JP7517668B2 - Health information provision system and health information provision program - Google Patents

Description

The present invention relates to a health information provision system and a health information provision program that provide health information to users.

Conventionally, there is known a body composition analyzer that measures body composition based on biometric information such as height, weight, age, and sex, and the bioimpedance of each part of the human body obtained by measurement (see, for example, Patent Document 1).

Meanwhile, health checkups involve blood tests and other procedures to measure biochemical test values such as HDL cholesterol, γ-GTP, and blood sugar. Health checkups may also include an assessment of health risks and advice on maintaining or restoring health based on these biochemical test values, biological information, and lifestyle habits such as smoking habits and alcohol intake.

JP 2017-23311 A

With conventional body composition scales, it was not possible to determine from the measurement results how changes in an individual's body water, body fat, and relative changes in muscle were specifically affecting that person's health condition.

In addition, disease risks such as lifestyle-related diseases and arteriosclerosis are things that change and accumulate daily due to changes in the body and body composition caused by changes in lifestyle. However, unless people receive annual health checkups or undergo medical checkups, they will not have the opportunity to notice changes in their internal risks, and so there is a risk that it will be too late to take action against the disease, and there has also been no awareness of the need to deal with disease risks.

Health risk assessments that combine past health checkup results trends with questionnaire surveys about lifestyle and prediction of future risk using model calculations are becoming more common in health checkup services. However, managing past data and answering questionnaires is tedious, and there is also the drawback that the evaluation lacks objectivity due to the inclusion of subjectivity. In other words, conventional predictions of future risk lack convenience, simplicity, and objectivity. Furthermore, as responses to questions are the main focus, ultimately it is not possible to detect abnormalities in the body unless one is aware of them.

The present invention provides a health information provision system and a health information provision program that can estimate biochemical test values in a simpler manner without relying on blood tests, etc., or that can evaluate health risks and provide advice for maintaining or restoring health.

The present disclosure employs the following technical means to solve the above problems. The claims and the symbols in parentheses described in this section are examples showing the corresponding relationship to the specific means described in the following embodiment as one aspect, and do not limit the technical scope of the present invention.

The inventors of the present application discovered a correlation between body composition information and biochemical test values measured during health checkups, leading to the invention of the present application.

To achieve the above object, the health information providing system of the present invention includes a body composition acquisition unit that acquires body composition information of a user, a biochemical test value estimation unit that estimates biochemical test values based on the body composition information, and an output unit that outputs biochemical test values or health information based on the biochemical test values.

With this configuration, the biochemical test value estimation unit estimates the biochemical test value based on the body composition information, so that the biochemical test value can be easily obtained without performing relatively expensive tests such as blood tests. The body composition information may be obtained, for example, using a body composition meter using the bioimpedance method or a weighing scale with a body composition measurement function, or may be obtained by measuring subcutaneous fat thickness using the caliper method. The user can know the estimated biochemical test value by a procedure similar to that for measuring body composition.

The device may further include a biometric information acquisition unit that acquires the user's biometric information, and the biochemical test value estimation unit may estimate the biochemical test value based on the biometric information.

This configuration allows biochemical test values to be accurately estimated based on biological information and body composition information.

The biometric information acquisition unit may acquire at least the user's bioimpedance as biometric information, and the body composition acquisition unit may acquire body composition information by performing calculations using the bioimpedance.

With this configuration, body composition information can be obtained by performing calculations using bioimpedance.

The biometric information acquisition unit may acquire biometric impedance through measurement.

Bioimpedance is relatively easy to measure, so this configuration makes it easy to obtain body composition information.

The body composition acquisition unit may acquire body composition information by accepting input of body composition information.

With this configuration, even if the health information system does not have the functionality to measure bioimpedance, biochemical test values can be estimated based on body composition information.

The biometric information acquisition unit may acquire at least the user's height, weight, age, and gender as biometric information.

With this configuration, the biochemical test value estimation unit can effectively estimate biochemical test values based on not only the user's body composition information, but also on the user's biometric information including the user's height, weight, age, and sex. Note that the biometric information acquisition unit may acquire weight information by measurement.

The device may further include a health risk assessment unit that assesses health risks based on biochemical test values, and the output unit may output the health risks as health information.

This configuration allows health risks assessed based on biochemical test values to be provided as health information, providing useful information to users who are unable to understand health risks based on estimated biochemical test values alone.

The device may further include a health advice determination unit that determines advice regarding maintaining or recovering health based on the biochemical test values, and the output unit may output the advice as health information.

This configuration makes it possible to provide advice on maintaining or restoring health based on biochemical test values as health information, and to provide useful information to users who are unsure of what to do based on estimated biochemical test values alone.

The biochemical test value estimation unit may estimate the biochemical test value using a learning model that estimates the biochemical test value using body composition information as input.

With this configuration, by inputting the user's body composition information into the learning model, it is possible to obtain biochemical test values corresponding to that body composition information. The learning model may be based on artificial intelligence such as machine learning, or may be obtained by other statistical methods. For example, it may be a model obtained by learning multiple combinations of body composition information and biochemical test values as learning data, or it may be a regression equation or algorithm that specifies the relationship between multiple types of body composition information and a certain biochemical test value, or it may be a neural network that outputs a certain biochemical test value when multiple types of body composition information are input.

The biochemical test value estimation unit may estimate the biochemical test value by referring to a table that specifies the relationship between body composition information and the biochemical test value.

This configuration makes it possible to easily and quickly estimate biochemical test values for body composition information without the need for complex calculations.

The biochemical test value estimation unit may set the user's biochemical test value obtained by actual measurement as the actual measured value, set the biochemical test value estimated using the learning model as the reference estimated value, and adjust the learning model based on the actual measured value and the reference estimated value.

With this configuration, the learning model can be adjusted and used based on actual measurements (e.g., values measured by a blood test or the like in an actual health check), and even if the estimation of the learning model does not match due to individual differences, the influence of such individual differences can be reduced. For example, if the actual measurement value of a certain biochemical test value is 100, but the biochemical test value estimated by the biochemical test value estimation unit is 90, this 90 can be used as a reference estimate, and the learning model can be adjusted to produce a value 10 (=100-90) larger than the original value. By adjusting the learning model in this way, the biochemical test value can be estimated based on the actual measurement value, and the biochemical test value according to changes in body composition and biological information can be estimated based on the actual measurement value. For this reason, it is desirable that the reference estimate value is a biochemical test value estimated using the user's biological information and body composition information at the same time as the actual measurement. For example, the reference estimate value may be a biochemical test value estimated using the user's biological information at the date and time closest to the actual measurement. Furthermore, the biological information and body composition information from the same period as the actual measurement may be estimated before the actual measurement, or may be estimated after the actual measurement.

The biochemical test value estimation unit may further adjust the learning model based on the time elapsed since the actual measurement was performed.

This configuration allows the learning model to be adjusted over time to approach the original learning model, for example.

The biochemical test value estimation unit may further adjust the learning model based on the difference between the actual measured value and the estimated biochemical test value.

With this configuration, for example, the learning model can be adjusted so that it approaches the original learning model as the difference between the actual measured value and the estimated biochemical test value increases.

The biochemical test value estimation unit may adjust the learning model based on multiple actual measurement values obtained by actual measurements at multiple different times.

With this configuration, when there are multiple actual measured values, they can be used to adjust the learning model. For example, when the results (actual measured values) of a yearly health check have been accumulated over several years, these multiple actual measured values can be used to adjust the learning model. In this case, weighting may be applied so that more important importance is placed on the newer actual measured values.

The biochemical test value estimation unit may determine a reference estimated value by performing statistical processing on multiple biochemical test values obtained by multiple estimations.

This configuration makes it possible to obtain an appropriate reference estimate. In other words, compared to obtaining a reference estimate using biochemical test values obtained by a single estimation, a highly accurate reference estimate with reduced variability can be obtained by statistically processing biochemical test values obtained by multiple estimations. Here, the statistical processing may involve simple averaging, or averaging after excluding outliers.

The biochemical test value estimation unit may set the user's biochemical test value obtained by actual measurement as the actual measured value, set the biochemical test value estimated using the learning model as a reference estimated value, and adjust the biochemical test value estimated by the learning model based on the actual measured value and the reference estimated value.

With this configuration, the learning model is not adjusted based on the actual measurement value, but the biochemical test value estimated by the learning model is adjusted. As with the above, the adjustment parameters for this adjustment may be decreased over time, may be decreased as the difference from the actual measurement value increases, or may be determined based on multiple actual measurement values.

The biochemical test value estimation unit may set the user's biochemical test value obtained by actual measurement as the actual measurement value, set the biochemical test value estimated by referring to the table as the reference estimate value, and adjust the estimated biochemical test value by referring to the table based on the actual measurement value and the reference estimate value.

With this configuration, the estimated biochemical test value is also adjusted based on the actual measurement value. As with the above, the adjustment parameter for this adjustment may be decreased over time, may be decreased as the difference from the actual measurement value increases, or may be determined based on multiple actual measurement values.

To achieve the above object, the health information providing system of the present invention includes a bioimpedance measuring unit that measures bioimpedance as the user's bioinformation, a biochemical test value estimating unit that estimates biochemical test values based on the bioimpedance, and an output unit that outputs biochemical test values or health information based on the biochemical test values.

According to this configuration, biochemical test values can be estimated by bioimpedance, and biochemical test values can be easily estimated. In other words, a body composition meter or a weight scale with a body composition measurement function can be used as a device capable of acquiring bioimpedance, and the user can obtain an estimate of a biochemical test value by a procedure similar to that for measuring body composition.

The biochemical test value estimation unit may estimate the biochemical test value using a learning model that estimates the biochemical test value using bioimpedance as input.

With this configuration, biochemical test values are estimated from bioimpedance using a learning model.

To achieve the above object, the health information provision program of the present invention causes a computer to acquire body composition information of a user, estimate biochemical test values based on the body composition information, and output biochemical test values or health information based on the biochemical test values.

With this configuration, the biochemical test value estimation unit estimates biochemical test values based on body composition information, making it possible to easily obtain biochemical test values without conducting relatively expensive tests such as blood tests.

To achieve the above object, the health information provision program of the present invention causes a computer to control a bioimpedance measurement unit that measures bioimpedance to acquire bioimpedance as the user's bioinformation, estimate biochemical test values based on the bioimpedance, and output biochemical test values or health information based on the biochemical test values.

With this configuration, biochemical test values can be estimated using bioimpedance, making it easy to estimate biochemical test values. For example, when a system or program is configured to estimate biochemical test values using body composition information calculated from bioimpedance, it becomes possible to estimate biochemical test values by taking advantage of established methods for calculating body composition information. Even when a system or program is configured to estimate biochemical test values directly from bioimpedance, a bioimpedance acquisition device that is common to a body composition scale or a weighing scale with a body composition measurement function can be used.

1 is a diagram showing a health information providing system according to a first embodiment of the present invention. 1 is a diagram showing a usage mode of a measurement device according to a first embodiment of the present invention; 1 is a block diagram showing a hardware configuration of a health information providing system according to a first embodiment of the present invention. 1 is a block diagram showing a functional configuration of a health information providing system according to a first embodiment of the present invention. 1 is a graph showing the correlation between HDL cholesterol (vertical axis), which is a biochemical test value, and fat mass (kg) (horizontal axis), which is body composition information, in the first embodiment of the present invention. 1 is a graph showing the correlation between HDL cholesterol (vertical axis) and visceral fat mass (cm ² ) (horizontal axis), which is body composition information, according to the first embodiment of the present invention. 4 is a graph showing the accuracy of estimation using multiple types of body composition information according to the first embodiment of the present invention. FIG. 2 is an image diagram of an internal risk assessment according to the first embodiment of the present invention. 1 is a graph showing a schematic diagram of adjustment of a learning model according to a first embodiment of the present invention; 13 is a graph illustrating an example of adjusting a learning model based on the elapsed time since actual measurement in the first embodiment of the present invention. 11 is a graph illustrating an example of adjusting a learning model based on a difference between an actual measurement value and an estimated value in the first embodiment of the present invention. 11 is a graph illustrating an example of a learning model in which the degree of adjustment is reduced based on the difference between an actual measured value and an estimated value in the first embodiment of the present invention. 11 is a graph showing an example of adjusting a learning model based on multiple actual measurement values obtained by actual measurements at different times in a first embodiment of the present invention. 1 is a graph showing an example of determining a reference estimated value by statistically processing a plurality of biochemical test values obtained by multiple estimations according to the first embodiment of the present invention. 13 is a graph showing an example of adjusting biochemical test values estimated by a learning model based on actual measured values and estimated values in the second embodiment of the present invention. FIG. 13 is a block diagram showing a functional configuration of a health information providing system according to a third embodiment of the present invention. FIG. 13 is a block diagram showing a functional configuration of a health information providing system according to a fourth embodiment of the present invention.

The following describes an embodiment of the present invention with reference to the drawings. Note that the embodiment described below is merely an example of how the present invention can be implemented, and the present invention is not limited to the specific configuration described below. When implementing the present invention, a specific configuration according to the embodiment may be appropriately adopted.

(First embodiment)
Fig. 1 is a diagram showing a health information providing system 50 according to a first embodiment of the present invention. Fig. 2 is a diagram showing a usage state of a measuring device 10 according to the first embodiment of the present invention. In this embodiment, the health information providing system 50 comprises the measuring device 10 and an information processing terminal (hereinafter referred to as a "user terminal") 20. The measuring device 10 is a body composition analyzer capable of measuring body composition such as body fat by measuring a weak electric current flowing in the body by applying a measurement principle called bioelectrical impedance analysis (BIA).

The measuring device 10 comprises a main body 11 and a handle unit 12. The main body 11 and the handle unit 12 are electrically connected by a connection cord 13. The handle unit 12 can be stored in a storage section 14 provided in the main body 11. When the handle unit 12 is stored in the storage section 14, the connection cord 13 is wound up by a winding mechanism (not shown) inside the main body 11 and stored inside the main body 11.

The main body 11 has a current-carrying electrode 111R and a measurement electrode 112R on the right side of the top surface, and a current-carrying electrode 111L and a measurement electrode 112L on the right side of the top surface.

The handle unit 12 has a roughly rod-like shape and includes a handle body 15 in the center, with grips 16R and 16L provided on either side of the handle body 15. The handle body 15 is provided with a display panel 17 and operation buttons 18A to 18D. Additionally, the grip 16R includes a current-carrying electrode 161R and a measurement electrode 162R, and the grip 16L includes a current-carrying electrode 161L and a measurement electrode 162L.

In the usage mode shown in FIG. 2, the user can measure body composition by standing upright on the main body 11 barefoot and gripping the handle unit 12 with both hands with both arms outstretched. At this time, the base of the right toes contacts the current-carrying electrode 111R, the heel of the right foot contacts the measurement electrode 112R, the base of the left toes contacts the current-carrying electrode 111L, the heel of the left foot contacts the measurement electrode 112L, the fingers of the right hand contacts the current-carrying electrode 161R, the palm of the right hand contacts the measurement electrode 162R, the fingers of the left hand contacts the current-carrying electrode 161L, and the palm of the left hand contacts the measurement electrode 162L.

The main body 11 also has a load cell for measuring body weight inside. As shown in FIG. 2, the weight of a user standing on the main body 11 can be measured. The load cell is composed of a metal member that deforms in response to a load, and a strain gauge attached to the strain gauge. When a user stands on the measuring device 10, the load cell's strain body bends due to the user's load, causing the strain gauge to expand and contract. The resistance value (output value) of the strain gauge changes in response to the expansion and contraction. The measuring device 10 calculates the body weight from the difference between the output value (zero point) of the load cell when no load is applied and the output value when a load is applied. The configuration for measuring body weight using a load cell may be the same as that of a general weighing scale.

The user terminal 20 is a portable terminal equipped with a computer capable of executing application programs, internal storage such as a flash memory, a touch panel, various connectors, and the like. The user terminal 20 also has a wireless communication device for connecting to the Internet, and a short-range communication device for connecting to other nearby devices. The measurement device 10 has a short-range communication device for connecting to other nearby devices. The measurement device 10 and the user terminal 20 can send and receive various information via short-range wireless communication (e.g., Bluetooth (registered trademark)) by pairing with each other.

Figure 3 is a block diagram showing the hardware configuration of a health information providing system 50 according to a first embodiment of the present invention. The health information providing system 50 includes an input unit 501, a weight measuring unit 502, a bioelectrical impedance measuring unit 503, a memory unit 504, a control unit 505, and an output unit 506. Of these components, the weight measuring unit 502 and the bioelectrical impedance measuring unit 503 are provided in the measuring device 10, but as described above, since the measuring device 10 and the user terminal 20 can communicate with each other, the input unit 501, the memory unit 504, the control unit 505, and the output unit 506 may be provided in either the measuring device 10 or the user terminal 20, or in both.

The input unit 501 accepts user operation input. In this embodiment, in particular, height, age, and gender are input to the input unit 501 for each user. The operation buttons 18A to 18D of the measuring device 20 and the touch panel of the user terminal 20 can all serve as the input unit 501. The weight measuring unit 502 corresponds to the load cell of the measuring device 10.

The bioimpedance measuring unit 503 also includes the electrodes 111R, 111L, 112R, 112L, 161R, 161L, 162R, and 162L provided in the measuring device 10, the current-carrying electrodes 161R, 161L, 111R, and 111L, and a current control circuit that passes a weak constant alternating current through these electrodes.

The memory unit 504 stores, for each user, information input from the input unit 501, weight measured by the weight measurement unit 502, bioimpedance measured by the bioimpedance measurement unit 503, body composition calculated by the control unit 505, etc. The memory unit 504 also stores a measurement program for measuring weight and body composition and a health information provision program of this embodiment, as well as data generated by these programs and various information used in these programs (for example, a learning model, described below).

The control unit 505 controls each unit of the health information provision system 50 according to the measurement program, and calculates body composition and obtains health information based on the information input to the input unit 501, the weight measured by the weight measurement unit 502, and the bioimpedance measured by the bioimpedance measurement unit 503 according to the health information provision program. The output unit 506 corresponds to the display panel 17 of the measurement device 20 or the touch panel of the user terminal 20. The output unit 506 displays, under the control of the control unit 505, a screen for inputting information to the input unit 501, a screen for controlling the control unit 505, a screen showing the results of calculations by the control unit 505, and the like.

Figure 4 is a block diagram showing the functional configuration of a health information provision system 50 according to a first embodiment of the present invention. In the health information provision system 50, various functions are realized by the control unit 505 executing various programs. Figure 4 particularly shows the functions realized by executing the health information provision program of this embodiment. The health information provision system 50 includes a height/age/gender acquisition unit 51, a weight acquisition unit 52, a bioelectrical impedance acquisition unit 53, a body composition acquisition unit 54, a biochemical test value estimation unit 55, a health risk assessment unit 56, a health advice determination unit 57, and an output unit 58.

The height/age/gender acquisition unit 51 acquires the user's biometric information, such as height, age, and gender, by accepting user operation input via the input unit 501. The weight acquisition unit 52 acquires the user's biometric information, such as weight, by measuring the user's weight via the weight measurement unit 502. The bioimpedance acquisition unit 53 measures the bioimpedance of the entire body and each body part, which is the user's biometric information.

The bioelectrical impedance acquiring unit 53 measures the bioelectrical impedance, for example, in the following manner.
(1) To measure the bioimpedance of the entire body, a current is supplied using current-carrying electrode 161L and current-carrying electrode 111L, and the potential difference between measurement electrode 162L in contact with the left hand and measurement electrode 112L in contact with the left foot is measured along a current path that flows through the left hand, left arm, chest, abdomen, left leg, and left foot.
(2) To measure the bioimpedance of the right leg, current is supplied using current-carrying electrode 161R and current-carrying electrode 111R, and the potential difference between measurement electrode 112L in contact with the left foot and measurement electrode 112R in contact with the right foot is measured in a current path that flows through the right hand, right arm, chest, abdomen, right leg, and right foot.
(3) To measure the bioimpedance of the left leg, current is supplied using current-carrying electrode 161L and current-carrying electrode 111L, and the potential difference between measurement electrode 112L in contact with the left foot and measurement electrode 112R in contact with the right foot is measured along a current path that flows through the left hand, left arm, chest, abdomen, left leg, and left foot.
(4) To measure the bioimpedance of the right arm, current is supplied using current-carrying electrode 161R and current-carrying electrode 111R, and the potential difference between measuring electrode 162L in contact with the left hand and measuring electrode 162R in contact with the right hand is measured in a current path that flows through the right hand, right arm, chest, abdomen, right leg, and right foot.
(5) To measure the bioimpedance of the left arm, current is supplied using current-carrying electrode 161L and current-carrying electrode 111L, and the potential difference between measuring electrode 162L in contact with the left hand and measuring electrode 162R in contact with the right hand is measured in a current path that flows through the left hand, left arm, chest, abdomen, left leg, and left foot.

In this way, the bioimpedance acquisition unit 53 passes a constant AC current from each current-carrying electrode to a specific part of the user's body and measures the potential difference generated in this current path. Then, based on each value of the current and potential difference, the bioimpedance of the user's whole body or each body part is calculated. The configuration for measuring the bioimpedance may be the same as that of a general body composition monitor. Here, the bioimpedance of the whole body and each body part is obtained when a constant AC current of a reference frequency (e.g., 50 kHz) is passed, when a constant AC current of a high frequency (e.g., 250 kHz) is passed, and when a constant AC current of a low frequency (e.g., 5 kHz) is passed.

The height/age/gender acquisition unit 51 acquires biometric information through operational input, while the weight acquisition unit 52 and bioimpedance acquisition unit 53 acquire biometric information through measurement. However, since all of them acquire the user's biometric information, they are collectively referred to as the biometric information acquisition unit.

The body composition acquisition unit 54 acquires the user's body composition information by performing calculations using bioinformation including the height, age, and sex acquired by the height/age/gender acquisition unit 51, the weight acquired by the weight acquisition unit 52, and the bioimpedance acquired by the bioimpedance acquisition unit 53. The body composition acquisition unit 54 acquires body composition information such as fat percentage, fat mass, fat-free mass, muscle mass, visceral fat mass, visceral fat level, visceral fat area, subcutaneous fat mass, basal metabolic rate, bone mass, total body water percentage, BMI (Body Mass Index), intracellular fluid volume, and extracellular fluid volume by applying a predetermined regression equation to the bioinformation for calculations. The configuration for calculating the body composition information can be the same as that of a general body composition scale.

The biochemical test value estimation unit 55 estimates a biochemical test value based on bioinformation including height, age, and sex acquired by the height/age/sex acquisition unit 51, weight acquired by the weight acquisition unit 52, and bioimpedance acquired by the bioimpedance acquisition unit 53, and on body composition information calculated by the body composition acquisition unit 54. Here, the biochemical test value refers to any of the test items of the biochemical test in the health checkup.

For example, according to the 2018 Basic Test Items for One-Day Medical Checkups from the Japanese Society of Health Checkups, biochemical tests include total protein, albumin, creatinine, eGFR, uric acid, total cholesterol, HDL cholesterol, LDL cholesterol, non-HDL cholesterol, triglycerides, total bilirubin, AST (GOT), ALT (GPT), γ-GT (γ-GTP), ALP, blood glucose (fasting), and HbA1c.

The biochemical tests estimated by the biochemical test value estimation unit 55 are not limited to these, and may include, for example, tests for triglyceride (TG), amylase, CRP (C-Reactive Protein), RF (Rheumatoid Factor), red blood cell count, white blood cell count, hemoglobin amount, hematocrit value, MCV (Mean Corpuscular Volume), MCH (Mean Corpuscular Hemoglobin), platelet count, etc.

Before explaining the operation of the biochemical test value estimation unit 55, we will explain the relationship between the biometric information and body composition information and the biochemical test values.

Figure 5 is a graph showing the correlation between HDL cholesterol (vertical axis), which is a biochemical test value, and fat mass (kg) (horizontal axis), which is body composition information, according to the first embodiment of the present invention. As can be seen from the graph in Figure 5, there appears to be a linear correlation between HDL cholesterol and fat mass, in which the greater the fat mass, the lower the HDL cholesterol, but the variability is relatively large, making it difficult to estimate HDL cholesterol from fat mass with high accuracy.

Fig. 6 is a graph showing the correlation between HDL cholesterol (vertical axis) and visceral fat mass ( ^cm2 ) (horizontal axis) which is body composition information according to the first embodiment of the present invention. As can be seen from the graph in Fig. 6, there appears to be a correlation between HDL cholesterol and the square of visceral fat mass, but even in this case the correlation is weak and there is too much variability to estimate HDL cholesterol from visceral fat mass.

Figure 7 is a graph showing the accuracy of estimation using multiple types of body composition information according to the first embodiment of the present invention. The vertical axis of the graph in Figure 7 is the measured HDL cholesterol, and the horizontal axis is the estimated HDL cholesterol estimated by multiple regression analysis using multiple types of body composition information. By performing an analysis method such as multiple regression analysis using multiple types of body composition information, HDL cholesterol can be estimated with high accuracy, as shown in Figure 7.

The biochemical test value estimation unit 55 estimates biochemical test values from the biochemical information and body composition information using a multiple regression equation obtained by performing multiple regression analysis on a large number of pairs of biochemical test values and body composition information. This multiple regression equation uses the biochemical test values as explanatory variables and the biochemical test values as objective variables. This multiple regression equation can also be said to be a learning model obtained by learning using a large number of pairs of biochemical test values and body composition information as teacher data. The learning model is not limited to the multiple regression equation, and may be, for example, a learning model generated by learning using a decision tree or a neural network. Age and gender may not be explanatory variables, and different multiple regression equations may be prepared and used for each age and gender.

For example, when fat mass (kg) is _x1 and visceral fat mass ( ^cm2 ) is _x2 , the biochemical test value estimation unit 55 estimates HDL cholesterol Y by multiple regression analysis using the following multiple regression equation (1): Note that, in order to improve the estimation accuracy of the biochemical test value, for example, the amount and/or direction of change in _x1 and _x2 may be taken into consideration and HDL cholesterol Y may be estimated using a multiple regression equation that also includes these as explanatory variables.
Y=ax ₁ /x ₂ ² +b (1)

The biochemical test value estimation unit 55 stores multiple multiple regression equations such as those described above for estimating multiple types of biochemical test values. The biological information and body composition information used when estimating each biochemical test value are generally different. The biochemical test value estimation unit 55 estimates each biochemical test value by substituting the biological information and body composition information required for the regression equation for estimating each biochemical test value.

The health risk assessment unit 56 assesses health risks based on the biochemical test values estimated by the biochemical test value estimation unit 55. For this purpose, the health risk assessment unit 56 stores a table that specifies the relationship between the range of biochemical test values and health risks. The health risk assessment unit 56 refers to the table to extract health risk assessments corresponding to the biochemical test values estimated by the biochemical test value estimation unit 55. The health risk assessments are combinations of possible diseases or syndromes (e.g., blood pressure) and the degree of possibility (risk) thereof (e.g., I: good, II: no change, III: bad), and the health risk assessment unit 56 assesses health risks, such as "Predicted blood pressure change I (↓ Good!)."

The health advice determination unit 57 determines advice regarding maintaining or recovering health based on the biochemical test values estimated by the biochemical test value estimation unit 55. For this reason, the health advice determination unit 57 stores a table that specifies the relationship between the range of biochemical test values and health advice.

The health advice determination unit 57 refers to the table to extract health advice corresponding to the biochemical test value estimated by the biochemical test value estimation unit 55. The health advice includes a comment about a possible disease or syndrome (e.g., metabolic syndrome), such as "<Comment on evaluation of internal changes> The risk of metabolic syndrome estimated from changes in body composition is on a downward trend, and it is highly likely that things are changing for the better. Keep it up!".

The output unit 58 displays the biochemical test values estimated by the biochemical test value estimation unit 56, as well as the health risk assessment obtained by the health risk assessment unit 56 and the health advice obtained by the health advice determination unit 57 as health information. The biochemical test values, health risk assessment, and health advice may be displayed in a switched manner, or only one of them may be displayed.

The health risk assessment and health information will be specifically explained using FIG. 8. FIG. 8 is an image diagram of the internal risk assessment according to the first embodiment of the present invention. In the example of FIG. 8, the health risk assessment 601 obtained by the health risk assessment unit 56 is displayed as "☆ Predicted blood pressure change...I (↓ Good!) Health check data available (HBP120 LBP80)", "☆ Predicted blood lipid change...I (↓ Good!) Health check data available (TC LDL HDL TG)", "☆ Predicted liver function change...II (- no change) Health check data available (GOT GPT γGTP)", and "☆ Predicted blood glucose (sugar metabolism) risk I (↓ Good) Health check data available (FBS HbA1c)".

In the example of FIG. 8, the health information 602 obtained by the health advice determination unit 57 displays "◎ The need for prevention of metabolic syndrome is low (maintain as is)" and "<Comment on evaluation of internal changes> The risk of metabolic syndrome estimated from changes in body composition is on a downward trend, and it is highly likely that changes are occurring in a positive direction. Keep it up!".

As described above, in the health information providing system 50 of this embodiment, the biochemical test value estimation unit 55 estimates biochemical test values based on body composition information, so that biochemical test values can be easily provided without performing relatively expensive tests such as blood tests.

The biometric information acquisition unit can also acquire body composition information by performing calculations using biometric impedance. Since biometric impedance can be easily measured using an existing body composition analyzer, the health information provision system 50 can easily acquire body composition information. Furthermore, the biochemical test value estimation unit 55 can effectively estimate biochemical test values based not only on the user's body composition information, but also on biometric information including the user's height, weight, age, and sex.

Furthermore, according to the health information providing system 50 of this embodiment, it is possible to provide health risks and health advice, and it is possible to provide useful information to users who are unable to understand health risks based on estimated biochemical test values alone, or to users who are unsure of what measures to take based on estimated biochemical test values alone.

The above-mentioned health information providing system 50 can estimate biochemical test values from biological information and body composition information by using a learning model. However, there are cases where biochemical test values (actual measurements) are obtained by actually conducting a health checkup including a blood test, etc.

In this case, highly accurate estimation can be made using the actual measured values. That is, for a person who has a health checkup, for example, once a year, a highly accurate estimation can be made by estimating biochemical test values from biological information and body composition information based on the results of the health checkup for one year until the next health checkup. The following describes how to estimate biochemical test values when actual measured values are available.

The biochemical test value estimation unit 55 adjusts the learning model based on the user's biochemical test value obtained by actual measurement (actual measurement value) and the biochemical test value estimated using the learning model (reference estimate value). This makes it possible to reduce the influence of individual differences even when the estimation by the learning model does not match due to individual differences. For example, if the actual measurement value of a certain biochemical test value is 100, but the biochemical test value estimated by the biochemical test value estimation unit is 90, the learning model can be adjusted so that 90 is used as the reference estimate value and a value 10 (=100-90) greater than the original value is obtained.

By adjusting the learning model in this way, it is possible to estimate biochemical test values based on actual measurements, and also to estimate biochemical test values according to changes in body composition and biological information using the actual measurements as a reference. For this reason, it is desirable for the reference estimated value to be a biochemical test value estimated using the user's biological information and body composition information from the same period as the actual measurement. For example, the reference estimated value may be a biochemical test value estimated using the user's biological information at the date and time closest to the actual measurement. Furthermore, the biological information and body composition information from the same period as the actual measurement may be estimated before the actual measurement, or after the actual measurement.

The actual measured value can be input by the user through the input unit 501, or if the input unit 501 has a communication module or a communication interface, it may be input from an external device through the communication module or communication interface.

Fig. 9 is a graph showing the adjustment of the learning model according to the first embodiment of the present invention. In the example of Fig. 9, an example is shown in which the input data and the estimated value of HDL cholesterol are in a linear relationship, but in reality, since multiple regression analysis is performed in this embodiment as described above, there are multiple explanatory variables. However, in the following, for the sake of convenience, the input data is considered to be one variable, and the input data and HDL cholesterol are in ^a linear correlation. In the case of the above formula (1), x = ( _x1 / _x22 ) can be used as the input data.

As a result of learning a large amount of learning data, the regression equation 81 shown in FIG. 9 is obtained. Furthermore, as a result of undergoing a health check, it is assumed that the result is that the HDL cholesterol is Yr. Assume that the HDL cholesterol estimated using the health information providing system 50 of this embodiment at the same time as the health check is Ye. In this case, the HDL cholesterol corresponding to the input data x obtained at the same time as the health check should be Yr. Therefore, the biochemical test value estimation unit 55 adjusts the learning model so that HDL cholesterol Yr is calculated for the input data x. In this embodiment, specifically, an adjustment is made to shift the learning model (regression equation) so that the output Ye becomes Yr without changing the profile of the learning model, and the adjusted learning model 82 is set.

If the body composition has changed between the time the input data x was obtained and the time of the health check, the above learning model cannot be adjusted correctly. For this reason, it is desirable to use the biochemical test value estimated using the input data x obtained as close as possible to the time of the health check as the reference estimate. On the other hand, since biochemical test values do not change significantly over a period of a few days, the two points in time do not necessarily have to be on the same day. For example, it is desirable to estimate the reference estimate and measure it in the health check with a time difference of about one week.

In the example of FIG. 9, the original learning model is shifted to adjust the learning model, which means that a fixed value based on the difference between the actual measurement value and the reference estimated value is added or subtracted from the HDL cholesterol value estimated by the original learning model. In other words, the biochemical test value estimation unit 55 may increase the accuracy of the estimation by adding or subtracting a fixed value (which may be the difference itself) based on the difference between the actual measurement value and the reference estimated value to the estimated value estimated by the original learning model.

Figure 10 is a graph illustrating an example of adjusting the learning model based on the elapsed time since the actual measurement according to the first embodiment of the present invention. In this example, the biochemical test value estimation unit 55 gradually reduces the degree of adjustment of the adjusted learning model as time passes since the actual measurement. In the example of Figure 10, after a predetermined period of time has passed since the actual measurement, the degree of adjustment of the learning model is reduced, and the adjusted learning model 82 is brought slightly closer to the original learning model 81 to become the adjusted learning model 83.

Figure 11 is a graph illustrating an example of adjusting a learning model based on the difference between an actual measurement value and an estimated value according to the first embodiment of the present invention. The estimated value here does not refer to a biochemical test value that is estimated at the same time as the actual measurement and used as a reference estimated value, but rather refers to an estimated value when an adjusted learning model is set based on such a reference estimated value and then the adjusted learning model is used to estimate a biochemical test value. When the estimated value is near the actual measurement value, a learning model that passes through the actual measurement value can be used to make a highly accurate estimation based on the actual measurement value, but when the deviation between the actual measurement value and the estimated value becomes large, the influence of the actual measurement value may be reduced, and conversely, it is considered that the reliability of the original learning model trained using a large amount of learning data will be increased.

Therefore, as shown in FIG. 11, the biochemical test value estimation unit 55 gradually reduces the degree of adjustment of the learning models in the order of learning model 82, learning model 84, learning model 85, and learning model 86 so that the further away from the actual measurement value the closer the learning model is to the original learning data, and finally adopts the original learning model.

Fig. 12 is a graph illustrating an example of a learning model in which the degree of adjustment decreases based on the difference between the actual measured value and the estimated value according to the first embodiment of the present invention. Instead of the example of Fig. 11, the biochemical test value estimation unit 55 may adjust the learning model at the time the reference estimated value is obtained, as shown in Fig. 12, so that the degree of adjustment from the original learning model is greatest for the actual measured value, and the degree of adjustment decreases as the value deviates from the actual measured value, thereby setting an adjusted learning model 87.

Figure 13 is a graph showing an example of adjusting the learning model based on multiple actual measurement values obtained by actual measurements at multiple different times according to the first embodiment of the present invention. When there are multiple actual measurement values, the biochemical test value estimation unit 55 uses them to adjust the learning model. For example, when the results (actual measurement values) of a yearly health check have been accumulated over several years, the learning model can be adjusted using those multiple actual measurement values. In the example of Figure 13, when a difference 91 between the latest actual measurement value and the latest reference estimated value, a difference 92 between the actual measurement value one year ago and the reference estimated value one year ago, and a difference 93 between the actual measurement value two years ago and the reference estimated value two years ago are obtained, the average value of these differences 91 to 93 is calculated, and the original learning model 81 is adjusted using this average value, and the adjusted learning model 88 is set. At this time, the biochemical test value estimation unit 55 may weight the newer the actual measurement value, and then take the average (weighted average).

Figure 14 is a graph showing an example of determining a reference estimated value by statistically processing multiple biochemical test values obtained by multiple estimations according to the first embodiment of the present invention. When there are multiple biochemical test values, the biochemical test value estimation unit 55 statistically processes them to adjust the reference estimated value. The statistical processing may be a simple average process, or an average process after excluding outliers. In the example of Figure 14, of the biochemical test values obtained by eight estimations over one week (1W), an outlier 94 is excluded and then averaged to obtain a 1W average value 95, which is determined as the reference estimated value.

Figure 15 is a diagram showing an example of adjusting a biochemical test value estimated by a learning model based on an actual measurement value and an estimated value according to the first embodiment of the present invention. In the example of Figure 9 above, the biochemical test value estimation unit 55 adjusted the learning model 81 based on the actual measurement value, but in the example of Figure 15, the biochemical test value 96 estimated by the learning model 81 is adjusted based on the actual measurement value. The adjustment parameters for this adjustment may be decreased over time, or may be decreased as the difference from the actual measurement value increases, as in the adjustments of Figures 10 to 13 above, or may be determined based on multiple actual measurement values.

As described above, the health information providing system 50 can obtain biochemical test values corresponding to the user's body composition information by inputting the body composition information into the learning model. This learning model can be adjusted based on actual measurements. This makes it possible to reduce the influence of individual differences even when the learning model's estimates do not match due to individual differences.

Adjustment of the learning model based on actual measurements can be made based on the time elapsed since the actual measurements were made, the difference between the actual measurements and the estimated biochemical test values, and multiple actual measurements obtained by actual measurements at multiple different times. This allows the learning model to be adjusted so that it approaches the original learning model over time, or the learning model to approach the original learning model as the difference between the actual measurements and the estimated biochemical test values increases, or, if there are multiple actual measurements, the learning model can be adjusted using these.

Furthermore, the health information providing system 50 can obtain an appropriate reference estimated value by statistically processing multiple biochemical test values obtained by multiple estimations to determine a reference estimated value. In other words, although an accurate estimation may not be possible with a single estimation of a biochemical test value, it is possible to obtain a biochemical test value that accurately reflects biological information and body composition information by estimating the biochemical test value multiple times. Furthermore, the health information providing system 50 can also adjust the biochemical test value estimated by the learning model, rather than adjusting the learning model based on the actual measured value.

Second Embodiment
The basic configuration of the health information providing system 50 according to the second embodiment is the same as that of the health information providing system 50 according to the first embodiment. However, the health information providing system 50 according to the second embodiment differs from the health information providing system 50 according to the first embodiment in that the biochemical test value estimating section 55 estimates the biochemical test values by referring to a table that defines the relationship between body composition information and biochemical test values. Only this difference will be described below.

The biochemical test value estimation unit 55 has a table that specifies the relationship between body composition information and biochemical test values. The biochemical test value estimation unit 55 refers to the table and estimates the biochemical test values from the biological information and body composition information.

The biochemical test value estimation unit 55 may, for example, set the user's biochemical test value obtained by actual measurement as the actual measurement value, set the biochemical test value estimated by referring to the table as the reference estimated value, and adjust the estimated biochemical test value by referring to the table based on the actual measurement value and the reference estimated value. The adjustment parameters for this adjustment may also be decreased over time, or may be decreased as the difference from the actual measurement value increases, as in the adjustments in Figures 10 to 13 above, or may be determined based on multiple actual measurement values.

As described above, the health information providing system 50 according to the second embodiment can estimate biochemical test values by referring to the table provided in the biochemical test value estimating unit 55, and can therefore easily and quickly estimate biochemical test values for body composition information without complex calculations. In addition, the estimated biochemical test values can be adjusted based on actual measurements.

Third Embodiment
The basic configuration of the health information providing system 50 according to the third embodiment is the same as that of the health information providing system 50 according to the first embodiment. However, the health information providing system 50 according to the third embodiment differs from the health information providing system 50 according to the first embodiment in that the health information providing system 50 does not include a body composition acquiring unit 54 and estimates biochemical test values based on biological information including height, age, and sex acquired by a height/age/sex acquiring unit 51, weight acquired by a weight acquiring unit 52, and bioimpedance acquired by a bioimpedance acquiring unit 53. Only this difference will be described below.

Figure 16 is a block diagram showing the functional configuration of a health information provision system according to a third embodiment of the present invention. The health information provision system 50 includes a height/age/gender acquisition unit 51, a weight acquisition unit 52, a bioelectrical impedance acquisition unit 53, a biochemical test value estimation unit 55, a health risk assessment unit 56, and an output unit 58.

The biochemical test value estimation unit 55 estimates biochemical test values based on bioinformation including the height, age, and sex acquired by the height/age/gender acquisition unit 51, the weight acquired by the weight acquisition unit 52, and the bioimpedance acquired by the bioimpedance acquisition unit 53, rather than based on the body composition information calculated by the body composition acquisition unit 54 as in the health information provision system 50 of the first embodiment. In other words, the biochemical test value estimation unit 55 estimates biochemical test values using bioimpedance without using body composition information.

Similarly, the biochemical test value estimation unit 55 may estimate biochemical test values using a learning model that estimates biochemical test values using bioimpedance as input, without basing the estimation on the body composition information calculated by the body composition acquisition unit 54 as in the health information provision system 50 of the first embodiment.

As described above, the health information providing system 50 according to the third embodiment can estimate biochemical test values using bioimpedance, and can easily estimate biochemical test values. That is, a body composition scale or a weighing scale with a body composition measuring function can be used as a device capable of acquiring bioimpedance, and the user can obtain an estimate of a biochemical test value using a procedure similar to that for measuring body composition. Furthermore, the health information providing system 50 according to the second embodiment can estimate a biochemical test value from bioimpedance using a learning model.

(Fourth embodiment)
The basic configuration of the health information providing system 50 according to the fourth embodiment is the same as that of the health information providing system 50 according to the first embodiment. However, the health information providing system 50 according to the fourth embodiment differs from the health information providing system 50 according to the first embodiment in that the health information providing system 50 does not include a bioelectrical impedance acquiring unit 53 and estimates biochemical test values based on height, age, and sex acquired by a height/age/sex acquiring unit 51, weight acquired by a weight acquiring unit 52, and body composition information acquired by a body composition acquiring unit 54. Only this difference will be described below.

Figure 17 is a block diagram showing the functional configuration of a health information providing system according to a fourth embodiment of the present invention. The health information providing system 50 includes a height/age/gender acquiring unit 51, a weight acquiring unit 52, a biochemical test value estimating unit 55, a health risk assessing unit 56, an output unit 58, and a body composition acquiring unit 54.

The body composition acquisition unit 54 acquires body composition information by accepting input from outside, without using the bioimpedance acquired by the bioimpedance acquisition unit 53 as in the health information provision system 50 of the first embodiment.

The acquired body composition information is used to estimate biochemical test values, similarly to the health information providing system 50 according to the first embodiment. That is, the biochemical test value estimation unit 55 estimates biochemical test values based on the bioinformation including height, age, and sex acquired by the height/age/sex acquisition unit 51 and weight acquired by the weight acquisition unit 52, and the body composition information acquired by the body composition acquisition unit 54.

As described above, the health information providing system 50 according to the fourth embodiment can estimate biochemical test values based on body composition information even if it does not have the function of measuring bioimpedance.

10: Measurement device 11: Main body 12: Handle unit 13: Connection cord 14: Storage section 15: Handle body 16L, R: Grip 17: Display panel 18A-D: Operation button 20: User terminal 50: Health information providing system 51: Height, age, and gender acquisition section 52: Weight acquisition section 53: Bioimpedance acquisition section 54: Body composition acquisition section 55: Biochemical test value estimation section 56: Health risk assessment section 57: Health advice determination section 58: Output section 111L, R: Current-carrying electrodes 112L, R: Measurement electrodes 161L, R: Current-carrying electrodes 162L, R: Measurement electrodes 501: Input section 502: Weight measurement section 503: Bioimpedance measurement section 504: Memory section 505: Control section 506: Output section

Claims (21)

a body composition acquisition unit that acquires body composition information of a user;
a biochemical test value estimation unit that estimates a predetermined biochemical test value that is correlated with a predetermined body composition information based on the predetermined body composition information among the body composition information acquired by the body composition acquisition unit ;
an output unit that outputs the biochemical test values or health information based on the biochemical test values;
Equipped with
A health information provision system in which the biochemical test value estimation unit estimates the specified biochemical test value using a learning model that estimates the specified biochemical test value using the specified body composition information as input, and sets the specified biochemical test value of the user obtained by actual measurement as an actual measurement value, sets the specified biochemical test value estimated using the learning model as a reference estimate value, and adjusts the learning model based on the actual measurement value and the reference estimate value. a body composition acquisition unit that acquires body composition information of a user;
a biochemical test value estimation unit that estimates a predetermined biochemical test value that is correlated with a predetermined body composition information based on the predetermined body composition information among the body composition information acquired by the body composition acquisition unit ;
an output unit that outputs the biochemical test values or health information based on the biochemical test values;
Equipped with
A health information provision system in which the biochemical test value estimation unit estimates the specified biochemical test value by referring to a table that specifies the relationship between the specified body composition information and the specified biochemical test value, sets the specified biochemical test value of the user obtained by actual measurement as an actual measurement value, sets the specified biochemical test value estimated by referring to the table as a reference estimated value, and adjusts the specified biochemical test value estimated by referring to the table based on the actual measurement value and the reference estimated value. A biometric information acquisition unit for acquiring biometric information of the user,
3 . The health information providing system according to claim 1 , wherein the biochemical test value estimating section estimates the predetermined biochemical test value based on predetermined biological information among the biological information acquired by the biological information acquiring section . The biometric information acquisition unit acquires at least a biometric impedance of the user as the biometric information,
The health information providing system according to claim 3 , wherein the body composition acquiring unit acquires the body composition information by a calculation using the bioelectrical impedance. The health information provision system according to claim 4, wherein the biometric information acquisition unit acquires the biometric impedance by measurement. The health information providing system according to any one of claims 1 to 3, wherein the body composition acquisition unit acquires the body composition information by accepting an input of the body composition information. The health information providing system according to any one of claims 3 to 5, wherein the biometric information acquisition unit acquires at least the user's height, weight, age, and sex as the biometric information. A health risk assessment unit that assesses health risks based on the biochemical test values,
The health information providing system according to claim 1 , wherein the output unit outputs the health risk as the health information. A health advice determination unit determines advice regarding health maintenance or health recovery based on the biochemical test values,
The health information providing system according to claim 1 , wherein the output unit outputs the advice as the health information. The health information provision system of claim 1, wherein the biochemical test value estimation unit further adjusts the learning model based on the elapsed time since the actual measurement was performed. 11. The health information providing system according to claim 1, wherein the biochemical test value estimating section further adjusts the learning model based on a difference between the actual measurement value and the estimated predetermined biochemical test value. The health information provision system according to claim 1, 10, or 11, wherein the biochemical test value estimation unit adjusts the learning model based on a plurality of actual measurement values obtained by actual measurements at a plurality of different times. 13. The health information providing system according to claim 1, wherein the biochemical test value estimating section determines the reference estimated value by performing statistical processing on a plurality of the predetermined biochemical test values obtained by a plurality of estimations. The health information provision system according to claim 10, or any one of claims 11 to 13 that cite claim 10, wherein the biochemical test value estimation unit adjusts the learning model so that the learning model approaches the original learning model as the elapsed time increases. 2. The health information provision system of claim 1, wherein the biochemical test value estimation unit sets the specified biochemical test value of the user obtained by actual measurement as an actual measurement value, sets the specified biochemical test value estimated using the learning model as a reference estimated value, and adjusts the biochemical test value estimated by the learning model based on the actual measurement value and the reference estimated value. A bioimpedance acquisition unit that acquires bioimpedance information of a user;
a biochemical test value estimation unit that estimates a predetermined biochemical test value associated with the bioelectrical impedance information based on the bioelectrical impedance information acquired by the bioelectrical impedance acquisition unit ;
an output unit that outputs the biochemical test values or health information based on the biochemical test values;
Equipped with
The biochemical test value estimation unit estimates the specified biochemical test value using a learning model that estimates the specified biochemical test value using the bioimpedance information as input , and sets the specified biochemical test value of the user obtained by actual measurement as an actual measurement value, sets the specified biochemical test value estimated using the learning model as a reference estimate value, and adjusts the learning model based on the actual measurement value and the reference estimate value. A bioimpedance acquisition unit that acquires bioimpedance information of a user;
a biochemical test value estimation unit that estimates a predetermined biochemical test value associated with the bioelectrical impedance information based on the bioelectrical impedance information acquired by the bioelectrical impedance acquisition unit ;
an output unit that outputs the biochemical test values or health information based on the biochemical test values;
Equipped with
The biochemical test value estimation unit estimates the specified biochemical test value by referring to a table that defines the relationship between the bioimpedance information and the specified biochemical test value, sets the specified biochemical test value of the user obtained by actual measurement as an actual measured value, sets the specified biochemical test value estimated by referring to the table as a reference estimated value, and adjusts the biochemical test value estimated by referring to the table based on the actual measured value and the reference estimated value. On the computer,
Acquire body composition information of the user;
estimating a predetermined biochemical test value that is correlated with the predetermined body composition information based on predetermined body composition information among the acquired body composition information , using a learning model that estimates a biochemical test value using the predetermined body composition information as an input;
outputting the biochemical test values or health information based on the biochemical test values;
the predetermined biochemical test value of the user acquired by actual measurement is defined as an actual measurement value, the predetermined biochemical test value estimated using the learning model is defined as a reference estimated value, and the learning model is adjusted based on the actual measurement value and the reference estimated value.
Health information programs. On the computer,
Acquire body composition information of the user;
estimating the predetermined biochemical test value by referring to a table that defines a relationship between the predetermined body composition information and a predetermined biochemical test value that is correlated with the predetermined body composition information, based on the predetermined body composition information among the acquired body composition information;
outputting the biochemical test values or health information based on the biochemical test values;
the predetermined biochemical test value of the user acquired by actual measurement is set as an actual measurement value, the predetermined biochemical test value estimated with reference to the table is set as a reference estimated value, and the predetermined biochemical test value estimated with reference to the table is adjusted based on the actual measurement value and the reference estimated value.
Health information programs. On the computer,
Controlling a bioimpedance acquisition unit that acquires bioimpedance information to acquire the bioimpedance information as bioinformation of the user;
estimating a predetermined biochemical test value associated with the bioimpedance information based on the acquired bioimpedance information using a learning model that estimates a predetermined biochemical test value using the bioimpedance information as an input;
outputting the biochemical test values or health information based on the biochemical test values;
the predetermined biochemical test value of the user acquired by actual measurement is defined as an actual measurement value, the predetermined biochemical test value estimated using the learning model is defined as a reference estimated value, and the learning model is adjusted based on the actual measurement value and the reference estimated value.
Health information programs. On the computer,
Controlling a bioimpedance acquisition unit that acquires bioimpedance information to acquire the bioimpedance information as bioinformation of the user;
and estimating the predetermined biochemical test value based on the acquired bioelectrical impedance information by referring to a table that defines a relationship between the bioelectrical impedance information and a predetermined biochemical test value associated with the bioelectrical impedance information ;
outputting the biochemical test values or health information based on the biochemical test values;
the predetermined biochemical test value of the user acquired by actual measurement is set as an actual measurement value, the predetermined biochemical test value estimated with reference to the table is set as a reference estimated value, and the predetermined biochemical test value estimated with reference to the table is adjusted based on the actual measurement value and the reference estimated value.
Health information programs.

Images