KR102657737B1 - Method and apparatus for analysis of oral in pets - Google Patents

Abstract

A method for analyzing companion animal oral information is disclosed. In one embodiment of the present disclosure, a method for analyzing oral information of a companion animal includes the steps of acquiring an oral image of a companion animal, inputting data from the acquired oral image into an oral information analysis model previously created through machine learning, Generating an oral information analysis result using data output from the oral information analysis model, wherein the oral information analysis model includes information on color uniformity of the first coronal region and color uniformity of the first gingival region. It is composed of an artificial neural network including a CNN (Convolutional Neural Network) that outputs information about the degree of recession of the first gum line, and a classifier that performs classification based on the information output from the CNN. , the first gingival area is an area adjacent to the first crown area, and is an area representing the gingiva corresponding to the first crown expressed by the first crown area, and the first gum line is the first gingival area and It may include a step that is the boundary line of the first crown region.

Description

Method and device for providing analysis of oral information of companion animals {METHOD AND APPARATUS FOR ANALYSIS OF ORAL IN PETS}

The present disclosure relates to a method for analyzing oral information of a companion animal and a device to which the method is applied. More specifically, it relates to a method of inferring the oral health and occurrence of periodontitis of a companion animal from a photograph of the pet's mouth and a system to which the method is applied.

Teeth are also one of the important body parts that determine the health of pets. In particular, dogs have a total of 42 teeth, and since they have many teeth, they have a wide range of teeth brushing. Unlike humans, they have a deep oral structure, so it is easy for dogs to develop tartar, making it difficult to maintain healthy teeth.

Currently, the dental health status of pets is measured through direct examination by experts using X-ray data.

Meanwhile, diagnostic assistance technologies based on artificial intelligence technology are provided. A representative example is machine learning-based artificial intelligence that predicts the likelihood of developing a specific disease or its severity by analyzing photos of problem areas. In situations where it is difficult to visit a veterinary hospital, artificial intelligence is provided that predicts the likelihood of developing a specific disease or its severity by analyzing images of the target pet taken by the pet's guardian using a smartphone, etc.

However, the basic and elemental technologies for photographing a pet's teeth using a terminal and inferring the dental health of the pet using artificial intelligence did not exist in Korea, and there was little discussion on the use of artificial intelligence. am.

Korean Patent Publication No. 10-2021-0107115 Korean Patent No. 10-2218752

Young-seon Han, Shin In-cheol, and Jang Won-du, “Tooth and tartar recognition using artificial neural networks” Master of Engineering thesis (2022.02)

The technical problem to be solved in some embodiments of the present disclosure is to provide an artificial intelligence-based companion animal oral information analysis method that accurately infers the dental health of a companion animal without an X-ray using an oral photograph of the companion animal.

The technical problems of the present disclosure are not limited to the technical problems mentioned above, and other technical problems not mentioned will be clearly understood by those skilled in the art from the description below.

In order to solve the above technical problem, a method of analyzing oral information of a companion animal according to an embodiment of the present disclosure includes the steps of acquiring an oral photographed image of a companion animal, and data of the acquired oral photographed image is previously generated using machine learning. inputting the data into an oral information analysis model and generating an oral information analysis result using data output from the oral information analysis model, wherein the oral information analysis model includes information about the color uniformity of the first crown region. A CNN (Convolutional Neural Network) that outputs information about the color uniformity of the first gingival area and information about the degree of recession of the first gum line, and a classifier that performs classification based on the information output from the CNN. ), wherein the first gingival area is an adjacent area to the first crown area, and is an area representing the gingiva corresponding to the first crown expressed by the first crown area, and the first gingival area is an area adjacent to the first crown area. The gum line may be a boundary line between the first gingival area and the first crown area.

In some embodiments, the classifier may perform classification using the result of feature fusion of information about the color uniformity of the first coronal region and information about the color uniformity of the first gingival region. there is.

In some embodiments, the result of feature fusion of information about the color uniformity of the first coronal region and information about the color uniformity of the first gingival region is a score for the color uniformity of the first coronal region and If the score for color uniformity of the first gingival area is the same and is higher than a predefined score, weight may be assigned.

In some embodiments, the classifier may perform classification using the results of scoring information about the degree of recession of the first gum line.

In some embodiments, the result of scoring information on the degree of recession of the first gum line is obtained by using a curvature model of the gum line generated by learning the degree of recession of the first gum line from a normal dog image. It may be derived.

In some embodiments, the result of scoring the information about the degree of recession of the first gum line is the first gingival area / (first It may be derived using a computational model (gingival area + first crown area).

In some embodiments, the classifier is based on the degree of recession of the first gum line calculated using the first gingival area/(first gingival area + first crown area) calculation model and the color uniformity of the first crown area. Classification may be performed using the result of feature fusion of information about and information about color uniformity of the first gingival area.

A companion animal oral information analysis system according to another embodiment of the present disclosure for solving the above technical problem includes a memory on which a companion animal oral information analysis program is loaded and a processor on which the companion animal oral information analysis program is executed. , the companion animal oral information analysis program inputs instructions for acquiring an oral image of a companion animal, data from the acquired oral image, into an oral information analysis model previously created through machine learning, and analyzes the oral information. Includes instructions for generating an oral information analysis result using data output from the model, wherein the oral information analysis model includes information on the color uniformity of the first coronal region, information on the color uniformity of the first gingival region, and , It is composed of an artificial neural network including a CNN (Convolutional Neural Network) that outputs information about the degree of recession of the first gum line, and a classifier that performs classification based on the information output from the CNN, and the first The gingival area is an area adjacent to the first crown area, and is an area representing the gingiva corresponding to the first crown expressed by the first crown area, and the first gum line is the first gingival area and the first crown. It may be the border of the area.

1 is a configuration diagram of a companion animal oral information analysis system according to an embodiment of the present disclosure.
Figure 2 is a flowchart of a method for analyzing oral information of a companion animal according to some embodiments of the present disclosure.
FIG. 3 is a diagram illustrating a CNN that derives information about color uniformity of a crown region used in some embodiments of the present disclosure.
Figure 4 is a diagram for explaining a CNN that derives information about the color uniformity of the gingival area used in some embodiments of the present disclosure.
Figure 5 is a diagram for explaining the process of feature fusion of the two pieces of information described using Figures 3 and 4.
Figure 6 is a diagram for explaining a CNN that scores information about the degree of recession of the gum line.
FIG. 7 is a diagram illustrating the artificial neural network architecture of a model used in some embodiments of the present disclosure.
8 is a hardware configuration diagram of a computing system described in some embodiments of the present disclosure.

Hereinafter, various embodiments of the present disclosure will be described in detail with reference to the attached drawings. The advantages and features of the present disclosure and methods for achieving them will become clear by referring to the embodiments described in detail below along with the accompanying drawings. However, the technical idea of the present disclosure is not limited to the following embodiments and may be implemented in various different forms. The following examples are merely intended to complete the technical idea of the present disclosure and to cover the technical field to which the present disclosure belongs. is provided to fully inform those skilled in the art of the scope of the present disclosure, and the technical idea of the present disclosure is only defined by the scope of the claims.

In describing various embodiments of the present disclosure, if it is determined that a detailed description of a related known configuration or function may obscure the gist of the present disclosure, the detailed description will be omitted.

Unless otherwise defined, terms (including technical and scientific terms) used in the following embodiments may be used in a meaning that can be commonly understood by those skilled in the art to which this disclosure pertains. It may vary depending on the intentions of engineers working in the related field, precedents, the emergence of new technologies, etc. The terminology used in this disclosure is for describing embodiments and is not intended to limit the scope of this disclosure.

The singular expressions used in the following embodiments include plural concepts, unless the context clearly specifies singularity. Additionally, plural expressions include singular concepts, unless the context clearly specifies plurality.

In addition, terms such as first, second, A, B, (a), (b) used in the following embodiments are only used to distinguish one component from another component, and the terms The essence, order, or order of the relevant components are not limited.

Hereinafter, various embodiments of the present disclosure will be described in detail with reference to the attached drawings.

First, the configuration and operation of the companion animal oral information analysis system according to an embodiment of the present disclosure will be described with reference to FIG. 1.

As shown in FIG. 1, the companion animal oral information analysis system according to this embodiment may be configured to include a companion animal oral information analysis assistance system 100 and a user terminal 200. The user of the user terminal 200 takes a picture so that the color and gum line of the teeth and crown of the target pet 10 are clearly visible, and the captured image resulting from the photograph is used in the pet oral information analysis assistance system (100). ) can be sent.

The user terminal 200 may have a companion animal oral information analysis application installed that is connected to the companion animal oral information analysis assistance system 100 or a separate service server, and the companion animal oral information analysis application may be installed on the companion animal's teeth and crown. It may be to provide a photo taking UI that guides photo taking so that the color and gum line are clearly visible.

Of course, the user terminal 200 is installed with a web browser that accesses the front-end web page served by the companion animal oral information analysis assistance system 100 or a separate service server, and the front-end web page is the companion animal's It may be to provide a photo taking UI that guides photo taking so that the color of teeth and crowns and gum lines are clearly visible.

In addition, the user terminal 200 is installed with a web browser that accesses a front-end web page served by the companion animal oral information analysis assistance system 100 or a separate service server, and the companion animal's information is displayed on the front-end web page. This may be providing a photo to the server that clearly shows the color of the teeth and crown and the gum line. The results analyzed by the server that received the photo are received back to the user terminal 200, and the user terminal 200 may provide a UI that provides the results to the user.

The companion animal oral information analysis assistance system 100 may refer to a virtual machine or container provisioned on one or more cloud nodes. Of course, the companion animal oral information analysis assistance system 100 may be implemented through one or more on-premise physical servers. The companion animal oral information analysis assistance system 100 may store definition data of a companion animal gum line scoring model previously created through machine learning. Machine learning runtime (ML runtime), which is software for executing the pet gum line scoring model, may be installed. That is, the companion animal oral information analysis assistance system 100 receives an oral photographed image of the companion animal from the user terminal 200, inputs the data of the obtained oral photographed image into the gum line scoring model, and Oral information analysis results are generated using data output from the line scoring model, and data including the generated oral information analysis results can be transmitted to the user terminal 200. The oral information analysis result includes information that can be used to derive the crown and gingiva of each tooth segmented, and the boundary line between the crown and the gingiva.

In some embodiments, the pet oral information analysis assistance system 100 may machine-learn the gum line scoring model periodically or aperiodically and deploy the updated gum line scoring model to the user terminal 200. There will be. That is, in this case, the pet oral information analysis assistance system 100 serves as a device that executes the machine learning pipeline of the gum line scoring model, and the user terminal 200 is a type of edge device that executes a machine learning application. It may also play the role of an edge device.

A method of analyzing companion animal oral information according to another embodiment of the present disclosure will be described with reference to FIG. 2. Analysis of companion animal oral information according to this embodiment may be performed by one or more computing devices. That is, in the companion animal oral information analysis method according to this embodiment, all operations may be performed by one computing device, or some operations may be performed by another computing device. For example, some operations may be performed by a first server system and other operations may be performed by a second server system.

Additionally, as the server system is implemented on a cloud computing node, operations performed by one server system may also be performed separately on a plurality of cloud computing nodes. Hereinafter, in explaining the method of analyzing companion animal oral information according to this embodiment, description of the performer of some operations may be omitted, and in this case, the performer is the companion animal oral information analysis assistant described with reference to FIG. 1. It may be system 100.

In step S100, training data is prepared. The learning data may include a plurality of photos taken so that the color and gum line of the pet's teeth and crowns are clearly visible, and labeling data for each photo.

The labeling data may include, for example, whether periodontitis occurs for each photo and one or more additional information. For example, the additional information may include information such as the pet's age and the latest scaling date.

The labeling data, for example, may consist only of whether periodontitis occurs for each photo.

In step S200, a companion animal oral information analysis model using the prepared learning data is machine learned. By automating the acquisition of learning data and establishing a metric for the situation of obtaining the learning data, machine learning for the companion animal oral information analysis model is repeated periodically or aperiodically, but the timing of such machine learning is It may be determined based on the value of the metric.

Next, when machine learning is completed, in step S300, the learned oral information analysis model can be deployed to a computing system running a service back-end instance. As described above, the computing system running the backend instance may be a service server.

When a request for oral information analysis including a captured image is received from the user terminal in step S400, data of the captured image may be input into the companion animal oral information analysis model in step S600. At this time, it goes without saying that pre-designated preprocessing, such as cropping, resizing, or applying a color filter, can be performed on the captured image.

In step S700, a response to the request for companion animal oral information analysis is generated using the data output from the oral information analysis model, and in step S800, a response to the request for companion animal oral information analysis is transmitted to the user terminal. You can.

According to this embodiment, the guardian of a companion animal can check the analysis results related to oral information simply by taking an image of the oral cavity of the companion animal. For example, information about the occurrence and progress of periodontitis can be displayed on the user terminal.

In order to analyze companion animal oral information more accurately, the present disclosure's unique artificial neural network architecture, which allows features that are reflected in the oral image of the companion animal to improve the accuracy of information analysis, be well reflected in the oral image of the companion animal, will be described later. .

The artificial neural network architecture of the oral information analysis model, which will be described later, is elaborately designed so that the oral information analysis model can identify periodontitis that may occur in the oral cavity of a companion animal as accurately as possible by integrating information from various perspectives. Hereinafter, with reference to FIGS. 3 to 7, examples of the artificial neural network architecture of an oral information analysis model that analyzes oral images of companion animals will be described.

Referring to FIG. 3, a CNN (Convolutional Neural Network) 110 that receives a captured image 10 and outputs a score (Z1) obtained by scoring information on the color uniformity of each segmented crown region will be described. The CNN 110 that receives the captured image 10 can output information about the color uniformity of individual crown regions as a scoring score.

The crown area is the visible part of the tooth and is the area that determines the shape and color of the tooth. The crown area is used to evaluate the size, shape, and alignment of teeth, and provides important information in dental care.

Color uniformity of the crown area is a standard that means whether the colors of the pixels that make up the crown area in the captured image are consistent and uniform. The pixels in the crown area have similar colors, and the smaller the color difference, the more uniform the color. there is.

The CNN 110 can output a result of scoring the degree of color uniformity using a color uniformity analysis model created by learning from normal dog images.

The CNN 110 can output a higher score as the color uniformity of one target crown region is more extreme. Extreme color uniformity means that the color is not uniform. In other words, a high color uniformity score may indicate that the color uniformity of a target crown region is more extreme as the colors included in one target crown region are further apart based on the RGB code. Based on the RGB code included in one target crown area, the closer the areas with distant colors are, the more extreme the color uniformity of the target crown area can be judged to be. The more extreme the color uniformity of the target crown area and the higher the score, the higher the probability that inflammation-related symptoms exist.

In the image 210 in which the score Z1, which scores information on the color uniformity of the crown area, is overlaid on the captured image 10, the more extreme the color uniformity, the higher the score may be output.

Referring to FIG. 4, a CNN (Convolutional Neural Network) 110 that receives a captured image 10 and outputs a score (Z2) obtained by scoring information on the color uniformity of the segmented individual gingival areas will be described. The CNN 110 that receives the captured image 10 can output information about the color uniformity of the gingival area as a scoring score.

The gingival area is the part of the gums adjacent to the teeth. The gingiva determines the color and tissue condition of the gums, and the health of the gingival area has a significant impact on oral hygiene and dental care.

Color uniformity of the gingival area is a standard that means whether the colors of the pixels that make up the gingival area in the captured image are consistent and uniform. The pixels in the gingival area have similar colors, and the smaller the color difference, the more uniform the color. there is.

The CNN 110 can output a result of scoring the degree of color uniformity using a color uniformity analysis model created by learning from normal dog images.

The CNN 110 can output a higher score as the color uniformity of one target gingival area is more extreme. The farther away the colors included in one target gingival area are based on the RGB code, the more extreme the color uniformity of the target gingival area can be judged to be. Based on the RGB code included in one target gingival area, the closer the areas with distant colors are, the more extreme the color uniformity of the target gingival area can be judged to be. The more extreme the color uniformity of the target gingival area and the higher the score, the higher the probability that inflammation-related symptoms exist.

In the image 220 in which the score Z2, which scores information on the color uniformity of the gingival area, is overlaid on the captured image 10, the more extreme the color uniformity, the higher the score may be output.

Figure 5 shows a feature fusion (310) that receives a score (Z1) in which the CNN scores information on the color uniformity of the crown area and a score (Z2) in which the CNN scores information on the color uniformity in the gingival area. It shows an image 230 on which the fusion score (Z(f)) output as a result of is overlaid.

Information about the color uniformity of the coronal area and the color uniformity of the gingival area, which are the targets of feature fusion, are information that can affect the probability of the presence of inflammation-related symptoms in the oral cavity.

Specifically, the further away the colors included in one target crown or gingival area are based on the RGB code, or the more similar the areas with distant colors are based on the included RGB code, the higher the probability that inflammation-related symptoms exist.

If the colors included in one target crown area are far apart based on the RGB code, this may mean that tartar has accumulated on the crown, corrosion has occurred, cavities have occurred, or foreign substances are present in the crown. If the colors included in one target gingival area are far apart based on the RGB code, it may mean that the gingiva is swollen, bleeding has occurred in the gingiva, or a foreign substance is present in the gingiva.

If the areas with distant colors based on the RGB code included in one target crown area are similar, the amount of tartar accumulated on the crown may be excessive or the decay may have progressed significantly. If the areas with distant colors based on the RGB code included in one target gingival area are similar, the degree of gingival swelling may be severe or excessive bleeding may have occurred in the gingiva.

As above, information affecting the color uniformity of the crown or gingival area may be generated by the presence of inflammatory symptoms in the oral cavity, may be generated temporarily by problems occurring in individual objects, or are generated by the presence of foreign substances. It may have happened.

Therefore, a score (Z1) that scores information on the color uniformity of the crown area and a score (Z2) that scores information on the color uniformity of the gingival area of the same tooth as the crown area are input to the feature fusion logic 310, A fusion score (Z(f)) may be output from the feature fusion logic 310.

If the color uniformity of the crown area and the color uniformity of the gingival area of the same tooth as the crown area are each more than a predefined score, and the difference between the color uniformity of the crown area and the color uniformity of the gingival area is within the standard value, the fusion score ( Z(f)) can be augmented according to a predefined method. The method in which the fusion score (Z(f)) of the tooth is augmented according to a predefined method includes a method of squaring the fusion score (Z(f)) of the tooth or a method of assigning a weighted score. You can. The above weighted score may be a score that cannot be derived using the color uniformity of the crown area or the color uniformity of the gingival area.

If the color uniformity of the crown area and the color uniformity of the gingival area of the same tooth as the crown area are each more than a predefined score, and the difference between the color uniformity of the crown area and the color uniformity of the gingival area exceeds the standard value, the crown area or gingival area An individual problem may have occurred temporarily in one of the areas, or a foreign substance may be present.

Additionally, although not shown, the object of feature fusion may be information about the color uniformity of the gingival area and information about the degree of recession of the gum line. Information about the color uniformity of the gingival area and the degree of recession of the gum line are information that can affect the probability of the presence of inflammation-related symptoms in the oral cavity. Information about the color uniformity of the gingival area and the degree of recession of the gum line are information that reflects the health status of the gums. Information on the degree of recession of the gum line will be described later with reference to FIG. 6.

Specifically, the further away the colors included in the target gingival area are based on the RGB code, or the more similar the areas with distant colors are based on the included RGB code, or the more severe the degree of gum line recession, the more likely it is that inflammation-related symptoms will exist. It is high.

If the colors included in one target gingival area are far apart based on the RGB code, it may mean that the gingiva is swollen, bleeding has occurred in the gingiva, or a foreign substance is present in the gingiva. If the areas with distant colors based on the RGB code included in one target gingival area are similar, the degree of gingival swelling may be severe or excessive bleeding may have occurred in the gingiva. Recession in one target gum line may be due to a physical accident occurring to the gums, leaving behind a scar. The gum line may differ depending on the oral structure of each individual.

As above, the information affecting the color uniformity of the gingival area may be generated by the presence of inflammatory symptoms in the oral cavity, may be generated by temporary problems occurring in individual objects, or may be generated by the presence of foreign substances. You can.

Therefore, a score that scores information on the color uniformity of the gingival area and a score that scores information on the degree of recession of the gum line identical to the corresponding tooth are input to the feature fusion logic 310, and the fusion score is calculated in the feature fusion logic 310. (Z(f)) can be output.

If the color uniformity of the gingival area and the degree of recession of the gum line of the same tooth as the gingival area are each more than a predefined score and have a difference within the standard value, the fusion score (Z(f)) of the corresponding tooth is calculated according to the predefined method. It can be augmented accordingly. The method in which the fusion score (Z(f)) of the tooth is augmented according to a predefined method includes a method of squaring the fusion score (Z(f)) of the tooth or a method of assigning a weighted score. You can. The above weighted score may be a score that cannot be derived using the color uniformity of the gingival area or the degree of recession of the gum line.

Additionally, although not shown, the object of feature fusion may be information about the color uniformity of the crown area, information about the color uniformity of the gingival area, and information about the degree of recession of the gum line. Information about the color uniformity of the crown area, information about the color uniformity of the gingival area, and information about the degree of recession of the gum line are information that can affect the probability of the presence of inflammation-related symptoms in the oral cavity.

Specifically, the further away the colors included in the target crown area and gingival area are based on the RGB code, or the more similar the area with the distant color is based on the included RGB code, or the more severe the degree of gum line recession, the more inflammation-related symptoms occur. There is a high probability that it exists.

If the colors included in one target crown area are far apart based on the RGB code, this may mean that tartar has accumulated on the crown, corrosion has occurred, cavities have occurred, or foreign substances are present in the crown. If the colors included in one target gingival area are far apart based on the RGB code, it may mean that the gingiva is swollen, bleeding has occurred in the gingiva, or a foreign substance is present in the gingiva.

If the areas with distant colors based on the RGB code included in one target crown area are similar, the amount of tartar accumulated on the crown may be excessive or the decay may have progressed significantly. If the areas with distant colors based on the RGB code included in one target gingival area are similar, the degree of gingival swelling may be severe or excessive bleeding may have occurred in the gingiva.

Recession in one target gum line may be due to a physical accident occurring to the gums, leaving behind a scar. The gum line may differ depending on the oral structure of each individual.

As above, information affecting the color uniformity of the crown or gingival area may be generated by the presence of inflammatory symptoms in the oral cavity, or may be generated temporarily by problems occurring in individual objects.

As described above, information affecting the occurrence of recession in the gum line may be generated by the presence of inflammatory symptoms in the oral cavity, or may be generated by a physical accident that occurs in an individual object.

Therefore, a score that scores information on the color uniformity of the crown area, a score that scores information on the color uniformity of the gingival area, and a score that scores information on the degree of recession of the gum line identical to the tooth in question are characteristic fusion logic (310). is input, and a fusion score (Z(f)) may be output from the feature fusion logic 310.

If the color uniformity of the crown area, the color uniformity of the gingival area, and the degree of recession of the gingival area and the gum line of the same tooth are each more than a predefined score and have a difference within the standard value, the fusion score (Z(f)) of the tooth Can be augmented according to a previously defined method. The method in which the fusion score (Z(f)) of the tooth is augmented according to a predefined method includes a method of squaring the fusion score (Z(f)) of the tooth or a method of assigning a weighted score. You can. The above weighted score may not be derived using the color uniformity of the crown or gingival area and the degree of recession of the gum line.

Referring to FIG. 6, a CNN 110 that receives a captured image 10 and outputs a score Z3 that scores information on the degree of recession of the gum line of each segmented tooth will be described. The CNN 110 that receives the captured image 10 can output information about the degree of recession of the gum line of each individual tooth as a scoring score.

Information about the degree of recession is information that can be used to derive the degree of recession of the gum line. The degree of gum line recession refers to how far the shape and position of the gums recede compared to the teeth. Ideally, the gum line surrounding the teeth should form the same horizontal line as the teeth, but due to various factors, the gum line between the teeth and gums may recede in a direction that exposes the tooth roots. One of the most common causes of gum line recession is periodontitis.

The CNN 110 can output a result of scoring the degree of recession of the gum line of the target tooth using a curvature model of the gum line of each individual tooth created by learning from a normal dog image. The CNN 110 may generate a curvature model of the gum line of each individual tooth by learning from an image containing information about the curvature of the gum line. The curvature of the gum line may be the curvature of a line receding in the direction of the tooth root based on the same horizontal line as the tooth.

The CNN (110) scores the degree of recession of the gum line of the target tooth using a ratio calculation model of the first crown area generated by learning from a normal dog image and the first crown area among the entire first gingival area. Can be printed. The CNN (110) learns with an image containing information about the ratio of the first crown area among the first crown area and the entire first gingival area, and selects the first crown area among the first crown area and the entire first gingival area. It may be that a ratio calculation model of the area has been created. In the ratio calculation model of the first crown area among the first crown area and the entire first gingival area, the first crown area may be the area of the visible part of the target tooth, and the first gingival area may be the area of the gum area adjacent to the target tooth. It may be an area.

The progression of recession of the gum line of one target tooth is determined by deriving the degree of curvature of the interface between the crown and the gingiva of the target tooth, and the degree of recession of the gum line of the target tooth is determined by the first crown area of the target tooth and the first crown area of the target tooth. 1 It can be classified by deriving the ratio of the first coronal area to the entire gingival area.

Depending on the curvature of the gum line covering other teeth for each tooth position, it can be determined that the gum line of the tooth is receding. When it is determined that the gum line of a tooth is in progress, the degree of recession of the gum line may be determined according to the area occupied by the gingival area of the target tooth among the combined gingival area and crown area. The curvature of the gum line is information that can affect the probability of gum line recession, and the area occupied by the gingival area of the target tooth among the combined gingival and crown areas is information that can determine the degree of gum line recession. am.

The determination of the degree of recession based on the curvature of the gum line may have been created by a problem currently temporarily occurring in an individual object, and the determination of the degree of recession based on the area occupied by the gingival area among the combined area of the gingival and coronal areas may be due to damage to the gingival area. It could be a trace of having been worn.

In the image 240 in which the score Z3, which scores information on the degree of recession of the gum line of the tooth, is overlaid on the captured image 10, a higher score may be output as the degree of recession of the gum line of the tooth is more extreme.

With reference to Figure 7, the architecture of the oral information analysis model will be described. According to this artificial neural network architecture, the CNN that receives the captured image 10 outputs a feature score, the feature scores Z1 and Z2 are input to the fusion logic 310, and the fusion score (Z(f)) in the fusion logic is It is output, and the feature score Z3 and the fusion score Z(f) are input to the classifier 320 that performs classification based on the feature score.

An image 250 in which information on teeth suspected of having periodontitis or inflammation is overlaid on the captured image may be output as a result of the classifier.

So far, the companion animal oral information analysis method according to some embodiments of the present disclosure has been described in detail. The embodiments described above should be understood in all respects as illustrative and not restrictive.

8 is a hardware configuration diagram of a computing system according to some embodiments of the present disclosure. The computing system 1000 of FIG. 8 includes one or more processors 1100, a system bus 1600, a communication interface 1200, and a memory 1400 that loads a computer program 1500 executed by the processor 1100. ) and a storage 1300 that stores the computer program 1500. For example, the computing system of FIG. 8 may be the companion animal oral information analysis assistance system described with reference to FIG. 1 .

The processor 1100 controls the overall operation of each component of the computing system 1000. The processor 1100 may perform operations on at least one application or program to execute methods/operations according to various embodiments of the present disclosure.

The memory 1400 stores various data, commands and/or information. The memory 1400 may load one or more computer programs 1500 from the storage 1300 to execute methods/operations according to various embodiments of the present disclosure. Additionally, the memory 1400 may load a runtime for executing the companion animal oral information analysis model described with reference to FIGS. 3 to 7 .

The definition data of the companion animal oral information analysis model is a CNN (Convolutional Neural Network) that outputs the first feature score using information about the color uniformity of the first coronal region and information about the color uniformity of the first gingival region. 1 Expressing an artificial neural network architecture including a CNN that outputs a second feature score using information about the degree of recession of the gum line and a classifier that performs classification based on the first feature score and the second feature score. It could be data.

The bus 1600 provides communication functions between components of the computing system 1000. The communication interface 1200 supports Internet communication of the computing system 1000. Storage 1300 may non-temporarily store one or more computer programs 1500. The computer program 1500 may include one or more instructions implementing methods/operations according to various embodiments of the present disclosure. When the computer program 1500 is loaded into the memory 1400, the processor 1100 can perform methods/operations according to various embodiments of the present disclosure by executing the one or more instructions.

In some embodiments, the computing system 1000 described with reference to FIG. 8 may be configured using one or more physical servers included in a server farm based on cloud technology, such as a virtual machine. there is. In this case, at least some of the processor 1100, memory 1400, and storage 1300 among the components shown in FIG. 8 may be virtual hardware, and the communication interface 1200 may also be a virtual switch. It may be composed of virtualized networking elements such as switches.

The computer program 1500 inputs instructions for acquiring an oral image of a companion animal and data from the obtained oral image into an oral information analysis model previously created through machine learning, and data output from the oral information analysis model. Includes instructions for generating oral information analysis results using, wherein the oral information analysis model includes information on the color uniformity of the first coronal region, information on the color uniformity of the first gingival region, and a first 1 It is composed of an artificial neural network including a CNN (Convolutional Neural Network) that outputs information about the degree of recession of the gum line and a classifier that performs classification based on the information output from the CNN, and the first gingival area is an adjacent area of the first crown area, and is an area representing the gingiva corresponding to the first crown expressed by the first crown area, and the first gum line is an area between the first gingival area and the first crown area. It could be a borderline.

So far, various embodiments of the present disclosure and effects according to the embodiments have been mentioned with reference to FIGS. 1 to 8 . The effects according to the technical idea of the present disclosure are not limited to the effects mentioned above, and other effects not mentioned can be clearly understood by those skilled in the art from the description below.

The technical ideas of the present disclosure described so far can be implemented as computer-readable code on a computer-readable medium. The computer program recorded on the computer-readable recording medium can be transmitted to another computing device through a network such as the Internet and installed on the other computing device, and thus can be used on the other computing device.

Although operations are shown in the drawings in a specific order, it should not be understood that the operations must be performed in the specific order shown or sequential order or that all illustrated operations must be performed to obtain the desired results. In certain situations, multitasking and parallel processing may be advantageous. Although embodiments of the present disclosure have been described above with reference to the attached drawings, those skilled in the art will understand that the present invention can be implemented in other specific forms without changing the technical idea or essential features. I can understand that there is. Therefore, the embodiments described above should be understood in all respects as illustrative and not restrictive. The scope of protection of the present invention should be interpreted in accordance with the claims below, and all technical ideas within the equivalent scope should be construed as being included in the scope of rights of the technical ideas defined by this disclosure.

Claims (13)

In a method performed by a computing device,
Obtaining an oral image of a companion animal; and
Inputting the data of the acquired oral photographed image into an oral information analysis model previously created through machine learning, and generating an oral information analysis result using the data output from the oral information analysis model,
The oral information analysis model is,
From the intraoral image, information on the color uniformity of the first crown area indicating whether the color of the pixels constituting the crown area is consistent and uniform, and the first gingival area indicating whether the color of the pixels constituting the gingival area are consistent and uniform. An artificial neural network that includes a CNN (Convolutional Neural Network) that outputs information about the color uniformity of the area and information about the degree of recession of the first gum line, and a classifier that performs classification based on the information output from the CNN. It is composed of a neural network,
The first gingival area is an area adjacent to the first crown area and is an area representing the gingiva corresponding to the first crown expressed by the first crown area,
The first gum line is a boundary line between the first gingival area and the first coronal area,
How to analyze pet oral information. According to claim 1,
The classifier is,
To perform classification using the result of feature fusion of information about the color uniformity of the first coronal region and information about the color uniformity of the first gingival region,
How to analyze pet oral information. According to clause 2,
The result of feature fusion of information about the color uniformity of the first coronal region and information about the color uniformity of the first gingival region is:
When the color uniformity of the first coronal region and the color uniformity of the first gingival region have a difference within the standard value, it is the result of applying a predefined transformation to the fused feature,
How to analyze pet oral information. According to clause 3,
If the color uniformity of the first coronal region and the color uniformity of the first gingival region have a difference within the standard value, the result of applying a predefined transformation to the fused feature is,
Augmenting according to a predefined method,
How to analyze pet oral information. According to clause 4,
What is augmented according to the previously defined method is,
A weighted score is given that cannot be derived using the color uniformity of the crown area and the color uniformity of the gingival area,
How to analyze pet oral information. According to clause 4,
What is augmented according to the previously defined method is,
The color uniformity of the first coronal region and the color uniformity of the first gingival region are squared, respectively.
How to analyze pet oral information. According to claim 1,
Information on the degree of recession of the first gum line,
Containing information about the curvature of the first gum line,
How to analyze pet oral information. According to claim 1,
Information on the degree of recession of the first gum line,
Containing information about the first crown area and the ratio of the first crown area among the entire first gingival area,
How to analyze pet oral information. According to claim 1,
The classifier is,
To perform classification using the result of feature fusion of information about the color uniformity of the first gingival area and information about the degree of recession of the first gum line,
How to analyze pet oral information. According to clause 9,
The result of feature fusion of information about the color uniformity of the first gingival area and information about the degree of recession of the first gum line is,
When the color uniformity of the first gingival area and the degree of recession of the first gum line are each above a predefined score and have a difference within the standard value, it is the result of applying a predefined transformation to the fused feature,
How to analyze pet oral information. According to claim 1,
The classifier is,
Performing classification using the result of feature fusion of information about the color uniformity of the first crown area, information about the color uniformity of the first gingival area, and information about the degree of recession of the first gum line. thing,
How to analyze pet oral information. According to claim 11,
The result of feature fusion of the information about the color uniformity of the first crown area, the information about the color uniformity of the first gingival area, and the information about the degree of recession of the first gum line is,
When the information on the color uniformity of the first crown area, the color uniformity of the first gingival area, and the degree of recession of the first gum line are each more than a predefined score and have a difference within the standard value, the fused features The result of applying a predefined transformation to
How to analyze pet oral information. A memory into which a companion animal oral information analysis program is loaded; and
Including a processor that executes the companion animal oral information analysis program,
The companion animal oral information analysis program is,
Instructions for acquiring oral images of companion animals; and
Including instructions for inputting the data of the acquired oral imaging image into an oral information analysis model previously created through machine learning and generating an oral information analysis result using the data output from the oral information analysis model,
The oral information analysis model is,
From the intraoral image, information on the color uniformity of the first crown area indicating whether the color of the pixels constituting the crown area is consistent and uniform, and the first gingival area indicating whether the color of the pixels constituting the gingival area are consistent and uniform. An artificial neural network that includes a CNN (Convolutional Neural Network) that outputs information about the color uniformity of the area and information about the degree of recession of the first gum line, and a classifier that performs classification based on the information output from the CNN. It is composed of a neural network,
The first gingival area is an area adjacent to the first crown area and is an area representing the gingiva corresponding to the first crown expressed by the first crown area,
The first gum line is a boundary line between the first gingival area and the first coronal area,
Pet oral information analysis system.

Images