Circulatory system physiological index calculation device, computer program, and recording medium

Abstract

To provide a technique for obtaining a circulatory system physiological index, which serves as an index for estimating blood pressure with higher accuracy.SOLUTION: A power spectrum obtained by frequency analysis of the time series data of the variation of the instantaneous heart rate is plot-displayed on a double logarithmic graph, multiple types of frequency bands to be analyzed are set for the plot display, and two types of approximation curves of different degrees are applied. Multiple approximation curves are obtained which are represented by a polynomial of the fourth degree or a polynomial of the third degree in which the standard deviation of the residual with the plot display of the power spectrum is the minimum. The smaller the standard deviation of the residual of an approximate curve, the higher the correlation between the frequency of the inflection points of the approximate curve and blood pressure, and on the basis of the fact, an approximate curve with a higher correlation to blood pressure is selected. For each of the tangents to the inflection points of the multiple approximate curves and the straight lines between the inflection points, the standard deviation of the residuals is also calculated, and the smallest one is adopted.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a technique for determining a physiological index of the circulatory system that correlates with blood pressure and blood flow. [Background technology]

[0002] The applicant has disclosed in Patent Document 1 a blood pressure estimation device that analyzes biosignal data obtained from a biosignal detection sensor using three-dimensional knitted fabric, captures vibration information within the body caused by blood flow fluctuations from the ventricular filling phase to the isovolumic contraction phase, determines an index showing the fluctuations by applying the Lorenz plot method, and estimates blood pressure values ​​from the index (Patent Document 1).

[0003] Specifically, the technology captures biosignals (acoustic pulse waves (APW)) from the surface of the body on the back using a biosignal detection sensor, filters the time-series waveform in a specified frequency band to obtain a filtered waveform that makes the cardiac cycle apparent, compares this filtered waveform with electrocardiogram waveform data obtained from an electrocardiograph measured at the same time, and identifies the waveform components from the ventricular filling period to the isovolumic contraction period. After that, the Lorentz plot method is applied, and using the slope of a large number of point clouds plotted at a specified measurement time as a reference, a new time series is constructed for the angle difference with the slope of the point clouds at a shorter measurement time. The time-series waveform is frequency-analyzed, the frequency analysis results are displayed on a double logarithmic axis, a regression line is found for the range from VLF to LF in the analyzed waveform, the slope of the regression line is defined as the fractal slope, and blood pressure is estimated from the fractal slope.

[0004] Patent Document 2 focuses on the fact that pulse wave velocity (PWV) depends on blood pressure and the elastic properties of arteries, and discloses a technique for measuring blood pressure assuming that blood pressure is proportional to PWV. It also describes that PWV is calculated from PAT (Pulse Arrival Time). Since PWV is inversely proportional to the derivative of PAT, blood pressure is estimated by calculating the derivative value of PAT. It also describes that heart rate and oxygen saturation are measured by an optical sensor used in mobile computers such as mobile phones.

[0005] Both Patent Documents 1 and 2 have in common the point that they are techniques capable of estimating blood pressure without using a cuff-based blood pressure monitor, i.e., an upper arm blood pressure monitor. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] JP 2019-122502 A [Patent Document 2] US2015 / 0320328A1 publication [Patent Document 3] Patent Publication No. 2022-71786 [Patent Document 4] Patent Publication No. 2022-71812 [Non-patent literature]

[0007] [Non-Patent Document 1] Guyton's Physiology, 13th edition, published March 20, 2018, author John E. Hall, general translator Yoshihiro Ishikawa, et al., publisher Elsevier Japan Co., Ltd. Summary of the Invention [Problem to be solved by the invention]

[0008] By the way, the complex control for maintaining blood pressure homeostasis has fluctuations due to energy fluctuations. Heart rate (HR) is derived from the RRI of the electrocardiogram (ECG), and heart rate variability (HRV) has fluctuations, and the fluctuation index is calculated from the results of frequency analysis of heart rate variability. Heart rate variability is reflected in afterload and baroreflex, and arterial pressure fluctuates due to afterload fluctuations and baroreflex.

[0009] In Patent Document 1, the fractal slope is used as an index showing this fluctuation, but the fractal slope is an index showing the characteristics of the analysis waveform only by a regression line. Therefore, it is difficult to say that it fully reflects the characteristics of blood pressure of a complex control structure having a complex control mechanism as described above.

[0010] In addition, as in Patent Document 2, blood pressure is calculated by detecting the pulse wave velocity. Although the pulse wave velocity is said to be correlated with blood pressure, the degree of correlation is not so high. Therefore, even if the pulse wave velocity or the pulse wave propagation delay time used as a substitute for the pulse wave velocity is simply estimated in comparison with the blood pressure, there are concerns about the accuracy.

[0011] Patent Documents 3 and 4 describe the determination of the apical pulsation wave (CAB) from the APW measured from the apex of the heart. Non-Patent Document 1 describes autoregulation.

[0012] The present invention has been made in consideration of the above, and an object of the present invention is to provide a technique for determining a circulatory system physiological index that serves as an index for estimating blood pressure with higher accuracy without using a cuff. [Means for solving the problem]

[0013] In order to solve the above problems, the circulatory system physiological index calculation device of the present invention comprises: A circulatory system physiological index calculation device for calculating a circulatory system physiological index correlated with blood pressure and blood flow using at least one of biosignal data of a heart rate and an acoustic pulse wave caused by a left ventricular pressure fluctuation obtained from a body surface, comprising: A biological signal receiving unit that receives the biological signal data; a time-series fluctuation data calculation unit for calculating time-series fluctuation data reflecting time-series fluctuation information of the biological signal data; a frequency analysis unit that plots and displays a power spectrum obtained by frequency analyzing the time-series fluctuation data on a double logarithmic graph; an approximation curve calculation unit that applies an n-th order polynomial or an n-1-th order polynomial to a plurality of types of analysis target frequency bands set in a range from VLF to LF in the plot display of the power spectrum, and obtains a plurality of approximation curves represented by the n-th order polynomial or the n-1-th order polynomial that minimizes the standard deviation of the residual from the plot display of the power spectrum for each analysis target frequency band; a frequency gradient calculation unit that calculates, for each of the approximation curves, a frequency gradient of a tangent to an inflection point, and, if there are two inflection points, a frequency gradient of a straight line between the two inflection points; an index determination unit that calculates a standard deviation of a residual between a tangent to the inflection points or a straight line between the inflection points and a plot display of the power spectrum, compares the standard deviations of the residuals, selects the tangent or straight line with the smallest standard deviation of the residual, and determines a frequency gradient of the selected tangent or straight line as the circulatory system physiological index; has.

[0014] The frequency band to be analyzed in the approximate curve calculation unit is It is preferable that the frequency range is from any of the power spectrum calculation points in the range of 0.006 to 0.034 Hz, including the fundamental frequency of 0.017 Hz, which is calculated using the time it takes for blood flow to return to a normal level when blood pressure rises due to the acute autoregulation of blood flow, to 0.08 Hz.

[0015] It is preferable that the approximation curve calculation unit determines whether or not a histogram of the residual between the approximation curve drawn in the frequency band to be analyzed and the plot display of the power spectrum is normally distributed, and if the histogram does not become normally distributed, adjusts the frequency band to be analyzed and calculates the approximation curve again.

[0016] It is preferable that the nth degree polynomial is a fourth degree polynomial, and the n-1th degree polynomial is a third degree polynomial. The approximation curve calculation unit is The analysis target frequency band is set to a range from any frequency of a power spectrum calculation point in a range of 0.006 to 0.034 Hz ​​to 0.08 Hz, and a first approximation curve represented by the third-order polynomial and a second approximation curve represented by the fourth-order polynomial are obtained; It is preferable to obtain a third approximation curve represented by the third order polynomial in a frequency range from the lower inflection point of two inflection points obtained from the second approximation curve of the fourth order polynomial to 0.08 Hz.

[0017] The index determination unit A tangent to an inflection point of the first approximation curve; Tangents to two inflection points of the second approximation curve; A straight line between two inflection points of the second approximation curve; Tangent to the inflection point of the third approximation curve From the total of five straight lines, It is preferable to select the straight line that has the smallest standard deviation of the residuals with the plot of the power spectrum.

[0018] The index determination unit A standard deviation of a residual of a tangent to each of the inflection points of the first approximation curve, the second approximation curve, and the third approximation curve is obtained within a range of a plot display of a predetermined power spectrum centered on each of the inflection points; For the second approximation curve, it is preferable to further obtain the standard deviation of the residuals for a predetermined range of the plot display of the power spectrum centered on the midpoint of the straight line between the two inflection points.

[0019] The computer program of the present invention comprises: A computer program that causes a computer to function as a circulatory system physiological index calculation device that calculates circulatory system physiological indexes that are correlated with blood pressure and blood flow using at least one of biosignal data of a heart rate and an acoustic pulse wave caused by a left ventricular pressure fluctuation obtained from a body surface, the computer program comprising: receiving said biosignal data; A step of obtaining time-series fluctuation data reflecting time-series fluctuation information of the biosignal data; a step of plotting and displaying a power spectrum obtained by frequency analyzing the time-series fluctuation data on a double logarithmic graph; a step of fitting an n-th degree polynomial or an n-1-th degree polynomial to a plurality of types of analysis target frequency bands set in a range from VLF to LF while the plot display of the power spectrum is being displayed, and obtaining a plurality of approximation curves represented by the n-th degree polynomial or the n-1-th degree polynomial that minimizes the standard deviation of the residual from the plot display of the power spectrum for each analysis target frequency band; calculating, for each of the approximation curves, a frequency gradient of a tangent to an inflection point, and, if there are two inflection points, a frequency gradient of a straight line between the two inflection points; a step of calculating a standard deviation of a residual between a tangent to the inflection point or a straight line between the inflection points and a plot display of the power spectrum, comparing the standard deviations of the residuals with each other, selecting the tangent or straight line with the smallest standard deviation of the residuals, and determining a frequency gradient of the selected tangent or straight line as the circulatory system physiological index; has.

[0020] The analysis target frequency band is It is preferable that the frequency range is from any of the power spectrum calculation points in the range of 0.006 to 0.034 Hz, including the fundamental frequency of 0.017 Hz, which is calculated using the time it takes for blood flow to return to a normal level when blood pressure rises due to the acute autoregulation of blood flow, to 0.08 Hz.

[0021] It is preferable to determine whether a histogram of the residual between the approximation curve drawn in the frequency band to be analyzed and the plot display of the power spectrum is normally distributed, and if it is not normally distributed, to adjust the frequency band to be analyzed and to obtain the approximation curve again.

[0022] It is preferable that the nth degree polynomial is a fourth degree polynomial, and the n-1th degree polynomial is a third degree polynomial. The analysis target frequency band is set to a range from any frequency of a power spectrum calculation point in a range of 0.006 to 0.034 Hz ​​to 0.08 Hz, and a first approximation curve represented by the third-order polynomial and a second approximation curve represented by the fourth-order polynomial are obtained; It is preferable to obtain a third approximation curve represented by the third order polynomial in a frequency range from the lower inflection point of two inflection points obtained from the second approximation curve of the fourth order polynomial to 0.08 Hz.

[0023] A tangent to an inflection point of the first approximation curve; Tangents to two inflection points of the second approximation curve; A straight line between two inflection points of the second approximation curve; Tangent to the inflection point of the third approximation curve From the total of five straight lines, From the total of five straight lines, It is preferable to select a straight line having a minimum standard deviation of the residual from the plot display of the power spectrum, and to determine the frequency gradient of the selected tangent line or the straight line as the circulatory system physiological index.

[0024] A standard deviation of residuals of tangents to each of the inflection points of the first approximation curve, the second approximation curve, and the third approximation curve is calculated within a range of a predetermined number of plots centered on each of the inflection points, For the second approximation curve, it is preferable to further obtain the standard deviation of the residuals for a predetermined range of the plot display of the power spectrum centered on the midpoint of the straight line between the two inflection points.

[0025] The present invention also provides a computer-readable recording medium on which the computer program is recorded. Effect of the Invention

[0026] According to the present invention, the power spectrum obtained by frequency analysis of time series data of the fluctuation of at least one of the biosignal data of the heart rate and the acoustic pulse wave caused by the fluctuation of the left ventricular pressure obtained from the body surface is plotted and displayed on a double logarithmic graph, and a plurality of frequency bands to be analyzed are set for the plot display, and two kinds of approximation curves of different orders are applied. As each approximation curve, a plurality of approximation curves represented by n-th order polynomials or n-1-th order polynomials that minimize the standard deviation of the residual with the plot display of the power spectrum are obtained. Since the correlation between the frequency of the inflection point of the approximation curve and the blood pressure is higher as the standard deviation of the residual is smaller, an approximation curve with a higher correlation with the blood pressure is selected based on the above, and the standard deviation of the residual is calculated for the tangents of the inflection points of each of the plurality of approximation curves and the straight lines between the inflection points, and the smallest one is adopted. As a result, the finally selected frequency gradient has a higher correlation with the blood pressure, and this is used as the circulatory system physiological index in the present invention. According to the present invention, it is possible to obtain a circulatory system physiological index with a high correlation with the blood pressure, and the accuracy of blood pressure estimation can be improved. [Brief description of the drawings]

[0027] [Figure 1] FIG. 1 is a block diagram showing the configuration of a circulatory system physiological index calculation device according to an embodiment of the present invention. [Diagram 2] FIG. 2 is a flowchart showing the process of frequency analyzing time-series fluctuation data (Δθi function) showing fluctuations in heart rate (instantaneous heart rate) and obtaining a power spectrum (spectral diagram) that is the result of the frequency analysis. [Diagram 3] FIG. 3 is a diagram showing data output in each step of FIG. 2 in accordance with the flowchart of FIG. [Figure 4] FIG. 4 is a flow chart showing the steps for obtaining the Δθi function using the APW. [Diagram 5] FIG. 5 is a diagram showing data output in each step of FIG. 4 in accordance with the flowchart of FIG. [Figure 6]Figure 6(a) is a scatter plot in which the frequency of the inflection point of the approximation curve represented by a fourth-order polynomial created using data from a total of 171 people (260 cases), including a group of healthy subjects (95 people) and a group of outpatients (76 people), in an experimental example is plotted in relationship to systolic blood pressure, and Figure 6(b) is a similar scatter plot when the frequency band to be analyzed is set to 0.006 to 0.08 Hz. [Figure 7] 7 is a flow chart for explaining the calculation steps in the approximation curve calculation unit, the frequency gradient calculation unit, and the index determination unit. [Figure 8] 8(a) to (c) are diagrams showing an example of data obtained by calculating an approximation curve and calculating a circulatory system physiological index (GT(GI)) in an experimental example. [Figure 9] 9(a) to (c) are diagrams showing an example of data obtained by calculating an approximation curve and calculating a circulatory system physiological index (GT(GI)) in an experimental example. [Figure 10] 10(a) to (c) are diagrams showing an example of data obtained by calculating an approximation curve and calculating a circulatory system physiological index (GT(GI)) in an experimental example. [Figure 11] 11(a) to (c) are diagrams showing an example of data obtained by calculating an approximation curve and calculating a circulatory system physiological index (GT(GI)) in an experimental example. [Figure 12] 12(a) to (c) are diagrams showing an example of data obtained by calculating an approximation curve and calculating a circulatory system physiological index (GT(GI)) in an experimental example. [Figure 13] 13(a) to (c) are diagrams showing an example of data obtained by calculating an approximation curve and calculating a circulatory system physiological index (GT(GI)) in an experimental example. [Figure 14] Figure 14(a) shows the correlation between the frequency gradient (GT(GI)) values ​​calculated for all 171 subjects (260 cases) and the actual systolic blood pressure (SBP) values ​​measured with an upper arm blood pressure monitor. Figure 14(b) shows a cross-tabulation table showing the relationship between systolic blood pressure and GT(GI). [Figure 15]FIG. 15(a) is a flowchart showing the steps for calculating an estimated blood pressure using the frequency gradient (GT or GI), which is a physiological index of the circulatory system, FIG. 15(b) is a diagram showing the correlation between the estimated blood pressure values ​​and the actually measured blood pressure values, and FIG. 15(c) is a cross-tabulation table thereof. [Figure 16] Figure 16(a) shows a three-dimensional response surface created using data from a group of healthy subjects (95 subjects), and Figure 16(b) shows a three-dimensional response surface created using data from a group of outpatients (76 subjects) with heart disease and risk factors. [Figure 17] 17(a) to (c) are diagrams showing each cross section of a three-dimensional response surface combining the data of the healthy subject group (95 subjects) and the data of the outpatient group (76 subjects). [Figure 18] Figure 18 shows the results of a verification experiment on a 34-year-old (healthy) subject, comparing the estimated blood pressure (eSBP) calculated using the past year's data on instantaneous heart rate with the actual brachial systolic blood pressure measured when measuring the instantaneous heart rate. [Figure 19] Figure 19 shows the results of a verification experiment on a 55-year-old subject (outpatient: taking medication for hypertension), comparing the estimated blood pressure (eSBP) calculated using the past year's data on instantaneous heart rate with the actual brachial systolic blood pressure measured when measuring the instantaneous heart rate. [Figure 20] Figure 20 shows the results of a verification experiment on a 67-year-old subject (outpatient: undergoing medical treatment for hypertension), comparing the estimated blood pressure (eSBP) calculated using the past year's data on instantaneous heart rate with the actual brachial systolic blood pressure measured when measuring the instantaneous heart rate. [Figure 21] 21(a)-(b) are diagrams for explaining application to a left ventricular pressure waveform. [Figure 22] FIG. 22 is a diagram for explaining the usefulness of an approximation equation for the left ventricular pressure waveform. [Figure 23] FIG. 23 is a diagram for explaining the reason why normal blood pressure often results in GI, while hypertension and hypotension often result in GT. [Figure 24]FIG. 24, like FIG. 23, is a diagram for explaining why normal blood pressure often results in GI, while hypertension and hypotension often result in GT. [Diagram 25] 25(a)-(b) are conceptual diagrams showing heart rate variability, blood pressure variability, and fluctuation variability. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0028] (Circulatory system physiological index calculation device) 1 shows a schematic configuration of one embodiment of a circulatory system physiological index calculation device 1. The circulatory system physiological index calculation device 1 of this embodiment is composed of a computer (the type of computer is not limited, and includes a personal computer, a microcomputer, a mobile information terminal such as a smartphone, and the like), and receives biosignal data and the like and executes a predetermined process.

[0029] Specifically, the circulatory system physiological index calculation device 1 has a computer program stored in a storage unit (memory) that executes procedures that function as a biosignal receiving unit 11, a time-series fluctuation data calculation unit 12, a frequency analysis unit 13, an approximation curve calculation unit 14, a frequency gradient calculation unit 15, and an index determination unit 16. The storage unit (memory) includes a recording medium such as a hard disk built into the circulatory system physiological index calculation device 1 composed of a computer, as well as various removable recording media and recording media of other computers connected by communication means. The circulatory system physiological index calculation device 1 can also be realized using an electronic circuit having a memory in which the above computer program is incorporated. That is, the biosignal receiving unit 11 can be configured as a transmission / reception circuit of a computer, and the time-series fluctuation data calculation unit 12, the frequency analysis unit 13, the approximation curve calculation unit 14, the frequency gradient calculation unit 15, and the index determination unit 16 can be configured as a processing circuit that calculates time-series fluctuation data from the biosignal data received in the transmission / reception circuit, performs frequency analysis, obtains an approximation curve, calculates a frequency gradient using the approximation curve, and obtains any of the frequency gradients as an index correlated with blood pressure.

[0030] The computer program can be provided by storing it in a recording medium. The recording medium storing the computer program may be a non-transient recording medium. There is no particular limitation on the non-transient recording medium, and examples of the non-transient recording medium include a flexible disk, a hard disk, a CD-ROM, an MO (magneto-optical disk), a DVD-ROM, and a memory card. It is also possible to transmit the computer program to a computer via a communication line and install it therein.

[0031] The biosignal receiving unit 11 receives at least one of biosignal data of the heart rate and the acoustic pulse wave caused by the left ventricular pressure fluctuation obtained from the body surface. The heart rate and the acoustic pulse wave are physiological indexes of the cardiovascular system, and their fluctuations affect the effective blood volume and become blood pressure control factors. The heart rate can be an instantaneous heart rate calculated from the RR interval (60 / Δt / min: δt=RRI) of an electrocardiogram (ECG) obtained from an electrocardiograph. The instantaneous heart rate may be calculated in an electrocardiograph and the biosignal receiving unit 11 may simply receive the data, or the biosignal receiving unit 11 may receive only the RRI and calculate the instantaneous heart rate. The biosignal receiving unit 11 of this embodiment may have either of these functions. In addition, a wristwatch type electrocardiograph, for example, a smart watch, may be used. The acoustic pulse wave (APW) is sound and vibration information collected from the back, and an acoustic pulse wave detection sensor used in a drowsy driving warning device (Sleep Buster (registered trademark)) manufactured by Delta Tooling Co., Ltd. In addition, the acoustic pulse wave (APW) uses data (APW20) that is obtained by filtering the received data at 10 to 30 Hz.

[0032] The time-series fluctuation data calculation unit 12 obtains time-series data (time-series fluctuation data) of fluctuations (fluctuations) of the biosignal data received by the biosignal receiving unit 11. The time-series fluctuation data is data constructed by comparing biosignal data at a specific time with biosignal data at the next specific time, and plotting data showing the fluctuation between the two. The fluctuation is based on the time phase of atrial contraction-ventricular contraction by ECG, i.e., the R wave, and in this embodiment, the fluctuation is converted into angle information by the Lorentz plot method, and a function with the angle as an argument ("Δθi function") is obtained.

[0033] That is, the time series fluctuation data calculation unit 12 obtains a reference slope from the Lorentz plot in the reference time range, performs moving average processing on the relative slope (characteristic of fluctuation), which is the difference between the slope of each Lorentz plot created at each specified time shorter than the reference time range and the reference slope, and obtains the time series data of the obtained relative slope as time series fluctuation data of the biosignal data.

[0034] For example, as shown in Fig. 2 and Fig. 3, after receiving the instantaneous heart rate (HR) from the RRI of the electrocardiogram (S401, S501), the horizontal axis is set to the i+1th value and the vertical axis is set to the i-th heart rate (instantaneous heart rate) value, and a Lorenz plot for a predetermined reference range time (e.g., 360 seconds) is created, and the slope of the regression line (reference slope) of the plotted points is obtained (S402, S502-1). The reference range time is preferably 180 seconds or more, and more preferably 360 seconds. The time series data of physiological indices includes information on the amount of fluctuation, the periodicity of the fluctuation, and the manner and pattern of the fluctuation.

[0035] The waves in blood vessels are a superposition of waves generated by various factors, and the LF component is by definition 0.04 to 0.15 Hz, but is usually distributed in the frequency band of 0.06 to 0.11 Hz. On the other hand, the very low frequency component (VLF component) defined as 0.003 to 0.04 Hz, which is the long-term heart rate fluctuation, contains information that reflects the regulation of the effective blood volume of the circulatory regulatory system and the health of the entire organism. To capture the frequency component of 0.003 Hz, a minimum measurement time of 1 / 0.003 = 333.4 seconds is required. If the measurement time is 360 seconds, frequency components of 0.0028 Hz and above can be captured.

[0036] For a time shorter than the reference range time (for example, when the reference range time is 360 seconds, a measurement time shorter than that (for example, 30 seconds)), the Lorentz plot method is applied to similarly plot, and the slope of the regression line of the point cloud plotted in this short time window (short-time point cloud slope) is obtained, and the deviation angle between this short-time point cloud slope (θi) and the reference slope (θ0) is obtained as the relative slope (Δθi) (S502-2 in FIG. 3). As described above, in this embodiment, the pulsation periods one beat before and one beat after are taken on the vertical and horizontal axes, respectively, and when the Lorentz plot falls on a 45-degree straight line drawn from the origin on the graph, this situation indicates that the fluctuation for each beat is zero. Therefore, the deviation angle (relative slope (Δθi)) indicates fluctuation information of the instantaneous heart rate.

[0037] The short-time point cloud slope (θi) is sequentially obtained for each time window by moving average processing. Since heart rate fluctuations respond more quickly to stimulation of the parasympathetic nervous system than to stimulation of the sympathetic nervous system, the time resolution is determined taking into consideration the parasympathetic nervous system's heart rate regulation function and respiratory regulation, and in this embodiment, the time resolution is set to 3 seconds, and the time window for the moving average is set to 30 seconds. Then, the short-time point cloud slope (θi) and relative slope (Δθi) are obtained for each time window, and a time series of the relative slope (Δθi) is created (S403 in FIG. 2, S503 in FIG. 3). This time series becomes time series data (time series fluctuation data) showing the fluctuation of the heart rate (instantaneous heart rate), that is, the above-mentioned "Δθi function". Note that 30 seconds is twice the average time of 15 seconds, which is the time in which the effect of sympathetic nerve enhancement / suppression is reflected.

[0038] On the other hand, when the biological signal data is APW20, three extreme values ​​(a0 to a2) between atrial contraction and ventricular contraction are extracted from the received APW20 data, and two inter-extreme value amplitudes (a2-a0, a1-a0) are calculated (S601 to S602 in FIG. 4, S701 to S702 in FIG. 5). These values ​​are plotted on the horizontal and vertical axes to create a Lorentz plot, and the reference slope (S603 in FIG. 4, S703 in FIG. 5) is calculated as described above. Furthermore, the relative slope (S604 in FIG. 4, S704 in FIG. 5) is calculated, and a Δθi function is created (S605 in FIG. 4, S705 in FIG. 5).

[0039] The VLF is an index of heart rate fluctuation for analyzing non-periodic components, and analysis is performed in the frequency band of 0.003 to 0.04 Hz. However, irregular noise-like components make the estimation of correlation inaccurate, so it is necessary to separate and evaluate the noise components. The movement of the baseline fluctuation, which will be described later, "non-perturbation" does not have a constant frequency, so no peak is formed in the power spectrum. Therefore, it is necessary to remove the influence of low-frequency fluctuation components that extend over minutes or more, such as the renin-angiotensin system, among the eight arterial pressure control mechanisms. For this reason, the frequency analysis unit 13 described below uses data obtained by setting a high-pass filter at 0.006 Hz for time-series data (time-series fluctuation data) showing the fluctuation of the heart rate (instantaneous heart rate), that is, the above-mentioned "Δθi function", and setting a low-pass filter at 0.08 Hz to remove respiratory fluctuation components in the frequency band of 0.1 Hz or more.

[0040] The frequency analysis unit 13 is a means for frequency analysis (S404 in FIG. 2, S504 in FIG. 3) of the Δθi function (applied with a 0.006 Hz high-pass filter and a 0.08 Hz low-pass filter), which is the time series data of the fluctuation of the biosignal data obtained from the time series fluctuation data calculation unit 12, and displaying it on a double logarithmic axis (S405 in FIG. 2, S505 in FIG. 3), thereby capturing the manner in which the biosignal data fluctuates. S505 in FIG. 3 shows an example of a double logarithmic graph display of the power spectrum as the frequency analysis result. This double logarithmic graph power spectrum is displayed with a large number of points plotted. In this specification, the display of the power spectrum displayed with a large number of points plotted is referred to as a "plot display".

[0041] The approximation curve calculation unit 14 performs a calculation to fit an approximation curve to the plot display of the double logarithmic graph obtained by the frequency analysis unit 13. Specifically, during the plot display of the power spectrum, an n-th order polynomial or an n-1-th order polynomial is fitted to a plurality of types of analysis target frequency bands set in the range from VLF to LF, and a plurality of approximation curves represented by an n-th order polynomial or an n-1-th order polynomial that minimizes the standard deviation of the residual with each plot of the plot display of the power spectrum are obtained for each analysis target frequency band. From the plot display of the power spectrum of the double logarithmic graph obtained by analyzing a large number of subjects, it is preferable to use a fourth order polynomial or a third order polynomial, which allows for almost accurate analysis. However, if the accuracy of the value of the circulatory system physiological index (GT or GI described later) is improved by fitting an approximation curve represented by a higher order polynomial, a higher order polynomial can also be applied.

[0042] The frequency band to which the approximation curve is applied in the approximation curve calculation unit 14 is preferably a range from VLF to LF, equal to or higher than the fundamental frequency calculated using the time required for blood flow to return to a normal level when blood pressure rises due to acute autoregulation of blood flow as a fundamental period. As a result, the index obtained by the index calculation unit 16 described later is a circulatory system physiological index that correlates with blood pressure and blood flow.

[0043] This is based on the following statement on page 187 of "Guyton's Physiology, 13th Edition, published March 20, 2018, author John E. Hall, general translator Yoshihiro Ishikawa, et al., publisher Elsevier Japan Co., Ltd.": "When arterial blood pressure rises suddenly in any tissue, the blood flow rate increases immediately. However, in less than a minute, blood flow rate returns to almost normal in many tissues even though arterial pressure remains high. This phenomenon of blood flow rate returning to normal is called autoregulation. When autoregulation occurs, the relationship between local blood flow rate and arterial pressure in many body tissues becomes like the "acute" curve shown by the solid line in Figure 17.5 (shown on page 187 of the same book)." This "acute" curve is a curve that changes linearly when blood pressure is between about 100 and 150 mmHg, and changes nonlinearly when blood pressure is below 100 mmHg and above 150 mmHg, with the vertical axis representing blood flow rate and the horizontal axis representing mean arterial pressure. In addition, because blood flow returns to normal levels in less than a minute, the basic period of arterial pressure fluctuation is 1 / 60 (second) ≒ 0.017 Hz. According to Guyton's previous research (Role of the Baroreceptor Reflex in Daily Control of Arterial Blood Pressure and Other Variables in Dogs), the recovery time of the baroreceptor reflex is 21 seconds (1 / 21 = 0.0476 Hz).

[0044] From the above, it can be seen that the fundamental period of arterial pressure fluctuations is 0.017 Hz when linearity is maintained, but when the patient is in a low or high blood pressure state, nonlinearity occurs, resulting in frequency components of 0.034 Hz, 0.051 Hz, 0.068 Hz, 0.085 Hz, 0.102 Hz, etc., which are harmonic components of 0.017 Hz.

[0045] On the other hand, the LF component commonly used in pulse wave measurement is usually 0.04 to 0.15 Hz, but the actual measured results are mainly distributed between 0.06 Hz and 0.11 Hz. Therefore, in order to capture the effects of arterial pressure fluctuations and baroreceptor reflex, the frequency band to be analyzed was mainly considered to be in the range up to 0.06 Hz, including 0.017 Hz, which is the fundamental period of arterial pressure fluctuations, its harmonic components 0.034 Hz ​​and 0.051 Hz, and 0.0476 Hz, which is the baroreflex.

[0046] The presence of non-periodic components is expected below 0.017 Hz. The frequency band in which non-periodic components appear in heart rate variability analysis is VLF, which is defined as 0.003 to 0.004 Hz. Therefore, between 0.017 and 0.06 Hz, periodic components and non-periodic components related to VLF are mixed. Non-periodic components below 0.017 Hz are expressed as 1 / f α This type of fluctuation is known as baseline perturbation. In this specification, this baseline perturbation is sometimes referred to as "non-perturbation," and short-term periodic fluctuations that appear in both directions across the baseline are referred to as "perturbation." When the baseline perturbation is white noise with no frequency dependency, α=0, and when α=1, it becomes so-called 1 / f fluctuation. When the amplitude of the baseline perturbation decreases inversely proportional to the power of the frequency, it approaches α=3. When α=2, it becomes signal noise (Brown noise) generated by Brownian motion.

[0047] The frequency band from 0.017 to 0.06 Hz shows the above-mentioned autoregulation function, and is a region where periodic and non-periodic components coexist, that is, where non-perturbations, which are baseline oscillations and short-term periodic fluctuations, and perturbations coexist, and where the 1 / f α It has frequency characteristics in which a peak is superimposed on the shape fluctuation.

[0048] From the above, it can be seen that normal blood pressure patients in a resting state have a non-periodic component of 1 / f 2 In hypertensive patients at rest, the 1 / f type fluctuation becomes shallower and the 1 / f type fluctuation appears.0 In subjects with hypotension at rest, the power spectrum of the baroreflex decreases to 1 / f 2 This results in fluctuations in shape.

[0049] As described above, the frequency band to be analyzed by the approximation curve calculation unit 14 is basically in the range of 0.017 to 0.06 Hz. However, since the occurrence of fluctuations varies from person to person, it can be adjusted to the range of 0.006 to 0.08 Hz when the Δθi function is calculated in the time-series fluctuation data calculation unit 12.

[0050] The approximate curve calculation unit 14 basically sets the approximate curve in the range of 0.017 to 0.06 Hz, but allows the analysis target frequency band to be adjusted around that range. Therefore, the standard deviation of the residual is used as a criterion for more appropriately setting the analysis target frequency band and the approximate curve. That is, the approximate curve calculation unit 14 obtains the residual from the plot display of the power spectrum obtained by the frequency analysis unit 13 for the approximate curve set in the range of 0.017 to 0.06 Hz, which is the basic analysis target frequency band, and the approximate curve set in the analysis frequency band ranging from any frequency around 0.017 Hz (however, 0.006 Hz or more as described above) to any frequency around 0.06 Hz (however, 0.08 Hz or less as described above), and calculates the standard deviation. In this embodiment, since a high-pass filter of 0.006 Hz is set as described above, the lower limit of the frequencies around 0.017 Hz is set to 0.006 Hz. Also, since data related to arterial pressure fluctuations is shown on the lower frequency side of the boundary of 0.03 Hz as described below, it is preferable to adjust the range to an upper limit of 0.034 Hz, which is the first harmonic component of the fundamental period of arterial pressure fluctuations, that is, in the range of 0.006 to 0.034 Hz.

[0051] Figure 6(a) shows a scatter plot of the frequency at the inflection point (IPF) calculated by creating an approximate curve represented by a fourth-order polynomial that minimizes the sum of squares of the residuals between 0.017 and 0.06 Hz using data (instantaneous heart rate data calculated using RRI obtained from ECG) from a total of 171 people (260 cases), consisting of a group of healthy people (95 people) and a group of outpatients (76 people) with cardiovascular diseases and their risk factors, such as heart disease, hypertension, gout, diabetes, and dyslipidemia. The vertical axis is systolic blood pressure and the horizontal axis is IPF. This scatter plot suggests the involvement of two control systems controlled by the nervous system between 0.017 and 0.06 Hz, with clusters around 0.025 Hz and 0.045 Hz, with the border at 0.03 Hz; the former is closer to 0.017 Hz and is therefore presumed to be associated with arterial pressure fluctuations, while the latter is closer to 0.0476 Hz and is therefore presumed to be associated with baroreflex.

[0052] Generally, when white noise is mixed with a linear component, it becomes a normal distribution, and when white noise is mixed with a nonlinear component, it becomes a biased normal distribution. Therefore, the frequency band to be analyzed was expanded to 0.006-0.08 Hz, and an approximation curve that reflects the linear and nonlinear components was created. The approximation curve was expressed as a fourth-order polynomial with linear or nonlinear components in the frequency band where the residual standard deviation is minimized and the residual becomes a normal distribution. Figure 6(b) shows the plot of two IPFs of the approximation curve shown by the fourth-order polynomial, and it can be seen from this figure that the two IPFs are weakly correlated but have blood pressure dependence.

[0053] An approximate solution can be calculated from the gradient approximation and linear approximation regarding the IPF. The gradient approximation characterizes the nonlinear component, and the linear approximation characterizes the linear component. Therefore, the frequency gradient regarding the inflection point (GT or GI below) becomes an approximate solution having blood pressure dependency regarding the inflection point of the approximate curve, and becomes a new index (circulatory system physiological index) regarding blood pressure and blood flow. Based on these, the approximate curve calculation unit 14 of this embodiment calculates three cubic and quartic approximate curves (first to third approximate curves) as follows. This will be explained based on the flowchart in FIG. 7.

[0054] (1) First Approximation Curve (3rd order polynomial approximation curve for "non-perturbation") This is an approximation curve shown by a third-order polynomial with 0.017 to 0.06 Hz as the basic frequency band to be analyzed. The approximation curve calculation unit 14 creates an approximation curve represented by a third-order polynomial in the range of 0.017 to 0.06 Hz (S1). It finds the residual between each plot of the displayed power spectrum and the approximation curve in the range of 0.017 to 0.06 Hz, and judges whether the histogram based on the residuals shows a normal distribution (S2). If the distribution is not normal (if "No" in S2), the approximation curve calculation unit 14 sets a new analysis target band using the power spectrum in the vicinity of 0.006 to 0.017 Hz (upper limit is 0.034 Hz) (S3), and obtains an approximation curve expressed by a third-order polynomial again (S1). Therefore, the first approximation curve becomes a third-order polynomial approximation curve biased toward the 0.006 Hz side, and is highly related to "non-perturbation". The calculation is repeated, and if there is one approximation curve whose residual histogram is normally distributed and whose standard deviation of the residuals is the smallest, the determination in S2 becomes "Yes" and that third-order polynomial approximation curve is determined to be the first approximation curve.

[0055] (2) The second approximation curve (a fourth-order polynomial approximation curve that combines "non-perturbation" and "perturbation") This is an approximation curve shown by a fourth-order polynomial with 0.017 to 0.06 Hz as the basic frequency band to be analyzed. The approximation curve calculation unit 14 creates an approximation curve represented by a fourth-order polynomial in the range of 0.017 to 0.06 Hz (S4). The residual between each plot of the displayed power spectrum and the approximation curve in the range of 0.017 to 0.06 Hz is calculated, and a histogram based on the residuals is determined to determine whether it is a normal distribution (S2). If the distribution is not normal (if "No" in S2), the approximation curve calculation unit 14 reselects the power spectrum at both ends of the frequency band near 0.017 Hz (0.006 to 0.034 Hz ​​range) and near 0.06 Hz (usually a range from a value exceeding 0.034 to 0.06 Hz, but the upper limit is a maximum of 0.08 Hz), sets a new analysis target band (S3), and again obtains an approximation curve represented by a fourth-order polynomial (S4). Therefore, the second approximation curve is a fourth-order polynomial approximation curve corresponding to the intermediate frequency band between 0.006 Hz and 0.06 Hz. The calculation is repeated, and if there is one approximation curve whose residual histogram is normally distributed and whose residual standard deviation is the smallest, the result of S2 is "Yes" and that fourth-order polynomial approximation curve is determined to be the first approximation curve.

[0056] (3) The third approximation curve (a third-order polynomial approximation curve for "perturbation") This is an approximation curve represented by a third-order polynomial, with the power spectrum from near the inflection point on the low-frequency side of the second approximation curve represented by a fourth-order polynomial to near 0.06 Hz as the frequency band to be analyzed. The approximation curve calculation unit 14 creates an approximation curve expressed by a third-order polynomial in the range from the vicinity of the inflection point on the low frequency side of the second approximation curve to 0.06 Hz (S5). The residual between each plot of the power spectrum plotted and displayed in the analysis target frequency band and the approximation curve is calculated. It is determined whether or not the histogram based on this residual is a normal distribution (S2), and if it is not a normal distribution (if "No" in S2), a new analysis target band is set (S3) using the power spectrum near the inflection point on the low frequency side of the second approximation curve and near 0.06 Hz, and an approximation curve expressed by a third-order polynomial is calculated again (S5). Therefore, the third approximation curve becomes a third-order polynomial approximation curve biased toward the 0.06 Hz side, and is highly related to "non-perturbation". The calculation is repeated, and if there is one approximation curve whose residual histogram is normally distributed and whose standard deviation of the residuals is the smallest, the result of S2 is "Yes" and that third-order polynomial approximation curve is determined to be the third approximation curve. The residual standard deviations of the three approximation curves (1) to (3) are almost the same value.

[0057] The frequency gradient calculation unit 15 calculates the frequency gradient for each of the approximation curves obtained by the approximation curve calculation unit 14. In the case of the first and third approximation curves represented by a third-order polynomial, a tangent to the inflection point (the equation representing this straight line is referred to as an "estimated regression equation" in the figure) is obtained, and the gradient is calculated as the frequency gradient (S6, S7 in FIG. 7). In the case of the second approximation curve represented by a fourth-order polynomial, a tangent to the two inflection points and a straight line connecting the two inflection points (the equation representing these three straight lines is also referred to as an "estimated regression equation" in the figure) are obtained, and the gradients of the three straight lines are calculated as the frequency gradient (S8 in FIG. 7). In this specification, depending on the case, regardless of the degree of the polynomial, the frequency gradient calculated as the gradient of the tangent to the inflection point is represented as GT (Gradient of Tangent), and the frequency gradient calculated as the gradient of the straight line between two inflection points is represented as GI (Gradient of Inflection).

[0058] The index determination unit 16 determines the frequency gradient to be finally adopted as the circulatory system physiological index from a total of five lines including the tangents of the four inflection points and the line between one inflection point obtained by the frequency gradient calculation unit 15. Specifically, for the five lines (regression equations of each line), one line is determined based on the residual standard deviation around the inflection point (S9 in FIG. 7), and the frequency gradient is set as the final circulatory system physiological index (S10 in FIG. 7). "Around the inflection point" refers to a range of a predetermined power spectrum centered on each inflection point in the case of the tangents of each inflection point of the first approximation curve, the second approximation curve, and the third approximation curve represented by a third-order polynomial. The power spectrum is obtained at an appropriate frequency interval, and as described above, is plotted and displayed for each obtained power spectrum. In this embodiment, the residual standard deviation of the tangent is obtained within a plot display of ±2 points around the inflection point, which is a range relatively close to the inflection point. This is merely an example, and the range is not limited, and the interval for obtaining the power spectrum is also arbitrary. However, it is preferable to obtain the residual standard deviation in a range as close as possible to the inflection point, as in this embodiment, in a frequency range within ±0.03 Hz, preferably within ±0.02 Hz, and more preferably within ±0.01 Hz, centered on the inflection point.

[0059] On the other hand, in the case of the second approximation curve represented by the approximation formula of a fourth-order polynomial, the residual standard deviation is calculated not only around the two inflection points, but also in a range of a plot display of a predetermined power spectrum centered on the midpoint of a line connecting the two inflection points. That is, in the case of the second approximation curve, the residual standard deviation is calculated at three points. In this specification, the "midpoint" which is the range of a predetermined power spectrum centered on this midpoint is also used as an element for determining one straight line from five straight lines, so for convenience, it is described as being included in the concept of "around the inflection point".

[0060] The index determination unit 16 compares the residual standard deviations around the five inflection points (including around the midpoint) for the five straight lines to determine the smallest one (S9 in FIG. 7). Then, it determines the frequency gradient (GT or GI) of the one straight line showing the smallest residual standard deviation (the equation showing this straight line is called the "true regression equation" in the figure) as the circulatory system physiological index (S10 in FIG. 7). However, if at least one of the residual standard deviations around the inflection points of the three frequency gradients (straight lines) of the second approximation curve of the fourth-order polynomial is smaller than the residual standard deviations around the inflection points of the frequency gradients (straight lines) of the first and third approximation curves of the third-order polynomial, it is preferable to consider the following judgment. That is, in this case, when the signs of the frequency gradients in the first and third approximation curves of the third-order polynomial are the same, the GT with the same gradient sign is selected from the gradients (GT) of the tangents of the two inflection points of the second approximation curve of the fourth-order polynomial. When the signs of the frequency gradients in the first and third approximation curves of the third-order polynomial are not the same, the gradient (GI) between the inflection points of the second approximation curve of the fourth-order polynomial is selected. This makes it easy to select one from the three frequency gradients (straight lines) of the second approximation curve of the fourth-order polynomial. Also, as is clear from the experimental example described later, even when this determination method is adopted, in most cases the residual standard deviation of the finally selected frequency gradient (straight line) is the smallest among the five.

[0061] In this specification, blood pressure classification is based on the classification defined in the "Guidelines for the Treatment of Hypertension 2019 (JSH2019)" (issued by the Japanese Society of Hypertension), and for systolic blood pressure, it is defined as "normal blood pressure: less than 120 mmHg", "high normal blood pressure: 120-129 mmHg", "high blood pressure: 130-139 mmHg", and "hypertension: 140 or higher (stage I hypertension: 140-159, stage II hypertension: 160-179, stage III hypertension: 180 or higher)". However, in this specification, systolic blood pressure of 100 mmHg or lower is classified as "low blood pressure" according to the WHO criteria.

[0062] (Experimental Example) The analysis results using the instantaneous heart rate data of the above 171 subjects (260 cases) are shown in Figures 8 to 13. Among (a), (b), and (c) of Figures 8 to 13, the figures shown in the upper rows are diagrams in which a Δθi function, which is time-series fluctuation data showing fluctuations in instantaneous heart rate, is calculated from the instantaneous heart rate data received by the biological signal receiving unit 11 in the time-series fluctuation data calculating unit 12, and frequency analysis is performed in the frequency analyzing unit 13, and the power spectrum is plotted.

[0063] The data of the 171 subjects was measured at the applicant's research facility and two hospitals. At the applicant's research facility, the subjects were seated in a car seat, and at the two hospitals, the subjects were seated in a relaxing chair. The experimental equipment, experimenters, subjects, analysts, and seasons in which the experiments were conducted were diverse.

[0064] In the applicant's research facility, systolic blood pressure was measured twice using an upper arm blood pressure monitor and the average value was recorded. However, if the difference between the two measurements was 5 mmHg or more, the measurement was performed three or four times and the average value of the two closest measurements was recorded. In the hospital, the measurement was performed only once.

[0065] In addition, heart rate data was collected for 360 seconds from 171 subjects (260 cases) using an electrocardiograph while they were sitting quietly.

[0066] The curves shown by solid lines in the upper diagrams of each of (a), (b), and (c) of Figures 8 to 13 are approximation curves calculated by the approximation curve calculation unit 14. Of these, each diagram (a) is a first approximation curve represented by a third-order polynomial, each diagram (b) is a second approximation curve represented by a fourth-order polynomial, and each diagram (c) is a third approximation curve represented by a third-order equation using an inflection point on the low-frequency side of the second approximation curve. In addition, in the diagrams shown in the upper diagrams, the first approximation curve, the second approximation curve, and the third approximation curve show only the approximation curve with the smallest residual standard deviation among the approximation curves represented by a plurality of third-order polynomials or fourth-order polynomials calculated by changing the analysis target frequency band as described above.

[0067] The diagram shown in the middle shows histograms of residuals calculated by the approximate curve calculation unit 14 for the first approximate curve, the second approximate curve, and the third approximate curve shown in the upper part.

[0068] The diagrams shown in the lower part show a total of five straight lines, including the tangents of the four inflection points calculated by the frequency gradient calculation unit 15 and the straight line between one inflection point, the residual standard deviation around the inflection points (including around the midpoint) calculated by the index determination unit 16, and the frequency gradient (GT or GI) value of the straight line with the smallest residual standard deviation. The residual standard deviation is obtained for the plot of the power spectrum at two points before and after the inflection point (including the midpoint). In addition, for the second approximation curve shown by the fourth-order polynomial in each figure (b), the symbol a indicates the inflection point on the low frequency side, and the symbol c indicates the frequency of the inflection point on the high frequency side. The symbol b indicates the straight line itself connecting the two inflection points, and the value to the right of "b:" at the bottom of the graph is the residual standard deviation around the midpoint between the inflection points a and c.

[0069] 8(a) to 8(c), comparing the residual standard deviations shown in the lower diagrams, the minimum value is the value around the midpoint of the line b between the inflection points of the second approximation curve: 0.0581. Therefore, the index determining unit 16 determines the frequency gradient of the line b between the inflection points: GI=-0.09913 as the circulatory system physiological index. The sign of the gradient of the first approximation curve biased toward the 0.006 Hz side is negative, while the sign of the gradient of the third approximation curve biased toward the 0.06 Hz side is positive or approximately 0. Therefore, since the signs of the two gradients do not match, the same result will be obtained even if the determination is made to select GI from the second approximation curve. The GI has two characteristics: a frequency gradient in the vicinity of 0.017 Hz and in the frequency band from 0.03 to 0.06 Hz. This subject had a normal blood pressure of 116 mmHg (heart rate at the time of measurement: 60 / min) as systolic blood pressure measured with an upper arm blood pressure monitor.

[0070] 9(a) to 9(c), comparing the residual standard deviations shown in the lower diagrams, the minimum value is the value around the inflection point c of the second approximation curve: 0.0551. Since the inflection point c is an inflection point on the high frequency side, the index determining unit 16 determines the frequency gradient of the tangent to the inflection point c: GT = 1.5691 as the circulatory system physiological index. In addition, the signs of the gradient of the first approximation curve biased toward the 0.006 Hz side and the gradient of the third approximation curve biased toward the 0.06 Hz side are both positive and the signs of the gradients match. Therefore, the same result is obtained by selecting the GT at the inflection point c with a positive gradient in the second approximation curve. This subject had a systolic blood pressure of 184 mmHg (heart rate at the time of measurement: 108 / min) measured with an upper arm blood pressure monitor, and was classified as a subject with stage III hypertension.

[0071] 10(a) to 10(c), comparing the residual standard deviations shown in the lower diagrams, the minimum value is the value around the inflection point a of the second approximation curve: 0.0653. Since the inflection point a is an inflection point on the low frequency side, the index determining unit 16 determines the frequency gradient of the tangent to the inflection point a: GT = -2.587 as the circulatory system physiological index. The gradients of the first approximation curve, which is biased toward the 0.006 Hz side, and the gradients of the third approximation curve, which is biased toward the 0.06 Hz side, are all negative and the gradient signs are the same. In addition, the gradients of the three straight lines of the second approximation curve are also negative. In this case, the GT of the tangent to the inflection point a of the smallest residual standard deviation in the second approximation curve is used. This subject had low blood pressure, with a systolic blood pressure of 76 mmHg (heart rate at the time of measurement: 108 / min) measured using an upper arm blood pressure monitor.

[0072] 11(a) to 11(c), comparing the residual standard deviations shown in the lower diagrams, the minimum value is 0.0922, which is the value around the midpoint of the line b between the inflection points of the second approximation curve. Therefore, the index determining unit 16 determines the frequency gradient of the line b between the inflection points: GI=-0.1025, as the circulatory system physiological index. Note that the gradient of the first approximation curve, which is biased toward 0.006 Hz, is negative, while the gradient of the third approximation curve, which is biased toward 0.06 Hz, is positive. Therefore, since the two gradients have different signs, the same result would be obtained if the GI of the second approximation curve were selected. This subject had a high systolic blood pressure of 137 mmHg (heart rate at the time of measurement: 58 / min) as measured with an upper arm blood pressure monitor.

[0073] 12(a) to 12(c), comparing the residual standard deviations shown in the lower diagrams, the minimum value is the value around the inflection point c of the second approximation curve: 0.1394. Since the inflection point c is an inflection point on the high frequency side, the index determining unit 16 determines the frequency gradient of the tangent to the inflection point c: GT = 0.3037 as the circulatory system physiological index. In addition, the signs of the gradient of the first approximation curve biased toward the 0.006 Hz side and the gradient of the third approximation curve biased toward the 0.06 Hz side are both positive and the signs of the gradients match. Therefore, the same result is obtained by selecting the GT at the inflection point c with a positive gradient in the second approximation curve. This subject had a systolic blood pressure of 149 mmHg (heart rate at the time of measurement: 74 / min) measured using an upper arm blood pressure monitor, and was classified as a subject with stage I hypertension.

[0074] 13(a) to 13(c), comparing the residual standard deviations shown in the lower diagrams, the minimum value is the value around the inflection point of the third approximation curve: 0.027. Therefore, the index determining unit 16 determines the frequency gradient of the tangent to this inflection point: GT = 1.2509 as the circulatory system physiological index. This subject had a systolic blood pressure of 168 mmHg (heart rate at the time of measurement: 83 / min) measured using an upper arm blood pressure monitor, and was classified as a subject with stage II hypertension.

[0075] FIG. 14(a) shows the correlation between the frequency gradient (GT(GI)) value calculated as described above and the actual systolic blood pressure (SBP) value measured with an upper arm blood pressure monitor for all 171 subjects (260 cases). The blood pressure estimation equation shown by the regression equation is y=27.085x+135.82, and the coefficient of determination is R 2=0.8663, showing a high correlation.

[0076] In the Guyton physiology cited above, after explaining that autoregulation occurs and that the relationship between local blood flow and arterial pressure in many body tissues becomes like the above-mentioned "acute" curve, the "acute" curve (hereinafter, in this specification, the characteristics of this curve will be referred to as "acute curve characteristics") is described as follows: "Please note that when arterial pressure is between about 70 and 175 mmHg, even if arterial pressure increases by 150%, blood flow increases by only 20 to 30%. In some tissues such as the brain and heart, this autoregulation is more precise." If this acute curve characteristic is combined with the regression equation shown in Figure 1 and blood pressure is eliminated, the above explanation can be rephrased as "When arterial pressure is between about 70 and 175 mmHg, blood flow is regulated within the range of -3 to 1.5 in frequency gradient (GT or GI), and during that time, blood flow increases by only 20 to 30%." In other words, the regression equation in FIG. 14(a) is considered to have been obtained by examining the relationship between autoregulatory function with acute curve characteristics and blood pressure regulation. From this point of view, it can be said that the frequency gradient (GT or GI) calculated in this embodiment is suitable as a circulatory system physiological index related to the regulatory function of blood flow and blood pressure.

[0077] In addition, Fig. 14(b) shows a cross-tabulation table showing the relationship between the measured systolic blood pressure (SBP), GT, and GI. The presence or absence of association was determined using Fisher's exact test, the strength of association was analyzed using Cramer's association coefficient, and an independence test was performed between the frequency slope (GT or GI) and systolic blood pressure (SBP).

[0078] The Cramer's correlation coefficient was calculated to be 0.558, indicating that there is a strong correlation between the frequency gradient (GT or GI) and systolic blood pressure (SBP). The Cramer's correlation coefficient is χ 2It is an index that shows the strength of association, with values ​​standardized between 0 and 1. The index showing the strength of association between items on the side and top of the cross-tabulation table is defined as follows: 0.1 to 0.25 indicates a weak association, 0.25 to 0.5 indicates a moderate association, and anything over 0.5 indicates a strong association.

[0079] Also, as shown in Figure 14(a), the y-intercept of the regression equation is 135.82 mmHg, and the cross-tabulation table also shows that the frequency gradient switches from GI to GT between 130 and 139 mmHg. When SBP exceeds 130 mmHg and the nervous control of the arterial pressure control mechanism is enhanced, the power spectrum in the frequency band of 0.03 to 0.05 Hz is enhanced and reflected on the approximation curve. This enhancement occurs when nonlinear components are mixed into the approximation curve, and the approximation solution of the nonlinear components is a positive GT. On the other hand, when the enhancement of the nervous control falls within the linear region, the linear approximation solution is GI. When blood pressure falls below 100 mmHg, nonlinear components are biased toward the low frequency band, and the approximation solution of the nonlinear components is a negative GT.

[0080] The frequency gradient (GT or GI), which is a circulatory system physiological index obtained by the index determination unit 16, is substituted into the above blood pressure estimation formula to obtain an estimated blood pressure value (eSBP) as shown in the flowchart of Fig. 15(a). When the GT or GI values ​​obtained in Figs. 8 to 13 are substituted into the above blood pressure estimation formula, the estimated blood pressures are 109mmHg, 178mmHg, 66mmHg, 133mmHg, 144mmHg, and 170mmHg, respectively, which are close to the actually measured blood pressure values.

[0081] Figure 15(b) shows the correlation between the estimated blood pressure (eSBP) obtained by calculating the frequency gradient (GT or GI) for 28 subjects (healthy subjects) other than the 171 subjects mentioned above and substituting it into the blood pressure estimation formula, and the actual systolic blood pressure (SBP) measured simultaneously for each subject using an upper arm blood pressure monitor. As shown in Figure 15(b), the coefficient of determination is R 2 =0.9293, and in the cross-tabulation table in Figure 15(c), the Cramer's correlation coefficient was 0.881, indicating a very high correlation.

[0082] The 28 subjects were men in their 30s to 80s who were not visiting a hospital for cardiocirculatory system disease. Measurements were taken in the sitting and lying positions (lateral or supine), and data measured over a period of 360 seconds using an electrocardiograph was used.

[0083] The 171 subjects were in a resting sitting position when the ECG was measured, and the measurement positions of the 28 subjects were a sitting position and a lying position (lateral position and supine position). Although the measurement was performed in various positions, the coefficient of determination of the estimation accuracy exceeded 0.9 as described above. Meanwhile, the actual measurement values ​​of the systolic blood pressure used to obtain the above blood pressure estimation formula were all data measured in a sitting position. Therefore, by using the frequency gradient (GT or GI) of this embodiment, blood pressure can be estimated without being limited by the measurement position, and it has robustness that is not easily affected by the measurement environment, and is highly useful for blood pressure measurement other than in a resting state.

[0084] In the cross-tabulation table of Figure 15(c), when the measured systolic blood pressure is less than 100mmHg and equal to or greater than 130mmHg, the frequency gradient is GT, and when the measured systolic blood pressure is in the range of 100mmHg to less than 130mmHg, the frequency gradient is GI. 100mmHg and 130mmHg are the borderline blood pressures where the frequency gradient changes between GT and GI.

[0085] (Evaluation of biological status using circulatory system physiological indices) Heart rate and blood pressure rise when an exercise load is applied, and then gradually return to normal when the exercise load is released and the patient is at rest. The index of blood flow control by these heart rate and blood pressure is the frequency gradient (GT or GI) calculated in this embodiment. Therefore, as shown in Figs. 16(a) and (b), a three-dimensional response surface was created using the data of the above 171 patients (260 cases), with the heart rate and blood pressure on the two horizontal axes and the frequency gradient (GT or GI (indicated as "GT(GI)" in the figure)) on the vertical axis. That is, this is a three-dimensional response surface with the heart rate (HR) and systolic blood pressure (actual measurement value: SBP) as explanatory variables and the frequency gradient (GT(GI)) as the objective variable. Note that Fig. 16(a) is a three-dimensional response surface created using data from a group of healthy subjects (95 patients), and Fig. 16(b) is a three-dimensional response surface created using data from a group of outpatients (76 patients) with heart disease and risk factors.

[0086] Figures 17(a)-(c) show each cross section of the three-dimensional response surface. The GT(GI)-SBP cross section in Figure 17(c) shows a strong correlation, while the SBP-HR scatter plot in Figure 17(a) and the GT(GI)-HR scatter plot in Figure 17(b) show no correlation, and are almost mirror-transformed with respect to the heart rate axis.

[0087] The three-dimensional response surface shown in Figures 16(a) and (b) is expressed by a fourth-order polynomial, and the data from the healthy subjects (95 people) is distributed in the center, which is approximately represented by a straight line, while the data from the outpatient group with heart disease or risk factors (76 people) is widely distributed all the way to the edges of the surface, which is represented by an approximate curve. It was found that this can be used to identify abnormal values ​​related to blood pressure and heart rate.

[0088] For example, in a group of healthy individuals with high blood pressure, where the curve shows nonlinearity, the heart rate fluctuates at 90 beats per minute or less, whereas in a group of healthy individuals with low blood pressure, where the curve shows linearity, the heart rate fluctuates at 60 beats per minute or more.

[0089] On the other hand, in the group of outpatients with high blood pressure, whose curve shows nonlinearity, cases were found where the heart rate fluctuated at 90 / min or more and were scattered on the four corners of the curve shown by a fourth-order polynomial. In the group with low blood pressure, the heart rate fluctuated at 60 / min or less and cases were found scattered on the curve showing nonlinearity, and the GT (GI) value fluctuated greatly when the blood pressure or heart rate fluctuated. Cases scattered on this nonlinear curve indicate the possibility of a disorder in the function of the autonomic nervous system, and indicate that the organs and organs of the biological system are approaching a state of loss of harmony due to modulation or dysfunction. From this, by using the frequency gradient (GT or GI), it is possible to estimate the process in which the functions of the nervous system control and the endocrine system control, including the renin-angiotensin-aldosterone mechanism (RAAS), which is a regulatory function that coexists with the nervous system control, decline together, although humans have the functions of the endocrine system control.

[0090] (Verification using data from the same person) Figures 18 to 20 show the results of a verification experiment using the past year's data on instantaneous heart rate for three males, age 34 (healthy), age 55 (outpatient: taking medication for hypertension), and age 67 (outpatient: taking medication for hypertension), comparing the estimated blood pressure (eSBP) calculated using the above blood pressure estimation formula with the actual brachial systolic blood pressure measured when measuring the instantaneous heart rate. All showed a high coefficient of determination, indicating that the above blood pressure estimation formula calculated from data on 171 subjects (260 cases) is also effective when calculating estimated blood pressure for the same person.

[0091] In addition, in the case of a 34-year-old man shown in Figure 18, the actual measured systolic blood pressure fluctuated within the normal blood pressure range of 100 to 129 mmHg. Cramer's correlation coefficient showed a weak correlation, and there was no independence between the GI and GT items and SBP. Fisher's exact test also showed no significant difference between the fluctuation of the actual measured systolic blood pressure and GI or GT. People who are able to maintain normal blood pressure in this state are considered to have good nervous system control.

[0092] On the other hand, in the case of a 55-year-old male subject and a 67-year-old male subject shown in Figures 19 and 20, the fluctuations in the measured systolic blood pressure were 104 to 164 mmHg and 108 to 150 mmHg, respectively. Both Fisher's exact test and Cramer's correlation coefficient showed a correlation. The correlation coefficient allowed the independence of GT and GI to be tested, and 100 mmHg and 130 mmHg were the boundaries for switching from GT to GI and from GI to GT. GT may be a blood pressure control in which the effects of aging and medication are reflected in the integrated effect. GI is a state in which blood pressure control is performed by nerve control, and is thought to be less affected by aging and medication.

[0093] (Application of Frequency Gradient (GT(GI))) The homeostasis of cardiac output is maintained by maintaining the homeostasis of heart rate and blood pressure. Therefore, the homeostasis of heart rate and blood pressure is determined by GT(GI), which means that the homeostasis of cardiac output is determined by GT(GI). If the fluctuations of heart rate and blood pressure can be captured in the control system related to cardiac function, for example, GT(GI) caused by the Frank-Starling mechanism that constitutes cardiac function can be calculated, and new clinically significant indicators related to circulatory dynamics can be extracted.

[0094] In the above embodiment, the systolic blood pressure is estimated by the instantaneous heart rate measured from the electrocardiogram, but the diastolic blood pressure cannot be estimated in principle. Therefore, in order to calculate the GT (GI) caused by the Frank-Starling mechanism, an approximation formula for the shape and propagation speed of the solitary wave is used to develop an equation that represents the left ventricular pressure waveform, and the left ventricular pressure waveform is created and compared with the measured waveform, the acoustic pulse wave (APW), to extract parameters for estimating the diastolic blood pressure. As described in the above embodiment, the systolic blood pressure can be estimated by using the APW.

[0095] Using an approximate equation obtained by Rayleigh's method for solving the two-dimensional wave generated on the xy plane using a series expansion of complex potential, the left ventricular pressure waveform generated during atrial contraction during the ventricular filling period was created, followed by a left ventricular pressure waveform representing the maximum rate of increase in internal pressure during isovolumic contraction, and a left ventricular pressure waveform during the ventricular ejection period after the aortic valve is released, which were normalized by the systolic blood pressure. Next, a left ventricular pressure waveform was created during the time phase from aortic valve closure to mitral valve release during isovolumic relaxation, and an attempt was made to create the apical pulsation wave (CAB). The approximate equation η(x, t) used to create the left ventricular pressure waveform during the ventricular cycle is expressed by the following equation.

[0096]

number

[0097] ε is the ejection fraction, and h is an index (m) that replaces the mechanical energy of blood flow with the height of potential energy. With a as the left ventricular pressure (mmHg) and the blood flow velocity as c (m / sec), a simulation of the left ventricular pressure waveform for each ventricular cycle shown in Figure 21(a) (Pseudo left ventricular pressure (analysis value: p-LVP) was created using the information in Table 1 for a heart rate of 60 / min, and the data was interpolated within the duration to create a p-LVP approximation curve (Figure 21(b)).

[0098] [Table 1]

[0099] In FIG. 22, the measured ECG, APW, and phonocardiogram (PCG) are plotted, and the p-LVP approximation curve and p-LVP are shown with a dashed line. max is the maximum rate of increase in the internal pressure of the APW (dP / dt max ), suggesting the usefulness of the approximation formula for the pressure waveform η(x, t). The APW shown here is the apical pulsation wave (CAB) measured from the apex of the heart (see Patent Documents 3 and 4), and a heart rate of 58 / min was used.

[0100] (A double-well potential system cardiac circulation model using an approximate curve based on an n-th degree polynomial) The non-periodic component involves the Frank-Starling mechanism and RAAS, while the periodic component involves various neurological regulatory factors. There is the baroreceptor feedback mechanism and central nervous ischemia mechanism, which operate on a second or minute basis, and the chemoreceptor mechanism normalizes arterial pressure in response to a drop in arterial pressure by increasing venous return through the joint action of these neural mechanisms, increasing heart rate and cardiac contractility, and inhibiting arterial blood flow through the constriction of peripheral arterioles. In response to a sudden increase in arterial pressure, it works in the opposite direction, lowering pressure to the normal level. This is the periodic component mentioned above.

[0101] These are the non-periodic components that operate on a minute or hourly basis, and include the renin-angiotensin system, which increases arterial pressure, the stress-relaxation system, which decreases arterial pressure and restores it to normal, and the capillary fluid shift mechanism, which is a mechanism for decreasing and increasing pressure by decreasing and increasing the blood volume in the circulatory system. These are the non-periodic components mentioned above. Arterial pressure is stabilized over the long term by the interaction between the RAAS and the renal-humoral system. This corresponds to the non-periodic components that cause individual differences.

[0102] These control systems are expressed as a double-well potential system model using an approximation curve based on an n-th degree polynomial, and are shown in Figure 23. Figure 24 shows the relationship between the degree of functional enhancement and GT (GI) in the double-well potential system model. In the low blood pressure state represented by "1:GT" in Figure 24, neural control works to suppress the arterial pressure from decreasing any further. The approximation curve of the power spectrum diagram is 1 / f 2 The frequency becomes close to fluctuation and a boundary frequency (BF4) where the main frequency component switches from periodic to aperiodic or vice versa appears near 0.02 Hz.

[0103] The normal blood pressure state represented by "2:GI" in Figure 24 is a bistable system model of "non-perturbation" and "perturbation" due to sympathetic nerve activity, which shows 1 / f type fluctuations. The regulation of the nervous system has relatively large fluctuations, alternating between inhibition and enhancement states.

[0104] The high blood pressure and hypertension states represented by "3: GI" and "4: GT" in Figure 24 are a system in which an approximation curve with an upward sloping trend occurs in the vicinity of 0.04 Hz, suggesting the activation of "perturbation" through neural control.

[0105] The frequency gradient (GT(GI) according to the present invention, which reflects the control state of each system, is calculated as the frequency gradient around the inflection point of an approximation curve expressed by an n-th degree polynomial, and it is suggested that it may become a control index for each system related to maintaining the homeostasis of heart rate and blood pressure. Note that FIG. 25(a) is an image diagram showing heart rate variability, blood pressure variability, and fluctuation variability in time series using the analysis results and measurement data, and FIG. 25(b) is a Lorentz plot diagram of blood pressure variability and heart rate variability created based on FIG. 25(a). [Explanation of symbols]

[0106] 1 Circulatory system physiological index calculation device 11 Biological signal receiving unit 12 Time series fluctuation data calculation section 13 Frequency analysis section 14 Approximate curve calculation section 15 Frequency gradient calculation section 16 Indicator determination section

Claims

1. A circulatory system physiological index calculation device for calculating a circulatory system physiological index correlated with blood pressure and blood flow using at least one of biosignal data of a heart rate and an acoustic pulse wave caused by a left ventricular pressure fluctuation obtained from a body surface, comprising: A biological signal receiving unit that receives the biological signal data; a time-series fluctuation data calculation unit for calculating time-series fluctuation data reflecting time-series fluctuation information of the biological signal data; a frequency analysis unit that plots and displays a power spectrum obtained by frequency analyzing the time-series fluctuation data on a double logarithmic graph; an approximation curve calculation unit that applies an n-th degree polynomial or an n-1-th degree polynomial to a plurality of types of analysis target frequency bands set in a range from VLF to LF while the plot display of the power spectrum is being displayed, and obtains a plurality of approximation curves represented by the n-th degree polynomial or the n-1-th degree polynomial that minimizes the standard deviation of the residual from the plot display of the power spectrum for each analysis target frequency band; a frequency gradient calculation unit that calculates, for each of the approximation curves, a frequency gradient of a tangent to an inflection point, and, if there are two inflection points, a frequency gradient of a straight line between the two inflection points; an index determination unit that calculates a standard deviation of a residual between a tangent to the inflection points or a straight line between the inflection points and a plot display of the power spectrum, compares the standard deviations of the residuals, selects the tangent or straight line with the smallest standard deviation of the residual, and determines a frequency gradient of the selected tangent or straight line as the circulatory system physiological index; A circulatory system physiological index calculation device having the above structure.

2. The frequency band to be analyzed in the approximate curve calculation unit is The range is from any frequency of the power spectrum calculation points in the range of 0.006 to 0.034 Hz, including the fundamental frequency of 0.017 Hz, which is calculated as the time it takes for blood flow to return to a normal level when blood pressure rises due to the acute autoregulation of blood flow, to 0.08 Hz. The circulatory system physiological index calculation device according to claim 1.

3. The approximation curve calculation unit determines whether a histogram of the residual between the approximation curve drawn in the analysis target frequency band and the plot display of the power spectrum is normally distributed, and if the histogram is not normally distributed, adjusts the analysis target frequency band and calculates the approximation curve again. The circulatory system physiological index calculation device according to claim 1.

4. The n-th degree polynomial is a fourth degree polynomial, and the n-1-th degree polynomial is a third degree polynomial. The circulatory system physiological index calculation device according to claim 1.

5. The approximation curve calculation unit is A first approximation curve represented by the third-order polynomial and a second approximation curve represented by the fourth-order polynomial are obtained by setting the analysis target frequency band to a range from any frequency of a power spectrum calculation point in a range of 0.006 to 0.034 Hz ​​to 0.08 Hz, A third approximation curve is obtained by the third-order polynomial in a range from the frequency of the inflection point on the low-frequency side of the two inflection points obtained by the second approximation curve of the fourth-order polynomial to 0.08 Hz. The apparatus for calculating a circulatory system physiological index according to claim 4.

6. The index determination unit A tangent to an inflection point of the first approximation curve; Tangents to two inflection points of the second approximation curve; A straight line between two inflection points of the second approximation curve; Tangent to the inflection point of the third approximation curve From the five straight lines, Select the line with the smallest standard deviation of the residuals from the plot of the power spectrum. The apparatus for calculating a circulatory system physiological index according to claim 5.

7. The index determination unit A standard deviation of a residual of a tangent to each of the inflection points of the first approximation curve, the second approximation curve, and the third approximation curve is calculated within a range of a plot display of a predetermined power spectrum centered on each of the inflection points; For the second approximation curve, the standard deviation of the residual is also calculated for a range of the plot display of a predetermined power spectrum centered on the midpoint of the straight line between the two inflection points. The apparatus for calculating a circulatory system physiological index according to claim 6.

8. A computer program that causes a computer to function as a circulatory system physiological index calculation device that calculates circulatory system physiological indexes that are correlated with blood pressure and blood flow using at least one of biosignal data of a heart rate and an acoustic pulse wave caused by a left ventricular pressure fluctuation obtained from a body surface, the computer program comprising: receiving said biosignal data; A step of obtaining time-series fluctuation data reflecting time-series fluctuation information of the biosignal data; a step of plotting and displaying a power spectrum obtained by frequency analyzing the time-series fluctuation data on a double logarithmic graph; a step of fitting an n-th degree polynomial or an n-1-th degree polynomial to a plurality of types of analysis target frequency bands set in a range from VLF to LF while the plot display of the power spectrum is being displayed, and obtaining a plurality of approximation curves represented by the n-th degree polynomial or the n-1-th degree polynomial that minimizes the standard deviation of the residual from the plot display of the power spectrum for each analysis target frequency band; a step of calculating, for each of the approximation curves, a frequency gradient of a tangent to an inflection point, and, if there are two inflection points, a frequency gradient of a straight line between the two inflection points; a step of calculating a standard deviation of a residual between a tangent to the inflection point or a straight line between the inflection points and a plot display of the power spectrum, comparing the standard deviations of the residuals with each other, selecting the tangent or straight line with the smallest standard deviation of the residuals, and determining a frequency gradient of the selected tangent or straight line as the circulatory system physiological index; A computer program that causes the computer to execute the above.

9. The analysis target frequency band is The range is from any frequency of the power spectrum calculation points in the range of 0.006 to 0.034 Hz, including the fundamental frequency of 0.017 Hz, which is calculated as the time it takes for blood flow to return to a normal level when blood pressure rises due to the acute autoregulation of blood flow, to 0.08 Hz.

9. A computer program product as claimed in claim 8.

10. A histogram of the residual between the approximation curve drawn in the frequency band to be analyzed and the plot display of the power spectrum is judged to be a normal distribution, and if the histogram is not a normal distribution, the frequency band to be analyzed is adjusted, and the approximation curve is calculated again.

10. A computer program according to claim 9.

11. The n-th degree polynomial is a fourth degree polynomial, and the n-1-th degree polynomial is a third degree polynomial.

9. A computer program product as claimed in claim 8.

12. A first approximation curve represented by the third-order polynomial and a second approximation curve represented by the fourth-order polynomial are obtained by setting the analysis target frequency band to a range from any frequency of a power spectrum calculation point in a range of 0.006 to 0.034 Hz ​​to 0.08 Hz, A third approximation curve is obtained by the third-order polynomial in a range from the frequency of the inflection point on the low-frequency side of the two inflection points obtained by the second approximation curve of the fourth-order polynomial to 0.08 Hz.

12. A computer program product as claimed in claim 11.

13. A tangent to an inflection point of the first approximation curve; Tangents to two inflection points of the second approximation curve; A straight line between two inflection points of the second approximation curve; Tangent to the inflection point of the third approximation curve From the five straight lines, From the five straight lines, A straight line having a minimum standard deviation of the residual from the plot display of the power spectrum is selected, and the frequency gradient of the selected tangent line or the straight line is set as the circulatory system physiological index.

13. A computer program according to claim 12.

14. A standard deviation of residuals of tangents to each of the inflection points of the first approximation curve, the second approximation curve, and the third approximation curve is calculated within a range of a predetermined number of plots centered on each of the inflection points; For the second approximation curve, the standard deviation of the residual is also calculated for a range of the plot display of a predetermined power spectrum centered on the midpoint of the straight line between the two inflection points.

14. A computer program according to claim 13.

15. A computer-readable recording medium on which the computer program according to claim 8 is recorded.

Images