Chewing assistance system

Abstract

The chewing assistance system of the present invention comprises: a chewing information storage unit (31b) for storing information on chewing quality; a muscle activity acquisition unit (21) for acquiring muscle activity signals of a person's chewing muscles; an analysis unit (22) for performing frequency analysis on the muscle activity signals acquired by the muscle activity acquisition unit and analyzing the chewing pattern based on the frequency analysis; a quality judgment unit (23) for judging the quality of the chewing pattern based on the information on the chewing pattern analyzed by the analysis unit; and an extraction unit (24) for extracting auxiliary information corresponding to the chewing quality judged by the quality judgment unit from the chewing information storage unit (32).

Description

Technical Field

[0001] The present invention relates to a system that assists in maintaining and improving the health of the oral cavity and throat area for the purpose of extending healthy lifespan, and in particular to a system that assists and supports improvement in chewing quality as a "delicious chewing and eating function." Background Art

[0002] Chewing and swallowing, as well as saliva production, significantly impact the brain and the entire body, significantly influencing both physical and mental health and healthy lifespan. Maintaining and improving the health of the oral and throat area can ultimately extend healthy lifespan.

[0003] In particular, thorough chewing of meals that require firmness promotes physical and mental growth, activates the brain, improves motor function, inhibits obesity and aging, and maintains sociality, contributing to a longer healthy lifespan. Insufficient chewing, such as ingesting meals that require little chewing, leads to decreased chewing function in growing children and oral weakness in the elderly (see Non-Patent Document 1).

[0004] Furthermore, the effects of "unilateral chewing," where one chews constantly on the same side, extend beyond shortening the lifespan of one tooth, soiling of unchewed teeth, strain on the jaw joint, and facial deformity. These effects on the teeth, jaw, and face can ultimately lead to physical deformities, stiff shoulders, and lower back pain, impacting the entire body. The imbalance of left and right bites (occlusal interference) is also thought to cause physical and emotional stress, impacting both the sympathetic and parasympathetic nervous systems.

[0005] In order to measure the quality of mastication, there are devices such as electromyographs that count chewings and devices that quantify bite force. However, a simple system that can accurately and precisely grasp the quality of complex, multifaceted chewing patterns has not yet been provided.

[0006] Prior art literature

[0007] Patent Literature

[0008] Patent Document 1: Japanese Patent Application Laid-Open No. 6-98865

[0009] Patent Document 2: Japanese Patent Application Laid-Open No. 2019-47859

[0010] Non-patent literature 1: Yoshinori Kobayashi, Commissioned paper: Healthy life span created by occlusion and mastication, Journal of the Japanese Society of Prosthodontics, Ann Jpn Prosthodont Soc 3, p189-219, 2011 Summary of the Invention

[0011] Problems to be solved by the invention

[0012] Therefore, in view of the above situation, the point to be solved by the present invention is to provide a chewing assistance system, which is a simple system that can grasp the quality of complex chewing patterns with many aspects in detail and accurately, and can accurately support the improvement of chewing quality, maintenance and promotion of health.

[0013] Technical solutions to solve problems

[0014] In view of the above-mentioned situation, the inventors conducted in-depth research and found the following situation, which ultimately led to the completion of the present invention: by performing frequency analysis on muscle activity signals during meals obtained by an electromyograph, and in particular making full use of the power values ​​of specific frequency bands that are particularly active during chewing, it is possible to analyze and judge the chewing pattern and its quality in detail and accurately, and based on the judgment results, it is possible to assist in improving chewing quality and maintaining and promoting health.

[0015] That is, the present invention includes the following inventions.

[0016] (1) A chewing assistance system including an information processing device, the information processing device including: a chewing information storage unit for storing information about chewing quality; a muscle activity acquisition unit for acquiring muscle activity signals of a person's chewing muscles; an analysis unit for performing frequency analysis on the muscle activity signals acquired by the muscle activity acquisition unit and analyzing chewing patterns based thereon; a quality judgment unit for judging the quality of the chewing patterns based on the information about the chewing patterns analyzed by the analysis unit; and an extraction unit for extracting auxiliary information corresponding to the chewing quality judged by the quality judgment unit from the chewing information storage unit.

[0017] (2) The chewing assist system according to (1), wherein the analyzing unit performs frequency analysis on the muscle activity signal and analyzes the chewing pattern based on changes in power values ​​in a specific frequency band.

[0018] (3) The chewing assistance system according to (2), wherein the analysis unit analyzes the chewing pattern based on an envelope obtained by performing a fast Fourier transform on each block of electromyographic data as a muscle activity signal, using the envelope as the change condition.

[0019] (4) The mastication assist system according to (2) or (3), wherein the analyzing unit determines that the mastication is performed when the change exceeds a predetermined threshold value.

[0020] (5) The mastication support system according to (4), wherein, regarding the threshold value, mastication is determined when an integral value calculated from the envelope as the state of change exceeds a predetermined threshold value.

[0021] (6) The mastication assist system according to (2) or (3), wherein the analysis unit analyzes the left and right mastication balance based on the changes in the muscle activity signals of the same left and right masticatory muscles.

[0022] (7) The chewing assistance system according to (3), wherein the analysis unit analyzes the characteristics of the chewed food based on the gradient and duration of the chewing interval determined to be chewing based on the change in the envelope.

[0023] (8) A chewing assist system according to (3), wherein the system comprises a user information storage unit for storing a correlation between a user's muscle activity value and a bite force value, wherein the correlation is obtained by obtaining a value of muscle activity when a prescribed food is eaten, the prescribed food being a food with a known characteristic that a bite force required to bite off the prescribed food is determined to be a certain value, and the analysis unit analyzes the bite force during chewing based on the value of the chewing interval judged to be being chewed from the change in the envelope line and the correlation.

[0024] (8) The mastication assist system according to any one of (1) to (7), wherein the analysis unit includes a machine learning unit and determines the mastication pattern by referring to a learning result of the machine learning unit.

[0025] (9) A chewing assist system according to any one of (1) to (8), wherein the chewing pattern analyzed by the analysis unit includes at least one of the following conditions: number of chewings, chewing rhythm, changes in biting movements during meals, degree of bite force, front-to-back / left-to-right chewing balance, and characteristics of the chewed food.

[0026] (10) A chewing assist system according to any one of (1) to (9), wherein the quality of the chewing pattern judged by the quality judgment unit includes at least one of the number of chewing times, the quality of the chewing rhythm, the quality of the change in the biting action, the quality of the bite force, the quality of the left and right chewing balance, whether the food is biased to one side, and the quality of the use of the masseter muscle.

[0027] (11) The chewing assistance system according to any one of (1) to (10), wherein the quality judgment unit judges whether there is improvement by comparing the chewing pattern with the past chewing pattern of the same person.

[0028] (12) The mastication assist system according to any one of (1) to (11), wherein the quality judgment unit includes a machine learning unit and judges the quality of the mastication pattern with reference to a learning result of the machine learning unit.

[0029] A chewing assistance program is a control program that enables an information processing device to function as a chewing assistance system described in any one of (1) to (12), and is used to enable the information processing device to function as the muscle activity acquisition unit, analysis unit, quality judgment unit and extraction unit.

[0030] Effects of the Invention

[0031] According to the present invention constructed as above, the muscle activity signal is frequency analyzed, the chewing pattern is analyzed based on the signal, the quality of the chewing pattern is judged, and auxiliary information corresponding to the judged chewing quality can be provided. Therefore, a simple system can be provided that can grasp the quality of complex chewing patterns with many aspects in detail, and can accurately support the improvement of chewing quality, maintenance and promotion of health.

[0032] According to this invention, it is particularly possible to provide accurate information on the healthy chewing quality of children during their developmental period, thereby providing a system that helps to improve the healthy chewing motor function of children during their developmental period. Furthermore, it is also possible to provide accurate auxiliary information corresponding to the chewing quality of the elderly, thereby providing a system that helps to maintain and improve the chewing motor function of the elderly. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 This is a block diagram showing the configuration of a mastication support system according to a representative embodiment of the present invention.

[0034] Figure 2A Raw data showing the muscle activity signal of the left masticatory muscle.

[0035] Figure 2B Raw data showing the muscle activity signal of the right masticatory muscle.

[0036] Figure 3A is a heat map showing the frequency distribution of muscle activity power values ​​on the left.

[0037] Figure 3B is a heat map showing the frequency distribution of muscle activity power values ​​on the right.

[0038] Figure 4 This is an explanatory diagram showing a method of determining a chewing zone using an envelope curve.

[0039] Figure 5 This is a graph showing raw data of a muscle activity signal of the masticatory muscles and an envelope curve obtained by performing FFT processing on the raw data.

[0040] Figure 6 This is a graph showing the raw data of a chewing interval, each area integral value of the envelope, and the bite force of the chewing interval.

[0041] Figure 7 This is a graph showing raw data of left and right muscle activity signals when chewing with a bias towards the left side of the teeth and an envelope curve obtained by performing FFT processing on the raw data.

[0042] Figure 8 This is a graph showing raw data of left and right muscle activity signals when chewing with a bias towards the right teeth and envelope curves obtained by performing FFT processing on the raw data.

[0043] Figure 9 This is a graph showing raw data of muscle activity signals of the temporalis and masseter muscles during mastication using molars, and envelope curves obtained by performing FFT processing on the raw data.

[0044] Figure 10 This is a graph showing raw data of muscle activity signals of the temporalis and masseter muscles during chewing using the incisors and envelope curves obtained by performing FFT processing on the raw data.

[0045] Figure 11A This is a graph showing raw data of muscle activity signals at different chewing speeds and rhythms, and envelopes obtained by performing FFT processing on the raw data.

[0046] Figure 11B This is a graph showing raw data of muscle activity signals at different chewing speeds and rhythms, and envelopes obtained by performing FFT processing on the raw data.

[0047] Figure 12 This is a flowchart showing the processing procedure of the mastication assistance system according to the representative embodiment. DETAILED DESCRIPTION

[0048] Next, embodiments of the present invention will be described in detail based on the drawings.

[0049] When chewing a meal, a person has preferences for the hardness and softness of the food, the movements of tearing and chewing it, the number of times the food is chewed, the duration of chewing, and the rhythm of chewing. Furthermore, the balance of the teeth during chewing is another factor. The quality of chewing is considered to be the quality of chewing. This system analyzes the frequency of the muscle activity signals of the masticatory muscles, and analyzes chewing patterns such as the number of chews, chewing rhythm, changes in bite movements, bite force, left-right chewing balance, food bias towards one side, and how the masseter muscles are used. The system then determines the quality of chewing. Furthermore, the system indicates changes over time based on the difference between the past and the present, thereby indicating improvements in chewing quality.

[0050] Specifically, if Figure 1As shown, the chewing aid system 1 of the present invention is composed of a single or multiple information processing devices 10 including a processing device 2, a storage unit 3, a muscle activity meter 4, and an information display unit 5. Specifically, the information processing device 10 is a computer or the like, which is centered around the processing device 2 and includes a storage unit, an indicator, an input unit such as a keyboard and a touch panel, a display unit such as a monitor, and other communication control units (not shown).

[0051] The processing device 2 is primarily composed of a CPU, such as a microprocessor, and includes a storage unit comprised of RAM and ROM (not shown), which stores programs defining the sequence of various processing operations and processing data. The storage unit 3 is comprised of memories and hard disks within and outside the information processing device 10. Alternatively, some or all of the contents of the storage unit may be stored in the memory or hard disk of another computer communicatively connected to the information processing device 10. This information processing device may be a dedicated device installed in a dental clinic, hospital, other facility, store, or the like, or a general-purpose personal computer installed in a home. Alternatively, it may be a smartphone carried by the user.

[0052] Functionally, the processing device 2 includes: a muscle activity acquisition unit 21 as a muscle activity acquisition unit, which performs the following processing: acquires the muscle activity signal of the user's masticatory muscles acquired and transmitted by the muscle activity meter 4, and stores the signal in the muscle activity data storage unit 31a within the user information storage unit 31; an analysis unit 22, which performs the following processing: performs frequency analysis on the above-mentioned muscle activity signal, analyzes the chewing pattern based on the signal, and stores the analyzed chewing pattern information in the chewing pattern storage unit 31b within the user information storage unit 31; a quality judgment unit 23 as a quality judgment unit, which performs the following processing: judges the chewing quality based on the above-mentioned chewing pattern information, and stores the judged chewing quality information in the judgment information storage unit 31c within the user information storage unit 31; an information extraction unit 24, which uses the judged chewing quality information as input and extracts recommended information from the chewing quality information stored in the chewing information storage unit 32; and an information output processing unit 25, which displays the information on a display (information display unit 5) and the like to prompt the user. These processing functions are realized by the above-mentioned program.

[0053] The muscle activity meter 4 is an electromyograph or the like, preferably having a communication unit capable of wirelessly transmitting and receiving data with the user's smartphone constituting the information processing device 10 at short distances. Alternatively, it may be an external electromyograph or the like connected via wired or wireless communication to a dedicated computer or the like constituting the information processing device 10. The muscle activity meter 4 acquires muscle activity signals of the user's masticatory muscles.

[0054] The muscle activity meter 4 acquires muscle activity signals for the masticatory muscles, measuring the activity of at least one of the four masticatory muscles on either side of the head: the temporalis and masseter muscles. To measure balance during chewing, the activity of at least two muscles is measured and compared. That is, to measure left-right balance, the activity of at least the left and right temporalis muscles or the left and right masseter muscles is acquired. To measure front-to-back balance, the activity of at least the left temporalis and masseter muscles, or the right temporalis and masseter muscles, is acquired.

[0055] The analysis unit 22 functions as an analysis unit, performing frequency analysis on the muscle activity signal acquired by the muscle activity acquisition unit 21. The analysis analyzes the chewing pattern based on changes in the power values ​​within a specific frequency band. By fully utilizing the power values ​​within a specific frequency band (e.g., 150 Hz to 450 Hz), which is particularly active during chewing, a more accurate analysis is possible.

[0056] In more detail, it includes: an FFT processing unit 22a, which extracts the data of the muscle activity signal from the muscle activity data storage unit 31a, performs high-speed Fourier transform on each block, obtains the average power value of a specific frequency band, stores these power values ​​in the power value storage unit 311, and generates an envelope of the obtained power value (hereinafter referred to as "envelope" in this specification) and stores it in the envelope storage unit 312; a pattern analysis processing unit 22b, which analyzes the chewing pattern and stores the result in the analysis result storage unit 313.

[0057] The specific processing performed by the FFT processing unit 22a is as follows. Assume that the muscle activity meter samples at 2000 samples / second. The FFT processing unit 22a first divides the raw muscle activity signal data (2000 samples / second) into blocks of a predetermined number of samples (here, 64 samples) and performs a fast Fourier transform on each block.

[0058] For the fast Fourier transform of each block, in this example, 32 pins (frequencies) are set, dividing the range from 0 to 1000 Hz into 32 equal parts, and the power value is calculated for each predetermined number of pins. Each pin corresponds to a frequency of 31.25 Hz (an integer multiple). The FFT processing unit 22a then calculates the average of, for example, eight power values ​​within a specific frequency band (here, between pins 7 and 14, i.e., 218.75 to 437.5 Hz) for each block, and outputs this average as the average power value for each block. The power value represents the amplitude of the spectrum at a specific frequency.

[0059] Figure 2A 、 Figure 2B This is the raw data of the muscle activity signal of the left and right masticatory muscles (2000 samples / second), Figure 3A 、 Figure 3BThis is a diagram for confirming the frequency distribution of power values ​​obtained by fast Fourier transforming the raw data by the FFT processing unit 22a for each block using the above-mentioned specific example as a heat map. Figure 2A and Figure 3A is the data / heatmap on the left, Figure 2B and Figure 3B The data / heat map on the right shows that averaging the power values ​​between 7 and 14 pins allows for more accurate judgment of chewing patterns based on muscle activity signals. However, other pin ranges, such as 6 to 14 or 6 to 15, are also preferred.

[0060] In addition, for example Figure 5 This example shows an envelope curve that connects the average power values ​​for each block (every 32 ms) output by the FFT processing unit 22a based on the above example. Compared to the graph of the original muscle activity data, this envelope curve further reflects the chewing pattern. Using this envelope curve, chewing pattern can be more accurately determined using data related to the chewing frequency band.

[0061] In order to reduce the storage area, it is preferable that the muscle activity data (raw data) stored in the muscle activity data storage unit 31 a is deleted from the storage unit 3 when the analysis result is stored in the analysis result storage unit 313 .

[0062] The pattern analysis processing unit 22b uses the envelope curve generated by the FFT processing unit 22a and other methods to analyze various chewing patterns and stores them in the analysis result storage unit 313. Examples of chewing patterns analyzed include the number of chewing times, chewing rhythm, changes in bite movements during meals, degree of bite force, front-to-back / left-to-right chewing balance, and characteristics of the chewed food. In this example, as a prerequisite for chewing pattern analysis, the pattern analysis processing unit 22b includes a chewing determination unit 221 that determines whether chewing is present. The chewing determination unit 221 determines chewing if the envelope curve exceeds a predetermined threshold. The details are as follows.

[0063] (Chew judgment)

[0064] Preferably, the background is first calculated using the envelope of the interval where muscle activity (average power value) is small and stable, clearly indicating non-chewing. This background is then multiplied by a coefficient to set the threshold for chewing determination. Then, chewing is determined when this threshold is exceeded under certain conditions.

[0065] Specifically, the background value can be calculated by analyzing the envelope and performing a low-pass filter. The filter can be a first-order autoregressive filter as shown in the following equation.

[0066] Y n=0.99Y n－1 +0.01X n－80

[0067] “X n－80 " is the envelope value 2.56 seconds ago. Here, "2.56" is the value obtained by 80samples / 31.25samples / s=2.56s. "Y n－1 " is the latest value of the background level, "Y n " is the new value of the background level. "0.99" is the filter constant, and "0.01" is the gain factor (gain coefficient) of the input signal, ensuring a full gain of 1.

[0068] In order to reduce the computational load of the embedded processor, the calculation is preferably performed by integer arithmetic. This can be done by multiplying the value from the FFT algorithm by a factor of 10,000 (8-bit algorithm). In addition, the above-mentioned filtering is calculated as follows.

[0069] Y n =(99Y n－1 +X n－80 ) / 100

[0070] The background value is preferably not calculated after the start of mastication is detected until a predetermined time has passed since the end of mastication is detected, and is maintained at the background level before the start of mastication is detected.

[0071] like Figure 4 As shown in FIG, the threshold is set to a value obtained by multiplying the background level by a predetermined value (e.g., a value 2.6 times the background level). The start / end of chewing is preferably detected when the envelope exceeds / falls below the threshold value between two sample times (64ms). The chewing determination unit 221 determines whether chewing occurs, for example, in the following manner. Subsequently, information on the number of chewings, speed, and rhythm can also be analyzed. Figure 11A 、 Figure 11B As shown, the speed and rhythm of chewing are varied.

[0072] Another method for determining chewing is to use the following formula to calculate the background as a moving average of a certain interval of the envelope. Similar to the above background calculation, the threshold value at the current moment can be set at 2.56 seconds (80 samples), and a background threshold (e.g., 1.2 times) can be set to be used as the average value.

[0073] Y n =X n－80 +4σ n－80

[0074] Here, "Y n " is the new value of the background level, X n－80This is the moving average of the envelope 2.56 seconds ago. Here, "2.56" is the value calculated by 80 samples / 31.25 samples / s = 2.56 seconds. The moving average is the envelope immediately before the calculation time (10 samples / 31.25 samples / s = 320 milliseconds ago), "σ" is the standard deviation, and "4" is the coefficient of variation.

[0075] In this case, it is also preferable that the threshold value for chewing judgment be a value obtained by multiplying the deviation from the background average value by a predetermined deviation coefficient (eg, 4), for example.

[0076] In addition, if the envelope exceeds the threshold value during a predetermined time, it is detected as the start of chewing. Figure 6 As shown, the area integral value of the envelope of the chewing interval is further calculated, and an integral threshold is set. If the integral value of the envelope of a chewing interval detected as the above chewing is lower than the integral threshold, it is considered as a non-chewing interval and excluded from the chewing. This is also a preferred embodiment. Figure 5 The short and weak chewing intervals judged by Figure 5 The interval enclosed by A in Figure 6 The intervals surrounded by A' are excluded from the chewing objects, and only reliable chewing actions are judged as chewing and counted.

[0077] However, the "background" also vibrates repeatedly within a certain range when the human body has no muscle activity. The potential (noise) at rest is defined as "background (noise)". However, as in the present invention, during chewing activities, muscle activities that do not exceed the judgment threshold are generated through human reactions other than chewing (shaking the head, deep breathing, etc.). When such muscle activities occur within a certain period of time, the threshold value rises only in the above-mentioned threshold calculation method, and depending on the degree of the rise, it may hinder the judgment of chewing.

[0078] Specifically, the threshold stops counting only when muscle activity exceeds the threshold. Therefore, the threshold for determining muscle activity events rises due to the following two factors: (1) situations where muscle activity events temporarily exceed the threshold but are short-lived and not counted as events, and (2) situations where small amounts of muscle activity occur over a short period of time. The resulting rise in the threshold in these situations can lead to misjudgments of chewing that should actually be counted and to undercounting chewing intensity.

[0079] Therefore, the calculation of the threshold is preferably performed as follows. That is, when the amplitude of the envelope fluctuation within a certain period of time is calculated, and the fluctuation amplitude produces a sharp change exceeding the specified value, the calculation of the threshold in that interval is stopped. In contrast, when the amplitude of the envelope fluctuation within a certain period of time remains at a gentle change that does not exceed the specified value, the calculation of the threshold in that interval is performed. In this way, the rise in the threshold caused by the influence of (1) and (2) above can be suppressed, and a stable threshold can be obtained. Moreover, the threshold for determining muscle activity events can be accurately tracked relative to a gentle rise in myoelectric potential or a long-term rise in myoelectric potential.

[0080] (Balance Analysis)

[0081] The pattern analysis processing unit 22b of this embodiment also includes a balance analysis unit 222 for analyzing front-to-back, left-to-right, and right-to-left chewing balance. For example, regarding left-to-right chewing balance, the balance analysis unit 222 calculates one or both of the integral value and the maximum peak value for each chewing interval of each envelope generated based on muscle activity data of the same left and right masticatory muscles (temporalis and masseter muscles), and compares these values ​​to determine the left and right magnitude of the chewing force. If the difference in these values ​​exceeds a predetermined threshold, or if chewing continues multiple times, it can be determined that the chewing is biased to the left or right.

[0082] Figure 7 The left and right envelopes are shown when chewing to the left. As can be seen from the figure, the integral value and peak value of each chewing interval are larger on the left side. Figure 8 In contrast, the envelope curve deviates to the right side when chewing, and the integral value and peak value are both larger when they are on the right side. Thus, it can be seen that by comparing the integral value or maximum peak value of the left and right envelope curves in the chewing interval, the left-right balance can be analyzed.

[0083] The masticatory muscles (temporalis and masseter) are involved in unilateral mastication, where food is continuously chewed with one tooth. In a healthy state, the left and right muscles exhibit roughly the same activity trend. Therefore, it is difficult to directly calculate left-right dominance from the raw signals (raw data) of muscle activity. By performing a balance analysis using frequency analysis, as in this example, it is possible to accurately determine which muscle is dominant.

[0084] For the front-to-back chewing balance, the maximum peak value of each chewing interval of each envelope generated based on the muscle activity data of the temporalis and masseter muscles is calculated and compared. When the peak value of the temporalis muscle becomes smaller than the peak value of the masseter muscle side to above the specified threshold, or when the chewing is continued for multiple times, it can be judged as chewing biased towards the incisor side.

[0085] Figure 9This figure shows the envelope of muscle activity data for the temporalis and masseter muscles during mastication using molars. As can be seen from this figure, the temporalis and masseter muscles are active at the same level, with similar peak values ​​for both muscles within each chewing interval. Figure 10 The envelope of chewing primarily with the incisors shows significantly less activity in the temporalis muscle, and the maximum peak values ​​for each chewing interval are significantly smaller than those for the masseter and temporalis muscles. This suggests that by comparing the maximum peak values ​​for the envelopes of the temporalis and masseter muscles in each chewing interval, it is possible to analyze whether chewing is biased toward the incisors.

[0086] (Analysis of chew characteristics)

[0087] The pattern analysis processing unit 22b of this embodiment also includes a chew property analysis unit 223 that analyzes the characteristics of the chewed food (texture: physical properties such as hardness and softness). The chew property analysis unit 223 analyzes these characteristics based on the gradient and duration of the chewing intervals of the envelope curve. The gradient of the chewing interval tends to be larger for harder ingredients, and this gradient can be used to determine the chewed food's characteristics, namely, the degree of hardness or softness of the food.

[0088] Furthermore, the shape of each chewing interval of the envelope curve, obtained from muscle activity data obtained when chewing a plurality of foods (prescribed foods) whose respective characteristics are known, is stored as a basic shape. Pattern analysis, etc., using this shape, can also determine the characteristics of the chewed food. Furthermore, it is also preferable to use the shape of the envelope curve (such as inclination, peaks, and other characteristic points) when a user chews the prescribed foods as training data, and to perform the determination by the machine learning unit.

[0089] (Bite Force Analysis Department)

[0090] The pattern analysis processing unit 22b of this embodiment further includes a bite force analysis unit 224 for analyzing the bite force during chewing. The relationship between muscle activity data and bite force varies from person to person. That is, even when chewing with the same bite force, muscle activity values ​​vary depending on the person. Therefore, in this embodiment, a correlation table between muscle activity values ​​(the aforementioned power values) and bite force values ​​is pre-generated for each user and stored in the user information storage unit 31.

[0091] This correlation table is generated by obtaining muscle activity data (power values) when a user bites through a variety of foods with known characteristics (prescribed foods). The hardness of the prescribed foods, or the bite force required to break them, is fixed to a constant value. Therefore, the muscle activity data (power values) when biting through the prescribed foods correspond one-to-one with the bite force values ​​obtained for the foods.

[0092] Therefore, the bite force analysis unit 224 can analyze the bite force during chewing by converting the power value (average value and maximum value) of the envelope of the chewing interval into bite force using the above-mentioned correlation table. Figure 6The diagram shows that Figure 5 In this example, the above correlation table is obtained by chewing a plurality of predetermined foods in advance. However, instead of this table, the correlation relationship can be approximated as a proportional relationship, and the correlation coefficient (proportional constant) can be obtained from one or more predetermined foods.

[0093] As a method of generating a correlation table, as a modified example, a muscle activity meter and a bite force meter may be used in combination and connected to directly obtain the correlation between the bite force and the muscle activity data.

[0094] (Chewing Movement Analysis)

[0095] The pattern analysis processing unit 22b of this embodiment further includes a chewing action analysis unit 225 for analyzing biting, crushing, or grinding movements. When the chewing interval gradient is small and the duration is long, the chewing action analysis unit 225 determines that chewing is a biting action. When the gradient is large and the duration is short, the chewing action analysis unit 225 determines that chewing is a crushing action. When chewing is determined based on data such as gradients, it is preferable to divide a chewing session into multiple intervals and perform a more detailed analysis using, for example, displacement within specific intervals or between intervals.

[0096] Analyzing these biting and chewing movements can further analyze changes in these chewing movements. Eating typically progresses from placing food in the mouth through biting, chewing, grinding or concentrating, and swallowing. Understanding this process allows understanding the entire process from placing food in the mouth to swallowing, and can help determine eating habits (behavioral characteristics) such as speed and habits.

[0097] The quality of the chewing pattern judged by the quality judgment unit 23 includes the number of chewing times, the quality of chewing rhythm, the quality of changes in biting movements, the quality of bite force, the quality of left and right chewing balance, whether the food is biased to one side, the quality of how the masseter muscles are used, etc.

[0098] Based on the obtained data and the user's past information and age-related statistical information stored in the judgment information storage unit 31c, information such as whether the user has improved their chewing quality compared to the past and whether they have a chewing quality appropriate for their age is preferably included. The quality judgment unit 23 preferably includes a machine learning unit 23a, which judges the quality of the chewing pattern by referring to the learning results of the machine learning unit 23a.

[0099] The information extraction unit 24 functions as an extraction unit. For example, if the chewing quality corresponding to age is not available, it is preferred to extract information such as oral function information corresponding to age, information on equipment used for oral hygiene and improvement, and information on specialists corresponding to the user's residence. Furthermore, it is preferred to provide suggestions for improvement, such as chewing more slowly or chewing harder foods.

[0100] Furthermore, the system can prompt users with any problems (how hard, how, and where to chew). This helps improve chewing quality for those who have teeth and the potential for healthy chewing, even if they have them. Chewing patterns are particularly important for children, and it can identify factors such as grinding and hard chewing to promote improvement. Furthermore, the use of masticatory muscles is particularly important for the elderly, and it can identify the type of masticatory muscles, the load on them, and provide guidance.

[0101] Figure 12 This is a flowchart showing the processing procedure performed by the mastication support system 1 according to this embodiment.

[0102] First, the muscle activity acquisition unit 21 obtains at least the muscle activity data of the user's chewing muscles from the time when the specified food (specified food) or ordinary food is put into the mouth to the time when it is swallowed (S101) from the muscle activity meter 4, and stores it in the muscle activity data storage unit 31a in the user information storage unit 31 (S102).

[0103] Next, the FFT processing unit 22a performs a high-speed Fourier transform on the muscle activity data for each block, thereby obtaining the average power value of a specific frequency band (S103), and storing them in the power value storage unit 311 (S104), and generating an envelope of the obtained power value (S105), and storing it in the envelope storage unit 312 (S106).

[0104] Next, the pattern analysis processing unit 22b analyzes the chewing pattern based on the envelope (S107) and stores the results in the analysis result storage unit 313 (S108). Next, the quality determination unit 23 determines the quality of the chewing pattern based on the analysis results (S109) and stores the determined chewing pattern quality information in the determination information storage unit 31c within the user information storage unit 31 (S110).

[0105] Next, the information extraction unit 24 takes the determined chewing quality information as input and extracts recommended information from the chewing quality information stored in the chewing information storage unit 32 (S111). The information output processing unit 25 then displays the extracted information on a display (information display unit 5) or the like to present it to the user (S112).

[0106] The above describes the embodiments of the present invention, but the present invention is not limited to such embodiments. For example, it is also preferable to constitute a part or all of the processing device by a hardware processing circuit instead of constituting the processing device by software processing performed by a computer. In this case, an artificial intelligence processing circuit can also be used as a machine learning unit. Of course, it can be implemented in various forms within the scope of the present invention.

[0107] Industrial applicability

[0108] The present invention can judge the quality of complex chewing patterns with many aspects in detail and accurately, and can provide auxiliary information corresponding to the judgment results. Therefore, by combining with instruments or products and services for chewing education and chewing training for children, it is possible to provide products and services that contribute to the healthy development of children. In addition, by combining with training instruments or services for beauty, such as chewing patterns with good balance in front, back, left and right, and how to use the chewing muscles, it is also possible to provide products and services for beauty that prevent facial deformation and obesity and maintain vivid and healthy expressions. In addition, by combining with products and services corresponding to oral weakness caused by reduced oral function of the elderly and physical aging, it is also possible to provide products and services that contribute to the extension of healthy life expectancy.

[0109] Description of Reference Numerals

[0110] 1 Chewing assistance system

[0111] 2 Processing device

[0112] 3 Storage Units

[0113] 4 Muscle activity meter

[0114] 5 Information display unit

[0115] 10 Information processing device

[0116] 21 Muscle activity acquisition unit

[0117] 22 Analysis Department

[0118] 22a FFT processing unit

[0119] 22b Sample Analysis and Processing Unit

[0120] 23 Quality Judgment Department

[0121] 23a Machine Learning Department

[0122] 24 Information Extraction Department

[0123] 25 Information output processing unit

[0124] 31 User information storage unit

[0125] 31a Muscle activity data storage unit

[0126] 31b Chewing pattern storage unit

[0127] 31c Judgment information storage unit

[0128] 32 Chewing information storage unit

[0129] 221 Chewing Judgment Department

[0130] 222 Balance Analysis Department

[0131] 223 Chewable Material Characteristics Analysis Department

[0132] 224 Bite Force Analysis Department

[0133] 225 Chewing Movement Analysis Department

[0134] 311 Power value storage unit

[0135] 312 Envelope Storage Unit

[0136] 313 Analysis result storage unit.

Claims

1. A chewing assistance system comprising an information processing device, characterized in that: The information processing device includes: a chewing information storage unit for storing information on chewing quality; A muscle activity acquisition unit, which acquires muscle activity signals of a person's masticatory muscles; an analyzing unit, configured to perform frequency analysis on the muscle activity signal acquired by the muscle activity acquiring unit, and analyze the chewing pattern based on the frequency analysis; a quality judging unit that judges the quality of the chewing pattern based on the information of the chewing pattern analyzed by the analyzing unit; and an extracting unit that extracts auxiliary information corresponding to the mastication quality determined by the quality determining unit from the mastication information storage unit, The analysis unit, Perform frequency analysis on muscle activity signals and analyze chewing patterns based on changes in power values ​​in specific frequency bands. Based on the envelope obtained by performing fast Fourier transform on each block of electromyographic data as muscle activity signals, the chewing pattern is analyzed as the above-mentioned change status. When the integral value calculated from the envelope as the state of change exceeds a predetermined threshold value, chewing is determined.

2. The chewing aid system according to claim 1, wherein: The analyzing unit analyzes the left and right chewing balance based on the changes in the muscle activity signals of the same left and right masticatory muscles.

3. The chewing aid system according to claim 1, wherein: The analyzing unit analyzes characteristics of the chewed food based on a gradient and a duration of a chewing interval determined to be chewing based on the change in the envelope.

4. The chewing aid system according to claim 1, wherein: A user information storage unit is provided for storing a correlation between a user's muscle activity value and a bite force value, wherein the correlation is obtained by obtaining the muscle activity value when a predetermined food is consumed, the predetermined food being a food with known characteristics such that a bite force required to bite the predetermined food is determined to be a certain value, The analysis unit analyzes the bite force during chewing based on the value of the chewing interval determined as being chewed based on the change in the envelope and the correlation.

5. The chewing aid system according to any one of claims 1 to 4, characterized in that: The analyzing unit includes a machine learning unit and determines the chewing pattern with reference to a learning result of the machine learning unit.

6. The chewing aid system according to any one of claims 1 to 5, characterized in that: The chewing pattern analyzed by the analysis unit includes at least one of the following: number of chewing times, chewing rhythm, changes in biting movements during meals, degree of bite force, front-back / left-right chewing balance, and characteristics of the chewed food.

7. The chewing aid system according to any one of claims 1 to 6, wherein: The quality of the chewing pattern judged by the quality judgment unit includes at least one of the number of chewing times, the quality of chewing rhythm, the quality of changes in biting movements, the quality of bite force, the quality of left and right chewing balance, whether the food is biased to one side, and the quality of how the masseter muscle is used.

8. The chewing aid system according to any one of claims 1 to 7, wherein: The quality judgment unit judges whether there is improvement by comparing the chewing pattern with the past chewing pattern of the same person.

9. The chewing aid system according to any one of claims 1 to 8, wherein: The quality judgment unit includes a machine learning unit and judges the quality of the chewing pattern with reference to a learning result of the machine learning unit.

10. A chewing assistance program is a control program that enables an information processing device to function as the chewing assistance system described in any one of claims 1 to 9, and is used to enable the information processing device to function as the muscle activity acquisition unit, analysis unit, quality judgment unit and extraction unit.

Images