Sphygmomanometer, blood pressure measurement method, blood pressure measurement program, learning model construction method, and learning model construction program

Abstract

Provided is a sphygmomanometer capable of achieving high-precision blood pressure measurement even with a small amount of teacher data. This sphygmomanometer is provided with: a cuff wound around the arm of a person to be measured; a control unit for controlling the pressure of the cuff on the arm; a detection unit that detects a cuff pressure applied to the cuff from the arm; a measuring unit for measuring the blood pressure of the subject by oscillography using the time-series change in the cuff pressure; a correction model, which is constructed by machine learning for a plurality of subjects, and which has, as an explanatory variable, waveform information relating to a time-series change in the cuff pressure, and which has, as a target variable, an error in the blood pressure measured by the measurement unit with respect to the blood pressure used as a true value; a correction unit that corrects the blood pressure measured by the measurement unit on the basis of an error output by the correction model by inputting, to the correction model, the information of the waveform relating to the time-series change in the cuff pressure detected by the detection unit; and an output unit that outputs the blood pressure corrected by the correction unit.

Description

Technical Field

[0001] This invention relates to a blood pressure monitor, a blood pressure measurement method, a blood pressure measurement procedure, a method for constructing a learning model, and a procedure for constructing a learning model. Background Technology

[0002] Blood pressure measurement is used to assess health status. Patent Document 1 describes a technique that stores the correlation between a subject's blood pressure measured using a cuff and the subject's blood pressure measured by auscultation, and uses the stored correlation to correct the subject's blood pressure. Furthermore, Patent Document 2 describes a technique for measuring blood pressure using photoplethysmography (PPG).

[0003] Existing technical documents

[0004] Patent documents

[0005] Patent Document 1: Japanese Patent Application Publication No. 5-309073

[0006] Patent Document 2: Japanese Patent Publication No. 2022-516820

[0007] In oscillometric blood pressure measurement, after inflating a cuff wrapped around the subject's upper arm or wrist, the pressure vibration superimposed on the cuff pressure is detected while gradually depressurizing, and blood pressure is measured based on the detected pressure vibration. However, depending on the subject's body movement, the method of cuff wrapping, and the position of the cuff, noise can sometimes be introduced into the detected pressure vibration. This noise can potentially reduce the accuracy of oscillometric blood pressure measurement. Machine learning has been considered to control for this reduction in accuracy, but collecting a large amount of teacher data that can suppress the effects of noise is not easy.

[0008] One aspect of the disclosed technology is to provide a blood pressure monitor, blood pressure measurement method, blood pressure measurement procedure, learning model construction method, and learning model construction procedure that enable high-precision blood pressure measurement even with limited teacher data. Summary of the Invention

[0009] One aspect of the disclosed technology is illustrated by the following blood pressure monitor. This blood pressure monitor includes: a cuff wrapped around the arm of a subject; a control unit for controlling the pressure of the cuff on the arm; a detection unit for detecting the cuff pressure received by the cuff from the arm; a measurement unit for measuring the subject's blood pressure using an oscillometric method based on the temporal variation of the cuff pressure; a calibration model constructed for the subject using machine learning, wherein the machine learning uses information about the waveform related to the temporal variation of the cuff pressure as an explanatory variable and uses the error of the blood pressure measured by the measurement unit relative to the blood pressure used as the true value as a target variable; a calibration unit for correcting the blood pressure measured by the measurement unit based on the error output by the calibration model by inputting the information about the waveform related to the temporal variation of the cuff pressure detected by the detection unit into the calibration model; and an output unit for outputting the blood pressure corrected by the calibration unit.

[0010] According to the aforementioned blood pressure monitor, the aforementioned error is used to correct the blood pressure measured by the oscillometric method, thus improving the accuracy of blood pressure measurement. Furthermore, in the aforementioned blood pressure monitor, the error between the blood pressure measured by the oscillometric method and the blood pressure used as the true value is set as the target variable. The variation range of the aforementioned error is considered to be approximately ±10 mmHg, and the variation range of the highest blood pressure in a person is considered to be approximately 80 mmHg to 200 mmHg. In machine learning, the amount of teacher data tends to increase proportionally to the width of the variation range of the target variable. In other words, by reducing the variation range of the target variable, the amount of teacher data used in learning can be reduced. The aforementioned blood pressure monitor sets the aforementioned error, whose variation range is narrower than the highest blood pressure, as the target variable, thus reducing the amount of teacher data used in learning. That is, this blood pressure monitor can achieve high-accuracy blood pressure measurement even with less teacher data. Additionally, the blood pressure used as the true value can be the blood pressure measured by auscultation.

[0011] Here, the information described above as the waveform can include, for example, information related to the cuff pressure. Alternatively, the measurement unit may acquire an oscilloscope waveform (OMW) based on the temporal variation of the cuff pressure, and the information described above for the waveform may include OMW information representing the waveform of the OMW.

[0012] Alternatively, the measurement unit may acquire the oscilloscope waveform envelope (OMWE) based on the OMW, and the waveform information may include OMWE information representing the waveform of the OMWE.

[0013] Alternatively, the blood pressure monitor may correct the blood pressure measured by the measuring unit based on the error output by the correction model obtained by inputting the temporal variation of the cuff pressure and the OMW information into the correction model. Furthermore, the blood pressure monitor may also correct the blood pressure measured by the measuring unit based on the error output by the correction model obtained by inputting the temporal variation of the cuff pressure and the OMW information into the correction model.

[0014] The disclosed technology can also be mastered from the aspects of blood pressure measurement methods, blood pressure measurement procedures, learning model construction methods, and learning model construction procedures.

[0015] According to the disclosed technology, even with a small amount of teacher data, high-precision blood pressure measurement can be achieved. Attached Figure Description

[0016] Figure 1 This is a diagram illustrating an example of the blood pressure monitor according to this embodiment.

[0017] Figure 2 This is a schematic diagram illustrating the correction of blood pressure based on the correction unit involved in the embodiment.

[0018] Figure 3 This is a diagram schematically illustrating the construction of the correction model involved in the implementation method.

[0019] Figure 4 This is a diagram illustrating an example of the processing flow of the blood pressure monitor according to this embodiment.

[0020] Figure 5 This is a graph illustrating the verification results of the accuracy of blood pressure measurement.

[0021] Figure 6 This is a diagram illustrating one example of the sphygmomanometer involved in the first variation. Detailed Implementation

[0022] <Application Examples>

[0023] An application example of the present invention will be described. The blood pressure monitor 100 involved in the application example is a device for measuring the blood pressure of a person by wrapping a cuff 120 around the arm of the person being measured. The cuff 120 may be wrapped around, for example, the upper arm or wrist of the person being measured. That is, the blood pressure monitor 100 may also be a wrist-type blood pressure monitor. In the blood pressure monitor 100, the measuring unit 114 measures the blood pressure of the person being measured using an oscillometric method. The measuring unit 114 inputs the measured blood pressure of the person being measured to the calibration unit 115.

[0024] The calibration unit 115 uses a calibration model 116 to correct the blood pressure measured by the measurement unit 114. The calibration model 116 is a learning model constructed using machine learning. This machine learning uses waveform information related to the temporal variation of cuff pressure 120 as an explanatory variable, and the error of the blood pressure measured by the measurement unit 114 relative to the blood pressure used as the true value as the target variable. Examples of waveform information related to the temporal variation of cuff pressure include information from oscillometric waveforms (OMW) and waveform information from the oscillometric waveform envelope (OMWE). The measurement unit 114 measures blood pressure, for example, using oscillometric methods. Alternatively, the blood pressure used as the true value can be, for example, blood pressure measured by auscultation. Furthermore, the machine learning includes deep learning.

[0025] If the calibration model 116 is input with waveform information related to the time-series changes in cuff pressure, it outputs the error of the blood pressure measured by the oscilloscope method relative to the blood pressure used as the true value. The calibration unit 115 uses the error output from the calibration model 116 to correct the blood pressure measured by the measuring unit 114. In the blood pressure monitor 100, through such correction, the blood pressure measured by the measuring unit 114 can be made as close as possible to the blood pressure actually measured by the auscultatory method.

[0026] Here, as described above, the calibration model 116 in this application example sets the error of the blood pressure measured by oscillometric method relative to the blood pressure used as the true value as the target variable. This error is considered to be approximately in the range of ±10 mmHg. On the other hand, the highest blood pressure value for a person is approximately in the range of 80 mmHg to 200 mmHg.

[0027] That is, if blood pressure is set as the target variable in the calibration model 116, the range of variation of the target variable is approximately 10 times larger compared to the case where the error is set as the target variable. In machine learning, the amount of teacher data tends to increase proportionally with the width of the range of variation of the target variable. Therefore, in this application example where the error is set as the target variable, the amount of teacher data used in constructing the calibration model 116 can be reduced compared to the case where blood pressure is set as the target variable. In other words, according to this application example, high-accuracy blood pressure measurement can be achieved even with a smaller amount of teacher data.

[0028] <Implementation Method>

[0029] Next, the implementation method will be described. Figure 1This is a diagram illustrating an example of the blood pressure monitor 100 according to this embodiment. The blood pressure monitor 100 includes a main body 110 and a cuff 120. The blood pressure monitor 100 measures the blood pressure of the subject by wrapping the cuff 120 around the upper arm or wrist (hereinafter also referred to as the arm) of the subject.

[0030] The cuff 120 is formed in the shape of a strip and has a pouch for supplying air to the inside. If air is supplied to the pouch inside the cuff 120 while the cuff 120 is wrapped around the arm of the person being measured, the pressure of the cuff on the arm increases. Conversely, if air is drawn in from the pouch inside the cuff 120 while the cuff 120 is wrapped around the arm of the person being measured, the pressure of the cuff on the arm decreases.

[0031] The main body 110 is equipped with a processor and a storage unit. The processor executes the program stored in the storage unit to implement various processing units such as the start switch 111, control unit 112, pressure sensor 113, measurement unit 114, calibration unit 115, calibration model 116, and output unit 117. In other words, the blood pressure monitor 100 can be regarded as a computer with a cuff 120. In addition, each processing unit such as the start switch 111, control unit 112, pressure sensor 113, measurement unit 114, calibration unit 115, calibration model 116, and output unit 117 can also be a dedicated hardware circuit. In the blood pressure monitor 100, the blood pressure monitor 100 is started by turning on the start switch 111, and blood pressure measurement begins.

[0032] The control unit 112 controls the pressure of the cuff 120 on the subject's arm by supplying air to the cuff 120 and drawing air from the cuff 120. If the start switch 111 is turned on, the control unit 112 begins to control the pressure of the cuff 120.

[0033] Pressure sensor 113 is a sensor that detects the pressure (cuff pressure) of cuff 120. Pressure sensor 113 outputs the detected cuff pressure to measuring unit 114. Cuff pressure includes, for example, the pressure applied by cuff 120 to the arm and the pressure vibration caused by the pulsation of blood flow in the arteries of the arm.

[0034] The measuring unit 114 performs oscillometric blood pressure measurement based on the pressure detected by the pressure sensor 113. If the pressure of the cuff 120 is higher than the highest blood pressure, the artery is blocked, and no pressure vibration caused by the pulsation of blood flow occurs. On the other hand, if the pressure of the cuff 120 is lower than the highest blood pressure, the artery begins to open, and the pressure vibration caused by the pulsation of blood flow is superimposed on the pressure of the cuff 120. Furthermore, if the pressure of the cuff 120 decreases to below the lowest blood pressure, the artery inside the arm moves by slightly expanding from the open state, without generating pressure vibration.

[0035] After the pressure is increased to a level higher than the highest blood pressure, the measuring unit 114 acquires the time-series change of the cuff pressure of the cuff 120 as it gradually decreases to a level lower than the lowest blood pressure from the pressure sensor 113. From the time-series change of the cuff pressure acquired by the pressure sensor 113, the measuring unit 114 extracts the time-series change of pressure oscillations caused by the pulsation of blood flow within the artery. The extracted time-series change of pressure oscillations is the OMW (Original Motion Wave). The measuring unit 114 also extracts the OMWE (Original Motion Wave Ending) from the OMW. The measuring unit 114 uses the characteristic quantity extracted from the OMW (hereinafter also referred to as the OMW characteristic quantity) and the characteristic quantity extracted from the OMWE (hereinafter also referred to as the OMWE characteristic quantity) to measure the blood pressure of the subject. Alternatively, the measuring unit 114 can also use either the OMW characteristic quantity or the OMWE characteristic quantity to measure blood pressure using an oscilloscope method. Furthermore, the measuring unit 114 outputs information about the waveform related to the time-series change of the cuff pressure to the correction unit 115. As waveform information related to the time-series changes in cuff pressure, examples include OMW information representing the waveform of OMW and OMWE information representing the waveform of OMWE. Hereinafter, we will explain, as an example, the use of at least one of OMW information and OMWE information as waveform information related to the time-series changes in cuff pressure.

[0036] The OMW and OMWE information output by the measurement unit 114 can be exemplified by the following information (1) to (28). The information in (1) to (14) can be said to be related to the waveform shape of OMW and OMWE. In addition, the information in (15) to (20) can be said to be related to the cuff pressure.

[0037] (1) Minimum value, average value, minimum value.

[0038] (2) The time axis is the value at time point X% (X is a value from 0 to 100).

[0039] (3) The value becomes the largest at time.

[0040] (4) The area of ​​the region enclosed by the waveforms and time axis of OMW and OMWE.

[0041] (5) Kurtosis and skewness of OMW and OMWE waveforms.

[0042] (6) The number of peaks in the waveforms of OMW and OMWE.

[0043] (7) The complexity of the waveforms of OMW and OMWE (e.g., the sum of squares of the differences between the time series data).

[0044] (8) The number of points that exceed the average value of the waveforms up to a predetermined time point (e.g., from the start to 1 / 3 of the time point).

[0045] (9) Divide the waveforms of OMW and OMWE into a predetermined number (e.g., “4”) of regions, and the slope of each region after division.

[0046] (10) Divide the time before the peak of the waveform of OMW and OMWE by the time after the peak.

[0047] (11) The height variation ratio of the waveforms of OMW and OWME.

[0048] (12) Divide the waveforms of OMW and OMWE into a predetermined number (e.g., “4”) of regions, and for each region after division, the ratio of the area of ​​the waveform to the area of ​​the circumscribed rectangle.

[0049] (13) The value obtained by subtracting the peak time of the lower envelope of OMW from the peak time of the upper envelope of OMW.

[0050] (14) The ratio of the area of ​​the upper envelope of OMW to the area of ​​the lower envelope of OMW.

[0051] (15) Cuff pressure at the peak of OMWE.

[0052] (16) The first derivative of OMWE becomes the cuff pressure at the point of maximum and the cuff pressure at the point of minimum.

[0053] (17) The cuff pressure at the point where the OMWE peak is N times lower than the peak (N is an integer greater than 0 and less than 1).

[0054] (18) The cuff pressure at the point where the peak of OMWE drops by N times (N is an integer greater than 0 and less than 1) after the peak.

[0055] (19) The time from the maximum value of the first derivative of OMWE to the peak value.

[0056] (20) The time from the peak of OMWE to the minimum of the first derivative.

[0057] (21) Length of time of OMWE.

[0058] (22) The value obtained by dividing the time up to the peak of OMWE by the length of OMWE.

[0059] (23) The position of the average "-σ" when performing Gaussian fitting on OMWE.

[0060] (24) The position of the average "+σ" when performing Gaussian fitting on OMWE.

[0061] (25) The value of OMWE at the position of the average "-σ" when performing Gaussian fitting on OMWE.

[0062] (26) The value of OMWE at the position of the average "+σ" when performing Gaussian fitting on OMWE.

[0063] (27) The maximum slope of OMWE.

[0064] (28) The minimum slope of OMWE.

[0065] The calibration unit 115 calibrates the blood pressure measured by the measurement unit 114. The calibration unit 115 inputs the OMW and OMWE information extracted by the measurement unit 114 into the calibration model 116. The calibration model 116 is a learning model; if OMW and OMWE information are input, it outputs the error between the blood pressure measured by the measurement unit 114 and the blood pressure measured by auscultation as a correction value. The calibration model 116 is stored, for example, in a storage unit housed in the main body 110. The calibration unit 115 uses the correction value obtained from the calibration model 116 to correct the blood pressure measured by the measurement unit 114.

[0066] Figure 2 This is a schematic diagram illustrating the correction of blood pressure based on the correction unit 115 according to the embodiment. (Refer to...) Figure 2 The correction of blood pressure based on the correction unit 115 will be explained. The measurement unit 114 extracts the OMW (Obstruction Waveform) of the subject C1 (step M1). The measurement unit 114 extracts OMW information representing the waveform of the OMW from the extracted OMW (step M2). Furthermore, the measurement unit 114 extracts the OMWE (Optical Waveform Ending) from the OMW (step M3), and measures the blood pressure of the subject C1 using the oscillometric method based on the extracted OMWE (step M4). Furthermore, the measurement unit 114 extracts OMWE information representing the waveform of the OMWE from the OMWE (step M5).

[0067] The calibration unit 115 inputs the OMW information extracted in step M2 and the OMWE information extracted in step M5 into the calibration model 116, obtains the error between the blood pressure measured by the oscillometric method and the blood pressure actually measured by the auscultatory method, and corrects the blood pressure measured in step M4 based on the obtained error (step M6). The output unit 117 outputs the corrected blood pressure in step M6 to a display or the like (step M7).

[0068] Here, the construction of the calibration model 116 used in the calibration based on the calibration unit 115 will be briefly described. The calibration model 116 is constructed by machine learning, which sets OMW information and OMWE information as explanatory variables and sets the error of the blood pressure measured by the measurement unit 114 relative to the blood pressure measured by auscultation as the target variable.

[0069] Figure 3This is a diagram schematically illustrating the construction of the correction model 116 involved in the implementation method. (Refer to...) Figure 3 The construction of the calibration model 116 will be described below. In constructing the calibration model 116, multiple subjects (E1, E2, ..., EN) are prepared. In this embodiment, multiple subjects are prepared, but a single subject may also be used. The measurement unit 114 extracts the OMW of subject E1 (step P1). The measurement unit 114 extracts OMW information representing the waveform of the OMW from the extracted OMW (step P2). Furthermore, the measurement unit 114 extracts the OMWE from the OMW (step P3), and measures the blood pressure of subject E1 using oscillometric measurement based on the extracted OMWE (step P4). Then, the measurement unit 114 extracts OMWE information representing the waveform of the OMWE from the OMWE (step P5).

[0070] In addition, physician D1 measures subject E1's blood pressure using auscultation (step P6). The error between the blood pressure measured in step P4 and the blood pressure measured in step P6 is calculated (step P7). Machine learning is then performed, using the OMW information extracted in step P2 and the OMWE information extracted in step P5 as explanatory variables, and the error calculated in step P7 as the target variable (step P8). Through machine learning in step P8, a correction model 116 is constructed.

[0071] Steps P1 to P8 are also performed for each of the subjects E2 to EN. That is, the calibration model 116 is not a learning model that improves the accuracy of blood pressure measurement for a single subject, but rather a model that can improve the accuracy of oscillometric blood pressure measurement by using measurement results from multiple subjects.

[0072] Return to Figure 1 The output unit 117 outputs the blood pressure corrected by the correction model 116. The output unit 117 can be, for example, a Liquid Crystal Display (LCD), a Plasma Display Panel (PDP), an Electroluminescence (EL) panel, or an Organic EL panel. The output unit 117 can also be a printer or a speaker.

[0073] <Processing Flow>

[0074] Figure 4 This is a diagram illustrating an example of the processing flow of the blood pressure monitor 100 according to this embodiment. Hereinafter, refer to... Figure 4 An example of the processing procedure for a blood pressure monitor 100 will be explained.

[0075] In step S1, the blood pressure monitor 100 is activated by turning on the start switch 111. In step S2, the control unit 112 delivers air to the cuff 120 to apply pressure to the arm of the person being measured. After the control unit 112 pressurizes the cuff 120 to a pressure higher than the maximum blood pressure, the pressure of the cuff 120 is gradually reduced.

[0076] In step S3, the measuring unit 114 extracts the OMW (Occupancy Mean Wavelength) based on the cuff pressure detected by the pressure sensor 113, and extracts OMW information from the extracted OMW. Furthermore, the measuring unit 114 extracts the OMWE (Occupancy Mean Wavelength Effect) from the OMW, and extracts OMWE information. The measuring unit 114 measures the subject's blood pressure using an oscillometric method of OMWE.

[0077] In step S4, the measurement unit 114 determines whether the blood pressure measurement of the subject is complete. For example, the measurement unit 114 may determine that the blood pressure measurement is complete if both the subject's highest and lowest blood pressure have been measured. If the measurement is complete (yes in step S4), the process proceeds to step S5. If the measurement is not complete (no in step S4), the process proceeds to step S2.

[0078] In step S5, the control unit 112 expels air from the cuff 120. In step S6, the calibration unit 115 inputs the OMW and OMWE information extracted in step S3 into the calibration model 116 to obtain the error between the blood pressure measured by the oscillometric method and the blood pressure measured by the auscultatory method. The calibration unit 115 uses the error obtained from the calibration model 116 to correct the blood pressure measured in step S3.

[0079] In step S7, the output unit 117 outputs the blood pressure corrected in step S6.

[0080] <Verification>

[0081] The blood pressure measurement accuracy of the sphygmomanometer 100 described above has been verified, and therefore will be described. Figure 5 This is a graph illustrating the verification results of blood pressure measurement accuracy. Figure 5 In order to compare with the implementation method, an example of not correcting the blood pressure (lowest blood pressure: DBP, highest blood pressure: SBP) measured by oscillometric method without using a correction model (uncorrected) and a comparative example of constructing a correction model with blood pressure as the target variable are prepared. Moreover, for each of the comparative example and the implementation method, as input data to the correction model, an example using OMWE information (Comparative Example 1, Implementation Method 1) and an example using both OMWE information and OME information are prepared (Comparative Example 2, Implementation Method 2).

[0082] In this validation, blood pressure was measured for each of multiple subjects, and the results were obtained. The obtained measurement results were evaluated based on the standard deviation of error (SDE).

[0083] Reference Figure 5 "Effective Method 1" and "Effective Method 2" show higher accuracy than "No Correction", "Comparative Example 1", and "Comparative Example 2". Furthermore, refer to... Figure 5 As input data to the calibration model, whether using OMWE information or both OMW information and OMWE information, "Implementation 1" and "Implementation 2" show high accuracy compared to "no calibration", "Comparative Example 1" and "Comparative Example 2".

[0084] <Effects of the Implementation Method>

[0085] According to this embodiment, since the blood pressure measured by the measuring unit 114 is corrected by the correction unit 115, the accuracy of blood pressure measurement can be improved as much as possible. Here, the correction model 116 sets the error between the blood pressure measured by the oscillometric method and the blood pressure actually measured by the auscultatory method as the target variable. As also explained in the application example, the variation range of the error is about 1 / 10 of the variation range of the blood pressure. The amount of teacher data tends to increase proportionally with the width of the variation range of the target variable. Therefore, according to this embodiment, which sets the error as the target variable, the amount of teacher data used in the construction of the correction model 116 can be reduced compared to the case where blood pressure is set as the target variable. That is, according to this embodiment, higher accuracy blood pressure measurement can be achieved with less teacher data.

[0086] Furthermore, in this embodiment, noise caused by the subject's body movement, the method of wrapping the cuff, and the position of the wrapped cuff are also included in the OMW and OMWE information. Here, the correction model 116 constructs the OMW and OMWE information as explanatory variables. Therefore, the correction model 116 takes into account such noise and can output the error between the blood pressure measured by the oscillometric method and the blood pressure measured by the auscultatory method. That is, according to this embodiment, blood pressure measurement based on the oscillometric method can be performed with higher accuracy.

[0087] In blood pressure measurements using cuffs, it is difficult to collect multiple blood pressure data points measured by the oscillometric method, which form the basis for calculating the aforementioned errors, due to the pressure applied to the body. In this embodiment, by setting the aforementioned errors as a target variable, the range of variation for the target variable can be narrowed, thereby suppressing the influence of deviations in the target variable. Therefore, even if the measured blood pressure data is biased due to the limited number of blood pressure data points measured by the oscillometric method, blood pressure measurements can be performed with higher accuracy.

[0088] <First Variation>

[0089] In the embodiments described above, the correction model 116 is stored in the storage unit within the main body 110. However, the correction model 116 may also be stored outside the storage unit within the main body 110. Figure 6 This is a diagram showing an example of the blood pressure monitor 100A according to the first modification. The blood pressure monitor 100A according to the first modification differs from the blood pressure monitor 100 according to the embodiment in that the calibration model 116 is not stored in the storage unit within the main body 110, but is stored in the server 200 connected to the main body 110 via the computer network N1, and a calibration unit 115A is provided instead of the calibration unit 115.

[0090] Computer network N1 is a network that interconnects information processing devices. Computer network N1 can be wired or wireless. An example of computer network N1 is the Internet.

[0091] The calibration unit 115A accesses the calibration model 116 stored in the server 200 via the computer network N1. That is, the calibration unit 115A sends the extracted OMW information and OMWE information to the server 200 via the computer network N1.

[0092] Server 200 inputs the OMW and OMWE information received from sphygmomanometer 100A into calibration model 116 to obtain the error between the blood pressure measured by oscillometric method and the blood pressure measured by auscultation method. Server 200 sends this error to main unit 110 via computer network N1. Calibration unit 115A uses the error received from server 200 to correct the blood pressure measured by measurement unit 114.

[0093] According to the first variation, the server 200 can perform computational processing that uses the correction model 116 to calculate errors. Therefore, the computational load within the main unit 110 can be reduced.

[0094] <Other Variations>

[0095] In the embodiments described above, refer to Figure 4The pressurized blood pressure measurement method has been described, but the blood pressure monitor 100 described in this embodiment can also measure blood pressure using the depressurized method.

[0096] The above-disclosed implementation methods and variations can also be combined separately.

[0097] <Computer-readable recording media>

[0098] An information processing program that enables a computer and other machines or devices (hereinafter referred to as computers, etc.) to perform any of the aforementioned functions can be recorded on a computer-readable recording medium. Furthermore, by causing the computer, etc., to read and execute the program on the recording medium, its functions can be provided.

[0099] Here, a computer-readable recording medium refers to a recording medium capable of storing and reading information such as data and programs through electrical, magnetic, optical, mechanical, or chemical processes. Examples of such recording media that can be removed from a computer include floppy disks, optical disks, compact disc read-only memory (CD-ROM), compact disc-recordable (CD-R), compact disc-rewriteable (CD-RW), digital versatile discs (DVD), Blu-ray discs (BD), digital audio tapes (DAT), 8mm magnetic tape, flash memory, external hard disk drives, and solid-state drives (SSDs). Additionally, recording media fixed to a computer include internal hard disk drives, SSDs, and ROMs.

[0100] <Appendix 1> A blood pressure monitor (100) comprising:

[0101] Cuff (120), wrapped around the arm of the person being measured;

[0102] The control unit (112) controls the pressure of the cuff (120) on the arm;

[0103] The detection unit (113) detects the cuff pressure (120) received by the arm from the cuff;

[0104] The measuring unit (114) measures the blood pressure of the subject by means of oscillometric method using the temporal variation of the cuff pressure;

[0105] The calibration model (116) is constructed for the subject through machine learning, wherein the machine learning takes information about the waveform related to the temporal change of the cuff pressure as the explanatory variable and the error of the blood pressure measured by the measuring unit relative to the blood pressure used as the true value as the target variable.

[0106] The correction unit (115) corrects the blood pressure measured by the measurement unit (114) based on the error output by the correction model (116) by inputting information about the waveform related to the temporal change of the cuff pressure detected by the detection unit (113) into the correction model (116); and

[0107] The output unit (117) outputs the blood pressure after it has been corrected by the correction unit (115).

[0108] <Appendix 2> According to the sphygmomanometer (100) described in Appendix 1, wherein,

[0109] The blood pressure used as the true value is the blood pressure measured by auscultation.

[0110] <Appendix 3> According to the sphygmomanometer (100) described in Appendix 1, wherein,

[0111] The waveform information includes information related to the cuff pressure.

[0112] <Appendix 4> According to any one of Appendices 1 to 3, the sphygmomanometer (100) wherein,

[0113] The measuring unit (114) acquires an oscilloscope waveform (OMW) based on the time-series change of the cuff pressure.

[0114] The waveform information includes OMW information representing the waveform of the OMW.

[0115] <Appendix 5> According to any one of Appendices 1 to 4, the sphygmomanometer (100) wherein,

[0116] The measurement unit (114) acquires the oscilloscope waveform envelope (OMWE) based on the OMW.

[0117] The waveform information includes OMWE information representing the waveform of the OMWE.

[0118] <Appendix 6> According to the sphygmomanometer (100) described in Appendix 4, wherein,

[0119] The correction unit (115) corrects the blood pressure measured by the measurement unit (114) based on the error output by the correction model (116) by inputting the temporal variation of the cuff pressure and the OMW information into the correction model (116).

[0120] <Appendix 7> According to the sphygmomanometer (100) described in Appendix 5, wherein,

[0121] The correction unit (115) corrects the blood pressure measured by the measurement unit (114) based on the error output by the correction model (116) by inputting the temporal variation of the cuff pressure and the OMWE information into the correction model (116).

[0122] Appendix 8: A blood pressure measurement method, executed by computer:

[0123] The control step involves controlling the pressure of the cuff (120) wrapped around the arm of the person being measured on the arm;

[0124] The detection step involves detecting the cuff pressure (120) exerted on the arm by the cuff.

[0125] The measurement procedure involves measuring the subject's blood pressure using an oscillometric method by measuring the temporal variation of the cuff pressure.

[0126] A correction step, based on the error output by the correction model (116) constructed for the subject through machine learning, corrects the blood pressure measured in the measurement step by inputting information of the waveform related to the temporal variation of the cuff pressure detected in the detection step into the correction model. The machine learning uses information of the waveform related to the temporal variation of the cuff pressure as an explanatory variable and the error of the blood pressure measured in the measurement step relative to the blood pressure used as the true value as a target variable.

[0127] The output step outputs the blood pressure after correction in the correction step.

[0128] Appendix 9: A blood pressure measurement program that is executed by a computer:

[0129] The control step involves controlling the pressure of the cuff (120) wrapped around the arm of the person being measured on the arm;

[0130] The detection step involves detecting the cuff pressure (120) exerted on the arm by the cuff.

[0131] The measurement procedure involves measuring the subject's blood pressure using an oscillometric method by measuring the temporal variation of the cuff pressure.

[0132] A correction step, based on the error output by the correction model (116) constructed for the subject through machine learning, corrects the blood pressure measured in the measurement step by inputting information of the waveform related to the temporal variation of the cuff pressure detected in the detection step into the correction model. The machine learning uses information of the waveform related to the temporal variation of the cuff pressure as an explanatory variable and the error of the blood pressure measured in the measurement step relative to the blood pressure used as the true value as a target variable.

[0133] The output step outputs the blood pressure after correction in the correction step.

[0134] <Appendix 10> A method for generating a learning model, executed by a computer:

[0135] The acceptance step accepts the input of the error between the blood pressure measured by oscillometric method and the blood pressure used as the true value; and

[0136] The generation step involves generating a learning model through machine learning, where the machine learning sets information about the waveform related to the temporal variation of the cuff pressure of the cuff wrapped around the arm of the person being measured as the explanatory variable, and sets the error as the target variable.

[0137] <Appendix 11> A procedure for generating a learning model, which a computer executes:

[0138] The acceptance step accepts the input of the error between the blood pressure measured by oscillometric method and the blood pressure used as the true value; and

[0139] The generation step involves generating a learning model through machine learning, where the machine learning sets information about the waveform related to the temporal variation of the cuff pressure of the cuff wrapped around the arm of the person being measured as the explanatory variable, and sets the error as the target variable.

[0140] Explanation of reference numerals in the attached figures

[0141] 100: Blood pressure monitor, 100A: Blood pressure monitor, 110: Main body, 111: Start switch, 112: Control unit, 113: Pressure sensor, 114: Measurement unit, 115: Calibration unit, 115A: Calibration unit, 116: Calibration model, 117: Output unit, 120: Cuff, 200: Server, C1: Subject, D1: Physician, E1: Subject, N1: Computer network.

Claims

1. A blood pressure monitor, comprising: A cuff is wrapped around the arm of the person being measured. The control unit controls the pressure of the cuff on the arm; The detection unit detects the cuff pressure exerted on the arm by the cuff. The measuring unit measures the subject's blood pressure using the temporal variation of the cuff pressure via oscillometric method; The calibration model is constructed for the subject through machine learning, wherein the machine learning takes information about the waveform related to the temporal variation of the cuff pressure as the explanatory variable and the error of the blood pressure measured by the measuring unit relative to the blood pressure used as the true value as the target variable. A calibration unit corrects for an error in the blood pressure measured by the measuring unit, the error being output by the calibration model by inputting information about the waveform related to the temporal variation of the cuff pressure detected by the detection unit into the calibration model; and The output unit outputs the blood pressure after it has been corrected by the correction unit.

2. The blood pressure monitor according to claim 1, wherein, The blood pressure used as the true value is the blood pressure measured by auscultation.

3. The blood pressure monitor according to claim 1, wherein, The waveform information includes information related to the cuff pressure.

4. The sphygmomanometer according to any one of claims 1 to 3, wherein, The measuring unit acquires the oscilloscope waveform, or OMW, based on the time-series change of the cuff pressure. The waveform information includes OMW information representing the waveform of the OMW.

5. The blood pressure monitor according to claim 4, wherein, The measurement unit acquires the oscilloscope waveform envelope, i.e., OMWE, based on the OMW. The waveform information includes OMWE information representing the waveform of the OMWE.

6. The blood pressure monitor according to claim 4, wherein, The correction unit corrects the blood pressure measured by the measurement unit based on an error, which is output by the correction model by inputting the temporal variation of the cuff pressure and the OMW information into the correction model.

7. The blood pressure monitor according to claim 5, wherein, The correction unit corrects the blood pressure measured by the measurement unit based on an error, which is output by the correction model by inputting the temporal variation of the cuff pressure and the OMWE information into the correction model.

8. A blood pressure measurement method, comprising the following calculation and execution: The control step involves controlling the pressure of the cuff wrapped around the arm of the person being measured on the arm. The detection step involves detecting the cuff pressure exerted on the arm by the cuff. The measurement procedure involves measuring the subject's blood pressure using an oscillometric method by measuring the temporal variation of the cuff pressure. A correction step corrects for an error in the measurement step, the error being output by a correction model that is constructed for the subject using machine learning. The machine learning uses information about the waveform related to the temporal variation of the cuff pressure as an explanatory variable and the error of the blood pressure measured in the measurement step relative to the blood pressure used as the true value as a target variable. The output step outputs the blood pressure after correction in the correction step.

9. A blood pressure measurement program that a computer executes: The control step involves controlling the pressure of the cuff wrapped around the arm of the person being measured on the arm. The detection step involves detecting the cuff pressure exerted on the arm by the cuff. The measurement procedure involves measuring the subject's blood pressure using an oscillometric method by measuring the temporal variation of the cuff pressure. A correction step corrects for an error in the measurement step, the error being output by a correction model that is constructed for the subject using machine learning. The machine learning uses information about the waveform related to the temporal variation of the cuff pressure as an explanatory variable and the error of the blood pressure measured in the measurement step relative to the blood pressure used as the true value as a target variable. The output step outputs the blood pressure after correction in the correction step.

10. A method for constructing a learning model, executed by a computer: The acceptance step accepts the input of the error between the blood pressure measured by oscillometric method and the blood pressure used as the true value; and The construction step involves building a learning model through machine learning, where information about the waveform related to the temporal variation of the cuff pressure of the cuff wrapped around the arm of the person being measured is set as an explanatory variable, and the error is set as the target variable.

11. A procedure for building a learning model, which enables a computer to execute: The acceptance step accepts the input of the error between the blood pressure measured by oscillometric method and the blood pressure used as the true value; and The construction step involves building a learning model through machine learning, where information about the waveform related to the temporal variation of the cuff pressure of the cuff wrapped around the arm of the person being measured is set as an explanatory variable, and the error is set as the target variable.

Images