Calibration method of blood pressure estimation value in wearable device, calibration assist device, and provision method of blood pressure estimation monitoring service

Abstract

To provide a calibration method of a blood pressure estimation value unique to a wearable device user, a calibration assist device, and a provision method of blood pressure estimation monitoring service.SOLUTION: There are provided a calibration method of a blood pressure estimation value in a wearable device, a calibration assist device that can be used for the calibration method, and a provision method of blood pressure estimation monitoring service, which acquire different blood pressure-related signals of the same individual output from the wearable device and determine a blood pressure estimation model unique to the user by using apparatus characteristics that a blood pressure measurement value of a blood pressure manometer authenticated as a medical apparatus changes according to the height from the heart.SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] The present invention relates to a method for calibrating a blood pressure estimated value in a wearable device, a calibration assistant device that can be used for the calibration method, and a method for providing a blood pressure estimation monitoring service. [Background technology]

[0002] Conventionally, the gold standard for measuring blood pressure is to attach a cuff-type blood pressure monitor to the upper arm or wrist, apply a certain amount of pressure to abstain from blood flow, and use a microphone to detect the sound waves (Korotkoff sounds) generated by the pressure waves when the blood vessels reopen, thereby measuring the systolic and diastolic blood pressure. A condition for these devices is that they must be measured at the same height as the heart.

[0003] Meanwhile, in recent years, a method called Photoplethysmography (PPG), which uses light to measure the expansion and contraction of capillaries in fingers and wrists, has become popular, as typified by wearable devices such as smart watches. These are non-invasive methods that use a photodiode to measure the absorption of light from an LED by hemoglobin in subcutaneous capillary blood. Although they are not certified as medical devices, devices that display measurement results such as heart rate, heart rate variability, and oxygen saturation on the wearable device itself, as well as on a smartphone or tablet with an installed app, are becoming more and more popular, and due to their ease of use and multifunctionality, they are being used to manage individuals' physical conditions. However, since optical methods cannot directly measure pressure, various blood pressure estimation models that correlate optical measurement results with blood pressure are being researched for blood pressure measurement, which is important for managing physical condition. For this reason, it has not yet been approved as a medical device in Japan, and blood pressure calculated by non-invasive optical measurement is considered to be a blood pressure estimate. Therefore, even on commercially available smart watches, blood pressure is displayed as a reference value, and the measurement accuracy has not been verified.

[0004] As a method for estimating blood pressure using a non-invasive wearable device, Patent Documents 1, 2, and 3 disclose an intravascular pressure estimation method and device, and non-invasive blood pressure estimation and monitoring. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Patent Publication No. 2017-35453 [Patent Document 2] Special table number 2017-521115 [Patent Document 3] Patent No. 5984088 [Non-patent literature]

[0006] [Non-Patent Document 1] AIST Anthropometric Database 1991-92 https: / / www.airc.aist.go.jp / dhrt / 91-92 / data / list.html [Non-Patent Document 2] Ken Kawakami, 2022 Hokkaido University Doctoral (Information Science) Thesis No. 15088. https: / / eprints.lib.hokudai.ac.jp / dspace / bitstream / 2115 / 85574 / 1 / Ken_Kawakami.pdf [Non-Patent Document 3] Matsura, K, Rolfe, P., Toda, S. et al.Cuffless blood pressure estimation using only a smartphone.Sci Rep 8,7298(2018). https: / / doi.org / 10.1038 / s41598-018-25681-5 Summary of the Invention [Problem to be solved by the invention]

[0007] In the past, the regression coefficients of blood pressure estimation models in wearable devices were determined by obtaining blood pressure measurements from a group of subjects with various physiological backgrounds. However, when actually estimating blood pressure, there was large individual variation, making it impossible to estimate blood pressure with high accuracy. In addition, it was necessary to collect a large number of subjects with different blood pressure levels and obtain their blood pressure measurements, which was a costly and cumbersome procedure. Even when obtaining different blood pressure measurements from the same individual, the regression coefficients of the blood pressure estimation model were determined by forcibly changing the blood pressure through exercise stress or psychological stress, so blood pressure measurements had to be performed under conditions other than medical conditions, that is, in a resting state. Therefore, it has been difficult to obtain multiple different blood pressure measurements from individuals with illnesses, elderly people, and children, who are difficult to subject to exercise or psychological stress. In view of these issues, the present invention provides a method for calibrating an estimated blood pressure value of an individual user in a wearable device, a calibration assistant device that can be used for the calibration method, and a method for providing a blood pressure estimation monitoring service. [Means for solving the problem]

[0008] In one aspect of the present invention, a blood pressure monitor, which is a medical device, and the wearable device to be calibrated are attached to a wrist or an ankle of a user of the wearable device, a blood pressure-related signal acquisition step of acquiring blood pressure measurements of the sphygmomanometer and blood pressure-related signals of the wearable device while aligning the height of the part where the sphygmomanometer is attached and the height of the part where the wearable device is attached; determining a regression coefficient of a blood pressure estimation model of the wearable device from the blood pressure measurement values ​​and the blood pressure related signals obtained by performing the blood pressure related signal acquisition step multiple times while changing the height of the attachment part from the heart; Setting a user-specific blood pressure estimation model based on the determined regression coefficients in the wearable device; Characterized in that it has A method for calibrating blood pressure estimates in a wearable device may be provided.

[0009] The blood pressure-related signal acquiring step acquires a plurality of blood pressure measurements and blood pressure-related signals from the same user while the user is at rest, without imposing any exercise load or psychological load on the user, and only by changing the height of the attachment part from the heart. A method for calibrating blood pressure estimates in a wearable device may be provided.

[0010] The blood pressure-related signal acquisition step includes the step of: having the user wear a blood pressure monitor, which is a medical device, on one wrist or ankle and a wearable device to be calibrated on the other wrist or ankle while in a sitting, standing or lying position; The peripheral blood pressure due to the hydrostatic pressure of blood is arbitrarily changed by raising and lowering the upper arm or forearm, or the thigh or lower leg, thereby obtaining a plurality of blood pressure measurements and blood pressure-related signals from the same user. A method for calibrating blood pressure estimates in the wearable device may be provided.

[0011] The step of determining the regression coefficients of the blood pressure estimation model of the wearable device includes a step of determining the regression coefficients of the blood pressure estimation model of the wearable device from a plurality of blood pressure measurement values ​​and blood pressure related signals obtained from the same user by performing the blood pressure related signal acquisition step a plurality of times using each part of the body whose height from the heart is fixed (buttocks, knees, nipples, ribs, shoulders, ears, head) as a landmark. A method for calibrating blood pressure estimates in the wearable device may be provided.

[0012] A calibration assistant device that can be used in the method for calibrating a blood pressure estimate in the wearable device, The forearm, on which the medical device, a blood pressure monitor, and the wearable device to be calibrated, are attached, is held at the same height, and both forearms are raised and lowered simultaneously. By simply changing the height from the heart and changing the peripheral blood pressure due to blood hydrostatic pressure, At rest, without any physical or psychological stress, A device that allows for the acquisition of multiple blood pressure measurements and blood pressure related signals from the same user. A calibration aid may be provided.

[0013] The calibration assistant device has a horizontal rotation axis that can be moved up and down to match the height of the user's heart (the fourth rib or the inferior angle of the scapula); The position and angle of the forearm holder can be adjusted, and by holding the user's forearm from below, Without straining your arms to maintain height, you can sit still. A calibration aid may be provided which features multiple blood pressure measurements and blood pressure related signal acquisition from the same user.

[0014] The calibration assistant device fixes the angle of the upper arm support at a multiple of 12.9 mm, which is the blood column height equivalent to 1 mmHg, according to the height of both arms, so that the vertical height relative to the heart position can be easily converted into blood pressure units. The lock hole is provided with a rotating holder and a horizontal rotating shaft. A calibration aid may be provided.

[0015] A method for providing a blood pressure estimation monitoring service, comprising a step of setting a user-specific blood pressure estimation model in the wearable device, comprising: Through an information terminal with a dedicated app installed, The user-specific regression coefficients and blood pressure estimation model determined by the calibration method of the blood pressure estimation value in the wearable device, The information is stored on the wearable device worn by the user and saved in a personal account in a database. Your daily blood pressure estimates and other physical information history It can be displayed on the web or on an information terminal upon request. It may include a method for providing a blood pressure estimation monitoring service.

[0016] The blood pressure estimation monitoring service providing method relates to a method for calibrating a blood pressure estimate in the wearable device. (1) Equipment provision services; (2) Proofreading and updating services; (3) Record display service; (4) Regular report service; (5) Reservation and billing payment services; It is composed of Connect the wearable device to an information terminal on which a dedicated application is installed; The main feature is that it includes a calibration update service that periodically updates the user's specific blood pressure estimation model. The blood pressure estimation monitoring service providing method may be included.

[0017] The said proofreading update service: For a user of the wearable device, The present invention provides a method of providing explanations by lending the proofreading assistance device, recording and viewing the proofreading method on a web server, watching YouTube (registered trademark), explaining the method through a paper manual, and interacting through a video conference. The blood pressure estimation monitoring service providing method may be included.

[0018] The regular report service visualizes blood pressure estimates and physical information history over a certain period of time using a calibrated wearable device worn by the user. Users can take the device with them to medical examinations of their own volition and refer to it as reference data when a doctor is not present. The blood pressure estimation monitoring service providing method may be included. Effect of the Invention

[0019] The regression coefficients of the blood pressure estimation model in the wearable device are determined by obtaining multiple different blood pressure measurements from an individual user, rather than being determined from a group of subjects with various physiological backgrounds as in the past, enabling highly accurate blood pressure estimation that is not dependent on individual differences. It also avoids the costly and cumbersome procedure of collecting blood pressure measurements from a large number of subjects with different blood pressures. By using the calibration assistance device and calibration method of the present invention, multiple blood pressure measurements can be obtained from the same individual by up and down movement of the part where the wearable device is attached, and a blood pressure estimation model and its regression coefficients can be determined for each user, so that the regression coefficients of the blood pressure estimation model can be determined without imposing any exercise or psychological stress. This will make it possible to provide a blood pressure estimation and monitoring service that is customized for individuals, such as elderly people and children, who have illnesses and for whom it is difficult to impose exercise or psychological stress. In addition, blood pressure estimates obtained from the results of wearable device calibration such as PPG are not recognized as medical blood pressures and are merely estimates. In the present invention, blood pressure is estimated by measuring peripheral blood pressure, which is highly correlated with blood pressure, using a wearable device. [Brief description of the drawings]

[0020] [Figure 1] FIG. 1 is an overall schematic diagram illustrating a method for calibrating a blood pressure estimated value in a wearable device according to one embodiment of the present invention, a method for using a calibration assistance device, and a method for providing a blood pressure estimation monitoring service. [Diagram 2] 1 is an overall schematic diagram illustrating a method for calibrating a blood pressure estimated value in a wearable device according to an embodiment of the present invention. FIG. [Diagram 3] 1 is an overall schematic diagram showing a limb supporting device which is a calibration assistance device according to one embodiment of the present invention; [Figure 4] FIG. 4 is a partial enlarged view of the limb support device of FIG. 3 according to one embodiment of the present invention. [Diagram 5] FIG. 4 is an overall schematic view showing a limb-supporting method using the limb-supporting device of FIG. 3 according to one embodiment of the present invention. [Figure 6] 1 is an overall schematic view showing a limb supporting device according to an embodiment of the present invention; [Figure 7] FIG. 1 is an overall schematic view showing a limb supporting method when a limb supporting device according to an embodiment of the present invention is not used. [Figure 8]FIG. 13 is an overall schematic view showing a different limb supporting method when a limb supporting device according to an embodiment of the present invention is not used. [Figure 9] 1 is an overall schematic view showing a limb support device according to an embodiment of the present invention when applied to a foot. [Figure 10] FIG. 2 is a diagram showing the relationship between heart position and blood hydrostatic pressure according to an embodiment of the present invention. [Figure 11] 1 is a graph showing the measurement results of the wrist-type blood pressure monitor according to one embodiment of the present invention. [Figure 12] 1 is a table showing measurement results of the wrist-type blood pressure monitor according to an embodiment of the present invention and an example of explanatory variable measurements of a PPG waveform. [Figure 13] 1 is a graph showing a comparison of blood pressure measurements and blood pressure estimates for an individual using a wrist blood pressure monitor according to an embodiment of the present invention. [Figure 14] 1 is a table showing the measurement results and the estimation accuracy of blood pressure estimates of an individual using a wrist-type blood pressure monitor according to an embodiment of the present invention. [Figure 15] FIG. 1 is an overall system configuration diagram showing an example of a blood pressure estimation monitoring service according to an embodiment of the present invention. [Figure 16] 1 is an overall schematic diagram showing an example of a blood pressure estimation monitoring service providing method according to an embodiment of the present invention; DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0021] Hereinafter, specific embodiments of the present invention will be described with reference to the drawings.

[0022] [Outline of the calibration method for estimated blood pressure values ​​in wearable devices, calibration assistance device, and method for providing estimated blood pressure monitoring services] (Calibration method and calibration assistance device for wearable devices)

[0023] FIG. 1 is a schematic overall diagram of a method for calibrating an estimated blood pressure value in a wearable device, a method for using a calibration assistant device, and a method for providing a blood pressure estimation monitoring service 100 according to an embodiment of the present invention. The method for calibrating a blood pressure estimated value in a wearable device, the method for using a calibration assistant device, and the method for providing a blood pressure estimated monitoring service 100 of the present invention are as follows: Wearable device calibration 101 for individuals who use it, Implementing 102 the calibration results on the wearable device; Recording physical condition using wearable devices 103 and Monitoring of information terminals such as smartphones, tablets, and PCs 104; Regularly update and store the database 105; It is equipped with:

[0024] The calibration flow chart 200 in FIG. 2 shows the main steps of wearable device calibration 101 performed for each individual. 201 to 206 show the setting steps of the calibration assistant device 300 in FIG. 3. After changing the wearable device 402 from the measurement mode to the calibration mode and selecting a blood pressure estimation model appropriate for the individual user 207, the sphygmomanometer and the wearable device are switched on, data acquisition is started 208, and the number of data is determined in (number of data sets>explanatory variable+1?) 209. If it is less than the (explanatory variable+1) of the regression analysis, the height and angle of the forearm holder is adjusted 206, and data is acquired again. After the data is acquired, regression analysis is performed 210, and regression coefficients and accuracy indexes are calculated 211. Then, the wearable device storage unit is updated and the database is saved 212, and these values ​​are transferred to the storage unit of the wearable device and the database, and are updated and saved. Finally, the wearable device is returned from the calibration mode to the measurement mode, two or more points at any height are selected, and in the reproducibility confirmation step 213, the blood pressure estimated by the wearable device is compared with the blood pressure measured by the sphygmomanometer, the reproducibility is confirmed, and the calibration process is completed. Here, a wrist-type blood pressure monitor that can easily move up and down is given as an example of a blood pressure monitor that is a medical device, but the present invention is not limited to this as long as it can move up and down while being worn.

[0025] An example of the above calibration method will be described in more detail below. An individual using a wearable device sits on chair 309 in FIG. 3, and maintains a resting state without physical or psychological stress using calibration assistance device 300 or any of the devices or methods in FIGS. 5 to 9. In this case, in the example of a sitting position, horizontal rotation axis 301 is adjusted to the height of heart 400 (as a guide, the nipple and the fourth rib in the front of the body, and the subscapular angle in the rear of the body) by horizontal rotation axis moving tool 306. The individual's upper arm is aligned with upper arm instruction part 304, and the angle and position of upper arm support part fixing tool 305 and forearm holding part fixing tool 308 are adjusted so that forearm holding part 307 is in a position where it can support the weight of the forearm from below the forearm.

[0026] A wrist-type blood pressure monitor 401 for measuring blood pressure is worn on one hand, and a wearable device 402 to be calibrated is worn on the other wrist or palm. Both hands are placed on the forearm holder 307, and the arms are relaxed. This minimizes physical and psychological disturbances other than blood pulsation during PPG measurement, allowing accurate pulse wave measurement.

[0027] The height range from the heart 400 is determined according to the range of blood pressure changes that can occur in daily life. For example, when the fluctuation from the mean blood pressure is +30mmHg and -20mmHg, 1mmHg corresponds to a blood hydrostatic pressure of 12.9mm (assuming a mercury density of 13.5g / mL and a blood density of 1.05g / mL), so the height is 387mm below the heart and 258mm above the heart, respectively. Since the length from the shoulder to the wrist of an average Japanese person is 700mm, this is not an impossible range (Non-Patent Document 1).

[0028] The horizontal rotation shaft 301 is turned using the horizontal rotation shaft handle 303 to determine the angle and position of the upper arm support part 304 and the forearm holding part 307, that is, the height of the attachment part of the measuring instrument from the heart. The relative height from the heart can be determined from the geometric positional relationship shown in Figure 10. In other words, the difference in altitude between the wearable device mounting position and the heart is expressed as the vertical distance from the heart, calculated by Zh = r × sinθ, where r is the distance from the heart and θ is the angle. The blood hydrostatic pressure is expressed as ρ × g × Zh (ρ: blood density 1.055 / mL, g: gravitational acceleration 9.8 m / s 2 , θ: angle between the horizontal plane and the line connecting the center of the heart and the position where the device is attached). The blood pressure measured at the position where the device is attached is the blood pressure value at the heart position plus or minus blood hydrostatic pressure.

[0029] As shown in Figure 4, in order to more easily grasp the relationship between the relative height from the heart and blood hydrostatic pressure, holes may be drilled at regular intervals in horizontal rotation shaft holder 302 so that the height of the device attachment part is a multiple of the blood hydrostatic pressure of 12.9 mm, and forearm holding part 307 may be fixed by passing horizontal rotation shaft holder lock pin 302a through horizontal rotation shaft holder lock hole 302b and horizontal rotation shaft lock hole 301a so that forearm holding part 307 comes to rest at a position where the blood hydrostatic pressure changes at regular intervals based on the angle and position of forearm holding part 307.

[0030] Furthermore, the calibration assist device 300 and limb holding method for changing the change in blood pressure hydrostatic pressure by only vertical movement of the upper arm and forearm in a resting state are not limited to those shown in FIG. For example, as shown in Fig. 6, a forearm holder 307 may be provided on a stand 404. In this case, if a heart position marker 403 is provided to indicate the heart height position and a scale is provided every 12.9 mm of the blood column height corresponding to 1 mmHg, the change in blood hydrostatic pressure can be easily grasped.

[0031] After determining the height and position of both hands, the wrist-type blood pressure monitor 401 is switched on, and at the same time, signal waveform recording of the wearable device 402 to be calibrated is started.

[0032] It is desirable to obtain the number of data required for calibration, i.e., the number of blood pressure measurement heights, at least (the number of explanatory variables + 1) in the regression analysis for the selected blood pressure estimation model, and preferably at least 5. Here, the case where there are two or more explanatory variables is particularly called multiple regression analysis. These regression coefficients are determined by regression analysis, and a blood pressure estimation model formula for each individual is determined. Since accuracy indicators such as correlation coefficient, standard error, and P value, which are indicators of the similarity between the measured value and the estimated value, are calculated from the regression analysis, these values ​​are also stored in the wearable device 402 main body and database S116 together with the selected blood pressure estimation model and regression coefficient every time calibration is updated. Furthermore, for verification purposes, blood pressure is measured again at two or more heights other than the height used for calibration, and the updated blood pressure estimate is compared with the measured blood pressure value to confirm the estimation accuracy.

[0033] Various types of blood pressure estimation models using PPG signals have been proposed depending on the selection of explanatory variables. These explanatory variables are closely related to blood vessel diameter, blood vessel length, blood vessel elasticity, intravascular pressure, intravascular resistance, blood viscosity, density, etc., which affect the pulse wave velocity transmitted to the periphery by the blood pumped from the heart 400 expanding the artery. By combining these, two physical quantities with different properties, namely, the LED light attenuation due to the change in the amount of hemoglobin in the pulsating blood, and the pressure in the blood vessel, are related.

[0034] For example, in Non-Patent Document 2, explanatory variables of intravascular pressure are read from the velocity pulse wave and acceleration pulse wave of the PPG signal, and are compared with the blood pressure, which is the objective variable, to calculate regression coefficients. In this model, five explanatory variables, mNPV, Qd1, Qd2, PR, and R, are read from PPG signal analysis, and seven regression coefficients (ad1, ad2, bd, cd, ap, bp, and cp) are calculated by multiple regression analysis. As a result, estimated values ​​of diastolic and systolic blood pressure are calculated from the following formulas. 1) ln(diastolic intravascular pressure) = ad1 × ln(Qd1) + ad2 × ln(Qd2) + bd × ln(R) + cd 2)ln(pulse pressure)=ap×ln(PR)+bp×ln(R)+cp 3) Systolic intravascular pressure = diastolic intravascular pressure + pulse pressure

[0035] In addition, Non-Patent Document 3 proposes a blood pressure estimation model with two sets of 3 regression coefficients (set 1: a1, a2, b; set 2: c1, c2, d) as follows, using heart rate PR and standardized ejection volume mNPV as two explanatory variables. 1) ln(systolic intravascular pressure) = a1 × ln(mNPV) + a2 × ln(PR) + b 2) ln(diastolic pressure) = c1 × ln(mNPV) + c2 × ln(PR) + d 3) Mean intravascular pressure = diastolic intravascular pressure + (systolic intravascular pressure - diastolic intravascular pressure) / 3

[0036] Furthermore, in Patent Document 3, the pulse wave transit time (PTT) is calculated as an explanatory variable from an ECG signal synchronized with the PPG signal, two regression coefficients (b1, b2) are determined, and the systolic intravascular pressure is estimated as follows: 1) Systolic intravascular pressure = b1 / PTT 2 + b2

[0037] In the present invention, the service provider selects the blood pressure estimation model most suitable for the individual user by referring to expert opinions, the latest reference literature, etc., and recommends it to the user. Alternatively, an algorithm that automatically recommends a model with high estimation accuracy based on user input information such as weight, height, age, sex, and medical history may be incorporated into a dedicated app. Then, the above regression coefficients are calculated not from a large group of subjects, but from an individual user of the wearable device, by varying the height of the device attachment point up and down, utilizing the changes in blood hydrostatic pressure to obtain multiple different blood pressure data, and a regression coefficient specific to that individual is calculated. In addition, in the calibration of the PPG wearable device of the present invention, a wearable device 402 capable of outputting PPG raw signal waveforms is used to accommodate multiple blood pressure estimation models.

[0038] The material that constitutes the calibration assistant device 300 is, for example, the arm, which is said to weigh 6% of the body weight, so if the subject weighs 60 kg, the weight would be 3.6 kg. When this weight is supported by the shoulder and wrist, a load of 1.8 kg is applied to the upper arm support part with a length of 700 mm, so any frame structural material that can withstand a moment of 1.26 kg m can be used, and examples that can be used include, but are not limited to, a PVC pipe with a diameter of 30 mm or a metal pipe with a diameter of 20 mm.

[0039] If there is no equipment such as the calibration assistant device 300 nearby, a measuring tape 405 may be attached to a wall or pillar and the height of the device attachment part may be changed for each scale, as shown in Fig. 7. In this case, if the heights are kept the same, the blood pressure measurement with the wrist-type blood pressure monitor 401 as the reference and the measurement with the wearable device 402 as the test device may be performed separately on the same hand.

[0040] Furthermore, if a scale cannot be prepared, the measurement results of the wrist blood pressure monitor and the wearable device may be recorded by placing both hands on landmarks on the body, such as the buttocks, above the knees, the solar plexus, the nipples, the shoulders, the ears, and above the head, in positions such as (a) forearms hanging down 406, (b) forearms above the knees 407, (c) arms crossed over the fourth rib 408, (d) forearms above the shoulders 409, (e) forearms above the ears 410, and (f) forearms above the head 411, as shown in Figure 8.

[0041] In the above calibration of the estimated blood pressure, the range of change in blood pressure is important, and the height itself is not used in the regression analysis.

[0042] In the case of an intractable neurological disease or a neurodegenerative disease, the wearable device may be attached to the lower leg or foot as shown in Fig. 9. In this case, too, the calibration assistant device 300 is modified so that the lower leg or foot is placed on the forearm holder 307 and the knee is bent to raise or lower the attachment part, so that multiple blood pressure measurements at different heights from the heart 400 can be obtained. (Example of PPG wearable device calibration method)

[0043] A calibration example of a PPG wearable device according to an embodiment of the present invention will now be described.

[0044] Figure 11 shows an example of obtaining multiple blood pressure measurements from one individual by changing the height of the part where the wearable device is attached from the heart using the calibration assistant device 300. Here, the wrist-type blood pressure monitor WS01 (managed medical device certification number: 301AGBZX00088000) made by ELECOM HEALTHCARE CO., LTD. was used as a medical device certified wrist-type blood pressure monitor.

[0045] The measurement results of a wrist blood pressure monitor, both systolic and diastolic intravascular pressure, change according to the relative height from the heart = blood hydrostatic pressure change. In this example, we show an example of obtaining eight sets of blood pressure data from one individual.

[0046] As shown in the table of FIG. 12, the heart rate HR hardly changes in response to changes in blood pressure. This is a major difference from the data of subjects subjected to conventional exercise and psychological stress. Since it is often ethically unacceptable to subject elderly people, children, and unhealthy individuals to exercise or psychological stress to cause changes in their blood pressure, the present invention makes it effective to measure blood pressure without a load.

[0047] Here, we show an example of a PPG wearable device, using the MAXM86161 from Analog Devices, Inc. and data collection software Optical Module EV System to obtain heart rate PR and standardized ejection volume mNPV.

[0048] After obtaining the above data set, for example, a blood pressure estimation model described in Non-Patent Document 3 can be used. In this blood pressure estimation model, as shown in the following formula, the heart rate PR and the standardized ejection volume mNPV are two explanatory variables, and two sets of three multiple regression coefficients ad1, ad2, cd, as1, ad2, and cs are obtained by multiple regression analysis. 1) ln(systolic intravascular pressure) = ad1 × ln(mNPV) + ad2 × ln(PR) + cd 2) ln(diastolic pressure) = as1 × ln(mNPV) + as2 × ln(PR) + cs 3) Mean intravascular pressure = diastolic intravascular pressure + (systolic intravascular pressure - diastolic intravascular pressure) / 3

[0049] Figure 13 shows (a) a comparison between measured blood pressure and estimated peripheral blood pressure, (b) the correlation between the two, and (c) a Bland-Altman plot, which is often used to measure intravascular pressure similarity. (a) shows that the average values ​​and variances of measured blood pressure and estimated peripheral blood pressure are in roughly the same range. (b) shows that the agreement between measured and estimated values ​​is high, with a correlation of 0.99 or more for systolic blood pressure, 0.94 or more for diastolic blood pressure, and 0.96 for mean blood pressure. (c) shows that the variance of the difference between measured and estimated values ​​is within ±2σ (standard deviation), indicating that more than 95% of estimated values ​​are in this range.

[0050] As shown in the table in Figure 14, the correlation coefficients obtained from the results of the regression analysis were 0.97 or more for systolic blood pressure, diastolic blood pressure, and mean blood pressure, indicating that there is a high correlation between the measured blood pressure and the estimated blood pressure. In addition, the P value of the main explanatory variable, ln(mNPV), was 0.05 or less, indicating that this explanatory variable has high discriminatory ability. In addition, the P value of the t-test between the measured and estimated values ​​was 0.97 or more, indicating that there was no significant difference between the two.

[0051] Although the wearable device that is the subject of this invention is not a medical device, the standard for medical device blood pressure monitors (JIS T 1115:2018) requires a correlation of 5mmHg ± 8mmHg, so it is possible to estimate the acceptable range of estimation accuracy based on the evaluation index.

[0052] The above results show that by using the blood pressure estimation calibration method and calibration assistance device in the wearable device of the present invention, it is possible to perform highly accurate blood pressure estimation tailored to a specific individual by referring to a medical device-certified wrist-type intravascular pressure gauge while the user is at rest, without imposing any particular exercise or psychological burden on the user.

[0053] In addition, by using the calibration method and calibration assistant device of the present invention, blood pressure can be estimated with high accuracy using regression coefficients obtained from an individual user even in other blood pressure estimation models having different explanatory variables. The explanatory variables of blood pressure-related signals may be selected from a group including normalized ejection volume mNPV, heart rate PR, pulse pressure, pulse wave transit time, pulse wave velocity, blood flow resistance index, blood flow index, vascular elasticity, vascular compliance, fluctuating components (AC components) and fixed components (DC components) that can be read from a PPG waveform, or the perfusion index (PI) or pulse-amplitude index (PI), which are the ratios of these. (System configuration for calibrating estimated blood pressure values ​​in wearable devices)

[0054] FIG. 15 shows an overall schematic diagram of an example system configuration for calibrating blood pressure estimates in a PPG wearable device. The wearable device 402 that acquires a PPG signal is composed of a host calculation unit 421, a sensor unit 422, and a device display unit 423. The sensor unit 422 further comprises a light emitting unit 441, a light receiving unit 442, a control unit 443, and a body signal acquisition unit 444. The host calculation unit 421 also comprises a start-up unit 431, a communication unit 432, a storage unit 433, and a signal processing unit 434. In the normal measurement mode, the signal acquired by the sensor unit 422 is sent to the signal processor 434, which displays the desired physical information, including a blood pressure estimate, on the device display unit 423 according to a pre-set algorithm.

[0055] In the calibration of the estimated blood pressure value, the wearable device 402 is put into calibration mode, and a blood pressure estimation model appropriate for the individual is selected using a dedicated application S111 on the information terminal S110. After acquiring the measurement value of the wrist-type blood pressure monitor and the wearable device signal obtained by the calibration method of the present invention, these are sent via a remote communication unit 452 to a Web server S117 consisting of an account management unit 491, a calculation unit 492, a data recording unit 493, a data transmission unit 494, and a data display unit 495, where a regression analysis is performed to calculate the regression coefficient and the accuracy index. The blood pressure estimation model, regression coefficients, and evaluation indexes obtained as a result of the above are transferred to the user's personal wearable device 402 via the remote communication unit 452, the information terminal S110, and the close-proximity communication unit 451, and then via the communication unit 432 of the wearable device 402, whereupon the calibration information in the memory unit 433 in the host calculation unit 421 is updated, and the calibration mode is terminated. After the updated calibration information is set in the wearable device 402, the calibration mode is returned to the measurement mode, and the PPG signal sent from the sensor unit 422 is again processed by the signal processing unit 434 using a predetermined algorithm, and an estimated blood pressure value is presented on the device display unit 423.

[0056] Furthermore, information on the user-specific blood pressure estimation model, regression coefficients, and accuracy index obtained in the above calibration process is stored in a calibration information recording section 483 linked to the account recording section 481 and physical condition recording section 482 of the database S116. In addition to the system configuration examples presented above, the regression analysis performed on the Web server S117 may be performed by the information terminal S110 processor or the wearable device 402 processor depending on the system configuration.

[0057] After the calibration process is completed, the user's personal blood pressure estimate can be measured in daily life using only the calibrated wearable device 402 in measurement mode. Also, the device is connected to an information terminal 461 such as a smartphone, tablet, or PC via USB or Bluetooth of the near-field communication unit 451 several times a week, preferably every day, and physical information data including the blood pressure estimate recorded and stored at regular intervals in daily life is stored in a personal account in the database S116 via the remote communication unit 452. Furthermore, the personal physical information including the estimated blood pressure value is output to the history display unit 464 or the report output unit 465.

[0058] In addition, the important thing about blood pressure estimation using the wearable device 402 that handles PPG signals alone is to stay still and keep the device at the same height as the heart, just like during calibration. By aligning the wearable device 402 with landmarks such as the nipple, the fourth rib, and the subscapular angle, it becomes possible to perform measurements of the same quality as a normal blood pressure monitor. In addition, it is known that the PPG signal changes significantly depending on the contact state between the skin and the PPG sensor and the load applied to the sensor. Therefore, it is desirable to reproduce the same measurement conditions as those used during calibration as much as possible. (Blood pressure estimation monitoring service provision method)

[0059] FIG. 16 shows an example of an overall configuration diagram of a physical condition monitoring service S100 including a blood pressure estimation monitoring service S101 according to one embodiment of the present invention.

[0060] The blood pressure estimation and monitoring service S101 of the present invention includes the following services. 1) Equipment provision service S102 2) Calibration Update Service S103 3) Record display service S104 4) Regular Report Service S105 5) Billing and settlement service S106

[0061] The equipment provision service S102 sells and rents a medical device certified wrist-type blood pressure monitor 401 that has a track record of calibrating blood pressure estimates in wearable devices, a wearable device 402 that can output PPG raw signal waveforms, an information terminal S110, a dedicated demo application S111, a calibration assistant device 300, etc. It also manages reservations, delivery, and return schedules for the above-mentioned rental equipment.

[0062] The proofreading update service S103 supports the user when carrying out proofreading work according to the proofreading flow chart 200 shown in FIG. In the task of periodically calibrating and updating the user-specific blood pressure estimation model and regression coefficients, a calibration and update service S103 may be provided in the following manner. 1) The calibration equipment required by the wearable device user is delivered to the user's home, and the service provider performs the calibration work on-site. 2) The calibration equipment required by the wearable device user is delivered to the user's home, and the user performs the calibration work themselves by following the enclosed instruction manual. 3) The calibration equipment required by the wearable device user is delivered to the user's home, and the user can view videos, illustrations, photos, and explanations posted on the Web and perform the calibration work themselves. 4) The calibration equipment required by wearable device users is delivered to their homes, and the calibration work is assisted through online interviews, with the users themselves performing the calibration work.

[0063] The record display service S104 can display physical information (activity level, number of steps, calories burned, heart rate, oxygen saturation, heart rate variability, respiratory rate, body temperature, bioimpedance, ECG signals, etc.) including blood pressure estimates that can be obtained by the wearable device on the device display unit 423 of the wearable device 402 and on the screen of the information terminal S110. In addition to the temporary recording results at the time of measurement, the trends of each piece of physical information over a certain period of time can be visualized on the information terminal S110 and the Web server S117 by calling up the daily periodic recording results, weekly, monthly, and yearly recording results that are recorded in a personal account in the database S116 when the information terminal S110 is accessed.

[0064] In the periodic report service S105, the trends of each piece of physical information for a certain period displayed in the record display service S104 can be summarized in a graph or table to create a periodic report. The user can also take the periodic report to a medical examination of their own volition, and use it as reference data when the doctor is not present.

[0065] The billing and settlement service S106 charges users for the dedicated wearable device, the dedicated app, the calibration work, and the physical information report creation service, and collects fees from the users. A support desk for the entire service is set up to handle problems and questions. [Explanation of symbols]

[0066] 100 Method and apparatus for calibrating estimated blood pressure in a wearable device, and method for providing estimated blood pressure monitoring service 101 Wearable Device Calibration 102 Implementation of calibration results 103 Physical Condition Record 104 Information terminal monitoring 105 Calibration data regularly updated and stored 200 Calibration Flowchart 201 Sitting in a forearm support device, resting 202 Align the horizontal rotation axis with the center of the heart 203 Medical wrist-type blood pressure monitor attached to one wrist 204 Wearable device attached to opposite wrist 205 Place your hands on the forearm support and relax 206 Forearm support height and angle adjustment 207 Calibration mode ON, blood pressure estimation model selected 208 Turn on the blood pressure monitor and wearable device and start acquiring data. 209 Number of data sets > explanatory variable + 1? 210 Regression analysis performed 211 Regression coefficient, accuracy index calculation 212 Update wearable device memory, save database, calibration mode OFF 213 Measurement mode ON, select two or more arbitrary height points, and check reproducibility 300 Calibration aids 301 Horizontal Rotation Axis 301a Horizontal rotation shaft lock hole 302 Horizontal Rotating Axis Holder 302a Horizontal Rotary Axis Holder Lock Pin 302b Horizontal rotation axis holder lock hole 303 Horizontal Rotating Axis Handle 304 Upper arm support 305 Upper arm support fixture 306 Horizontal Rotation Axis Movement Tool 307 Forearm holder 308 Forearm holder fixture 309 Chair 400 Heart 401 Wrist Sphygmomanometer 402 Wearable Devices 403 Heart Position Marker Stand 404 405 Major 406 Forearm Drop 407 Forearm Above Knee 408 Forearm Cross Above 4th Rib 409 Forearm Above Shoulder 410 Forearm above ear 411 Forearm Overhead 421 Host Calculation Unit 422 Sensor section 423 Device display section 431 Starting section 432 Communications Department 433 Storage section 434 Signal Processing Unit 441 Light emitting part 442 Light receiving part 443 Control Unit 444 PPG signal acquisition section 445 Body signal acquisition unit 451 Near Field Communication Unit 452 Remote Communications Department 461 Information terminal 464 History display section 465 Report Output Section 481 Accounts Record Division 482 Physical Information Recording Department 483 Calibration Information Record Section 491 Account Management Department 492 Arithmetic section 493 Data Recording Unit 494 Data Transmission Unit 495 Data display section S100 Physical Condition Monitoring Service S101 Blood pressure estimation monitoring service S102 Equipment provision services S103 Calibration Renewal Service S104 Record display service S105 Regular Report Service S106 Billing and settlement service S110 Information terminal S111 dedicated app S116 Database S117 Web Server

Claims

1. A blood pressure monitor, which is a medical device, and the wearable device to be calibrated are attached to the wrist or ankle of a user of the wearable device; a blood pressure-related signal acquisition step of acquiring blood pressure measurements of the sphygmomanometer and blood pressure-related signals of the wearable device while aligning the height of the part where the sphygmomanometer is attached and the height of the part where the wearable device is attached; determining a regression coefficient of a blood pressure estimation model of the wearable device from the blood pressure measurement values ​​and the blood pressure related signals obtained by performing the blood pressure related signal acquisition step multiple times while changing the height of the attachment part from the heart; Setting a user-specific blood pressure estimation model based on the determined regression coefficients in the wearable device; Characterized in that it has A method for calibrating blood pressure estimates in wearable devices.

2. the blood pressure-related signal acquiring step acquires a plurality of blood pressure measurement values ​​and blood pressure-related signals from the same user while the user is in a resting state, without imposing any exercise load or psychological load on the user, and only by changing the height of the attachment part from the heart. A method for calibrating a blood pressure estimate in a wearable device according to claim 1.

3. The blood pressure-related signal acquisition step includes the step of: having the user wear a blood pressure monitor, which is a medical device, on one wrist or ankle and a wearable device to be calibrated on the other wrist or ankle while in a sitting, standing or lying position; The peripheral blood pressure due to the hydrostatic pressure of blood is arbitrarily changed by raising and lowering the upper arm or forearm, or the thigh or lower leg, thereby obtaining a plurality of blood pressure measurements and blood pressure-related signals from the same user. A method for calibrating a blood pressure estimate in a wearable device according to claim 1.

4. The step of determining the regression coefficients of the blood pressure estimation model of the wearable device includes a step of determining the regression coefficients of the blood pressure estimation model of the wearable device from a plurality of blood pressure measurement values ​​and blood pressure related signals obtained from the same user by performing the blood pressure related signal acquisition step a plurality of times using each part of the body whose height from the heart is fixed (buttocks, knees, nipples, ribs, shoulders, ears, head) as a landmark. A method for calibrating a blood pressure estimate in a wearable device according to claim 1.

5. A calibration assistance device that can be used in the method for calibrating a blood pressure estimated value in a wearable device according to claim 1, The forearm, on which the medical device, a blood pressure monitor, and the wearable device to be calibrated, are attached, is held at the same height, and both forearms are raised and lowered simultaneously. By simply changing the height from the heart and changing the peripheral blood pressure due to blood hydrostatic pressure, At rest, without any physical or psychological stress, A device that allows for the acquisition of multiple blood pressure measurements and blood pressure related signals from the same user. Calibration aid.

6. The calibration assistant device has a horizontal rotation axis that can be moved up and down to match the height of the user's heart (the fourth rib or the inferior angle of the scapula); The position and angle of the forearm holder can be adjusted, and by holding the user's forearm from below, Without straining your arms to maintain height, you can sit still. A device that allows for the acquisition of multiple blood pressure measurements and blood pressure related signals from the same user. The calibration assistant device according to claim 5.

7. The calibration assistant device fixes the angle of the upper arm support at a multiple of 12.9 mm, which is the blood column height equivalent to 1 mmHg, according to the height of both arms, so that the vertical height relative to the heart position can be easily converted into blood pressure units. The lock hole is provided with a rotating holder and a horizontal rotating shaft. The calibration assistant device according to claim 5.

8. A method for providing a blood pressure estimation monitoring service, comprising a step of setting the user-specific blood pressure estimation model according to claim 1 in the wearable device, Through an information terminal with a dedicated app installed, The regression coefficients and blood pressure estimation model specific to a user determined by the method for calibrating a blood pressure estimate in a wearable device according to claim 1 are The information is stored on the wearable device worn by the user and saved in a personal account in a database. Your daily blood pressure estimates and other physical information history It can be displayed on the web or on an information terminal upon request. A method for providing a blood pressure estimation monitoring service.

9. The blood pressure estimation and monitoring service providing method relates to a method for calibrating a blood pressure estimated value in a wearable device according to claim 1. (1) Equipment provision services; (2) Proofreading and updating services; (3) Record display service; (4) Regular report service; (5) Billing and settlement services; It is composed of Connect the wearable device to an information terminal on which a dedicated application is installed; The main feature is that it includes a calibration update service that periodically updates the user's specific blood pressure estimation model. The blood pressure estimation monitoring service providing method according to claim 8.

10. The said proofreading update service: For a user of the wearable device, The present invention provides a method of providing explanations by lending the proofreading assistance device, recording and viewing proofreading methods on a web server, viewing on YouTube (registered trademark), explaining the method through a paper manual, and interacting through a video conference. The blood pressure estimation monitoring service providing method according to claim 8.

11. The regular report service visualizes blood pressure estimates and physical information history over a certain period of time using a calibrated wearable device worn by the user. Users can take the device with them to medical examinations of their own volition and refer to it as reference data when a doctor is not present. The blood pressure estimation monitoring service providing method according to claim 8.

Images