JP2021032770A - Food product evaluation device and food product evaluation method - Google Patents

Abstract

To use for development of a food product of higher precision.SOLUTION: A waveform data generation unit 20 generates waveform data of surface myogenic potential from change in surface myogenic potential in deglutition muscle when a subject swallows food. A parameter calculation means 30 calculates a parameter regarding swallowing by analyzing waveform data generated by the waveform data generation unit 20. A sensory evaluation data storage unit 40 stores sensory evaluation data indicating to what extent flavor of a food product swallowed by the subject is similar to a taste of the target food along with the parameter calculated by the parameter calculation means 30. A correlation analyzing unit 50 analyzes a correlation between the parameter calculated by the parameter calculation unit 30 and the sensory evaluation data stored in the sensory evaluation data storage unit 40. An evaluation method storing means 60 stores the analysis result of the correlation between the parameter analyzed by the correlation analysis unit 50 and the sensory evaluation data as an evaluation expression.SELECTED DRAWING: Figure 1

Description

The present invention relates to a food and drink evaluation device and a food and drink evaluation method.

In recent years, consumer tastes for food have become more diverse. In general, factors for evaluating the taste of food include the taste, aroma, texture, chewyness, texture, and throat of the food. Among these, it is known that the texture such as the chewy texture and chewy texture of food and the swallowing sensation such as the throat of a beverage can be evaluated by measuring the myoelectric potential.

Patent Document 1 discloses an apparatus for evaluating the texture of food and drink by measuring the myoelectric potential of muscles involved in mastication and swallowing.

Japanese Unexamined Patent Publication No. 2012-139442

However, in the apparatus of Patent Document 1, although it is described that the texture is evaluated, the evaluation of the flavor other than the texture is not examined.

Conventionally, sensory evaluation that relies exclusively on human senses has been emphasized for the flavor of foods and drinks. Sensory evaluation is suitable for comprehensive evaluation, but subjective factors such as individual differences, sensory fatigue, and changes in physical condition may affect the evaluation. The QDA method (quantitative descriptive analysis method) is a method that gives objectivity to the subjective evaluation, but it takes a lot of time and cost to select common terms and train the panel. For this reason, food and drink developers have longed for new food and drink evaluation devices and food and drink evaluation methods that are more objective and useful for developing highly accurate food and drink while keeping costs down.

Therefore, the present invention provides a food and drink evaluation device and a food and drink evaluation method that are useful for developing more accurate food and drink.

The food and drink evaluation device of the present invention has a waveform data generation unit, a parameter calculation unit, a sensory evaluation data storage unit, a correlation analysis unit, and an evaluation formula storage unit.

The waveform data generation unit generates waveform data of the surface myoelectric potential from the change in the surface myoelectric potential of the swallowing muscle when the subject swallows food and drink. The parameter calculation unit analyzes the waveform data generated by the waveform data generation unit and calculates parameters related to swallowing (hereinafter, also simply referred to as parameters). The sensory evaluation data storage unit generates sensory evaluation data indicating how similar the flavor of the food or drink swallowed by the subject is to the flavor of the target food or drink, together with the parameters calculated by the parameter calculation unit. Remember. The correlation analysis unit analyzes the correlation between the sensory evaluation data stored in the sensory evaluation data storage unit and the parameters calculated by the parameter calculation unit. The evaluation formula storage unit stores the analysis result of the correlation between the parameters analyzed by the correlation analysis unit and the sensory evaluation data as an evaluation formula.

The other food and drink evaluation device of the present invention has a waveform data generation unit, a parameter calculation unit, an evaluation formula storage unit, and an estimation unit.

The waveform data generation unit generates waveform data of the surface myoelectric potential from the change in the surface myoelectric potential of the swallowing muscle when the subject swallows food and drink. The parameter calculation unit analyzes the waveform data generated by the waveform data generation unit and calculates parameters related to swallowing. The evaluation formula storage unit stores the evaluation formula of the correlation between the parameter and the sensory evaluation data. The estimation unit uses the parameters calculated by the parameter calculation unit and the evaluation formula stored in the evaluation formula storage unit, and the flavor of the food or drink swallowed by the subject is different from the flavor of the target food or drink. Estimate if you feel similar in degree.

The food and drink evaluation method of the present invention includes a stage of generating surface myoelectric potential waveform data from changes in the surface myoelectric potential of the swallowing muscle when a subject swallows food and drink, and a parameter related to swallowing by analyzing the generated waveform data. And the stage where the sensory evaluation data indicating how similar the flavor of the food and drink swallowed by the subject to the flavor of the target food and drink is stored together with the calculated parameters. It includes a step of analyzing the correlation between the calculated parameter and the stored sensory evaluation data, and a step of storing the analysis result of the correlation between the analyzed parameter and the sensory evaluation data as an evaluation formula.

Another food and drink evaluation method of the present invention includes a step of generating surface myoelectric potential waveform data from changes in the surface myoelectric potential of the swallowing muscle when a subject swallows food and drink, and a step of analyzing the generated waveform data to swallow. How similar is the flavor of the food or drink swallowed by the subject to the flavor of the target food or drink using the stage of calculating the parameters related to? Including the stage of estimating and.

According to the food and drink evaluation device and the food and drink evaluation method of the present invention, it becomes easy to obtain an objective index as to how similar the flavor of the food and drink is to the flavor of the target food and drink, and the accuracy becomes higher. Can develop expensive food and drink.

It is a block diagram of the food and drink evaluation apparatus of Embodiment 1. It is a figure which shows the procedure of the food evaluation method of Embodiment 1. FIG. It is a figure which shows an example of the waveform data generated by the waveform data generation part. It is a figure which provides the explanation of the parameter calculated by the parameter calculation unit. It is a figure which shows the waveform obtained by full-wave rectifying the waveform shown in FIG. It is a figure which shows an example of the questionnaire form for acquiring the sensory evaluation data. It is a figure which shows an example of the map obtained from the correlation between the parameter analyzed by the correlation analysis part, and the sensory evaluation data. It is a figure which shows an example of the plot figure based on the evaluation formula obtained from the correlation between the parameter analyzed by the correlation analysis part, and the sensory evaluation data. It is a figure which shows the verification result of the evaluation formula of FIG. It is a block diagram of the food and drink evaluation apparatus of Embodiment 2. It is a figure which shows the procedure of the food evaluation method of Embodiment 2.

Hereinafter, embodiments of the food and drink evaluation device and the food and drink evaluation method of the present invention will be described separately in [Embodiment 1] and [Embodiment 2].

In the first embodiment, the waveform data when the subject tastes the food or drink under development and the subject swallows the food or drink under development is similar to the flavor of the target food or drink. It is an embodiment about a food and drink evaluation device and a food and drink evaluation method that finally stores an evaluation formula from sensory evaluation data regarding whether or not the person feels that the person is feeling.

Further, in the second embodiment, the subject is asked to sample the food or drink under development, and the food or drink under development swallowed by the subject using the waveform data at that time and the evaluation formula stored in the first embodiment. The present invention relates to a food and drink evaluation device and a food and drink evaluation method for estimating how similar the flavor of the above is to the flavor of the target food and drink.

[Embodiment 1]
[Configuration of food and beverage evaluation device]
FIG. 1 is a block diagram of the food and drink evaluation device of the first embodiment. The food and drink evaluation device 100 includes a waveform data generation unit 20, a parameter calculation unit 30, a sensory evaluation data storage unit 40, a correlation analysis unit 50, and an evaluation formula storage unit 60.

The myoelectric potential measurement electrode 10 is connected to the waveform data generation unit 20. The waveform data generation unit 20, the parameter calculation unit 30, the sensory evaluation data storage unit 40, the correlation analysis unit 50, and the evaluation formula storage unit 60, which constitute the food and drink evaluation device 100, are provided in one housing. For example, the food and drink evaluation device 100 is provided in a general personal computer (PC). In this embodiment, all the elements of the waveform data generation unit 20, the parameter calculation unit 30, the sensory evaluation data storage unit 40, the correlation analysis unit 50, and the evaluation formula storage unit 60 are housed in one housing. However, the present invention is not limited to this, and the elements may be arbitrarily combined and housed in separate housings.

The myoelectric potential measuring electrode 10 measures a change in the surface myoelectric potential of the swallowing muscle when the subject swallows food and drink. The myoelectric potential measuring electrode 10 is attached to either the lower part of the chin or the anterior neck of the subject, or both the lower part of the chin and the anterior neck.

The swallowing muscle may be any muscle involved in swallowing and is not particularly limited. Preferably, it is the suprahyoid muscle group and / or the infrahyoid muscle group. The suprahyoid muscle group is a muscle group including the geniohyoid muscle, the mylohyoid muscle, the digastric muscle, and the stylohyoid muscle. Is. The infrahyoid muscle group is a muscle group including the sternothyroid muscle, the thyrohyoid muscle, the scapulohyoid muscle, and the sternohyoid muscle, and is directly or indirectly connected to the lower part of the hyoid body. The muscles of the anterior jaw and its surroundings.

Specifically, the myoelectric potential measuring electrode 10 has an electrode for grounding attached to one place of the subject's ear canal, and 1 to 3 places of the surface myoelectric potential measuring part of the lower part of the chin and / or the anterior jaw. 1 to 2 pairs of electrodes are attached to the chin. It is said that the electric potential generated in the muscle is attenuated to 1/1000 or less by the time it conducts the subcutaneous tissue and reaches the body surface, and the magnitude of the electric potential actually obtained on the body surface is several tens of μV or more. It is about several mV. Therefore, the surface myoelectric potential to be measured is preferably in the range of about 5 μV to 5 mV. The frequency of the surface potential to be sampled is preferably in the range of 0 to 1000 Hz, and in the case of the surface myoelectric potential, a large amount of information on muscle activity is included in the range of 5 to 500 Hz.

The waveform data generation unit 20 generates waveform data of the surface myoelectric potential from the change in the surface myoelectric potential of the swallowing muscle when the subject swallows food and drink, which is measured by the myoelectric potential measurement electrode 10. Specifically, the waveform data generation unit 20 generates waveform data by arranging the surface myoelectric potentials of the swallowing muscles measured at a constant sampling time by the myoelectric potential measurement electrode 10 in chronological order and connecting them with a line.

FIG. 3 is a diagram showing an example of waveform data generated by the waveform data generation unit 20. The waveform data generation unit 20 generates waveform data as shown in FIG. 3 as an example. As for the waveform data, when the same subject swallows foods and drinks having the same flavor, the same waveform data is obtained, but when the same subject swallows foods and drinks with different flavors, different waveform data is obtained. The present invention was reached based on the discovery by the present inventors that the subtle differences in the flavors of foods and drinks can be understood by recognizing the differences. That is, when the present inventors were studying a new sensory evaluation method, the present inventors defined swallowing as a series of movements from the oral cavity to the stomach of the food and drink in the mouth. , A complex movement that occurs when multiple swallowing muscles in the pharynx and throat contract and relax in order, but the data obtained from the muscle activity at this time includes the amount of swallowing muscle activity and the type of muscle fiber (fast muscle, slow muscle). It is considered that various information such as muscle or intermediate muscle), movement form (dynamic, static) and muscle fatigue are included, and characteristic quantities of muscle activity such as amplitude, area, and frequency element of waveform data of muscle activity during swallowing. If it is possible to read the effect of the flavor of food and drink on the movement of swallowing by extracting and utilizing an analysis method such as an artificial neural network (mathematical model that imitates the nerve cell network in the brain), it is possible to read from the waveform data. The present invention has been completed with the technical idea that the similarity of the flavors of foods and drinks can be quantified, or the flavor of foods and drinks swallowed can be predicted based on the similarity. Physiological response measurements such as surface myoelectric potential measurements have the great advantage of being able to estimate the subject's true intentions, honest responses and subconscious minds (which affect decision making and behavioral choices), giving objectivity to sensory evaluation. Can be done.

The food and drink to which the present invention is applied is not particularly limited. Specific examples of foods and drinks include confectionery, beverages, marine products, processed foods, livestock products, vegetables, and seasonings. Flavors may be added to foods and drinks. A flavor is a compound or composition that can add or enhance a flavor to a food or drink by adding it to the food or drink. For example, there are perfume compounds, perfume compositions, extracts from animals and plants, and natural essential oils. Specific examples include "Patent Agency Gazette, Well-known and Conventional Techniques (Fragrances) Part II Food Fragrances, January 14, 2000. "Published daily", "Survey on the actual use of food flavor compounds in Japan" (2000 Health Science Research Report, Japan Flavor Industry Association, published in March 2001), and "Synthetic Perfume Chemistry and Product Knowledge" (2016) There are, but are not limited to, natural essential oils, natural fragrances, synthetic fragrances, etc. listed in (December 20, supplementary new edition issued, edited by Synthetic Fragrance Editing Committee, Kagaku Kogyo Nihonsha).

The food or drink subject to the sensory evaluation of the present invention has a texture (so-called "texture", which is a mechanical property such as hardness, tactile sensation, elasticity, and throatiness) and a subject in a plurality of sensory evaluations. It is preferable that the amount of food and drink (weight and / or volume) is unified. This is because the flavor (aroma and / or taste) of food and drink can be evaluated purely. Even if flavor is added to food and drink, the mechanical properties of food and drink are usually hard to change, and if the texture of food and drink and the amount of food and drink that the subject eats and drinks are unified, the influence of the flavor of food and drink will be affected by the subject. Appears obediently as a change in the surface myoelectric potential of. When the food and drink have different flavors depending on the temperature, it is preferable to set the temperature of the food and drink to the same temperature for evaluation.

The "flavor of food and drink" to be evaluated in the present invention refers to a sensation that can be perceived by at least the sense of smell, the sense of taste, or both. That is, at least aroma, taste, or both.

The parameter calculation unit 30 analyzes the waveform data generated by the waveform data generation unit 20 and calculates parameters related to swallowing. Examples of parameters for swallowing include temporal factors, quantitative factors, and frequency factors. As a temporal factor, the activity time of the swallowing muscle can be exemplified. The activity time is the time during which the swallowing muscle is active during swallowing. For example, a portion of the generated waveform data in which the swallowing muscle is active can be selected, and the width of that portion can be used as the muscle activity time. (In this specification, it may be referred to as a selection time width). Examples of the quantitative factors include the maximum amplitude of the waveform (maximum spectrum amplitude), the integrated value of the waveform (spectral area), and the root mean square (RMS). The maximum amplitude of the waveform (maximum spectrum amplitude) is the maximum muscle strength exerted during swallowing. The integrated value (spectral area) of the waveform is the amount of muscle activity, and is the total integrated value of the waves of the waveform data of the surface myoelectric potential. The root mean square (RMS) indicates the amount of muscle activity. Examples of frequency factors include power spectrum, power spectral density (PSD), and central power frequency. The frequency factor is determined by, for example, the Fourier transform. The power spectrum indicates which frequency the force is distributed in, and is a power spectrum before standardization at a unit frequency (1 Hz width). Power spectral density (PSD) is a spectral function normalized by a unit frequency. The central power frequency is used as an indicator of muscle fatigue. Alternatively, as a parameter related to swallowing, the brightness of the waveform data when the waveform data is imaged may be used. For example, the parameters are the spectral area, the maximum spectral amplitude, the muscle activity time, the power spectrum, the power spectral density, and the central power frequency, or the waveform data obtained by analyzing the waveform data generated by the waveform data generation unit 20. It is at least one of the brightness of the waveform data when imaged. The power spectrum and the power spectrum density are obtained by Fourier transforming or wavelet transforming the waveform data. Further, the brightness of the waveform data when the waveform data is imaged is obtained by quantifying the brightness of each pixel of the imaged waveform data. In the present embodiment, the waveform obtained from the electrode (hereinafter referred to as the upper right electrode) attached to the upper right (lower right region of the otogai) of the surface myoelectric potential measuring electrode of the swallowing muscle in the parameter calculation unit 30 Maximum spectral amplitude of the data, spectral area of the waveform data obtained from the upper right electrode, central power frequency of the waveform data obtained from the upper right electrode, power of a specific frequency (eg 20-45 Hz, 46-80 Hz, 81-350 Hz). The spectral density is calculated.

For these parameters, when a plurality of electrodes for measuring the surface myoelectric potential of the swallowing muscle are attached to the subject, the waveform data obtained by any of the electrodes is used as the waveform data used for calculating the parameters related to swallowing. It may be calculated using the waveform data obtained from all the electrodes.

The parameters of frequency factors such as power spectrum and PSD can also be calculated by separating each frequency bandwidth. The surface myoelectric potential data includes a large amount of myoelectric potential in a specific frequency band depending on the type of active muscle fiber. Therefore, the activity of the muscle that controls endurance is reflected in the power spectrum and PSD of the low frequency band, and the activity of the muscle that controls the instantaneous force is reflected in the power spectrum and PSD of the high frequency band. For example, at least one of each bandwidth of 20 to 45 Hz, 46 to 80 Hz, and 81 to 350 Hz is adopted. Each of these bandwidths is the frequency band that best represents the muscle activity of the slow, intermediate, and fast muscles.

FIG. 4 is a diagram provided for explaining the parameters calculated by the parameter calculation unit 30. FIG. 5 is a diagram showing a waveform obtained by full-wave rectifying the waveform shown in FIG. For example, if a part of the waveform data of FIG. 3 is taken out as the waveform data portion at the time of swallowing, the waveform as shown in FIG. 4 can be obtained. When this waveform is full-wave rectified, the waveform becomes as shown in FIG. For example, the parameter calculation unit 30 calculates the peak value 210 of the waveform of FIG. 5 to obtain the maximum amplitude of the spectrum, calculates the area surrounded by the envelope 200 surrounding the outer shape of the waveform to obtain the spectrum area, and determines the center of the waveform spectrum. The number of waves in the above is the central power frequency, the power spectrum density is used as the power value per unit frequency width, and the waveform width 220 is calculated and used as the muscle activity time.

The sensory evaluation data storage unit 40 obtains sensory evaluation data indicating how similar the flavor of the food or drink, which is the subject swallowed by the subject, to the flavor of the target food or drink, as a parameter calculation unit. Stored together with the parameters calculated by 30.

Sensory evaluation data is obtained by having the subject fill out a sensory evaluation questionnaire. The subjects to be collected when taking the sensory evaluation questionnaire can be selected according to the purpose. For example, if you want to make a prediction about a specific age, collect subjects with as many characteristics as possible in the specific age and suppress bias other than the age. Alternatively, people with different genders, ages, and tastes can be gathered evenly to obtain an evaluation formula that can be applied to all people and to predict the degree of similarity in flavor. The content of the sensory evaluation questionnaire can be arbitrarily set according to the content of the flavor to be evaluated. It is preferable that the answer to the degree of similarity in flavor of food and drink can be quantified. For example, it is desirable that the flavor of the food or drink to be evaluated can be scored for the degree of similarity with the flavor of the target food or drink, and the sensory evaluation data can be obtained as a score (numerical value). For example, the score increases according to the degree of similarity, "not very similar" is 0 points, "not similar" is 1 point, "not very similar" is 2 points, and "slightly similar". Can be used with a scale bar having 3 points, 4 points for "similar", and 5 points for "similar".

FIG. 6 is a diagram showing an example of a questionnaire as a means for acquiring sensory evaluation data. The questionnaire for sensory evaluation data shown in FIG. 6 numerically indicates the degree of similarity between the flavors of the food and drink swallowed by the subject and the flavor of the target food and drink in multiple stages. It is the converted data. The questionnaire for acquiring the sensory evaluation data shown in FIG. 6 is capable of acquiring data when the subject actually eats and drinks food and drink and evaluates what the subject feels on a 6-point scale from 0 to 5. Is. The mode of the means for acquiring the sensory evaluation data is not limited to the mode shown in FIG. For example, as shown in FIG. 6, the stages from dissimilarity to similarity may be 7 stages from -3 to 3 instead of 6 stages from 0 to 5. Alternatively, words such as "strongly similar", "moderately similar", and "weakly similar" may be selected as to how similar they are, or any other evaluation may be used. In addition, the response to the sensory evaluation questionnaire may be at the time when swallowing is completed or at other times.

The sensory evaluation data of food and drink and the parameters calculated by the parameter calculation unit 30 are stored in the sensory evaluation data storage unit 40 as data related to a set of the parameters calculated from the waveform data of the surface myoelectric potential and the sensory evaluation data. .. That is, the data set means a set in which the sensory evaluation data related to a certain food or drink and the parameters calculated from the waveform data generated when the food or drink are swallowed are linked.

The content of the sensory evaluation in the sensory evaluation data acquisition relates to sensations other than texture such as swallowing sensation, and at least to the degree of similarity to the flavor perceived based on the sensations felt by taste and / or olfaction. ..

For example, the content of the sensory evaluation in the acquisition of sensory evaluation data is the degree of similarity to the flavor of the target food or drink, the degree of similarity to the flavor of a specific food or drink including a specific product or a specific product, or a specific part. The similarity to the flavor of a specific ingredient, the similarity to the flavor of a specific component, the similarity to the flavor of a specific processing method, or the similarity to a flavor reminiscent of a specific impression.

Processing including specific fruits, vegetables, livestock products, fresh foods including marine products, retort foods, frozen foods, as examples of the similarity to the flavor of specific foods and drinks and specific products including specific products. Foods, seasonings, beverages, etc., alcoholic beverages, beer, plum wine, wine, chewy, butter, milk, pork bones, cream, coffee, tea, green tea, cocoa. Like chocolate, like chocolate, like honey, like maple syrup, like malt, like tea leaves, like mint, like yogurt, like soy milk, like beef, like pork, like chicken, like sheep, like nuts, like pepper, like herbs ..

Examples of the similarity to the flavor of a specific part include fruit juice, pulp, or peel, and livestock skin. Examples of the similarity of a specific component to the flavor include the fat feeling, oily feeling (oily feeling), and alcohol (ethanol) feeling of milk and animals. The degree of similarity to the flavor of a specific processing method includes a roasted feeling, a roasted feeling, a burnt feeling, a caramel feeling, and a brewed feeling such as coffee or tea. Examples of similarities to flavors that evoke a particular impression include flavor impressions such as lightness, body, moisture, freshness, light flavor, heavy flavor, and freshness. In addition, as another example of a specific impression, there is a specific culture or regionality such as Western style, Japanese style, Japanese style as an example of expressing the target product image of the developed product, and it is gorgeous or refreshing as expressing the mood. There are floral and green as examples of those that express the incense tone.

As described above, the sensory evaluation data is collected in advance from one or more subjects as follows and stored in the sensory evaluation data storage unit 40.

Specifically, when the content of the sensory evaluation is beer-like, the subject is asked to drink a beverage refrigerated at a predetermined temperature, and the flavor of the beverage is, for example, beer-like, as shown in FIG. As described above, the parameter calculated by the parameter calculation unit 30 from the waveform data generated when the subject swallows the beverage is associated with how many of the six levels the degree corresponds to. It is stored in the storage unit 40. This sensory evaluation is preferably performed by as many subjects as possible. Alternatively, when there is only one subject, it is preferable to perform the procedure as many times as possible. As a result, the parameters calculated by the parameter calculation unit 30 and the sensory evaluation data of many subjects are accumulated in the sensory evaluation data storage unit 40.

The correlation analysis unit 50 analyzes the correlation between the sensory evaluation data stored in the sensory evaluation data storage unit 40 and the parameters calculated by the parameter calculation unit 30. By associating the parameters stored in the sensory evaluation data storage unit 40 with the sensory evaluation data in which the degree of similarity of flavor is quantified in a plurality of steps, the correlation between the parameters and the sensory evaluation data is analyzed.

The correlation analysis unit 50 uses statistical analysis, machine learning, or deep learning to analyze the correlation between the sensory evaluation data and the parameters.

For statistical analysis, any one of multivariate analysis such as mean value test, discriminant analysis, multiple regression analysis, and PLS regression analysis is used. Machine learning includes linear regression, logistic regression, regression analysis using support vector machines (SVM), tree analysis using decision tree (CART), regression tree, random forest, gradient boosting tree, etc., Perceptron. For analysis using neural networks such as convolutional neural network (CNN), recurrence type neural network (RNN), residual network (ResNet), self-organization map (SOM), deep learning, and simple bays (naive bays). Bayesian analysis, k-nearest neighbor (KNN), hierarchical clustering (Euclidean distance, ward method, etc.), non-hierarchical clustering (k-means, etc.), clustering analysis using topic model (LDA, etc.), boosting, bagging Use one of ensemble analysis, association analysis, and collaborative filtering (item-based, user-based, etc.) analysis using. Deep learning also increases the number of hidden layers located between the input layer and the output layer. In order to improve the learning accuracy, it is preferable to increase the number of hidden layers as much as possible in consideration of calculation cost and practicality.

When analyzing the correlation, the correlation analysis unit 50 may select a parameter that increases the correlation coefficient, which will be described later. As a result, a more accurate evaluation formula can be obtained. Which parameter has a high correlation with the evaluation formula can be obtained by trial and error by experience or experiment.

The evaluation formula storage unit 60 stores the analysis result of the correlation between the parameters analyzed by the correlation analysis unit 50 and the sensory evaluation data as an evaluation formula.

If a map representing the correlation is created using a known mapping method based on the evaluation formula expressing the correlation, a map of the degree of similarity in the flavor of food and drink can be derived. For example, by plotting the flavor similarity derived from the parameters in the area corresponding to each similarity in the map, it can be visualized and easily understood visually and intuitively. By creating such a map, the degree of similarity to the flavor of the food target can be grasped at a glance. This map can be used for food and beverage advertising and marketing. The type of map is arbitrary, and the optimum one is selected according to the solution method and the desired visualization form. For example, contour maps, plot maps, and bi-plot maps can be exemplified as the types of maps.

FIG. 7 is a map obtained by plotting the correlation analysis result between the parameters acquired for the non-alcoholic beer under development and the sensory evaluation data regarding the beer-likeness. This figure shows an example in which a map is created by dividing the degree of feeling "beer-likeness" (that is, the degree of similarity to beer flavor) into three stages of "strong", "medium", and "weak".

FIG. 7 is created based on the statistics calculated when canonical discriminant analysis is used as a method for discriminating multiple groups. Specifically, plot the multi-group data on a multi-dimensional scatter plot, and find the discriminant axis (canonical axis 1) that minimizes the overlap of the plots of the two most distant groups in the multi-group. , A multi-group plot is projected on that axis to obtain a discrimination score of 1 (canonical level 1). Next, find the discrimination axis 2 (canonical axis 2) that minimizes the overlap of the other two groups that had a large overlap on that axis, and this time, the plot of multiple groups is projected orthogonally on that axis for discrimination. Obtain a score of 2 (canonical level 2). Then, the obtained canonical 1 and canonical 2 values are plotted. The map of FIG. 7 is obtained through such a procedure.

In each of the "strong", "medium", and "weak" circles in FIG. 7, regarding the beer-likeness, which is the objective variable, "strong" has a score of 1 or more and less than 4, and "medium" has a score of 4 or more and less than 7. , "Weak" was 7 or more and less than 10, and the beer-like score was obtained using the sensory evaluation questionnaire as shown in FIG. 6, with 1 being the lowest and 10 being the highest. The explanatory variables are those in the power spectral density (PSD) frequency band 0-500 Hz. Each circle is a probability ellipse calculated by the correlation analysis unit 50, and is a contour line indicating where the probability densities are equal. Specifically, it is a range that includes about 90% of the data plots belonging to each group, assuming a bivariate normal distribution.

As shown in this map, by creating a map, the positional relationship between the flavor of the food and drink under development and the flavor of the food and drink targeted is clarified on the map. For this reason, maps are of great help in the development of food and drink to bring the flavor of the food and drink under development closer to the flavor of the target food and drink.

FIG. 8 is a diagram showing an example of a plot diagram based on an evaluation formula obtained from the correlation between the parameters analyzed by the correlation analysis unit 50 and the sensory evaluation data.

In FIG. 8, as the degree of similarity to the flavor of the target food and drink, the beer-likeness (similarity to beer) of the non-alcoholic beer under development was targeted for sensory evaluation, and the non-alcoholic under development was selected from five subjects. The parameters calculated from the waveform data during swallowing of beer and the sensory evaluation data are acquired, the evaluation formula is obtained as a result of the analysis of the correlation, and the plot diagram relating to the similarity of flavors created based on the evaluation formula is used. is there. In this evaluation formula, the parameter is a variable, and if the parameter is introduced, the similarity can be predicted.

In FIG. 8, the vertical axis (beer-like learning) is the measured value of the “beer-likeness” of the sensory evaluation, and the horizontal axis (beer-likeness predicted value) is predicted by the prediction model constructed by the neural network based on the myoelectric data. It is a predicted value. This prediction model corresponds to the evaluation formula of the present invention. In preparing FIG. 8, the subjects were asked to score the "beer-likeness" of the sensory evaluation using a scale bar as shown in FIG. 6, assuming that 1 was the lowest and 10 was the highest. For the predicted value of "beer-likeness", the objective variable is the measured value of "beer-likeness" of the sensory evaluation, and the parameter which is the explanatory variable is 20-45 Hz and 46-80 Hz for the power spectral density (PSD) of the geniohyoid muscle. , 81-350Hz frequency band, muscle activity time, maximum amplitude (maximum muscle activity), root mean square (RMS), central power frequency, integral value (muscle activity) Got Then, the result of the plot of the black circle as shown in FIG. 8 showing the actually measured value and the predicted value of the beer-likeness based on this evaluation formula was obtained. In FIG. 8, the straight line is an auxiliary line for making it easy to see the correlation between the measured value of the similarity and the predicted value. This line indicates y (measured value) = x (predicted value), and the more plots closer to this line of y = x, the more accurate the evaluation formula is obtained.

FIG. 9 is a diagram showing a verification result of the evaluation formula of FIG. Specifically, an operation was performed to verify the presence or absence of overfitting in the neural network. This is because if there is overfitting, only the data used in the evaluation formula will be explained well, but it will not fit other data, that is, it will not be versatile and the prediction accuracy may be low. When making a discrimination from a prediction model using AI, it is preferable to verify whether or not overfitting is performed, and FIG. 9 shows the result of the verification. Specifically, in FIG. 8, a new set of data (hereinafter, also referred to as unknown data) is shown in FIG. 8 instead of the set of data (set of sensory evaluation data and parameters) used to obtain the evaluation formula of FIG. It is applied to the obtained evaluation formula. That is, in FIG. 9, the measured value, which is the value on the horizontal axis of each plot, represents the sensory evaluation data in the new set of data, and the predicted value, which is the value on the vertical axis of each plot, is in the evaluation formula of FIG. It is a value obtained by introducing a parameter in a new set of data. If the new set of data used in the verification of overfitting is not the set of data used for deriving the evaluation formula, it is arbitrary when the acquired data is used. That is, even if the evaluation formula is derived using a part of the acquired data set and the set of data not used for the derivation of the evaluation formula is used as unknown data for the verification of overtraining. Alternatively, a set of data newly acquired after calculating the evaluation formula may be used. As shown in FIG. 9, even if the data is unknown, the evaluation formula of FIG. 8 has high prediction accuracy. Therefore, it was confirmed that the evaluation formula of FIG. 8 is a versatile and useful evaluation formula, that is, a prediction model. .. The correlation coefficient of this evaluation formula was 0.89 when the set of data used in FIGS. 8 and 9 was combined and calculated. It can be seen that the reliability of the evaluation formula is quite high. This correlation coefficient represents the degree of approximation between the similarity (beer-likeness) predicted based on the parameters from the evaluation formula and the similarity (beer-likeness) obtained by the sensory evaluation.

That is, as shown in FIGS. 8 and 9, the food and drink evaluation device 100 of the present embodiment proves that the created prediction model is effective even for unknown data. That is, it is shown that the prediction accuracy for the data set used at the time of model creation (FIG. 8) and the prediction accuracy for unknown data (FIG. 9) are about the same. The correlation coefficient records a high numerical value of 0.8 or more even when each is calculated independently, indicating that the prediction model is accurate.

The configuration of the food and drink evaluation device 100 of the first embodiment is as described above.

[Operation of food and drink evaluation device (procedure of food and drink evaluation method)]
Next, the schematic operation of the food and drink evaluation device 100 will be described. FIG. 2 is a diagram showing a procedure of a food and drink evaluation method. The procedure of the food and drink evaluation method shown in FIG. 2 also shows the operation of the food and drink evaluation device 100. The food and drink evaluation device 100 operates according to this procedure. Since the operation of each part constituting the food / drink evaluation device 100 is as described above, the procedure of the food / drink evaluation method will be briefly described.

First, the surface myoelectric potential of the swallowing muscle when the subject swallows food and drink is measured by the myoelectric potential measuring electrode 10 (S100).

Next, the waveform data generation unit 20 generates waveform data of the surface myoelectric potential from the change in the surface myoelectric potential measured by the myoelectric potential measurement electrode 10 (S110).

The waveform data generated by the waveform data generation unit 20 is analyzed, and the parameter calculation unit 30 calculates parameters related to swallowing (S120).

Sensory evaluation data storage unit that shows how similar the flavor of food and drink swallowed by the subject is to the flavor of the target food and drink, together with the parameters calculated by the parameter calculation unit 30. 40 memorizes (S130).

The correlation analysis unit 50 analyzes the correlation between the sensory evaluation data stored in the sensory evaluation data storage unit 40 and the parameters (S140).

The evaluation formula storage unit 60 stores the analysis result of the correlation between the parameters analyzed by the correlation analysis unit 50 and the sensory evaluation data as an evaluation formula (S150).

By the above processing, the evaluation formula is finally stored in the evaluation formula storage unit 60. By substituting the parameters into the evaluation formula and predicting the flavor of food and drink using this evaluation formula, the flavor felt by the subject can be predicted with fairly high accuracy. In the next second embodiment, a food and drink evaluation device capable of predicting the flavor of food and drink using an evaluation formula will be described.

[Embodiment 2]
[Configuration of food and beverage evaluation device]
FIG. 12 is a block diagram of the food and drink evaluation device of the second embodiment. The food and drink evaluation device 300 includes a waveform data generation unit 20, a parameter calculation unit 30, an evaluation formula storage unit 60, an estimation unit 70, and a display unit 80. The same components as those of the food and drink evaluation device 100 of the first embodiment are designated by the same reference numerals.

The myoelectric potential measurement electrode 10 is connected to the waveform data generation unit 20. The waveform data generation unit 20, the parameter calculation unit 30, the evaluation formula storage unit 60, the estimation unit 70, and the display unit 80, which constitute the food and drink evaluation device 300, are provided in one housing. For example, the food and drink evaluation device 300 is provided in a general personal computer (PC).

The myoelectric potential measuring electrode 10 is attached to either the lower part of the chin or the anterior neck of the subject, or both the lower part of the chin and the anterior neck, as in the first embodiment. The myoelectric potential measuring electrode 10 measures a change in the surface myoelectric potential of the swallowing muscle when the subject swallows food and drink.

The waveform data generation unit 20 generates waveform data of the surface myoelectric potential from the change in the surface myoelectric potential of the swallowing muscle when the subject swallows food and drink, which is measured by the myoelectric potential measurement electrode 10. Specifically, the waveform data generation unit 20 generates waveform data by arranging the surface myoelectric potentials of the swallowing muscles measured at a constant sampling time by the myoelectric potential measurement electrode 10 in chronological order and connecting them with a line. The details of the operation of the waveform data generation unit 20 are the same as those in the first embodiment.

The parameter calculation unit 30 analyzes the waveform data generated by the waveform data generation unit 20 and calculates parameters related to swallowing. The parameters are the same as in the first embodiment, for example, the spectral area, the maximum spectral amplitude, the muscle activity time, the power spectrum, the power spectral density, and the central power obtained by analyzing the waveform data generated by the waveform data generation unit 20. It is at least one of the frequency or the brightness of the waveform data when the waveform data is imaged. The power spectrum and power spectral density are obtained by Fourier transforming or wavelet transforming the waveform data. Further, the brightness of the waveform data when the waveform data is imaged is obtained by quantifying the brightness of each pixel of the imaged waveform data.

The evaluation formula storage unit 60 stores the evaluation formula of the correlation between the parameter calculated by the parameter calculation unit 30 and the sensory evaluation data. The evaluation formula storage unit 60 stores the same evaluation formula as the evaluation formula storage unit 60 of the food and drink evaluation device 100 of the first embodiment. For example, an evaluation formula obtained as a result of repeated sensory evaluations of the flavor of a certain food or drink by a plurality of subjects with the same sensory evaluation content (that is, similarity to the flavor of the target food or drink) is stored. ..

The estimation unit 70 uses the parameters calculated by the parameter calculation unit 30 and the evaluation formula stored in the evaluation formula storage unit 60 to make the flavor of the food or drink swallowed by the subject relative to the flavor of the target food or drink. Estimate how similar they feel.

The display unit 80 digitizes and displays the similarity estimated by the estimation unit 70, or displays it in the form of mapping as shown in FIG. The display unit 80 is embodied as an LCD display or an organic EL display.

The configuration of the food and drink evaluation device 300 of the second embodiment is as described above.

[Operation of food and drink evaluation device (procedure of food and drink evaluation method)]
Next, the schematic operation of the food and drink evaluation device 300 will be described. FIG. 11 is a diagram showing a procedure of a food and drink evaluation method. The procedure of the food and drink evaluation method shown in FIG. 11 also shows the operation of the food and drink evaluation device 300. The food and drink evaluation device 300 operates according to this procedure. Since the operation of each part constituting the food / drink evaluation device 300 is as described above, the procedure of the food / drink evaluation method will be briefly described.

First, the surface myoelectric potential of the swallowing muscle when the subject swallows food and drink is measured by the myoelectric potential measuring electrode 10 (S400).

Next, the waveform data generation unit 20 generates waveform data of the surface myoelectric potential from the change in the surface myoelectric potential measured by the myoelectric potential measurement electrode 10 (S410).

The parameter calculation unit 30 calculates the parameters related to swallowing by analyzing the waveform data generated by the waveform data generation unit 20 (S420).

The estimation unit 70 uses the parameters calculated by the parameter calculation unit 30 and the evaluation formula stored in the evaluation formula storage unit 60 to make the flavor of the food or drink swallowed by the subject relative to the flavor of the target food or drink. To estimate how similar they feel (S430).

The display unit 80 digitizes and displays the similarity estimated by the estimation unit 70. Alternatively, it is displayed in the form of mapping as shown in FIG. 7 (S440).

By the above processing, finally, the display unit 80 displays how much the flavor of the food or drink swallowed by the subject feels to be similar to the flavor of the target food or drink. Food and beverage developers can determine how similar the flavor of the food and drink currently under development is to the flavor of the target food and drink by looking at the displayed similarity, and the direction of food development. Can be judged. As a display form of the degree of similarity, it is conceivable to display the degree of flavor (for example, beer-likeness) of the food and drink that the subject eats and drinks in 10 stages. For example, the beer feeling of the third beer under development that the subject drank is displayed by quantifying the degree of similarity, such as the flavor of the target food or drink (for example, beer-likeness) “9”.

Further, by having the subject swallow various foods and drinks, mapping the similarity obtained at that time, and displaying the mapping result on the display unit 80 as shown in FIG. 7, the positions of various foods and drinks are displayed. You can grasp the relationship at a glance.

As described above, in the food and drink evaluation device and the food and drink evaluation method of the present invention, the evaluation formula obtained as a result of the analysis by the correlation analysis unit 50 and the evaluation formula stored in the evaluation formula storage unit 60 are the flavors of the food and drink. It can be seen that the accuracy of the prediction of is greatly affected.

We asked the subjects to actually sample the foods and drinks under development, and sensory-evaluated the degree of similarity in how the flavors of the foods and drinks that the subjects tasted were similar to the flavors of the target foods and drinks. Next, the similarity that the subject answered was predicted using the evaluation formula from the parameters obtained when the subject tasted the food and drink. Then, an experiment was conducted in which the correlation coefficient between the similarity answered by the sensory evaluation of the subject and the similarity answered by the food and drink evaluation device of the present invention was calculated.

This experiment was performed as follows. This experiment will be described separately for Example 1 and Example 2.

[Example 1]
In Example 1, the flavor of food and drink targeting "butter-likeness" is defined as butter-likeness from sensory evaluation data of butter-likeness (similarity to butter flavor) and waveform data obtained from the surface myoelectric potential of the swallowing muscle. This is an example in which the correlation with the calculated parameters is analyzed to obtain an evaluation formula. In this example, as food and drink, commercially available margarine and flavored margarine containing a butter flavor manufactured by Hasegawa Fragrance Co., Ltd. were used.

First, in analyzing the correlation, a commercially available margarine (unscented product) was applied to the crumb portion of 5 grams of bread, and the subjects were asked to sample it. A myoelectric potential measuring electrode 10 was attached to the neck (lower part of the man and the anterior neck) of the subject, and the swallowing myoelectric potential when swallowing a commercially available margarine (unscented product) was measured to obtain waveform data.

On the other hand, the subjects were asked to perform a sensory evaluation of the "butter-likeness" of commercially available margarine (unscented product). The sensory evaluation was based on the score related to "butter-likeness", and the same questionnaire as in FIG. 6 was used.

Next, a flavored margarine was applied to the crumb portion of 5 grams of bread, and the subjects were asked to sample it. The myoelectric potential measuring electrode 10 was attached to the subject's neck at the same position as above, and the swallowing myoelectric potential when swallowing the perfume margarine was measured to obtain waveform data.

On the other hand, the subjects were asked to perform a sensory evaluation of the "butter-likeness" of the flavored margarine in the same manner as the commercially available margarine (unflavored product).

Then, from the obtained waveform data, parameters related to swallowing were calculated. The parameters are 20-45Hz, 46-80Hz, 81-350Hz frequency bands, muscle activity time, maximum amplitude (maximum muscle activity), root mean square (RMS) for the power spectral density (PSD) of the geniohyoid muscle. , Central power frequency, integrated value (muscle activity) was used. The data of commercially available margarine (unscented product) was corrected so that the median value was 1, and normalized.

For each margarine-coated bread, the correlation analysis between the calculated parameters and the sensory evaluation score was performed by a neural network, and the evaluation formula was calculated. The specifically obtained evaluation formula is as follows. In the evaluation formula, TanH: hyperbolic tangent is the following function.
TanH (x) = ex-ex / ex + ex.
"Beer-ness" =

H1_12 =

H1_22 =

H1_22 =

In the case of Example 1, a correlation coefficient of 0.67 was obtained. This correlation coefficient indicates the degree of similarity between the flavor similarity (butter-likeness) predicted from the parameters related to swallowing by the neural network and the similarity actually obtained by the sensory evaluation.

[Example 2]
In Example 2, the flavor of the target food and drink is beer-like, and the same as in Example 1 except that the data on the lower right side of the otogai is used for the correlation analysis, the beer-like sensory evaluation score and the surface muscle of the swallowing muscle. This is an example in which the correlation with the parameter obtained from the potential is analyzed to obtain an evaluation formula.

First, in analyzing the correlation, the subjects swallowed 60 cc of non-alcoholic beer under development that had been refrigerated at a predetermined temperature. A myoelectric potential measuring electrode 10 was attached to the subject's neck (lower right side of the man and the anterior neck), and the swallowing myoelectric potential when swallowing the beer under development was measured to obtain waveform data. The fact that the myoelectric potential measurement electrode 10 is attached to the right side of the lower part of the chin means that the electrode is attached to the region on the right half of the subject in the lower part of the chin.

On the other hand, the subjects were asked to perform a sensory evaluation of the "beer-likeness" of the non-alcoholic beer under development. The sensory evaluation was based on the score related to "beer-likeness", and the same questionnaire as in FIG. 6 was used.

Next, the subjects swallowed 60 cc of water refrigerated at the same temperature as the beer under development. A myoelectric potential measuring electrode 10 was attached to the subject's neck at the same position as above, and the swallowing myoelectric potential when swallowing beer for sale was measured to obtain waveform data.

Then, from the obtained waveform data, parameters related to swallowing were calculated using the data on the lower right side of the chin. The parameters are 20-45Hz, 46-80Hz, 81-350Hz frequency bands, muscle activity time, maximum amplitude (maximum muscle activity), root mean square (RMS) for the power spectral density (PSD) of the geniohyoid muscle. , Central power frequency, integrated value (muscle activity) was used. The water data was corrected so that the median value was 1, and normalization was performed.

For each beverage, the correlation analysis between the calculated parameters and the sensory evaluation data was performed by a neural network, and the evaluation formula was calculated. The specifically obtained evaluation formula is as follows. In the evaluation formula, TanH: hyperbolic tangent is the following function.
TanH (x) = ex-ex / ex + ex.
"Beer-ness" =

H1_1 =

H1-2 =

H1_3 =

8 and 9 above are diagrams showing the analysis result of the correlation of Example 2 and the verification result of overfitting thereof. As shown in the figure, it can be seen that the correlation between the evaluation formula and the predicted value of the subject is quite good. In the case of Example 2, a correlation coefficient of 0.89 was obtained. This correlation coefficient indicates the degree of similarity between the flavor similarity (beer-likeness) predicted from the parameters related to swallowing by the neural network and the similarity actually obtained by the sensory evaluation.

As described above, according to Examples 1 and 2, the reliability of the evaluation formula is considerably high, and by using the present invention, the development of foods and drinks can be efficiently performed, and the development with high accuracy can be performed. I understand.

As described above, in the food and drink evaluation device and the food and drink evaluation method of the first embodiment, an evaluation formula that can predict the sensory evaluation of the subject can be obtained from the parameters obtained from each subject and the sensory evaluation of each subject. In the food and drink evaluation device and the food and drink evaluation method of 2, the evaluation formula can be used to estimate the sensory evaluation of the subject from the parameters obtained from the subject. Therefore, according to the food and drink evaluation device and the food and drink evaluation method of the present invention, the food and drink evaluation device and the food and drink evaluation method that are objective and useful for the development of more accurate food and drink have been longing for. It can meet the expectations of product developers.

Further, according to the food and drink evaluation device and the food and drink evaluation method of the present invention, by displaying the map as shown in FIG. 7, an index for bringing the flavor of the food and drink closer to the target flavor of the food and drink can be obtained. It becomes easy to develop foods and drinks that are objective and more accurate. Therefore, product development of foods and drinks can be performed efficiently, and the cost of product development is reduced.

Although the food and drink evaluation device and the food and drink evaluation method of the present invention have been described above, the present invention is not limited to the description of the above embodiments and examples, and is modified within the technical scope of the present invention. Of course it is possible.

10 Myoelectric potential measurement electrode,
20 Waveform data generator,
30 Parameter calculation unit,
40 Sensory evaluation data storage unit,
50 Correlation Analysis Department,
60 Evaluation type storage unit,
70 estimation part,
80 display section,
100, 300 food and drink evaluation device.

Claims (17)

A waveform data generator that generates waveform data of the surface myoelectric potential from changes in the surface myoelectric potential of the swallowing muscle when the subject swallows food and drink,
A parameter calculation unit that analyzes the waveform data generated by the waveform data generation unit and calculates parameters related to swallowing.
Sensory evaluation data that stores sensory evaluation data indicating how similar the flavor of food and drink swallowed by the subject is to the flavor of the target food and drink together with the parameters calculated by the parameter calculation unit. Memory and
A correlation analysis unit that analyzes the correlation between the parameters calculated by the parameter calculation unit and the sensory evaluation data stored in the sensory evaluation data storage unit, and a correlation analysis unit.
An evaluation expression storage unit that stores the analysis result of the correlation between the parameter analyzed by the correlation analysis unit and the sensory evaluation data as an evaluation expression, and an evaluation expression storage unit.
Food and beverage evaluation device. A waveform data generator that generates waveform data of the surface myoelectric potential from changes in the surface myoelectric potential of the swallowing muscle when the subject swallows food and drink,
A parameter calculation unit that analyzes the waveform data generated by the waveform data generation unit and calculates parameters related to swallowing.
An evaluation formula storage unit that stores an evaluation formula for the correlation between the parameters and the sensory evaluation data,
Using the parameters calculated by the parameter calculation unit and the evaluation formula stored in the evaluation formula storage unit, the flavor of the food or drink swallowed by the subject becomes the flavor of the target food or drink swallowed by the subject. On the other hand, the estimation unit that estimates how similar it feels,
Food and beverage evaluation device with. The waveform data generation unit
The first or second claim, wherein the change in the surface myoelectric potential of the swallowing muscle when the subject swallows food or drink is obtained from the myoelectric potential measuring electrode attached to the lower part of the otogai and / or the anterior neck of the subject. Food and beverage evaluation device. The parameter calculation unit
By analyzing the waveform data, among the spectral area, the maximum spectral amplitude, the muscle activity time, the power spectrum, the power spectral density, and the central power frequency, or the brightness of the waveform data when the waveform data is imaged. The food and drink evaluation device according to claim 1 or 2, wherein at least one of the parameters is obtained. The food and drink evaluation device according to claim 4, wherein the power spectrum and the power spectrum density are obtained by performing a Fourier transform or a wavelet transform on the waveform data. The food and drink evaluation device according to claim 4, wherein the brightness of the waveform data when the waveform data is imaged is obtained by quantifying the brightness of each pixel of the imaged waveform data. The sensory evaluation data is
It is data obtained by quantifying the degree of similarity of the degree of similarity between the flavor of the food and drink swallowed by the subject to the flavor of the target food and drink swallowed by the subject in a plurality of steps. The food and drink evaluation device according to claim 1 or 2. The content of the sensory evaluation in the sensory evaluation data is
The food and drink evaluation device according to claim 7, wherein the food and drink evaluation device relates to a sensation other than the texture such as a swallowing sensation, and relates to a flavor including a sensation felt by a taste, an olfaction, and / or a cold or warm sensation. The content of the sensory evaluation in the sensory evaluation data is
Similarity to the flavor of a specific food or drink or a specific product, the similarity to the flavor of a specific part, the similarity to the flavor of a specific ingredient, the similarity to the flavor of a specific processing method, One of the similarities to the flavor that evokes a particular impression,
The food or drink evaluation device according to claim 7, wherein the specific food or drink is a fruit, vegetable, livestock product, fresh food including marine products, retort food, processed food including frozen food, seasoning, or beverage. The content of the sensory evaluation in the sensory evaluation data is
Similarity to the flavor of certain foods and drinks and certain products, including certain products,
Similarity to the flavor of a particular part,
Similarity to the flavor of a particular ingredient,
Similarity to the flavor of a particular processing method,
Similarity to flavors that evoke a particular impression,
The food and drink evaluation device according to claim 9, which is selected from at least one of the two. The correlation analysis unit
By associating the sensory evaluation data in which the degree of similarity with the parameter stored in the sensory evaluation data storage unit is quantified in a plurality of steps, the correlation between the parameter and the sensory evaluation data is analyzed. Item 1. The food and drink evaluation device according to item 1. The food and drink evaluation device according to claim 1, wherein the correlation analysis unit uses statistical analysis, machine learning, or deep learning to analyze the correlation between the sensory evaluation data and the parameters. The statistical analysis uses any of multivariate analysis such as mean value test, discriminant analysis, multiple regression analysis, and PLS regression analysis.
The machine learning includes linear regression, logistic regression, regression analysis using a support vector machine (SVM), tree analysis using a decision tree (CART), a regression tree, a random forest, a gradient boosting tree, etc., Perceptron, Analysis using neural networks such as convolutional neural network (CNN), recurrence type neural network (RNN), residual network (ResNet), self-organization map (SOM), deep learning, and simple bays (naive bays) were used. Bayesian analysis, k-nearest neighbor (KNN), hierarchical clustering (Euclidean distance, ward method, etc.), non-hierarchical clustering (k-means, etc.), clustering analysis using topic model (LDA, etc.), boosting, bagging, etc. Using one of the following: Ensemble analysis, association analysis, or collaborative filtering (item-based, user-based, etc.) analysis using
The food and drink evaluation device according to claim 12, wherein deep learning has a plurality of hidden layers located between an input layer and an output layer. The estimation unit
When estimating the degree of similarity between the flavor of the food and drink swallowed by the subject and the flavor of the target food and drink swallowed by the subject, the evaluation formula is used. The food and drink evaluation device according to claim 2, wherein a parameter having a high correlation with the above is selected and the degree of similarity is estimated. The food and drink evaluation device according to claim 14, further comprising a display unit that digitizes or displays the similarity estimated by the estimation unit as a mapping. The stage of generating waveform data of the surface myoelectric potential from the change of the surface myoelectric potential of the swallowing muscle when the subject swallows food and drink, and
The stage of analyzing the generated waveform data and calculating the parameters related to swallowing,
A step of storing sensory evaluation data indicating how similar the flavor of the food or drink swallowed by the subject to the flavor of the target food or drink is stored together with the calculated parameters.
At the stage of analyzing the correlation between the stored sensory evaluation data and the parameters,
A step of storing the analysis result of the correlation between the analyzed parameter and the sensory evaluation data as an evaluation formula, and
Food and beverage evaluation methods, including. The stage of generating waveform data of the surface myoelectric potential from the change of the surface myoelectric potential of the swallowing muscle when the subject swallows food and drink, and
The stage of analyzing the generated waveform data and calculating the parameters related to swallowing,
Using the calculated parameters and the evaluation formula, it is estimated how much the flavor of the food and drink swallowed by the subject feels similar to the flavor of the target food and drink swallowed by the subject. And the stage to do
Food and beverage evaluation method including.