact location of the detected dental lesion without user intervention. The lesion location identified using this method can be widely used in various fields of dental practice such as dental treatment and dental restoration manufacturing.

[0086] Will refer to Figure 3 Continue with the description.

[0087] Reference Figure 3 and Figure 5 In steps S300 and S400, in response to the presence of the CT image when the user selects the region of interest on the panoramic view, the dental lesion information visualization system 100 may generate a panoramic view consisting of a dental arch of the CT image, and display the panoramic image including an indicator 512 of the region of interest on the user terminal 300. According to some embodiments of the present disclosure, the lesion type and detection confidence score 513 of the lesion in the region of interest may also be displayed.

[0088] Next, in step S500 , a user input related to the indicator on the user terminal 300 may be detected.

[0089] In step S600, the dental lesion information visualization system 100 may display a CT cross-sectional image 511 of the patient's oral cavity in response to detecting a user input to an indicator. For example, when a user clicks an area of ​​interest indicator 512 displayed on a panoramic area on a user terminal, a corresponding CT cross-sectional image 511 of the area of ​​interest is displayed, such as Figure 5 Here, a CT cross-sectional image aligned with the angle of the tooth axis of the segmented tooth in the region of interest may be provided.

[0090] In some embodiments related to step S600, the displayed CT image may be shown with a field of view (FoV) initially set based on the location of the region of interest. Figure 5 If the region of interest indicator 512 shown in FIG. 5 is shown, the dental lesion information visualization system 100 may display the CT cross-sectional image 511 at an angle that provides the clearest view of the lesion in the region of interest.

[0091] In some other embodiments related to step S600, in response to determining that there is no corresponding CT image for the region of interest indicator 612, the dental lesion information visualization system 100 may activate a CT imaging module for acquiring a CT image. The CT image acquired by the CT imaging module may be in the same Figure 5 Displayed in the same manner as shown in .

[0092] According to the present embodiment, the user can view both the CT image and the panoramic image on a single display device without the need for a separate terminal or system to compare the lesion location between the two images. In addition, if the overlapping anatomical structures in the patient's panoramic image prevent the lesion analysis model from accurately identifying the dental lesion, the system can immediately switch to a user-based diagnosis method.

[0093] So far, a detailed description of a dental lesion information visualization method according to one embodiment of the present disclosure has been provided, wherein an indicator of a region of interest containing a lesion in a panoramic image is displayed on a user terminal together with the panoramic image. In addition, it should be noted that each of the steps described so far can be used independently or in combination in various ways if necessary.

[0094] In the following, reference is made to Figures 9 to 12 , a more detailed description of a dental lesion information visualization method according to another embodiment of the present disclosure will be provided.

[0095] Fig. 9 is a flowchart illustrating a dental lesion information visualization method according to another embodiment of the present disclosure.

[0096] In step S910 , the dental lesion information visualization system 100 may generate a lesion detection list based on the data on the panoramic image output by the lesion analysis model in S900 .

[0097] The lesion detection list may include the tooth number of the tooth where the lesion is detected, the name or type of the detected lesion, the confidence score of the lesion detection result, and whether a region of interest is set.

[0098] In the following, reference is made to Fig.10 , the lesion detection list will be described. In the lesion detection list 1012, for example, the first detection result 1011 may indicate that the area of ​​tooth number 27 in the panoramic image is detected as periodontal disease with a probability of 99.5%, and no region of interest is set. Here, the step of setting the region of interest can be clearly understood by referring to the detailed description of step S230. When the region of interest is set, an indicator corresponding to the region of interest may appear on the panoramic image, as shown in reference numeral 1013.

[0099] In step S920, refer to Fig.12 The dental lesion information visualization system 100 can be combined with the patient management system to display the lesion detection list and the patient's panoramic image in an electronic chart format on the user terminal. Fig.12The lesion detection list shown in an electronic chart format is merely an example provided to aid in understanding the present disclosure, and it should be noted that any method of displaying data output by the lesion detection model may be used without limitation.

[0100] In step S930, the dental lesion information visualization system 100 may evaluate whether a user input to the lesion record displayed on the user terminal can be detected. For example, the user input may involve clicking on an area containing lesion name text displayed on the display device of the user terminal, such as Fig.12 as shown in .

[0101] In step S940 , the dental lesion information visualization system 100 may display a CT image of the patient's oral cavity on the user terminal in response to detecting a user input to a lesion record displayed on the user terminal.

[0102] Reference Fig.12 , the user terminal can display the lesion type contained in the lesion record clicked by the user, the tooth number containing the lesion, the lesion detection confidence score, whether the region of interest is set, the position of the detected region of interest in the panoramic image, and the CT image corresponding to the position of the region of interest.

[0103] According to this embodiment, Fig.11 As shown, the dental lesion information visualization system 100 allows the user to view both the CT image and the panoramic image at the same time by simply clicking on the displayed detected lesion, which can help reduce patient treatment time. In addition, the tooth axis information obtained by segmenting the CT image can be used to construct the dental arch and generate it as a panoramic view. The lesion location can then be displayed on the panoramic view corresponding to the tooth number.

[0104] So far, a detailed description of a dental lesion information visualization method according to one embodiment of the present disclosure has been provided, in which a dental lesion information visualization system 100 displays a CT image on a user terminal in response to a user input of a lesion record displayed on a user terminal. In addition, it should be noted that each of the steps described so far can be used independently or in combination in various ways if necessary.

[0105] In the following, reference is made to Figures 13 to 23 , a detailed description of a dental lesion information visualization method according to another embodiment of the present disclosure will be provided.

[0106] Fig.13 is a flow chart illustrating a method for visualizing dental lesion information according to yet another embodiment of the present disclosure.

[0107] In step S1300, the dental lesion information visualization system 100 may obtain data of a lesion analysis model, which is trained to output data about the type and location of lesions contained in a panoramic image of a patient's oral cavity. The lesion analysis model may include first to third artificial neural networks. The data of the lesion analysis model may define each of the first to third artificial neural networks, including parameters or hyperparameters of each artificial neural network.

[0108] Each of the first to third artificial neural networks may be configured using a known neural network architecture such as U-NET, a convolutional neural network (CNN), a recurrent neural network (RNN), a deep belief network, or a restricted Boltzmann machine.

[0109] The first to third artificial neural networks of the lesion analysis model can be trained using data about panoramic images for which lesions have been identified or data about panoramic images for neural network training separately input by a user and training data obtained by data enhancement of these panoramic images. Here, the enhanced training data can be generated by inverting the color of the original panoramic image, rotating its angle, flipping the image direction, or adjusting the window level or window width. In this case, the window level refers to the center value of the grayscale, and the window width can represent the range of HU values ​​that can be represented in grayscale.

[0110] The above-mentioned enhancement methods for the original training data (i.e., panoramic images) reflect various situations in which dental lesions are further distinguished from normal areas. For example, in the case of medical images in DICOM format, adjusting the window level and window width may cause specific lesions to appear or disappear. In addition, applying an inversion effect can enhance the black-and-white contrast, making lesion detection easier. By performing comprehensive training on these different modifications of the same panoramic image, the dental lesion information visualization model will achieve a high level of accuracy, sensitivity, and precision.

[0111] According to some embodiments of the present disclosure, the training data may include caries training data 1911, training data 1912 with masked regions of interest for caries, periodontal disease training data 2111, training data 2112 with masked regions of interest for periodontal disease, apical periodontitis training data 2311, and training data 2312 with masked regions of interest for apical periodontitis, such as Fig.19 , Fig.21 and Fig.23 shown.

[0112] According to this embodiment, in addition to the original training data, enhanced training data and masked training data are used to generate a lesion analysis model, thereby allowing the model to reflect the information contained in the training data from a wide range of perspectives. In the case of medical images (e.g., panoramic images), where data is highly scarce, training a lesion analysis model by modifying a limited amount of training data differently as described above can achieve the construction of a highly accurate lesion analysis model at a lower cost.

[0113] Next, in step S1310, the dental lesion information visualization system 100 may input the panoramic image of the patient as the lesion identification object into the first artificial neural network to obtain a panoramic image in which each tooth region is segmented. Figures 14 to 18 .

[0114] In some embodiments related to step S1310, refer to Fig.14 and Fig.15 , the dental lesion information visualization system 100 may identify a region including a tooth region within the panoramic image 1411 as a region of interest 1412. In addition, a panoramic image 1511 that is cropped to display only the region of interest may be obtained.

[0115] In some other implementations related to step S1310, refer to Fig.16 As shown by reference numeral 1611 in the figure, the dental lesion information visualization system 100 can mask the area corresponding to the teeth in the cropped panoramic image. In addition, as shown by reference numeral 1612, the dental lesion information visualization system 100 can also mask the area corresponding to the crown in the cropped panoramic image.

[0116] Here, the action of masking a specific area may refer to adding different marks to ensure that the specified area can be distinguished from other areas. Figure 16 to Figure 17 In the present invention, marking is expressed by assigning a different color to a specific area compared to other areas, but it should be noted that any method of clearly distinguishing a specific area from other areas may be used without limitation.

[0117] Additionally, masking can be performed by running an image processing algorithm that selects pixels that meet certain criteria or by feeding the image into a machine learning neural network.

[0118] In some other implementations related to step S1310, refer to Fig.17Referring to the reference numeral 1711, the dental lesion information visualization system 100 may mask the region corresponding to the dental pulp in the cropped panoramic image. In addition, referring to the bone level masking result 1712, the dental lesion information visualization system 100 may also mask the bone level region in the cropped panoramic image.

[0119] In some other implementations related to step S1310, refer to Fig.18 The dental lesion information visualization system 100 can obtain a segmented panoramic image 1811, in which the masked areas are integrated with different markers to distinguish them from each other. The segmented panoramic image 1811 can also be displayed on a display device of a user terminal according to user input.

[0120] According to this embodiment, the dental lesion information visualization system 100 can receive a panoramic image of a patient and acquire a panoramic image in which regions corresponding to teeth, crowns, pulp, and bone levels are divided, and each dental region is assigned a tooth identification number after segmentation processing. Of course, the panoramic image with segmented regions can be applied to various purposes in the technical field.

[0121] Next, in step S1320, the dental lesion information visualization system 100 may input data from the segmented panoramic image into a second artificial neural network to identify lesions contained in each tooth region in the panoramic image.

[0122] In some embodiments related to step S1320, the dental lesion information visualization system 100 may divide each tooth region into first to third sub-regions. More specifically, the dental lesion information visualization system 100 may divide the tooth region into first to third sub-regions by sequentially dividing the tooth region from the crown toward the root in a direction perpendicular to the tooth axis. The method for dividing the tooth region may be clearly understood by referring to the embodiments related to S220.

[0123] In some other embodiments related to step S1320, the dental lesion information visualization system 100 may input the data of the first sub-region into the second artificial neural network to identify the presence of caries lesions in each tooth region within the panoramic image. Hereinafter, the artificial neural network that receives the data of the first sub-region and outputs data about the presence of caries lesions will be referred to as a "first type of second artificial neural network". The data about the presence of caries lesions output from the first type of second artificial neural network may include information related to the severity of the caries lesions.

[0124] For example, refer to Fig. 20, based on the progression of the caries lesions, the data related to the severity of the caries lesions may be classified into at least one stage from C0 to C4. Fig. 20 In the present invention, C0 may represent the initial cavity form, C1 may be a cavity extending to the enamel layer, C2 may be a cavity reaching the dentin, C3 may represent a cavity progressing to the nerve, and C4 may represent a severe cavity forming an abscess at the root of the tooth.

[0125] In some other embodiments related to step S1320, if the data output from the second artificial neural network of the first type indicates that the caries lesion cannot be identified, it may be due to overlap with adjacent tooth areas that the caries lesion cannot be identified. In order to convey data related to this situation, the dental lesion information visualization system 100 can superimpose and display an indicator pointing to a sub-area where identification of the caries lesion is considered impossible on the panoramic image.

[0126] In some other embodiments related to step S1320, the dental lesion information visualization system 100 may input the information of the second sub-region into the second artificial neural network to identify the presence of periodontal lesions in each tooth region within the panoramic image. Hereinafter, the artificial neural network that receives the data of the second sub-region and outputs the data on the presence of periodontal lesions will be referred to as a "second type of second artificial neural network".

[0127] Here, the second artificial neural network of the second type may output data on the degree of depression of the bone level region relative to the crown region in the segmented panoramic image. At this time, the dental lesion information visualization system 100 may calculate the degree of depression using the data output from the second artificial neural network of the second type, and if the calculated degree of depression is greater than or equal to the reference value relative to the segmented crown, it may be identified that the tooth region containing the second sub-region includes a periodontal lesion.

[0128] For example, refer to Fig. 22 If the depression distance 2211 of the crown region in the left maxillary second molar region relative to the bone horizontal region in the panoramic image is greater than or equal to the reference value, the dental lesion information visualization system 100 can identify that the left maxillary second molar region in the panoramic image includes a periodontal lesion. In addition, the second artificial neural network can output data related to the severity of the periodontal lesion based on the depression distance 2211.

[0129] In some other embodiments related to step S1320, the dental lesion information visualization system 100 may input information of the third sub-region into the second artificial neural network to identify the presence of apical periodontitis lesions in each tooth region within the panoramic image. Hereinafter, the artificial neural network that receives data of the third sub-region and outputs data on the presence of apical periodontitis lesions will be referred to as a "third type of second artificial neural network".

[0130] A second artificial neural network of the third type may receive data regarding a third sub-region and output data indicating an inflamed region that is adjacent to the tooth and lacks Hounsfield Unit (HU) values ​​corresponding to the bone and the tooth.

[0131] By using the data output from the second artificial neural network of the third type, the dental lesion information visualization system 100 can identify the inflammation area within the third sub-region. By using the shape of the inflammation area, the dental lesion information visualization system 100 can determine whether the third sub-region includes the apical periodontitis lesion. For example, the dental lesion recognition system can measure the roundness of the inflammation area, and if the roundness exceeds a reference value, it is determined that the third sub-region contains the apical periodontitis lesion.

[0132] The panoramic image of the patient represents the patient's arched oral structure on a plane, which may introduce distortion in the image. Therefore, the panoramic image may include some overlapping anatomical structures. Therefore, there may be errors in the dentist's visual diagnosis. According to this embodiment, the lesion analysis model pre-processes the patient's panoramic image in a manner suitable for identifying dental lesions, and determines whether the lesion is identified based on a standard standard with a quantitative value, thereby achieving accurate detection of dental lesions.

[0133] In step S1330, the dental lesion recognition system inputs the panoramic image of the tooth area containing the identified lesion into the third artificial neural network to obtain data on the type, location and detection confidence score of the lesion. Preferably, the dental lesion information visualization system 100 can use the obtained information to generate a lesion detection list and display the lesion detection list on the display device of the user terminal. The method of displaying the lesion detection list will be clearly understood by referring to the description of some embodiments of step S920 above.

[0134] In addition, the dental lesion information visualization system 100 can use the obtained data about the lesion location to mark a specific location containing the lesion in the patient's panoramic image as an area of ​​interest, and can display it on the display device of the user terminal. Figure 5 The method of displaying data about the lesion location will be clearly understood from some embodiments of step S400 described.

[0135] As described above, the first to third artificial neural networks sequentially perform tooth segmentation and the process of analyzing the presence of caries lesions, periodontal disease lesions, and periapical periodontitis lesions in each tooth region, thereby providing high accuracy in detecting dental lesions. In other words, in order to achieve the ultimate goal of visualizing dental lesion information, the described embodiment provides an optimal network architecture by sequentially linking the first neural network, the second neural network, and the third neural network.

[0136] Furthermore, through the sequential linking of the first neural network, the second neural network, and the third neural network, the lesion type, lesion location, and lesion detection confidence score of the tooth region containing periodontal lesions can all be output.

[0137] Furthermore, the computing system enhances the visibility of information by displaying the panoramic image together with the tooth region containing the periodontal lesion and the lesion type information, so that the dentist can easily identify data related to the visualization of the dental lesion.

[0138] In addition, the computing system provides data about the lesion location where the lesion detection confidence score is less than or equal to the reference value and therefore requires CT image inspection. When the user selection input occurs, such as clicking on the lesion location, the computing system can display a CT image with a FoV setting that provides an optimal view of the lesion, or activate the CT imaging module, thereby guiding the user to view or capture a CT image of an area of ​​interest where lesion identification is unclear. Therefore, the possibility that the dentist will misidentify the lesion based only on the panoramic image can be reduced.

[0139] So far, a detailed description of the dental lesion information visualization method according to an embodiment of the present disclosure has been provided. In addition, it should be noted that each of the steps described so far can be used independently or in combination in various ways if necessary.

[0140] Fig.24 The hardware configuration of the dental lesion information visualization system 100 according to some embodiments of the present disclosure is shown. Fig.24 The dental lesion information visualization system 100 shown in FIG. 1 may be referred to as Figure 1 The dental lesion information visualization system 100 is described. The dental lesion information visualization system 100 may include: at least one processor 1100; a system bus 1600; a communication interface 1200; a memory 1400 configured to load a computer program 1500 executed by the processor 1100; and a storage device 1300 configured to store the computer program 1500.

[0141] The processor 1100 controls the overall operation of each configuration of the dental lesion information visualization system 100. The processor 1100 may perform calculations for at least one application or program for performing methods and / or operations according to various embodiments of the present disclosure. The memory 1400 stores various types of data, commands, and / or information. The memory 1400 may load one or more programs 1500 from the storage device 1300 to perform methods and / or operations according to various embodiments of the present disclosure. The bus 1600 provides communication functions between components of the dental lesion information visualization system 100. The communication interface 1200 supports Internet communication of the dental lesion information visualization system 100. The storage device 1300 may store one or more computer programs 1500 in a non-transient manner. The computer program 1500 may include one or more instructions for implementing methods and / or operations according to various embodiments of the present disclosure. When the computer program 1500 is loaded onto the memory 1400, the processor 1100 may run one or more instructions to perform methods and / or operations according to various embodiments of the present disclosure.

[0142] In some implementations, the dental lesion detection system may be configured based on cloud technology (eg, virtual machines) using one or more physical servers included in a server farm.

[0143] The computer program 1500 may include instructions for: inputting data about a panoramic image of a patient's oral cavity into a machine-learned lesion analysis model; determining a region of interest in the panoramic image using data output by the lesion analysis model; superimposing and displaying an indicator pointing to the region of interest on the panoramic image; generating a lesion detection list using data output by the lesion analysis model; displaying the lesion detection list together with the panoramic image; searching for a CT image of the patient's oral cavity in response to a user's selection input of an indicator; and running a CT imaging module in response to determining that no CT image of the patient's oral cavity exists based on the search results.

[0144] So far, reference has been made to Figures 1 to 24 Various embodiments and their effects are described in which the technical features of the present disclosure are implemented. The effects according to the technical features of the present disclosure are not limited to the above effects, and those skilled in the art will clearly understand other effects not described from the following description.

[0145] The technical features of the present disclosure described above can be implemented as computer-readable codes on a computer-readable medium. The computer program recorded on the computer-readable recording medium can be sent to other computing devices through a network such as the Internet and installed in other computing devices, thereby being used in other computing devices.

[0146] Although operations are shown in a specific order in the accompanying drawings, it should not be understood that the operations must be performed in the specific order shown or in a continuous order, or that all the operations shown must be performed to achieve the desired result. In some cases, multitasking and parallel processing may be advantageous. In summarizing the detailed description, those skilled in the art will appreciate that many changes and modifications may be made to the exemplary embodiments without substantially departing from the principles of the present disclosure. Therefore, the disclosed exemplary embodiments of the present invention are intended only for general and descriptive purposes, and not for limiting purposes. The scope of the present disclosure should be interpreted by the appended claims, and any technical features within the scope equivalent to the claims should be interpreted as falling within the scope of the technical features defined by the present disclosure.

Claims

1. A method for visualizing dental lesion information performed by a computing system, comprising the following steps: inputting data regarding the panoramic image of the patient's oral cavity into a machine learning lesion analysis model; determining a region of interest containing a lesion in the panoramic image by using data output by the lesion analysis model; In response to a user's selection input of an indicator, searching a computed tomography (CT) image of the patient's oral cavity; In response to determining that there is no CT image of the oral cavity of the patient based on the search results, executing a CT imaging module; as well as An indicator pointing to the region of interest is superimposed and displayed on the panoramic image.

2. The dental lesion information visualization method according to claim 1, wherein: The step of determining the region of interest comprises the following steps: Determining teeth requiring verification whose lesion detection confidence scores are less than or equal to a reference value by using data output by the lesion analysis model; identifying the area of ​​the tooth that needs to be verified in the panoramic image; and A portion of the area of ​​the tooth that needs to be verified is designated as the region of interest.

3. The dental lesion information visualization method according to claim 2, wherein: The region of the tooth to be verified includes a first sub-region, a second sub-region, and a third sub-region obtained by sequentially dividing the region from the crown toward the root in a direction perpendicular to the tooth axis, and The step of designating a portion of the area of ​​the tooth that needs to be verified as the area of ​​interest includes: the step of designating a sub-area corresponding to the lesion type of the tooth that needs to be verified among the first sub-area to the third sub-area as the area of ​​interest, wherein the lesion type of the tooth that needs to be verified is determined using data output by the lesion analysis model.

4. The dental lesion information visualization method according to claim 3, wherein: The step of designating the sub-region corresponding to the lesion type of the tooth that needs to be verified as the region of interest includes the following steps: when the lesion type of the tooth that needs to be verified is a caries type, designating the first sub-region as the region of interest; when the lesion type of the tooth that needs to be verified is a periodontal disease type, designating the second sub-region as the region of interest; and when the lesion type of the tooth that needs to be verified is a periapical lesion type, designating the third sub-region as the region of interest.

5. The dental lesion information visualization method according to claim 2, wherein: The region of the tooth to be verified includes a first sub-region, a second sub-region, and a third sub-region obtained by sequentially dividing the region from the crown toward the root in a direction perpendicular to the tooth axis, The first sub-region includes a 1-1th sub-region to a 1-Nth sub-region obtained by dividing the first sub-region into N sub-regions in a direction parallel to the direction of the tooth axis (where N is a natural number greater than or equal to 2), The second sub-region includes 2-1th sub-region to 2-Nth sub-region obtained by dividing the second sub-region into N sub-regions in a direction parallel to the direction of the tooth axis (where N is a natural number greater than or equal to 2), The third sub-region includes a 3-1th sub-region to a 3-Nth sub-region obtained by dividing the third sub-region into N sub-regions in a direction parallel to the direction of the tooth axis (where N is a natural number greater than or equal to 2), and The step of designating a portion of the area of ​​the tooth that needs to be verified as the area of ​​interest includes: using the data output by the lesion analysis model to designate one of the 1-1 sub-area to the 1-N sub-area, the 2-1 sub-area to the 2-N sub-area and the 3-1 sub-area to the 3-N sub-area as the lesion location of the tooth that needs to be verified.

6. The dental lesion information visualization method according to claim 1, further comprising: The step of displaying a CT image of the patient's oral cavity in response to the user's selection input of an indicator.

7. The dental lesion information visualization method according to claim 6, wherein: The step of displaying the CT image of the patient's oral cavity includes displaying the CT image with a field of view (FoV) initially set based on the position of the region of interest.

8. A method for visualizing dental lesion information performed by a computing system, comprising the following steps: acquiring data of a lesion analysis model, the lesion analysis model outputting data on the type and location of a lesion included in a panoramic image of the patient's oral cavity, the lesion analysis model comprising a first artificial neural network, a second artificial neural network, and a third artificial neural network; inputting data of the panoramic image to the first artificial neural network, and performing segmentation processing on each tooth region included in the panoramic image by using the data output from the first artificial neural network; inputting data of the segmented panoramic image to the second artificial neural network, and identifying a lesion included in the panoramic image of each tooth by using the data output from the second artificial neural network; as well as An image of a tooth region having an identified lesion from the segmented panoramic image is input to the third artificial neural network, and data on the type and location of the lesion and a lesion detection confidence score are output by using the data output from the third artificial neural network.

9. The dental lesion information visualization method according to claim 8, wherein: The step of performing segmentation processing on each tooth region included in the panoramic image comprises the following steps: Cropping the panoramic image to display only a region of interest; and The tooth region, the crown region, the pulp region and the bone level region are masked in the cropped panoramic image, and the masked panoramic image is input to the first artificial neural network.

10. The dental lesion information visualization method according to claim 8, wherein: The step of identifying the lesions included in the panoramic image of each tooth comprises the following steps: Dividing each tooth region included in the segmented panoramic image into first to third sub-regions by sequentially dividing the tooth region from a crown toward a root in a direction perpendicular to a tooth axis; inputting data of the first sub-region into the second artificial neural network of the first type, and identifying the presence of a caries lesion in the panoramic image by using the data output from the second artificial neural network of the first type; inputting the data of the second sub-region into the second artificial neural network of the second type, and identifying the presence of a periodontal disease lesion in the panoramic image by using the data output from the second artificial neural network of the second type; and The data of the third sub-region is input to the second artificial neural network of the third type, and the presence of a periapical periodontitis lesion in the panoramic image is identified by using the data output from the second artificial neural network of the third type.

11. The dental lesion information visualization method according to claim 10, wherein: The step of identifying the presence of periodontal lesions in the panoramic image includes: identifying the presence of periodontal lesions in the panoramic image by using data about the degree of depression of the bone level area included in the second sub-area relative to the crown area, which data is output from the second artificial neural network of the second type.

12. The dental lesion information visualization method according to claim 10, wherein: The step of identifying the presence of apical periodontitis lesions in the panoramic image includes: identifying the presence of apical periodontitis lesions in the panoramic image by using data indicating an inflammatory area in the third sub-area, wherein the inflammatory area is adjacent to the teeth and lacks Hounsfield unit (HU) values ​​corresponding to bones and teeth, and the data is output from the second artificial neural network of the third type.

13. The dental lesion information visualization method according to claim 10, wherein: The step of identifying the presence of caries lesions in the panoramic image includes: identifying the presence of caries lesions in the panoramic image by using data indicating an inflammatory area having an HU value different from the HU value of the crown area included in the first sub-area, the data being output from the second artificial neural network of the first type.

14. A dental lesion information visualization system, comprising: one or more processors; as well as a memory configured to load one or more instructions, The one or more processors perform operations by running one or more stored instructions, and the operations include: inputting data regarding the panoramic image of the patient's oral cavity into a machine learning lesion analysis model; determining a region of interest in the panoramic image by using data output by the lesion analysis model; superimposing and displaying an indicator pointing to the region of interest on the panoramic image; generating a lesion detection list using data output by the lesion analysis model; and The lesion detection list is displayed together with the panoramic image.

Images