Information processing device, information processing method, and program

Abstract

The present invention accurately predicts and notifies the degree of daily discomfort after food intake. The present invention comprises: an acquisition unit that acquires blood glucose-related data related to the blood glucose level of a subject; a prediction unit that, on the basis of the blood glucose-related data, predicts the degree of daily discomfort after a prescribed amount of time from food intake by the subject; and an output unit that outputs the result of prediction by the prediction unit.

Description

Information Processing Apparatus, Information Processing Method, and Program

[0001] The present disclosure relates to an information processing apparatus, an information processing method, and a program.

[0002] Blood glucose level is the concentration of glucose in the blood, which is regulated by diet, exercise, insulin secretion, etc., and it is desirable to be maintained within a certain range.

[0003] Patent Document 1 proposes a method including obtaining blood glucose measurement values provided by a continuous glucose monitoring (CGM) system worn by a user, obtaining additional data associated with the user obtained from one or more sources different from the CGM system, and predicting a health index of the user by processing the blood glucose measurement values and the additional data using one or more models. Patent Document 1 discloses predicting health indexes such as prediabetes, type 1 diabetes, type 2 diabetes, etc.

[0004] Japanese Patent Translation Publication No. 2023-504398

[0005] By the way, carbohydrates are a typical food that raises blood glucose levels, and when a user ingests carbohydrates, the user's fatigue may increase along with the increase in blood glucose levels. In such a current situation, there is a demand to accurately know the degree of daily discomfort such as fatigue after ingesting foods such as carbohydrates.

[0006] An object of the present disclosure is to provide an information processing apparatus, an information processing method, and a program that can accurately predict and notify the degree of daily discomfort after food intake.

[0007] To solve such problems, the information processing apparatus of the present disclosure includes an acquisition unit that acquires blood glucose-related data related to a subject's blood glucose level, a prediction unit that predicts the degree of daily discomfort at a predetermined time after the subject's food intake based on the blood glucose-related data, and an output unit that outputs the prediction result by the prediction unit.

[0008] In this disclosure, it is preferable that the system further includes a calculation unit that calculates the highest value or the slope of the blood glucose rise after the start of food intake from blood glucose-related data, and that the prediction unit uses the highest value or the slope of the blood glucose rise after the start of food intake to predict the degree of daily discomfort experienced by the subject a predetermined time after food intake.

[0009] In this disclosure, it is preferable that the prediction unit predicts the degree of daily discomfort in the subject after a predetermined time after food intake, using a trained model that has been trained to output a prediction result predicting the degree of daily discomfort in the subject after a predetermined time after food intake, when information regarding the highest value or the slope of the blood glucose rise after the start of food intake related to the subject's blood glucose level is input.

[0010] To solve the aforementioned problems, the information processing method disclosed herein is characterized by including acquiring blood glucose-related data related to the subject's blood glucose level, predicting the degree of daily discomfort the subject will experience a predetermined time after food intake based on the blood glucose-related data, and outputting the predicted result.

[0011] To solve these problems, the program of this disclosure causes a computer to acquire blood glucose-related data related to the subject's blood glucose level, predict the degree of daily discomfort the subject will experience a predetermined time after food intake based on the blood glucose-related data, and output the predicted result.

[0012] The information processing device, information processing method, and program related to this disclosure can accurately predict and inform the degree of everyday discomfort after food ingestion.

[0013] This figure shows the schematic configuration of the information processing system according to the first embodiment of this disclosure. This figure shows the schematic configuration of the server device. This is a graph showing the classification results by a trained model (support vector machine). This is a sequence chart that predicts the degree of fatigue in a subject and notifies the prediction result in the first embodiment. This is a flowchart that predicts the degree of fatigue in a subject. This figure shows the schematic configuration of the information processing system according to the second embodiment of this disclosure. This is a sequence chart that predicts the degree of fatigue in the second embodiment and notifies the prediction result. This figure shows the prediction result of classifying the degree of fatigue from unknown data (B) using a trained model generated using known data (A) in one embodiment.

[0014] <First Embodiment> Hereinafter, an information processing device, information processing method, and program relating to one aspect of the first embodiment will be described with reference to the figures. However, it should be noted that the technical scope of this disclosure is not limited to these embodiments, but extends to the disclosures described in the claims and their equivalents.

[0015] Figure 1 is a diagram showing the schematic configuration of the information processing system 1 according to the first embodiment of this disclosure.

[0016] As shown in Figure 1, the information processing system 1 is used by an administrator, separate from the subject, to predict the degree of fatigue experienced by one or more subjects after food intake, and comprises one or more blood glucose meters LB, one or more terminal devices 200, and a server device 100. The subject possesses both a blood glucose meter LB and a terminal device 200.

[0017] In the information processing system 1, the blood glucose meter LB, terminal device 200, and server device 100 are connected to each other via network N so that they can communicate with one another. Network N is a wired network such as the Internet or an intranet. Network N may also be a wireless network such as a wireless LAN (Local Area Network).

[0018] A blood glucose meter (LB) is a device that measures the concentration of glucose (blood sugar) in a subject's blood or interstitial fluid (i.e., blood glucose level). Examples of blood glucose meters (LB) that measure blood glucose levels include the FreeStyle Libre® and the FreeStyle Libre 2 (both manufactured by Abbott). After measuring the subject's blood glucose level, the blood glucose meter (LB) transmits that blood glucose level to a server device (100) via the network (N). Alternatively, the blood glucose meter (LB) may transmit the measured blood glucose level to a terminal device (200), which in turn transmits it to the server device (100) via the network (N).

[0019] The terminal device 200 is an example of an information processing device. The terminal device 200 is, for example, a smartphone. However, the terminal device 200 may also be a personal computer, a notebook PC, a tablet PC, etc. The terminal device 200 acquires the input results of a questionnaire regarding the degree of the subject's daily discomfort and transmits the input results to the server device 100 via the network N.

[0020] "Daily discomfort" refers to temporary unhealthy states in daily life. Specific examples of daily discomfort include increased fatigue, increased drowsiness, decreased concentration, decreased attention span, lethargy, stress, decreased performance, decreased vitality, increased sluggishness, and decreased mood. Daily discomfort may consist of one of these, or a combination of two or more. In this embodiment, the degree of fatigue will be described below as an example of daily discomfort.

[0021] The fatigue questionnaire administered via the terminal device 200 uses a visual analogue scale called VAS (Visual Analogue Scale). In the VAS, the degree of fatigue of the subject is indicated by the length of a 100 mm straight line, and the subject specifies a position on the straight line corresponding to the degree of fatigue. In this case, a 100 mm straight line is displayed on the display screen of the terminal device 100, and the subject specifies a position on the straight line corresponding to their degree of fatigue.

[0022] In the VAS questionnaire, if the subject feels not tired, they specify a position, for example, between 0 mm and 20 mm on a 100 mm line; if they feel slightly tired, they specify a position, for example, between 21 mm and 50 mm on a 100 mm line; and if they feel very tired, they specify a position, for example, greater than 51 mm on a 100 mm line. The subject completes the VAS questionnaire via the terminal device 200 at the time their blood glucose level is measured by the blood glucose meter LB.

[0023] The terminal device 200 transmits to the server device 100, as the input result of the questionnaire, a predetermined degree of fatigue ("not tired", "tired", or "very tired") according to the VAS value specified by the subject. The server device 100 stores the blood glucose value received from the blood glucose meter LB and the input result of the questionnaire (degree of fatigue) received from the terminal device 200 in association with each other.

[0024] In addition to the VAS questionnaire, it is also acceptable to use various other questionnaires, such as the NRS (Numerical Rating Scale) or face scales, to obtain the degree of fatigue as input from the questionnaire.

[0025] Figure 2 shows a schematic configuration of the server device 100.

[0026] The server device 100 includes a communication unit 101, a storage unit 110, and a processing unit 120, etc. The communication unit 101, the storage unit 110, and the processing unit 120 are interconnected via a CPU (Central Processing Unit) bus or the like.

[0027] The communication unit 101 has an antenna for transmitting and receiving wireless signals and a wireless communication interface circuit that conforms to a communication protocol such as a wireless LAN (Local Area Network), and communicates with the network N in accordance with a communication standard such as a wireless LAN.

[0028] The communication unit 101 sends data received from the blood glucose meter LB, terminal device 200, and external devices (not shown) via the network N to the processing unit 120. The communication unit 101 also transmits data received from the processing unit 120 to the terminal device 200, etc., via the network N. The communication unit 101 may have a wireless communication interface circuit conforming to communication standards such as LTE (Long Term Evolution) or 5G, and may communicate with the network N via a base station. Alternatively, the communication unit 101 may have a wired communication interface circuit conforming to communication protocols such as TCP / IP (Transmission Control Protocol / Internet Protocol), and may communicate with the network N according to communication standards such as Ethernet (registered trademark).

[0029] The storage unit 110 includes memory devices such as RAM (Random Access Memory) and ROM (Read Only Memory), fixed disk devices such as hard disks, or portable storage devices such as flexible disks and optical disks.

[0030] Furthermore, the storage unit 110 stores various data such as computer programs, databases, and tables used for various processes of the server device 100. The computer programs may be installed into the storage unit 110 from a computer-readable portable recording medium using a known setup program or the like. Examples of portable recording media include CD-ROMs (compact disc read-only memory) and DVD-ROMs (digital versatile disc read-only memory). The computer programs may also be stored on a recording medium owned by a predetermined server and installed via the network N.

[0031] In addition to the above, the memory unit 110 stores a prediction program 111 and a trained model 112, etc. The prediction program 111 is an example of the program of this disclosure.

[0032] The prediction program 111 is software that causes the processing unit 120 to perform a series of processes to notify the subject of the prediction result by having the processing unit 120 predict the degree of fatigue of the subject using a trained model 112 and then transmitting the prediction result to the terminal device 200.

[0033] The trained model 112 is a model that classifies the degree of fatigue experienced by a subject between food intake and a predetermined time, based on blood glucose-related data acquired by the subject. The trained model 112 is, for example, a support vector machine and is pre-trained through supervised learning. The trained model 112 is not limited to a support vector machine; it may also be a classification model based on algorithms such as random forests, logistic circuits, or k-nearest neighbors.

[0034] Furthermore, the blood glucose-related data input to the trained model 112 is, for example, the concentration of glucose (blood glucose level) in the blood. Blood glucose level is just one example of blood glucose-related data. However, it is not limited to this; insulin, HbA1c, HOMA-IR, and HOMA-β may also be used as blood glucose-related data.

[0035] The food consumed by the subjects is a carbohydrate-rich diet, primarily consisting of carbohydrates. Carbohydrates are sugars that are absorbed into the body and used as an energy source. Examples of sugars include monosaccharides such as glucose, fructose, and galactose; disaccharides such as sucrose, lactose, and maltose; oligosaccharides of three or more types; polysaccharides such as starch and dextrin; and sugar alcohols such as sorbitol and xylitol. Carbohydrates may consist of one type of sugar or a combination of two or more types of sugars. In addition, the food may also consist of proteins, lipids, etc., other than carbohydrates, or combinations thereof.

[0036] The trained model (support vector machine) 112 is pre-programmed with training data including the peak blood glucose elevation fluctuation ΔCmax (mg / dL) and the degree of fatigue experienced by multiple subjects after the start of food intake. The peak blood glucose elevation fluctuation ΔCmax (mg / dL) after the start of food intake is the explanatory variable, and the degree of fatigue is the dependent variable.

[0037] The peak value of blood glucose elevation fluctuation ΔCmax (mg / dL) after the start of food intake is the difference between the blood glucose level BG0 immediately before food intake (at 0 minutes) and the maximum value of the highest blood glucose level (hereinafter referred to as "maximum blood glucose level") Cmax that occurred between the start of food intake and 170 minutes later. In this disclosure, the peak value of blood glucose elevation ΔCmax (mg / dL) is the difference between the blood glucose level BG0 immediately before food intake (at 0 minutes) and the highest blood glucose level (hereinafter referred to as "maximum blood glucose level") Cmax that occurred after the start of food intake. After the start of food intake, blood glucose levels usually rise within about 15 minutes, and reach their peak value by 170 minutes. In one embodiment, it is preferable that the peak value of blood glucose elevation fluctuation up to 170 minutes after the start of food intake be defined as ΔCmax (mg / dL).

[0038] As mentioned above, the degree of fatigue is categorized into three predetermined groups: "not tired," "tired," or "very tired," based on the position on the Visual Analog Scale (VAS). For example, if the VAS value is within the range of 0 mm to 20 mm, it is classified into category A, "not tired"; if the VAS value is within the range of 21 mm to 50 mm, it is classified into category B, "tired"; and if the VAS value is greater than 51 mm, it is classified into category C, "very tired." The range of VAS values ​​corresponding to the degree of fatigue can be changed as needed.

[0039] Figure 3 is a graph showing the classification results of the degree of fatigue experienced by multiple subjects, as determined by a trained model (support vector machine).

[0040] A support vector machine, for example, when performing classification in two dimensions, plots the data for multiple subjects on a horizontal axis with the maximum blood glucose elevation fluctuation ΔCmax (mg / dL) after the start of food intake and the maximum time to reach Tmax on the vertical axis, and classifies the data into three groups using two decision boundaries L1 and L2. In this case, the subject's data is the maximum blood glucose elevation fluctuation ΔCmax (mg / dL) after the start of food intake, and the maximum time to reach Tmax is the time to reach the maximum blood glucose elevation fluctuation ΔCmax (mg / dL) obtained between the time immediately before food intake (0 minutes) and 170 minutes after food intake.

[0041] In this case, the subjects in the data group of the area smaller than the decision boundary L1 are classified into the "not tired" category A, the subjects in the data group of the area larger than the decision boundary L1 and smaller than the decision boundary L2 are classified into the "tired" category B, and the subjects in the data group of the area larger than the decision boundary L2 are classified into the "very tired" category C. The decision boundaries L1 and L2 are examples of boundary data.

[0042] Therefore, when the maximum value ΔCmax (mg / dL) of the blood glucose increase fluctuation after the start of food intake of a plurality of subjects is input as an explanatory variable, and the degree of fatigue associated with the maximum value ΔCmax (mg / dL) of the blood glucose increase fluctuation after the start of food intake is input as an objective variable, the learned model 112 is pre-learned to classify (predict) and output the degree of fatigue of the subjects by the decision boundaries L1 and L2.

[0043] The processing unit 120 operates based on a computer program stored in the storage unit 110 in advance. The processing unit 120 is, for example, a CPU. As the processing unit 120, a DSP (digital signal processor), LSI (large scale integration), ASIC (Application Specific Integrated Circuit), FPGA (Field-Programmable Gate Array), or the like may be used.

[0044] The processing unit 120 is connected to the communication unit 101 and the storage unit 110 and controls each unit. The processing unit 120 functions as an acquisition unit 121, a calculation unit 122, a prediction unit 123, and an output unit 124 by operating according to a computer program stored in the storage unit 110.

[0045] The processing unit 120 classifies the degree (category) of fatigue of the subject from food intake to a predetermined time later based on the blood glucose-related data (blood glucose value) acquired by the subject by the learned model 112 stored in the storage unit 110 in advance, and outputs the classification result as a prediction result to the terminal device 200 via the communication unit 101.

[0046] FIG. 4 is a sequence chart that predicts the degree of fatigue that occurs within 180 minutes after a subject's food intake and notifies the prediction result in the first embodiment.

[0047] Next, an example of the processing in the information processing system 1 that predicts the degree of fatigue of the subject and notifies the prediction result will be described while referring to the sequence chart of FIG. 4. The operations of the server device 100 described below are mainly executed by the processing unit 120 in cooperation with the respective elements of the server device 100 based on the prediction program 111 stored in the storage unit 110 in advance.

[0048] First, the blood glucose meter LB measures the blood glucose level of the subject (step L1). The subject in this case is a subject who does not currently know how much fatigue will occur within a predetermined time (for example, 180 minutes) after food intake.

[0049] The blood glucose meter LB measures the blood glucose level of the subject that occurs from immediately before food intake (0 minutes) to a predetermined time after food intake (for example, 170 minutes) at a predetermined time interval. The blood glucose meter LB measures the blood glucose level of the subject at, for example, 10-minute intervals. However, the blood glucose meter LB may measure the blood glucose level at 1-minute intervals or 30-minute intervals.

[0050] Next, the blood glucose meter LB transmits the blood glucose level of the subject measured at a predetermined time interval and the respective arrival times T when the blood glucose levels are acquired at a predetermined time interval to the server device 100 via the network N (step L2). The arrival time T is the respective time from immediately before food intake (0 minutes) until the blood glucose level is measured at a predetermined time interval. Therefore, the arrival time T from immediately before food intake (0 minutes) to the maximum blood glucose level Cmax is the maximum arrival time Tmax. Note that the blood glucose meter LB may transmit a plurality of blood glucose levels measured at a predetermined time interval from the subject to the server device 100 in a lump at 170 minutes later.

[0051] The acquisition unit 121 of the processing unit 120 in the server device 100 acquires the blood glucose level of the subject and the arrival time T from the blood glucose meter LB via the communication unit 101 and temporarily stores them in the storage unit 110 (step S101).

[0052] Next, the calculation unit 122 calculates the maximum value of blood glucose elevation fluctuation after the start of food intake, ΔCmax (mg / dL), which is the difference between the blood glucose level BG0 immediately before food intake (at 0 minutes) and the maximum blood glucose level max that rose the most during the period from food intake to 170 minutes later, and the maximum time to reach the maximum value of blood glucose elevation fluctuation after the start of food intake, ΔCmax (mg / dL), Tmax (step S102).

[0053] The peak value of blood glucose elevation fluctuation ΔCmax (mg / dL) after the start of food intake is the difference obtained by subtracting the blood glucose level BG0 immediately before food intake (at 0 minutes) from the maximum blood glucose level Cmax after food intake. In other words, the peak value of blood glucose elevation fluctuation ΔCmax (mg / dL) after the start of food intake is the maximum increase when the maximum blood glucose level max is reached after food intake. The peak value of blood glucose elevation fluctuation ΔCmax (mg / dL) after the start of food intake is calculated by the following equation (1).

[0054] ΔCmax = Cmax - BG0 …………………………………………(1) Cmax: Maximum blood glucose level (mg / dl) BG0: Blood glucose level at 0 minutes (mg / dl)

[0055] Next, the prediction unit 123 predicts the degree of fatigue experienced by the subject by inputting the highest value ΔCmax (mg / dL) of the blood glucose elevation fluctuation after the subject starts eating into the trained model 112 (step S103).

[0056] Figure 5 is a flowchart for predicting the degree of fatigue experienced by the subjects.

[0057] The prediction unit 123 inputs the highest value ΔCmax (mg / dL) of blood glucose elevation fluctuation after the start of food intake in the subject to the trained model (support vector machine) 112, and classifies the subject's data corresponding to the highest value ΔCmax (mg / dL) of blood glucose elevation fluctuation after the start of food intake into one of the regions (categories) divided by the decision boundaries L1 and L2 (step S1031).

[0058] In other words, the prediction unit 123 can predict the degree of fatigue experienced by a subject after a predetermined time has passed since food intake by utilizing decision boundaries L1 and L2, which are pre-set by the trained model 112 and classify the degree of fatigue experienced after a predetermined time in accordance with the highest value ΔCmax (mg / dL) of blood glucose elevation fluctuation after the subject starts eating food.

[0059] Next, the prediction unit 123 determines whether the subject's data corresponding to the highest value ΔCmax (mg / dL) of blood glucose elevation fluctuation after the start of food intake can be classified into a region smaller than the decision boundary L1 (step S1032). If the prediction unit 123 predicts that the subject's data can be classified into a region smaller than the decision boundary L1 (step S1032: YES), it obtains a classification result in which subjects whose blood glucose elevation fluctuation after the start of food intake reached the highest value ΔCmax (mg / dL) within 170 minutes after food intake are classified into category A.

[0060] In this case, the prediction unit 123 predicts that the subject is classified as Category A ("not tired") (step S1033). This means that the subject can remain active without experiencing an increase in fatigue for 180 minutes after food intake.

[0061] Next, the prediction unit 123 predicts that if the subject's data cannot be classified into a region smaller than the decision boundary L1 (step S1032: NO), the subject will not be classified into category A ("not tired").

[0062] Next, the prediction unit 123 determines whether the subject's data can be classified into a region greater than the decision boundary L1 and smaller than the decision boundary L2 (step S1034). If the prediction unit 123 predicts that the subject's data can be classified into a region greater than the decision boundary L1 and smaller than the decision boundary L2 (step S1034: YES), it obtains a classification result indicating that the subject is classified into category B when the highest value ΔCmax (mg / dL) of blood glucose elevation fluctuation after the start of food intake occurs within 170 minutes after food intake.

[0063] In this case, the prediction unit 123 predicts that the subject is classified as category B ("tired") (step S1035). This means that the subject will experience increased fatigue and become tired within 180 minutes of food intake.

[0064] Next, the prediction unit 123 can predict that if the subject's data cannot be classified into a region greater than the decision boundary L1 and less than the decision boundary L2, the subject's data can be classified into a region greater than the decision boundary L2 (step S1034: NO). Therefore, it obtains a classification result that the subject's data is classified into category C when the highest value ΔCmax (mg / dL) of blood glucose elevation fluctuation after the start of food intake occurs within 170 minutes after food intake.

[0065] In this case, the prediction unit 123 predicts that the subject is classified as category C ("very tired") (step S1036). This means that the subject will experience a significant increase in fatigue within 180 minutes of food intake and will become very tired.

[0066] Next, the prediction unit 123 determines which category the degree of fatigue in the subject falls into based on the classification result of step S1033, step S1035, or step S1036 (step S1037). After the prediction unit 223 stores the classification result for the subject in the storage unit 110, it transmits the classification result as a prediction to the subject's terminal device 200 via the communication unit 101 (step S104).

[0067] The terminal device 200 receives the subject's prediction result (classification result) from the server device 100 via the network N (step S201). The terminal device 200 displays the received prediction result (classification result), which is category A, B, or C, on the display unit (step S202).

[0068] This allows subjects to easily obtain the degree of fatigue experienced during the period from food intake up to 180 minutes later from the server device 100 via the terminal device 200, simply by transmitting blood glucose-related data (blood glucose levels) measured by the blood glucose meter LB to the server device 100.

[0069] By visually checking the category of the predicted result displayed on the terminal device 200, the subject can easily recognize whether, within 180 minutes after food intake, their fatigue level will remain unchanged and they will be in a "not tired" state, their fatigue level will increase and they will be in a "tired" state, or their fatigue level will increase significantly and they will be in a "very tired" state. Thus, the subject can predict how their physical condition will change within 180 minutes after food intake and take preventative measures.

[0070] For example, if a subject is a vehicle driver and receives a Category C ("very tired") prediction within 180 minutes of food intake, they can take preventative measures, such as taking fatigue-recovering medication, to avoid such physical discomfort.

[0071] As detailed above, the server device 100 acquires the subject's blood glucose-related data (blood glucose level), predicts the degree of fatigue a predetermined time after food intake based on that blood glucose level, and can notify the subject of the prediction result via the terminal device 200.

[0072] <Second Embodiment> Hereinafter, an information processing device, information processing method, and program relating to one aspect of the second embodiment will be described with reference to the figures. However, it should be noted that the technical scope of this disclosure is not limited to these embodiments, but extends to the disclosures described in the claims and their equivalents.

[0073] Figure 6 is a diagram showing the schematic configuration of the information processing system 1A according to the second embodiment of this disclosure.

[0074] As shown in Figure 6, where the corresponding parts in Figure 1 are denoted by the same reference numerals, the information processing system 1A is used by an administrator, separate from the subject, to predict the degree of fatigue experienced by the subject after food intake, and includes one or more blood glucose meters LB and one or more terminal devices 400.

[0075] The configuration of the blood glucose meter LB in the information processing system 1A is the same as in the first embodiment, but the terminal device 400 differs from the terminal device 200 in the first embodiment in that it has a communication unit 101, a storage unit 110, and a processing unit 120 of the server device 100. Therefore, in this information processing system 1A, the degree of the subject's daily discomfort can be predicted by the terminal device 400 without using the server device 200 in the first embodiment, and the prediction result can be output.

[0076] Figure 7 is a sequence chart that predicts the degree of fatigue and notifies the user of the prediction result in the second embodiment.

[0077] First, the blood glucose meter LB measures the subject's blood glucose-related data (blood glucose level) (Step L1). The blood glucose meter LB measures the blood glucose level of a new subject at predetermined time intervals (for example, every 10 minutes).

[0078] Next, the blood glucose meter LB transmits the blood glucose levels of new subjects measured at predetermined time intervals, and the arrival time T for each of these blood glucose levels at predetermined time intervals, to the server device 200 via the network N (step L2).

[0079] The acquisition unit 121 of the processing unit 120 in the terminal device 400 acquires the blood glucose level and arrival time T of a new subject from the blood glucose meter LB via the communication unit 101, and temporarily stores them in the storage unit 110 (step S401).

[0080] Next, the calculation unit 122 of the terminal device 400 calculates, as blood glucose-related indicators, for example, the difference between the blood glucose level BG0 immediately before food intake (at 0 minutes) and the maximum blood glucose level Cmax that rises to its highest point between food intake and 170 minutes later, which is the highest value of blood glucose elevation fluctuation after the start of food intake, ΔCmax (mg / dL), and the maximum time to reach the highest value of blood glucose elevation fluctuation ΔCmax (mg / dL) after the start of food intake, Tmax (step S402).

[0081] Next, the prediction unit 123 of the terminal device 400 predicts the degree of fatigue of a new subject based on blood glucose-related indicators, such as the highest value ΔCmax (mg / dL) of the blood glucose rise fluctuation after the start of food intake and the time to reach the maximum Tmax (step S403). The prediction in step S403 is the same as in step S103 (Figure 4) described above.

[0082] The terminal device 400 displays the category of the new subject, which is the prediction result (classification result) predicted in step S403, on the display unit (step S404).

[0083] In this way, the terminal device 400 can predict the degree of fatigue that a new subject will experience up to 180 minutes after food intake, based on the blood glucose level received from the blood glucose meter LB, and notify the patient of the prediction result.

[0084] This allows new subjects to easily recognize the degree of fatigue they are expected to experience within 180 minutes of food intake by visually checking the category displayed on the terminal device 400.

[0085] <One Example> Figure 8 shows the prediction results obtained by classifying the degree of fatigue from unknown data (B) using a trained model generated using known data (A) in one example.

[0086] In this study, 19 healthy men and women were used as subjects to predict the degree of daily discomfort that would occur between the carbohydrate food load (ingestion) and a predetermined time interval. (A) The training data consisted of data from 15 of the 19 subjects, and (B) the classification results consisted of data from 4 of the 19 subjects. Subjects were prohibited from eating or drinking anything other than water for 5 hours prior to the carbohydrate food load (ingestion). The carbohydrate food was prepared by combining thin udon noodles (dried noodles) (manufactured by Sanuki Bussan Co., Ltd.) and udon soup (powder) (manufactured by Higashimaru Shoyu Co., Ltd.) (test food). One serving of the test food (containing 85.1 g of carbohydrates) was administered to the subjects within 10 minutes.

[0087] Blood glucose levels were measured using a blood glucose meter (LB) before carbohydrate loading (ingestion) and from the start of ingestion until 170 minutes later. Specifically, blood glucose levels were measured immediately before carbohydrate loading (ingestion) (0 minutes), and then every 15 minutes from the start of carbohydrate loading (ingestion) until 170 minutes later. Based on the blood glucose levels, the maximum value of the blood glucose rise fluctuation after the start of ingestion of the test food, ΔCmax (mg / dL), and the time to reach Tmax were determined. In Figures 8(A) and (B), the time to reach Tmax is omitted. Blood glucose levels were measured using a FreeStyle Libre® (Abbott) blood glucose meter (LB).

[0088] Immediately before carbohydrate loading (ingestion) (0 minutes) and from the start of ingestion to 170 minutes later, the peak value ΔCmax (mg / dL) and the time to reach it Tmax for four new subjects were input into a trained model 112. As a result, new subjects whose peak value ΔCmax (mg / dL) after the start of food ingestion reached 162 were classified into category B, "tired," while new subjects whose peak value ΔCmax (mg / dL) after the start of food ingestion reached 132 were classified into category A, "not tired."

[0089] In fact, the fatigue levels of four new subjects were evaluated using a Visual Analog Scale (VAS). The new subjects were asked to mark their fatigue level on a straight line of the VAS.

[0090] The time Tmax (minutes) from the start of carbohydrate diet loading to reaching ΔCmax for the subjects ranged from 15 to 165 minutes.

[0091] As a result, the VAS score of a new subject whose peak blood glucose fluctuation ΔCmax (mg / dL) after the start of food intake was 162 fell within the range corresponding to "tired," while the VAS score of a new subject whose peak blood glucose fluctuation ΔCmax (mg / dL) after the start of food intake was 132 fell within the range corresponding to "not tired." As a result, the F1 score, which is the harmonic mean of precision and recall, was 0.75. Incidentally, an F1 score of 0.7 or higher is generally considered to indicate a good learning model.

[0092] <Other Embodiments> In the first and second embodiments, the degree of fatigue was predicted by the trained model 112 using the maximum value ΔCmax (mg / dL) of blood glucose elevation fluctuation after the start of food intake, which was calculated by the calculation unit 122. However, the degree of fatigue may be predicted using the slope Gs of blood glucose elevation calculated by the calculation unit 122. In this case, the slope Gs of blood glucose elevation is the value obtained by subtracting the blood glucose value BG0 immediately before food intake (at 0 minutes) from the maximum blood glucose value Cmax when the maximum value ΔCmax (mg / dL) of blood glucose elevation fluctuation after the start of food intake was obtained, and dividing the result by the time to reach the maximum value Tmax. The slope Gs of blood glucose elevation is calculated by the following equation (2).

[0093] Gs = (Cmax - BG0) / Tmax ………………………………(2) BG0: Blood glucose level at 0 minutes (mg / dl) Tmax: Time to reach maximum (minutes)

[0094] In the first and second embodiments, the degree of fatigue in the subject was predicted using the highest value ΔCmax (mg / dL) of the blood glucose rise after the start of food intake. However, the degree of fatigue may be predicted by the prediction unit 123 using the maximum blood glucose value Cmax after the start of food intake. Alternatively, the prediction unit 123 may predict the degree of fatigue using a table in which the range of the highest value ΔCmax (mg / dL) of the blood glucose rise after the start of food intake or the range of the maximum blood glucose value Cmax is pre-associated with the degree of fatigue, without using the trained model 112.

[0095] In the first and second embodiments, the degree of fatigue experienced by the subject was predicted up to 180 minutes after food intake. However, the embodiment is not limited to this, and the degree of fatigue experienced up to 60 minutes, 120 minutes, 240 minutes, or other predetermined time periods after food intake may also be predicted.

[0096] In the first and second embodiments, the data was classified into three groups using two decision boundaries L1 and L2 generated by a support vector machine. However, the invention is not limited to this, and subjects may be classified into "not tired" category A and "tired" category B using one decision boundary, or further classified using three or more decision boundaries.

[0097] In the first and second embodiments, the terminal device 200 transmits to the server device 100, based on the input result of the questionnaire, a predetermined degree of fatigue ("not tired", "tired", or "very tired") according to the range of VAS values ​​specified by the subject. However, the system is not limited to this, and the terminal device 200 may transmit the subject's AVS value to the server device 100, and the processing unit 120 may set the degree of fatigue ("not tired", "tired", or "very tired") according to the range of VAS values.

[0098] 1, 1A... Information processing system, LB... Blood glucose meter, 200, 400... Terminal device, 100... Server device, 101... Communication unit, 110... Storage unit, 111... Prediction program, 112... Trained model, 120... Processing unit, 121... Acquisition unit, 122... Calculation unit, 123... Prediction unit, 124... Output unit.

Claims

1. An information processing device comprising: an acquisition unit that acquires blood glucose-related data relating to the blood glucose level of a subject; a prediction unit that predicts the degree of daily discomfort of the subject a predetermined time after food intake based on the blood glucose-related data; and an output unit that outputs the prediction results from the prediction unit.

2. The information processing apparatus according to claim 1, further comprising a calculation unit that calculates the highest value of the blood glucose rise fluctuation or the slope of the blood glucose rise after the start of food intake from the blood glucose-related data, wherein the prediction unit uses the highest value of the blood glucose rise fluctuation after the start of food intake or the slope of the blood glucose rise to predict the degree of the subject's daily discomfort after a predetermined time has passed since food intake.

3. When information relating to the highest value of the fluctuation in blood glucose elevation after the start of food intake or the slope of the blood glucose elevation related to the subject's blood glucose level is input, the prediction unit predicts the degree of the subject's daily discomfort after a predetermined time after food intake, using a trained model that has been trained to output a prediction result predicting the degree of the subject's daily discomfort after a predetermined time after food intake, as described in claim 2.

4. An information processing method characterized by including: acquiring blood glucose-related data related to the blood glucose level of a subject; predicting the degree of daily discomfort of the subject a predetermined time after food intake based on the blood glucose-related data; and outputting the predicted result.

5. A program characterized by causing a computer to perform the following actions: acquire blood glucose-related data related to the blood glucose level of a subject; predict the degree of daily discomfort of the subject a predetermined time after food intake based on the blood glucose-related data; and output the predicted result.

Images