JP7770522B2 - Electronic device, estimation system, estimation method, and estimation program - Google Patents

Description

The present invention relates to an electronic device, estimation system, estimation method, and estimation program that estimate the health condition of a subject from measured biological information.

Traditionally, blood components and blood fluidity have been measured as a means of estimating the health status of a subject (user). These measurements are performed using blood drawn from the subject. Electronic devices that measure biological information from a test site such as the subject's wrist are also known. For example, Patent Document 1 describes an electronic device that measures the subject's pulse when worn on the subject's wrist.

Japanese Patent Application Laid-Open No. 2002-360530

However, drawing blood is painful, making it difficult to estimate one's own health condition on a daily basis. Furthermore, the electronic device described in Patent Document 1 only measures pulse rate and cannot estimate the subject's health condition other than pulse rate.

In light of these circumstances, the object of the present invention is to provide an electronic device, estimation system, estimation method, and estimation program that can easily estimate the health condition of a subject.

In order to solve the above problem, an electronic device according to one embodiment of the present invention includes a control unit that estimates the blood glucose level of a subject based on an estimation equation created based on the blood glucose level and the subject's pulse wave associated with the blood glucose level, and the subject's age. The control unit uses the estimation equation created based on the results of a regression analysis using the subject's age and an index AI that represents the ratio between the forward wave of the pulse wave and the reflected wave that appears later than the forward wave.

An electronic device according to one embodiment of the present invention includes a control unit that estimates the glucose metabolism of a subject based on an estimation equation created based on the blood glucose level and the subject's pulse wave associated with the blood glucose level, and the subject's age. The control unit uses the estimation equation created based on the results of a regression analysis using the subject's age and an index AI that represents the ratio between the forward wave of the pulse wave and the reflected wave that appears later than the forward wave.

An electronic device according to one embodiment of the present invention includes a control unit that estimates the lipid level of a subject based on an estimation equation created based on the lipid level and the subject's pulse wave associated with the lipid level, and the subject's age. The control unit uses the estimation equation created based on the results of a regression analysis using the subject's age and an index AI that represents the ratio between the forward wave of the pulse wave and the reflected wave that appears later than the forward wave.

An electronic device according to one embodiment of the present invention includes a control unit that estimates the lipid metabolism of a subject based on an estimation equation created based on lipid values and the subject's pulse waves associated with the lipid values, and the subject's age. The control unit uses the estimation equation created based on the results of a regression analysis using the subject's age and an index AI that represents the ratio between the forward wave of the pulse wave and the reflected wave that appears later than the forward wave.

An estimation method according to one embodiment of the present invention includes an estimation step of estimating the blood glucose level or glucose metabolism of a subject based on an estimation formula created based on the blood glucose level and the subject's pulse wave associated with the blood glucose level, and the subject's age, and the estimation step uses an estimation formula created based on the results of a regression analysis using the subject's age and an index AI expressed as the ratio between the forward wave of the pulse wave and the reflected wave that appears later than the forward wave.

An estimation method according to one embodiment of the present invention includes an estimation step of estimating the lipid level or lipid metabolism of a subject based on an estimation formula created based on lipid levels and the subject's pulse waves associated with the lipid levels, and the subject's age, wherein the estimation step uses an estimation formula created based on the results of a regression analysis using the subject's age and an index AI expressed as the ratio between the forward wave of the pulse wave and the reflected wave that appears later than the forward wave.

An estimation program according to one embodiment of the present invention causes a computer to function as a control unit that estimates a subject's blood glucose level or glucose metabolism based on an estimation formula created based on the blood glucose level and the subject's pulse wave associated with the blood glucose level, and the subject's age. The control unit uses the estimation formula created based on the results of a regression analysis using the subject's age and an index AI that represents the ratio between the forward wave of the pulse wave and the reflected wave that appears later than the forward wave.

An estimation program according to one embodiment of the present invention causes a computer to function as a control unit that estimates a subject's lipid level or lipid metabolism based on an estimation equation created based on the subject's lipid level and the subject's pulse wave associated with the lipid level, and the subject's age. The control unit uses the estimation equation created based on the results of a regression analysis using the subject's age and an index AI that represents the ratio between the forward wave of the pulse wave and the reflected wave that appears later than the forward wave.

The present invention provides an electronic device, estimation system, estimation method, and estimation program that can easily estimate the health condition of a subject.

1 is a schematic diagram illustrating a general configuration of an electronic device according to a first embodiment of the present invention. FIG. 2 is a cross-sectional view showing a schematic configuration of a main body portion of FIG. 1 . FIG. 2 is a diagram illustrating an example of a usage state of the electronic device of FIG. 1. FIG. 2 is a functional block diagram showing a schematic configuration of the electronic device of FIG. 1. 3A to 3C are diagrams illustrating an example of an estimation method based on changes in pulse waves in the electronic device of FIG. 1 . FIG. 10 is a diagram showing an example of an acceleration pulse wave. FIG. 4 is a diagram showing an example of a pulse wave acquired by a sensor unit. 1. FIG. 4 is a diagram illustrating another example of an estimation method based on a change in a pulse wave in the electronic device of FIG. FIG. 2 is a flow chart for creating an estimation formula used by the electronic device of FIG. 1 . FIG. 10 is a flow diagram for estimating a subject's postprandial blood glucose level using the estimation equation created by the flow of FIG. 9 . FIG. 10 is a diagram showing a comparison between postprandial blood glucose levels estimated using the estimation formula created by the flow of FIG. 9 and actually measured postprandial blood glucose levels. FIG. 11 is a flowchart illustrating the creation of an estimation formula used by an electronic device according to a second embodiment of the present invention. FIG. 13 is a flow diagram for estimating a subject's postprandial blood glucose level using the estimation equation created by the flow of FIG. 12. FIG. 11 is a flowchart illustrating the creation of an estimation formula used by an electronic device according to a third embodiment of the present invention. FIG. 15 is a flow diagram for estimating a subject's postprandial lipid level using the estimation formula created by the flow of FIG. 14. FIG. 15 is a diagram showing a comparison between postprandial lipid levels estimated using the estimation formula created by the flow of FIG. 14 and actually measured postprandial lipid levels. FIG. 2 is a diagram schematically illustrating communication between an electronic device and a blood glucose meter. 1 is a schematic diagram showing a general configuration of a system according to an embodiment of the present invention;

Embodiments of the present invention will be described in detail below with reference to the drawings.

(First embodiment)
Fig. 1 is a schematic diagram showing the general configuration of an electronic device according to a first embodiment of the present invention. Electronic device 100 includes a mounting unit 110 and a measurement unit 120. Fig. 1 is a view of electronic device 100 observed from a rear surface 120a that contacts a test portion.

Electronic device 100 measures the subject's biometric information while the subject is wearing electronic device 100. The biometric information measured by electronic device 100 is the subject's pulse wave, which can be measured by measurement unit 120. In the present embodiment, the following description will be given assuming that electronic device 100 is worn on the subject's wrist and acquires the pulse wave, as an example.

In this embodiment, the attachment unit 110 is a straight, elongated band. The pulse wave is measured, for example, with the subject wearing the attachment unit 110 of the electronic device 100 around their wrist. Specifically, the subject wraps the attachment unit 110 around their wrist so that the back surface 120a of the measurement unit 120 is in contact with the area to be measured, and then measures the pulse wave. The electronic device 100 measures the pulse wave of blood flowing through the ulnar artery or radial artery at the subject's wrist.

Figure 2 is a cross-sectional view showing the schematic configuration of the measurement unit 120 in Figure 1. In addition to the measurement unit 120, Figure 2 also illustrates the attachment unit 110 around the measurement unit 120.

The measuring unit 120 has a back surface 120a that comes into contact with the subject's wrist when worn, and a front surface 120b opposite the back surface 120a. The measuring unit 120 has an opening 111 on the back surface 120a side. The sensor unit 130 is supported by the measuring unit 120 with one end protruding from the opening 111 toward the back surface 120a when the elastic body 140 is not pressed. One end of the sensor unit 130 is provided with a pulse contact portion 132. One end of the sensor unit 130 is displaceable in a direction approximately perpendicular to the plane of the back surface 120a. The other end of the sensor unit 130 is supported by the measuring unit 120 via a support portion 133 so that one end of the sensor unit 130 can be displaced.

One end of the sensor unit 130 comes into contact with the measuring unit 120 via the elastic body 140 and is displaceable. The elastic body 140 is, for example, a spring. However, the elastic body 140 is not limited to a spring and can be any other elastic body, such as resin or sponge.

Although not shown, the measurement unit 120 may also include a control unit, memory unit, communication unit, power supply unit, notification unit, circuits for operating these units, connecting cables, etc.

The sensor unit 130 includes an angular velocity sensor 131 that detects the displacement of the sensor unit 130. The angular velocity sensor 131 may be capable of detecting the angular displacement of the sensor unit 130. The sensor included in the sensor unit 130 is not limited to the angular velocity sensor 131, and may be, for example, an acceleration sensor, an angle sensor, or another motion sensor, or may include multiple of these sensors.

The electronic device 100 includes an input unit 141 on the surface 120b side of the measurement unit 120. The input unit 141 accepts operational input from the subject and is composed of, for example, operation buttons (operation keys). The input unit 141 may also be composed of, for example, a touch screen.

Figure 3 shows an example of how the electronic device 100 is used by a subject. The subject wears the electronic device 100 by wrapping it around their wrist. The electronic device 100 is worn with the back surface 120a of the measuring unit 120 in contact with the area to be measured. With the wearing unit 110 wrapped around the wrist, the position of the measuring unit 120 can be adjusted so that the pulse contact unit 132 comes into contact with the location where the ulnar artery or radial artery is located.

In Figure 3, when the electronic device 100 is worn, one end of the sensor unit 130 is in contact with the skin above the radial artery, which is an artery on the thumb side of the subject's left hand. The elastic force of the elastic body 140, which is arranged between the measurement unit 120 and the sensor unit 130, causes one end of the sensor unit 130 to be in contact with the skin above the radial artery. The sensor unit 130 displaces in response to the movement of the subject's radial artery, i.e., pulsation. The angular velocity sensor 131 detects the displacement of the sensor unit 130 to obtain a pulse wave. A pulse wave is a waveform captured from the body surface that represents the change in blood volume over time caused by the inflow of blood.

Referring again to FIG. 2 , one end of the sensor unit 130 protrudes from the opening 111 when the elastic body 140 is not pressed. When the electronic device 100 is worn on the subject, one end of the sensor unit 130 contacts the skin above the subject's radial artery. In response to the pulse, the elastic body 140 expands and contracts, displacing the one end of the sensor unit 130. The elastic body 140 has an appropriate elastic modulus so as not to impede the pulse and to expand and contract in response to the pulse. The opening width W of the opening 111 is sufficiently larger than the diameter of the blood vessel, which in this embodiment is the diameter of the radial artery. By providing the opening 111 in the measuring unit 120, the rear surface 120a of the measuring unit 120 does not compress the radial artery when the electronic device 100 is worn. This enables the electronic device 100 to acquire pulse waves with less noise, improving measurement accuracy.

While Figure 3 shows an example in which the electronic device 100 is worn on the wrist and acquires a pulse wave in the radial artery, the present invention is not limited to this. For example, the electronic device 100 may be worn on the subject's neck to acquire a pulse wave of blood flowing through the carotid artery. Specifically, the subject may measure the pulse wave by lightly pressing the pulse contact portion 132 against the position of the carotid artery. Alternatively, the subject may wear the attachment portion 110 by wrapping it around their neck so that the pulse contact portion 132 is positioned at the position of the carotid artery.

Figure 4 is a functional block diagram showing the general configuration of electronic device 100. Electronic device 100 includes sensor unit 130, input unit 141, control unit 143, power supply unit 144, memory unit 145, communication unit 146, and notification unit 147. In this embodiment, control unit 143, power supply unit 144, memory unit 145, communication unit 146, and notification unit 147 are included within measurement unit 120 or attachment unit 110.

The sensor unit 130 includes an angular velocity sensor 131 and detects pulsation from the test area to acquire a pulse wave.

The control unit 143 is a processor that controls and manages the entire electronic device 100, including each functional block of the electronic device 100. The control unit 143 is also a processor that estimates the subject's blood glucose level from the acquired pulse wave. The control unit 143 is composed of a processor such as a CPU (Central Processing Unit) that executes a program that defines the control procedure and a program that estimates the subject's blood glucose level, and such programs are stored in a storage medium such as the storage unit 145. The control unit 143 also estimates the subject's state related to glucose metabolism, lipid metabolism, etc. based on the calculated indices. The control unit 143 also notifies the notification unit 147 of the data.

The power supply unit 144 includes, for example, a lithium-ion battery and a control circuit for charging and discharging the battery, and supplies power to the entire electronic device 100.

The memory unit 145 stores programs and data. The memory unit 145 may include any non-transitory storage medium, such as a semiconductor storage medium or a magnetic storage medium. The memory unit 145 may include multiple types of storage media. The memory unit 145 may include a combination of a portable storage medium, such as a memory card, optical disk, or magneto-optical disk, and a storage medium reader. The memory unit 145 may include a storage device used as a temporary storage area, such as RAM (Random Access Memory). The memory unit 145 stores various information and programs for operating the electronic device 100, and also functions as a work memory. The memory unit 145 may store, for example, pulse wave measurement results acquired by the sensor unit 130.

The communication unit 146 transmits and receives various data by performing wired or wireless communication with an external device. For example, the communication unit 146 communicates with an external device that stores the subject's biometric information to manage the subject's health condition, and transmits the pulse wave measurement results measured by the electronic device 100 and the health condition estimated by the electronic device 100 to the external device.

The notification unit 147 notifies information using sound, vibration, images, etc. The notification unit 147 may include a speaker, a vibrator, and a display device such as a liquid crystal display (LCD), an organic electro-luminescence display (ELD), or an inorganic electro-luminescence display (ELD). In this embodiment, the notification unit 147 notifies, for example, the state of the subject's glucose metabolism or lipid metabolism.

The electronic device 100 estimates the blood glucose level of the subject based on an estimation formula created by regression analysis. The electronic device 100 stores estimation formulas for estimating blood glucose levels based on pulse waves, for example, in advance in the memory unit 145. The electronic device 100 estimates blood glucose levels using these estimation formulas.

Here, we will explain the estimation theory for estimating blood glucose levels based on pulse waves. After a meal, an increase in blood glucose levels in the blood causes a decrease in blood fluidity (increased viscosity), vascular dilation, and an increase in circulating blood volume. Vascular dynamics and hemodynamics are determined so that these states are in equilibrium. A decrease in blood fluidity occurs, for example, due to an increase in plasma viscosity or a decrease in red blood cell deformability. Vascular dilation occurs due to the secretion of insulin and digestive hormones, an increase in body temperature, and other factors. Vascular dilation suppresses a decrease in blood pressure, thereby increasing the pulse rate. An increase in circulating blood volume also compensates for the blood consumption required for digestion and absorption. Changes in vascular dynamics and hemodynamics between before and after a meal due to these factors are also reflected in the pulse wave. Therefore, electronic device 100 can acquire a pulse wave and estimate blood glucose levels based on changes in the waveform of the acquired pulse wave.

Based on the above estimation theory, an estimation formula for estimating blood glucose levels can be created by performing regression analysis on sample data of pre- and post-prandial blood glucose levels and pulse waves obtained from multiple subjects. During estimation, the created estimation formula can be applied to an index based on the subject's pulse wave to estimate the subject's blood glucose level. In particular, by performing regression analysis using sample data in which the variation in blood glucose levels is close to a normal distribution, the blood glucose level of the subject being tested can be estimated regardless of whether it is before or after a meal.

Figure 5 is a diagram illustrating an example of an estimation method based on changes in the pulse wave, showing an example of a pulse wave. The estimation formula for estimating blood glucose levels is created, for example, by regression analysis of an index (rising index) S1 that indicates the onset of the pulse wave, an Augmentation Index (AI), and the pulse rate PR.

The rise index Sl is derived based on the waveform shown in region D1 in Figure 5. Specifically, the rise index Sl is the ratio of the first minimum value to the first maximum value in an acceleration pulse wave derived by differentiating the pulse wave twice. For example, in the acceleration pulse wave shown as an example in Figure 6, the rise index Sl is expressed as -b/a. The rise index Sl decreases after a meal due to factors such as a decrease in blood fluidity, insulin secretion, and blood vessel dilation (relaxation) caused by an increase in body temperature.

AI is an index expressed as the ratio of the magnitude of the forward wave to the reflected wave of the pulse wave. A method for deriving AI will be explained with reference to Figure 7. Figure 7 is a diagram showing an example of a pulse wave acquired at the wrist using electronic device 100. Figure 7 shows a case where angular velocity sensor 131 is used as the pulsation detection means. Figure 7 shows the time integration of the angular velocity acquired by angular velocity sensor 131, with the horizontal axis representing time and the vertical axis representing angle. The acquired pulse wave may contain noise caused by, for example, the subject's body movement, so correction may be performed using a filter that removes the DC (Direct Current) component, and only the pulsation component may be extracted.

Pulse wave propagation is a phenomenon in which the pulsation of blood pumped from the heart travels through the walls of arteries and blood. The pulsation of blood pumped from the heart reaches the extremities of the limbs as a forward wave, and a portion of it is reflected by branching points of blood vessels, points of change in blood vessel diameter, etc., and returns as a reflected wave. AI is the magnitude of this reflected wave divided by the magnitude of the forward wave, and is expressed as AI _n = (P _Rn - P _Sn ) / (P _Fn - _{P Sn} ). Here, AI n is the AI for each pulse. AI may be calculated, for example, by measuring the pulse wave for several seconds and calculating the average value AI _ave of AI _n (n = an integer from 1 to n) for each pulse. AI is derived based on the waveform shown in region D2 in Figure 5. AI decreases after eating due to factors such as decreased blood fluidity and vascular dilation caused by elevated body temperature.

The pulse rate PR is derived based on the period _TPR of the pulse wave shown in Fig. 5. The pulse rate PR increases after a meal.

The electronic device 100 can estimate blood glucose levels using an estimation formula created based on these rising indexes Sl, AI, and pulse rate PR.

Figure 8 is a diagram illustrating another example of an estimation method based on changes in the pulse wave. Figure 8(a) shows a pulse wave, and Figure 8(b) shows the result of performing an FFT (Fast Fourier Transform) on the pulse wave in Figure 8(a). An estimation formula for estimating blood glucose levels is created, for example, by regression analysis of the fundamental wave and harmonic components (Fourier coefficients) derived by FFT. The peak value in the FFT result shown in Figure 8(b) changes based on changes in the pulse wave waveform. Therefore, blood glucose levels can be estimated using an estimation formula created based on the Fourier coefficients.

The electronic device 100 estimates the subject's blood glucose level using an estimation formula based on the above-mentioned rising index Sl, AI, pulse rate PR, Fourier coefficients, etc.

Here, we will explain how the electronic device 100 creates an estimation formula used when estimating a subject's blood glucose level. The estimation formula does not need to be created by the electronic device 100, but may be created in advance using a separate computer, etc. In this specification, the device that creates the estimation formula will be referred to as an estimation formula creation device. The created estimation formula is stored in advance, for example, in the memory unit 145, before the subject estimates their blood glucose level using the electronic device 100.

Figure 9 is a flow diagram for creating an estimation formula used by the electronic device 100 of Figure 1. The estimation formula is created by measuring the subject's pre- and post-prandial pulse waves using a pulse wave meter, and measuring the subject's pre- and post-prandial blood glucose levels using a blood glucose meter, and performing regression analysis based on the sample data obtained by the measurements. Note that "pre-prandial" refers to the subject's fasting state, and "post-prandial" refers to the time a predetermined time after a meal when blood glucose levels rise (for example, about one hour after starting a meal). The sample data obtained is not limited to pre- and post-prandial data, and can be data from a time period when blood glucose levels fluctuate greatly.

When creating an estimation formula, first, information about the subject's pre-meal blood glucose level and the pulse wave associated with the blood glucose level, measured using a blood glucose meter and a pulse wave meter, respectively, is input into the estimation formula creation device (step S101).

In addition, information regarding the subject's postprandial blood glucose level and the pulse wave associated with the blood glucose level, measured by a blood glucose meter and a pulse wave meter, respectively, is input to the estimation formula creation device (step S102). The blood glucose level input in steps S101 and S102 is measured by a blood glucose meter, for example, by taking a blood sample. In addition, in step S101 or step S102, the age of the subject for each sample data is also input.

The estimation formula creation device determines whether the number of samples of the sample data input in steps S101 and S102 is equal to or greater than N, which is sufficient for performing regression analysis (step S103). The number of samples N can be determined as appropriate and can be set to 100, for example. If the estimation formula creation device determines that the number of samples is less than N (No), it repeats steps S101 and S102 until the number of samples is equal to or greater than N. On the other hand, if the estimation formula creation device determines that the number of samples is equal to or greater than N (Yes), it proceeds to step S104 and calculates the estimation formula.

To calculate the estimation formula, the estimation formula creation device analyzes the input pre- and post-prandial pulse waves (step S104). In this embodiment, the estimation formula creation device analyzes the onset index Sl, AI, and pulse rate PR of the pre- and post-prandial pulse waves. Note that the estimation formula creation device may also perform FFT analysis to analyze the pulse waves.

The estimation formula creation device then performs regression analysis (step S105). The objective variable in the regression analysis is the postprandial blood glucose level. The explanatory variables in the regression analysis are the age input in step S101 or step S102, and the preprandial and postprandial pulse wave onset indices S1, AI, and pulse rate PR analyzed in step S104. If the estimation formula creation device performs FFT analysis in step S104, the explanatory variables may be, for example, Fourier coefficients calculated as a result of the FFT analysis.

The estimation formula creation device creates an estimation formula for estimating postprandial blood glucose levels based on the results of the regression analysis (step S106). An example of an estimation formula for estimating postprandial blood glucose levels is shown in equation (1) below.

In equation (1), GLa represents the postprandial blood glucose level. Furthermore, age represents age, PRb represents the preprandial pulse rate PR, AIb represents the preprandial AI, Slb represents the preprandial rise index Sl, PRa represents the postprandial pulse rate PR, AIa represents the postprandial AI, Sla represents the postprandial rise index Sl, and BLG represents the blood glucose level entered by the subject (measured by drawing blood). The blood glucose level BLG entered by the subject is a blood glucose level measured at a different time than the estimated blood glucose level GLa. In this embodiment, the blood glucose level BLG entered by the subject is a blood glucose level measured by drawing blood before a meal. By using the blood glucose level BLG measured by drawing blood in the estimation equation, the accuracy of blood glucose level estimation is improved.

Next, we will explain the flow for estimating a subject's blood glucose level using the estimation formula. Figure 10 is a flow diagram for estimating a subject's postprandial blood glucose level using the estimation formula created by the flow in Figure 9. Here, we will explain the case where the subject inputs their preprandial blood glucose level, measured using a blood glucose meter, through the input unit 141 of the electronic device 100.

First, the electronic device 100 inputs the subject's age based on the subject's operation of the input unit 141 (step S201).

In addition, the electronic device 100 inputs the pre-meal blood glucose level measured by the subject using a blood glucose meter based on the subject's operation of the input unit 141 (step S202).

The electronic device 100 also measures the subject's pre-meal pulse wave based on the subject's operation (step S203).

Then, after the subject has eaten, the electronic device 100 measures the subject's postprandial pulse wave based on the subject's operation (step S204).

Next, the electronic device 100 analyzes the measured pulse wave (step S205). Specifically, the electronic device 100 analyzes, for example, the rising index Sl, AI, and pulse rate PR related to the measured pulse wave.

The electronic device 100 estimates the subject's postprandial blood glucose level by applying the preprandial blood glucose level input in step S202, the rising index S1, AI, and pulse rate PR analyzed in step S205, and the subject's age to, for example, the above-mentioned equation (1) (step S206). The estimated postprandial blood glucose level is notified to the subject, for example, by the notification unit 147 of the electronic device 100.

Figure 11 shows a comparison between postprandial blood glucose levels estimated using the estimation formula created using the flow in Figure 9 and actually measured postprandial blood glucose levels. In the graph shown in Figure 11, the horizontal axis shows the measured postprandial blood glucose levels (actual measured values), and the vertical axis shows the estimated postprandial blood glucose levels. The measured blood glucose levels were measured using a Medisafe Fit blood glucose meter manufactured by Terumo Corporation. As shown in Figure 11, the measured and estimated values are generally within a range of ±20%. In other words, the estimation accuracy using the estimation formula can be said to be within 20%.

In this way, the electronic device 100 can non-invasively and quickly estimate postprandial blood glucose levels based on the preprandial blood glucose levels measured by blood sampling from the subject. In this embodiment, the estimation formula was created using preprandial and postprandial blood glucose levels and pulse waves, but the creation of the estimation formula is not limited to this, and the estimation formula may also be created using either preprandial or postprandial blood glucose levels and pulse waves. Furthermore, the electronic device 100 may estimate the subject's blood glucose level at any timing, not just postprandial blood glucose levels. The electronic device 100 can also estimate blood glucose levels at any timing non-invasively and quickly.

The electronic device 100 according to this embodiment may update the estimation formula stored in the memory unit 145 based on the subject's pre-meal blood glucose level and pulse wave acquired in steps S202 and S203 when estimating the blood glucose level. That is, the electronic device 100 can use the pre-meal blood glucose level and pulse wave acquired when estimating the blood glucose level as sample data for updating the estimation formula. This allows the estimation formula to be updated each time the subject estimates their blood glucose level, improving the accuracy of estimating post-meal blood glucose levels using the estimation formula.

(Second embodiment)
In the first embodiment, a case where an estimation formula is created based on the subject's pre- and post-prandial blood glucose levels and pulse waves is described. In the second embodiment, an example where an estimation formula is created based on the subject's own pre- and post-prandial blood glucose levels and pulse waves is described.

Figure 12 is a flow diagram for creating an estimation formula used by electronic device 100 according to this embodiment. In this embodiment, the estimation formula is described as being created by electronic device 100. Note that, as described in the first embodiment, the estimation formula may also be created by an estimation formula creation device different from electronic device 100.

First, the electronic device 100 inputs the pre-meal blood glucose level measured by the subject using a blood glucose meter based on the subject's operation of the input unit 141 (step S301).

The electronic device 100 also measures the subject's pre-meal pulse wave based on the subject's operation (step S302).

Then, after the subject has eaten, the electronic device 100 inputs the postprandial blood glucose level measured by the subject using a blood glucose meter based on the subject's operation of the input unit 141 (step S303). The blood glucose level input in steps S301 and S303 is measured by the blood glucose meter, for example, by the subject taking a blood sample.

The electronic device 100 also measures the subject's postprandial pulse wave based on the subject's operation (step S304).

The electronic device 100 determines whether the number of samples of the sample data input in steps S301 to S304 is equal to or greater than N, which is sufficient for performing regression analysis (step S305). The number of samples N can be determined as appropriate and can be set to 5, for example. If the estimation formula creation device determines that the number of samples is less than N (No), it repeats steps S301 to S304 until the number of samples is equal to or greater than N. On the other hand, if the estimation formula creation device determines that the number of samples is equal to or greater than N (Yes), it proceeds to step S306 and calculates the estimation formula.

The method for calculating the estimation formula in steps S306 to S308 is the same as steps S104 to S106 in Figure 9, so a detailed description thereof will be omitted here. The estimation formula created by electronic device 100 according to the flow shown in Figure 12 is, for example, a formula in which each coefficient in formula (1) is different.

Next, we will explain the flow for estimating a subject's blood glucose level using the estimation formula. Figure 13 is a flow diagram for estimating a subject's postprandial blood glucose level using the estimation formula created by the flow in Figure 12. Here, we will explain the case where the subject inputs the blood glucose level measured using a blood glucose meter through the input unit 141 of the electronic device 100.

First, the electronic device 100 inputs the subject's age based on the subject's operation of the input unit 141 (step S401).

In addition, the electronic device 100 inputs the pre-meal blood glucose level measured by the subject using a blood glucose meter based on the subject's operation of the input unit 141 (step S402).

The electronic device 100 also measures the subject's pre-meal pulse wave based on the subject's operation (step S403).

Then, after the subject has eaten, the electronic device 100 measures the subject's postprandial pulse wave based on the subject's operation (step S404).

Next, the electronic device 100 analyzes the measured pulse wave (step S405). Specifically, the electronic device 100 analyzes, for example, the rising index Sl, AI, and pulse rate PR related to the measured pulse wave.

The electronic device 100 estimates the subject's postprandial blood glucose level by applying the rising index Sl, AI, and pulse rate PR analyzed in step S405, as well as the subject's age, to the estimation formula created using the flow diagram in FIG. 12 (step S406). The estimated postprandial blood glucose level is notified to the subject, for example, by the notification unit 147 of the electronic device 100.

In this way, the electronic device 100 can non-invasively and quickly estimate postprandial blood glucose levels based on the preprandial blood glucose levels measured by blood sampling from the subject. In this embodiment, the estimation formula for estimating postprandial blood glucose levels is created based on sample data obtained from the subject, improving the accuracy of estimating the subject's postprandial blood glucose levels.

As with the first embodiment, the electronic device 100 according to this embodiment may also update the estimation formula stored in the memory unit 145 based on the subject's pre-prandial blood glucose level and pulse wave acquired in steps S402 and S403 during blood glucose level estimation. This allows the estimation formula to be updated each time the subject estimates their blood glucose level, improving the accuracy of post-prandial blood glucose level estimation using the estimation formula.

Furthermore, if a sufficient number of sample data samples can be collected from the subject, the electronic device 100 may estimate the blood glucose level based on the subject's pulse wave, rather than using the blood glucose level measured by blood sampling. For example, the electronic device 100 estimates the subject's pre-prandial blood glucose level based on the subject's pre-prandial pulse wave. In this way, the subject measures their pre-prandial pulse wave using the electronic device 100, and the electronic device 100 can estimate the subject's pre-prandial blood glucose level using an estimation formula based on the pre-prandial pulse wave. In this case, the electronic device 100 can also estimate the pre-prandial blood glucose level non-invasively and in a short time. Note that sufficient sample data refers to an amount of data that allows for the creation of an estimation formula that can estimate the subject's pre-prandial blood glucose level with an accuracy equal to or higher than a predetermined accuracy based on the pre-prandial pulse wave. Furthermore, the estimated blood glucose level is not limited to pre-prandial; post-prandial blood glucose levels may be estimated based on post-prandial pulse waves. Furthermore, the estimated blood glucose level is not limited to pre-prandial, but may be estimated at any timing based on pulse waves measured at any timing.

(Third embodiment)
In the first embodiment, a case where the electronic device 100 estimates a subject's postprandial blood glucose level has been described. In the third embodiment, an example where the electronic device 100 estimates a subject's postprandial lipid level will be described. Here, the lipid level includes triglycerides, total cholesterol, HDL cholesterol, LDL cholesterol, etc. In the description of this embodiment, the same points as in the first embodiment will be omitted as appropriate.

The electronic device 100 stores, in advance, for example, in the storage unit 145, estimation formulas for estimating lipid levels based on pulse waves. The electronic device 100 estimates lipid levels using these estimation formulas.

The estimation theory for estimating lipid levels based on pulse waves is the same as the estimation theory for blood glucose levels described in the first embodiment. That is, changes in blood lipid levels are also reflected in changes in the waveform of the pulse wave. Therefore, electronic device 100 can acquire pulse waves and estimate lipid levels based on changes in the acquired pulse waves. By inputting blood glucose levels along with the pulse waves during lipid estimation, electronic device 100 improves the accuracy of lipid level estimation.

Figure 14 is a flow diagram for creating an estimation formula used by electronic device 100 according to this embodiment. In this embodiment, too, the estimation formula is created by performing regression analysis based on sample data. In this embodiment, the estimation formula is created based on sample data such as pre-meal pulse waves, lipid levels, and blood glucose levels. In this embodiment, "pre-meal" refers to the subject's fasting state. "Post-meal" refers to the period a predetermined time after a meal when lipid levels are high (for example, approximately three hours after starting a meal). In particular, by performing regression analysis using sample data in which the variation in lipid levels is close to a normal distribution, it is possible to estimate lipid levels at any timing, whether before or after a meal, for the subject being tested.

In creating the estimation equation, first, information regarding the subject's pre-prandial blood glucose level, as well as the pulse wave and lipid levels associated with the blood glucose level, measured by a blood glucose meter, a pulse wave meter, and a lipid measurement device, is input into the estimation equation creation device (step S501).

In addition, information regarding the subject's postprandial blood glucose level, as well as the pulse wave and lipid levels associated with the blood glucose level, measured by the blood glucose meter, pulse wave meter, and lipid measurement device, respectively, is input to the estimation formula creation device (step S502). The blood glucose level input in steps S501 and S502 is measured by the blood glucose meter, for example, by taking a blood sample. In addition, in step S501 or step S502, the age of the subject for each sample data is also input.

The estimation formula creation device determines whether the number of samples of the sample data input in steps S501 and S502 is equal to or greater than N, which is sufficient for performing regression analysis (step S503). The number of samples N can be determined as appropriate and can be set to 100, for example. If the estimation formula creation device determines that the number of samples is less than N (No), it repeats steps S501 and S502 until the number of samples is equal to or greater than N. On the other hand, if the estimation formula creation device determines that the number of samples is equal to or greater than N (Yes), it proceeds to step S504 and calculates the estimation formula.

In calculating the estimation formula, the estimation formula creation device analyzes the input pre- and post-prandial pulse waves (step S504). In this embodiment, the estimation formula creation device analyzes the onset index Sl, AI, and pulse rate PR of the pre- and post-prandial pulse waves. Note that the estimation formula creation device may also perform FFT analysis to analyze the pulse waves.

The estimation formula creation device then performs regression analysis (step S505). The objective variable in the regression analysis is the postprandial lipid level. The explanatory variables in the regression analysis are the age input in step S501 or step S502, and the preprandial and postprandial pulse wave onset indices S1, AI, and pulse rate PR analyzed in step S504. If the estimation formula creation device performs FFT analysis in step S504, the explanatory variables may be, for example, Fourier coefficients calculated as a result of the FFT analysis.

The estimation formula creation device creates an estimation formula for estimating postprandial lipid levels based on the results of the regression analysis (step S506).

Next, we will explain the flow for estimating a subject's lipid levels using the estimation formula. Figure 15 is a flow diagram for estimating a subject's postprandial lipid levels using the estimation formula created by the flow in Figure 14. Here, we will explain the case where the subject inputs the blood glucose level measured using a blood glucose meter through the input unit 141 of the electronic device 100.

First, the electronic device 100 inputs the subject's age based on the subject's operation of the input unit 141 (step S601).

In addition, the electronic device 100 inputs the pre-meal blood glucose level measured by the subject using a blood glucose meter based on the subject's operation of the input unit 141 (step S602).

The electronic device 100 also measures the subject's pre-meal pulse wave based on the subject's operation (step S603).

Then, after the subject has eaten, the electronic device 100 inputs the postprandial blood glucose level measured by the subject using the blood glucose meter based on the subject's operation (step S604).

The electronic device 100 also measures the subject's postprandial pulse wave based on the subject's operation (step S605).

Next, the electronic device 100 analyzes the measured pulse wave (step S606). Specifically, the electronic device 100 analyzes, for example, the rising index Sl, AI, and pulse rate PR related to the measured pulse wave.

The electronic device 100 estimates the subject's postprandial lipid level by applying the rising index Sl, AI, and pulse rate PR analyzed in step S606, as well as the subject's age, to the estimation formula created using the flow diagram in FIG. 14 (step S607). The estimated postprandial lipid level is notified to the subject, for example, by the notification unit 147 of the electronic device 100.

Figure 16 shows a comparison between postprandial lipid values estimated using the estimation formula created according to the flow in Figure 14 and actually measured postprandial lipid values. In the graph shown in Figure 16, the horizontal axis shows the measured postprandial lipid values (actual measured values), and the vertical axis shows the estimated postprandial lipid values. The measured lipid values were measured using a Cobas b101 manufactured by Roche Diagnostics. As shown in Figure 16, the measured and estimated values are generally within a range of ±20%. In other words, the estimation accuracy using the estimation formula can be said to be within 20%.

In this way, the electronic device 100 can estimate postprandial lipid levels based on the preprandial and postprandial blood glucose levels measured by blood sampling from the subject.

In addition, electronic device 100 estimates lipid levels using pre- and post-prandial blood glucose levels. Therefore, electronic device 100 can estimate lipid levels by correcting (removing) the effect of post-prandial blood glucose levels on pulse waves. This improves the accuracy of lipid level estimation using electronic device 100.

In this embodiment, the estimation formula was created using pre- and post-prandial blood glucose levels, pulse waves, and lipid levels, but the creation of the estimation formula is not limited to this, and the estimation formula may also be created using either pre- or post-prandial blood glucose levels, pulse waves, and lipid levels. Furthermore, the electronic device 100 may estimate the subject's lipid levels at any timing, not just post-prandial lipid levels. The electronic device 100 can also estimate lipid levels at any timing non-invasively and in a short time.

In the electronic device 100 according to this embodiment, as described in the first embodiment, the estimation formula stored in the memory unit 145 may be updated based on the subject's pre-meal blood glucose level and pulse wave and post-meal blood glucose level and pulse wave acquired in steps S602 to S605 when estimating lipid levels. This allows the estimation formula to be updated each time the subject estimates their blood glucose level, improving the accuracy of estimating post-meal lipid levels using the estimation formula.

In the above first and second embodiments, an example has been described in which, when using the electronic device 100 to estimate a postprandial blood glucose level, the subject inputs the preprandial blood glucose level measured using a blood glucose meter using the input unit 141 of the electronic device 100. However, the preprandial blood glucose level may also be automatically input to the electronic device 100 from, for example, the blood glucose meter.

Figure 17 is a diagram showing a schematic diagram of communication between the electronic device 100 and the blood glucose meter 160. The blood glucose meter 160 has a communication unit and is capable of sending and receiving information via the communication unit 146 of the electronic device 100. For example, when the blood glucose meter 160 measures a blood glucose level (pre-meal blood glucose level) based on the subject's operation, it transmits the blood glucose level as the measurement result to the electronic device 100. The electronic device 100 uses the blood glucose level acquired from the blood glucose meter 160 to estimate the subject's post-meal blood glucose level, for example, according to the flow shown in Figure 10 or Figure 13, etc.

Similarly, in the third embodiment, the electronic device 100 may acquire blood glucose levels from a blood glucose meter 160 with which it can communicate. In this case, the electronic device 100 can estimate lipid levels based on the blood glucose levels acquired from the blood glucose meter 160.

Furthermore, in the above embodiment, an example was described in which the blood glucose level and lipid level estimation was performed by the electronic device 100, but the blood glucose level and lipid level estimation does not necessarily have to be performed by the electronic device 100. An example in which the blood glucose level and lipid level estimation is performed by a device other than the electronic device 100 will be described.

Figure 18 is a schematic diagram showing the overall configuration of a system according to one embodiment of the present invention. The system of the embodiment shown in Figure 18 includes an electronic device 100, a server 151, a mobile terminal 150, and a communication network. As shown in Figure 18, the pulse wave measured by the electronic device 100 is transmitted to the server 151 via the communication network and stored in the server 151 as the subject's personal information. The server 151 estimates the subject's blood glucose or lipid levels by comparing them with the subject's previously acquired information and various databases. The server 151 may also create advice optimized for the subject. The server 151 returns the estimation results and advice to the mobile terminal 150 owned by the subject. A system can be constructed in which the mobile terminal 150 reports the received estimation results and advice on the display of the mobile terminal 150. By utilizing the communication function of the electronic device 100, the server 151 can collect information from multiple users, further improving the accuracy of the estimation. Furthermore, because the mobile terminal 150 is used as the notification means, the electronic device 100 does not require the notification unit 147, allowing for further miniaturization. Furthermore, because the subject's blood glucose level or lipid level is estimated by the server 151, the computational burden on the control unit 143 of the electronic device 100 can be reduced. Furthermore, because the server 151 can store previously acquired information about the subject, the burden on the storage unit 145 of the electronic device 100 can be reduced. This allows for further miniaturization and simplification of the electronic device 100. The computational processing speed is also improved.

In the system according to this embodiment, the electronic device 100 and the mobile terminal 150 are connected via a communications network via the server 151, but the system according to the present invention is not limited to this. The electronic device 100 and the mobile terminal 150 may also be directly connected via a communications network without using the server 151.

While specific examples have been described to provide a complete and clear disclosure of the present invention, the appended claims should not be construed as being limited to the above-described embodiments, but should be construed to embody all modifications and alternative arrangements that may be conceived by those skilled in the art within the scope of the fundamental subject matter presented herein.

For example, in the above embodiment, the sensor unit 130 is described as having an angular velocity sensor 131, but the electronic device 100 according to the present invention is not limited to this. The sensor unit 130 may be an optical pulse wave sensor consisting of a light-emitting unit and a light-receiving unit, or may be a pressure sensor. Furthermore, the electronic device 100 does not have to be worn on the wrist. The sensor unit 130 may be placed on an artery in the neck, ankle, thigh, ear, or other area.

DESCRIPTION OF SYMBOLS 100 Electronic device 110 Wearing part 120 Measurement part 120a Back side 120b Front side 111 Opening 130 Sensor part 131 Angular velocity sensor 132 Pulse contact part 133 Support part 140 Elastic body 141 Input part 143 Control part 144 Power supply part 145 Storage part 146 Communication part 147 Notification part 150 Portable terminal 151 Server 160 Blood glucose meter

Claims (16)

A method for generating an estimation model for estimating a blood glucose level or glucose metabolism of a subject based on a pulse wave acquired from the subject, comprising:
creating the estimation model based on sample data in which information on the pulse wave acquired from the subject is associated with information on the blood glucose level measured from the subject;
A method for generating an estimation model, wherein the sample data uses at least one of a rising index SI based on an accelerated pulse wave derived by differentiating the pulse wave twice, an AI based on the forward wave and reflected wave of the pulse wave, and a pulse rate PR calculated based on the period of the pulse wave . The method for generating an estimation model according to claim 1 , wherein when the AI is used for the sample data, the AI is based on the forward wave of the pulse wave minus the reflected wave. A method for generating an estimation model for estimating a blood glucose level or glucose metabolism of a subject based on a pulse wave acquired from the subject, comprising:
A method for generating an estimation model, which creates the estimation model based on sample data in which information about pulse waves obtained from a subject is associated with information about blood glucose levels measured from the subject , by performing regression analysis on the pulse waves that have been subjected to a fast Fourier transform . The method for generating an estimation model according to claim 1 , wherein the sample data is obtained from a plurality of subjects. The method for generating an estimation model according to claim 1 , wherein the subject is the target person. The method for generating an estimation model according to claim 1 , wherein the sample data uses data in which blood glucose level variations are close to a normal distribution. The method for generating an estimation model according to claim 1 , wherein the sample data is acquired at least either before or after a meal by the subject. 8. The method for generating an estimation model according to claim 1, wherein the sample data is obtained by measuring the pulse wave with a pulse wave meter and the blood glucose level with a blood glucose meter. The method for generating an estimation model according to claim 1 , wherein the sample data is obtained by measuring the blood glucose level by drawing blood from the subject. The method for generating an estimation model according to claim 1 , wherein the sample data further includes an age of the subject. A program that causes a computer to execute creation of an estimation model that estimates a blood glucose level or glucose metabolism of a subject based on a pulse wave acquired from the subject, based on sample data in which information on a pulse wave acquired from the subject and information on a blood glucose level measured from the subject are associated with each other,
The sample data uses at least one of a rising index SI based on an accelerated pulse wave derived by differentiating the pulse wave twice, an AI based on a forward wave and a reflected wave of the pulse wave, and a pulse rate PR calculated based on a period of the pulse wave.
program. and causing a computer to execute a regression analysis based on the pulse wave obtained from the subject, the estimation model estimating the blood glucose level or glucose metabolism of the subject based on the pulse wave obtained from the subject, based on sample data in which information on the pulse wave obtained from the subject is associated with information on the blood glucose level measured from the subject .
program. A recording medium storing a program for causing a computer to execute creation of an estimation model for estimating a blood glucose level or glucose metabolism of a subject based on a pulse wave acquired from the subject, based on sample data in which information on a pulse wave acquired from the subject and information on a blood glucose level measured from the subject are associated with each other,
The sample data uses at least one of a rising index SI based on an accelerated pulse wave derived by differentiating the pulse wave twice, an AI based on a forward wave and a reflected wave of the pulse wave, and a pulse rate PR calculated based on a period of the pulse wave.
Recording medium. A recording medium storing a program for causing a computer to execute the creation of an estimation model for estimating a blood glucose level or glucose metabolism of a subject based on a pulse wave obtained from the subject, based on sample data in which information on a pulse wave obtained from the subject is associated with information on a blood glucose level measured from the subject , by regression analysis based on the pulse wave obtained from the subject that has been subjected to fast Fourier transform . 1. An electronic device comprising: a control unit that creates an estimation model for estimating a blood glucose level or glucose metabolism of a subject based on a pulse wave acquired from the subject, based on sample data in which information about a pulse wave acquired from the subject and information about a blood glucose level measured from the subject are associated with each other;
The sample data uses at least one of a rising index SI based on an accelerated pulse wave derived by differentiating the pulse wave twice, an AI based on a forward wave and a reflected wave of the pulse wave, and a pulse rate PR calculated based on a period of the pulse wave.
electronic equipment. An electronic device comprising: a control unit that creates an estimation model for estimating a blood glucose level or glucose metabolism of a subject based on a pulse wave acquired from the subject, using sample data in which information about the pulse wave acquired from the subject is associated with information about the blood glucose level measured from the subject , by performing regression analysis on the pulse wave that has been subjected to a fast Fourier transform .