Meal evaluation device, meal evaluation method, and meal evaluation program

Abstract

A contributing factor of an evaluation subject is acquired, and a predicted result of a sensory evaluation value of the evaluation subject with respect to a meal is acquired from the contributing factor of the evaluation subject, by using a machine learning model in which the contribution factor is used as an explanatory variable and the sensory evaluation value with respect to the meal is used as an objective variable.

Description

Dietary evaluation device, dietary evaluation method, and dietary evaluation program

[0001] The present invention relates to a meal evaluation device, a meal evaluation method, and a meal evaluation program.

[0002] Patent Literature 1 discloses a technology for generating user-based recommendations from user behavior and attribute information, and recommending items that match the user's preferences.

[0003] Patent Document 2 discloses a technology that generates appropriate advice messages based on the diet and health condition of a subject, thereby improving motivation for a health management curriculum.

[0004] Patent Document 3 discloses a technology that can manage health taking into consideration individual differences and outputs food components that improve or worsen vital data.

[0005] Patent Document 4 discloses a technology that outputs a blood glucose level prediction after eating a meal and the calorie intake that will keep it within a standard, and allows a user to select a meal menu based on the blood glucose level prediction and preferences.

[0006] JP 2018-181135 A JP 2017-62854 A JP 2022-91962 A JP 2012-181804 A

[0007] However, conventional inventions have had the problem that they are unable to predict what kind of diet an individual should eat to prevent excessive overeating and optimize the content of their meals, so that they will be healthy and feel full and satisfied.

[0008] The present invention has been made in consideration of the above-mentioned problems, and aims to provide a meal evaluation device, a meal evaluation method, and a meal evaluation program that can provide information and communication technology for predicting subjective post-meal sensations using meal content and individual difference information as explanatory variables.

[0009] In order to solve the above-mentioned problems and achieve the objective, a meal evaluation device is provided which comprises a memory unit and a control unit, wherein the memory unit comprises a meal storage means which stores a machine learning model in which contributing factors including user characteristics, environmental characteristics and / or dietary characteristics are used as explanatory variables and a sensory evaluation value for a meal is used as a target variable, and the control unit comprises a factor acquisition means which acquires the contributing factors of the person being evaluated, and a prediction result acquisition means which uses the machine learning model to acquire a prediction result of the sensory evaluation value for the meal of the person being evaluated from the contributing factors of the person being evaluated.

[0010] In addition, in the diet evaluation device according to the present invention, the sensory evaluation value is a subjective sensory evaluation value, and / or an amount of an in vivo factor related to the subjective sensory evaluation value and / or a measurement result.

[0011] In addition, in the meal evaluation device according to the present invention, the subjective sensory evaluation value is hunger, fullness, satisfaction, appetite, bloating, heartburn, stomach discomfort, snacking desire, drowsiness, or fatigue.

[0012] In the diet evaluation device according to the present invention, the amount of the biological factor is a blood glucose level transition and / or a blood hormone concentration.

[0013] In the diet evaluation device according to the present invention, the blood hormone is leptin, ghrelin, insulin, cortisol, and / or a gastrointestinal hormone.

[0014] In the diet evaluation device according to the present invention, the measurement results are electroencephalograms.

[0015] In addition, in the diet evaluation device of the present invention, the machine learning model is a model consisting of at least one of multiple regression analysis, random forest, neural network, XGBoost, LightGBM, support vector regression, and linear regression including ElasticNet.

[0016] In the meal evaluation device according to the present invention, the sensory evaluation value is a score obtained by sensory evaluation.

[0017] In the meal evaluation device according to the present invention, the sensory evaluation value is a VAS score.

[0018] In addition, in the diet evaluation device according to the present invention, the contributing factors are factors acquired from at least one of a diet record, a questionnaire, an internet log, and device measurement.

[0019] In addition, in the dietary evaluation device of the present invention, the dietary characteristics are the amounts of at least one of dietary fiber and carbohydrates including sugars, proteins, lipids, salt, minerals, vitamins, amino acids, micronutrients, and organic acids in the diet.

[0020] In addition, in the meal evaluation device according to the present invention, the prediction result acquisition means is further characterized in that it designs a suggested meal content based on the prediction result and displays the suggested meal content.

[0021] Furthermore, the dietary evaluation method of the present invention is a dietary evaluation method to be executed by a dietary evaluation device having a memory unit and a control unit, wherein the memory unit comprises a dietary storage means for storing a machine learning model in which contributing factors including user characteristics, environmental characteristics and / or dietary characteristics are used as explanatory variables and a sensory evaluation value for the meal is used as a target variable, and the dietary evaluation method includes a factor acquisition step executed in the control unit for acquiring the contributing factors of the person being evaluated, and a prediction result acquisition step for acquiring a prediction result of the sensory evaluation value for the meal of the person being evaluated from the contributing factors of the person being evaluated.

[0022] In addition, the meal evaluation program of the present invention is a meal evaluation program to be executed by a meal evaluation device having a memory unit and a control unit, wherein the memory unit comprises a meal memory means for storing a machine learning model in which contributing factors including user characteristics, environmental characteristics and / or meal characteristics are used as explanatory variables and a sensory evaluation value for the meal is used as a target variable, and the control unit executes a factor acquisition step of acquiring the contributing factors of the person being evaluated, and a prediction result acquisition step of using the machine learning model to acquire a predicted result of the sensory evaluation value for the meal of the person being evaluated from the contributing factors of the person being evaluated.

[0023] The present invention has the effect of being useful in preventing lifestyle-related diseases such as diabetes by optimizing carbohydrate intake in meals. Furthermore, the present invention has the effect of enabling the design of meal contents and products that provide a high level of satisfaction and satiety for each individual, thereby enabling stress-free continuation of dietary improvements. Furthermore, the present invention has the effect of defining postprandial satiety and satiety and blood glucose level fluctuations based on meal contents and blood glucose level fluctuations. Furthermore, the present invention has the effect of avoiding excessive dietary corrections and providing information for providing a diet that is easy for the individual to continue without straining themselves. Furthermore, the present invention has the effect of enabling stress-free continuation of dietary optimization, leading to the prevention of lifestyle-related diseases (and ultimately to an extension of an individual's healthy lifespan and a reduction in medical costs).

[0024] FIG. 1 is a diagram showing an example of meal evaluation processing in this embodiment. FIG. 2 is a block diagram showing an example of the configuration of a meal evaluation device in this embodiment. FIG. 3 is a flowchart showing an example of meal evaluation processing in this embodiment. FIG. 4 is a diagram showing an example of meal evaluation processing in this embodiment. FIG. 5 is a diagram showing an example of meal evaluation processing in this embodiment. FIG. 6 is a diagram showing an example of meal evaluation processing in this embodiment. FIG. 7 is a diagram showing an example of meal evaluation processing in this embodiment. FIG. 8 is a diagram showing a feeling of fullness in this embodiment. FIG. 9 is a diagram showing a feeling of hunger in this embodiment. FIG. 10 is a diagram showing prediction accuracy verification in this embodiment.

[0025] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described in detail with reference to the accompanying drawings. However, the present invention is not limited to this embodiment.

[0026] [1. Overview] First, an overview of the present invention will be described with reference to Fig. 1. Fig. 1 is a diagram showing an example of a meal evaluation process in this embodiment.

[0027] Conventionally, there have been techniques for predicting postprandial blood glucose levels based on meal contents and blood glucose fluctuations, and techniques for optimizing meal contents based on meal contents and an individual's blood pressure, weight, and muscle mass.

[0028] However, the food that satisfies and fills people varies from time to time, and if the quantity or quality of food is insufficient, stress builds up, making it difficult to continue with the optimized suggested meal plan.

[0029] Therefore, as shown in FIG. 1 , this embodiment provides a mechanism for predicting satiety and the like through machine learning using food records, internet logs, survey results, and device measurement results as inputs. Here, in this embodiment, the user's schedule and external factors such as weather may also be added to the inputs. As a result, this embodiment makes it possible to obtain information to support behavioral changes in consumers based on personal device measurement results, including blood glucose levels.

[0030] [2. Configuration of the meal evaluation device 100] The meal evaluation device 100 according to this embodiment can be configured in any unit, functionally or physically, by being distributed or integrated (either as a stand-alone type or a system type). In this embodiment, an example of the configuration of the meal evaluation device 100 will be described with reference to Fig. 2. Fig. 2 is a block diagram showing an example of the configuration of the meal evaluation device 100 according to this embodiment.

[0031] 2 , the meal evaluation device 100 may be an information processing device such as a personal computer or a workstation. The meal evaluation device 100 includes a control unit 102, a storage unit 106, and an input / output unit 112, and the units included in the meal evaluation device 100 are communicably connected via any communication path. The meal evaluation device 100 is communicably connected to other devices via a network 300.

[0032] The input / output unit 112 may have a function for inputting and outputting data (I / O). Here, the input / output unit 112 may be, for example, a key input unit, a touch panel, a control pad (e.g., a touch pad, a game pad, etc.), a mouse, a keyboard, a microphone, etc. The input / output unit 112 may also be a display unit (e.g., a display, monitor, and touch panel made of liquid crystal or organic electroluminescence, etc.) that displays (input / output) information of application software, etc. The input / output unit 112 may also be an audio output unit (e.g., a speaker, etc.) that outputs audio information as audio. The input / output unit 112 may also be an image input unit (e.g., a camera, etc.) that records images (still images and videos) captured by an imaging element such as a CCD image sensor or a CMOS image sensor as digital data. The input / output unit 112 may also be a fingerprint sensor, a camera (e.g., an infrared camera, etc.) that can be used for iris authentication or face authentication, etc., and / or a biometric sensor such as a vein sensor.

[0033] The storage unit 106 stores various databases, tables, and / or files. The storage unit 106 stores computer programs that cooperate with an operating system (OS) to issue commands to a central processing unit (CPU) to perform various processes. The storage unit 106 may be, for example, a random access memory (RAM), a read-only memory (ROM), a hard disk drive (HDD), and / or a solid state drive (SSD). The storage unit 106 may store image data recorded by the input / output unit 112, data received via the network 300, and / or input data input via the input / output unit 112. The storage unit 106 conceptually includes a meal database 106a.

[0034] The meal database 106a stores meal data. The meal database 106a may also store a machine learning model related to meals. The meal database 106a may also store a machine learning model using contributing factors, including user characteristics, environmental characteristics, and / or dietary characteristics, as explanatory variables and a sensory evaluation value for the meal as a response variable. The sensory evaluation value may be a subjective sensory evaluation value and / or an amount of a biological factor related to the subjective sensory evaluation value and / or a measurement result. The subjective sensory evaluation value may be hunger, satiety, satisfaction, appetite, bloating, heartburn, stomach discomfort, snacking desire, sleepiness, or fatigue. The amount of a biological factor may be a blood glucose level transition and / or a blood hormone concentration. The blood hormone may be leptin, ghrelin, insulin, cortisol, and / or a gastrointestinal hormone (e.g., GLI-1, GIP, or cholecystokinin). The measurement result may be an electroencephalogram. The machine learning model may be a model including at least one of multiple regression analysis, random forest, neural network, XGBoost, LightGBM, support vector regression, and linear regression including ElasticNet. The sensory evaluation value may be a visual analogue scale (VAS) score. The contributing factors may be factors obtained from at least one of a food record, a questionnaire, an internet log, and device measurement. The dietary characteristics may be the amount of at least one of dietary fiber and carbohydrates (including sugars), protein, lipids, salt, minerals, vitamins, amino acids, micronutrients, and organic acids in the diet. The diet database 106a may store the contributing factors, the predicted results of the sensory evaluation value, and the suggested meal contents.

[0035] Here, the sensory evaluation value may be a score obtained by a sensory evaluation. The scale used for the sensory evaluation may be an ordinal scale or an interval scale. The sensory evaluation value may be a score obtained by a Labeled Magnitude Scale (LMS), a Semantic Differential (SD), a Qualitative Descriptive Analysis (QDA), or a scoring method.

[0036] The control unit 102 is a CPU or the like that comprehensively controls the diet evaluation device 100. The control unit 102 has an internal memory for storing control programs such as an OS, programs that define various processing procedures, required data, etc., and executes various information processing operations based on these stored programs. Functionally, the control unit 102 conceptually includes a model construction unit 102a, a factor acquisition unit 102b, and a prediction result acquisition unit 102c.

[0037] The model construction unit 102a constructs a machine learning model. Here, the model construction unit 102a may train the machine learning model using training data. The model construction unit 102a may also register data sets that serve as explanatory variables and target variables for the machine learning in the meal database 106a.

[0038] The factor acquiring unit 102b acquires the contributing factors of the person to be evaluated. Here, the factor acquiring unit 102b may register the contributing factors in the diet database 106a.

[0039] The prediction result acquisition unit 102c acquires a prediction result of the sensory evaluation value of the subject's meal. Here, the prediction result acquisition unit 102c may use a machine learning model to acquire a prediction result of the subject's sensory evaluation value of the subject's meal from the subject's contributing factors. The prediction result acquisition unit 102c may also design a suggested meal content based on the prediction result and display the suggested meal content.

[0040] 3. Meal Evaluation Process An example of the meal evaluation process according to this embodiment will be described with reference to Fig. 3 to Fig. 7. Fig. 3 is a flowchart showing an example of the meal evaluation process according to this embodiment.

[0041] As shown in Figure 3, when the user sets the person to be evaluated and contributing factors including user characteristics, environmental characteristics, and / or dietary characteristics via the input / output unit 112, the factor acquisition unit 102b acquires the contributing factors of the person to be evaluated and registers the contributing factors in the diet database 106a (step SA-1).

[0042] Then, the prediction result acquisition unit 102c uses the machine learning model to acquire the subjective sensory evaluation value for the subject's meal and the prediction results of the blood glucose level trend from the subject's contributing factors (step SA-2).

[0043] Then, the prediction result acquisition unit 102c designs a proposed meal plan based on the prediction result, displays the proposed meal plan on the input / output unit 112 (step SA-3), and ends the process.

[0044] Here, in this embodiment, an appropriate acquisition method or preprocessing (data input burden reduction processing) may be applied to the contributing factors of the sensory evaluation value, including user characteristics, environmental characteristics, and / or dietary characteristics, which enables a reduction in the amount of data input by the user and the burden thereof.

[0045] For example, in the data entry burden reduction process (1), VAS transition data for a certain time period after a meal may be used as the objective variable for predicting subsequent VAS transition data. Here, since the VAS transition data reflects user characteristics, environmental characteristics, and / or dietary characteristics, using this data can reduce the amount of data entry for user characteristics, environmental characteristics, and / or dietary characteristics.

[0046] Here, in the data entry burden reduction process (1), at least one autoregressive model such as a long-term memory or a moving average model such as an ARMA model (autoregressive moving average model), an ARIMA model (autoregressive sum moving average model), or a SARIMA model (seasonal autoregressive sum moving average model) may be applied as a machine learning model suitable for predicting time series data.

[0047] Furthermore, in the data input burden reduction process (2), the dietary characteristics set by the user do not necessarily need to be detailed data, and data related to the number of ingredients or menu items, or the amount of nutrients, particularly data limited to only ingredients that contribute greatly to the sensory evaluation value, may be used. In this way, when limited dietary characteristics are used in the data input burden reduction process (2), the average or median value of the actual dietary characteristics when the data is input may be used as a provisional value, and a preset input value may be used based on the nutritional value of the ingredients in a food composition table, etc.

[0048] In addition, as a data input burden reduction process (3), limited user characteristics, environmental characteristics, and / or dietary characteristics are acquired, and random numbers are generated within that range for the input data to acquire a sensory evaluation value, thereby making it possible to acquire the range, upper limit, and lower limit of the variation in the sensory evaluation value.

[0049] Furthermore, as the data entry burden reduction process (4), a data set relating to the contributing factors of the sensory evaluation value, including user characteristics, environmental characteristics, and / or dietary characteristics, and any one of the sensory evaluation values, may be utilized to perform clustering or classification, and then determine which cluster a new user is similar to, thereby estimating the sensory evaluation value of the user.

[0050] Here, in the data input burden reduction process (4), the similarity of the clusters may be determined without inputting contributing factors of the sensory evaluation value, including user characteristics, environmental characteristics, and / or dietary characteristics, or by describing or illustrating the characteristics of the clusters themselves, or by setting a person or character that represents the characteristics of the cluster and allowing the user to select one.

[0051] In this embodiment, the machine learning model used to predict satiety and the like and the machine learning model used to propose meals or develop foods based on the prediction results may be combined separately. Here, the proposal of meals or development of foods may be evaluated not only based on the prediction results of the postprandial sensory evaluation value, but also in combination with prediction information from other machine learning models or existing information including nutritional information and health effects.

[0052] A specific example of the meal evaluation process in this embodiment will be described with reference to Figures 4 to 7. Figures 4 to 7 are diagrams showing an example of the meal evaluation process in this embodiment.

[0053] As shown in FIG. 4 , in this embodiment, a machine learning model is used to predict human characteristics (e.g., blood glucose fluctuations (Patrick Wyatt et al., Nat Metab., 2021, Postprandial glycaemic dips predict appetite and energy intake in healthy individuals.), preferences (Wenting Yin et al., Appetite, 2017. Effects of aroma and taste, independently or in combination, on appetite sensation and subsequent food intake. ), amount and type of exercise (Broom DR et al., Am J Physiol Regul Integr Comp Physiol., 2009. Influence of resistance and aerobic exercise on hunger, circulating levels of acylated ghrelin, and peptide YY in healthy males), medication status (I Chapman et al., Diabetologia, Effect of medication on satiety and food intake in obese subjects and subjects with type 2 diabetes. ), metabolic capacity (Lucia Camacho-Barcia et al., Nutrients. 2021. Circulating Metabolites Associated with Postpranial Satiety in Overweight / Obese Participants: The SATIN Study), Digestive Capacity (Christine Feinle et al., Am J Physiol Gastrointest Liver Physiol., 2003. Effects of fat digestion on appetite, APD motility,and got hormones in response to dual fat infusion in humans. ), gastric emptying (Daniel Gonzalez-Izundegui et al., Obesity (Silver Spring). 2021. Association of gastric emptying with postprandial appetite and satiety sensations in obesity), sleep quality and duration (Charli Sargent et al., Int J Environ Res Public Health., 2016. Daily Rhythms of Hunger and Satiety in Healthy Men during One Week of Sleep Restriction and Circadian Misalignment. ), stress (Born JM et al., Int J Obes (Lond)., 2010, Acute stress and food-related reward activation in the brain during food choice during eating in the absense of hunger), hormone secretion (Mari Hotta et al., Endocr J., 2009. and Food intake in patients with restricting-type anorexia nervosa: A pilot study. ), nervous system status (Ashim Maharjan et al., Appetite. 2022. The effects of frequency-specific, non-invasive, median nerve stimulation on food-related attention and appetite), receptor activity (P. A. Sargent et al.,Psychopharmacology (Berl). 1997. 5-HT2C receptor activation decreases appetite and body weight in obese subjects. ), gene (Anestis Doukas et al., Br J Nutr. 2013. The impact of obesity-related SNP on appetite and energy intake. ), muscle mass (Masahiro Okada et al., Int J Environ Res Public Health., 2018. Influence of Muscle Mass and Outdoor Environmental Factors on Appetite and Satiety Feeling in Young Japanese Women. ), fat mass (Ilse M T Nijs et al., Appetite, 2010, Differences in attention to food and food intake between overweight / obese and normal-weight females under conditions of hunger and satiety), gut bacteria (Marina Sanchez et al., Nutrients. 2017. Effects of a Diet-Based Weight-Reducing Program with Probiotic Supplementation on Satiety Efficiency, Eating Behaviour Traits, and Psychosocial Behaviours in Obese Individuals. ), age / gender (Parker BA et al., Eur J Clin Nutr., 2004,Relationship between food intake and visual analog scale ratings of appetite and other sensations in healthy older and young subjects), drinking (S J Caton et al., Physiol Behav., 2004. Dose-dependent effects of alcohol on appetite and food intake.), smoking history (Nikolaj T. Gregersen et al., Food Nutr Res. 2011. Determinants of appetite ratings: the role of age, gender, BMI, physical activity, smoking habits, and diet / weight concerns, illness (Camilla Klastrup et al., Eat Weight Disord., 2020. Hunger and satiety perception in patients with severe anorexia nervosa.), lifestyle habits such as night-time or diurnal habits (Bruno Simao Teixeira et al., Eur J Nutr. , 2023. Influence of fasting during the night shift on next day eating behavior, hunger, and glucose and insulin levels: a randomized, three-condition, crossover trial.), environmental characteristics (e.g., people eating together (Ruddock HK et al., Sci Rep. 2021. People serve themselves larger portions before a social meal), eating time (Leonie C. Ruddick-Collins et al., Cell Metab.,2022. Timing of daily calorie loading affects appetite and hunger responses without changes in energy metabolism in (Elizabeth A Thomas et al., Obesity (Silver Spring), 2015. Usual breakfast eating habits affect response to breakfast skipping in overweight women), plate size (Samira Rabiei et al., Iran J Public Health., 2023. Effects of Food Plate Size and Color on Visual Perception of Satiate in Adolescents; a New Strategy toward Weight), plate color (Asli Akyol et al., Nutr J., 2018. Impact of three different plate colors on short-term satiate and energy intake: a randomized controlled trial. ), lighting (Konstantin V Danilenko et al., Obes Facts., 2013. Bright light for weight loss: results of a controlled crossover trial.), temperature / weather (Masahiro Okada et al., Int J Environ Res Public Health. 2018. Influence of Muscle Mass and Outdoor Environmental Factors on Appetite and Satisfaction Feeling in Young Japanese Women, number of meals per day (E Papakonstantinou et al.,Diabetes Metab. 2018. Effects of 6 vs 3 eucaloric meal patterns on glycaemic control and satiety in people with impaired glucose tolerance or overt type 2 diabetes: A randomized trial. The order in which meals are eaten (Alissa M. Mori et al., Nutr Metab (London) 2011. Acute and second-meal effects of almond form in impaired glucose tolerant adults: a randomized crossover trial), and the speed at which meals are eaten (Andrade AM et al., Int J Behav Nutr Phys Act. 2012. Does eating slowly influence appetite and energy intake when Is water intake controlled?), number of chews (Marion M Hetherington et al., Appetite, 2011. Mastication of almonds: effects of lipid bioaccessibility, appetite, and hormone response.), etc.), and dietary characteristics (calories (Mantzavinou A et al., Physiol Behav., 2023, Apple versus chocolate: Evidence for discrimination of carbohydrates (B J Rolls et al., Am J Clin Nutr., 2014) and dietary fiber (B J Rolls et al., Am J Clin Nutr., 2014) contribute to the regulation of postprandial fullness and hunger, and the quality and location of other body sensations.1994. 饱腹感与不同量脂肪和碳水化合物预负荷:对肥胖的影响)、蛋白质(奥利维拉·CLP等人,《欧洲营养学杂志》,2022年,高蛋白全饮食替代对健康、正常体重成年人食物摄入和能量稳态的调节作用)、脂质(科齐莫尔·A等人,《食欲》,2013年,高脂肪餐的膳食脂肪酸组成对饱腹感的影响)、氨基酸(安东内洛·E·里加蒙蒂等人,《临床医学杂志》,2020年,评估氨基酸混合物对肥胖青少年胃肠道肽分泌、葡萄糖代谢稳态和食欲的影响,给予固定剂量或随意进餐)、矿物质·维生素(温迪·陈·谢·平德尔福斯等人,《临床营养学》,2011年,饮食诱导的产热、脂肪氧化和连续餐后的食物摄入:钙和维生素D的影响)、有机酸类(海尔达·图通奇等人,《国际临床实践杂志》,2023年,羟基柠檬酸补充对身体成分、肥胖指标、食欲、瘦素的影响以及热量限制饮食的非酒精性脂肪性肝病女性的脂联素()、酒精(SJ Caton等人,《生理行为》,2004年。酒精对食欲和食物摄入的剂量依赖性影响)、味道(Yin Wentin等人,《食欲》,2017年。香气和味道单独或联合对食欲感觉和随后食物摄入的影响。)、口感(Stribitcaia E等人,《分享食物质地对饱腹感的影响:系统评价和荟萃分析。科学报告》,2020年)、气味(Yin Wentin等人,《食欲》,2017年。香气和味道单独或联合对食欲感觉和随后食物摄入的影响。)、食物温度(BJ Rolls等人,《食欲》,1990年。果汁的温度和呈现方式对人类饥饿、口渴和食物摄入的影响。)、粘度(Zhu Yong等人,《公共科学图书馆·综合》,2013年。食物粘度对进食速度、主观食欲、血糖反应和胃排空率的影响)、食物颜色(Suzuki Maki等人,《食欲》,2017年。热汤的颜色调节餐后饱腹感、热感觉,The VAS prediction results for hunger, fullness, satisfaction, or appetite were obtained from factors such as body temperature in young women, and size of ingredients (Maryam S. Hafiz et al., Food Funct., 2022. Impact of food processing on postprandial glycemic and appetite responses in healthy adults: a randomized, controlled trial).

[0054] As shown in FIG. 5, in this embodiment, the meal contents that will give a sufficient feeling of fullness are predicted using the factors contributing to a feeling of fullness as input data.

[0055] Furthermore, as shown in Figure 6, in this embodiment, a satiety architect using a machine learning model is developed to predict and display meal contents that will sufficiently satisfy various consumers.

[0056] As shown in FIG. 7, in this embodiment, a service is provided that utilizes a satiety architect to suggest meal plans based on data input at retail stores.

[0057] [Example 1] An example of this embodiment will be described with reference to Fig. 8 to Fig. 10. Fig. 8 is a diagram showing a feeling of fullness in this embodiment. Fig. 9 is a diagram showing a feeling of hunger in this embodiment. Fig. 10 is a diagram showing verification of prediction accuracy in this embodiment.

[0058] In Example 1, seven subjects were evaluated for satiety, hunger, etc. using a visual analog scale (VAS), and then multiple regression analysis was performed using the food record results and VAS evaluation time as explanatory variables and the VAS evaluation results for hunger as the dependent variable.

[0059] (1) Implementation of VAS Evaluation First, in Example 1, VAS evaluation results regarding hunger, fullness, satisfaction, and appetite during ad libitum eating were obtained. Here, Figures 8 and 9 show the VAS evaluation results of fullness (Figure 8) and hunger (Figure 9) for the same person over three days, and the VAS evaluation results fluctuate on each evaluation day because factors affecting fullness, such as diet and lifestyle, change.

[0060] (2) Data Preprocessing In Example 1, in order to standardize the trends in VAS evaluations between subjects, the average value of the VAS evaluation results for each subject was calculated from the VAS evaluation results on different evaluation days, and standardized VAS evaluation results for hunger, fullness, satisfaction, and appetite were obtained using the formula [standardized individual VAS evaluation result = individual VAS evaluation result ÷ average VAS evaluation result ÷ AUC evaluation time]. The area under the curve (AUC) of the transition of the VAS evaluation results was calculated, and the value of AUC corrected by the evaluation time was calculated as the VAS intensity.

[0061] (3) Multiple Regression Analysis As shown in FIG. 10 , in Example 1, a multiple regression analysis was performed on 16 data sets that were amenable to AUC analysis, using the VAS intensity of the hunger VAS assessment results from after lunch to before dinner, standardized by the preprocessing in (2), with the food record results (carbohydrate intake, protein intake, and lipid intake) and the time from after lunch to before dinner as explanatory variables, and fullness as the objective variable. Cross-validation was then performed to verify prediction accuracy. Specifically, in Example 1, approximately 10% of the validation data was used, and the remaining data was used as training data. The validation data was then sequentially swapped and validated, thereby validating all data. As a result, as shown in FIG. 10 , a correlation coefficient of 0.818 and a coefficient of determination of 0.564 were obtained in Example 1. A correlation between the predicted and measured VAS intensity values ​​was confirmed using these four parameters. The p-value of a two-sided test for the null hypothesis that the gradient was zero was 5.9*10E-5, indicating a significant trend.

[0062] (4) Various Analyses of VAS Evaluation Results As shown in FIGS. 8 and 9, the VAS evaluation results in Example 1 can be analyzed not only based on AUC but also by various other analyses and predictions. For example, prediction of VAS at a fixed time after a meal, the time at which the VAS evaluation results rise to a certain slope or more after a meal, or a method of smoothing the VAS evaluation results and then determining the slope or maximum value may be used.

[0063] In this embodiment, AUC is a comprehensive evaluation of post-meal fluctuations: it is a fairly rounded value, and is an overall score that roughly evaluates the content of a certain meal and how it is eaten. For example, to measure satisfaction with a frozen lunch box, the score is converted into a score up to a certain time (for example, 3 hours after eating) and displayed to the consumer.

[0064] In this embodiment, the slope of the VAS from the maximum and minimum values ​​after a meal indicates how much the VAS fluctuates, and is somewhat similar in concept to AUC, but because it is a slope, it is a parameter that excludes information about the minimum and maximum values. For example, in this embodiment, when evaluating an index such as a feeling of fullness, the slope of the VAS from the maximum and minimum values ​​after a meal can be used to evaluate how quickly it improves.

[0065] In this embodiment, the prediction of a fixed point after a certain time period predicts the VAS level at any given time, and can be used when you want to be in the best condition after a certain meal or eating method before doing something, for example, to replenish energy before a sports game, reduce satiety (or improve mood), and achieve the best performance.Furthermore, in this embodiment, the prediction of a fixed point after a certain time period can also be used to develop sports diet menus for athletes and diet development for students taking exams.

[0066] In addition, in this embodiment, the time of the inflection point indicates the time when the VAS evaluation result rises or falls. For example, when a person is suddenly overcome by hunger before dinner, the inflection point can be evaluated to assess how long the person can go without feeling hungry, thereby enabling the person to eat dinner without being suddenly overcome by hunger before dinner.

[0067] In this embodiment, the maximum and minimum values ​​of the VAS represent the impact of postprandial fluctuations or the peak of postprandial VAS fluctuations, and the impact of how much fluctuation there is after a meal can be evaluated. For example, the postprandial satisfaction can be scored and the maximum value can be observed, or the time at which the feeling of fullness peaks can be evaluated and the time at which this occurs can be evaluated. In this embodiment, the maximum and minimum values ​​of the VAS can suggest eating habits that are less likely to cause drowsiness, and for example, the occurrence rate of snacking desires can be evaluated based on whether the desire to eat sweets peaks after dinner.

[0068] [4. Other Embodiments] In addition to the above-described embodiments, the present invention may be implemented in various different embodiments within the scope of the technical concept described in the claims.

[0069] For example, among the processes described in the embodiments, all or part of the processes described as being performed automatically can be performed manually, or all or part of the processes described as being performed manually can be performed automatically using known methods.

[0070] Furthermore, the processing procedures, control procedures, specific names, information including parameters such as registered data and search conditions for each process, screen examples, and database configurations shown in this specification and drawings can be changed as desired unless otherwise specified.

[0071] Furthermore, with regard to the diet evaluation device 100 and the like, the components shown in the drawings are functional concepts, and do not necessarily have to be physically configured as shown in the drawings.

[0072] For example, all or any part of the processing functions of the diet evaluation device 100, particularly the processing functions performed by the control unit, may be implemented by a CPU and a program interpreted and executed by the CPU, or may be implemented as hardware using wired logic. The program is recorded on a non-transitory computer-readable recording medium containing programmed instructions for causing the information processing device to execute the processes described in this embodiment, and is mechanically read as needed. That is, a computer program is recorded in a storage unit such as a ROM or HDD (hard disk drive) for working with an OS to issue instructions to the CPU and perform various processes. This computer program is executed by being loaded into RAM and works with the CPU to form the control unit.

[0073] In addition, this computer program may be stored in an application program server connected to the meal evaluation device 100, etc. via an arbitrary network 300, and it is also possible to download all or part of it as needed.

[0074] Furthermore, the program for executing the processes described in this embodiment may be stored in a non-transitory computer-readable recording medium, or may be configured as a program product. Here, this "recording medium" includes memory cards, USB (Universal Serial Bus) memories, SD (Secure Digital) cards, flexible disks, magneto-optical disks, ROMs, EPROMs (Erasable Programmable Read Only Memory), EEPROMs (registered trademark) (Electrically Erasable and Programmable Read Only Memory), CD-ROMs (Compact Disk Read Only Memory), MOs (Magneto-Optical disks), DVDs (Digital Versatile Disks), and more. This includes any "portable physical medium" such as a Blu-ray Disc, a DVD player, a DVD player, a Blu-ray Disc, etc.

[0075] Furthermore, a "program" is a data processing method written in any language or description method, and does not matter whether it is in the form of source code or binary code. Note that a "program" is not necessarily limited to a single structure, but also includes a structure that is distributed as multiple modules or libraries, or a structure that achieves its function by cooperating with a separate program, such as an OS. Note that the specific configuration and reading procedure for reading a recording medium in each device shown in this embodiment, as well as the installation procedure after reading, can use well-known configurations and procedures.

[0076] The various databases stored in the memory unit are storage means such as memory devices such as RAM and ROM, fixed disk devices such as hard disks, flexible disks, and optical disks, and store various programs, tables, databases, and web page files used for various processes and providing websites.

[0077] The diet evaluation device 100 may be configured as an information processing device such as a known personal computer or workstation, or may be configured as an information processing device connected to any peripheral device. The diet evaluation device 100 may also be realized by installing software (including programs or data) that causes the device to perform the processing described in this embodiment.

[0078] Furthermore, the specific form of distribution and integration of the devices is not limited to that shown in the drawings, and all or part of them can be functionally or physically distributed and integrated in any unit depending on various additions or functional loads. In other words, the above-mentioned embodiments can be implemented in any combination, or embodiments can be implemented selectively.

[0079] REFERENCE SIGNS LIST 100 Diet evaluation device 102 Control unit 102a Model construction unit 102b Factor acquisition unit 102c Prediction result acquisition unit 106 Storage unit 106a Diet database 112 Input / output unit 300 Network

Claims

1. A meal evaluation device comprising a memory unit and a control unit, wherein the memory unit comprises: a meal memory means for storing a machine learning model in which contributing factors including user characteristics, environmental characteristics and / or meal characteristics are explanatory variables and a sensory evaluation value for a meal is an objective variable, and the control unit comprises: a factor acquisition means for acquiring the contributing factors of a person to be evaluated, and a prediction result acquisition means for acquiring a prediction result of the sensory evaluation value for the meal of the person to be evaluated from the contributing factors of the person to be evaluated using the machine learning model.

2. The dietary evaluation device as described in claim 1, characterized in that the sensory evaluation value is a subjective sensory evaluation value and / or the amount of an in vivo factor related to the subjective sensory evaluation value and / or a measurement result.

3. The meal evaluation device according to claim 2, characterized in that the subjective sensory evaluation value is hunger, satiety, satisfaction, appetite, bloating, heartburn, stomach discomfort, snacking desire, drowsiness, or fatigue.

4. The dietary evaluation device according to claim 2, characterized in that the amount of the intrinsic biological factor is blood glucose level trends and / or blood hormone concentrations.

5. The dietary evaluation device according to claim 4, wherein the blood hormones are leptin, ghrelin, insulin, cortisol, and / or gastrointestinal hormones.

6. The dietary evaluation device according to claim 2, characterized in that the measurement result is an electroencephalogram.

7. A dietary evaluation device as described in any one of claims 1 to 6, characterized in that the machine learning model is a model consisting of at least one of multiple regression analysis, random forest, neural network, XGBoost, LightGBM, support vector regression, and linear regression including ElasticNet.

8. A meal evaluation device as described in any one of claims 1 to 6, characterized in that the sensory evaluation value is a score obtained by sensory evaluation.

9. The meal evaluation device according to claim 8, wherein the sensory evaluation value is a VAS score.

10. A dietary evaluation device as described in any one of claims 1 to 6, characterized in that the contributing factors are factors obtained from at least one of a diet record, a questionnaire, an internet log, and a device measurement.

11. A dietary evaluation device as described in any one of claims 1 to 6, characterized in that the dietary characteristics are the amount of at least one of dietary fiber and carbohydrates including sugars, protein, lipids, salt, minerals, vitamins, amino acids, micronutrients, and organic acids in the meal.

12. The meal evaluation device according to claim 1, characterized in that the prediction result acquisition means further designs a suggested meal content based on the prediction result and displays the suggested meal content.

13. A meal evaluation method to be executed by a meal evaluation device having a memory unit and a control unit, wherein the memory unit comprises: a meal memory means for storing a machine learning model in which contributing factors including user characteristics, environmental characteristics and / or meal characteristics are explanatory variables and a sensory evaluation value for a meal is an objective variable; and the meal evaluation method includes: a factor acquisition step, executed in the control unit, for acquiring the contributing factors of a person to be evaluated; and a prediction result acquisition step, executed in the control unit, for acquiring a predicted result of the sensory evaluation value for the meal of the person to be evaluated from the contributing factors of the person to be evaluated, using the machine learning model.

14. A meal evaluation program to be executed by a meal evaluation device having a memory unit and a control unit, wherein the memory unit comprises: a meal memory means for storing a machine learning model in which contributing factors including user characteristics, environmental characteristics and / or meal characteristics are explanatory variables and a sensory evaluation value for a meal is an objective variable; and the control unit executes: a factor acquisition step for acquiring the contributing factors of a person to be evaluated; and a prediction result acquisition step for acquiring a predicted result of the sensory evaluation value for the meal of the person to be evaluated from the contributing factors of the person to be evaluated using the machine learning model.

Images