Blood glucose level management system and blood glucose level management method

Abstract

The blood glucose level management system according to the present disclosure includes a blood glucose level fluctuation learning unit that learns a user's blood glucose level fluctuations and constructs a blood glucose level fluctuation model based on biometric data including at least the user's blood glucose level fluctuations and meal data including at least the user's meal menu, meal amount, meal start time, and meal duration; a blood glucose level fluctuation prediction unit that, before the user starts eating, predicts the user's blood glucose level fluctuations after consuming the meal based on the meal data of the meal to be ingested and the blood glucose level fluctuation model and calculates a blood glucose level fluctuation prediction value; and a result presentation unit that presents the blood glucose level fluctuation prediction value to the user.

Description

[Technical Field]

[0001] The present disclosure relates to a blood glucose level management system, and more particularly to a blood glucose level management system that manages blood glucose levels based on biological data and dietary data. [Background technology]

[0002] Diet has a significant impact on health, but the effects of dietary content and eating behavior on the body are difficult to understand. Therefore, several technologies that provide advice on healthy eating are being investigated.

[0003] For example, Patent Document 1 describes a method for estimating a user's behavior based on behavioral information and biometric information, applying the estimated behavior results to a metabolic model to estimate blood glucose levels, and presenting recommended behaviors to the user based on the estimated blood glucose levels. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Laid-Open No. 2012-235869 Summary of the Invention [Problem to be solved by the invention]

[0005] Blood glucose levels fluctuate, and it is important to estimate and present time-series changes in order to encourage users to manage their blood glucose levels. Furthermore, factors that affect blood glucose level fluctuations include multiple conditions such as diet, an individual's constitution, health condition, and exercise history, and the technology disclosed in Patent Document 1 has the problem of not being able to adequately manage blood glucose levels.

[0006] The present disclosure has been made to solve the above problems, and aims to provide a blood glucose level management system that allows users to adequately manage their blood glucose levels. [Means for solving the problem]

[0007] The blood glucose level management system according to the present disclosure includes a blood glucose level fluctuation learning unit that learns the blood glucose level fluctuations of a user based on biometric data including at least the blood glucose level fluctuations of the user and dietary data including at least the user's meal menu, meal amount, meal start time, and meal duration, and constructs a blood glucose level fluctuation model; a blood glucose level fluctuation prediction unit that, before the user starts eating, predicts the blood glucose level fluctuations of the user after consuming the meal based on the dietary data of the meal to be ingested and the blood glucose level fluctuation model, and calculates a blood glucose level fluctuation prediction value; and a result presentation unit that presents the blood glucose level fluctuation prediction value to the user. [Effects of the Invention]

[0008] According to the blood glucose management system of the present disclosure, the blood glucose fluctuations of the user are learned based on the user's biometric data and dietary data, a blood glucose fluctuation model is constructed, and a prediction of the user's blood glucose fluctuations after consuming a meal is made based on the dietary data of the meal to be consumed and the blood glucose fluctuation model, thereby making it possible to make highly accurate blood glucose predictions and effectively encourage the user to manage their blood glucose levels. [Brief explanation of the drawings]

[0009] [Figure 1] 1 is a functional block diagram showing the configuration of a blood glucose level management system according to a first embodiment. [Figure 2] 4 is a flowchart illustrating processing in the blood glucose level management system of the first embodiment. [Figure 3] 10 is a flowchart illustrating a method for creating a blood glucose level fluctuation model. [Figure 4] 10 is a flowchart illustrating a method for predicting blood glucose fluctuations. [Figure 5] FIG. 10 is a table showing the results of calculations using a blood glucose level fluctuation model. [Figure 6] FIG. 10 is a graph showing the results of calculations using a blood glucose fluctuation model. [Figure 7] FIG. 1 shows predicted blood glucose levels in a table. [Figure 8]FIG. 10 is a graph showing predicted blood glucose levels. [Figure 9] FIG. 10 is a table showing the recalculated predicted results of postprandial blood glucose fluctuations. [Figure 10] FIG. 10 is a diagram showing an example of a display screen on a result presentation unit. [Figure 11] FIG. 10 is a functional block diagram showing the configuration of a blood glucose level management system according to a second embodiment. [Figure 12] 10 is a flowchart illustrating processing in a blood glucose level management system according to the second embodiment. [Figure 13] FIG. 10 is a diagram showing an example of a display screen on a result presentation unit. [Figure 14] FIG. 11 is a functional block diagram showing the configuration of a blood glucose level management system according to a third embodiment. [Figure 15] 11 is a flowchart illustrating processing in a blood glucose level management system according to the third embodiment. [Figure 16] FIG. 10 is a diagram showing an example of a display screen on a result presentation unit. [Figure 17] FIG. 10 is a functional block diagram showing the configuration of a blood glucose level management system according to a fourth embodiment. [Figure 18] 10 is a flowchart illustrating processing in a blood glucose level management system according to a fourth embodiment. [Figure 19] FIG. 10 is a functional block diagram showing the configuration of a blood glucose level management system according to a fifth embodiment. [Figure 20] 13 is a flowchart illustrating processing in a blood glucose level management system according to a fifth embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0010] <First Embodiment> <System configuration> 1 is a functional block diagram showing the configuration of a blood glucose level management system 100 according to a first embodiment of the present disclosure. As shown in FIG. 1, the blood glucose level management system 100 includes a biological data collection unit 1, a dietary data collection unit 2, a blood glucose level fluctuation learning unit 3, a blood glucose level fluctuation model storage unit 4, a blood glucose level fluctuation prediction unit 5, a dietary method advice creation unit 6, and a result presentation unit 7.

[0011] The biological data collection unit 1 sequentially collects biological data of a specific user (hereinafter referred to as "user") to be managed. The biological data collected from the user includes at least blood glucose fluctuations [mg / dL], and can also include life log data related to daily life and health checkup data measured during health checkups.

[0012] Examples of life log data include the number of steps [steps], sleep time [hours], sleep score, heart rate [bpm], whether or not oral care was performed, stress level, and working hours.

[0013] Examples of health check data include height [cm], weight [kg], BMI [kg / m2], waist circumference [cm], blood pressure [mmHg], triglycerides [mg / dL], cholesterol [mg / dL], HbA1c [%], and smoking status.

[0014] Examples of methods for inputting biometric data include manual input, as well as input methods using invasive sensors, non-invasive sensors, wearable devices, cameras, and video cameras.

[0015] An example of manual input is a method in which weight, blood pressure, and blood glucose levels are recorded and saved on a personal computer using spreadsheet software, and then uploaded to a server computer via a public communication network such as the Internet, and then downloaded from there to the blood glucose level management system 100. Blood glucose levels can be measured using a commercially available personal blood glucose meter or the like that measures the blood glucose level of blood obtained by pricking a fingertip or the like with an attached measuring needle.

[0016] An example of an input method using an invasive sensor is to attach a wearable invasive sensor equipped with a measuring needle to the arm on a daily basis, store the blood glucose data obtained from the sensor as a file in the measuring device, and then upload the file to a server computer.

[0017] An example of an input method using a non-invasive sensor is a method in which blood glucose levels are measured non-invasively using laser technology in a commercially available stationary sensor, the acquired blood glucose data is saved as a file in the meter, and the file is then uploaded to a server computer.

[0018] An example of an input method using a wearable device is to use a commercially available wristwatch-type sensor to acquire daily activity data such as heart rate and number of steps, as well as sleep recording data, export the acquired data, and upload the output data file to a server computer.

[0019] An example of an input method using a camera is to take an image of the display showing the measurement results of a weight scale and a blood glucose sensor on a smartphone, read the numbers using the character recognition function, save them to a file, and then upload the file to a server computer.

[0020] An example of an input method using a video camera is to take a video of a smartphone displaying the blood glucose sensor measurement results or a video of the user exercising, and then use a character recognition function or a behavior recognition service that analyzes the subject's biometric data and behavior from the video to read the numbers and exercise history, save them in a file, and upload them to a server computer.

[0021] The meal data collection unit 2 collects meal data for one meal. The meal data includes at least the meal menu (a list of the items in each dish), the amount of food eaten [g, servings], the time the meal started, and the time required to eat [min], and can also include data such as the nutritional value of the meal, the energy [kcal], and the people who ate with the meal.

[0022] Examples of nutritional values ​​include carbohydrates [g], protein [g], fat [g], sugars [g], dietary fiber [g], salt [g] and GI (Glycemic Index) value.

[0023] Examples of methods for inputting meal data include manual input, as well as input methods using wearable devices, cameras, and video cameras.

[0024] An example of manual input is a method in which a meal menu, calories, and meal start time are entered into an input app on a smartphone, and the app outputs the data to a file, which is then uploaded to a server computer.

[0025] An example of a wearable device is a method that analyzes eating speed from arm movements measured with a wristwatch-type sensor, outputs the data to a file, and uploads the file to a server computer.In addition, research is being conducted on technology that estimates meal content from acceleration data from a wristwatch-type sensor worn on the dominant hand, and this technology could also be used.

[0026] An example of an input method using a camera is to take a photo of food with a smartphone, analyze the menu and calories of the food in the photo using an app, record it in a file, and upload the file to a server computer. Services that analyze food data from camera images are commonly used.

[0027] An example of an input method using a video camera is to film the eating process with a smartphone camera, analyze the food content and eating speed from the video, record the results in a file, and upload the file to a server.

[0028] The blood glucose fluctuation learning unit 3 uses each user's past meal data and past biological data to extract features that have a large impact on blood glucose fluctuations, and uses these features to build a blood glucose fluctuation model for each user. The construction of the blood glucose fluctuation model will be described in more detail later.

[0029] The blood glucose level fluctuation model storage unit 4 is a storage device that accumulates the blood glucose level fluctuation model constructed by the blood glucose level fluctuation learning unit 3.

[0030] Before starting a meal, the blood glucose fluctuation prediction unit 5 predicts the user's blood glucose fluctuation after consuming the meal using the meal data of the meal to be consumed and the blood glucose fluctuation model stored in the blood glucose fluctuation model storage unit 4, and calculates a blood glucose fluctuation prediction value. For example, when the time required for the meal, the meal menu, and the amount of food eaten are input, the time change in blood glucose level is output as a graph of the blood glucose fluctuation prediction value. Note that the time required for the meal in the meal data cannot be obtained before the meal starts, so it is input as an estimate. The method for predicting blood glucose fluctuation will be explained further below.

[0031] If the predicted blood glucose fluctuation value deviates from a preset reference value based on the predicted postprandial blood glucose fluctuation value predicted by the blood glucose fluctuation prediction unit 5, the dietary advice creation unit 6 creates dietary advice that prevents the predicted blood glucose fluctuation value from deviating from the reference value. An example of dietary advice will be given later.

[0032] The result presentation unit 7 presents to the user the dietary advice created by the dietary advice creation unit 6, the blood glucose fluctuation predicted value predicted by the blood glucose fluctuation prediction unit 5 when the dietary advice is not applied, and the blood glucose fluctuation predicted value recalculated by the blood glucose fluctuation prediction unit 5 when the dietary advice is applied.

[0033] <System processing> FIG. 2 is a flowchart explaining the processing in the blood glucose level management system 100 of the first embodiment shown in FIG. 1. When the blood glucose level management system 100 starts processing, the meal data collection unit 2 collects meal data (step S1), and the biological data collection unit 1 collects biological data (step S2). These processes can be performed in either order, or simultaneously. Furthermore, these processes are data collection for building a blood glucose level fluctuation model of the user, and the biological data also includes data on blood glucose level fluctuations before and after the user's meal.

[0034] Next, a blood glucose fluctuation model is created in the blood glucose fluctuation learning unit 3 (step S3). Here, a method for creating the blood glucose fluctuation model will be described with reference to FIG.

[0035] Figure 3 is a flowchart explaining a method for creating a blood glucose level fluctuation model. As shown in Figure 3, the blood glucose level fluctuation learning unit 3 first analyzes data items that affect blood glucose level fluctuations from the collected biological data and dietary data (step S31). The analysis method uses statistical techniques or machine learning, such as inputting the biological data and dietary data and correlating them with blood glucose level fluctuations before and after a meal.

[0036] For example, the input data includes blood glucose fluctuations, meal data A and B, life log data C, D and E, and health check data F, G, H, and I, and the output shows that blood glucose fluctuations are strongly influenced by meal data A, life log data C, and health check data H.

[0037] Next, the blood glucose fluctuation learning unit 3 constructs a blood glucose fluctuation model using data items with a high degree of influence (step S32).

[0038] For example, a fluctuation model is assumed in which a predicted blood glucose level F(t) at time t after a meal is expressed by the following equation (1).

[0039]

number

[0040] In the above formula (1), A(t) is the value of the dietary data A at time t, C(t) is the value of the life log data C at time t, H(t) is the value of the health check data H, and J is an arbitrary constant. The above formula (1) shows a case where the fluctuation model is expressed as a nonlinear model, but it may also be expressed as a linear model. Note that AI (artificial intelligence) or machine learning can be used to create the blood glucose fluctuation model.

[0041] Now, returning to the explanation of the flowchart in Fig. 2, after the blood glucose level fluctuation model is created in step S3, the meal data collection unit 2 collects meal data for the meal that the user will take in the future (step S4).

[0042] Next, the blood glucose fluctuation prediction unit 5 predicts the user's postprandial blood glucose fluctuation using the blood glucose fluctuation model created by the blood glucose fluctuation learning unit 3 (step S5). Here, the method for predicting blood glucose fluctuation will be described with reference to FIG.

[0043] 4 is a flowchart illustrating a method for predicting blood glucose fluctuations. As shown in Fig. 4, the blood glucose fluctuation prediction unit 5 first receives data on meals that the user plans to take from the meal data collection unit 2 (step S51).

[0044] An example of meal data that a user plans to eat is a meal menu of katsudon (pork cutlet rice bowl), cauliflower, pumpkin, boiled green beans, and tea, with the meal starting time at 12:30, the meal time (planned) being 20 minutes, and the calories being 983 kcal.

[0045] Next, the blood glucose fluctuation model created by the blood glucose fluctuation learning unit 3 is used to predict postprandial blood glucose fluctuation (step S52).

[0046] For example, a fluctuation model is assumed in which a predicted blood glucose level f(t) at time t after a meal is expressed by the following equation (2).

[0047]

number

[0048] In the above formula (2), a(t) is the calorie intake per unit of time elapsed since the meal (calorie intake / meal time) [kcal], b(t) is the number of steps since the previous meal [thousands of steps], c(t) is the muscle percentage, and d(t) is the average resting blood glucose level.

[0049] The results of calculations using the above formula (2) are shown in the table in Figure 5 and as a graph in Figure 6. In Figures 5 and 6, the time unit of time t is 5 minutes, with time t=0 representing the start of the meal and -1 representing the time immediately before the start of the meal. In this case, since a(t) is 0, the predicted blood glucose level f(t) is 77, which is the average resting blood glucose level.

[0050] 5 and 6 show that the predicted blood glucose level peaked at 156 mg / dL 15 minutes after the meal started. Also, at t=4, a(t)=0, indicating that the meal ended 20 minutes after the meal started.

[0051] Returning now to the explanation of the flowchart in Fig. 2, after predicting postprandial blood glucose fluctuations in step S5, the blood glucose fluctuation prediction unit 5 determines whether the blood glucose fluctuations deviate from the reference value based on the prediction results (step S6).

[0052] Here, the reference value is an arbitrarily set value, and can be, for example, 140 mg / dL, above which a blood glucose spike is determined, or 126 mg / dL, the lower limit of fasting blood glucose used as an indicator of diabetes.

[0053] In step S6, if it is determined that the value deviates from the standard value (if Yes), the process proceeds to step S7 in the dietary method advice creation unit 6, and if it is determined that the value does not deviate from the standard value (if No), the process proceeds to step S8.

[0054] Here, the reference value used is 140 mg / dL, which is the standard for determining blood glucose spikes. Figure 7 is a table showing the predicted blood glucose levels f(t) from the table shown in Figure 5, and Figure 8 is a graph of this. Figure 8 shows the reference value of 140 mg / dL, and it can be seen that it exceeds 140 mg / dL when t=3.

[0055] In step S7, the blood glucose fluctuation model used for the prediction is referenced, and values ​​of changeable items that affect blood glucose fluctuation are reset.

[0056] That is, the items used in the blood glucose fluctuation model are checked, and changeable items, i.e., items related to the food data to be consumed and items related to the life log, are identified. Here, we use a(t) used in formula (2), which is "calorie intake per unit of time elapsed since eating."

[0057] Then, the value of "calorie intake per unit of elapsed time since meal" is changed and input to the blood glucose fluctuation prediction unit 5, and prediction of postprandial blood glucose fluctuation is performed again in step S5.

[0058] Figure 9 is a table showing the recalculated predicted postprandial blood glucose fluctuations. Figure 9 shows the recalculated predicted postprandial blood glucose level f(t), calorie intake per unit of time elapsed since eating a(t), and the meal time and calorie intake. Figure 9 shows that the blood glucose level is 114 mg / dL at t=3. This shows that taking more time to eat lowers the peak blood glucose level.

[0059] After predicting postprandial blood glucose fluctuations in step S5, a determination is again made in step S6 as to whether the blood glucose fluctuations deviate from the reference value. As shown in Figure 9, the blood glucose peak is 114 mg / dL, so it is determined in step S6 that the blood glucose fluctuations do not deviate from the reference value, and the process proceeds to step S8 in the blood glucose fluctuation prediction unit 5.

[0060] In step S8, it is determined whether the prediction has been made two or more times, that is, whether the prediction has been redone, and if the prediction has been made once (No), no advice on eating habits is necessary, so the series of processes ends without creating any advice.

[0061] On the other hand, if it is determined in step S8 that prediction has been performed two or more times (if Yes), the items changed during the re-prediction, i.e., the improvement value, are converted into advice for the user, and the dietary advice creation unit 6 creates dietary advice (step S9).

[0062] That is, as can be seen from Figure 9, by changing the calorie intake a(t) per unit of time elapsed since the meal, the blood glucose peak will not deviate from the reference value, and therefore the eating advice creation unit 6 creates eating advice such as "Take 25 minutes to eat your meal."

[0063] Returning now to the explanation of the flowchart in Fig. 2, after the eating method advice has been created in step S9, the result presenting unit 7 presents the blood glucose level fluctuations before and after the eating method improvement and the eating method advice to the user (step S10).

[0064] Figure 10 shows the display screen of a smartphone as an example of the result presentation unit 7. The display screen displays an image of the meal the user is about to eat, taken with the smartphone, and the eating advice "Take 25 minutes to eat it," as well as a graph of blood glucose level fluctuations before the eating method was improved, when the food was eaten as is (dashed line), and a graph of blood glucose level fluctuations after the eating method was improved, when the food was eaten as advised (solid line), along with the reference value of 140 mg / dL.

[0065] <Effects> By using the blood glucose level management system 100 of embodiment 1 described above, it is possible to construct a blood glucose level fluctuation model that takes into account individual differences in factors that affect blood glucose level fluctuation, and to perform highly accurate blood glucose level predictions, thereby encouraging the user to manage their blood glucose level effectively.

[0066] <Embodiment 2> <System configuration> Fig. 11 is a functional block diagram showing the configuration of a blood glucose level management system 200 according to a second embodiment of the present disclosure. As shown in Fig. 11, the blood glucose level management system 200 is basically the same as the blood glucose level management system 100 shown in Fig. 1, but is configured so that the blood glucose fluctuation prediction unit 5 also receives input of biological data collected by the biological data collection unit 1.

[0067] That is, in the blood glucose level management system 200, the biological data collected by the biological data collection unit 1 while the user is taking a meal is input to the blood glucose level fluctuation prediction unit 5.

[0068] The biometric data collected by the biometric data collection unit 1 and the collection method thereof are basically the same as the biometric data and collection method described in embodiment 1, but also include items that can only be obtained during a meal, such as the number of chews (number of times chewed), intake amount, and intake speed.

[0069] Then, by inputting biological data while the user is eating into the blood glucose level fluctuation prediction unit 5, blood glucose level fluctuation prediction is repeated even while the user is eating. For example, by adding the number of chews, the chewing side (i.e., which side of the teeth is used for chewing), the heart rate during meals, and the amount of food per mouthful to the blood glucose level fluctuation model, if a small number of chews per mouthful affects blood glucose level fluctuations, the eating method advice creation unit 6 can create eating method advice such as "chew your food more thoroughly."

[0070] In addition, if postprandial blood glucose levels are predicted before a meal and estimated values ​​of dietary data and biological data during the meal are used in the prediction, a process is also envisioned in which the estimated values ​​are replaced with actual measured values ​​obtained during the meal.

[0071] <System processing> Figure 12 is a flowchart explaining the processing in the blood glucose level management system 200 of embodiment 2 shown in Figure 11. As shown in Figure 12, the processing in the blood glucose level management system 200 is basically the same as the processing in the blood glucose level management system 100 shown in Figure 2, but after the processing of step S10 in the result presenting unit 7, processing of step S12 is added in which the biological data collecting unit 1 collects biological data of the user while the user is consuming a meal and inputs the data to the blood glucose level fluctuation predicting unit 5.

[0072] Figure 13 shows the display screen of a smartphone as an example of the result presentation unit 7. The display screen displays an image of the meal the user is about to eat, taken with the smartphone, and the eating advice "chew your food 20 to 30 times," as well as a graph of blood glucose level fluctuations before the dietary improvement, when the food is eaten as is (dashed line), and a graph of blood glucose level fluctuations after the dietary improvement, when the food is eaten as advised (solid line), along with the reference value of 140 mg / dL.

[0073] <Effects> By using the blood glucose level management system 200 of the second embodiment described above, in addition to the effects of the blood glucose level management system 100 of the first embodiment, it is possible to provide advice to the user on their behavior during meals by repeating blood glucose level predictions even during meals.

[0074] <Third Embodiment> <System configuration> Fig. 14 is a functional block diagram showing the configuration of a blood glucose level management system 300 according to the third embodiment of the present disclosure. As shown in Fig. 14, the blood glucose level management system 300 includes a biological data storage unit 8 that accumulates biological data collected by the biological data collection unit 1, in addition to the configuration of the blood glucose level management system 200 shown in Fig. 11.

[0075] That is, in the blood glucose level management system 300, the biological data collected by the biological data collection unit 1 while the user is eating is input to the blood glucose level fluctuation prediction unit 5 and stored in the biological data storage unit 8.

[0076] The biological data collected by the biological data collection unit 1 while the user is eating includes data on meals that took 25 minutes or more, and this data is stored as biological data for the past few days in the biological data storage unit 8. This biological data for the past few days is read out by the dietary advice creation unit 6, processed so that it can be used as reference information, and presented to the user by the result presentation unit 7 together with dietary advice.

[0077] <System processing> Figure 15 is a flowchart explaining the processing in the blood glucose level management system 300 of the third embodiment shown in Figure 14. As shown in Figure 15, in addition to the processing in the blood glucose level management system 200 shown in Figure 12, the processing in the blood glucose level management system 300 further includes, following step S2, a process of storing biological data in the biological data storage unit 8 (step S21), and in step S10, past biological data is also presented as reference information in addition to the blood glucose fluctuations before and after the improvement of the dietary method and dietary advice.

[0078] Figure 16 shows the display screen of a smartphone as an example of the result presentation unit 7. The display screen displays an image of the meal the user is about to eat, taken with the smartphone, along with the eating advice "Take 25 minutes to eat," a graph of blood sugar level fluctuations before the dietary improvement, showing the food eaten as is (dashed line), and a graph of blood sugar level fluctuations after the dietary improvement, showing the food eaten as advised (solid line), as well as a record of compliance with the dietary improvement suggestions for the past five days.

[0079] The dietary improvement suggestion compliance record shows the dates of the last five days and whether or not the suggestion was implemented using a symbol. For example, on April 1st (4 / 1), the suggestion was not implemented and is marked with an X, but on April 2nd (4 / 2), April 3rd (4 / 3), and April 3rd (4 / 3), the suggestion was implemented and is marked with a O. Also, on April 4th (4 / 4), the suggestion was not implemented sufficiently and is marked with a △.

[0080] In this way, the user can flexibly choose their actions by presenting the results of past advice execution along with dietary advice on the result presentation unit 7. For example, after eating healthy meals for three days in accordance with dietary improvement suggestions, the user can choose to "eat as much as I want of whatever I particularly want today."

[0081] <Effects> By using the blood glucose level management system 300 of the third embodiment described above, in addition to the effects of the blood glucose level management system 100 of the first embodiment, it is possible to more flexibly support the user's dietary behavior selection by presenting the results of past dietary improvement suggestions to the user. Furthermore, it is possible to keep the user motivated to improve their diet.

[0082] <Fourth Embodiment> <System configuration> Fig. 17 is a functional block diagram showing the configuration of a blood glucose level management system 400 according to a fourth embodiment of the present disclosure. As shown in Fig. 17, the blood glucose level management system 400 includes, in addition to the configuration of the blood glucose level management system 100 shown in Fig. 1, a similar blood glucose level fluctuation model selection unit 9 for selecting a blood glucose level fluctuation model of another user having similar biometric data to that of the user.

[0083] The biological data collected by the biological data collection unit 1 is input to the blood glucose fluctuation learning unit 3 and also to the similar blood glucose fluctuation model selection unit 9. The similar blood glucose fluctuation model selection unit 9 selects a blood glucose fluctuation model of another user who has biological data similar to the inputted user's biological data, and inputs the selected model to the blood glucose fluctuation prediction unit 5 as a similar blood glucose fluctuation model.

[0084] That is, instead of generating one's own blood glucose level fluctuation model, the blood glucose level model of another user with similar biometric data to one's own is used to predict blood glucose level fluctuations. This makes it possible to accurately predict blood glucose level fluctuations even before one's own blood glucose level fluctuation model is constructed.

[0085] Here, the similarity is determined by focusing on, for example, items measured in a regular health checkup or characteristic items that deviate from the average value among the biometric data acquired from the user. A blood glucose level model of another user with similar items is selected as the similar blood glucose fluctuation model. In addition, methods for determining similarity can use, for example, the magnitude of the Euclidean distance of the biometric data or the magnitude of the correlation coefficient of the items used to determine similarity.

[0086] The blood glucose level management system according to the present disclosure is not a system for managing the blood glucose level of only a specific user, but a system for managing the blood glucose levels of multiple users, and stores the biometric data of multiple users and blood glucose fluctuation models generated based on the data, from which a similar blood glucose level fluctuation model can be selected. The biometric data of multiple users can be stored in a storage unit provided inside the similar blood glucose level fluctuation model selection unit 9, and the blood glucose level fluctuation models of multiple users can also be stored in a storage unit provided inside the similar blood glucose level fluctuation model selection unit 9. The blood glucose level fluctuation models of multiple users can be read out and selected from the blood glucose level fluctuation model storage unit 4.

[0087] To generate a blood glucose level fluctuation model, at least one correct answer data, i.e., the actual measured value of blood glucose level fluctuation after eating, is used, so it is necessary to record blood glucose level fluctuation after eating at least once, which takes time. Furthermore, to improve the accuracy of the blood glucose level fluctuation model, it is necessary to collect a lot of correct answer data, which takes even more time because many meals are eaten.

[0088] However, by using the similar blood glucose fluctuation model, it becomes possible to start managing one's own blood glucose level without actually measuring blood glucose fluctuations after meals, thereby increasing the convenience of the blood glucose management system.

[0089] The initial blood glucose level management uses the similar blood glucose level fluctuation model, but after constructing the own blood glucose level fluctuation model, the similar blood glucose level fluctuation model is not used and the own blood glucose level fluctuation model is used instead.

[0090] The blood glucose fluctuation models of others stored in the blood glucose fluctuation model storage unit 4 are constructed by performing actual measurements of blood glucose fluctuations multiple times, so there is a lot of learning data and they are highly accurate, but the initial blood glucose fluctuation model of oneself is constructed at a stage when there are few actual measurements of blood glucose fluctuations, so there is little learning data and they are less accurate, so a similar blood glucose fluctuation model is used for initial blood glucose management.

[0091] However, since a blood glucose fluctuation model of oneself, constructed by actually measuring blood glucose fluctuations multiple times, is more accurate than a blood glucose fluctuation model of another person, it is ultimately desirable to use one's own blood glucose fluctuation model for blood glucose management. Therefore, in this embodiment, the blood glucose fluctuation model to be used is changed.

[0092] <System processing> Figure 18 is a flowchart explaining the processing in the blood glucose level management system 400 of the fourth embodiment shown in Figure 17. As shown in Figure 18, in addition to the processing in the blood glucose level management system 100 shown in Figure 2, the processing in the blood glucose level management system 400 further includes, following step S2, a process (step S22) in which, if a blood glucose level fluctuation model has not been constructed, the similar blood glucose level fluctuation model selection unit 9 selects a similar blood glucose level fluctuation model.

[0093] The display on the result presentation unit 7 is the same as the display on the result presentation unit 7 of the blood glucose level management system 100 of the first embodiment shown in FIG.

[0094] <Effects> By using the blood glucose level management system 400 of the fourth embodiment described above, in addition to the effects of the blood glucose level management system 100 of the first embodiment, new users can perform blood glucose level predictions even before creating their own blood glucose level fluctuation model, and can receive suggestions for improving their diet.

[0095] <Fifth Embodiment> <System configuration> Fig. 19 is a functional block diagram showing the configuration of a blood glucose level management system 500 according to the fifth embodiment of the present disclosure. As shown in Fig. 19, the blood glucose level management system 500 has basically the same configuration as the blood glucose level management system 400 shown in Fig. 17, except that the function of the blood glucose level fluctuation learning unit 3 has changed to a blood glucose level fluctuation learning unit 31.

[0096] In other words, the blood glucose management system 500 is the same as the blood glucose management system 400 of embodiment 4 in that it uses a similar blood glucose fluctuation model to predict blood glucose fluctuations instead of generating its own blood glucose fluctuation model, but differs in that the blood glucose fluctuation learning unit 31 re-learns the similar blood glucose fluctuation model based on the results of analyzing the user's own biological data.

[0097] That is, rather than constructing a blood glucose fluctuation model from scratch, if there is a similar blood glucose fluctuation model selected by the similar blood glucose fluctuation model selection unit 9, the blood glucose fluctuation learning unit 31 uses the similar blood glucose fluctuation model for initial blood glucose management, and then re-learns the similar blood glucose fluctuation model to obtain a blood glucose fluctuation model that is more suited to the user. The blood glucose fluctuation model obtained by re-learning is stored in the blood glucose fluctuation model storage unit 4.

[0098] To retrain the similar blood glucose fluctuation model, AI or machine learning is used to analyze data items that affect blood glucose fluctuations from one's own biological data and dietary data, and based on the analysis results, the similar blood glucose fluctuation model is retrained and used as one's own blood glucose fluctuation model.

[0099] <System processing> Figure 20 is a flowchart explaining the processing in the blood glucose level management system 500 of the fifth embodiment shown in Figure 19. As shown in Figure 20, in the processing in the blood glucose level management system 500, the processing in step S3 in the blood glucose level management system 400 shown in Figure 18 is replaced by processing in step S23 in which, if a similar blood glucose fluctuation model has already been selected, the model is re-learned using the actual measurement data of the device itself.

[0100] The display on the result presentation unit 7 is the same as the display on the result presentation unit 7 of the blood glucose level management system 100 of the first embodiment shown in FIG.

[0101] <Effects> By using the blood glucose level management system 500 of the fifth embodiment described above, in addition to the effects of the blood glucose level management system 100 of the first embodiment, new users can perform blood glucose level predictions even before creating their own blood glucose level fluctuation model, and by relearning similar blood glucose level fluctuation models, they can manage their blood glucose levels using a blood glucose level fluctuation model that is more suited to them.

[0102] In addition, since there is no need to build a blood glucose fluctuation model from scratch, the cost of building a blood glucose fluctuation model can be reduced. Furthermore, by storing the retrained blood glucose fluctuation model in the blood glucose fluctuation model storage unit 4, the number of blood glucose fluctuation models available to users who will be using the blood glucose management system 500 for the first time will increase. This allows users to select a blood glucose fluctuation model that is more suited to them.

[0103] <How the blood glucose management system is used> In the blood glucose level management systems 100 to 500 of the first to fifth embodiments described above, components other than the biological data collection unit 1, the dietary data collection unit 2, and the result presentation unit 7 can be constructed on a server computer constituting a cloud environment. Then, biological data and dietary data are acquired from the user of the system via a communication network such as the Internet, a blood glucose level fluctuation model is constructed and blood glucose level prediction is performed on the server computer, and dietary advice, predicted blood glucose level fluctuation values, etc. are presented via the communication network to the result presentation unit 7, for example, the display screen of the user's smartphone. Note that other usage forms are also possible.

[0104] The above description is illustrative in all respects, and it is understood that countless variations not illustrated can be envisioned.

[0105] It is possible to freely combine the embodiments, and to modify or omit the embodiments as appropriate.

[0106] The present disclosure described above will be summarized as an appendix.

[0107] (Appendix 1) Biometric data including at least fluctuations in the user's blood glucose level; a blood glucose fluctuation learning unit that learns the blood glucose fluctuation of the user based on meal data including at least the user's meal menu, meal amount, meal start time, and meal duration, and constructs a blood glucose fluctuation model; a blood glucose level fluctuation prediction unit that, before the user starts eating, predicts blood glucose level fluctuations of the user after consuming the meal based on the meal data of the meal to be ingested and the blood glucose level fluctuation model, and calculates a blood glucose level fluctuation prediction value; a result presentation unit that presents the predicted blood glucose fluctuation value to the user.

[0108] (Appendix 2) a dietary advice creating unit that creates, when the blood glucose fluctuation predicted value deviates from a preset reference value, dietary advice to prevent deviation from the reference value; The result presentation unit The dietary advice; and The blood glucose level fluctuation prediction value before the dietary method improvement based on the dietary method advice; and the predicted value of blood glucose fluctuation after the dietary method is improved based on the dietary method advice.

[0109] (Appendix 3) The blood glucose fluctuation learning unit 3. The blood glucose level management system according to claim 1, wherein data items contained in the dietary data and the biological data that have a large impact on the blood glucose level fluctuations of the user are used as features, and the blood glucose level fluctuation model is constructed using the features.

[0110] (Appendix 4) The dietary advice creation unit A blood glucose management system as described in Appendix 2, wherein, when the predicted blood glucose fluctuation value deviates from a predetermined reference value, the blood glucose fluctuation model used to predict the user's blood glucose fluctuation is referenced, and the dietary advice is created by resetting values ​​of changeable data items among data items that affect the blood glucose fluctuation and are included in the dietary data and the biological data.

[0111] (Appendix 5) The blood glucose level fluctuation prediction unit Calculating the blood glucose level fluctuation predicted value during the meal based on the biological data of the user during the meal; The dietary advice creation unit The blood glucose level management system according to claim 2, wherein the dietary advice is created based on the predicted blood glucose level fluctuation value calculated during the meal.

[0112] (Appendix 6) a biometric data storage unit that stores the biometric data of the user during the meal, The dietary advice creation unit 3. The blood glucose level management system according to claim 2, wherein the biometric data stored in the biometric data storage unit is processed so as to be presented to the user as reference information.

[0113] (Appendix 7) a similar blood glucose level fluctuation model selection unit that selects a similar blood glucose level fluctuation model of another user having the biometric data similar to that of the user, The blood glucose level fluctuation prediction unit 3. The blood glucose management system according to claim 2, wherein the similar blood glucose fluctuation model is used when calculating the predicted blood glucose fluctuation value for the first time.

[0114] (Appendix 8) The blood glucose fluctuation learning unit The blood glucose level management system according to claim 7, wherein the similar blood glucose level fluctuation model is re-learned using the biometric data and dietary data of the user to be used as the blood glucose level fluctuation model of the user.

[0115] (Appendix 9) 3. The blood glucose level management system according to claim 2, further comprising a blood glucose level fluctuation model storage unit that stores the blood glucose level fluctuation model constructed by the blood glucose level fluctuation learning unit.

[0116] (Appendix 10) (a) learning the user's blood glucose fluctuations based on biometric data including at least the user's blood glucose fluctuations and meal data including at least the user's meal menu, meal amount, meal start time, and meal duration, and constructing a blood glucose fluctuation model; (b) before the user starts eating, predicting the user's blood glucose fluctuation after consuming the meal based on the meal data of the meal to be ingested and the blood glucose fluctuation model, and calculating a blood glucose fluctuation prediction value; (c) presenting the predicted blood glucose fluctuation value to the user.

[0117] (Appendix 11) (d) if the blood glucose fluctuation prediction value deviates from a preset reference value, creating dietary advice to prevent deviation from the reference value, The step (c) The dietary advice; and The blood glucose level fluctuation prediction value before the dietary method improvement based on the dietary method advice; 11. The blood glucose management method according to claim 10, further comprising a step of presenting to the user the predicted blood glucose fluctuation value after the dietary method has been improved based on the dietary method advice.

[0118] (Appendix 12) The step (a) 12. The blood glucose management method according to claim 10, further comprising a step of using data items included in the dietary data and the biological data that have a large impact on the blood glucose fluctuation of the user as features, and constructing the blood glucose fluctuation model using the features.

[0119] (Appendix 13) The step (d) A blood glucose management method according to claim 11, wherein, when the predicted blood glucose fluctuation value deviates from a predetermined reference value, the blood glucose fluctuation model used to predict the user's blood glucose fluctuation is referenced, and the dietary advice is created by resetting values ​​of changeable data items among data items that affect the blood glucose fluctuation and are included in the dietary data and the biological data.

[0120] (Appendix 14) The step (d) 12. The blood glucose level management method according to claim 11, wherein the dietary advice is created when the blood glucose level fluctuation prediction in step (b) is performed two or more times.

Claims

1. Biometric data including at least fluctuations in the user's blood glucose level; a blood glucose fluctuation learning unit that learns the blood glucose fluctuation of the user based on meal data including at least the user's meal menu, meal amount, meal start time, and meal duration, and constructs a blood glucose fluctuation model; a blood glucose level fluctuation prediction unit that, before the user starts eating, predicts blood glucose level fluctuations of the user after consuming the meal based on the meal data of the meal to be ingested and the blood glucose level fluctuation model, and calculates a blood glucose level fluctuation prediction value; a result presentation unit that presents the predicted blood glucose fluctuation value to the user.

2. a dietary advice creating unit that creates, when the blood glucose fluctuation predicted value deviates from a preset reference value, dietary advice to prevent deviation from the reference value; The dietary advice creation unit When the blood glucose fluctuation prediction value deviates from the predetermined reference value, the blood glucose fluctuation model used for predicting the blood glucose fluctuation of the user is referenced, changing values ​​of the data items of the dietary data and the data items of the life log included in the biological data that can be changed, and re-predicting the blood glucose level fluctuation of the user after consuming the meal in the blood glucose level fluctuation prediction unit, and recalculating the blood glucose level fluctuation prediction value; If the recalculated blood glucose level fluctuation prediction does not deviate from the reference value, the value of the data item used for the recalculation is set as the dietary advice, The result presentation unit The dietary advice; and The blood glucose level fluctuation prediction value before the dietary method improvement based on the dietary method advice; The blood glucose level management system according to claim 1 , further comprising: a blood glucose level prediction value after the dietary method is improved based on the dietary method advice;

3. The blood glucose fluctuation learning unit 3. The blood glucose level management system according to claim 2, wherein the blood glucose level fluctuation model is constructed by correlating the data items of the dietary data and the data items of the biological data with the blood glucose level fluctuation after a meal using statistical methods or machine learning, and using the data items that affect the blood glucose level fluctuation in the user as features.

4. The blood glucose level fluctuation prediction unit Calculating the blood glucose level fluctuation predicted value during the meal based on the biological data of the user during the meal; The dietary advice creation unit The blood glucose level management system according to claim 2 , wherein the dietary advice is prepared based on the predicted blood glucose level fluctuation value calculated during the meal.

5. a biometric data storage unit that stores the biometric data of the user during the meal, The dietary advice creation unit The blood glucose level management system according to claim 2 , wherein the biological data stored in the biological data storage unit is processed so as to be presented to the user as reference information.

6. a similar blood glucose level fluctuation model selection unit that selects a similar blood glucose level fluctuation model of another user having the biometric data similar to that of the user, The blood glucose level fluctuation prediction unit The blood glucose level management system according to claim 2 , wherein the similar blood glucose level fluctuation model is used when the predicted blood glucose level fluctuation value is calculated for the first time.

7. The blood glucose fluctuation learning unit The blood glucose level management system according to claim 6 , wherein the similar blood glucose level fluctuation model is re-learned using the biometric data and dietary data of the user to be used as the blood glucose level fluctuation model of the user.

8. The blood glucose level management system according to claim 2 , further comprising a blood glucose level fluctuation model storage unit that stores the blood glucose level fluctuation model constructed by the blood glucose level fluctuation learning unit.

9. A blood glucose level management method using computer software, comprising: (a) learning the user's blood glucose fluctuations based on biometric data including at least the user's blood glucose fluctuations and meal data including at least the user's meal menu, meal amount, meal start time, and meal duration, and constructing a blood glucose fluctuation model; (b) before the user starts eating, predicting the user's blood glucose fluctuation after consuming the meal based on the meal data of the meal to be ingested and the blood glucose fluctuation model, and calculating a blood glucose fluctuation prediction value; (c) presenting the predicted blood glucose fluctuation value to the user.

10. (d) if the blood glucose fluctuation prediction value deviates from a preset reference value, creating dietary advice to prevent deviation from the reference value, The step (d) When the blood glucose fluctuation prediction value deviates from the predetermined reference value, the blood glucose fluctuation model used for predicting the blood glucose fluctuation of the user is referenced, changing values ​​of the data items of the diet data and the data items of the life log included in the biological data that are changeable; In the step (b), the blood glucose level fluctuation prediction of the user after ingesting the meal is performed again, and the blood glucose level fluctuation prediction value is recalculated; If the recalculated blood glucose level fluctuation prediction does not deviate from the reference value, the value of the data item used for the recalculation is set as the dietary advice, The step (c) The dietary advice; and The blood glucose level fluctuation prediction value before the dietary method improvement based on the dietary method advice; 10. The blood glucose level management method according to claim 9, further comprising a step of presenting to the user the predicted value of blood glucose fluctuation after the dietary method has been improved based on the dietary method advice.

11. The step (a) 11. The blood glucose management method according to claim 10, further comprising a step of constructing the blood glucose fluctuation model using the data items that affect the blood glucose fluctuation in the user as features by correlating the data items of the dietary data and the data items of the biological data with the blood glucose fluctuation after a meal using statistical methods or machine learning.

12. The step (d) 11. The blood glucose level management method according to claim 10, wherein the dietary advice is created when the recalculation of the blood glucose level fluctuation prediction in step (b) has been performed at least once.

Images