Oral cavity state estimation apparatus

Abstract

To provide a technology capable of readily estimating an oral cavity state.SOLUTION: An oral cavity state estimation apparatus (10) comprises: means (21) for acquiring a tooth image in which one or more specific teeth of a person to be estimated is or are included; means (23) for extracting information on a natural tooth area of a tooth area, which indicates the specific tooth, of the acquired tooth image; and means (24) for estimating an oral cavity state of the person to be estimated, on the basis of the extracted information on the natural tooth area.SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] The present invention relates to a technique for estimating an oral cavity condition using images of teeth. [Background technology]

[0002] The following Patent Document 1 discloses a data generation system that can efficiently generate training data for a machine learning model that estimates the condition of the oral cavity from oral cavity images. Patent Document 2 below discloses that oral cavity images P are analyzed using a machine learning learning model M to obtain future information regarding the condition of the user's oral cavity after a specified period of time has passed.The disclosure discloses that the oral cavity condition includes information regarding the presence or absence or degree of gingival inflammation, caries, and staining, as well as information regarding the condition of the teeth, gums, and oral mucosa. The following Patent Document 3 discloses a system that obtains a dentition image that includes at least a portion of the dentition in the oral cavity, determines the oral condition related to the teeth and gums from the dentition image by referring to the learning results of a machine learning mechanism, and extracts recommended information regarding oral care based on the determined oral condition. The following Patent Document 4 discloses an oral health prediction device that uses a machine learning algorithm to analyze a user's periodontal images, analyzes whether or not they have orthodontic treatment, the state of their caries, and the state of their prosthetic appliances, and provides predicted information on the user's oral health condition. The following Patent Document 5 discloses a mobile device that extracts image data of the gingival region and the dental region by comparing image data of the oral region captured by a camera with a judgment model previously constructed by machine learning, and estimates the condition of the oral region based on the image data of the oral region. Examples of the condition of the oral region include the degree of swelling of the gums, the condition of the periodontal pockets, the condition of dirt in the oral cavity, and the condition of the gum wounds. The following Patent Document 6 discloses an intraoral condition assessment device that identifies each tooth from an image, determines the gum condition of the portion of the gums in the image that corresponds to the identified tooth, and assesses the overall condition of the gums captured in the image. The following Patent Document 7 discloses an oral disease diagnosis system that diagnoses diseases in a user's oral cavity based on intraoral photographs consisting of multiple images showing specific areas in the oral cavity. The system determines the presence and state of periodontal disease based on the characteristics of at least one image of the gum region and the tooth region, and determines the presence and state of dental caries based on the characteristics of the image of the tooth region. The following Patent Document 8 discloses a support system for extending healthy lifespan that analyzes information capable of grasping the current and past functions and health conditions of biological tissues in the oral and pharyngeal areas, the actual state of care, and the actual state of awareness and knowledge, based on user information, and at the time of providing the analyzed information, generates information useful for extending the user's healthy lifespan and provides this information to the user's terminal. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent Publication No. 2022-69880 [Patent Document 2] JP 2022-68486 A [Patent Document 3] JP 2020-179173 A [Patent Document 4] Special Publication No. 2022-508923 [Patent Document 5] Patent Publication No. 2021-53175 [Patent Document 6] JP 2020-54646 A [Patent Document 7] JP 2019-155027 A [Patent Document 8] JP 2019-67302 A Summary of the Invention [Problem to be solved by the invention]

[0004] As described above, there are techniques for estimating the condition of teeth or gums by analyzing oral cavity images. In the analysis of oral cavity images in such existing techniques, the oral cavity images are input to a machine learning model, or a partial region of a target part (the teeth themselves or gums) whose condition is to be estimated in the oral cavity image is analyzed.

[0005] However, there is still room for improvement in these existing methods. For example, in order to obtain an image that reliably captures the gum area in order to estimate the state of the gums, the gums must be largely exposed during shooting, which can cause problems such as difficulty in capturing the image. There is also room for improvement in the accuracy of the estimation.

[0006] The present invention has been made from this perspective, and provides a technique that enables the state of the oral cavity to be easily estimated. [Means for solving the problem]

[0007] According to the present invention, an oral condition estimation device can be provided which includes an image acquisition means for acquiring a tooth image in which one or more specific teeth of a person being evaluated are captured, an image analysis means for extracting information about a natural tooth area within a tooth area representing the specific tooth in the acquired tooth image, and an estimation means for estimating the oral condition of the person being evaluated based on the information about the extracted natural tooth area.

[0008] Further, an oral condition estimation method can be provided that is executed by one or more computers having one or more processors and memory, in which the one or more processors execute an image acquisition step of acquiring a tooth image in which one or more specific teeth of the subject are captured, an image analysis step of extracting information about the natural tooth area excluding discolored areas and restored areas from within the tooth area representing the specific teeth in the acquired tooth image, and an estimation step of estimating the oral condition of the subject based on the information about the extracted natural tooth area.

[0009] It is also possible to provide a computer program for causing one or more processors to execute the above-described oral cavity state estimation method, or a recording medium for recording the computer program. Effect of the Invention

[0010] According to the present invention, a technique that enables easy estimation of the state of the oral cavity can be provided. [Brief description of the drawings]

[0011] [Figure 1] FIG. 2 is a diagram conceptually illustrating an example of a hardware configuration of the oral cavity condition estimating device according to the present embodiment. [Diagram 2] FIG. 2 is a diagram conceptually illustrating an example of the software configuration of the oral cavity condition estimating device according to the present embodiment. [Diagram 3] FIG. 2 shows an example of a tooth image and a natural tooth region. [Figure 4] FIG. 2 is a diagram showing an example of a natural tooth region, a discolored region, and a restored region in a maxillary central incisor. [Diagram 5] 4 is a flowchart showing an example of operation of the oral cavity condition estimating device according to the embodiment. [Figure 6] 1 is a graph showing a correlation between age and saturation according to classification of gingival conditions. [Figure 7] 13 is a graph showing a method for estimating the state of gingival inflammation using the correlation between the subject's age and the saturation of the natural tooth area. [Figure 8] 1 is a graph showing the relationship between the hue angle of the natural tooth region and periodontal pocket depth classification. [Figure 9] 1 is a graph showing the accuracy of estimating the progress state of periodontal disease. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0012] Hereinafter, preferred embodiments of the present invention (hereinafter, referred to as the present embodiment) will be described. The following embodiments are merely examples, and the present invention is not limited to the configurations of the following embodiments.

[0013] [Oral condition estimation device] FIG. 1 is a diagram conceptually illustrating an example of the hardware configuration of an oral cavity condition estimating device (hereinafter sometimes abbreviated as the estimating device) 10 according to this embodiment. The estimation device 10 is a so-called computer (information processing device) and includes a CPU 11, a memory 12, an input / output interface (I / F) 13, a communication unit 14, etc. The estimation device 10 may be a stationary PC (Personal Computer), a portable terminal such as a portable PC, a smartphone, or a tablet, or may be a dedicated computer.

[0014] The CPU 11 is a so-called processor, and may include an application specific integrated circuit (ASIC), a digital signal processor (DSP), a graphics processing unit (GPU), etc. in addition to a general central processing unit (CPU). The memory 12 is a random access memory (RAM), a read only memory (ROM), or an auxiliary storage device (such as a hard disk).

[0015] The input / output I / F 13 can be connected to user interface devices such as a display device 15 and an input device 16. The display device 15 is a device that displays a screen corresponding to drawing data processed by the CPU 11 or the like, such as an LCD (Liquid Crystal Display) or CRT (Cathode Ray Tube) display. The input device 16 is a device that accepts input of user operations, such as a keyboard, a mouse, etc. The display device 15 and the input device 16 may be integrated and realized as a touch panel. The communication unit 14 communicates with other computers via a communication network, and exchanges signals with other devices such as a printer, etc. A portable recording medium, etc. can also be connected to the communication unit 14.

[0016] The hardware configuration of the estimation device 10 is not limited to the example in Fig. 1. The estimation device 10 may include other hardware elements not shown. Furthermore, the number of each hardware element is not limited to the example in Fig. 1. For example, the estimation device 10 may have multiple CPUs 11. Furthermore, the estimation device 10 may be realized by multiple computers consisting of multiple housings.

[0017] The estimation device 10 can execute the oral cavity condition estimation method by executing a computer program stored in the memory 12 by the CPU 11. The computer program is installed from a portable recording medium such as a CD (Compact Disc) or a memory card or from another computer on a network via the input / output I / F 13 or the communication unit 14, and stored in the memory 12.

[0018] Next, an overview of the processing configuration of the estimation device 10 will be described, and details will be described later. The estimation device 10 includes at least an image acquisition means, an image analysis means, and an estimation means. The image acquisition means acquires a tooth image in which one or more specific teeth of the person being evaluated are captured. The "specific tooth" may be one or more teeth in the upper or lower jaw, but is preferably either one or both of the upper central incisors. This is because the upper central incisors are easily exposed by opening the mouth wide to the side with the jaws closed, and because each tooth is large, it is easy to extract information from them. The image acquiring means can acquire tooth images captured by a camera from the camera via the input / output I / F 13 or the communication unit 14, or can acquire tooth images stored in the memory 12 of the estimation device 10 or another computer. In this way, the method of acquiring tooth images by the image acquiring means is not limited.

[0019] The image analysis means extracts information about a natural tooth region from a tooth region indicating a specific tooth in the tooth image acquired by the image acquisition means. The "natural tooth region" refers to a region that indicates a healthy tooth state, and can also be said to be the remaining region of the tooth region of a specific tooth excluding at least discolored regions and restored regions. "Discolored area" refers to an area that has changed from the original (natural) color of the tooth, and includes white spots, areas discolored by tetracycline antibiotics, caries, areas stained due to eating, drinking, smoking, etc. Areas stained due to eating, drinking, smoking, etc. are areas where exogenous stains such as tannins and polyphenols derived from food and drink have adhered to the tooth surface, and therefore can also be called areas discolored by exogenous stains. Furthermore, the term "repaired area" refers to an area that has been repaired through treatment, etc., and includes areas treated with resin (including metal), pulp-extracted teeth, dentures, etc. The "natural tooth area information" extracted by the image analysis means may be an image showing only the natural tooth area, or it may be information that identifies the range of the natural tooth area within the tooth image, or it may be color information of the natural tooth area.

[0020] The estimation means estimates the oral cavity condition of the person to be evaluated based on the information of the natural tooth area extracted by the image analysis means. The "oral condition" estimated by the estimation means may be the condition of the entire oral cavity, or may be the condition of one or more parts that constitute the oral cavity, such as the teeth, gingiva (gums), tongue, palate, lips, etc. In addition, examples of the "condition" of the oral cavity that can be estimated include, but are not limited to, the presence or absence of symptoms, classification of symptoms, and progression of symptoms.

[0021] In this manner, in this embodiment, attention is focused on information about the natural tooth region within a tooth region showing one or more specific teeth in the tooth image of the subject, and the oral condition of the subject is estimated based on the information about the natural tooth region. In the existing technology for estimating the condition of teeth or gums, estimation is performed using the tooth region including the discolored region, but in this embodiment, estimation accuracy is improved by focusing on the information of the natural tooth region to estimate the oral cavity condition. In addition, the tooth image used for estimating the oral cavity condition only needs to include one or more specific teeth of the subject, so it is possible to estimate the oral cavity condition more easily than with the existing method.

[0022] The following provides a more detailed explanation of the processing configuration of the estimation device 10. In the following explanation, the same content as that outlined above will be omitted as appropriate.

[0023] FIG. 2 is a diagram conceptually illustrating an example of the software configuration of the oral cavity condition estimating device (estimating device) 10 according to this embodiment. 2 by causing the CPU 11 to execute a computer program stored in the memory 12. Specifically, the estimation device 10 has, as its software configuration, an image acquisition unit 21, an information acquisition unit 22, an image analysis unit 23, an estimation unit 24, an output processing unit 25, and the like. However, since each software component shown in FIG. 2 is shown conceptually separated for ease of explanation, the software configuration realized by the estimation device 10 does not need to be clearly divided into each component as shown in FIG. 2.

[0024] The image acquisition unit 21 generally corresponds to the image acquisition means described above. That is, the image acquisition unit 21 acquires a tooth image in which one or more specific teeth of the subject are captured. As described above, the method of acquiring the tooth image by the image acquisition unit 21 is not limited, but for example, the image acquisition unit 21 acquires the tooth image via the communication unit 14 from the camera that captured the tooth image.

[0025] FIG. 3 is a diagram showing an example of a tooth image and a natural tooth region. For example, the image acquisition unit 21 acquires a tooth image as exemplified in Fig. 3. In the example of Fig. 3, a plurality of teeth of the subject are captured, including two upper central incisors. In addition, it is preferable that the acquired tooth image is a non-reflective image, as exemplified in Fig. 3. The tooth image may be made a non-reflective image by a photographing method such as using a polarizing plate, or may be made a non-reflective image by removing the reflected light component from the tooth image including the reflected light component through image processing by the image acquisition unit 21. An existing method may be used as the image processing method for removing the reflected light component.

[0026] The information acquiring unit 22 acquires the age information of the person being evaluated. Naturally, the person being evaluated here is the same person whose teeth are reflected in the tooth image acquired by the image acquiring unit 21. The acquired age information may indicate the actual age of the person being evaluated, or may indicate an age range, such as 30s, 40s, etc. The information acquiring unit 22 can acquire, for example, age information input by a user operation using the input device 16. However, the method of acquiring the age information is not limited, and the information acquiring unit 22 can also acquire age information transmitted from another computer via the communication unit 14.

[0027] The image analysis unit 23 corresponds to the image analysis means described above in summary. That is, the image analysis unit 23 extracts information on the natural tooth region from the tooth region indicating the specific tooth in the tooth image acquired by the image acquisition unit 21. The tooth region indicating a specific tooth in the tooth image may be identified using an image recognition technology, or may be identified by a user operation using the input device 16 after displaying the tooth image on the display device 15. In identification using image recognition technology, each tooth may be recognized from a tooth image captured from the front using existing image recognition technology, and from among them, for example, a maxillary central incisor may be recognized based on the positional relationship of each tooth. The natural tooth region in the tooth region recognized in this way may also be identified using image recognition technology, or may be identified by a user operation using the input device 16 after displaying a tooth image on the display device 15. When identifying using image recognition technology, the natural tooth region may be identified based on the color distribution in the tooth region of a specific tooth, etc.

[0028] In addition, an AI model can be constructed that takes a tooth image as input and outputs information that enables identification of a natural tooth area within a tooth area representing a specific tooth from the tooth image, and then inputs a tooth image into the trained AI model to identify the natural tooth area from the information obtained. Also, an AI model may be used that takes a tooth image (which may be an image showing only a specific tooth) as an input and outputs information on the natural tooth region used by the estimation unit 24 described later. In this case, the AI ​​model may be constructed to output color information of the natural tooth region (e.g., any one, more than one, or all of the hue angle information, saturation information, and brightness information of the natural tooth region) as described later as information on the natural tooth region. In this case, training data including a plurality of pairs of a tooth image and a correct answer for color information of the natural tooth region in the tooth region showing the specific tooth in the tooth image may be used, and learning may be performed by an existing learning method. In this case, the image analysis unit 23 inputs the tooth image acquired by the image acquisition unit 21 into the AI ​​model, thereby extracting color information of the natural tooth region in the tooth region showing the specific tooth in the tooth image. The AI ​​model may be composed of an AI model that receives a tooth image as an input and outputs information that enables identification of a natural tooth region in a tooth region that indicates a specific tooth from the tooth image, and an AI model that receives information that enables identification of the natural tooth region and the tooth image as input and outputs information on the natural tooth region. Such an AI model may be a neural network model obtained by deep learning, etc., and the data structure, learning algorithm, etc. of the model are not limited.

[0029] 4 is a diagram showing an example of a natural tooth region, a discolored region, and a restoration region of a maxillary central incisor. In FIG. 4, the tooth region of the maxillary central incisor is shown by a long chain line, and the natural tooth region is shown by a dashed line. FIG. 4(a) shows an example in which the natural tooth region is specified from the tooth region of the maxillary central incisor by excluding the restoration region of the treated part such as resin. FIG. 4(b) shows an example in which the natural tooth region is specified from the tooth region of the maxillary central incisor by excluding the white spot part (discolored region). FIG. 4(c) shows an example in which the natural tooth region is specified from the tooth region of the maxillary central incisor by excluding the discolored region such as caries. FIG. 4(d) shows an example in which the natural tooth region is specified from the tooth region of the maxillary central incisor by excluding the restoration region of the pulp extraction tooth. FIG. 4(e) shows an example in which the entire tooth region of the maxillary central incisor is excluded as the region discolored by tetracycline antibiotics (discolored region). FIG. 4(f) shows an example in which the entire tooth region of the maxillary central incisor is excluded as the region of the denture (restoration region).

[0030] In this embodiment, the image analysis unit 23 extracts, as the information of the natural tooth region, any one, any plurality, or all of the hue angle information, saturation information, and lightness information of the natural tooth region. The hue angle information indicates the hue angle, e.g., L * a * b * In color space a * The red (+) axis is set to 0 degrees. * It is expressed as the angle h shifted counterclockwise relative to the hue with the yellow (+) axis at 90 degrees. Saturation information indicates the vividness of a color, e.g., L * a * b * In color space C * That is, saturation (C * If the value of saturation (C * ) is small, the color is closer to the center of the circle in the color space and becomes dull. The lightness information indicates the degree of lightness or darkness of a color. For example, L * a * b * In color space, L * That is, the lightness (L * ) is larger, the color becomes closer to white, and the lightness (L * ) is small, the color becomes closer to black. The method of calculating the hue angle, saturation, and brightness from the natural tooth region of the tooth image is not limited in any way, and various existing methods can be used. For example, the average value or representative value of the hue angle, saturation, and brightness calculated from each pixel value of the natural tooth region are calculated as the hue angle information, saturation information, and brightness information of the natural tooth region.

[0031] The estimation unit 24 corresponds to the estimation means described above in brief. That is, the estimation unit 24 estimates the oral cavity state of the person to be evaluated based on the information of the natural tooth region extracted by the image analysis unit 23. The inventors have found that there is a correlation between specific information of the natural tooth region and a specific oral cavity condition of a person, and the estimation unit 24 uses such correlation to estimate the oral cavity condition of the person being evaluated.

[0032] Specifically, the estimation unit 24 can estimate the oral cavity condition indicating the gingival inflammation state of the person being evaluated, based on the age information acquired by the information acquisition unit 22 and the saturation information of the natural tooth region extracted by the image analysis unit 23. As the gingival inflammation state, for example, the presence or absence of gingival inflammation in one or more places can be estimated. The present inventors have found that there is a correlation between the saturation information of the natural tooth region and age depending on the presence or absence of gingival inflammation. For this reason, the estimation unit 24 generates a correlation equation between saturation and age in advance from a sample without gingival inflammation, inputs the age of the subject into the correlation equation, and compares the estimated value of saturation obtained with the actual measured value of saturation of the natural tooth region extracted by the image analysis unit 23. If the actual measured value is greater than the estimated value by a predetermined threshold, the estimation unit 24 can estimate that the subject has gingival inflammation. As a similar method, for example, the estimation unit 24 holds in advance a boundary equation (the above correlation equation) that can divide the two-axis coordinate of saturation and age into a region with gingival inflammation and a region without gingival inflammation, and plots the age of the subject and the saturation of the natural tooth region extracted by the image analysis unit 23 on the two-axis coordinate to estimate the presence or absence of gingival inflammation of the subject. In this way, if there is a dental image that shows a specific tooth and age information, it is possible to estimate the gingival inflammation state of the person being evaluated, so according to this embodiment, the oral cavity state can be easily estimated.

[0033] The estimation unit 24 can also estimate the oral cavity condition indicating the periodontal pocket depth state of the person being evaluated based on the hue angle information extracted by the image analysis unit 23. As the periodontal pocket depth state, it is possible to estimate whether or not there is a periodontal pocket depth of 4 mm or more, which is an index used to determine the degree of periodontitis. For example, as described below as an example, the estimation unit 24 can determine and store a predetermined threshold value in advance based on a subject sample, and if the hue angle is equal to or greater than the threshold value, determine that there is a periodontal pocket depth of 4 mm or more, and if the hue angle is less than the threshold value, determine that there is no periodontal pocket depth of 4 mm or more. Thus, if there is a tooth image that captures a specific tooth, it is possible to estimate the periodontal pocket depth state of the person being evaluated, and therefore, according to this embodiment, the state of the oral cavity can be easily estimated.

[0034] Furthermore, the estimation unit 24 can estimate the progression state of periodontal disease of the person being evaluated based on the saturation information and hue angle information extracted by the image analysis unit 23 and the age information acquired by the information acquisition unit 22. The progression state of periodontal disease may be indicated in three levels, namely, healthy state, tendency to gingivitis, and tendency to periodontitis, or in four levels, namely, healthy state, tendency to gingivitis, tendency to moderate periodontitis, and tendency to severe periodontitis, or in two levels, namely, tendency to periodontitis or not, or caution required for periodontal disease or not, For example, the estimation unit 24 can estimate the presence or absence of gingival inflammation based on the age information and the saturation information of the natural tooth region as described above, and can also estimate the presence or absence of a periodontal pocket depth of 4 mm or more based on the hue angle information as described above, and can estimate the progression state of periodontal disease of the subject by combining these estimation results. The estimation unit 24 estimates a healthy state when there is no gingival inflammation and no periodontal pocket depth of 4 mm or more, estimates a tendency to gingival inflammation when there is gingival inflammation and no periodontal pocket depth of 4 mm or more, and estimates a tendency to periodontitis when there is no gingival inflammation or when there is periodontitis and a periodontal pocket depth of 4 mm or more. The estimation unit 24 can also estimate a tendency to moderate periodontitis when there is no gingival inflammation and a periodontal pocket depth of 4 mm or more, and estimate a tendency to severe periodontitis when there is periodontitis and a periodontal pocket depth of 4 mm or more. In this way, if there is a tooth image that shows a specific tooth and age information, it is possible to estimate the progression of periodontal disease in the person being evaluated, so according to this embodiment, the condition of the oral cavity can be easily estimated.

[0035] The estimation unit 24 can also estimate the oral cavity condition indicating the saliva condition of the subject based on either or both of the brightness information and saturation information extracted by the image analysis unit 23. As the saliva condition, either or both of the saliva composition and the amount of saliva can be estimated. As the saliva composition, the ratio of calcium components and phosphorus components in saliva, the amount of calcium components and phosphorus components, the total protein concentration in saliva, etc. can be estimated, as described in detail in the examples. The inventors have found that there is a correlation between the lightness and chroma information of the natural tooth region and the saliva state. Therefore, the estimation unit 24 generates and stores in advance a correlation equation between the lightness and saliva state of the natural tooth region, a correlation equation between chroma and saliva state, or a correlation equation between lightness, chroma and saliva state based on a subject sample, and inputs either or both of the lightness information and chroma information extracted by the image analysis unit 23 into this correlation equation, thereby estimating the saliva state of the subject. On the other hand, there is knowledge that dental caries can be suppressed when the amount of saliva is large, when the ratio of calcium and phosphorus components in saliva is high, when the amounts of calcium and phosphorus components in saliva are high, and when the total protein concentration in saliva is high. Therefore, according to this embodiment, if there is a tooth image that captures a specific tooth, it is possible to estimate such a saliva condition, and therefore it is possible to easily estimate the oral cavity condition.

[0036] The output processing unit 25 displays recommended information on either or both of oral care products and oral care methods on the display device 15 based on the oral condition of the person to be evaluated estimated by the estimation unit 24. However, the output form of the recommended information by the output processing unit 25 is not limited to this example, and the recommended information may be printed on a printer device, or the recommended information may be converted into an electronic file and the electronic file may be transmitted to another computer, recording medium, or the like. The recommended information for oral care products may present toothpaste, toothbrushes, etc. that are suited to the oral condition of the person being evaluated, and the recommended information for oral care methods may present tooth brushing methods and other care methods that are suited to the oral condition of the person being evaluated. For example, the output processing unit 25 stores oral care product information and oral care method information in association with each oral condition that can be estimated by the estimation unit 24, and based on the oral condition estimated by the estimation unit 24 for the subject, it can read out either the oral care product information or the oral care method information or both associated with that oral condition and generate them as recommended information. However, the method of generating the recommendation information is not limited to this example.

[0037] The output processing unit 25 can also output, as the oral cavity condition information of the subject, any one, any plurality, or all of the gingival inflammation condition, periodontal pocket depth condition, periodontal disease progression condition, and saliva condition of the subject estimated by the estimation unit 24. The output form of the oral cavity condition information is not limited in any way, and may be, for example, displayed on the display device 15, printed on a printer device, or transmitted as an electronic file to another computer or recording medium.

[0038] [Oral condition estimation method] FIG. 5 is a flowchart showing an example of the operation of the oral cavity condition estimating device 10 according to the present embodiment. Hereinafter, the oral cavity condition estimation method according to the present embodiment will be described with reference to Fig. 5. Each step shown in Fig. 5 is executed by the estimation device 10 or the CPU 11 (processor). Note that each step shown in Fig. 5 is similar to the processing content by each processing module of the estimation device 10, and therefore, the details of each step will be omitted below as appropriate.

[0039] The estimation device 10 (image acquisition unit 21) acquires a tooth image in which one or more specific teeth of the subject are captured (S51). As described above, the acquired tooth image preferably includes both maxillary central incisors as the specific teeth and is a non-reflective light image.

[0040] The estimation device 10 (image analysis unit 23) identifies (S52) a natural tooth region within the tooth region indicating the specific tooth in the tooth image acquired in step (S51). If the estimation device 10 fails to identify the natural tooth region (S53; NO), the estimation device 10 ends the process. At this time, the estimation device 10 can also display on the display device 15 that the estimation device 10 has failed to identify the natural tooth region.

[0041] When the estimation device 10 (image analysis unit 23) has succeeded in identifying the natural tooth region (S53; YES), it extracts information on the identified natural tooth region. In the operation example of FIG. 5, the estimation device 10 extracts saturation information, hue angle information, and lightness information of the natural tooth region as information on the natural tooth region (S54) (S57) (S60). For example, the estimation device 10 can calculate saturation, hue angle, and lightness based on pixel values ​​of the natural tooth region identified in the tooth image.

[0042] Furthermore, the estimation device 10 (information acquisition unit 22) acquires age information of the person being evaluated (S55). As described above, the acquired age information may indicate an actual age or a generation. For example, the estimation device 10 can cause an input screen for the person being evaluated to input their age to be displayed on the display device 15, and acquire the age information entered into the input screen.

[0043] Next, the estimation device 10 (estimation unit 24) estimates the gingival inflammation state of the person being evaluated (S56) based on the saturation information of the natural tooth area extracted in step (S54) and the age information of the person being evaluated obtained in step (S55). Furthermore, the estimation device 10 (estimation unit 24) estimates the periodontal pocket depth state of the subject (S58) based on the hue angle information of the natural tooth region extracted in step (S57). Furthermore, the estimation device 10 (estimation unit 24) can also estimate (S59) the progression state of periodontal disease of the person being evaluated based on the gingival inflammation state estimated in step (S56) and the periodontal pocket depth state estimated in step (S58). In addition, the estimation device 10 (estimation unit 24) estimates the saliva condition of the person to be evaluated (S61) based on one or both of the saturation information and the brightness information of the natural tooth region extracted in step (S60). Specific methods for estimating these oral cavity conditions are as described above.

[0044] The estimation device 10 (output processing unit 25) outputs (S62) oral cavity condition information of the subject indicating the gingival inflammation state, periodontal pocket depth state, periodontal disease progression state, and saliva state estimated in steps (S56), (S58), (S59), and (S61). The output form of the oral cavity condition information is not limited as described above.

[0045] Furthermore, the estimation device 10 (output processing unit 25) outputs (S63) recommendation information for either or both of an oral care product and an oral care method based on the gingival inflammation state, periodontal pocket depth state, periodontal disease progression state, and saliva state estimated in steps (S56), (S58), (S59), and (S61). The output form of this recommendation information is not limited as described above.

[0046] [Variations] The contents of the above-described embodiment can be modified as appropriate. For example, the cavity state estimation method and the operation of the estimation device 10 in this embodiment, as well as the order of execution of each step, are not limited to the example in FIG. 5, and can be changed as appropriate. For example, any one or more of steps (S56), (S58), (S59), and (S61) may be omitted. In this case, among steps (S54), (S55), (S57), and (S60), steps that are not necessary for the estimation in each omitted step may also be omitted. Furthermore, either or both of step (S62) and step (S63) may be omitted.

[0047] In addition, the image analysis unit 23 may further extract information about non-natural tooth areas indicating either or both of discolored areas or restored areas within the tooth area indicating a specific tooth in the tooth image acquired by the image acquisition unit 21. In this case, the image analysis unit 23 can identify the non-natural tooth region in the same way as the natural tooth region. That is, the non-natural tooth region may be identified using an image recognition technique, may be identified by a user operation using the input device 16 by displaying a tooth image on the display device 15, or may be identified using a trained AI model. When both discolored regions and repaired regions are identified as non-natural tooth regions, the image analysis unit 23 may assign a flag to each identified region to indicate whether it corresponds to a natural tooth region, a discolored region, or a repaired region. Furthermore, when identifying discolored regions, they may be distinguished by type, such as white spot region (see FIG. 4(b)), caries region (see FIG. 4(c)), tetracycline region (see FIG. 4(e)), and when identifying repaired regions, they may be distinguished by type, such as resin or other treatment region (see FIG. 4(a)), pulp extraction tooth region (see FIG. 4(d)), denture region (see FIG. 4(f)).

[0048] In this case, the estimation unit 24 can estimate an oral cavity condition that further indicates either or both of the caries state or discoloration state of the subject's teeth based on the information of the non-natural tooth area extracted by the image analysis unit 23 and the saliva state estimated by the estimation unit 24. For example, the caries risk can be estimated as the caries state, and the staining risk can be estimated as the staining state. When information on a non-natural tooth region showing a repaired region or a discolored region is extracted, the caries risk and staining risk can be estimated to be high. Even when information on a non-natural tooth region showing a repaired region or a discolored region is not extracted, the staining risk can be estimated to be high if the total protein concentration in saliva is estimated to be high as the oral cavity state. Even when information on a non-natural tooth region is not extracted, the caries risk can be estimated to be high if the ratio of calcium components and phosphorus components in saliva is low or the amount of saliva is low as the oral cavity state. And when information on a non-natural tooth region is not extracted and the oral cavity state is estimated to be low, the total protein concentration in saliva is low, the ratio of calcium components and phosphorus components in saliva is high, or the amount of saliva is high, the caries risk and staining risk can be estimated to be low. In addition, when white spot areas, tetracycline areas, areas treated with resin or the like, pulp extraction areas, denture areas, etc. can be identified in a distinguishable manner, the information on these areas can be used as is as estimated information.

[0049] In this way, the oral cavity state can be estimated by taking into account not only information from the natural tooth area but also information from the non-natural tooth area, thereby making it possible to estimate the oral cavity state with high accuracy and to obtain more detailed information about the oral cavity state.

[0050] Furthermore, although not specifically mentioned in the above embodiment, it is possible to observe changes in the oral condition by having the estimation device 10 estimate the oral condition based on multiple dental images of the same subject and determining the differences between the estimated oral conditions. In this case, the image acquisition unit 21 acquires a first tooth image and a second tooth image captured at different timings as the tooth images of the subject. For example, the subject refers to the recommendation information for the oral care product or oral care method output as described above based on the first tooth image, and the second tooth image is captured after a predetermined period of time (one week, one month, etc.) has elapsed. The first and second tooth images may be captured before and after changing to a new oral care product, or before and after receiving oral care advice from a dentist or the like.

[0051] The image analysis unit 23 extracts information on the natural tooth region for each of the first and second tooth images acquired by the image acquisition unit 21. The estimation unit 24 estimates the first and second oral cavity states of the person being evaluated for the first and second oral cavity images based on the information on the natural tooth region extracted by the image analysis unit 23. The output processing unit 25 generates difference information between the first and second oral cavity states of the person being evaluated, and displays it on the display device 15. The output form of the difference information is not limited to display.

[0052] According to this modification, the subject can understand the change in his / her own oral condition by referring to the difference information. In addition, the subject can also understand the effects of the recommended information output by the estimation device 10, the new oral care product that he / she started to use, the oral care method that he / she started to use based on the advice, etc.

[0053] Some or all of the above-described embodiments and modifications may be specified as follows: However, the above-described embodiments and modifications are not limited to the following descriptions.

[0054] <1> An image acquisition means for acquiring a tooth image in which one or more specific teeth of the subject are captured; an image analysis means for extracting information on a natural tooth region from a tooth region representing the specific tooth in the acquired tooth image; An estimation means for estimating the oral cavity condition of the subject based on the extracted information of the natural tooth region; An oral cavity state estimation device comprising:

[0055] <2> The specific teeth are either or both of the maxillary central incisors. <1> The oral cavity state estimation device according to claim 1. <3> An information acquisition means for acquiring age information of the assessee; Further comprising: The image analysis means extracts saturation information of the natural tooth region, The estimation means estimates the oral cavity condition indicating the gingival inflammation state of the person to be evaluated based on the acquired age information and the extracted saturation information. <1> or <2> The oral cavity state estimation device according to claim 1. <4> The image analysis means extracts hue angle information of the natural tooth region, The estimation means estimates the oral cavity condition indicating the periodontal pocket depth state of the subject based on the extracted hue angle information. <1> from <3> 13. The oral cavity state estimating device according to claim 12, <5> An information acquisition means for acquiring age information of the assessee; Further comprising: The image analysis means extracts saturation information and hue angle information of the natural tooth region, The estimation means estimates a progression state of periodontal disease of the subject based on the extracted saturation information and hue angle information and the acquired age information. <1> from <4> 13. The oral cavity state estimating device according to claim 12, <6> The image analysis means extracts either or both of lightness information and chroma information of the natural tooth region, The estimation means estimates the oral cavity state indicating the saliva state of the person to be evaluated based on either one or both of the extracted lightness information or saturation information. <1> from <5> 13. The oral cavity state estimating device according to claim 12, <7> The image analysis means further extracts information on a non-natural tooth region indicating either or both of a discolored region and a restored region in the tooth region indicating the specific tooth in the acquired tooth image, The estimation means estimates the oral cavity condition further indicating either or both of a caries state and a staining state of the teeth of the person to be evaluated based on the information of the extracted non-natural tooth region and the estimated saliva state. <6> The oral cavity state estimation device according to claim 1. <8> an output processing means for outputting recommendation information of either or both of an oral care product or an oral care method to an output device based on the oral condition estimated by the estimation means; Further comprising <1> from <7> 13. The oral cavity state estimating device according to claim 12, <9> an output processing means for outputting difference information of the two oral cavity states estimated by the estimation means to an output device; Further comprising: The image acquisition means acquires a first tooth image and a second tooth image having different imaging timings as the tooth images of the subject, The image analysis means extracts information of the natural tooth region from each of the acquired first tooth image and second tooth image, The estimation means estimates a first oral cavity state and a second oral cavity state of the person to be evaluated regarding the first tooth image and the second tooth image based on the information of the extracted natural tooth region, The output processing means generates difference information between the first oral cavity state and the second oral cavity state of the subject. <1> from <8> 13. The oral cavity state estimating device according to claim 12,

[0056] <10> 1. A method for estimating oral cavity state, executed by one or more computers comprising one or more processors and a memory, comprising: the one or more processors An image acquisition step of acquiring a tooth image in which one or more specific teeth of the subject are captured; an image analysis step of extracting information on a natural tooth region excluding a discolored region and a restored region from a tooth region representing the specific tooth in the acquired tooth image; An estimation step of estimating the oral cavity condition of the subject based on the extracted information of the natural tooth region; The oral cavity state estimation method performs the above. <11> The specific teeth are either or both of the maxillary central incisors. <10> The oral cavity state estimation method according to claim 1, <12> the one or more processors: Obtaining age information of the assessee; Further execute In the image analysis step, saturation information of the natural tooth region is extracted, In the estimation step, the oral condition indicating the gingival inflammation state of the person to be evaluated is estimated based on the acquired age information and the extracted saturation information. <10> or <11> The oral cavity state estimation method according to claim 1, <13> The image analysis step includes extracting hue angle information of the natural tooth region, In the estimation step, the oral cavity condition indicating the periodontal pocket depth state of the subject is estimated based on the extracted hue angle information. <10> from <12> 13. The method for estimating an oral cavity state according to claim 12, <14> the one or more processors: Obtaining age information of the assessee; Further execute In the image analysis step, saturation information and hue angle information of the natural tooth region are extracted, In the estimation step, a progress state of periodontal disease of the subject is estimated based on the extracted saturation information and hue angle information and the acquired age information. <10> from <13> 13. The method for estimating an oral cavity state according to claim 12, <15> In the image analysis step, either or both of lightness information and chroma information of the natural tooth region are extracted, In the estimation step, the oral cavity state indicating the saliva state of the person to be evaluated is estimated based on either one or both of the extracted lightness information or saturation information. <10> from <14> 13. The method for estimating an oral cavity state according to claim 12, <16> In the image analysis step, information on a non-natural tooth region indicating either or both of a discolored region and a restored region in the tooth region indicating the specific tooth in the acquired tooth image is further extracted, In the estimation step, the oral cavity condition further indicating either or both of a caries state and a staining state of the teeth of the subject is estimated based on the information of the extracted non-natural tooth region and the estimated saliva state. <15> The oral cavity state estimation method according to claim 1, <17> the one or more processors: A step of outputting recommendation information of either or both of an oral care product or an oral care method to an output device based on the oral condition estimated in the estimation step; Run the following again <10> from <16> 13. The method for estimating an oral cavity state according to claim 12, <18> the one or more processors: an output processing step of outputting difference information of the two oral cavity conditions estimated by the estimation means to an output device; Further execute In the image acquisition step, a first tooth image and a second tooth image having different imaging timings are acquired as the tooth images of the subject, In the image analysis step, information on the natural tooth region is extracted from each of the acquired first tooth image and second tooth image, In the estimation step, a first oral cavity state and a second oral cavity state of the subject regarding the first tooth image and the second tooth image are estimated based on information of the extracted natural tooth region, In the output processing step, difference information between the first oral cavity state and the second oral cavity state of the subject is generated. <10> from <17> 13. The method for estimating an oral cavity state according to claim 12,

[0057] <19> The one or more processors <10> from <18> 2. A computer program for executing the oral cavity state estimation method according to claim 1 . <20> <19> A recording medium for recording the computer program described above.

[0058] The above content will be described in more detail below with reference to examples, but the description of the following examples does not limit the above content in any way. EXAMPLES

[0059] In this example, 465 dental images were used, taken of 465 subjects aged between 20 and 60. These dental images were taken by placing a lip hook in the mouth of each subject so that multiple upper and lower teeth were captured, and were taken as non-reflective light images using a photography technique that uses a polarizing plate. Then, information on the natural tooth region was extracted from each of these 465 tooth images using the above-mentioned method. In identifying the natural tooth region, the natural tooth region in the region of the two upper central incisors was designated by the user's operation using the input device 16 for the tooth image displayed on the display device 15. As information on the natural tooth region, saturation and hue angle were calculated.

[0060] Meanwhile, information on the actual ages of 465 subjects was obtained, and a dentist also classified their gingival status. In the gingival status classification, subjects were classified into a healthy gingival group (hereafter referred to as group H) in which no gingival inflammation was observed and the periodontal pocket depth of all teeth was less than 4 mm, a gingival inflammation group (hereafter referred to as group G) in which gingival inflammation was observed in one or more places and the periodontal pocket depth of all teeth was less than 4 mm, a moderate periodontitis group (hereafter referred to as group P) in which gingival inflammation was observed in one or more places and the periodontal pocket depth of one or more places was 4 mm or more but less than 6 mm, and a severe periodontitis group (hereafter referred to as group SP) in which gingival inflammation was observed in one or more places and the periodontal pocket depth of one or more places was 6 mm or more. As a result, of the 465 people evaluated, 5 were judged to be in the H group, 188 were judged to be in the G group, 211 were judged to be in the P group, and 61 were judged to be in the SP group.

[0061] FIG. 6 is a graph showing correlation equations between age and saturation according to classification of gingival condition. For each gum condition classification, a correlation analysis was conducted between the age of the subjects in each classification and the saturation of the natural tooth area. As a result, as shown in Figure 6, a positive correlation was found between age and the saturation of the natural tooth area for each gum condition classification. The lines in Figure 6 show the correlation equations for each gingival condition classification. Looking at each correlation equation, it can be seen that the correlation equations for classifications other than Group H (Group G, Group P, and Group SP) are quite similar, but the correlation equation for Group H is distinguishable from the correlation equations for the other classifications.

[0062] FIG. 7 is a graph showing a method for estimating the state of gingival inflammation using the correlation between the age of the subject and the saturation of the natural tooth area. As shown in Figure 7, in a two-axis coordinate system of age and saturation of the natural tooth area, if the correlation equation for group H shown in Figure 6 (y = 0.2422x + 4.683 (y: saturation, x: age)) is used as the boundary line, it is possible to divide the area into areas with gingival inflammation and areas without gingival inflammation. As a result, the estimation device 10 (estimation unit 24) can estimate the presence or absence of gingival inflammation in the subject by pre-storing the correlation equation for group H and plotting the age of the subject and the saturation of the natural tooth area extracted by the image analysis unit 23 on the two-axis coordinate system.

[0063] FIG. 8 is a graph showing the relationship between the hue angle of the natural tooth region and periodontal pocket depth classification. Figure 8 shows the distribution of hue angles in the natural tooth area of ​​subjects belonging to each of the three categories (less than 4 mm, 4 to 6 mm, and 6 mm or more) used as indicators of periodontal pocket depth in the gingival condition classification. There were 193 subjects in the group with a periodontal pocket depth of less than 4 mm, 211 subjects in the group with a periodontal pocket depth of 4 to 6 mm, and 61 subjects in the group with a periodontal pocket depth of 6 mm or more.

[0064] Also, Figure 8 shows that a statistically significant difference was found between the two classifications of periodontal pocket depth. Specifically, the significance probability (p value) between the group with a periodontal pocket depth of less than 4 mm and the group with a periodontal pocket depth of 4 mm to less than 6 mm was smaller than the significance level (p = 0.01), indicating that the difference between the two was highly significant (**). Furthermore, the significance probability (p value) between the group with a periodontal pocket depth of less than 4 mm and the group with a periodontal pocket depth of 6 mm or more was smaller than the significance level (p = 0.05), indicating that there was a significant difference between the two (*).

[0065] The median hue angle of the group with a periodontal pocket depth of less than 4 mm was 54.5, while the median hue angle of both the group with a periodontal pocket depth of 4 mm to less than 6 mm and the group with a periodontal pocket depth of 6 mm or more was 55.7. As a result, the estimation device 10 (estimation unit 24) can store a threshold value of 55.7 in advance, and if the hue angle of the person being evaluated is equal to or greater than that threshold, it can determine that the periodontal pocket depth is 4 mm or more, and if the hue angle is less than the threshold, it can determine that the periodontal pocket depth is not 4 mm or more.

[0066] In this embodiment, the state of gingival inflammation and periodontal pocket depth were estimated from dental images of 465 subjects using the above-mentioned method (see Figures 7 and 8), and the progression state of periodontal disease was estimated based on these estimation results as follows. Specifically, gums were presumed to be healthy if there was no gingival inflammation and no periodontal pocket depth of 4 mm or more, gums were presumed to be prone to gingivitis if there was gingival inflammation and no periodontal pocket depth of 4 mm or more, and gums were presumed to be prone to periodontitis if there was no gingival inflammation or periodontitis and a periodontal pocket depth of 4 mm or more.

[0067] FIG. 9 is a graph showing the estimation accuracy of the progress state of periodontal disease. The graph in Figure 9 shows the percentage of progression of periodontal disease estimated as described above from dental images of subjects belonging to each category: Group H (healthy gingiva group), Group G (gingivitis group), and Group P (moderate periodontitis group). As a result, the percentage of cases in which the dentist's clinical results and the estimated results of the progression of periodontal disease matched was 60% in group H, 48.9% in group G, and 32.4% in group P. On the other hand, the percentage of cases in which the clinical results showed that the gums were not healthy (group G or group P) but the gums were estimated to be healthy as the progression of periodontal disease was 16.0% in group G and 16.5% in group P. Conversely, the percentage of cases in which the clinical results showed that the gums were not healthy (group G and group P) and the gums were estimated to be not healthy as the progression of periodontal disease, that is, the percentage of cases in which the estimation of not being healthy gums matched the clinical results was 84.0% in group G and 83.5% in group P. It was thus demonstrated that the above-mentioned method can be used to estimate the progression of periodontal disease. [Explanation of symbols]

[0068] 10 Oral cavity condition estimation device (estimation device) 11 CPU 12. Memory 13 Input / Output Interface 14 Communication unit 15 Display device 16 Input Devices 21 Image acquisition unit 22 Information Acquisition Department 23 Image Analysis Unit 24 Estimation part 25 Output Processing Section

Claims

1. Image acquisition means for acquiring a tooth image in which one or more specific teeth of the person being evaluated are captured, Image analysis means for extracting information on the natural tooth region within the tooth region representing the specific tooth in the acquired tooth image, An estimation means for estimating the oral condition of the person being evaluated based on the extracted information of the natural tooth region, An oral condition estimation device equipped with the following features.

2. The aforementioned specific tooth is either one or both of the maxillary central incisors. The oral cavity condition estimation device according to claim 1.

3. Information acquisition means for acquiring the age information of the person being evaluated, Furthermore, The image analysis means extracts saturation information of the natural tooth region, The estimation means estimates the oral condition indicating the gingival inflammation state of the person being evaluated, based on the acquired age information and the extracted saturation information. The oral cavity condition estimation device according to claim 1 or 2.

4. The image analysis means extracts hue angle information of the natural tooth region, The estimation means estimates the oral condition indicating the periodontal pocket depth state of the person being evaluated based on the extracted hue angle information. The oral cavity condition estimation device according to claim 1 or 2.

5. Information acquisition means for acquiring the age information of the person being evaluated, Furthermore, The image analysis means extracts saturation information and hue angle information of the natural tooth region, The estimation means estimates the progression of periodontal disease of the person being evaluated based on the extracted saturation information and hue angle information, as well as the acquired age information. The oral cavity condition estimation device according to claim 1 or 2.

6. The image analysis means extracts either the brightness information or the saturation information of the natural tooth region, or both. The estimation means estimates the oral condition indicating the saliva condition of the person being evaluated based on either or both of the extracted brightness information or saturation information. The oral cavity condition estimation device according to claim 1 or 2.

7. The image analysis means further extracts information on non-natural tooth regions that show either a discolored region or a restored region, or both, within the tooth region representing the specific tooth in the acquired tooth image. The estimation means estimates the oral condition that further indicates either or both of the caries state or discoloration state of the person being evaluated, based on the extracted non-natural tooth region information and the estimated salivary state. The oral cavity condition estimation device according to claim 6.

8. Based on the oral condition estimated by the estimation means, an output processing means causes an output device to output recommended information for either or both oral care products or oral care methods. The oral condition estimation device according to claim 1 or 2, further comprising the following:

9. Output processing means that causes the output device to output the difference information of the two oral conditions estimated by the estimation means, Furthermore, The image acquisition means acquires a first tooth image and a second tooth image, respectively, as the tooth images of the person being evaluated, with the acquisition timings being different from each other. The image analysis means extracts information about the natural tooth region from the acquired first tooth image and second tooth image, respectively. The estimation means estimates the first and second oral conditions of the person being evaluated with respect to the first tooth image and the second tooth image, based on the extracted information of the natural tooth region. The output processing means generates difference information between the first oral state and the second oral state of the person being evaluated. The oral cavity condition estimation device according to claim 1 or 2.

10. A method for estimating oral condition, which is performed by one or more computers equipped with one or more processors and memory, The one or more processors An image acquisition process to obtain a dental image in which one or more specific teeth of the person being evaluated are captured, An image analysis step of extracting information about the natural tooth region from the tooth region representing the specific tooth in the acquired tooth image, excluding the discolored region and the restored region. An estimation step of estimating the oral condition of the person to be evaluated based on the extracted information of the natural tooth region, A method for estimating oral condition.

Images