Food bolus forming device, chewing state evaluation method, food texture evaluation method, and food bolus manufacturing method

Abstract

The present invention provides a food bolus forming device which can be used for assessing the texture of food by reproducing a person's chewing. A food bolus forming device (1) according to the present invention is provided with: lower artificial teeth (4) in an artificial oral cavity space (S); upper artificial teeth (5) disposed at positions opposing the lower artificial teeth (4); an artificial tongue (6) disposed in parallel with the lower artificial teeth (4) or the upper artificial teeth (5); a wall surface-shaped artificial cheek (7) disposed lateral to the lower artificial teeth (4) and / or the upper artificial teeth (5); and a robot arm (2a) and a robot hand (2b) that drive the lower artificial teeth (4) or the upper artificial teeth (5) to carry out a biting motion. The food bolus forming device (1) chews food in the artificial oral cavity space (S) by a biting motion, and forms a bolus of the food.

Description

[Technical field]

[0001] The present invention relates to a food bolus forming device that forms a food bolus by masticating food in an artificial oral space, a method for evaluating a mastication state, a method for evaluating food texture, and a method for producing a food bolus. [Background technology]

[0002] Conventionally, when evaluating the texture of food, the evaluator would chew the food, spit it out, and evaluate it. Recently, evaluation systems that use a pressing mechanism and a pressure sensor to press the food and evaluate the texture have become known.

[0003] For example, the food texture evaluation system in Patent Document 1 comprises a pressing device for pressing a sample whose texture is to be evaluated, a measuring device (pressure distribution sensor) that measures the change over time in the pressure distribution received from the sample when the sample is pressed, and a food texture evaluation means (control PC) that controls the pressing operation by the pressing device and evaluates the texture of the sample based on the pressure distribution data from the pressure distribution sensor.

[0004] This pressing device has a pair of upper and lower plates, a pressure distribution sensor is placed on the upper surface of the lower plate, and a sample is placed on the pressure distribution sensor. The upper plate is provided at a position facing the pressure distribution sensor in the vertical direction, and is connected to a linear slider. The linear slider is driven in the vertical direction with respect to the pressure sensor surface in response to a pressing operation control signal from a control PC, so that the pressing operation of the sample can be controlled (paragraphs 0022 to 0028, Figure 1). [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Patent No. 6324741 Summary of the Invention [Problem to be solved by the invention]

[0006] The food texture evaluation system in Patent Document 1 is configured to evaluate food texture only based on feature quantities obtained from changes in pressure distribution over time. However, actual human chewing behavior varies greatly from person to person in terms of chewing style, saliva infiltration, and other factors, making it difficult to accurately evaluate food texture.

[0007] The present invention has been made in consideration of the above circumstances, and has an object to provide a food bolus forming device that can be used to evaluate the texture of food by reproducing human mastication. [Means for solving the problem]

[0008] In order to achieve the above-mentioned object, the food bolus forming device of the present invention comprises a first artificial tooth provided in an artificial oral space, a second artificial tooth arranged in a position opposite the first artificial tooth, an artificial tongue arranged in parallel with the first artificial tooth or the second artificial tooth, a wall-shaped artificial cheek arranged on at least one side of the first artificial tooth and the second artificial tooth, and a driving means for driving the first artificial tooth or the second artificial tooth to perform an occlusal motion, and is characterized in that food placed in the artificial oral space is chewed by the occlusal motion to form a bolus of the food.

[0009] In the food bolus forming device of the present invention, the first artificial tooth or the second artificial tooth in the artificial oral space is driven by the driving means, and the first artificial tooth and the second artificial tooth bite food. Since the food bolus forming device has an artificial tongue and artificial cheek, food that is bitten in the same way as in a human oral cavity is pushed back to the center of the artificial oral space by the artificial cheek acting as a wall, and its position is adjusted by the artificial tongue. As a result, a food bolus of the food is formed in the artificial oral space, and this device can reproduce human mastication.

[0010] In the food bolus forming device of the present invention, it is preferable that the driving means performs an operation of horizontally shifting one artificial tooth relative to the other artificial tooth, in addition to the occlusion operation.

[0011] The drive means of the food bolus forming device can operate to shift one artificial tooth (for example, the first artificial tooth) horizontally relative to the other artificial tooth (the second artificial tooth). This enables the device to form a food bolus while grinding food in the same way as in the human mouth.

[0012] It is also preferable that the food bolus forming device of the present invention further comprises food collecting means for collecting the food dispersed by the biting action at a predetermined position in the artificial oral cavity space.

[0013] In the food bolus forming device of the present invention, food is scattered into the artificial oral space due to the bite action. Therefore, the food bolus forming device is provided with a food collecting means to collect the scattered food at a predetermined position in the artificial oral space (for example, the upper surface of the artificial tongue). This allows the device to form a food bolus by more faithfully reproducing human mastication.

[0014] In the food bolus forming apparatus of the present invention, it is preferable to further comprise moisture supplying means for supplying moisture to the food.

[0015] By providing a moisture supplying means in the food bolus forming device, moisture equivalent to saliva is supplied to the artificial oral cavity space. This makes it easier for food to form a bolus with the moisture, and the device can mimic the state of human chewing.

[0016] It is also preferable that the food bolus forming device of the present invention further comprises an imaging means for imaging the food or the food bolus in the artificial oral space, and an evaluation means for evaluating the chewing state from the image of the food bolus captured by the imaging means.

[0017] According to this configuration, the imaging means captures an image of the food or food bolus in the artificial oral cavity space, and the evaluation means evaluates the chewing state from the image by image analysis, etc. This allows the device to quantitatively evaluate the progress of the chewing state.

[0018] In the food bolus forming apparatus of the present invention, it is preferable that the evaluation means evaluates the food bolus based on one or two properties selected from local change and overall uniformity.

[0019] The evaluation means of the food bolus formation device evaluates the food bolus in the artificial oral cavity space in one (one) or two (both) ways: local change and overall uniformity. This allows the device to accurately evaluate the progress of the chewing state.

[0020] Moreover, it is preferable that the food bolus forming device of the present invention further comprises a machine learning means for inputting images of food boluses in different chewing states and performing machine learning, and that the food bolus is evaluated using a judgment method obtained by the machine learning means.

[0021] According to this configuration, the machine learning means learns from images of many food boluses with different chewing states and establishes a judgment method, which enables the evaluation means of the present device to accurately evaluate the state of the food bolus.

[0022] The method for evaluating the mastication state of the present invention is a method for evaluating the mastication state of the food using the food bolus forming device described in any one of claims 1 to 7, and the method for evaluating the food texture of the present invention is a method for evaluating the food texture of the food using the food bolus forming device described in any one of claims 1 to 7.

[0023] Further, a food texture method of the present invention is a food bolus production method using the food bolus formation apparatus according to any one of claims 1 to 4. Effect of the Invention

[0024] According to the present invention, the texture of various foods can be evaluated with high accuracy. [Brief description of the drawings]

[0025] [Figure 1] FIG. 1 is a schematic diagram illustrating a food bolus forming device of the present invention. [Diagram 2] FIG. 2 is a perspective view of the food bolus forming device of the present invention. [Diagram 3]1 is a flowchart of a food bolus formation process by the food bolus formation device. [Figure 4A] FIG. 4 is a diagram illustrating the chewing trajectory of the food bolus forming device. [Figure 4B] A diagram explaining five types of chewing trajectories. [Diagram 5] FIG. 13 is a diagram for explaining an experiment on food bolus formation using the food bolus formation device. [Figure 6A] Graph showing the results (Contrast) when evaluating food using a food bolus formation device. [Figure 6B] Graph showing the results (Angular Second Moment) when evaluating food using a food bolus formation device. [Figure 7A] Graph showing the results (normalized contrast) when evaluating foods using a food bolus formation device. [Figure 7B] Graph showing the results (normalized Angular Second Moment) when evaluating foods using a food bolus formation device. [Figure 8A] Graph showing the results (normalized contrast) of evaluating two types of food using a food bolus formation device. [Figure 8B] A graph showing the results (normalized Angular Second Moment) of evaluating two types of food using a food bolus formation device. [Figure 9A] A graph showing the results of evaluating two types of food using a texture analyzer. [Figure 9B] A diagram showing the results of a sensory evaluation of two types of food. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0026] [First embodiment] Hereinafter, a first embodiment of a food bolus forming device according to the present invention will be described with reference to the drawings.

[0027] First, an overview of a food bolus forming device 1 according to the present invention will be described with reference to Figs. 1 and 2.

[0028] Fig. 1 is a schematic diagram of a food bolus forming device 1. As shown in the figure, the food bolus forming device 1 is mainly composed of a robot arm 2a, a robot hand 2b, and a chewing mechanism unit 3 that reproduces the inside of a human oral cavity. The robot arm 2a and the robot hand 2b correspond to the "driving means" of the present invention.

[0029] The tip of the robot arm 2a corresponds to the lower jaw, and the lower artificial tooth 4 is attached to it. The movement of the lower artificial tooth 4 has two degrees of freedom (V Ax ,V Ay ) and can perform biting (occlusion) and grinding movements. The aluminum frame 11 corresponds to the upper jaw, and upper artificial teeth 5 are attached to it.

[0030] The biting operation is performed by moving the robot arm 2a vertically (V Ax ) to bite food with the lower artificial teeth 4 and the upper artificial teeth 5. The mortar grinding operation is performed by driving the robot arm 2a in the horizontal direction (V Ay ) to shift the lower artificial teeth 4 and the upper artificial teeth 5, thereby grinding up the food.

[0031] Both the lower artificial tooth 4 and the upper artificial tooth 5 are made of resin and are produced using a 3D printer. As shown in the figure, a groove and a protrusion are formed in the center of the lower artificial tooth 4 and the upper artificial tooth 5, respectively.

[0032] An artificial tongue 6 is attached to the tip (gripper moving part) of the robot hand 2b. The movement of the artificial tongue 6 is a vertical movement (V By The artificial tongue 6 has only one degree of freedom: the vertical movement of the artificial tongue 6. The artificial tongue 6 is made of silicone, and an elastic sheet is attached to its surface, which expands and contracts in accordance with the vertical movement of the artificial tongue 6.

[0033] Additionally, an artificial cheek 7 is attached to the portion adjacent to the lower artificial tooth 4. The artificial cheek 7 is made of silicone, and is operated by the robot arm 2a in the same manner as the lower artificial tooth 4. The chewing mechanism 3 has the above-mentioned configurations, and food delivered to the artificial oral cavity space is crushed and mashed through the chewing motion, mixed with saliva, and transformed into a food bolus.

[0034] Next, Fig. 2 shows an overall perspective view of the food bolus forming device 1. As shown in the figure, a frame 11 is disposed above the chewing mechanism 3. In addition to the upper artificial teeth 5, a collecting tongue 8 (the "food collecting means" of the present invention) that collects chewed food, and a camera 9 (the "imaging means" of the present invention) are attached to the frame 11. A commercially available cotton swab is used as the collecting tongue 8, which collects food scattered in the artificial oral cavity space S due to the biting action, and moves it to, for example, the upper surface of the lower artificial teeth 4 or the artificial tongue 6.

[0035] Furthermore, the camera 9 is preferably a web camera connected to a computer, and the camera 9 can automatically capture images of the food being chewed. As will be described in detail later, a computer (the "evaluation means" of the present invention) performs image analysis on the images captured by the camera 9 and further evaluates the food bolus (degree of progress of chewing). This allows the food bolus forming device 1 to quantitatively evaluate the chewing state of food. Each step of chewing, food collection, and image capture is automatically controlled by a computer program.

[0036] Also, acrylic plates are used to create walls 7' on the sides of the lower artificial teeth 4 and the artificial tongue 6 to prevent the chewed food from spilling out. Furthermore, in order to accurately reproduce the inside of a human mouth, a saliva supply unit (the "moisture supplying means" of the present invention) is required. In this example, a spray was used to add moisture (0.6 ml per time) to the food bolus, but a water storage unit 12 may be provided in the upper artificial teeth 5 to supply moisture to the food bolus at predetermined intervals. This helps form a food bolus, as the food dispersed in the artificial oral cavity space S becomes small lumps due to the moisture. With this configuration, the food bolus forming device 1 can faithfully reproduce human chewing in the artificial oral cavity space S.

[0037] Next, the food bolus formation process by the food bolus formation device 1 will be described with reference to a flowchart in FIG.

[0038] First, in the food bolus forming device 1, the counter i is set to "1" and the number of chewing times n is set to "0" (step S01). In the food bolus forming device 1, the lower jaw moves toward the fixed upper jaw, so the number of chewing times is also called the number of mandibular cycles. Then, the process proceeds to step S02.

[0039] In step S02, imaging is started by the camera 9. Specifically, the camera 9 images the state of chewing of food, which will start later. Then, the process proceeds to step S03.

[0040] Next, it is determined whether or not the counter i has reached the upper limit cycle number N (step S03). If the upper limit cycle number N has been reached, the process proceeds to step S08, and if not, the process proceeds to step S04.

[0041] If the counter i has not reached the upper limit cycle number N ("NO" in step S03), moisture supply is executed (step S04). This is a process for supplying moisture to replace saliva to the artificial oral space S. Then, the process proceeds to step S05.

[0042] In step S05, the food is chewed. Here, the number of chewing times n is added by 5, so the lower jaw is moved five times in succession. Before and after the fifth lower jaw movement, the artificial tongue 6 is moved up and down to stir the food bolus. As a result, the food is chewed in the artificial oral cavity space S, and a food bolus is gradually formed. Then, the process proceeds to step S06.

[0043] In step S06, food collection is performed. This is a process in which the collection tongue 8 (cotton swab) scrapes up food scattered within the artificial oral cavity space S. Then, the process proceeds to step S07.

[0044] In step S07, the counter i is incremented by 1, and the process returns to step S03. If the counter i has not reached the upper limit cycle number N ("NO" in step S03), the process of steps S04 to S07 is repeated again.

[0045] On the other hand, if the counter i reaches the upper limit cycle number N ("YES" in step S03), the camera 9 stops capturing images (step S08). After that, the processing of the series of food bolus formation processes ends. As described above, the food bolus formation device 1 captures and analyzes the state of food bolus formation (degree of progress of mastication) for each food item.

[0046] Next, the chewing trajectory of the food bolus forming device 1 will be described with reference to Figs. 4A and 4B.

[0047] In the food bolus formation device 1 of the present invention, the trajectory of a person's mandible is linearized and given as a parallelogram-type trajectory. As shown in Fig. 4A, the position where the upper jaw teeth (upper artificial teeth 5) and the lower jaw teeth (lower artificial teeth 4) mesh is taken as the origin O, the x-axis is taken horizontally, and the y-axis is taken vertically. If the movements in the x-axis and y-axis directions are expressed as a function of time t, the trajectory will have the form of the following Fourier series. Here, f [Hz] is the frequency, X [mm] is the mortise length, Y [mm] is the bite length, and α (0≦α≦2) is a parameter that defines the ratio of the mortise length to the bite length.

[0048] (1) When 0≦α<1 x(t) = αXA(t) (1A) y(t) = YB(t) (1B) (2) When 1≦α≦2 x(t) = XA(t) (2A) y(t) = (2-α)YB(t) (2B) however,

number

number

number

[0049] Figure 4B shows five types of trajectories with different parameters (α = 0, 0.5, 1.0, 1.5, 2.0), and the arrows indicate the direction of movement. In this study, we observed the food bolus (chewing state) when each of these trajectories was adopted.

[0050] Next, an experiment on food bolus formation using the food bolus formation device 1 will be described with reference to FIG.

[0051] This time, the upper limit number of cycles N was fixed at 6 [times], frequency f = 1.0 [Hz], mortar length X = 10.0 [mm], and bite length Y = 8.5 [mm], and an experiment on food bolus formation using the food bolus formation device 1 was conducted. Using a sample food (a doughnut from company A, 5.0 g), the experiment was conducted 10 times for each of the chewing trajectories described above, and food bolus images were obtained for the number of chews n = [0, 5, 10, 15, 20, 25, 30].

[0052] Figure 5 shows images of food according to the number of times it is chewed, from top to bottom, for (a) α = 0, (b) α = 0.5, (c) α = 1.0, (d) α = 1.5, and (e) α = 2.0. As shown in the figure, the food bolus gradually breaks down as the number of times it is chewed increases. It can also be seen that the way the food breaks down and sticks together differs depending on the chewing trajectory.

[0053] Next, to quantitatively evaluate the differences between these food bolus images, we perform image texture analysis using a Grey-Level Co-occurrence Matrix (GLCM), which is a matrix that represents the frequency with which pairs of pixels with a specified spatial relationship occur in an image (the derivation method is omitted).

[0054] The size of each food bolus image in Fig. 5 is 1920 × 1080px, and a portion of 224 × 224px is cropped from it (area T). The cropped image is then converted to a grayscale image, after which the co-occurrence matrix is ​​calculated and f is used as the image texture feature. C : Contrast (local change) and f A : Calculate the Angular Second Moment (overall uniformity).

[0055] Contrast indicates the local variation in gray levels of an image, and Angular Second Moment indicates the uniformity of an image.

[0056] Fig. 6A is a graph showing the results of contrast (local change) from the above experiment. This result includes data on a "human food bolus" when a person (subject) chewed a food (a donut from company A, 10.0 g) (Human), for comparison with the "artificial food bolus" produced by the food bolus formation device 1. The "human food bolus" data is from when the subject chewed naturally in time with a metronome at a frequency of 1.0 [Hz].

[0057] In Figure 6A, the horizontal axis represents the number of chewing cycles n [cycles], and the vertical axis represents the local change f C (average). C (Average) is the average value of all 10 trials, normalized by the value of the number of chewing trials n = 0. From this result, it can be seen that, especially when horizontal grinding movements are included such as α = 0.5-2.0, f C It can be seen that (average) increases once and then tends to gradually decrease. The reason for this is thought to be that heterogeneous parts appear at the beginning of chewing, but as chewing progresses, these parts gradually decrease.

[0058] In addition, (i), (ii), and (iii) in Fig. 6A show the number of chewing cycles n = 0, 10, and 30, respectively, when α = 1.0. The corresponding food bolus images (star marks) in Fig. 5 also show the changes in the food bolus caused by chewing. In comparison with the graph of the "human food bolus," f C Although there were differences in the (average) size, among the "artificial food boluses" produced by the food bolus formation device 1, α=1.0 tended to be closest to the "human food bolus."

[0059] FIG. 6B is a graph showing the results of the Angular Second Moment (overall uniformity) from the above experiment. In FIG. 6B, the horizontal axis represents the number of chewing cycles n [cycles], and the vertical axis represents the overall uniformity f A (average).A (Average) is the average value of all 10 trials, normalized by the value for number of chews n=0.

[0060] The graph trends can be divided into three groups based on the parameters of the trajectory close to the biting movement (α = 0, 0.5), the trajectory close to the grinding movement (α = 1.5, 2.0), and the intermediate trajectory (α = 1.0). A It can be seen that (average) decreases up to the number of chews n = 10, and then tends to increase monotonically.

[0061] Here too, among the "artificial food boluses" produced by the food bolus formation device 1, α=1.0 tended to be closer to the "human food bolus." These results suggest that the food bolus formation device 1 of the present invention can reproduce a human food bolus by providing an appropriate mandibular trajectory.

[0062] In this study, the food bolus (chewing state) was evaluated by extracting Contrast (local change) and Angular Second Moment (overall uniformity) as features from the image captured by camera 9, but it is also possible to evaluate the food bolus using at least one of these.

[0063] Furthermore, the food bolus may be evaluated by making the image captured by the camera 9 judged by a machine learning model. Specifically, about 1000 input images of the same food (donut) in different chewing states are prepared. The input images are food bolus images captured according to the number of chews (e.g., 0 to 30), and these are classified into classes based on the number of chews to be used as training data.

[0064] Then, a convolutional neural network (CNN) model is trained to create an estimation model for the chewing class. In particular, a convolutional neural network can input images as 2D data and automatically extract effective features during the learning process. This makes it possible to easily and quickly determine the extent to which chewing has progressed based on the state of the food bolus when a new food bolus image is input.

[0065] [Second embodiment] A second embodiment of the food bolus forming device according to the present invention will be described below.

[0066] In the first embodiment, the food bolus obtained by the food bolus formation device 1 was photographed by the camera 9, and the food bolus was evaluated by image analysis using Contrast (local change) and Angular Second Moment (overall uniformity). The second embodiment is similar in that the food bolus is image analyzed, but the food bolus is evaluated using improved evaluation values ​​called "normalized Contrast" and "normalized Angular Second Moment."

[0067] First, the normalized Contrast f C Explain the definition of (n) (average). C (n) (average) is the average of all 10 trials of contrast, f C This value is calculated by subtracting (0) (average) (difference value) and then normalizing it using the maximum absolute value of the difference value.

[0068] f C (n) (average) is given by the following formula:

number

[0069] Next, the normalized Angular Second Moment f A Explain the definition of (n) (average). A(n) (average) is the average value of all 10 trials of Angular Second Moment, f A This value is calculated by subtracting (0) (average) (difference value) and then normalizing it using the maximum absolute value of the difference value.

[0070] f A (n) (average) is given by the following formula:

number

[0071] Fig. 7A is a graph showing the results of contrast (local change) when food X (a donut from company A) was evaluated using the food bolus formation device 1. The parameter used was α=1.0, which is close to the "human food bolus" formed by humans. In addition, in (Equation 6), the cases of number of chews n=m={0, 5, 10, 15, 20, 25, 30} were plotted and graphed.

[0072] The data for the "human food bolus" is shown with a dashed line for comparison with the "artificial food bolus" (solid line) produced by the food bolus formation device 1. The "human food bolus" data is from a case in which the subject chewed naturally at a frequency of 1.0 [Hz] in time with the metronome.

[0073] In this case, the conditions are almost the same as α = 1.0 in Figure 6A, so the normalized local change f C (n) (average) initially increases monotonically, reaches a maximum at n = 10-20 [times] of chewing, and then tends to decrease. In other words, heterogeneous parts appear at the beginning of chewing, but as chewing progresses, these parts gradually decrease. The same is true for the "human food bolus."

[0074] Fig. 7B is a graph showing the Angular Second Moment (overall uniformity) when food X (a donut from company A) was evaluated using the food bolus formation device 1. Again, the parameter α=1.0, which is close to that of a "human food bolus," was used. In addition, in (Equation 7), the cases of number of chews n=m={0, 5, 10, 15, 20, 25, 30} were plotted and graphed (solid line).

[0075] f A (n) (average) initially decreased monotonically, reaching a minimum when the number of chewing cycles n was 10-15, and then tended to increase gradually. In other words, uniformity is suddenly lost at the beginning of chewing, but as chewing progresses, uniformity is gradually restored. The same is true for the "human food bolus" (dashed line). These results suggest that the food bolus forming device 1 of the present invention can reproduce a human food bolus by providing an appropriate mandibular trajectory, even if the definition of local changes and overall uniformity are changed.

[0076] Next, the results of an investigation into whether foods with different textures can be evaluated using the food bolus forming device 1 will be described with reference to Figs. 8A and 8B.

[0077] In Fig. 8A, the solid line graph shows the results of contrast (local change) when food X (a doughnut from company A) was evaluated using the food bolus formation device 1. Again, the parameter α = 1.0, which is close to that of a "human food bolus," was used. Also, in (Equation 6), the cases of number of chews n = m = {0, 5, 10, 15, 20, 25, 30} were plotted.

[0078] The dashed line graph shows the Contrast results when food Y (a doughnut from company B) was evaluated using the food bolus forming device 1. The parameters and number of chewing cycles n (m) were the same as for food X, so if the two waveforms can be distinguished, it means that the food bolus forming device 1 can distinguish the texture of food X and food Y. Here, it can be seen that chewing progresses faster for food Y than for food X.

[0079] In addition, in Fig. 8B, the solid line graph shows the results of the Angular Second Moment (overall uniformity) when food X (a donut from company A) was evaluated using the food bolus formation device 1. Again, the parameter α = 1.0, which is close to that of a "human food bolus," was used. In addition, in (Equation 7), the cases where the number of chews n = m = {0, 5, 10, 15, 20, 25, 30} were plotted.

[0080] The dashed line graph shows the results of the Angular Second Moment when food Y (a donut from company B) was evaluated using the food bolus formation device 1. As shown in the figure, it can be seen that food Y recovers uniformity faster than food X, and mastication progresses faster.

[0081] Finally, the results of the evaluation by a texture analyzer and the sensory evaluation of Food X and Food Y will be described with reference to Figs. 9A and 9B.

[0082] Fig. 9A is a graph showing the evaluation of food X and food Y using a texture analyzer (TA.XTplus: manufactured by Stable Micro Systems). In Fig. 9A, the horizontal axis represents the number of mastications n [times], and the vertical axis represents the maximum stress [g] when compressed.

[0083] Specifically, the food boluses of Food X and Food Y were measured using a texture analyzer after 5 to 30 chews to measure the maximum stress [g] during compression. This showed that the maximum stress of Food Y was always smaller than the maximum stress of Food X, indicating that Food Y was softer.

[0084] Fig. 9B is a graph showing the sensory evaluation of Food X and Food Y. The sensory evaluation was performed by five evaluators, who chewed in time with a metronome (set at 100 times per minute). The evaluators evaluated the food on a scale of 0 to 10 using the VAS method, and the average value of six repeated evaluations by the five evaluators was taken as the sensory evaluation score.

[0085] According to this, the sensory evaluation also showed that food Y had a score (melt in the mouth) that was more than twice as high (P value indicates a significant difference in a multiple comparison test). These results demonstrate that even though they are both donuts, food Y (company B) has a better melt in the mouth, and the food bolus formation device 1 of the present invention was able to distinguish the textures of food X and food Y.

[0086] In this way, the food bolus forming device 1 according to the present invention can form a food bolus by chewing food in the artificial oral space S just like in a human mouth. By providing an appropriate mandibular trajectory to the food bolus forming device 1, we have succeeded in reproducing the process by which a human forms a food bolus. By using the food bolus forming device 1, it is possible to investigate, for example, how a newly developed food is chewed by a human.

[0087] Furthermore, since the degree of progress of mastication can be easily evaluated by image recognition of the food bolus, the food bolus forming device 1 can also compare the texture of multiple foods. Although a donut was used as the food this time, the evaluation method is the same for other foods. However, when evaluating a food bolus image using a machine learning model, it is necessary to have the model learn images corresponding to the mastication state of the target food in advance. [Explanation of symbols]

[0088] Reference Signs List: 1...food bolus formation device, 2a...robot arm, 2b...robot hand, 3...chewing mechanism, 4...lower artificial teeth, 5...upper artificial teeth, 6...artificial tongue, 7...artificial cheek, 7'...wall, 8...collecting tongue, 9...camera, 11...frame, 12...water storage section, S...artificial oral space.

Claims

1. A first artificial tooth provided in the artificial oral space; A second artificial tooth disposed at a position opposite to the first artificial tooth; An artificial tongue arranged in parallel with the first artificial tooth or the second artificial tooth; a wall-shaped artificial cheek disposed on at least one side of the first artificial tooth and the second artificial tooth; a driving means for driving the first artificial tooth or the second artificial tooth to perform an occlusal movement; a food collecting means for collecting food scattered by the bite action at a predetermined position in the artificial oral space, the food placed in the artificial oral space is chewed by the occlusion action, and the food dispersed by the occlusion action is collected by the food collecting means at a predetermined position in the artificial oral space, thereby forming a bolus of the food; A food bolus forming device, characterized in that, during an occlusion movement, the food collecting means is positioned outside the artificial cheek.

2. 2. The food bolus forming device according to claim 1, wherein the drive means performs an operation of horizontally shifting one artificial tooth relative to the other artificial tooth in addition to the occlusal operation.

3. 3. The food bolus forming apparatus according to claim 1, further comprising a moisture supplying means for supplying moisture to the food.

4. an imaging means for imaging the food or the food bolus in the artificial oral space; an evaluation means for evaluating a chewing state from an image of the food bolus captured by the imaging means; The food bolus forming device according to any one of claims 1 to 3, further comprising:

5. 5. The food bolus forming apparatus according to claim 4, wherein the evaluation means evaluates the food bolus based on one or two characteristics selected from local change and overall uniformity.

6. A machine learning means is provided for performing machine learning by inputting images of food boluses in different chewing states, The food bolus forming device according to claim 4 or 5, characterized in that the food bolus is evaluated using a judgment method obtained by the machine learning means.

7. A food bolus forming device as described in any one of claims 1 to 6, wherein one of the first artificial tooth and the second artificial tooth constitutes a lower jaw tooth and the other constitutes an upper jaw tooth, the lower jaw tooth moves up and down toward the upper jaw tooth, and the artificial tongue is located between the lower jaw teeth and is configured to move up and down independently of the lower jaw teeth.

8. A method for evaluating a mastication state, comprising: evaluating a mastication state of the food using the food bolus forming device according to any one of claims 1 to 7.

9. A food texture evaluation method for evaluating the food texture using the food bolus forming device according to any one of claims 1 to 7.

10. A method for producing a food bolus using the food bolus forming apparatus according to any one of claims 1 to 4.

Images