JP2022097176A - Oral cavity state evaluation system - Google Patents

Abstract

To provide an oral cavity state evaluation system capable of suitable analysis and estimation of an oral cavity image.SOLUTION: The oral cavity state evaluation system includes an information analysis part which analyzes an oral cavity image included in an input image with first analysis means. The information analysis part classifies the oral cavity image into a first oral cavity image which can be analyzed with the first analysis means and a second oral cavity image which cannot be analyzed with the first analysis means. The first oral cavity image is analyzed with the first analysis means by the information analysis part and in a case where the oral cavity image is the second oral cavity image, the oral cavity image is analyzed with second analysis means different from the first analysis means.SELECTED DRAWING: Figure 1

Description

The present disclosure relates to an oral condition evaluation system for evaluating a user's oral condition.

There is known a system that analyzes the oral condition from an oral image including the entire oral cavity and assists a judgment by a specialist such as a dentist. Patent Document 1 discloses a dental analysis system that extracts a lesion site from an X-ray image of the entire oral cavity of a user.

Japanese Unexamined Patent Publication No. 2019-208831

The oral condition becomes more diverse for each user as the user gets older. Diversity arises, in one case, from acquired tooth loss, caries, and treatment. Oral images that reflect diverse oral conditions may not be properly analyzed and evaluated by one means.

The oral condition evaluation system according to the present invention is an oral condition evaluation system including an information analysis unit that analyzes the oral image included in the input image by the first analysis means, and the information analysis unit analyzes the oral image by the first analysis unit. The first oral cavity image that can be analyzed by the analysis means and the second oral cavity image that cannot be analyzed by the first analysis means are classified, and the first oral cavity image is the first analysis means by the information analysis unit. When the oral cavity image is the second oral cavity image, it is analyzed by a second analysis means different from the first analysis means.

According to the oral condition evaluation system according to the present invention, appropriate analysis and evaluation of oral images can be performed.

The block diagram which shows the structure of the oral condition evaluation system of embodiment. The figure which shows an example of the 1st oral cavity image. The figure which shows an example of the 2nd oral cavity image. The flowchart which shows an example of the analysis of the oral cavity image performed by the information analysis part.

(Example of possible forms of oral condition evaluation system)
(1) According to the oral condition evaluation system according to the present disclosure, the oral condition evaluation system includes an information analysis unit that analyzes the oral image included in the input image by the first analysis means, and the information analysis unit is the oral cavity. The images are classified into a first oral cavity image that can be analyzed by the first analysis means and a second oral cavity image that cannot be analyzed by the first analysis means, and the first oral cavity image is obtained by the information analysis unit. It is analyzed by the first analysis means, and when the oral image is the second oral image, it is analyzed by a second analysis means different from the first analysis means.
According to the oral condition evaluation system, the oral image is classified into a first oral image and a second oral image, the first oral image is analyzed by the first analysis means, and in the case of the second oral image, the second analysis means. Analyze with. Therefore, appropriate analysis and evaluation of oral images are performed.

(2) According to an example of the oral condition evaluation system, the first analysis means includes analysis by an analysis learning model in which an image relating to the oral cavity is learned in advance, and the analysis result of the first oral image by the first analysis means. The first analysis result is output.
According to the oral condition evaluation system, the user can appropriately recognize the oral condition from the first analysis result by the first analysis means.

(3) According to an example of the oral condition evaluation system, the second analysis means includes analysis by at least one of a specialist, an inspection device, and an inspection instrument, and the second analysis means by the second analysis means. 2 The second analysis result, which is the analysis result of the oral image, is output.
According to the oral condition evaluation system, the user can appropriately recognize the oral condition by the second analysis result by the second analysis means.

(4) According to an example of the oral condition evaluation system, at least one of the second oral image and the second analysis result is stored in a database.
According to the oral condition evaluation system, information regarding the second oral image is preferably stored.

(5) According to an example of the oral condition evaluation system, the information analysis unit executes classification of the first oral image and the second oral image by a classification learning model in which images related to the oral cavity are learned.
According to the oral condition evaluation system, the first oral image and the second oral image can be suitably classified.

(6) According to an example of the oral condition evaluation system, at least one of the analysis learning model and the classification learning model is updated based on the analysis of the second oral image.
According to the oral condition evaluation system, the learning model is updated by the information about the second oral image, so that the accuracy of the learning model is improved.

(7) According to an example of the oral condition evaluation system, an external terminal that displays the second oral image and inputs the analysis result of the second oral image is further included.
According to the oral condition evaluation system, for example, analysis by an expert can be easily performed by an external terminal different from the interface unit.

(8) According to an example of the oral condition evaluation system, the information analysis unit further classifies the input image into the oral image and the non-oral image not including the oral cavity, and the input image is the non-oral image. If so, the third analysis result is output.
According to the oral condition evaluation system, the user can easily recognize that the input image is not appropriate based on the third analysis result.

(9) According to an example of the oral condition evaluation system, an interface unit for acquiring the input image is further provided.
According to the oral condition evaluation system, the user can suitably input the input image.

(10) According to an example of the oral condition evaluation system, the interface unit includes a display unit, and the display unit has at least one of a first analysis result, a second analysis result, and a third analysis result. indicate.
According to the oral condition evaluation system, the user can easily recognize at least one of the first analysis result, the second analysis result, and the third analysis result.

(Embodiment)
The oral condition evaluation system 10 of the present embodiment will be described with reference to FIGS. 1 to 4. The oral condition evaluation system 10 is a system that acquires and analyzes input information I regarding the oral condition and outputs output information O. The main element constituting the oral condition evaluation system 10 is the server 30. Preferably, the oral condition evaluation system 10 includes an interface unit 20 for executing information exchange with the server 30. The interface unit 20 is, for example, a smart device configured to be portable by a user. Smart devices include tablet devices or smartphones. In another example, the interface unit 20 is a personal computer. The personal computer is installed in the user's residence, store or dental office. Stores include stores that sell oral care items or other products. The interface unit 20 and the server 30 are configured to be able to communicate with each other using, for example, the Internet line N. In another example, the interface unit 20 and the server 30 are integrally configured.

The interface unit 20 includes a control unit 21, a storage unit 22, an acquisition unit 23, a communication unit 24, and a display unit 25. The control unit 21 includes an arithmetic processing device that executes a predetermined control program. The arithmetic processing unit includes, for example, a CPU, GPU or MPU.

The storage unit 22 stores various control programs executed by the control unit 21 and information used for various control processes. The storage unit 22 includes, for example, a non-volatile memory and a volatile memory. The non-volatile memory includes, for example, at least one of ROM, EPROM, EEPROM, and flash memory. Volatile memory includes, for example, RAM.

The acquisition unit 23 acquires the input information I from the user. The acquisition unit 23 includes an arbitrary configuration for acquiring the input information I of the user. In one example, the input information I is a camera capable of acquiring an oral image P of the user's oral cavity as an input image. Input images include still images and moving images. The input image includes a three-dimensional image having depth information or a panoramic image composed by laminating a plurality of images. The camera may be a camera mounted on a smart device, or may be a separately independent camera. The independent camera is, for example, a pen-type camera provided with a camera at the tip suitable for photographing in the oral cavity or a camera capable of photographing in a range of 360 °. The independent camera is configured to enable wired or wireless communication with the control unit 21. The input image taken by the independent camera is transmitted to the control unit 21 by wired or wireless communication.

The communication unit 24 is configured to be able to communicate with the outside of the interface unit 20 based on the control of the control unit 21. The communication unit 24 is configured to be able to communicate via the Internet line N. The communication unit 24 may be configured to be able to communicate with the server 30 of the interface unit 20 by wired communication or wireless communication. For example, the communication unit 24 transmits the user input information I acquired by the acquisition unit 23 under the control of the control unit 21, and receives the information from the server 30.

The display unit 25 displays various information based on the control of the control unit 21. The various information is, for example, information about the user's input information I and information from the server 30. In one example, the display unit 25 is composed of a display. The display of the display unit 25 may be composed of a touch panel. When a part of the display unit 25 is composed of a touch panel, a part of the touch panel may also serve as a user interface function of the acquisition unit 23.

The user communicates with the server 30 by, for example, inputting a predetermined URL in the interface unit 20 or reading a QR code (registered trademark) in the interface unit 20. Communication with the server 30 may be started by the user selecting an icon displayed on the display unit 25.

The server 30 includes an information acquisition unit 31, an information output unit 32, an information analysis unit 33, and an information storage unit 34. The information acquisition unit 31 is configured so that information can be acquired. In one example, the information acquisition unit 31 acquires information from the communication unit 24 of the interface unit 20. The information output unit 32 is configured to be able to output information. In one example, the information output unit 32 outputs information to the communication unit 24 of the interface unit 20.

The information analysis unit 33 executes various analyzes and controls. The information analysis unit 33 includes an arithmetic processing device that executes a predetermined control program. The arithmetic processing unit includes, for example, a CPU or an MPU. The information analysis unit 33 is configured to be able to analyze the input information I from the user. In one example, the information analysis unit 33 analyzes the oral image P included in the input information I by the first analysis means. The first analysis means includes analysis by a learning model by machine learning. The learning model is, for example, a supervised learning model, which is one of the models of machine learning. In another example, the learning model is an unsupervised learning model or a reinforcement learning model.

The information storage unit 34 stores at least one of the learning model and various information. The information analysis unit 33 refers to a learning model stored in the information storage unit 34 and various information as needed.

The oral condition evaluation system 10 further includes a database 40 and an external terminal 50. The database 40 stores at least one of the second oral image P2 and the second analysis result. The database 40 is configured to be communicable with at least one of the interface unit 20, the server 30, and the external terminal 50.

The external terminal 50 is used to analyze the second oral image P2 by the second analysis means. The external terminal 50 is, for example, a smart device or a personal computer. The external terminal 50 includes a control unit 51, a storage unit 52, a communication unit 53, and a display unit 54. The control unit 51 includes an arithmetic processing device that executes a predetermined control program. The arithmetic processing unit includes, for example, a CPU, GPU or MPU.

The storage unit 52 stores various control programs executed by the control unit 51 and information used for various control processes. The storage unit 52 includes, for example, a non-volatile memory and a volatile memory. The non-volatile memory includes, for example, at least one of ROM, EPROM, EEPROM, and flash memory. Volatile memory includes, for example, RAM.

The communication unit 53 is configured to be able to communicate with an element other than the external terminal 50 based on the control of the control unit 51. In one example, the communication unit 53 is configured to be able to communicate via the Internet line N. The communication target is at least one of the interface unit 20, the server 30, and the database 40.

The display unit 54 displays information based on the control of the control unit 51. The information to be displayed is, for example, the second oral cavity image P2. In one example, the display unit 25 is composed of a display. The display of the display unit 25 may be composed of a touch panel.

The information analysis unit 33 classifies the input image into an oral image P or a non-oral image. In one example, a trained model suitable for detecting a face is used to determine whether it is an oral image P or a non-oral image depending on whether the oral region R can be detected from the input image. The oral region R includes a first oral region R1 including the teeth of the user's upper jaw and a second oral region R2 including the teeth of the user's lower jaw. The trained model for detecting the face is introduced into the information analysis unit 33 by, for example, API. The non-oral image includes, for example, an image in which the oral region R cannot be recognized because the oral region R is not included, the image is blurred, or the brightness of the image is not appropriate.

When the input image is a non-oral image, the information analysis unit 33 outputs information requesting the user to re-input the input information I including the oral image P to the display unit 25 of the interface unit 20. When one of the first oral region R1 and the second oral region R2 can be acquired and the other cannot be acquired, the information analysis unit 33 may be configured to determine that the input image is the oral image P. It may be configured to be judged as an oral image.

The oral image P includes a first oral image P1 and a second oral image P2. FIG. 2 shows an example of the first oral cavity image P1. The first oral cavity image P1 is an image that can be analyzed by the analysis learning model MA. FIG. 3 shows an example of the second oral cavity image P2. The second oral cavity image P2 is an image that is determined to be impossible to analyze by the analysis learning model MA. The second oral image P2 includes, for example, a tooth defect, involution, caries, and padding in the oral region R. In the second oral cavity image P2 shown in FIG. 3, some teeth are missing. The first oral image P1 is analyzed by the first analysis means.

The first oral image P1 includes at least the first specific site in the oral cavity. The first specific site includes a middle incisor T1, a tooth corresponding to the middle incisor T1, a side incisor T2, a tooth corresponding to the side incisor T2, a canine T3, and a tooth stem corresponding to the canine T3. The first oral image P1 includes a tooth image and a gum image in the oral cavity. Tooth images include intraoral central incisor T1 images, lateral incisor T2 images, and canine T3 images. The gingival image includes an image of the gingiva corresponding to the central incisor T1, the lateral incisor T2, and the canine T3. The first oral image P1 includes at least one first specific site among the four first specific sites on the top, bottom, left, and right in the user's oral cavity. The analysis result of the first oral cavity image P1 by the first analysis means is output as the first analysis result.

The analysis of the oral cavity image P using the learning model executed by the information analysis unit 33 will be described. The learning model includes at least an analysis learning model MA that analyzes the oral region R of the first oral image P1 and a classification learning model MC that classifies the oral image P into the first oral image P1 and the second oral image P2.

Each learning model was developed based on approximately 10,000 oral images P. The developer classified the oral image P into a learning image, a verification image, and a test image. The learning image is supervised data at the time of creating the learning model. The verification image is an image for modifying the behavior of the learning model based on the learning image. The test image is an image for finally confirming the operation of the learning model. The test image is used, for example, to confirm whether the learning model is overfitting. The learning image is about 56% of the whole. The verification image is about 24% of the whole. The test image is 20% of the total.

The first oral image P1 for creating the analytical learning model MA was supervised by an expert. Experts are, for example, dentists, dental hygienists, researchers who study oral conditions, or developers who develop oral care products. Experts who analyze the oral cavity determine the degree of crowding, the degree of gingival recession, and the degree of interdental space for at least the first oral region R1 and the second oral region R2 of the first oral image P1. gone. Experts analyzing the oral cavity made decisions based on the type of crowding in determining the degree of crowding. Experts who analyze the oral cavity made multiple stages of judgment of gingival recession, from no gingival recession to total gingival recession. Experts analyzing the oral cavity performed multiple stages of evaluation of the interdental space, from no interdental space to severe interdental space, for each of the upper and lower jaws.

The analysis learning model MA includes a plurality of trained models. The analytical learning model MA includes a first trained model M1 for determining whether or not the oral region R can be evaluated and a second trained model M2 for actually evaluating the oral region R.

The first trained model M1 is a learning model M11 that determines whether or not the oral region R of the oral image P can be analyzed, a learning model M12 that determines whether crowding evaluation is possible, and an interdental space in the upper jaw teeth. A learning model M13 for determining whether or not the void evaluation is possible, a learning model M14 for determining whether or not the interdental void evaluation in the lower jaw teeth is possible, and a learning model for determining whether or not the gingival retraction evaluation is possible. Includes at least one of M15.

The second trained model M2 is a learning model M21 for determining the presence or absence of crowding in the upper jaw teeth, a learning model M22 for determining the presence or absence of crowding in the lower jaw teeth, and a determination for the presence or absence of interdental gaps in the upper jaw teeth. It includes at least one of a learning model M23 for determining the presence or absence of interdental space in the lower jaw teeth, and a learning model M25 for determining the presence or absence of interdental retraction.

The analysis learning model MA that analyzes the first oral image P1 outputs the result of classification. The analytical learning model MA is an information on the interdental space, gingival retraction, crowding, gingival inflammation, unpolished, chewing, caries, hypersensitivity, halitosis, and the presence or absence or degree of coloring in the first specific site, in the user's oral cavity. It outputs information on whether or not periodontal disease is present, information on whether or not masticatory function is normal, information on whether or not occlusion is normal, and information on whether or not swallowing function is normal.

The classification learning model MC is created by learning the first oral image P1 in one example. In this case, the classification learning model MC determines whether the oral image P has a feature amount peculiar to the first oral image P1. In another example, the classification learning model MC is created by learning the second oral image P2. In this case, the classification learning model MC makes a judgment based on whether the oral image P has a feature amount peculiar to the second oral image P2. The classification learning model MC may be created by learning both the first oral image P1 and the second oral image P2. The classification of the first oral image P1 and the second oral image P2 of the oral image P used for learning is performed by a specialist.

When the oral image P is the second oral image P2, it is analyzed by a second analysis means different from the first analysis means. The second analysis means includes analysis means that do not depend on the learning model. In one example, the second analysis means is the analysis of the second oral image P2 by an expert. The analysis by the expert is carried out in the same manner as the supervision for the first oral image P1. The expert performs the analysis with reference to, for example, the second oral cavity image P2 displayed on the external terminal 50. The second analysis means may include analysis by at least one of the inspection device and the inspection instrument. The analysis by the inspection device is, for example, X-ray photography for the user of the oral cavity shown in the second oral cavity image P2. The analysis using the inspection instrument is, for example, an inspection using a probe for the user of the oral cavity shown in the second oral cavity image P2.

The information analysis unit 33 outputs the second analysis result and the second oral cavity image P2, which are the analysis results by the second analysis means, to the outside of the server 30. In one example, the information analysis unit 33 outputs to at least one of the database 40 and the external terminal 50.

The information analysis unit 33 outputs the output information O based on the input information I. The output information O includes at least one of the first analysis result, the second analysis result, and the third analysis result. The first analysis result and the second analysis result include information regarding the presence / absence of a predetermined state, the degree, and at least one of the occurrence probabilities. The first analysis result and the second analysis result are information on the interdental space, gingival recession, crowding, gingival inflammation, unpolished, gingival, caries, hypersensitivity, halitosis, and the presence or absence or degree of coloring in the first specific site. , Or at least one piece of information regarding the condition of the teeth, gums, and oral mucosa. Information about the condition of the teeth, gums, and oral mucosa includes, for example, at least one of tooth wear, tooth defects, presence or absence or degree of tooth fracture, complexion, dryness, and texture. The texture includes a plurality of textures including, for example, a hard texture that feels firm when touched with a finger and a soft texture that feels bumpy when touched with a finger. The first analysis result and the second analysis result include information on whether or not periodontal disease is present in the user's oral cavity, information on whether or not masticatory function is normal, information on whether or not occlusion is normal, and swallowing function. It may contain at least one piece of information as to whether it is normal or not. Information regarding the presence or absence of periodontal disease is information regarding at least one of the size, depth, and number of so-called periodontal pockets that occur between the gingiva and the tooth. The information on whether or not the masticatory function is normal is, for example, information on masticatory force, presence or absence of uneven mastication, and masticatory sound. Information on whether or not the occlusion is normal is, for example, information on the occlusal force and the engagement of the upper jaw tooth and the lower jaw tooth. The information regarding whether or not the swallowing function is normal is, for example, information regarding the presence or absence of dysphagia. The third analysis result includes information that the input input image is a non-oral image.

The analysis learning model MA is periodically updated by the second oral image P2 and the second analysis result stored in the database 40. One example is a case where a predetermined number or more of the second oral images P2 are stored in the database 40. The predetermined number of sheets is included in the range of, for example, 100 to 1000 sheets. Specifically, a predetermined number of second oral images P2 to which the second analysis result is associated are classified into a learning image, a verification image, and a test image, and the steps of learning, verification, and testing are performed. To update the analysis learning model MA.

The classification learning model MC is periodically updated by the second oral image P2 stored in the database 40 and the second analysis result. One example is a case where a predetermined number or more of the second oral images P2 are stored in the database 40. The predetermined number of sheets is included in the range of, for example, 100 to 1000 sheets. Specifically, the second oral cavity image P2 to which the second analysis result is associated is used as a learning image, a verification image, and a test image, and the steps of learning, verification, and test are executed and classified. Update the learning model MC.

An example of the analysis procedure executed by the information analysis unit 33 will be described with reference to FIG. The information analysis unit 33 executes the steps shown below periodically or every time an input image is input.

In step S11, the information analysis unit 33 determines whether or not the oral image P has been input. In the case of affirmative determination, the information analysis unit 33 executes the process of step S12. In the case of a negative determination, the information analysis unit 33 executes the process of step S17.

In step S12, the information analysis unit 33 classifies the oral image P into the first oral image P1 or the second oral image P2 by the classification learning model MC.

In step S13, the information analysis unit 33 determines whether or not the oral image P is the first oral image P1. In the case of affirmative determination, the information analysis unit 33 executes the process of step S14. In the case of a negative determination, the process of step S16 is executed.

In step S14, the information analysis unit 33 analyzes the first oral image P1 by the first analysis means. Specifically, the first oral cavity image P1 is analyzed by the analysis learning model MA. After the end of step S14, the information analysis unit 33 executes the process of step S15.

In step S15, the information analysis unit 33 outputs the first analysis result. Specifically, it is output to the interface unit 20. The control unit 21 of the interface unit 20 causes the display unit 25 to display the first analysis result.

In step S16, the information analysis unit 33 outputs the second oral image P2 to the outside. Specifically, it is output to at least one of the database 40 and the external terminal 50. The external terminal 50 causes the display unit 54 to display the second oral cavity image P2, and an expert analyzes it by the second analysis means. The expert outputs the second analysis result, which is the analysis result of the second oral cavity image P2, to at least one of the server 30 and the database 40.

In step S17, the information analysis unit 33 outputs a third analysis result including information indicating that the oral image has not been input and the non-oral image has been input. Specifically, it is output to the interface unit 20.

In step S18, the information analysis unit 33 determines whether or not the second analysis result has been received. Specifically, the information analysis unit 33 determines whether or not the second analysis result corresponding to the oral image P has been received from the external terminal 50. In the case of affirmative determination, the information analysis unit 33 executes the process of step S19. In the case of a negative determination, the information analysis unit 33 ends the analysis.

In step S19, the information analysis unit 33 outputs the second analysis result. Specifically, it is output to the interface unit 20. The control unit 21 of the interface unit 20 causes the display unit 25 to display the second analysis result.

The operation of the oral condition evaluation system 10 of the present embodiment will be described. The user inputs an input image to the acquisition unit 23 of the interface unit 20. The interface unit 20 outputs an input image to the server 30. The information analysis unit 33 of the server 30 analyzes the input image. When the input image is the first oral cavity image P1, the information analysis unit 33 outputs the first analysis result to the interface unit 20. When the input image is the second oral cavity image P2, the information analysis unit 33 outputs the input image to the outside. The output destination is the database 40 and the external terminal 50. When the input image is a non-oral image other than the oral image P, the third analysis result is output to the interface unit 20. The second oral image is analyzed by an expert via the external terminal 50. The external terminal 50 outputs the second analysis result to the server 30. The information analysis unit 33 outputs the second analysis result to the interface unit 20. The user recognizes the oral cavity state by referring to the analysis result displayed on the display unit 25 of the interface unit 20.

According to the oral condition evaluation system 10 of the present embodiment, the following effects can be further obtained.
The second analysis result by the second analysis means is output to, for example, the interface unit 20. Therefore, even in the case of the input information I that cannot be analyzed by the first analysis means, the user can easily recognize the oral cavity state by referring to the second analysis result.

At least one of the analysis learning model MA and the classification learning model MC is updated according to the second analysis result. Therefore, the accuracy of the analysis learning model MA and the classification learning model MC is improved as the second oral image and the second analysis result are stored in the database.

(Modification example)
The description of the embodiments is an example of possible forms of the oral condition evaluation system according to the present invention, and is not intended to limit the forms. In addition to the embodiments, the present invention may take, for example, a modification of the embodiment shown below and a combination of at least two modifications that do not contradict each other.

-The analysis learning model MA may be configured to output the result of regression analysis. In this case, the analysis learning model MA quantifies and outputs at least one of the degree of crowding, the degree of interdental space, and the degree of gingival recession.

-At least one of the trained model M1 and the trained model M2 may be a model trained by unsupervised learning or reinforcement learning. At least one of each learning model M11 to M15 and each learning model M21 to M25 may be a model learned by unsupervised learning or reinforcement learning.

-The analysis learning model MA may determine whether or not evaluation is possible without distinguishing between the first oral region R1 and the second oral region R2 in the oral image P. The analysis learning model MA may recognize a set of teeth and gums in the oral image P and determine whether or not the evaluation is possible. Further, it may be an analysis learning model MA that recognizes a set of teeth and gums in the oral cavity region R and executes an analysis when the majority can be evaluated.

-The analysis learning model MA may be configured to display the oral image P in a pseudo color. In one example, the region used for analysis by the learning model in the oral image P, or the region where interdental space, gingival recession, and crowding occur is displayed in red. The user can easily recognize the area used for the analysis and the area where the oral condition has a problem.

The configuration of the acquisition unit 23 is not limited to the embodiment. In one example, the acquisition unit 23 is a user interface configured so that the user can input or select characters or the like. The user input information I is at least one of information about the user's lifestyle, information about the user's oral secretions, information about the user's oral flora, answer information to a predetermined question, and information about the user's own attributes. May further be included. Information about the user's lifestyle includes information about the user's dietary content, oral care behavior, wake-up or sleep time. Information about the user's oral secretions includes information about the amount of saliva, viscosity, hydrogen ion concentration, amount of antibacterial component, and amount of component related to tooth remineralization. Information about the user's oral flora includes information about the amount and type of major bacteria present in saliva, plaque, or mucous membranes and information about the type and amount of antibodies of the major bacteria. The information on the oral secretions of the user and the information on the bacterial flora in the oral cavity of the user may be the result of an interview with the user, and are the result information of inspecting the oral secretions and the like existing in the oral cavity of the user by a predetermined means. There may be. The answer information for a given question is the user's gum condition question, preferred brushing method for oral care, number of brushes per day, time to brush each time, time to brush, and dentures. Including the presence or absence. Questions about the condition of the user's gums include, for example, the presence or absence of bleeding from the gums when brushing or eating or drinking. Information about the user's own attributes is, for example, the user's own age, gender, height, weight, dominant arm, and medical history. In another example, the acquisition unit 23 is a sensor that detects a state in the oral cavity. The sensor is, for example, a fluorescence sensor and a temperature sensor. The fluorescence sensor irradiates light of a predetermined wavelength, quantifies the amount of light, and measures the distribution and amount of a predetermined substance in the oral cavity. Predetermines are, for example, stains or plaques. The temperature sensor measures the temperature in the oral cavity. The sensor is configured to enable wired or wireless communication with the control unit 21. The information obtained by the sensor is transmitted to the control unit 21 by wired or wireless communication. The sensor may be configured to be able to further measure at least one of masticatory force, quotient force, gum blood flow, bad breath, brush pressure strength during brushing, or toothbrush movement during brushing. The information analysis unit 33 may analyze the oral image P by adding the information acquired from the acquisition unit 23.

-The input information I may be acquired via the IoT device. In one example, input information I is acquired by connecting an IoT device to an oral care item used for brushing. In one example, input information I includes information about the number of tooth brushes per day and the frequency of use of oral care items. The IoT device may transmit the input information I to the acquisition unit 23, or may transmit the input information I to the server 30.

The input information I may include a first input information I1 including the current state information of the user and a second input information I2 which is the past state information of the user. When the first input information I1 and the second input information I2 include the oral image P of the user, the oral image P included in the second input information I2 is the user's oral image P included in the second input information I2 at a predetermined time before the acquisition of the first input information I1. It is an image in the oral cavity. When the first input information I1 and the second input information I2 include the information about the oral care of the user, the information about the oral care contained in the second input information I2 is a predetermined time before the acquisition of the first input information I1. Information about oral care. The predetermined time is, in one example, one month or more. In another example, the predetermined time is one year or more. The second input information I2 may be transmitted to the server 30 via the communication unit 24 before the acquisition unit 23 acquires the first input information I1, and the second input information I2 may be transmitted to the server 30 at the same time as the transmission of the first input information I1 via the communication unit 24. May be transmitted to the server 30. In this case, the interface unit 20 further includes a storage unit that stores at least the second input information I2. The input information I may further include the third input information I3 acquired before the second input information I2 by a predetermined time. The information analysis unit 33 calculates the estimation information E using at least one of the first input information I1 to the third input information I3. The accuracy is further improved by adding the third input information I3. The interval of the predetermined time may be changed by each input information I. The input information I may further include the input information I after the fourth input information.

The database 40 may be provided in the server 30. The database 40 is provided in the information storage unit 34 in one example. Each learning model and at least one of various information may be stored in a place other than the information storage unit 34 of the server 30. In one example, it is stored in a storage unit provided in the interface unit 20. In another example, it is stored in the database 40.

The first analysis result and the second analysis result may include estimation information which is information other than the first specific site included in the oral image P. In one example, the first analysis result and the second analysis result include information about the second specific site. The second specific site comprises at least one of the premolars and molars in the oral cavity. Estimated information includes at least one piece of information about premolars and molars in the oral cavity. The state at the first specific site in the oral cavity correlates with the state at the second specific site different from the first specific site in the oral cavity. In one example, in a study of 75 women in their 40s and 70s, the gingival condition in the anterior teeth and the gingival condition in the back teeth were highly correlated. That is, when gingival recession occurs at the first specific site, gingival recession also occurs at the second specific site. The same applies to the degree of interdental space and crowding.

The first analysis result and the second analysis result contain at least one of information on the prediction of the future oral condition of the user, information on the oral care method for the user, and information on the health condition of the user affected by the oral condition of the user. It may be included. Information about the user's future oral prediction includes an estimated image showing the user's oral condition after a predetermined period of time. Information on oral care methods for the user includes information on oral care products and usage methods suitable for the user's oral condition. Information about the user's health condition affected by the user's oral condition includes, for example, information about non-oral health related to periodontal disease or oral condition.

When the learning of the analysis learning model MA and the classification learning model MC is completed, the processes of step S13, step S16, step S18, and step S19 executed by the information analysis unit 33 may be omitted. In this case, all oral images P are analyzed by the first analysis means. Therefore, the user can recognize the oral condition by referring to the first analysis result.

In place of or in addition to the external terminal 50, other terminals locally connected to the server 30 may be provided. Other terminals are used to analyze the second oral image P2 by the second analysis means. The other terminal may be a smart device or a personal computer. The interface unit 20 may analyze the second oral cavity image P2 by the second analysis means in place of or in addition to the external terminal 50.

10: Oral condition evaluation system 20: Interface unit 30: Server 31: Information acquisition unit 32: Information output unit 33: Information analysis unit 34: Information storage unit 40: Database 50: External terminal I: Input information O: Output information MA: Analytical learning model MC: Classification learning model

Claims (10)

An oral condition evaluation system including an information analysis unit that analyzes an oral image included in an input image by a first analysis means.
The information analysis unit classifies the oral image into a first oral image that can be analyzed by the first analysis means and a second oral image that cannot be analyzed by the first analysis means.
The first oral cavity image is analyzed by the information analysis unit by the first analysis means.
When the oral image is the second oral image, the oral condition evaluation system is analyzed by a second analysis means different from the first analysis means. The first analysis means includes analysis by an analysis learning model in which an image relating to the oral cavity is learned in advance.
The oral condition evaluation system according to claim 1, wherein the first analysis result, which is the analysis result of the first oral image by the first analysis means, is output. The second analysis means includes analysis by at least one of a specialist, an inspection device, and an inspection instrument.
The oral condition evaluation system according to claim 2, wherein the second analysis result, which is the analysis result of the second oral image by the second analysis means, is output. The oral condition evaluation system according to claim 3, wherein at least one of the second oral image and the second analysis result is stored in a database. The oral condition evaluation system according to any one of claims 2 to 4, wherein the information analysis unit executes classification of the first oral image and the second oral image by a classification learning model in which an image relating to the oral cavity is learned. The oral condition evaluation system according to claim 5, wherein at least one of the analysis learning model and the classification learning model is updated based on the analysis of the second oral image. The oral condition evaluation system according to claim 6, further comprising an external terminal that displays the second oral image and inputs the analysis result of the second oral image. The information analysis unit further classifies the input image into the oral cavity image and the non-oral cavity image that does not include the oral cavity.
The oral condition evaluation system according to any one of claims 1 to 7, wherein when the input image is the non-oral image, the third analysis result is output. The oral condition evaluation system according to any one of claims 1 to 8, further comprising an interface unit for acquiring the input image. The interface unit includes a display unit and has a display unit.
The oral condition evaluation system according to claim 9, wherein the display unit displays at least one of a first analysis result, a second analysis result, and a third analysis result.

Images