JP7286093B2 - Eating Related Proposal Device and Eating Related Proposing System - Google Patents

Description

The present disclosure relates to an eating-related suggestion device and an eating-related suggestion system.

2. Description of the Related Art Conventionally, there has been known a device that analyzes the behavior of an animal, including humans, using information obtained from an intraoral sensor provided in the oral cavity of the animal (see Patent Document 1, for example).

JP 2012-191994 A

However, the device described in Patent Literature 1 only analyzes the behavior of animals, and there is room for improvement in terms of feeding back the analysis results of a certain user to the user. In addition, even if the above analysis results are made available to the user, if the user does not have the necessary expertise to understand the analysis results, the user may not be able to obtain useful information from the analysis results. can't get

Accordingly, an object of the present disclosure is to provide an eating-related proposal device and an eating-related proposal system capable of appropriately providing useful information to a user.

An eating-related proposal device according to an aspect of the present disclosure includes an acquisition unit that acquires jaw motion information related to jaw motion during eating by a user, and an acquisition unit that acquires jaw motion information related to jaw motion during eating by a user; An analysis unit that acquires an analysis result, a generation unit that generates suggestion information related to eating for the user based on the analysis result, and an output unit that outputs the suggestion information generated by the generation unit.

The eating-related proposal device generates and outputs the eating-related proposal information for the user from the analysis result of the jaw movement during eating of the user. As a result, it is possible to appropriately present a user or the like with an eating-related proposal based on the analysis result. Therefore, according to the eating-related proposal device, it is possible to appropriately provide useful information to the user.

The generator may generate an improvement proposal for improving the user's eating behavior as the proposal information.

The obtaining unit may further obtain health condition data of the user, and the generating unit may generate the proposal information based on the health condition data as well.

The jaw movement information is time-series information of the jaw movement during eating by the user, and the analysis result may include attribute information indicating the attribute of the food that the user is supposed to masticate during eating. Also, the food attributes may include at least one of food size, hardness, and type. This makes it possible to present the user with suggested information according to the attributes of the food that the user is presumed to have ingested.

The analysis unit extracts mastication information, which is time-series information indicating jaw movement for a single mouthful, from the jaw movement information, inputs the mastication information into an analysis model generated by machine learning, and outputs a mouthful from the analysis model. Attributes of food presumed to be masticated by the user by jaw movements are acquired as attribute information, and the analysis model is the first information, which is time-series information indicating past jaw movements acquired for the user. and the second information indicating the attribute of the food chewed by the user by the jaw movement of the past mouthful, and machine learning may be performed using teacher data. As a result, it is possible to easily and accurately acquire analysis results (attributes of food estimated to be masticated by the user) using an analysis model that has learned features of jaw movement during eating by the user.

The analysis unit extracts mastication information, which is time-series information indicating jaw movement for a single mouthful, from the jaw movement information, inputs the mastication information into an analysis model generated by machine learning, and outputs a mouthful from the analysis model. Attributes of food presumed to be chewed by the user by jaw movements are acquired as attribute information, and the analysis model is time-series information indicating the jaw movements of past mouthfuls acquired for unspecified users. It may be a learned model machine-learned using teacher data including the first information and the second information indicating attributes of food chewed by an unspecified user by jaw movements for past mouthfuls. As a result, there is an advantage that the analysis model can be shared among a plurality of users, and there is no need to create and manage the analysis model for each user.

The analysis unit further inputs profile information indicating user attributes to the analysis model, and the analysis model is a learned model that has undergone machine learning using teacher data that further includes unspecified user profile information. good too. The user's jaw movement during eating is considered to differ depending on the user's attributes (eg, gender, age, health condition, etc.). Therefore, according to the above configuration, by including the attribute of the user in the feature amount, it is possible to obtain the attribute of the food that the user is estimated to have chewed with higher accuracy.

The analysis model includes data corresponding to the first masticatory information, which is time-series information indicating the jaw movement corresponding to the first chewing motion among the first information, and data corresponding to the second chewing motion among the first information. It is a learned model machine-learned using teacher data including the second information and data corresponding to the second mastication information, which is time-series information indicating movement, and the analysis unit extracts the first mastication information from the mastication information. and the second mastication information, and inputting the extracted information into the analysis model, the attribute of the food estimated to be masticated by the user by one mouthful of jaw movement may be obtained. In particular, the jaw movements corresponding to the first chewing motion and the second chewing motion are likely to exhibit characteristics according to the attributes of the masticated food. Therefore, according to the above configuration, by using the analysis model machine-learned using the jaw movements corresponding to the first chewing motion and the second chewing motion as feature amounts, the food estimated to be masticated by the user can be obtained. attributes can be obtained with higher accuracy.

The eating-related suggestion device may further include a display unit that displays the suggestion information output by the output unit. By displaying the proposal information on the display unit, it is possible to appropriately present the content of the proposal to the user or the like according to the analysis result.

The output unit may notify the terminal owned by the user of the proposal information. According to the above configuration, for example, when the eating-related proposal device is a server or the like on the cloud, the proposal information can be appropriately presented to the user.

The output unit may notify the proposed information to a terminal owned by another user different from the user and set in advance as a notification destination. According to the above configuration, other users can grasp the eating status of the user. As a result, for example, it is possible for the user's family/close relatives to watch over the user's eating situation.

The analysis result includes an amount of opening that indicates the size of the user's mouth that is opened when eating, and the generation unit generates proposal information including information on the necessity of diagnosing temporomandibular disorder based on the amount of opening. good. According to the above configuration, it is possible to make a proposal to prompt early diagnosis for a user suspected of having temporomandibular disorder based on the extent to which the user opens his/her mouth during eating.

An eating-related proposal system according to one aspect of the present disclosure includes a denture that is provided with a sensor for acquiring jaw movement information related to jaw movement during eating of the user and is attached to the user's lower jaw; and the eating-related proposal device for analyzing the jaw movement information. According to the eating-related proposal system, the same effects as those of the eating-related proposal device described above can be obtained.

According to the present disclosure, it is possible to provide an eating-related proposal device and an eating-related proposal system capable of appropriately providing useful information to a user.

It is a block diagram which shows the structure of the eating-related suggestion system which concerns on one Embodiment. It is a block diagram showing a schematic configuration of the denture and the base device. FIG. 4 is a diagram schematically showing an outline of an analysis model; It is a figure which shows the outline|summary of a proposal model typically. It is a figure which shows an example of the process of an analysis part and a production|generation part. FIG. 10 is a diagram showing an example of a mastication segment identified by a mastication segment identifying module; It is a figure which shows an example of mastication information for one mouthful. FIG. 4 is a diagram showing an example of sections corresponding to first and second chewing motions in mastication information for one mouthful. (A) is a graph showing simulation results when almonds are chewed. (B) is a graph showing simulation results when rice is chewed. (C) is a graph showing simulation results when chewing strawberries. It is a figure which shows an example of the processing flow of an eating-related suggestion system. It is a block diagram which shows the structure of the eating-related suggestion system which concerns on a modification. It is a figure which shows typically the outline|summary of the analysis model which concerns on a modification. It is a block diagram which shows the structure of the eating-related suggestion system which concerns on a modification.

Hereinafter, embodiments of the present disclosure will be described with reference to the drawings. In each figure, the same or corresponding parts are denoted by the same reference numerals, and redundant explanations are omitted.

[Overall configuration of eating-related proposal system]
As shown in FIG. 1, an eating-related proposal system 1 according to one embodiment includes a denture 2, a base device 3 (storage device), and a user terminal 4, which is an eating-related proposal device according to one embodiment. , a server 5; The denture 2 is a plate denture that is attached to the lower jaw of a denture wearer (hereinafter "user"). The denture 2 has a function of acquiring measurement data (jaw movement information) relating to the jaw movement of the user during eating. The base device 3 is a device that stores the denture 2 removed from the user's lower jaw. The user terminal 4 is a terminal used by the user. The user terminal 4 is, for example, a mobile terminal such as a smart phone or tablet owned by the user. Alternatively, the user terminal 4 may be a personal computer or the like installed in the user's home or the like. The user terminal 4 mainly has a function of analyzing measurement data acquired by the denture 2 and a function of generating suggested information related to eating for the user based on the analysis results. The server 5 generates an analysis model used when the user terminal 4 analyzes measurement data and a proposal model used when the user terminal 4 generates proposal information, and sends the generated analysis model and proposal model to the user terminal 4. It is a device that provides

[Configuration of denture]
As shown in FIG. 1, the denture 2 has a base portion 2a attached to the user's lower jaw and tooth portions 2b (teeth) detachably provided on the base portion. Moreover, as shown in FIG. 2, the denture 2 has a built-in sensor 20 . The sensor 20 is configured by, for example, circuits such as a processor (for example, CPU) and memory (for example, ROM and RAM) mounted on a substrate (not shown). The sensor 20 includes a sensor section 21 , a calculation section 22 , a storage section 23 , a communication section 24 , a battery 25 and a wireless charging antenna 26 . As one configuration example, the calculation unit 22 is configured by circuits such as a processor (eg, CPU) and memory (eg, ROM and RAM) mounted on a substrate (not shown), and the storage unit 23 is a memory chip or the like. It consists of an electronic storage element or the like.

The sensor unit 21 includes various sensors for detecting the movement of the upper and lower jaws of the user or the movement of the lower jaw. A sensor that detects at least one of the acceleration and angular velocity of the movement of the upper and lower jaws or the lower jaw, that is, at least one of an acceleration sensor and an angular velocity sensor can be used to detect the movement of the upper and lower jaws or the lower jaw. As the acceleration sensor, a sensor capable of measuring acceleration in three axial directions is suitable. In addition, since the movement of the mandible often involves rotation elements other than simple one-axis rotation, a sensor capable of measuring angular velocities in three-axis directions is also suitable for the angular velocity sensor. In this embodiment, the sensor unit 21 includes an acceleration sensor that detects acceleration in each of three axial directions (X-axis direction, Y-axis direction, and Z-axis direction). In this case, the sensor unit 21 continuously detects the acceleration according to the movement of the user's lower jaw (jaw movement). However, the configuration of the sensor unit 21 is not limited to the above. For example, the sensor unit 21 may include an angular velocity sensor (gyro sensor) that detects angular velocities in each of three axial directions instead of the acceleration sensor, or may include both the acceleration sensor and the angular velocity sensor.

The calculation unit 22 acquires measurement data continuously acquired by the sensor unit 21 (in this embodiment, triaxial acceleration data indicating acceleration in each of the three axial directions), and stores the measurement data in the storage unit 23. . In the present embodiment, the measurement data is time-series information indicating temporal changes in acceleration (angular velocity if the sensor unit 21 includes an angular velocity sensor) according to the user's jaw movement. The storage unit 23 is configured by, for example, the memory described above. However, the measurement data is not limited to the above example, and may be, for example, time-series information indicating changes over time in the speed of jaw movement. Moreover, the calculation unit 22 may have a time count function. For example, when the computing unit 22 is configured by a processor or the like, it can be configured to incorporate a timer/counter as many processors are equipped with. This timer/counter can be used for time allocation of time-series information.

The calculation unit 22 adds the acquisition time of the measurement data to the measurement data at each time point acquired by the sensor unit 21 by using, for example, the timer/counter described above. Then, the calculation unit 22 stores the measurement data to which the acquisition time is added in the storage unit 23 . As a result, the storage unit 23 stores time-series information in which the measurement data at each point in time are arranged in order of acquisition time.

Also, when the denture 2 is removed from the user's lower jaw and set in the storage space 34 provided in the base device 3 , the calculation unit 22 executes processing according to the control signal transmitted from the base device 3 . For example, the calculation unit 22 reads out the measurement data stored in the storage unit 23 and transmits it to the base device 3 via the communication unit 24, or deletes the measurement data stored in the storage unit 23.

Further, the calculation unit 22 may control the operation mode of the sensor unit 21 as follows. For example, the calculation unit 22 stores in advance an acceleration pattern according to the user's eating motion. Such patterns may be stored in the storage unit 23, for example. Then, the calculation unit 22 may operate the sensor unit 21 in a power saving mode with a relatively low sampling rate (that is, a relatively long sampling period) as the initial mode. In addition, the calculation unit 22 constantly monitors measurement data continuously acquired by the sensor unit 21, and detects an acceleration pattern (for example, acceleration equal to or greater than a predetermined threshold value) according to the user's eating action. The sensor unit 21 may be operated in a data acquisition mode with a higher sampling rate (that is, a shorter sampling period) than in the power saving mode, using the event as a trigger. After that, the calculation unit 22 detects the sensor unit 21 as a trigger when a predetermined set time (for example, 30 seconds or the like) has passed since the pattern of acceleration corresponding to the user's eating motion is no longer detected. The operation mode may be switched to the power saving mode again.

In this way, by setting the operation mode of the sensor unit 21 to the power saving mode during periods other than when the user is eating, the power consumption of the sensor unit 21 can be reduced. As a result, even if the user wears the denture 2 continuously for a long time, it is possible to continuously acquire the measurement data. Further, by setting the operation mode of the sensor unit 21 to the data acquisition mode when the user eats, it is possible to acquire highly accurate measurement data indicating the jaw movement of the user when the user eats.

Alternatively, the computing unit 22 stores in advance the user's scheduled meal time zone (for example, the time zone corresponding to each of breakfast, lunch, snack, and dinner), and detects the sensor during a time zone other than the scheduled meal time zone. The unit 21 may be operated in the power saving mode, and the sensor unit 21 may be operated in the data acquisition mode during the scheduled meal time period. The same effects as those described above can also be obtained by such control of the operation mode. Note that the above-described timer/counter or the like can be used for the above-described control regarding the scheduled meal time.

The communication unit 24 has a function of communicating with the base device 3 by wireless communication while the denture 2 is set in the storage space 34 of the base device 3 . The communication unit 24 is configured by, for example, an antenna for transmitting and receiving information. Wireless communication includes, for example, radio wave communication, optical communication, and sound wave communication. Radio wave communication is communication using, for example, Bluetooth (registered trademark), Wi-Fi (registered trademark), or the like. Optical communication is communication using, for example, visible light or infrared rays. Sound wave communication is communication using, for example, ultrasonic waves. Upon receiving the control signal from the base device 3 , the communication section 24 transmits the control signal to the calculation section 22 . Further, when receiving the measurement data from the calculation unit 22 , the communication unit 24 transmits the measurement data to the base device 3 .

The battery 25 is a storage battery that supplies electric power to the sensor section 21, the calculation section 22, the storage section 23, and the like. The wireless charging antenna 26 is a component for charging the battery 25 . With the denture 2 set in the storage space 34 of the base device 3 , power is supplied from the base device 3 to the battery 25 via the wireless charging antenna 26 . That is, the battery 25 is charged by wireless charging. Thus, in the denture 2 of this embodiment, the battery 25 can be charged in a non-contact manner, so there is no need to expose the circuit (charging terminal) for charging the battery 25 to the outside of the denture 2 .

The communication operation with the base device 3 by the communication unit 24 may be performed together with the charging operation of the battery 25 . As described above, in this embodiment, both the communication operation and the charging operation are performed while the denture 2 is set in the storage space 34 of the base device 3 . Here, the communication operation by the communication unit 24 consumes a certain amount of electric power regardless of which wireless communication method is adopted. can be appropriately compensated for.

At least part of each part of the sensor 20 described above (especially the sensor part 21 and the battery 25, which are frequently replaced) may be embedded in a specific tooth portion 2b. In this case, the electrical connection between the parts provided in the sensor 20, that is, the part embedded in the tooth part 2b (for example, the battery 25 embedded in the individual tooth part 2b1 of the tooth part 2b) and the other part (floor The electrical connection with the portion embedded in the portion 2a or the portion embedded in another tooth portion 2b) may be ensured, for example, at the boundary between the tooth portion 2b and the floor portion 2a. According to the above configuration, replacement work of each part provided in the sensor 20 is facilitated. For example, the tooth portion 2b in which the sensor portion 21 is embedded (for example, the individual tooth portion 2b2) is removed from the floor portion 2a, and the new tooth portion 2b (the tooth portion including the new sensor portion 21 (individual tooth portion 2b2)) is replaced with the floor portion 2a. The sensor section 21 can be easily replaced by simply attaching it to the sensor.

[Configuration of base device]
As shown in FIG. 2 , the base device 3 includes a wireless charging antenna 31 (charging section), a communication section 32 and a cleaning section 33 . The base device 3 is provided, for example, at the user's home. Focusing on the storage function of the base device 3 , the base device 3 further comprises a storage space 34 for storing the denture 2 .

The wireless charging antenna 31 wirelessly charges the battery 25 of the denture 2 while the denture 2 is set in the storage space 34 of the base device 3 . Specifically, the wireless charging antenna 31 supplies power to the battery 25 via the wireless charging antenna 26 described above. The wireless charging antenna 31 constitutes a charging section. Specifically, the wireless charging antenna 31 constitutes a charging section 31A on the base device 3 side, and the wireless charging antenna 26 constitutes a charging section 26A on the sensor 20 side. Charging section 31A and charging section 26A constitute charging section CH.

The communication unit 32 has a function of communicating with the sensor 20 of the denture 2 and the user terminal 4 . Specifically, the communication unit 32 communicates with the communication unit 24 of the denture 2 by wireless communication while the denture 2 is set in the storage space 34 of the base device 3 . Also, the communication unit 32 communicates with the user terminal 4 through wired communication or wireless communication. For example, upon receiving a control signal indicating a request for measurement data from the user terminal 4 , the communication section 32 transmits the control signal to the communication section 24 of the denture 2 set on the base device 3 . Then, upon receiving the measurement data from the communication unit 24 , the communication unit 32 transmits the measurement data to the user terminal 4 . Thereby, the measurement data acquired by the sensor 20 of the denture 2 is transmitted to the user terminal 4 via the base device 3 . Then, the user terminal 4 analyzes the measurement data and generates proposal information according to the analysis result. The communication unit 32 is a communication unit on the base device 3 side, and the communication unit 24 is a communication unit on the sensor 20 side. The communication section 32 and the communication section 24 constitute a communication section TR.

The cleaning unit 33 has a function of cleaning the denture 2 while the denture 2 is set in the storage space 34 of the base device 3 . The cleaning unit 33 performs, for example, a process of introducing a liquid containing a dedicated drug into the storage space 34, a process of cleaning the surface of the denture 2 by immersing the denture 2 in the liquid, and removing the liquid from the storage space 34 after cleaning. A process of extracting and drying the denture 2 is executed. That is, the storage space 34 also serves as a washing space in which the denture 2 can be washed.

As described above, the base device 3 stores the denture 2, charges the sensor 20, communicates with the sensor 20 (acquisition of measurement data from the sensor 20, transmission of a control signal to the calculation unit 22, etc.), and denture 2 cleaning function. As a result, the user can store the denture 2, charge the sensor 20, and receive power from the sensor 20 simply by removing the denture 2 from the lower jaw of the user before going to bed, for example, and setting the denture 2 in the storage space 34 of the base device 3. Acquisition of measurement data (and transmission to the user terminal 4) and cleaning of the denture 2 can be performed. According to such a base device 3, it is possible to acquire measurement data relating to jaw movements in the user's daily life without imposing an excessive burden on the user.

As described above, the eating-related proposal system 1 includes the denture 2 that is provided with the sensor 20 for acquiring jaw motion information related to the jaw motion when the user eats and is attached to the user's lower jaw, and the sensor 20 and an eating-related proposal device (user terminal 4) that analyzes the acquired jaw movement information. The eating-related suggestion system 1 further includes a base device 3 that stores the denture 2 removed from the user's lower jaw, the base device 3 acquires jaw movement information from the sensor 20 by communicating with the sensor 20, and It has a communication unit 32 that transmits jaw movement information to the eating-related proposal device (user terminal 4) by communicating with the eating-related proposal device (user terminal 4). In the eating-related suggestion system 1, the base device 3 further has a charging section (charging section 31A including the wireless charging antenna 31) that charges the sensor 20. FIG. The base device 3 further has a cleaning section 33 for cleaning the denture 2 .

[Configuration of user terminal]
As shown in FIG. 1, the user terminal 4 includes an acquisition unit 41, an analysis unit 42 (mastication information acquisition unit, estimation unit), an analysis model storage unit 43, a generation unit 44, and a proposed model storage unit 45. , an output unit 46 and a display unit 47 . The user terminal 4 is a computer device including a processor (such as CPU) and memory (such as ROM and RAM). It should be noted that the user terminal 4 may have a timer/counter in the same manner as the calculation unit 22 of the sensor 20 . The acquisition unit 41, the analysis unit 42, the generation unit 44, the output unit 46, and the display unit 47 may each be configured by a circuit, and the processor may configure part or all of these circuits. Each of the acquisition unit 41, the analysis unit 42, the generation unit 44, the output unit 46, and the display unit 47 may constitute part or all of the processor, and each circuit may include a memory.

Acquisition unit 41 acquires measurement data relating to jaw movement during eating by the user. As described above, the acquisition unit 41 acquires the measurement data acquired and stored by the sensor 20 of the denture 2 via the base device 3 . However, the method by which the acquisition unit 41 acquires the measurement data is not limited to the above. For example, the acquisition unit 41 may acquire measurement data by directly communicating with the sensor 20 (communication unit 24) without going through the base device 3. FIG.

In the present embodiment, the measurement data is time-series information (three-axis acceleration data) indicating temporal changes in acceleration in each of the three-axis directions detected by the sensor 20 provided on the denture 2 worn on the user's lower jaw. be. The triaxial acceleration data is information indicating the X-axis component, Y-axis component, and Z-axis component of the acceleration at each time continuously acquired in the sensor unit 21 at a predetermined sampling period.

For example, when the user sets the denture 2 in the storage space 34 of the base device 3 while the user is asleep, the acquisition unit 41 collects data for one day before the user goes to bed (that is, after the user wakes up and wears the denture 2). period until the denture 2 is removed before going to bed). Here, the acquisition unit 41 may acquire all the measurement data (that is, data for one day) stored in the storage unit 23 of the sensor 20, or may acquire the measurement data acquired by the sensor unit 21 in the data acquisition mode. Data only (ie, measurement data when the user eats) may be acquired. For example, the acquisition unit 41 transmits a control signal requesting only the measurement data acquired in the data acquisition mode to the sensor 20 (calculation unit 22) via the base device 3, and transmits the measurement data acquired in the data acquisition mode. may be obtained from the sensor 20 only. According to the latter, it is possible to efficiently acquire only data necessary for analysis by the analysis unit 42 (that is, data measured when the user eats). By omitting communication of unnecessary measurement data in this way, it is possible to reduce the amount of data communication and to acquire measurement data necessary for analysis in a short time.

The acquisition unit 41 may be configured to further acquire data other than the measurement data of jaw movements during eating of the user. For example, the acquisition unit 41 may be configured to acquire health condition data of the user (one of attributes of the user), which will be described later. In this case, it is possible to grasp as data that the jaw movement indicated in the measurement data was performed during the user's disease state. Hereinafter, data indicating the user's health condition, etc. (that is, data that may be related to eating but is not measurement data of jaw movements during eating) is referred to as indirect eating-related data. write. Indirect feeding-related data may be entered by the user or by someone other than the user. Indirect eating-related data may also be acquired in conjunction with equipment, devices, etc. for measuring the user's health condition. For example, the acquisition unit 41 may be linked with a thermometer or a sphygmomanometer. That is, when the user measures body temperature or blood pressure using a thermometer or a sphygmomanometer, the acquisition unit 41 automatically acquires the result of the measurement from the thermometer or the sphygmomanometer as indirect eating-related data. good too. In addition, the acquisition unit 41 may work in conjunction with a user observation device that automatically analyzes the user's facial expression, body movement, etc., and acquires measurement results regarding the user's mental state from the user observation device as indirect eating-related data. good.

Based on the measurement data acquired by the acquisition unit 41, the analysis unit 42 acquires an analysis result of the eating state of the user at the time of eating. In this embodiment, the analysis result includes attribute information indicating the attribute of the food estimated to be masticated (ingested) by the user at the time of eating. Food attributes include at least one of size, hardness, and type of food. The type of food is information specifying food such as almonds, rice (rice), and strawberries. In this embodiment, the analysis results include attribute information including all of the size, hardness, and type of food that the user is estimated to have masticated, an estimated value of the user's masticatory muscle activity during eating, and including. In addition to the above, the analysis results may include various information obtained by analyzing measurement data using, for example, statistical methods (e.g., number of chews per bite, bite force per bite, posture per bite (information such as head posture when eating)). The estimated food attribute information may be information including at least hardness. Although the details will be described later, the hardness attribute of the object to be masticated is preferably estimated by detecting the acceleration and/or angular velocity of the movement of the mandible during the first chewing motion and the second chewing motion of the mastication object. . By estimating the hardness of the object to be masticated, it becomes easier to make suggestions for improving eating habits, including recommending the intake of harder foods.

The analysis unit 42 uses the analysis model (analysis model M1) stored in the analysis model storage unit 43 to obtain part or all of the analysis results from the measurement data. In this embodiment, the analysis unit 42 extracts mastication information (time-series data D1 and D2 shown in FIGS. 7 and 8), which is time-series information indicating the jaw movement of the user for one mouthful, from the measurement data. Then, the analysis unit 42 inputs mastication information for one mouthful into the analysis model, and acquires information output from the analysis model as part of the analysis result. In this embodiment, the information output from the analysis model indicates the attributes of the food (in this embodiment, the size, hardness, and type of food) estimated to be chewed by the user through jaw movements for a single mouthful. Contains attribute information and estimates of masticatory muscle activity for a single mouthful. Details of such an analysis model will be described later.

Here, the jaw movement for one bite means that the user opens the mouth to put a mouthful of food in the oral cavity, puts the food in the mouth, and then opens the mouth again to put the next food in the oral cavity. It is jaw movement during the period until it opens. Alternatively, the user usually opens the mouth to put a mouthful of food in the oral cavity, puts the food in the mouth, chews, swallows, and then opens the mouth again to put the next food in the oral cavity. Therefore, the jaw movement for a mouthful may be jaw movement during the period from opening the mouth to put a mouthful of food into the oral cavity until the end of chewing. That is, the jaw movement for a mouthful is jaw movement during the period of mastication of the mouthful of food without putting new food into the mouth after putting a mouthful of food into the oral cavity. Jaw movement for one mouthful is performed in the oral cavity between the opening of the mouth to chew one mouthful of food and the opening of the mouth to chew the next food during one meal. It may be considered as a jaw movement of a period including at least the period from the start to the end of mastication of a mouthful of food. A single meal can be defined as above, as it usually consists of multiple chewings of a mouthful of food. It is also possible to determine that mastication is completed when a certain period of time elapses after the end of mastication in a state in which there is no jaw movement, so that the jaw movement information for the last mouthful can be processed. In addition, the jaw movement for a single mouthful is defined as the mastication period from opening the mouth to masticate a single mouthful of food to swallowing the mouthful of food placed in the oral cavity without adding new food to the mouth. can be thought of as a jaw movement.

The generation unit 44 generates suggestion information related to eating for the user based on the analysis result acquired by the analysis unit 42 . The suggested information includes, for example, advice on meal content (e.g., alerting people to chew less often, not eating rice, etc.), evaluation of recent dietary trends (evaluation compared to standard eating habits), and eating habits. (e.g., chewing more often, eating at regular times, eating longer, eating less hard foods recently positive proposals such as increasing the frequency of ingestion of The proposal information may be, for example, one template selected from a plurality of prepared templates (for example, data containing text according to the contents of the proposal), or a combination of several templates (i.e., It may be a composite proposal content). Advice on meal content or advice for improving eating habits can be said to be an eating behavior improvement proposal (direct eating behavior improvement proposal) that directly proposes better eating behavior to the user. An evaluation of recent dietary trends also has the effect of motivating people to improve their eating habits when the evaluation is poor, so proposals to improve eating behavior indirectly encourage the user to take better eating behaviors (indirect proposals to improve eating behavior). You can say that.

The generating unit 44 uses the proposed model (proposed model M2) stored in the proposed model storage unit 45 to extract part or all of the above-described proposal information from part or all of the analysis results obtained by the analysis unit 42. Generate (obtain). As an example, the generation unit 44 inputs the analysis result corresponding to the predetermined period to the proposal model, and acquires information output from the proposal model as part of the above-described proposal information. Here, the predetermined period is, for example, a meal period for one meal, a meal period for one day (that is, a period including breakfast, lunch, snack, and dinner), or the like. The details of such a proposed model will be described later.

The output unit 46 outputs the proposal information generated by the generation unit 44 . In this embodiment, the output unit 46 outputs the proposal information to the display unit 47 . Specifically, the output unit 46 generates display data for displaying the contents of the proposal information on the screen, and outputs the display data to the display unit 47 .

The display unit 47 displays proposal information. The display unit 47 is configured by, for example, a display device included in the user terminal 4 . The display unit 47 presents the content of the proposal information to the user by displaying the display data acquired from the output unit 46 . Thereby, the user can confirm the content of the proposal according to the analysis result of the measurement data acquired by the sensor 20 of the denture 2 .

A mechanical configuration example of the acquisition unit 41, the analysis unit 42, the analysis model teaching data storage unit 52, the analysis model storage unit 43, the generation unit 44, and the output unit 46 will be described below. The acquisition unit 41 is composed of an interface for acquiring measurement data. When processing the measurement data, the acquisition unit 41 consists of an interface and a circuit. The circuitry may be part or all of the processor. The analysis unit 42 is composed of a program for analysis and a circuit for executing the program. The circuitry may be part or all of the processor. The program that performs the analysis consists of electrical signals stored in memory. The analytical model teaching data storage unit 52 is composed of a memory. The analysis model storage unit 43 consists of memory. The generating unit 44 includes a program for generating and a circuit for executing the program. The circuitry may be part or all of the processor. The generating program consists of electrical signals stored in memory. The output unit 46 includes a program for outputting and a circuit for executing the program. The circuitry may be part or all of the processor. The program that produces the output consists of electrical signals stored in memory.

[Server configuration]
As shown in FIG. 1, the server 5 includes an analysis model teacher data generation unit 51, an analysis model teacher data storage unit 52, and an analysis model learning unit 53 as functional elements for generating the above-described analysis model. And prepare. The server 5 also includes a proposed model teacher data generation unit 54, a proposed model teacher data storage unit 55, and a proposed model learning unit 56 as functional elements for generating the above-described proposed model. The server 5 is a computer device including a processor (such as a CPU) and memory (such as a ROM and a RAM). The server 5 may be composed of one or more devices provided on the cloud, for example.

The analytic model teaching data generation unit 51 generates teaching data for learning the analytic model. The analytical model teaching data storage unit 52 stores the teaching data generated by the analytical model teaching data generating unit 51 . The analytic model learning unit 53 generates an analytic model, which is a learned model, by executing machine learning using the teacher data stored in the analytic model teacher data storage unit 52 .

FIG. 3 is a diagram schematically showing an overview of the analysis model. As shown in FIG. 3 , first, the analysis model teaching data generating unit 51 generates mastication information (first information), which is chronological information indicating jaw movements for a past mouthful acquired about the user, and the past mastication information. A plurality of sets of teacher data are generated including information on the user's actual meal content corresponding to the jaw movement for one mouthful. The information on the content of the user's actual meal is information including the attribute (second information) of the food that the user has actually masticated through the jaw movement for the past mouthful. In this embodiment, the information on the actual meal content also includes the amount of masticatory muscle activity corresponding to the jaw movement for the past mouthful. This amount of mastication muscle activity was measured when the user actually had a meal. In addition to the above, the information on the actual meal content may also include the results of measuring the occlusion force during the eating motion, the posture of the head, and the like.

Such training data can be obtained, for example, by having the user eat various foods and correlating chewing information for each mouthful extracted from measurement data at that time with the user's actual meal content. However, the method of acquiring teacher data is not limited to the above. The teacher data may be acquired by, for example, having the user wear smart glasses capable of recording an image of the user's mouth, and logging the user's actual eating behavior with the smart glasses. Specifically, mastication information for each mouthful during a meal is acquired from the measurement data acquired by the sensor 20 of the denture 2, and the user's actual meal content is grasped from the image acquired by the smart glasses. Then, by associating the user's actual meal content and mastication information grasped from the image acquired by the smart glasses for each mouthful, the above-described teacher data can be obtained. The teacher data generated in this manner is accumulated in the analytical model teacher data storage unit 52 .

Subsequently, the analytic model learning unit 53 generates an analytic model M1, which is a learned model, by executing machine learning using the teacher data accumulated in the analytic model teacher data storage unit 52 . The machine learning method executed by the analysis model learning unit 53 is not limited to a specific method, and various methods such as SVM, neural network, and deep learning can be used, for example. For example, when the analytical model M1 is configured by a neural network, a trained model in which the intermediate layer parameters of the neural network are tuned by teacher data is obtained as the analytical model M1.

The analysis model M1 obtained by such machine learning is a learning system configured to input one mouthful of mastication information of the user and output an estimation result of the user's meal content corresponding to the one mouthful of mastication information. It is a finished model. Here, the estimation result of the content of the user's meal includes the estimation result of food attributes such as the size, hardness, and type of food for one mouthful, and the estimated value of masticatory muscle activity for one mouthful. . The analytical model M<b>1 generated by the analytical model learning unit 53 is provided from the server 5 to the user terminal 4 and stored in the analytical model storage unit 43 of the user terminal 4 .

The proposed model teaching data generating unit 54 generates teaching data for learning the proposed model. The proposed model teaching data storage unit 55 stores the teaching data generated by the proposed model teaching data generating unit 54 . The proposed model learning unit 56 generates a proposed model, which is a trained model, by executing machine learning using the teacher data stored in the proposed model teacher data storage unit 55 .

FIG. 4 is a diagram schematically showing an overview of the proposed model. As shown in FIG. 4, first, the proposed model teaching data generating unit 54 includes analysis results corresponding to the past predetermined period acquired for the user, and proposal information regarding eating according to the analysis results. Generate multiple sets of training data. Here, as an example, the predetermined period is a meal period for one meal (for example, breakfast). Further, the analysis results corresponding to the predetermined period are the attribute information (size, hardness, and type) of the food chewed by the user during the predetermined period, the amount of masticatory muscle activity during the predetermined period, the total number of times of mastication during the predetermined period, the predetermined It may include biting force in a period, head posture during eating motion in a predetermined period, and the like. The occlusal force may be actually measured, but may also be analyzed from standard average values based on, for example, sex, age, build, and the like. That is, the analysis result corresponding to the predetermined period input to the proposed model is the information obtained using the above-described analysis model M1 (attribute information of the food presumed to be masticated by the user, estimated masticatory muscle activity, etc.). ) as well as information obtained by a statistical method or the like without using the analysis model M1. Note that the posture of the head may be measured based on the detection result of the sensor section 21 .

The suggestion information regarding eating included in the teacher data may be the above-mentioned eating behavior improvement suggestion. Alternatively, the suggestion information may be, for example, suggestion information according to the user's health condition (for example, cold, stomach ache, headache, etc.) during the predetermined period in the past. The teacher data generated by the proposed model teacher data generation unit 54 is accumulated in the proposed model teacher data storage unit 55 . The user's health condition is considered to be reflected in the user's eating behavior. Although it is considered that there are individual differences for each user, for example, when a user is suffering from a cold, stomach ache, etc., the amount of food eaten is less than usual, the food taken is uneven, and the bite force is reduced. It may be smaller than usual. Also, when the user is suffering from a headache, the head posture may be worse than usual (for example, slanted). A user's state of health may also include a user's mental state in a broad sense. This is because you may eat less when you're feeling down, or eat more when you're feeling happier than usual. Thus, it can be said that there is a certain correlation between the user's health condition and the user's eating behavior. For this reason, it can be said that there is a certain correlation between suggested information according to the user's health condition (that is, suggested information considered useful for the user based on the user's health condition) and the user's eating behavior. . Therefore, according to the above-described machine learning based on teacher data, a proposal configured to input an analysis result corresponding to a predetermined period and output proposal information according to the user's health condition estimated from the analysis result A model can be generated. Examples of suggestions include, when a disease is suspected, presenting the cause of the disease to the user and proposing an appropriate menu, and suggestions related to healing support such as recommending a doctor to see a doctor. A recommendation to see a doctor or the like is a proposal (health improvement proposal) related to eating to direct the user toward a better state of health.

Subsequently, the proposed model learning unit 56 generates the proposed model M2, which is a trained model, by executing machine learning using the teacher data accumulated in the proposed model teacher data storage unit 55 . The machine learning technique executed by the proposed model learning unit 56 is not limited to a specific technique, and various techniques such as SVM, neural network, and deep learning can be used, for example. For example, when the proposed model M2 is composed of a neural network, a trained model in which the intermediate layer parameters of the neural network are tuned by teacher data is obtained as the proposed model M2.

The proposed model M2 obtained by such machine learning is a trained model configured to input the analysis result corresponding to the user's predetermined period and to output the proposal information corresponding to the analysis result. The proposed model M<b>2 generated by the proposed model learning unit 56 is provided from the server 5 to the user terminal 4 and stored in the proposed model storage unit 45 of the user terminal 4 .

Mechanical configuration examples of the analytical model teacher data generator 51, the analytical model learner 53, the proposed model teacher data generator 54, the proposed model teacher data storage 55, and the proposed model learner 56 will be described below. The analytic model teaching data generating unit 51 includes a program for generating analytic model teaching data and a circuit for executing the program. The circuitry may be part or all of the processor. A program for generating analysis model training data consists of electrical signals stored in a memory. The analytical model learning unit 53 is composed of a program for performing analytical model learning and a circuit for executing the program. The circuitry may be part or all of the processor. A program for analytic model learning consists of electrical signals stored in memory. The proposed model teaching data generation unit 54 includes a program for generating proposed model teaching data and a circuit for executing the program. The circuitry may be part or all of the processor. A program for generating training data for the proposed model consists of electrical signals stored in memory. The proposed model teaching data storage unit 55 is composed of a memory. The proposed model learning unit 56 includes a program for performing proposed model learning and a circuit for executing the program. The circuitry may be part or all of the processor. A program for performing proposed model learning consists of electrical signals stored in memory.

[Processing of analysis unit and generation unit]
An example of the processing of the analysis unit 42 and the generation unit 44 will be described with reference to FIGS. 5 to 8. FIG. FIG. 5 is a diagram showing an example of a processing flow up to generating proposal information from measurement data (measurement data acquired by the acquisition unit 41) during eating. As shown in FIG. 5, in the present embodiment, the analysis unit 42 includes a mastication segment identification module 42a, a filtering/calculation module 42b, and an analysis module 42c. The generator 44 also includes a proposal module 44a. These modules are composed of programs and the like for executing various processes. In FIG. 5, an analysis model M1 is shown in the analysis module 42c. When the analysis model M1 is stored in the memory of the analysis module 42c, and when accessing the analysis model M1 in the analysis model storage unit 43, the analysis model M1 is transferred from the analysis model storage unit 43 to the analysis module 42c. It shows a configuration in which the analysis model M1 is used in some form, such as when it is loaded into an internal memory.

[Mastication section identification module]
First, the analysis unit 42 uses the mastication segment identification module 42a to determine the amount of dentures for a certain period of time (for example, one day from when the user wears the dentures 2 after waking up to when they remove the dentures 2 before going to bed. The period from wearing the denture 2 to removing the denture 2 may be called a denture wearing period. Specifically, the mastication section identification module 42a identifies a mastication section for each mouthful included in the measurement data (that is, a section in which the user puts a mouthful of food in his mouth and performs a mastication action). During the chewing interval for each bite, multiple chewing motions (chewing motions) are included to break up the food in the cavity. Then, the mastication section identification module 42a extracts the measurement data corresponding to each identified mastication section as data for one mouthful.

FIG. 6 is a diagram showing an example of the mastication segment CS identified by the mastication segment identification module 42a. X, Y, and Z in the figure indicate the acceleration in the X-axis direction, the Y-axis direction, and the Z-axis direction, respectively. Here, it is known that a certain user's jaw movement pattern during mastication is almost constant regardless of the type of food. By utilizing this fact, the mastication segment identification module 42a identifies the mastication segment, for example, as follows.

First, the mastication section identification module 42a acquires in advance the period of jaw movement during mastication by the user (hereinafter referred to as "jaw movement period"). For example, the jaw movement period can be obtained (calculated) based on a preliminary measurement (eg, video observation of the user's chewing movement, etc.). Further, for example, if there is a section in the measurement data where it is known that the user is performing a chewing motion, by extracting the periodic motion component observed in that section, The period can be obtained as the jaw motion period. During the jaw movement cycle, in addition to the movement that actually moves the jaw, a maintenance movement that does not move the jaw may be included. In the following, the motion of moving the jaw during the jaw motion cycle is referred to as positive jaw motion, and the motion of maintaining the jaw without moving is referred to as negative jaw motion. Periodic jaw movements can consist of positive jaw movements or a combination of positive and passive jaw movements. Thus, the aforementioned mouthful jaw movements may consist of positive jaw movements or a combination of positive and negative jaw movements.

Subsequently, the mastication segment identification module 42a identifies, as mastication segments, a segment in which a periodic motion component with a period matching or similar to the jaw motion cycle is observed from the measurement data. For example, when the mastication segment identification module 42a detects a segment in which the periodic motion component continues for a predetermined threshold time (for example, several seconds) or longer, the mastication segment identification module 42a identifies the detected segment as a mastication segment. Then, the measurement data corresponding to the mastication section specified in this way is extracted as data for one mouthful. By such processing of the mastication segment identification module 42a, as shown in FIG. 6, the mastication segment CS is identified from the measurement data (three-axis acceleration data in this case). In this way, the mastication section CS for each mouthful is specified for each meal period of the user (for example, period of breakfast, lunch, snack, dinner, etc.) included in the measurement data. That is, a plurality of mastication intervals CS are specified for each meal period included in the measurement data. Then, the measurement data included in each mastication section CS is extracted as triaxial acceleration data for each mouthful. The mastication section identification module 42a may extract jaw movements for mastication from among the user's jaw movements, and identify the initiation stage of mastication. Further, the mastication segment identification module 42a may determine that mastication is completed when a certain period of time elapses after the end of mastication with no jaw movement. Then, the mastication section specifying module 42a may separate the start and end of one meal, set the interval between one meal as a meal section, and specify the mastication section CS in the meal section.

[Filtering/calculation module]
Next, the analysis unit 42 uses the filtering/calculation module 42b to remove the effects of noise components and gravitational acceleration contained in the three-axis acceleration data for one mouthful. In this embodiment, the filtering/calculation module 42b performs, for example, a low-pass filter to remove high-frequency components contained in the three-axis acceleration data for one mouthful, and performs offset correction on the three-axis acceleration data for one mouthful. Offset correction is processing for removing a signal component due to gravitational acceleration acquired by an acceleration sensor. Such processing for removing or reducing noise components that may adversely affect analysis may be referred to as “adverse component countermeasure processing”. The filtering/calculation module 42b may be called an "adverse effect component countering module" that performs adverse effect component countering processing.

As shown in FIG. 7, the filtering/calculation module 42b acquires time-series data D1 indicating the norm of acceleration based on the three-axis acceleration data for one mouthful after noise removal and offset correction. Subsequently, the filtering/calculation module 42b subtracts the gravitational acceleration (1 g) from the acceleration norm and integrates the subtraction result. Thereby, time-series data D2 is obtained. The time-series data D2 is waveform data similar to the user's jaw movement (angular velocity of the lower jaw).

The time-series data D1 and D2 are used as mastication information for one mouthful input to the analysis model M1. That is, the analysis unit 42 extracts the time-series data D1, which is mastication information indicating the user's jaw movement for one mouthful, from the measurement data acquired by the acquisition unit 41 by the mastication segment identification module 42a and the filtering/calculation module 42b. , D2. As described above, the mastication information (time-series data D1, D2) is extracted for each mastication segment CS identified by the mastication segment identification module 42a.

From the time-series data D1 and D2, information on the size of the user's mouth opening (that is, the size of opening the lower jaw with respect to the upper jaw for chewing food) and bite movement (biting movement) can be grasped. For example, from the characteristics of the waveforms of the time-series data D1 and D2, it can be seen that the user opens the mouth wide (that is, separates the lower jaw from the upper jaw), closes the mouth, chews food, and then swallows. , the interval until the mouth is opened wide again to put the next mouthful of food in the mouth can be extracted. This section is a section in which one chewing motion (in this embodiment, the mouth is once opened, the food is chewed, and the mouth is opened again for the next chewing motion).

FIG. 8 is a diagram showing an interval B1 corresponding to the first biting motion and an interval B2 corresponding to the second biting motion extracted from the time-series data D1 and D2. The occlusal movement of the first chewing motion (data in interval B1) and the occlusal motion of the second chewing motion (data in interval B2) are specified from the type (hardness, cohesiveness, etc.) of the food (masticated material). type of food eaten), size, and masticatory muscle activity. In other words, the data in the sections B1 and B2 are portions in which features tend to appear according to the type and size of the food and the amount of masticatory muscle activity.

FIG. 9 shows an example of data obtained by the sensor 20 of the denture 2 when the denture 2 is attached to a robot that simulates the jaw movement during mastication of an actual person. Here, the sensor unit 21 of the sensor 20 includes an angular velocity sensor for detecting the angular velocity of rotational motion of the mandible around the axis connecting the reference points of the mandible in addition to the acceleration sensors in the three axial directions. A specific example of such an angular velocity sensor is an angular velocity sensor that detects the angular velocity of the rotational movement of the mandible about the axis connecting the kinematic condyle points of the left and right mandibular condyles of the mandible. As the angular velocity sensor, a triaxial angular velocity sensor is preferably used. When a robot is used, it is easy to set which point to use as the reference point. The time-series data D3 is data obtained by integrating the norms of angular velocities obtained from data obtained by angular velocity sensors in three axial directions. That is, the time-series data D3 is data corresponding to the angle of the lower jaw with respect to the upper jaw at each time point. The time-series data D4 is data corresponding to the above-described time-series data D2 (that is, data obtained by integrating the result of subtracting the gravitational acceleration from the acceleration norm). In the graph shown in FIG. 9, the smaller the value of the time-series data D3 (the lower the value), the larger the opening of the mouth (opening of the lower jaw with respect to the upper jaw). In the examples of FIGS. 5 to 8, the analysis unit 42 acquires mastication information for one mouthful (here, time-series data D1 and D2) only from the three-axis acceleration data, but the sensor unit 21 acquires angular velocity in three-axis directions. When a sensor is provided, the analysis unit 42 may acquire data corresponding to the time-series data D3 as mastication information for one mouthful.

FIG. 9A is a graph showing simulation results when almonds are chewed. (B) of FIG. 9 is a graph showing a simulation result when chewing a mouthful of rice (rice). (C) of FIG. 9 is a graph showing a simulation result when chewing a strawberry. Here, the physical characteristics of each food are as follows.
• Almonds: Relatively small and relatively hard.
・Rice: Relatively small and relatively soft.
・Strawberries: Relatively large and relatively soft.

As shown in FIG. 9, in the overall waveforms of the time-series data D3 and D4, differences between foods (that is, differences according to the differences in physical characteristics) appear. In particular, the difference between the foods is conspicuous in the portions corresponding to the first chewing motion (section B1) and the second chewing motion (section B2) in the time-series data D3 and D4. For example, in the second chewing motion of almonds, once the jaw stops moving, it is characterized by further chewing motion. In addition, the number of times of chewing per unit time (the number of times of chewing motion) is larger for highly sticky rice than for the other two. In addition, for strawberries, the size of the opening in the second and subsequent chewing motions is relatively small. This is presumed to be due to eating while moving the bolus with the tongue.

In this way, it can be seen that the user's mastication information for a single mouthful has characteristics corresponding to the physical characteristics of the food to be masticated. Although it is considered that there are individual differences in jaw movements during mastication, the analysis model M1 can be trained using a plurality of teacher data (correct data) obtained from the user's past meals as described above. , it is possible to obtain an analysis model M1 corresponding to the characteristics of jaw movements during mastication unique to the user.

The input data for the analysis model M1 includes time-series information (first mastication information) indicating the jaw movement corresponding to the first chewing action and time-series information (second chewing information) indicating the jaw movement corresponding to the second chewing action. information) may be included. The first mastication information is data included in section B1, and the second mastication information is data included in section B2. That is, the data included in the interval B1 is explicitly input to the analysis model M1 as data corresponding to the first biting motion, and the data included in the interval B2 is explicitly input to the second biting motion. It may be entered as corresponding data. In this case, the above-described analysis model teaching data generating unit 51 generates data corresponding to the first chewing motion (that is, information corresponding to the first ) and data corresponding to the second chewing motion (that is, information corresponding to the second mastication information), and these extracted data are included in the training data for the analysis model. Thus, preferably, the teacher data includes at least data corresponding to the first biting motion and data corresponding to the second biting motion. In this case, the analysis model M1 outputs analysis results for the first mastication information and the second mastication information (for example, attribute information of food that the user is presumed to have chewed, an estimated value of masticatory muscle activity for one mouthful, etc.). is learned as The data corresponding to the first chewing motion may be called "first chewing information analysis training data", and the data corresponding to the second chewing motion may be called "second chewing information analysis training data".

[Analysis module]
Next, the analysis module 42c of the analysis unit 42 acquires the above-described analysis result from the mastication information for one mouthful. The analysis unit 42 inputs mastication information for one mouthful into the analysis model M1 stored in the analysis model storage unit 43 . As described above, the mastication information for a single mouthful input to the analysis model M1 may include only the time-series data D1 and D2 in the mastication section CS (hereinafter referred to as "whole section data"). Among them, only the time-series data D1 and D2 included in the intervals B1 and B2 (hereinafter “specific interval data”) may be included, or both the entire interval data and the specific interval data may be included. Further, the mastication information for one mouthful input to the analysis model M1 may include only one of the time-series data D1 and D2. Further, as described above, when the mastication information for one mouthful includes time-series data corresponding to the time-series data D3 (that is, data indicating the angle of the mandible), it is input to the analysis model M1. The chronological data may be included in the mastication information for one mouthful. The reason why the first and second chewing motions are particularly noted is that the first and second chewing motions clearly show the characteristics of the object to be masticated. You may pay attention to the biting motion up to a plurality of times (n-th time). That is, mastication information corresponding to the first to n-th chewing motions may be used as input data for the analysis model M1. Needless to say, the n number of times is set to be smaller than the total number of chewing motions during the mastication section CS.

Then, the analysis unit 42 uses the analysis module 42c to determine the attributes of the food estimated to be masticated by the user by the jaw movement for one mouthful (in this embodiment, the size and hardness of the food), which are output from the analysis model M1. , and type) are acquired as the attribute information described above. That is, the analysis unit 42 (analysis module 42c) performs attribute estimation using the analysis model M1. The attribute information of the food output from the analysis model M1 may be called "estimated attribute information", and the attribute of the food estimated to be masticated may be called "estimated attribute". Further, as in the present embodiment, the analysis model M1 may output an estimated value of the amount of masticatory muscle activity for one mouthful along with the attribute of the food. Also, the analysis model M1 may be configured to output information other than the above.

Further, the analysis unit 42 may acquire information obtained without using the analysis model M1 as part of the analysis result. Information obtained without using the analysis model M1 is, for example, information such as bite force for each bite and posture for each bite (posture of the user's head during eating). Methods for obtaining information without using the analysis model M1 include, for example, analyzing measurement data using a statistical method or the like, using the sensor unit 21 to obtain the information, and the like.

For example, the pose of the user's head during an eating motion can be derived from the 3-axis acceleration data. Specifically, the analysis unit 42 derives an estimation formula for estimating the orientation of the user's head (orientation of the face) from the triaxial acceleration data in advance according to the following procedure.
1. The denture 2 is rested so as to face each of the first direction, the second direction and the third direction which are orthogonal to each other.
2. First acceleration data in three axial directions measured by the acceleration sensor of the sensor unit 21 when the denture 2 is stationary so as to face the first direction, sensor when the denture 2 is stationary so as to face the second direction 2nd acceleration data in three axial directions measured by the acceleration sensor of the part 21, and third acceleration data in three axial directions measured by the acceleration sensor of the sensor part 21 when the denture 2 is held stationary so as to face the third direction. Get acceleration data.
3. Three directions of the denture 2 (first direction, second direction, third direction) and three-axis acceleration data corresponding to each direction (first acceleration data, second acceleration data, third acceleration data) and Based on the relationship, a calculation formula for estimating the posture of the user's head is derived from the three-axis acceleration data corresponding to an arbitrary state in which the user's head (including the lower jaw) is stationary.

Then, the analysis unit 42 extracts the three-axis acceleration data of the time-series data D1 and D2 when the user is momentarily stationary (for example, when the upper teeth and the lower teeth are engaged) and the above calculations. The posture of the user's head can be estimated (calculated) based on the equations.

The analysis unit 42 uses the analysis model M1 and/or the statistical method, etc., as described above, for all mastication information included in a predetermined period (in this embodiment, the eating time for one meal), It is possible to acquire the analysis result corresponding to the predetermined period (that is, the information obtained by collecting the analysis results corresponding to each mastication section CS included in the predetermined period).

[Proposed module]
Next, the generation unit 44 inputs part or all of the analysis results corresponding to the predetermined period to the proposal model M2 by the proposal module 44a. Thereby, the generation unit 44 acquires the information output from the proposed model M2 as the proposed information. In addition, the generation unit 44 analyzes the analysis results corresponding to the predetermined period using a statistical method or the like, without using the proposed model M2, to obtain the amount of one meal, the nutritional balance, the number of times of mastication, the meal time, Information such as bite force may be acquired, and proposal information may be generated based on the acquired information and a predetermined proposal generation rule. For example, the generating unit 44 predetermines a preferable range for the amount for one meal, and if the amount for one meal (estimated value) obtained from the analysis result is outside the range, the amount for one meal is set to the relevant range. Suggestion information may be generated to prompt the user to stay within the range. For example, when the amount (estimated value) for one meal is less than the range, the generation unit 44 extracts text data such as "Let's eat a little more" from template information prepared in advance. , the text data may be generated as the proposal information. The proposal generation rule described above is information that defines such a rule in advance. Note that the proposal information may be audio data. For example, if a sound generator such as a speaker is provided as the output unit 46, the suggested information may be output as audio data. The above suggestions such as "Let's eat a little more." Suggestions such as “Because you are biased, eat soft foods as well” are examples of suggestions for improving food quality. In this case, the suggestions for improving food quality are, in particular, suggestions for improving hardness of food. The output suggestion information preferably includes at least one of such improvement suggestions for meal amount and food quality improvement suggestions.

Further, the proposal information generated by the generating unit 44 includes, in addition to the above-described information, information grasped by statistical processing of the analysis results, past meals of the user (information grasped from the analysis results). Various information that may be useful to the user, such as history, may be included. An example of the proposal information (display items) generated by the generation unit 44 is shown in the table below.

[Processing Flow of Eating Related Proposal System]
FIG. 10 is a diagram showing an example of the processing flow of the eating-related proposal system 1. As shown in FIG. As shown in FIG. 10, the user first wears the denture 2 (step S1). Then, the user carries out daily life while wearing the denture 2 (step S2). Daily life includes eating periods during which the user consumes food. As a result, the sensor 20 of the denture 2 acquires and stores the user's jaw movement information (in this embodiment, triaxial acceleration data and/or triaxial angular velocity data acquired by the sensor unit 21) during the meal period as measurement data. be done. The user removes the denture 2 from the user's lower jaw and puts it in the storage space 34 of the base device 3 at a timing such as before going to bed (step S3).

Subsequently, the base device 3 acquires measurement data from the sensor 20 of the denture 2 through the communication unit 32 . The base device 3 also charges the sensor 20 with the wireless charging antenna 31 and cleans the denture 2 with the cleaning unit 33 (step S4). After the processing of step S4 is completed (for example, after waking up), the user takes out the denture 2 from the base device 3 (step S5), and puts on the denture 2 again (step S1). On the other hand, the base device 3 transmits the measurement data acquired with the denture 2 set in the storage space 34 in step S4 to the user terminal 4 (step S6).

Subsequently, in the user terminal 4, the acquisition unit 41 acquires the measurement data from the base device 3 (step S7). Subsequently, the analysis unit 42 analyzes the measurement data (step S8). Specifically, as shown in FIG. 5, the analysis unit 42 uses the mastication segment identification module 42a to identify the mastication segment CS (see FIG. 6) for each mouthful, and the filtering/calculation module 42b identifies each mastication segment CS. Remove noise, etc. As a result, mastication information for each mouthful (time-series data D1 and D2 as an example in this embodiment) is obtained. Then, the analysis module 42c of the analysis unit 42 acquires an analysis result for each mouthful from the mastication information for each mouthful. The analysis result includes information obtained using the analysis model M1, but may further include information obtained without using the analysis model M1. In this embodiment, the information obtained using the analysis model M1 includes the attribute of the food estimated to be chewed by the user for each bite and the estimated value of the masticatory muscle activity for each bite. Information obtained without using the analysis model M1 is information obtained by a statistical method or the like for mastication information for each bite. posture of the head at the time) and the like.

Subsequently, the generation unit 44 generates proposal information as described above based on the analysis result by the analysis unit 42 (step S9). Subsequently, the output unit 46 outputs the proposal information generated by the generation unit 44 (step S10). In this embodiment, the output unit 46 outputs display data to the display unit 47 . The display unit 47 displays the display data (proposed information) acquired from the output unit 46 (step S11). Thereby, the proposal information is presented to the user. The proposal information may be presented only to the user who wears the denture 2, or may be presented to people related to the user who wears the denture 2, such as the family of the user who wears the denture 2, the attending physician, and the attending physician. good. Alternatively, the suggested information may be presented only to those associated with the user wearing the denture 2 .

[Effect]
The user terminal 4 (eating-related proposal device) includes an acquisition unit 41 , an analysis unit 42 , a generation unit 44 , and an output unit 46 . The eating-related proposal system 1 also includes a false tooth 2 and a user terminal 4 as an eating-related proposal device that analyzes measurement data acquired by the sensor 20 . The eating-related proposal system 1 and the user terminal 4 generate and output proposal information related to eating for the user from the analysis result of the jaw movement during eating of the user. As a result, it is possible to appropriately present a user or the like with an eating-related proposal based on the analysis result. Therefore, according to the eating-related proposal system 1 and the user terminal 4, useful information can be appropriately provided to the user.

In addition, in the present embodiment, since measurement data analysis, proposal information generation, and proposal information display are performed locally (that is, on the user terminal 4), real-time feedback (presentation of proposal information) is provided to the user. ) becomes possible. In addition, for example, there is no need to transmit the measurement data to a server for analyzing the measurement data (for example, a device installed in a data center that provides a cloud service) or cause the server to perform analysis processing. Therefore, the communication load, the processing load of the server, and the like can be suppressed.

Also, the measurement data (jaw movement information) is information indicating changes over time in at least one of three-axis direction acceleration and three-axis direction angular velocities detected by the sensor 20 provided on the denture 2 attached to the user's lower jaw. contains. In the present embodiment, the measurement data is information (three-axis acceleration data) indicating temporal changes in acceleration in three-axis directions. According to this configuration, by using the sensor 20 provided in the denture 2 worn by the user during daily life including eating, the acceleration and / Or the information which shows the time change of angular velocity can be acquired appropriately as measurement data. For example, conventionally, the amount of masticatory muscle activity is measured by attaching electrodes to the user's outer skin instead of inside the cavity, but in this embodiment, by using the sensor 20 provided on the denture 2 , the attachment of such electrodes can be dispensed with.

Also, the measurement data is time-series information of jaw movement during eating by the user, and the analysis result by the analysis unit 42 includes attribute information indicating the attribute of the food estimated to be masticated by the user during eating. there is In this embodiment, the attributes of food are size, hardness, and type of food. This makes it possible to present the user with suggested information according to the attributes of the food that the user is presumed to have ingested.

The analysis unit 42 also extracts mastication information (time-series data D1 and D2 in this embodiment), which is time-series information indicating jaw movements for one mouthful, from the measurement data. Then, the analysis unit 42 inputs the mastication information into the analysis model M1 generated by machine learning, and the attribute of the food that is estimated to be chewed by the user by jaw movement for one mouthful, which is output from the analysis model M1, is Acquired as attribute information. The analysis model M1 shows mastication information (first information), which is chronological information indicating a past mouthful of jaw movement acquired for the user, and attributes of food chewed by the user through the past mouthful of jaw movement. It is a trained model machine-learned using teacher data including information (second information). With the above configuration, it is possible to easily and accurately acquire analysis results (attributes of food presumed to be masticated by the user) using the analysis model M1 that has learned the features of jaw movement during eating unique to the user. can. In this embodiment, as an example, the analysis model M1 is configured to output an estimated value of the amount of masticatory muscle activity for one mouthful in addition to the attributes of the food. That is, the analysis model M1 is a learned model that has been machine-learned using teacher data that further includes the amount of masticatory muscle activity for the past mouthful. In other words, the output target of the analysis model M1 and the acquisition target of the analysis unit 42 may be only the attribute information of the food, but may further include the attribute information of the food and the estimated value of the masticatory muscle activity amount for one mouthful. When the analysis unit 42 also acquires the estimated value of the masticatory muscle activity amount for one mouthful, the teacher data is configured to include the masticatory muscle activity amount for one mouthful as the second information.

Further, as described above, the analysis model M1 is the first mastication information (in this embodiment, the first mastication information (interval B1 in FIG. 8 data in section B2 in FIG. 8); It may be a learned model that has been machine-learned using teacher data containing two pieces of information. The analysis unit 42 extracts the information corresponding to the first mastication information and the information corresponding to the second mastication information from the mastication information, and inputs the extracted information to the analysis model M1. Attributes of food presumed to be chewed by the user may be obtained. In particular, the jaw movements corresponding to the first chewing motion and the second chewing motion are likely to exhibit characteristics according to the attributes of the masticated food. Therefore, according to the above configuration, it is estimated that the user masticates by using the analysis model M1 machine-learned using the jaw movements corresponding to the first chewing motion and the second chewing motion as feature amounts. Food attributes can be acquired more accurately.

The user terminal 4 also includes a display section 47 that displays the proposal information output by the output section 46 . By displaying the proposal information (for example, part or all of the information corresponding to the display items shown in Table 1 above) on the display unit 47, it is possible to appropriately present the user with the proposal content according to the analysis result. .

The eating-related suggestion system 1 also includes a base device 3 that stores the denture 2 removed from the user's lower jaw. The base device 3 has a communication unit 32 that acquires measurement data from the sensor 20 by communicating with the sensor 20 and that transmits the measurement data to the user terminal 4 by communicating with the user terminal 4 . According to the above configuration, the user simply sets the denture 2 on the base device 3 , and the measurement data is automatically transmitted from the base device 3 to the user terminal 4 . Thereby, the user's convenience can be improved regarding acquisition (sucking) of measurement data from the sensor 20 provided in the denture 2 .

The base device 3 also has a wireless charging antenna 31 for charging the sensor 20 . According to the above configuration, it is possible to perform communication by the communication unit 32 and charging by the wireless charging antenna 31 . As a result, the power consumed by the sensor during the communication operation of the communication unit 32 can be compensated appropriately.

Moreover, the base device 3 further has a cleaning section 33 for cleaning the denture 2 . According to the above configuration, the user can transmit the measurement data to the user terminal 4 and clean the denture 2 only by setting the denture 2 on the base device 3 . Thereby, user convenience can be improved.

[Modification]
An embodiment of the present disclosure has been described above, but the present disclosure is not limited to the above-described embodiment.

For example, in the above embodiment, the terminal of the user wearing the denture 2 (the user terminal 4) is configured as the eating-related suggestion device, but the eating-related suggestion device is configured on a terminal different from the user terminal 4. may For example, the eating-related proposal device (user terminal 4) described above may be a terminal of the user's attending physician (for example, a terminal in a hospital, etc.). In this case, the base device 3 may be provided within the hospital. In this case, when the user regularly visits a hospital, by setting the denture 2 on the base device 3, the measurement data can be downloaded to the terminal of the attending physician. Also, at the time of medical examination, the attending physician can give advice on eating to the user while referring to the analysis result of the measurement data and the proposal information corresponding to the analysis result on the terminal. Note that the presentation target from a terminal different from the user terminal 4 may be the family of the user, the doctor in charge, etc., in addition to the attending physician.

Moreover, like the eating-related proposal system 1A according to the modification shown in FIG. 11, the function of the above-described eating-related proposal device (user terminal 4) may be configured on a server. The eating-related proposal system 1A is different from the eating-related proposal system 1 in that it includes a user terminal 4A and a server 5A instead of the user terminal 4 and the server 5 . Specifically, in the eating-related proposal system 1A, some functions of the user terminal 4 (acquisition unit 41, analysis unit 42, analysis model storage unit 43, generation unit 44, and proposal model storage unit 45) is provided by the server 5A. The server 5A has an output unit 46A. The user terminal 4A has a display section 47A similar to the display section 47. FIG.

The user terminal 4A is configured to, when acquiring measurement data from the base device 3, transmit the measurement data to the server 5A without executing analysis processing on the measurement data by itself. The server 5A, like the user terminal 4 described above, is configured to analyze measurement data and generate proposal information based on the analysis results.

The output unit 46A may notify the user terminal 4A owned by the user wearing the denture 2 of the proposal information. As a result, the proposal information notified from the output unit 46A is displayed on the display unit 47A of the user terminal 4A. According to the above configuration, when the eating-related proposal device is configured as, for example, the server 5A on the cloud as in this modified example, the proposal information can be appropriately presented to the user.

In addition, the output unit 46A may also notify a terminal, which is owned by another user who is different from the user who wears the denture 2 and is set in advance as a notification destination, of the proposal information. For example, the server 5A may preliminarily register and store contact information (e.g., e-mail address, etc.) of other users associated with the user for each user. In this modification, terminals 4B and 4C are set as notification destinations. The terminal 4B is owned by another user who is a family member or close relative of the user. The terminal 4C is a terminal owned by another user who is the user's attending physician. As a result, the proposal information notified from the output unit 46A is displayed on each of the display unit 47B of the terminal 4B and the display unit 47C of the terminal 4C. According to the above configuration, other users can grasp the eating status of the user. This makes it possible for the user's family/close relatives, attending physician, or the like to monitor the user's eating condition.

It is also conceivable that the user may leave the denture 2 on for a long period of time. In this case, the analysis of the measurement data and the generation of the proposal information by the server 5A are not executed, and the notification of the proposal information to the user's family/close relatives, attending physician, etc. is not performed. Therefore, for example, the server 5A (for example, the output unit 46A), when a state in which no proposal information is generated for a certain user continues for a predetermined period or longer, the other user's terminal 4B, which is associated with the user in advance, 4C may be notified accordingly. Such notification may also be made by the base device 3 or the user terminal 4A. For example, the base device 3 may be in a state where the denture 2 is not set in the storage space 34 of the base device 3 after the denture 2 has been previously set in the storage space 34 of the base device 3 and communication, charging, cleaning, etc. have been performed. When a predetermined period of time or more has elapsed, the terminals 4B and 4C of the other users may be notified to that effect. Similarly, when the user terminal 4A cannot acquire measurement data from the base device 3 for a predetermined period or longer, the user terminal 4A may notify the terminals 4B and 4C of other users to that effect. Further, the user terminal 4A may display notification information for prompting the user to set the denture 2 on the base device 3 on the display unit 47A.

Further, in the above embodiment, the analysis model M1 is prepared for each user. That is, the analysis model M1 is generated based only on teacher data obtained from past eating actions of the user to be analyzed. However, the analysis model M1 is based on mastication information (first information) indicating past jaw movements acquired for an unspecified user (a user wearing a denture provided with a sensor similar to the denture 2) and past It may be a learned model machine-learned using teacher data including the attribute (second information) of the food chewed by the unspecified user by the jaw movement of one mouthful. In other words, the teacher data used for learning the analysis model M1 may be obtained from the eating actions performed by a plurality of unspecified users in the past. When teacher data is obtained from past eating actions performed by a plurality of unspecified users, when a certain user becomes an analysis target, analysis corresponding to the second information that most applies to the user among the first information. It may be trained to output results. In this case, the analysis model M1 can be shared among a plurality of users. This has the advantage of eliminating the need to create and manage the analysis model M1 for each user. In addition, compared to the case where the analysis model M1 is prepared for each user as in the above embodiment, there is also the advantage that it is easier to collect teacher data for the analysis model M1.

Furthermore, as shown in FIG. 12, the analytical model M1 may be a learned model that has been machine-learned using teacher data that further includes a user profile that indicates unspecified user attributes. In this case, the analysis unit 42 is configured to further input a user profile (profile information) indicating the attributes of the user (the user to be analyzed) to the analysis model M1. Here, the attributes of the user are, for example, sex, age (or generation), health condition, and the like. It is conceivable that the user's jaw movement during eating differs depending on such attributes of the user. Therefore, according to the above configuration, by including the attribute of the user in the feature amount, it is possible to obtain the attribute of the food that the user is estimated to have chewed with higher accuracy.

The analysis result may also include an opening amount indicating the size of the user's mouth that is opened during eating. Such an opening amount can be grasped, for example, from the waveform of the mastication information (time-series data D3 in FIG. 9) described above. Then, the generation unit 44 may generate proposal information including information regarding the necessity of temporomandibular disorder diagnosis based on the amount of the mouth opening. For example, the generation unit 44 obtains the maximum value (or average value, etc.) of the user's mouth opening amount in the predetermined period from the analysis result corresponding to the predetermined period (for example, one day), and determines the maximum value in advance. and the specified threshold. Then, when the maximum value is smaller than the threshold, the generation unit 44 may determine that temporomandibular disorder is suspected and generate proposal information prompting diagnosis of temporomandibular disorder. According to such a configuration, it is possible to make a proposal to prompt early diagnosis for a user who is suspected of having temporomandibular disorder based on the size of the user's mouth opening during eating.

In addition, the generating unit 44 holds information about the user's usual meal amount in advance, and based on the analysis result (mainly the size of food estimated for each bite) corresponding to the predetermined period, A total meal amount may be calculated, the total meal amount may be compared with the usual meal amount, and proposal information may be generated according to the comparison result. For example, when the total amount of meals is less than the usual amount of meals by a predetermined threshold or more, the generating unit 44 generates suggestion information to prompt an increase in the amount of meals, or generates information suggesting that a stomach or intestinal disease is suspected. It may be determined that there is, and suggestion information may be generated prompting to check the condition of the stomach or intestines. Alternatively, when the total meal amount is greater than the usual meal amount by a predetermined threshold or more, the generating unit 44 may determine that the person is overeating, and may generate suggestion information to encourage the user to reduce the meal amount. .

Further, for example, when the user wears full dentures, the dentures worn on the upper jaw may also be provided with sensors similar to the sensor 20 of the dentures 2 worn on the lower jaw. Then, the measurement data acquired by the acquisition unit 41 is detected by a sensor provided on the denture attached to the user's upper jaw together with the measurement data detected by the sensor 20 (hereinafter referred to as "lower jaw measurement data"). It may also include information indicating temporal changes in at least one of axial acceleration and angular velocity in three axial directions (hereinafter, “upper jaw measurement data”). For example, when the user is riding in a vehicle such as a car or train, the vibration component from the vehicle may be detected by the sensor 20, and may be erroneously detected as the user's jaw movement during eating. be. Moreover, when the user eats while riding in a vehicle, the vibration component from the vehicle will be mixed as noise in the jaw movement of the user during eating. On the other hand, according to the above configuration, the analysis unit 42 detects not only the detection result of the sensor 20 provided on the denture 2 attached to the user's lower jaw, but also the sensor provided on the denture attached to the user's upper jaw. Analysis can be performed using the results. Thereby, for example, the analysis unit 42 obtains the relative value of the detection result of the lower jaw sensor (that is, the lower jaw measurement data) with respect to the detection result of the upper jaw sensor (that is, the upper jaw measurement data), thereby obtaining the vibration component from the vehicle. (that is, components that are common to both the maxillary sensor and the mandibular sensor) can be cancelled. As a result, the problems described above can be resolved.

Also, like the eating-related proposal system 1B according to the modified example shown in FIG. The server may be omitted by providing it as an eating-related proposal device. The user device 4D may be provided independently, for example, in the user's home or the like. Alternatively, the user device 4D connects to a terminal 4E similar to the terminal 4B shown in FIG. 11, a terminal 4F similar to the terminal 4C shown in FIG. Information (for example, information output by the output unit 46) may be notified via. Since the functions of the user device 4D, the terminal 4E, and the terminal 4F are the same as the functions of the user terminal 4, the server 5, the terminal 4B, and the terminal 4C, detailed description thereof will be omitted. When the user device 4D is connected to the terminal 4E and the terminal 4F, it is possible to appropriately notify the related parties of the user (for example, family members, attending physicians, etc.) of suggested information, warnings, and the like. In particular, when the user's related parties live separately from the user, early detection and early handling of problems occurring in the user are possible.

1, 1A, 1B ... eating-related proposal system, 2 ... denture, 3 ... base device, 4 ... user terminal (eating-related proposal device), 4A ... user terminal, 4B, 4C ... terminal, 4D ... device for user ( Eating-related proposal device), 5 Server, 5A Server (Eating-related proposal device), 20 Sensor, 31 Wireless charging antenna (charging unit), 32 Communication unit, 33 Washing unit, 41 Acquisition Section 42...Analysis part 44...Generation part 46, 46A...Output part 47, 47A, 47B, 47C...Display part CS...Mastication section M1...Analysis model M2...Proposal model.

Claims (12)

an acquisition unit that acquires jaw movement information , which is time-series information of jaw movement during eating by a user;
an analysis unit that acquires an analysis result including attribute information indicating an attribute of the food estimated to be chewed by the user at the time of eating, based on the jaw movement information;
Based on the analysis result, generation of proposal information including at least one of advice on the user's dietary content, evaluation of the user's dietary tendency, and advice for improving the user's dietary habits. Department and
an output unit that outputs the proposal information generated by the generation unit;
with
The analysis unit is
extracting mastication information, which is chronological information indicating jaw movements for a mouthful, from the jaw movement information;
The mastication information is input to an analysis model generated by machine learning, and the attribute of the food estimated to be chewed by the user by the jaw movement for the mouthful is output from the analysis model as the attribute information. death,
The analysis model includes first information, which is time-series information indicating a past mouthful of jaw movement acquired for the user, and second information, which is time-series information indicating the past mouthful of jaw movement acquired for the user, and second information indicating attributes of food chewed by the user by the past mouthful of jaw movement. An eating-related proposal device , which is a trained model machine-learned using teacher data including information . an acquisition unit that acquires jaw movement information, which is time-series information of jaw movement during eating by a user;
an analysis unit that acquires an analysis result including attribute information indicating an attribute of the food estimated to be chewed by the user at the time of eating, based on the jaw movement information;
Based on the analysis result, generation of proposal information including at least one of advice on the user's dietary content, evaluation of the user's dietary tendency, and advice for improving the user's dietary habits. Department and
an output unit that outputs the proposal information generated by the generation unit;
with
The analysis unit is
extracting mastication information, which is chronological information indicating jaw movements for a mouthful, from the jaw movement information;
The mastication information is input to an analysis model generated by machine learning, and the attribute of the food estimated to be chewed by the user by the jaw movement for the mouthful is output from the analysis model as the attribute information. death,
The analysis model includes first information, which is time-series information indicating a past mouthful of jaw movement acquired for an unspecified user, and information on food masticated by the unspecified user by the past mouthful of jaw movement. an eating-related proposal device, which is a learned model machine-learned using teacher data including second information indicating an attribute; The analysis unit further inputs profile information indicating attributes of the user to the analysis model,
3. The eating-related proposal device according to claim 2, wherein said analysis model is a learned model machine-learned using said teacher data further including profile information of said unspecified user. The acquisition unit further acquires data on the health condition of the user,
The eating-related proposal device according to any one of claims 1 to 3 , wherein said generating unit generates said proposal information also based on said health condition data. The eating-related proposal device according to any one of claims 1 to 4 , wherein the food attributes include at least one of food size, hardness, and type. The analysis model includes data corresponding to first mastication information, which is time-series information indicating a jaw movement corresponding to a first chewing motion among the first information, and data corresponding to the second chewing motion among the first information. A learned model machine-learned using teacher data including data corresponding to second mastication information, which is time-series information indicating corresponding jaw movements, and the second information,
The analysis unit extracts the first mastication information and the second mastication information from the mastication information, and inputs the extracted information to the analysis model so that the user masticates by the jaw movement for the mouthful. The eating-related proposal device according to any one of claims 1 to 5 , which acquires the attribute of the food estimated to have been eaten. The analysis model uses the teacher data including data from the first mastication information to the n-th mastication information corresponding to each of the first to n-th (n≧3) chewing motions in the first information. is a trained model that has been machine-learned using
The analysis unit extracts each piece of information from the first mastication information to the n-th mastication information from the mastication information, and inputs the extracted information to the analysis model to perform jaw movement for the mouthful. 7. The eating-related proposal device according to claim 6, which acquires the attributes of food presumed to be chewed by said user. The eating-related proposal device according to any one of claims 1 to 7 , further comprising a display unit for displaying said proposal information output by said output unit. The eating-related proposal device according to any one of claims 1 to 8 , wherein said output unit notifies a terminal owned by said user of said proposal information. The input unit according to any one of claims 1 to 9 , wherein the output unit notifies the proposed information to a terminal owned by another user different from the user and set in advance as a notification destination. Meal-related suggestion device. an acquisition unit that acquires jaw motion information related to jaw motion during eating by a user;
an analysis unit that acquires, based on the jaw movement information, an opening amount that indicates the size of the mouth of the user that is opened during eating;
a generation unit that generates information regarding the necessity of diagnosing temporomandibular disorder based on the opening amount;
an output unit for outputting information on whether or not to diagnose the temporomandibular disorder generated by the generation unit;
A feeding-related suggestion device comprising: a denture that is attached to the user's lower jaw and that is provided with a sensor for acquiring jaw movement information regarding the user's jaw movement during eating;
An eating-related proposal system, comprising: the eating-related proposal device according to any one of claims 1 to 11 , which analyzes the jaw movement information acquired by the sensor.

Images