JP6847489B1 - Blood pressure estimation device, blood pressure estimation method, and blood pressure estimation program - Google Patents

Abstract

The blood pressure estimator has a detection unit that detects the first parameter indicating the length of the heartbeat cycle of the living body, and a first frequency component that does not depend on the characteristics of the baroreceptor of the living body from the change of the first parameter with time. The second frequency component that depends on the characteristics of the baroreceptor of the living body is extracted, and based on the first frequency component, the first resistance ratio due to the first frequency component and the second frequency component are extracted. The second resistance ratio is determined based on the above, and the second parameter related to the resistance of the blood circulation system of the living body is determined based on the determined first resistance ratio and the second resistance ratio, and the determination is made. It is provided with a processing unit for estimating the pressure of blood pressure based on the second parameter.

Description

The present invention relates to a blood pressure estimation device, a blood pressure estimation method, and a blood pressure estimation program.

Obtaining highly accurate daily continuous blood pressure is extremely significant for preventing illness and promoting health. Existing methods for obtaining blood pressure include measurement with a cuff-type sphygmomanometer, a method using a correlation between pulse waveform and blood pressure or a correlation between pulse rate and blood pressure using an optical device, and a method using a circulatory system simulation. However, with the general cuff method for measuring blood pressure, the load due to compression is large and continuous measurement is difficult. In the method based on the correlation between the pulse waveform of the optical device and the blood pressure, the influence of daily body movement is large and an error is likely to occur. On the other hand, the method based on the correlation between pulse rate and blood pressure is inadequate in accuracy. Moreover, since the method for setting parameters has not been established in the method by circulatory system simulation, the conventional method cannot be used as a method for estimating daily continuous blood pressure.

The inventor of the present application has previously used a simplified mathematical model to represent blood flowing through the circulatory system of a living body by a fluid flowing through a flow path in which a plurality of elastically deformable containers are connected in an annular shape, and the pressure (pressure) of blood ( In other words, we have proposed a blood pressure estimation device that estimates blood pressure) (see Patent Document 1). Specifically, the simplified circulatory system model in the circulatory system blood pressure estimation device proposed by the inventor of the present application comprises a circulatory dynamic system model and a circulatory regulation system inverse model. The inverse model of the circulatory regulation system of the blood pressure estimation device proposed above uses the pulse rate as an input value, and the peripheral resistance and the stroke amount are controlled by the low frequency component of the pulse rate change. More specifically, the ratio of the difference between the pressure containers of the fluid in the pair of containers to the flow rate of the fluid between the pair of containers connected to each other among the plurality of containers (in other words, in other words). The target value of (resistance corresponding to peripheral resistance) is determined based on the pulse rate. Further, the blood pressure estimator determines the resistance so that the resistance approaches the determined target value with a delay, and estimates the blood pressure.

Patent No. 6580158

It was confirmed that the blood pressure estimator of Patent Document 1 estimates an appropriate continuous blood pressure with a certain accuracy as compared with the blood pressure actually measured by a cuff type automatic sphygmomanometer. However, there were some parts where the accuracy was not very good. As a result of diligent studies, the inventor focused on the fact that the actual circulatory system includes a circulatory regulation system, which is a very complicated control system including short-term regulation by autonomic nerves and long-term regulation by hormones. The blood pressure estimator of Patent Document 1 does not consider the function of the circulatory regulation system, particularly the short-time regulation function of blood vessels. Therefore, it was concluded that it is necessary to reflect the short-term regulatory function of blood vessels in order to estimate blood pressure with high accuracy.

One of the objects of the present invention is to estimate blood pressure with high accuracy.

On one side,
A detector that detects the first parameter that represents the length of the heartbeat cycle of the living body,
From the change of the first parameter with respect to time, a first frequency component that does not depend on the characteristics of the baroreceptor of the living body is extracted, and a second frequency component that depends on the characteristics of the baroreceptor of the living body is extracted. Based on the frequency component, the first resistance ratio due to the first frequency component and the second resistance ratio based on the second frequency component are determined, and the determined first resistance ratio and second resistance ratio are obtained. Based on this, a processing unit that determines a second parameter related to the resistance of the blood circulation system of the living body and estimates the pressure of blood pressure based on the determined second parameter.
A blood pressure estimation device, a blood pressure estimation method, and a blood pressure estimation program.

Blood pressure can be estimated with high accuracy.

It is a block diagram which shows the structure of the blood pressure estimation apparatus of 1st Embodiment. It is a block diagram which shows the structure of the processing part of FIG. It is explanatory drawing which shows the mathematical model used by the processing part of FIG. It is a graph which shows the change with time of the reference no-load volume which the processing part of FIG. 1 stores. It is a block diagram for demonstrating the circulatory regulation system reverse model used by the processing part of FIG. It is a block diagram which shows the function of the processing part of FIG. It is a flowchart which shows the process which the processing part of FIG. 1 executes. It is a flowchart which shows the process which the processing part of FIG. 1 executes. It is a graph which shows the change with respect to time of the blood pressure estimated by the processing part of FIG. It is a graph which shows the change with time of the blood pressure estimated by the conventional blood pressure estimator. It is a graph which shows the relationship between the change of slow pulse rate change time constant T _{c, and the error of blood pressure.} It is a block diagram which shows the structure of the blood pressure estimation apparatus of 2nd Embodiment. FIG. 13 (A) is a correlation diagram between the low frequency component extracted from the time change of the pulse rate and blood pressure, and FIG. 13 (B) is a correlation diagram between the time change of the pulse rate and blood pressure.

[I. First Embodiment]
Hereinafter, embodiments of the present invention relating to the blood pressure estimation device, the blood pressure estimation method, and the blood pressure estimation program will be described with reference to FIGS. 1 to 8.

In the invention previously proposed, the present inventor has found a method for estimating blood pressure with high accuracy by constructing a circulatory system of a living body with a circulatory dynamic system model and a circulatory regulation system inverse model. The circulatory dynamics model is represented by a simple mathematical model consisting of eight elastic vessels and eight linear resistors connecting them. On the other hand, the circulatory regulation reverse model is represented by a mathematical model in which the time change of the pulse rate is input and the peripheral vascular resistance of the two arteries and the unloaded volume change of the two ventricles are output. Hereinafter, a blood pressure estimation device, a blood pressure estimation method, and a blood pressure estimation program as embodiments will be described.

[1. Constitution]
As shown in FIG. 1, the blood pressure estimation device 1 of the first embodiment includes a detection unit 10 and a processing unit 20. In this example, the blood pressure estimation device 1 is a wristwatch type. The blood pressure estimation device 1 may be of a type different from the wristwatch type (for example, an adhesive plaster type). Further, the detection unit 10 and the processing unit 20 may be integrally configured, or the detection unit 10 and the processing unit 20 may be configured as separate bodies via a wireless line, a wired line, or the like.

The detection unit 10 detects a first parameter representing the length of the cycle of the heartbeat (in other words, the pulse) of the living body. In this example, the living body is a human living body. The living body may be a living body of an animal other than a human.

In this example, the first parameter is the pulse rate. The pulse rate is the number of pulses of the living body per predetermined unit time (1 minute in this example). In this example, the pulse rate may be calculated by dividing the unit time by the cycle length of one pulse. The first parameter may be the cycle length.

In this example, the detection unit 10 may use any means as long as the pulse rate and the cycle length of one pulse can be obtained. For example, the arteries of a living body are irradiated with light, the intensity of the light reflected by the living body is detected, and the pulse rate is detected based on the change in the detected intensity with time. The detection unit 10 detects the pulse rate based on the time between two consecutive peaks in the detected change in intensity.
Further, the pulse rate may be detected from a non-contact sensor such as a video pulse wave, not limited to the wristwatch type contact type.

The detection unit 10 is not limited to optical means, and includes, for example, a member that presses the surface of a living body in the vicinity of an artery (for example, a wrist) and detects the pressure that the member receives from the surface. May detect the pulse rate. In this case, the detection unit 10 may detect the pressure by using a piezoelectric element.

Further, the detection unit 10 may include an electrode attached to the surface of the living body in the vicinity of the heart, and may detect the pulse rate by detecting the potential on the surface via the electrode. In this case, the detection unit 10 may be an adhesive plaster type.

As shown in FIG. 2, the processing unit 20 includes a processing device 21, a storage device 22, an input device 23, and an output device 24, which are connected to each other via a bus BU. The processing unit 20 is an example of an information processing device.

The processing device 21 controls each element constituting the processing unit 20 by executing a program (blood pressure estimation program) stored in the storage device 22. As a result, the processing unit 20 realizes a function described later. In this example, the processing device 21 includes a CPU (Central Processing Unit). The processing device 21 is not limited to the CPU, and may be configured by other known alternative means.

The storage device 22 stores information in a readable and writable manner. In this example, as the storage device 22, known storage means such as a RAM (Random Access Memory), a semiconductor memory, and an organic memory can be used.

The input device 23 inputs information from the outside of the blood pressure estimation device 1. In this example, the input device 23 includes a key-type button. The input device 23 may include a microphone.
The output device 24 outputs information to the outside of the blood pressure estimation device 1. In this example, the output device 24 includes a display. The output device 24 may include a speaker.
The processing unit 20 may include a touch panel type display that constitutes both the input device 23 and the output device 24.
Here, the detection signal from the detection unit 10 is input to the processing unit 20 via an interface (not shown), but may be input via the input device 23.

[2. function]
[2.1 Function of detector]
As described above, the detection unit 10 detects the pulse rate (first parameter).

[2.2 Functions of processing unit]
The processing unit 20 uses the change in the pulse rate detected by the detection unit 10 with respect to time, and the pressure held by the blood flowing through the circulatory system of the blood of the living body based on the circulatory regulation system inverse model and the circulatory dynamic system model. Estimate (in other words, blood pressure).

[2.2.1 Circulatory dynamics model]
First, a circulatory dynamic system model will be described. In this example, the circulatory dynamics model represents blood flowing through the circulatory system of a living body by a fluid flowing through a flow path in which a plurality of elastically deformable containers are connected in an annular shape.

In this example, as shown in FIG. 3, the flow path in the hemodynamic system model is formed by the first to eighth containers FV1 to FV8 and the first to eighth communication pipes FC1 to FC8. To. The flow path may be formed by 9 or more containers.

The first to eighth containers FV1 to FV8 are connected in an annular shape. In this example, the first to seventh containers FV1 to FV7 and the second to eighth containers FV2 to FV8 are connected by the second to eighth communication pipes FC2 to FC8, respectively. The eighth container FV8 and the first container FV1 are connected by a first communication pipe FC1.

In the flow path, the fluid flows from the first to seventh containers FV1 to FV7 to the second to eighth containers FV2 to FV8, respectively. Further, in the flow path, the fluid flows from the eighth container FV8 to the first container FV1.

In this example, the first to eighth vessels FV1 to FV8 are, in order, the left atrium, the left ventricle, the aorta and the artery downstream of the aorta, the vena cava and the vein upstream of the vena cava, the right atrium, and the right. Represents the ventricle, pulmonary artery, and pulmonary vein, respectively.
Each container FV1 to FV8 is a spherical shell that elastically deforms. Of the first to eighth containers FV1 to FV8, the differential dPi / dt with respect to the time t of the pressure Pi of the fluid in the i-th container FVi is expressed by Equation 1. i represents each integer from 1 to 8.

Qi represents the flow rate flowing into the i-th container FVi. Q _i +1 represents the flow rate flowing out from the container FVi of the i. V _i is (in other words, no-load volume) pressure _{P i} of the fluid in the container FVi i th volume of the container FVi of the i in the case of 0 represents a. E _i represents a predetermined coefficient associated with the i-th container FVi. The coefficient E _i may be regarded as a parameter representing the ratio of the change in the volume of the third container FVi with respect to the time to the change in the pressure Pi of the fluid in the third container FVi with respect to time. The coefficient E _i, the amount of fluid flowing from the container FVi of the i per unit time, and the change with time of the no-load volume Vi of the container FVi of the i, per unit time in a container FVi of the i in relation to the amount obtained by subtracting from the amount of fluid flowing, it may be regarded as a parameter representing the ratio of the change with time of the pressure P _i of the fluid in the container FVi of the i.

V _i is the pressure Pi of the container FVi of the i is a volume of 0, referred to as no-load volume.
Of container FV1~FV8 the first to eighth, second and sixth container FV2, FV6 other container FV1 corresponding to the left and right ventricles, FV3～FV5, FV7, unloaded volume FV8 _{V 1} , V _{3 to} V ₅ , V ₇ , V ₈ do not change with time. On the other hand, the no-load volume _V 2, _{V 6} of the container FV2, FV6 the second and sixth, as represented in Equation 2, time-varying. That is, the change with time of the second and sixth container FV2, unloaded volume V ₂ of _FV6, V ₆ may be regarded as representing the beats of the left and right ventricles.

a (t) is the ratio of the unloaded volume change of the left and right ventricles, which represents the magnitude of the cardiac output of the living body, to the value at the reference pulse rate, which is determined by the processing unit 20 (in other words, the unloaded single heart). Cardiac output ratio (pulse amplitude)). In this example, the unloaded single cardiac output ratio is the ratio of the difference between the minimum value and the maximum value in one pulse to the value at the reference pulse rate of the signal representing the amount of blood in the artery. The unloaded single cardiac output ratio is a signal representing the width of the artery, the cross-sectional area of the artery, the amount of blood in the artery, the flow rate of blood in the artery, the flow velocity of blood in the artery, or at least one of these. May be the ratio of the difference between the minimum value and the maximum value in one pulse to the value at the reference pulse rate. The unloaded single cardiac output ratio a (t) is determined by the inverse model of the circulatory regulation system described later. The no-load volume other than the left and right ventricles did not change with time.
b represents the pulse rate detected by the detection unit 10. τ represents the time (in other words, the time within the cycle) from the time when the cycle starts in one pulse cycle.

f _i represents the no-load volume _{V i.} No-load volume f _i is unloaded once cardiac output ratio a, pulse rate b, and has a predetermined value depending on the time tau. In this example, no-load volume _V i (t) is represented by Equation 3. In this example, the no-load volume _{V 2} of the second container FV2 (left ventricle), a sixth container FV6 change _V 2 (t) versus time for the no-load volume _{V 6} (right _{ventricle), V} 6 ( t) is determined based on the unloaded single cardiac output ratio a and the time change b (t) of the pulse rate.

Here, τ (t) represents the elapsed time from the start time of each pulse including the time t. b (t) represents the pulse rate of the pulse included in the time t, and b ₀ represents the reference pulse rate. _fi0 (τ) represents the time variation of the unloaded volume of the left or right ventricle or the reference unloaded volume at the reference pulse rate. In this example, the second and sixth container FV2, criteria for FV6 unloaded volume _{f 20, f 60} is the curve VL in FIG. 4, respectively represented by VR.

As expressed by Equation 4, the flow rate Q ₉ flowing out of the eighth container FV 8 is equal _{to the flow rate Q 1} flowing out of the first container FV 1. The flow rate _{Q i} flowing into the container FV1~FV8 the first to eighth, respectively, may be regarded as a flow rate in the communicating pipe FC1~FC8 the first to eighth.

Flow rate _{Q i} of each communication pipe FC1~FC8 is represented by Equation 5.

In the function C _i , as represented by Equation 6, the pressure _Pi-1 of the upstream container connected to the i-th communication pipe FCi is connected to the downstream communication pipe FCi of the i-th communication pipe FCi. If less than the pressure P _i of it represents 0. Further, in the function C _{i , as represented by Equation 6, the pressure Pi-1} of the container on the upstream side connected to the communication pipe FCi of the i-th is connected to the communication pipe FCi of the i-th communication pipe FCi on the downstream side. of the case greater than or equal to the pressure P _i of the vessel represents one. The function C may be regarded as representing a valve (in other words, a check valve) that prevents the fluid from flowing back. Indexes i = 2,3,6,7 correspond to mitral valve, aortic valve, tricuspid valve, and pulmonary valve, respectively. There is no possibility of backflow with other resistors.

_{As represented by Equation 7, the pressure P 0} of the upstream container connected to the first communication pipe FC1 is equal to _{the pressure P 8} of the downstream container connected to the eighth communication pipe FC 8.

R _i represents the container of the i upstream connected to the communicating tube FCi of resistance to flow of fluid to the i communicating pipe FCi linked downstream of the container of the (resistance factor). Resistor R _i is, of the plurality of containers FV1～FV8, seen as for fluid flow between a pair of containers which are connected to one another, the ratio of the difference between the container pressure with the fluid at the pair of container You can.

Of the first to eighth communicating pipe FC1~FC8 of the fourth and eighth communicating pipe FC4, FC8 than communicating pipe FC1～FC3, resistors _R 1 _{to R} 3 for _FC5~FC7, R 5 ~R ₇ Is not changed with time, as expressed in Equation 8.

On the other hand, among the communicating pipe FC1~FC8 the first to eighth resistor _R 4, _{R 8} for the fourth and eighth communicating pipe FC4, FC8 of, as represented by Equation 9, with respect to time Change. Resistors _R 4, _{R 8} for the fourth and eighth communicating pipe FC4, FC8 of may be regarded as peripheral vascular resistance. Specifically, the resistance R ₄ is the resistance of the peripheral blood vessels that communicate between the aorta and the arterial FV 3 downstream of the aorta and the vein FV 4 upstream of the vena cava and the vena cava, and the resistance R ₈ Is the resistance of the peripheral blood vessels that communicate between the pulmonary artery FV7 and the pulmonary vein FV8.

Resistor R _{i (i} = 4,8) is the peripheral resistance R _i0 o'clock reference pulse rate b _0, it was extracted from the time variation of the pulse rate, the frequency component that does not depend on the characteristics of the baroreceptor biological (first The value obtained by multiplying the peripheral resistance ratio r (t) based on the frequency component (frequency component) and the frequency component (second frequency component) that depends on the characteristics of the baroreceptor of the living body is multiplied by the unloaded cardiac output ratio a (t). Obtained by dividing by. The determination of the peripheral resistance ratio r (t) is determined by the inverse model of the circulatory regulation system described later. For the peripheral vascular resistance value Ri ₀ at the reference pulse rate, a predetermined value is used. Therefore, the resistance _R 4, _{R 8} for the fourth and eighth communicating pipe FC4, FC8 of the peripheral resistance ratio r (t), is determined according to the ratio of the no-load once cardiac output a (t) Value.

[2.2.2 Circulation regulation system reverse model]
Next, the inverse model of the circulatory regulation system will be described. The inverse model of the circulatory regulation system is simplified in FIG. In this example, as shown in FIG. 5, the circulatory regulation reverse model is represented by the process surrounded by the alternate long and short dash line, and the time change b (t) of the pulse rate is input as the peripheral resistance ratio r (t). ) And the unloaded one-time cardiac output ratio a (t) are output. Furthermore, the output of the circulatory regulatory system inverse model is an input hemodynamics system model (Equation 9 and Equation 3), peripheral vascular resistance R ₄ of the body artery and pulmonary artery _(t),
R ₈ (t) and the no-load volume changes V ₂ (t) and V ₆ (t) of the left and right ventricles are finally output.
In the actual circulatory regulation system, the signal from the autonomic nerves from the brain or the signal from the brain that promotes hormone secretion is used as an input variable, and the pulse rate and peripheral resistance are output variables. It can be said that it is the reverse system.

In the actual circulatory regulation system, very complicated control is performed, but in this example, a simple inverse model of the circulatory regulation system was constructed using only the following basic characteristics of the circulatory regulation system. ..
Characteristic 1: The circulatory regulation system controls to keep the blood pressure constant. Characteristic 2: The baroreceptor response of blood pressure has a differential characteristic and responds well to short-term changes. Characteristic 3: For an increase in pulse rate. Focusing on these characteristics, the parameters of the inverse model of the circulation regulation system are the rate of change of the peripheral resistance coefficient with respect to the change of the low frequency component of the pulse rate s _r (corresponding to characteristic 1). _{The time constant T c} (corresponding to characteristic 2) of the low-pass filter characteristic in which the pulse rate is input in the circulation regulation inverse system model, and the rate _{of change s a} of the ventricular volume with respect to the change in pulse rate s a (corresponding to characteristic 3). Using these parameters, the resistance of the mathematical model of the circulatory dynamics based model R _i and for determining a no-load volume Vi, peripheral resistance ratio r (t) and unloaded once cardiac output ratio a (t ) Is determined.

<(1) Determination of peripheral resistance ratio r (t)>
_{First, using the time constant T c} of the low-pass filter characteristic that inputs the pulse rate in the inverse model of the circulation regulation system, which is the parameter corresponding to the characteristic 2, _{the low frequency component b LF} of the change of the pulse rate with time is used. Find t).
The pulse rate of a living body changes during the day due to a decrease in blood pressure at night and fluctuations in blood pressure during the day. For example, when the pulse rate of a living body per second is plotted over 24 hours, the pulse rate can be roughly divided into a slow change for a long time and a rapid change for a short time. From a frequency point of view, a long, slow change in pulse rate can be described as a low frequency component of the time change in pulse rate, and a short, rapid change in pulse rate is a high frequency of time change in pulse rate. It can be expressed as an ingredient. From the above characteristics, the baroreceptor that senses the change in blood pressure in the living body responds to the high frequency component of the time change of the pulse rate and exerts the blood pressure regulating function. Therefore, it is considered that the influence of the high frequency component on the estimated blood pressure is small. On the other hand, baroreceptors do not respond rapidly to low frequency components. Therefore, it is considered that the estimated blood pressure includes the influence of low frequency components.

Therefore, the low frequency component is extracted from the time change of the pulse rate. In this example, a low-pass filter that passes through a low frequency band is applied to a signal that represents a change in pulse rate over time. In this example, the process of extracting the time change from the pulse rate detected by the detection unit 10 and the process of extracting the low frequency component from the time change of the pulse rate are performed in parallel. _{In this example, as represented by Equation 10, the low frequency component b LF} (t) (first frequency component) is conveniently used by a second-order low-pass filter with the time change b (t) of the pulse rate as an input. Is output. ω _c is the break point frequency. The break point frequency means a reference frequency of a low-pass filter capable of suppressing a high frequency component which is a sudden change and extracting a low frequency component which is a slow change. Further, T _c = 1 / ω _c is a time constant of slow pulse rate change (referred to as “slow pulse rate change time constant”).

The time constant T _c is within the range that satisfies the condition that the average error between systolic blood pressure and diastolic blood pressure becomes small and that the waveform of the 24-hour change of the blood pressure calculation result appropriately represents the characteristics of the measurement result. It was determined that the value was between 10 and 1000 seconds, preferably between 80 and 1000 seconds, more preferably between 100 and 300 seconds, and optimally around 200 seconds. In this example, a series of processing was performed with _{T c = 200 seconds fixed.}

Next, a parameter corresponding to the characteristic 1, using the rate of change s _r of peripheral resistance coefficient for the change of the low frequency components of the pulse rate, the peripheral resistance coefficient for the change of the low frequency components of the pulse rate by circulating regulatory system Determine the impact of change. _{Here, the influence r LF} (t) of the (first frequency component) low frequency component is within a constant value with the inverse proportional function _{of the low frequency component b LF} (t) of the pulse rate change, as represented by Equation 11. As a minute, it is modeled using a _{weighting coefficient s r} called the rate of change in peripheral resistance. The _{closer s r} is to 1, the stronger the effect of regulation that keeps blood pressure constant, _{and the closer s r} is to 0, the weaker the influence of regulation that keeps blood pressure constant.
Since the pulse rate b (t) _{can be expressed by the sum of the low frequency component b LF} (t) and the high frequency component b _HF (t), the high frequency component b _HF (t) has a pulse wave number b (t) to a low frequency. It can be extracted by subtracting the component b _{LF (t). Therefore, the influence r HF} (t) of the (second frequency component) high frequency component is the pulse rate b (t) (b _LF (t) + high frequency component b _HF (t)) as expressed in Equation 12. Modeled by an inverse proportional function, it means that the effect of regulation that keeps blood pressure constant is strong. In Equation 12, when the high frequency component b _HF (t) is 0 (when b (t) = b _LF (t)), r _HF (t) = 1 and the high frequency component b _HF (t) is positive ( If it has a negative) value, _rHF (t) is less than (larger) than 1. As described above, _rHF (t) represents the influence of the change in peripheral resistance due to the (second frequency component) high frequency component.

From the characteristic 2, it can be seen that the circulatory regulation system controls the blood pressure to be kept constant by responding to the high-frequency component of the pulse rate change by the baroreceptor. Therefore, in blood pressure estimation, it is necessary to consider the effects of both low-frequency and high-frequency components of pulse rate changes. Therefore, the peripheral resistance ratio r (t) is modeled as the product of the effects of the low-frequency component and the high-frequency component of the pulse rate change, as represented by Equation 13. Peripheral resistance ratio r (t) is the second parameter associated with baroreceptor resistance.

<(2) Determination of unloaded cardiac output ratio a (t)>
Further, the unloaded single cardiac output ratio a (t) is calculated using the rate of change in ventricular volume with respect to the change in pulse rate, which corresponds to characteristic 3. The unloaded one-time cardiac output ratio a (t) is called the one-time cardiac output change rate as a linear function of the pulse rate b (t) and a constant value as shown in Equation 14. It is modeled using the weighting factor s _a. As for this weighting coefficient s _a , the closer s _a is to 1, the stronger the regulation of the increase in cardiac output with respect to the increase in pulse rate, and the closer _{s a is to 0, the stronger the regulation of the increase in cardiac output with respect to the increase in pulse rate.} Means weak. The unloaded single cardiac output ratio a (t) is a third parameter relating to the adjustment of cardiac output.

<(3) Determination of the resistance _{R i>}
Substituting the peripheral resistance ratio r (t) determined based on the above formulas 10 to 13 and the unloaded single cardiac output ratio a (t) determined based on the formula 14 into the formula 9. in the resistance _R i (i = 4,8) is determined.
<(4) Determination of the no-load volume _V i (t)>
Cardiac output ratio a (t) no load once determined based on the equation 14, by substituting in equation 2, no-load volume _{V i (t) (i =} 2,6) is determined To.

[2.2.3 Functions of processing unit]
As shown in FIG. 6, the processing unit 20 includes a storage unit 210, an extraction unit 220, a determination unit 230, and an estimation unit 240.

_{Based on Equations 1 to 9, the processing unit 20 has pressures P 1} (t) to P ₈ (t) of the fluid in the first to eighth containers FV1 to FV8 at time (in other words, time point) t. _{Therefore, the process of estimating the pressures P 1} (t + Δt) to P ₈ (t + Δt) of the fluids in the first to eighth containers FV1 to FV8 at the time t + Δt after the predetermined step time Δt elapses from the time t is performed. Execute repeatedly.

The storage unit 210 stores in advance the reference no-load volume _fi0 (τ) and the peripheral vascular resistance _{Ri0 at the reference pulse rate.} The processing unit 20, instead of storing the reference no-load volume f _{i0 (tau),} may store a function to calculate the reference no-load volume f _{i0 (τ).} Similarly, the processing unit 20, instead of the storage of the reference pulse rate at the peripheral vascular resistance R _i0, may store a function for calculating a peripheral resistance R _i0 o'clock reference pulse rate.
Further, the storage unit 210 determines _{the influence of the low frequency component b LF} (t) determined in each process, the influence of the resistance due to the low frequency component b _{LF (t) r LF} (t), and the resistance caused by the high frequency component. effect r _{HF (t),} peripheral resistance ratio r (t), other values such as no-load 1 stroke volume ratio a (t), finally estimated stored value P _i of the blood pressure as a reference value You may.

The extraction unit 220 includes a frequency component extraction unit 221, and the frequency component extraction unit 221 has a first frequency component (low frequency component b) that does not depend on the characteristics of the baroreceptor of the living body from the change of the pulse rate with respect to time. _LF) is extracted. Here, the time change of the pulse rate is a change in the number of pulses of the living body per minute detected by the detection unit 10 every second in a predetermined time (24 hours in this example). For example, by plotting the pulse rate per minute in a time series with the pulse rate as the X-axis and the 24-hour time axis as the Y-axis, the change in the pulse rate with respect to time is acquired. The extraction unit 220 extracts the first frequency component based on the mathematical formula 10.
Further, the extraction unit 220 acquires the time change of the first frequency component.

Determination unit 230, using the obtained pulse rate b (t), based on Equation 1 to 14, to calculate the resistance _{R i} and volume _{V i,} further determines a flow rate _{Q i} based on the resistance _{R i} .. Determining unit 230 determines the resistance R _i on the basis of the first frequency component extracting section 220 extracts.
Specifically, the determination unit 230 includes a frequency resistance ratio determination unit 231, a second parameter determination unit 232, a third parameter determination unit 233, a resistance determination unit 234, a flow rate determination unit 235, and a volumetric time change determination. Includes part 236 and.

Based on the first frequency component (low frequency component) extracted by the extraction unit 220, the frequency resistivity determination unit 231 has the first resistance ratio _rLF (t) caused by the first frequency component and the baroreceptor of the living body. _{The second resistivity rHF} (t) due to the second frequency component (high frequency component) that depends on the characteristics of is determined.

The second parameter determining unit 232 is a second resistance ratio r (t) (peripheral vascular resistance ratio) of peripheral blood vessels based on _{the first resistance ratio r LF} (t) and the second resistance ratio r _{HF (t).} Determine the parameters.

The third parameter determination unit 233 determines the third parameter, which is the unloaded one-time cardiac output ratio a (t), which represents the magnitude of the heartbeat, from the time change b (t) of the pulse rate detected by the detection unit 10. To do.

Resistance determining unit 234, and peripheral vascular resistance ratio r (t) (second parameter), and a resistor _{R i} (fourth and based on the no-load stroke volume ratio a (t) (3 parameters) 8 communicating pipe of FC4, peripheral vascular resistance _R 4 for FC8, _{R 8)} determined.

Flow rate determination unit 235, based on the determined resistance R _i, of a plurality of containers, determining the flow rate Q _i of the fluid between the pair of containers which are connected to each other.

Volume time change determination unit 236, the no-load stroke volume ratio a (t) (third parameter) is based, the change with time of the at least one container of volume of the plurality of containers V _{i (t)} To determine.

Further, the estimation unit 240 sets the time t, the pressures P _{1 to} P ₈ , and the resistors R _{1 to} R ₈ to the initial value t _ini , the initial value P _{1, ini to} P _{8, ini} , and the initial value R _{1. , Ini} ~ R8 _{, ini} respectively.

The estimation unit 240 acquires the pulse rate b (t) at time t based on the pulse rate detected by the detection unit 10. In this example, the estimation unit 240 performs interpolation (for example, linear interpolation) based on the time when the pulse rate is detected by the detection unit 10, the time t, and the pulse rate detected by the detection unit 10. Thereby, the pulse rate b (t) at time t is acquired.

The estimation unit 240 may acquire the pulse rate detected by the detection unit 10 at the time closest to the time t as the pulse rate b (t) at the time t without performing interpolation.

Then, the estimating unit 240, based on the flow rate _{Q i} and the volume _{V i} determined by the determination unit 230 estimates the blood pressure _{P i.}

The estimation unit 240 calculates the _{cycle length τ e} at time t based on the acquired pulse rate b (t) and the mathematical formula 15.

The estimation unit 240 sets the cycle start time t ₀ to the time t. The cycle start time t ₀ is the time at which the cycle of the pulse for each pulse starts.
The estimation unit 240 calculates the in-cycle time τ by subtracting the _{cycle start time t 0 from the time t.}

Further, the estimation unit 240 determines the time based on the pulse rate b (t), the unloaded one-time cardiac output ratio a (t), and the in-period time τ (t) calculated by the estimation unit 240. _{The time derivatives Φ 2} (τ) and Φ ₆ (τ) of the unloaded volumes V2 and V6 of the second and sixth containers FV2 and FV6 at t are calculated. Time differential [Phi _i unloaded volume V _i, as represented in Equation 16 is an example of a change with time of the no-load volume V _i. Calculating the time derivative [Phi _i is an example of a determination of the time derivative [Phi _i.

The storage unit 210, instead of the reference no-load volume f _i0, or in addition to the reference no-load volume f _i0, may store the time derivative of the reference no-load volume f _i0. In this case, the determination unit 230 uses _{the time derivative of the reference no-load volume fi0} stored in the storage unit 210, the pulse number b (t) acquired by the estimation unit 240, and the period calculated by the estimation unit 240. Based on the internal time τ (t), the time derivatives Φ ₂ (τ) and Φ ₆ (τ) of the _{unloaded volumes V 2} and V ₆ of the second and sixth containers FV2 and FV6 at time t are obtained. It may be calculated. In this case, the processing unit 20, instead of the storage of the time derivative of the reference no-load volume f _i0, may store a function for calculating a time derivative of the reference no-load volume f _i0.

As described above, in this example, of the first to eighth containers FV1 to FV8, the no-load volume V of the containers FV1, FV3 to FV5, FV7, FV8 other than the second and sixth containers FV2 and FV6. ₁ , V _{3 to} V ₅ , V ₇ , and V ₈ do not change with time. Thus, out of the container FV1~FV8 the first to eighth, second and sixth container FV2, FV6 other container FV1, FV3~FV5, FV7, unloaded volume _V 1 of the _FV8, V 3 ~V ₅ , V ₇ , V ₈ time differentials Φ ₁ (τ), Φ ₃ (τ) to Φ ₅ (τ), Φ ₇ (τ), Φ ₈ (τ) are 0.

By applying the fourth-order Runge-Kutta method to the differential equation represented by the equation 1, the estimation unit 240 uses the first to eighth containers at the time t + Δt after the step time Δt elapses from the time t. _{The pressures P 1} (t + Δt) to P ₈ (t + Δt) of the fluid in FV1 to FV8 are calculated. Calculation of the pressure _{P 1 (t + Δt) ~P} 8 (t + Δt) are an example of estimation of the pressure _{P 1 (t + Δt) ~P} 8 (t + Δt). As the fourth-order Runge-Kutta method, for example, the well-known method described in Japanese Patent No. 6580158 may be used.

Estimation unit 240, after estimating the pressure _{P 1 (t + Δt) ~P} 8 (t + Δt), the time t, to update the time t + Delta] t after a lapse of step time Delta] t from the time t.
The estimation unit 240 uses the time differential Φ ₂ (τ), Φ ₆ (τ), and the pressure P of the in-period time τ, the resistance R ₄ (t + Δt), R ₈ (t + Δ t), the no-load volumes V ₂ , and V _{6. The in-} period processing including the calculation of 1 (t + Δt) to P ₈ (t + Δt) and the update of the time t is repeatedly executed while the in-period time τ is less than or equal to the period length τe.

When the in-cycle time τ becomes larger than the cycle length τe, the estimation unit 240 _{re-acquires the pulse rate b (t), calculates the cycle length τe, and sets the cycle start time t 0} , and then performs the cycle start time t 0 again. , Perform the in-period processing again.

In this example, the processing unit 20, for each time t is updated, and time t, the pressure _P 1 in the time t and (t) _{to P} 8 (t), in association with. Further, in this example, the processing unit 20 outputs (for example, displays on a display) the latest calculated blood pressure via the output device 24 every time a predetermined display cycle (for example, 5 seconds) elapses. In this example, the processing unit 20 outputs the pressure P ₃ of the fluid in the third container FV3 as blood pressure.

The estimation unit 240 is the unloaded volume of the second and sixth containers FV2 and FV6 in the period between the time within the period τ and the time τ + Δt after the step time Δt elapses from the time within the period τ. A constant value (for example, the time derivative Φ _i (τ) at the _{time τ in the period) may be used as the time derivative of V 2} and V _6.

[3. motion]
Next, the operation of the blood pressure estimation device 1 will be described.
The detection unit 10 detects the pulse rate every time a predetermined detection cycle (for example, 1 second) elapses.
Further, the processing unit 20 executes the processing shown in FIG. 7. Hereinafter, the processing of FIG. 7 will be described.

The processing unit 20 sets the time t, the pressures P _{1 to} P ₈ , and the resistors R _{1 to} R ₈ to the initial values t _ini , the initial values P _{1, ini to} P _{8, ini} , and the initial values R _{1, ini.} ~ R _{8 and ini} are set respectively (step S101).
Next, the processing unit 20 acquires the pulse rate b (t) at time t based on the pulse rate detected by the detection unit 10 (step S102).

Next, the processing unit 20 calculates the cycle length τe at the time t based on the pulse rate b (t) acquired in step S102, _{and sets the cycle start time t 0} to the time t (step S103). ..

_{Then, the processing unit 20 calculates a value obtained by subtracting the cycle start time t 0} set in step S103 from the time t as the in-cycle time τ (step S104). Next, the processing unit 20 determines whether or not the in-period time τ calculated in step S105 is equal to or less than the period length τe calculated in step S104 (step S105).

First, a case where the time within the period τ is equal to or less than the period length τe will be described. In this case, the processing unit 20 determines “Yes” in step S105, and proceeds to step S106. Then, in step S106, the processing unit 20 based on the pulse rate b (t), calculates the calculated and volume _{V i} of the resistor _{R i,} further calculates the pressure _{P i.} The process of step S106 will be described later.

_{Next, the processing unit 20 calculates the pressures P 1} (t + Δt) to P ₈ (t + Δt) of the fluids in the first to eighth containers FV1 to FV8 at the time t + Δt after the step time Δt has elapsed from the time t. (Step S107). For the process of step S107, a well-known method, for example, a fourth-order Runge-Kutta method or the like may be used.

After that, the processing unit 20 updates the time t to the time t + Δt after the step time Δt has elapsed from the time t (step S108). Then, the processing unit 20 returns to step S104 and repeatedly executes the processes from step S104 to step S108 until the time within the period τ becomes larger than the period length τe.

When the time within the cycle τ becomes larger than the cycle length τe, the processing unit 20 determines “No” in step S105 and returns to step S102. Then, the processing unit 20 executes the processing after step S102 again for the time t updated in step S108. In this example, when the processing unit 20 proceeds to step S102, the latest time among the times when the pulse rate is detected by the detection unit 10 is earlier than the time t updated in step S108 (future). Wait until it's time.

The process of step S106 of FIG. 7 will be further described.
The processing unit 20 executes the process shown in FIG. 8 as the process in step S106 of FIG. Hereinafter, the process of FIG. 8 will be described. The flowchart of FIG. 8 corresponds to the block diagram of FIG. 5 and 8 are related by the same mathematical formula number, and the processing of the circulatory regulation system inverse model is surrounded by a dash-dotted line in FIG. 8 as in FIG.

_{The processing unit 20 extracts the first frequency component (low frequency component) b LF} (t) from the time change (first parameter) b (t) of the pulse rate (step S201; formula 10 calculation unit in FIG. 5). ..

Next, the processing unit 20 _{calculates the first resistance ratio r LF} (t) caused by the first frequency component based on _{the first frequency component b LF} (t) (step S202; the mathematical expression 11 calculation unit in FIG. 5). ). _{Further, the processing unit 20 calculates the second resistivity rHF} (t) due to the second frequency component (high frequency component) that depends on the characteristics of the baroreceptor of the living body (step S202; the formula 12 calculation in FIG. 5). Department).

Next, the processing unit 20 calculates the peripheral vascular resistance ratio (second parameter) r (t) based on _{the first resistance ratio r LF} (t) and the second resistance ratio r _{HF (t) (step S203;} Formula 13 calculation unit in FIG. 5).

Further, the processing unit 20 has a pulse-free one-time cardiac output ratio a (t) (no-load one-time cardiac output) indicating the magnitude of the heartbeat from the time change b (t) of the pulse rate detected by the detection unit 10. The output (third parameter) is determined (step S204; formula 14 calculation unit in FIG. 5).
It should be noted that step S204 may be performed in parallel with steps S201 to 203, or may be performed before and after steps S201 to S203.

Furthermore, the processing unit 20, based on the no-load 1 stroke volume ratio a (t) calculated in step S204, and peripheral vascular resistance ratio r calculated in step S203 (t), the resistance _{R i} (peripheral resistance _R 4, _{R 8)} determining (step S205; equation 9 arithmetic unit of FIG. 5).

Furthermore, the processing unit 20 based on the resistance _{R i} calculated in step S205, determines the flow rate _{Q i} of a fluid between a pair of containers connected to each other (step S206).

Furthermore, the processing unit 20, based on the no-load 1 stroke volume ratio calculated in the step S204 a (t), the change with time of the volume of the at least one container V _i of the plurality of containers ( t) (V ₄ (t), V ₆ (t)) is determined (step S207; formula 3 calculation unit in FIG. 5).

Then, the processing unit 20 based on the capacitance change _{V i} calculated in the flow rate _{Q i} and step S207 which is calculated in step S206, calculates the pressure _{P i} (step S208).

[4. inspection result]
Using the blood pressure estimation device 1 of the first embodiment of the present invention, the pulse rate of a male volunteer subject in his 60s is measured every 24 hours and 1 second by a commercially available wearable pulse rate meter (Wristable GPS, SF-810, manufactured by EPSON). I measured it. For comparison, a commercially available cuff-type automatic blood pressure monitor (HEM-1025, manufactured by OMRON) is used to measure pulse rate, systolic blood pressure, and diastolic blood pressure in a sitting position for approximately 30 minutes (when waking up) or 1 hour (sleeping). It was done at intervals of hour). The experiment was conducted with the approval of the Ethics Committee of Tohoku University. The differential equation of this model was integrated by a computer using the pulse rate data from the pulse rate meter as an input. The calculation time step was Δt = 0.0002 s. The calculation used a server (HPCT W215s, Intel Xeon Gold 6132, 2.6 GHz 14 Core x 2, 192 GB memory, HPC Tec, Japan). Half and parameter values of the measurement results of the automatic blood pressure meter and compared with the corresponding results of the calculations set (reference pulse _{rate, E 30, R 40, s} r, s a) to variously determine the value of the parameter did. Furthermore, the effectiveness of this blood pressure estimation method was examined by comparing the measurement results of the other half with the calculation results of the determined model parameters.

FIG. 9 shows a 24-hour calculation result (line) and a measurement result (circle) by an automatic sphygmomanometer using the blood pressure estimation device 1 of the present invention. The 24-hour calculation results (lines) represent systolic blood pressure (systolic blood pressure), mean blood pressure, diastolic blood pressure (diastolic blood pressure), and pulse pressure in order from the top. With reference to FIG. 9, the 24-hour calculation result (line) of the blood pressure estimation device 1 and the measurement result of the automatic sphygmomanometer are in good agreement.

In order to show the effectiveness of the present invention, as a comparative example, a 24-hour calculation result (line) by the method described in Japanese Patent No. 6580158 (a method that does not consider the second frequency component (high frequency component) with respect to the time change of the pulse rate). And the measurement result (marked with a circle) by the automatic sphygmomanometer are shown in FIG. 10 in the same format as in FIG. Also in FIG. 10, although they match with a certain accuracy, there are some places where they do not match the measurement results of the automatic blood pressure monitor.

FIG. 11 shows a graph in which the change in the slow pulse rate change time constant T _c is plotted on the horizontal axis and the blood pressure error is plotted on the vertical axis. _{The slow pulse rate change time constant T c} corresponding to FIG. 9 of FIG. 11 is shown by α, and the slow pulse rate change time constant T _c corresponding to FIG. 10 is shown by β.
As shown in FIG. 11, α in consideration of the second frequency component with respect to the time change of the pulse rate showed an improvement in accuracy with respect to β in which it was not considered, and was 7 mmHg or less using the blood pressure estimation device 1 of the present invention. It has become. This error was about the same as the standard value for medical devices.

[5. effect]
As described above, the blood pressure estimation device 1 of the first embodiment detects the first parameter (pulse rate b) representing the length of the heartbeat cycle of the living body. _{Further, the blood pressure estimation device 1 extracts a first frequency component (b LF} (t)) independent of the characteristics of the baroreceptor of the living body from the change (b (t)) of the first parameter with time, and extracts the first frequency component (b LF (t)) of the living body. The second frequency component (high frequency component b _{HF (t)) that depends on the characteristics of the baroreceptor is extracted, and the first resistance ratio (r LF} (t)) due to the first frequency component is based on the first frequency component. ) And the second resistance ratio ( _rHF (t)) based on the second frequency component, and the resistance of the blood circulation system of the living body based on the determined first resistance ratio and second resistance ratio. The second parameter (peripheral vascular resistance ratio r (t)) related to (R _i : particularly R ₄ , R _{8 ) is determined, and the pressure (Pi} ) of blood pressure is estimated based on the determined second parameter. To do.

The first parameter representing the cycle length is more likely to be detected with higher accuracy than the arterial diameter and pulse wave shape. _{Furthermore, the first frequency component (bLF} (t)), which is determined based on the highly accurate first parameter and does not depend on the characteristics of the baroreceptor of the living body, and the second frequency, which depends on the characteristics of the baroreceptor of the living body. The pressure (Pi) of blood pressure is estimated using the component (high frequency component b _{HF (t)).} As a result, it is possible to express both a long-term slow change in the pulse rate from a low frequency component of the time change of the pulse rate and a short-time rapid change in the pulse rate from the high frequency component of the time change of the pulse rate. The behavior of blood flowing through the circulatory system of a living body can be expressed with high accuracy. Thus, the blood pressure estimation apparatus 1 can estimate the pressure P _i which the blood has with high accuracy.

Further, the blood pressure estimation device 1 is connected to each other among a plurality of containers in a mathematical model representing blood flowing through a circulatory system of a living body by a fluid flowing through a flow path in which a plurality of elastically deformable containers are connected in an annular shape. _{The resistance (R i} ), which is the ratio of the difference in pressure of the fluid in the pair of containers between the containers to the flow rate (Q _{i) of the fluid between the pair of containers, and is related to the second parameter.} Determined based on the second parameter, the pressure ( _Pi ) of the blood pressure is estimated based on the determined resistance.

Thereby, the influence on the blood pressure caused by the low frequency component and the high frequency component can be considered based on the low frequency component of the time change of the pulse rate. In addition, the mathematical model can express the behavior of peripheral blood vessels in a living body with higher accuracy than when _{the resistances R 4} and R _{8 are maintained constant with respect to the conventional high-frequency component of the pulse rate.} As a result, the mathematical model can represent the behavior of blood flowing through the circulatory system of the living body with high accuracy. As a result, the blood pressure estimation device 1 can estimate the pressure Pi of the blood with high accuracy.

Further, the blood pressure estimation device 1 of the first embodiment determines and determines a third parameter (no-load single cardiac output a) representing the magnitude of the heartbeat based on the change of the first parameter with time. The resistance is determined based on the third parameter.

By the way, there is a strong correlation between heart rate and resistance in peripheral blood vessels. Therefore, according to the blood pressure estimation device 1, the behavior of peripheral blood vessels in a living body can be expressed with even higher accuracy. As a result, the mathematical model can represent the behavior of blood flowing through the circulatory system of a living body with high accuracy. As a result, the blood pressure estimation device 1 can estimate the pressure Pi of the blood flowing through the circulatory system of the living body with high accuracy.

Moreover, the blood pressure estimation apparatus 1 of the first embodiment, the third on the basis of the parameter change over time of the volume of the at least one container of the plurality of containers _(V i _(t): particularly V _{4 (t)} , V ₆ (t)) is determined.

By the way, there is a strong correlation between the magnitude of the heartbeat and the volumes of the left and right ventricles. Therefore, according to the blood pressure estimation device 1, the mathematical model can represent the behavior of the left ventricle and the right ventricle in the living body with even higher accuracy. As a result, the mathematical model can represent the behavior of blood flowing through the circulatory system of the living body with high accuracy. As a result, the blood pressure estimation apparatus 1, the pressure P _i with the blood flowing through the circulatory system of a living body can be estimated with high accuracy.

Further, in the blood pressure estimation device 1 of the first embodiment, the plurality of containers are the left atrium, the left ventricle, the aorta and the artery downstream of the aorta, the aorta and the vein upstream of the aorta, the right atrium, and the right. It contains eight vessels, each representing the ventricles, pulmonary arteries, and pulmonary veins, and two of the resistances between the vessels are the aorta and the arteries downstream of the aorta and the aorta and upstream of the aorta. It is a peripheral vascular resistance that represents the resistance of a peripheral blood vessel that communicates with a vein, and is a peripheral vascular resistance that represents the resistance of a peripheral blood vessel that communicates between a pulmonary artery and a pulmonary vein.

According to this, the behavior of blood flowing through the circulatory system of the living body and the resistance of peripheral blood vessels can be expressed with high accuracy. As a result, the blood pressure estimation apparatus 1, the pressure P _i with the blood flowing through the circulatory system of a living body can be estimated with high accuracy.

[II. Second Embodiment]
[1. Constitution]
The blood pressure estimation device 1'of the second embodiment includes a storage unit 30 and an extraction unit 40, as shown in FIG. In this example, the detection unit 10'and the blood pressure estimation device 1'are separate bodies, but have the same function as the detection unit 10 of the first embodiment, and detect the pulse rate. In the second embodiment, the blood pressure estimated by the blood pressure estimation device 1 described in the first embodiment described above is stored in the storage unit 30 as a reference value, and this reference value and the low frequency component of the pulse rate or the pulse rate are stored. It is different from the above-described first embodiment in that the blood pressure can be easily estimated without using the mathematical model of the first embodiment.

Since the blood pressure varies depending on various conditions such as the sex, age, and physical condition of the living body, the estimated blood pressure may be stored in the storage unit 30 in association with the various conditions. Such an associated blood pressure can be used as a reference value when adjusting each parameter or when verifying data estimated by another blood pressure estimator. 13 (A) and 13 (B) show the values of the daily calculation result and the measurement result by the cuff type automatic blood pressure. In FIG. 13 (A), the correlation between the low frequency component extracted from the time change of the pulse rate and the blood pressure is obtained by plotting the first frequency component (low frequency component) on the horizontal axis and the blood pressure on the vertical axis in a graph. did. In FIG. 13B, the pulse rate is plotted on the X-axis and the blood pressure is plotted on the Y-axis to obtain the correlation between the time change of the pulse rate and the blood pressure.

In the correlation diagram of FIG. 13 (A), α represented by 〇 or ● is systolic blood pressure (systolic blood pressure), Δ or β represented by filled Δ is diastolic blood pressure (diastolic blood pressure), □ or filled. Γ represented by □ indicates pulse pressure (mean blood pressure), open 〇, △, □ indicate measurement results by cuff-type automatic blood pressure, closed ●, filled △, filled □ indicate calculation results. There is. As shown in the correlation diagram of FIG. 13 (A), the systolic blood pressure value, the diastolic blood pressure value, and the systolic blood pressure value and the diastolic blood pressure value of the measurement result estimated by the blood pressure estimation device 1 of the first embodiment are , It can be seen that it shows a good correlation with the low frequency component. Therefore, if the low frequency component is known based on the correlation, the blood pressure can be easily estimated by combining it with the reference blood pressure, for example, the estimated blood pressure obtained in the first embodiment or the blood pressure obtained by the cuff-type automatic blood pressure. It is possible.
On the other hand, in the correlation diagram of FIG. 13 (A), the correlation between the pulse pressure and the low frequency component is rather poor.

On the other hand, also in the correlation diagram of FIG. 13 (B), it is shown in the same format as that of FIG. 13 (A). As shown in the correlation diagram of FIG. 13B, it can be seen that the pulse pressure estimated by the blood pressure estimation device 1 of the first embodiment and the pulse pressure of the measurement result show a good correlation with the pulse rate. Therefore, if the pulse rate is known based on the correlation, the value of the pulse pressure can be easily calculated by combining it with the reference blood pressure, for example, the estimated blood pressure obtained in the first embodiment or the blood pressure obtained by the cuff-type automatic blood pressure. It is possible to estimate.
On the other hand, in the correlation diagram of FIG. 13B, the correlation between the value of systolic blood pressure, the calculated value of diastolic blood pressure, and the measurement result and the pulse rate is rather poor.

From these results, if the value of systolic blood pressure and the value of diastolic blood pressure are required, the low frequency component is used, and if the value of pulse pressure is required, the pulse rate component is used. The value of systolic blood pressure, the value of diastolic blood pressure, and the value of pulse pressure can be estimated.

[2. motion]
In the blood pressure estimation device 1'of the second embodiment, the extraction unit 40 extracts the pulse rate and the time change of the pulse rate when the pulse rate is input from the detection unit 10', and also from the time change of the pulse rate. Extract low frequency components.
When the information on the reference blood pressure is acquired from the storage unit 30 that stores the blood pressure estimated by the blood pressure estimation device 1 as a reference value and the information on the reference blood pressure is obtained as the pulse pressure, the pulse pressure is used. When calculating the systolic blood pressure and diastolic blood pressure based on the correlated pulse rate, it is possible to easily obtain information on blood pressure with a certain accuracy by calculating based on the low frequency component of the time change of the pulse rate. It will be possible.
In this example, the reference value of the blood pressure is obtained from the storage unit 30 for the blood pressure estimated by the blood pressure estimation device 1, but the reference value is not limited to this and is obtained by using a cuff type sphygmomanometer. You may enter it as blood pressure to get information about blood pressure.

[3. Other application examples]
Further, since the pulse pressure is the difference between the systolic blood pressure and the diastolic blood pressure and the mean blood pulse pressure is the average value of the systolic blood pressure and the diastolic blood pressure, the systolic blood pressure and the diastolic blood pressure can be obtained from the pulse pressure and the mean blood pressure. It may be possible to obtain the accurate systolic blood pressure and diastolic blood pressure from the average blood pressure and the pulse pressure after obtaining the average blood pressure from the correlation with the low frequency component of the pulse rate and the pulse pressure from the correlation with the pulse rate. is there.

[4. effect]
According to this, the pulse rate is acquired from the detection unit 10'without performing the calculation process (steps S202 to S208 in FIG. 8) according to the first embodiment, and the pulse rate is low from the time change by the extraction unit 40. By simply extracting the frequency component (step S201 in FIG. 8), it is possible to easily estimate the systolic blood pressure value, diastolic blood pressure value, and pulse pressure value based on the correlation (slope of the function derived from the correlation). ..

[III. Other perspectives]
In other words, the blood pressure estimation device, blood pressure estimation method, and blood pressure estimation program according to one aspect described above are
From the change (b (t)) of the first parameter with respect to time, a first frequency component (b _LF (t)) that does not depend on the characteristics of the baroreceptor of the living body is extracted, and the above is based on the first frequency component. resistance of the circulatory system of blood of a living body (R _i: particularly R _4, R ₈₎ to the associated second parameter (peripheral vascular resistance ratio r (t)) to determine the blood pressure based on the second parameter the determined A processing unit for estimating the pressure ( _{Pi) of the blood pressure (Pi) is provided.}
Further, the processing unit
Based on the first frequency component, the first resistance ratio ( _rLF (t)) caused by the first frequency component and the second frequency component (high frequency component) depending on the characteristics of the baroreceptor of the living body are obtained. Determine the resulting resistivity ( _rHF (t)) and
The second parameter is determined based on the determined first resistance ratio and the second resistance ratio.
In a mathematical model representing blood flowing through the circulatory system of a living body by a fluid flowing through a flow path in which a plurality of elastically deformable containers are connected in an annular shape, between a pair of containers connected to each other among the plurality of containers. to the flow rate (Q _i) of said fluid, said resistance a ratio of the difference is the resistance between the container pressure the fluid in the pair of container has associated with the second parameter (R _i), the first Determined based on 2 parameters
_{The pressure (Pi} ) of blood pressure is estimated based on the determined resistance.

1 Blood pressure estimation device (first embodiment)
10 Detection unit 20 Processing unit 21 Processing device 22 Storage device 23 Input device 24 Output device 210 Storage unit 220 Extraction unit 221 Frequency component extraction unit 230 Determination unit 231 Frequency resistance ratio determination unit 232 Second parameter determination unit 233 Third parameter determination unit 234 Resistance determination unit 235 Flow rate determination unit 236 Volumetric time change determination unit 240 Estimator unit 1'Blood rate estimation device (second embodiment)
10'Detection unit 30 Storage unit 40 Extraction unit

Claims (10)

A detector that detects the first parameter that represents the length of the heartbeat cycle of the living body,
From the change of the first parameter with respect to time, a first frequency component that does not depend on the characteristics of the baroreceptor of the living body is extracted, and a second frequency component that depends on the characteristics of the baroreceptor of the living body is extracted. Based on the frequency component, the first resistance ratio due to the first frequency component and the second resistance ratio based on the second frequency component are determined.
The second parameter related to the resistance of the blood circulation system of the living body is determined based on the determined first resistance ratio and the second resistance ratio, and the pressure possessed by the blood pressure is determined based on the determined second parameter. Estimating processing unit and
To prepare
Blood pressure estimator. The processing unit
In a mathematical model representing blood flowing through the circulatory system of a living body by a fluid flowing through a flow path in which a plurality of elastically deformable containers are connected in an annular shape, between a pair of containers connected to each other among the plurality of containers. The resistance, which is the ratio of the difference in pressure of the fluid in the pair of containers to the flow rate of the fluid between the containers, and which is related to the second parameter, is determined based on the second parameter.
Estimate the pressure that blood pressure has based on the determined resistance,
The blood pressure estimation device according to claim 1. The processing unit
A third parameter representing the magnitude of the heartbeat is determined based on the change of the first parameter with time, and the resistance is determined based on the determined third parameter.
The blood pressure estimation device according to claim 1 or 2. The processing unit
The change over time in the volume of at least one of the plurality of containers is determined based on the third parameter.
The blood pressure estimation device according to claim 3. The plurality of containers represent the left atrium, the left ventricle, the aorta and the artery downstream of the aorta, the vena cava and the vein upstream of the vena cava, the right atrium, the right ventricle, the pulmonary artery, and the pulmonary vein, respectively. Includes 1 container
Two of the resistances between the vessels are the resistances of the aorta and the peripheral blood vessels that communicate between the arteries downstream of the aorta and the veins upstream of the vena cava. It is a peripheral vascular resistance that represents the resistance of a peripheral blood vessel that communicates between the pulmonary artery and the pulmonary vein.
The blood pressure estimation device according to any one of claims 1 to 3. From the change with time of the first parameter representing the length of the heartbeat cycle of the living body, the first frequency component independent of the characteristics of the baroreceptor of the living body was extracted.
A second frequency component that depends on the characteristics of the baroreceptor of the living body is extracted, and based on the first frequency component, the first resistivity due to the first frequency component and the second frequency component are used. Determine the second resistivity and
Based on the determined first resistance ratio and the second resistance ratio, the second parameter related to the resistance of the blood circulation system of the living body is determined.
A processing unit for estimating the pressure of blood pressure based on the determined second parameter is provided.
Blood pressure estimator. A storage unit that stores the pressure information of the blood pressure estimated by the blood pressure estimation device according to any one of claims 1 to 6 or the blood pressure value obtained by the sphygmomanometer as a reference value.
From the change of the first parameter with respect to time and the change of the first parameter with respect to time, the change with respect to the time of the first parameter or the first frequency component independent of the characteristics of the baroreceptor of the living body is extracted. Extractor and
With
The blood pressure of the living body is estimated from the reference value stored in the storage unit based on the change of the first parameter extracted in the extraction unit with time or the first frequency component.
Blood pressure estimator. The blood pressure estimation device according to claim 7.
The average blood pressure is calculated based on the first frequency component, the pulse pressure is calculated based on the first parameter, and the systolic blood pressure and the diastolic blood pressure are estimated from the calculated average blood pressure and the pulse pressure.
Blood pressure estimator. From the change with time of the first parameter representing the length of the heartbeat cycle of the living body, the first frequency component independent of the characteristics of the baroreceptor of the living body was extracted.
A second frequency component that depends on the characteristics of the baroreceptor of the living body is extracted.
Based on the first frequency component, the first resistance ratio due to the first frequency component and the second resistance ratio based on the second frequency component are determined.
Based on the determined first resistance ratio and the second resistance ratio, the second parameter related to the resistance of the blood circulation system of the living body is determined.
Estimate the pressure that blood pressure has based on the determined second parameter.
Blood pressure estimation method. From the change with time of the first parameter representing the length of the heartbeat cycle of the living body, the first frequency component that does not depend on the characteristics of the baroreceptor of the living body is extracted, and it depends on the characteristics of the baroreceptor of the living body. The second frequency component is extracted, and the first resistance ratio due to the first frequency component and the second resistance ratio based on the second frequency component are determined based on the first frequency component. Based on the determined first resistance ratio and second resistance ratio, the second parameter related to the resistance of the blood circulation system of the living body is determined.
Estimate the pressure that blood pressure has based on the determined second parameter.
A blood pressure estimation program that lets a computer perform the process.

Images