JP4415466B2 - Biological signal detection device and non-invasive blood pressure monitor - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a biological signal detection device capable of detecting a biological signal that does not include disturbance from biological information, and a principle / system non-invasive blood pressure monitor that measures only blood pressure fluctuations.
[0002]
[Prior art]
Conventionally, a sphygmomanometer that can continuously measure blood pressure using a cuff is known. However, with this sphygmomanometer, if the measurement interval is shortened, tightening with the cuff is frequently repeated, which places a heavy burden on the subject. Giving and sometimes bleeding inside the cuff.
On the other hand, there is a sphygmomanometer of the principle / method that measures only the blood pressure fluctuation, not directly measuring the absolute value of the blood pressure. For example, there is a sphygmomanometer that captures a parameter by capturing a speed at which a pulse wave travels through an artery (pulse wave propagation speed), a characteristic amount of the pulse wave waveform, and the like, and converts the change into a change in blood pressure. Since this type of sphygmomanometer does not require repeated cuff tightening, it does not place a heavy burden on the subject, and in principle, blood pressure can be calculated from a single heartbeat electrocardiogram or pulse wave signal. It can be performed non-invasively.
[0003]
However, since the above sphygmomanometer (a sphygmomanometer that measures only blood pressure fluctuations) calculates blood pressure based on the amount of change in the pulse wave, if a disturbance is placed on the pulse wave waveform due to body movement (body movement) or the like, There is a problem that accurate blood pressure measurement is not possible.
The following conventional techniques are known as countermeasures.
(1) It has a judgment means for judging the occurrence of body movement noise. When the judgment means judges the occurrence of body movement noise, the reading of the body movement noise is made by waiting for the reading of the sensor signal for a predetermined time. Prevents the blood pressure measurement accuracy from being reduced (see Patent No. 2666983).
[0004]
(2) A method of calculating a basic period of a current pulse wave using a plurality of pulse wave periods and correcting an amplitude of an abnormal pulse wave according to variations in the pulse wave period calculated from the basic period No. 5-184547).
(3) A method for removing disturbance caused by body movement from the biological information of the human body. In other words, the influence of general artifacts including body motion is removed from the pulse wave and the electrocardiogram, and the past pulse wave of 64 beats from the observation time of the pulse wave is seen, and the part where the dispersion of the pulse wave amplitude is large And other signal parts are used for analysis (comparison of the relationship between pulse wave propagation time and blood pressure at different measurement sites, 38th Annual Meeting of the ME Society of Japan Sendai, 21-23 April, 1999 pp.373 ).
[0005]
[Problems to be solved by the invention]
However, the method (1) is not suitable for blood pressure measurement under free action in which body motion noise frequently appears.
Although the method (2) can deal with a single abnormal pulse wave, it is difficult to obtain the fundamental frequency of the pulse wave under free action in which the pulse wave is frequently disturbed. Similarly, it is not suitable for blood pressure measurement under free action.
In the method {circle around (3)}, a portion containing a disturbance such as body movement is not used, so that biological signals that can be used for analysis are limited. In addition, the temporal continuity of the biological signal is lost. Furthermore, if a disturbance such as body movement is applied for a long time, analysis of the biological signal becomes impossible.
As a result of the above, a non-invasive sphygmomanometer that measures only blood pressure fluctuations cannot ensure sufficient blood pressure estimation accuracy when the pulse wave waveform is greatly disturbed by body movement that frequently occurs under free action .
[0006]
The present invention has been made based on the above circumstances, and its first object is to detect a biological signal that does not include disturbance from biological information and maintain the temporal continuity of the biological signal. Providing a detection device. In addition, the second purpose is to separate noise caused by body movement from a pulsating component synchronized with a heartbeat in a non-invasive sphygmomanometer of the principle and method for measuring only blood pressure fluctuations, and accurately and continuously even under free action. It is to enable blood pressure measurement.
[0007]
[Means for Solving the Problems]
The biological signal detection apparatus of the present invention has the following configuration as technical means for solving the above-described problems.
In claim 1, biological information is input as an analog signal from two or more sensors that detect biological information, A / D conversion means for converting the analog signal into a digital signal, and body information detected by the sensor It is considered that disturbance due to motion is included, and only the amplitude changes until the biological information including the disturbance is detected by the sensor, so that the disturbance due to the body motion and the biological information not including the disturbance are mixed. Separation matrix calculation means for calculating one separation matrix for separating a disturbance from biological information including a disturbance by using statistical independence between the biological information not including the disturbance and the disturbance when being deceived And a signal separation calculation means for separating the digital signal into a biological signal not including disturbance and a body motion signal indicating disturbance by applying a separation matrix to the digital signal converted by the A / D conversion means. Obtain.
[0008]
According to a second aspect of the present invention, biological information is input as an analog signal from two or more sensors that detect biological information, A / D conversion means for converting the analog signal into a digital signal, and body information detected by the sensor. Disturbances due to movement are included, and the biological information including the disturbances is linearly changed in frequency and phase until the biological information including the disturbances is detected by the sensor. Two or more separation matrices for separating disturbances from biological information including disturbances by using statistical independence between the biological information not including disturbances and disturbances when it is considered mixed And a separation matrix calculation means for calculating the digital signal, and two or more separation matrices are applied to the digital signal converted by the A / D conversion means to convert the digital signal into a biological signal not including disturbance and a body motion signal indicating disturbance. And a signal separation calculation means for releasing.
[0009]
According to a third aspect of the present invention, the biological information is input as an analog signal from two or more sensors that detect the biological information, and the analog signal is converted into a digital signal. Non-biological information is input and probability density function estimating means for estimating the probability density function of the biological information, and the biological information detected by the sensor includes disturbance due to body movement, and the biological information including the disturbance is Uses a probability density function and does not include disturbances when it is assumed that only the amplitude changes until it is detected by the sensor, and disturbances associated with body movements and biological information that does not include disturbances are mixed. Utilizing statistical independence between biological information and disturbance, separation matrix calculation means for calculating one separation matrix for separating disturbance from biological information including disturbance, and conversion by A / D conversion means By the action of separation matrix into a digital signal, and a signal separation calculation means for separating the body motion signal indicating the biological signal and disturbance without the disturbance of the digital signal.
[0010]
According to a fourth aspect of the present invention, the biological information is input as an analog signal from two or more sensors that detect the biological information, and the analog signal is converted into a digital signal. Non-biological information is input and probability density function estimating means for estimating the probability density function of the biological information, and the biological information detected by the sensor includes disturbance due to body movement, and the biological information including the disturbance is When the frequency and phase change linearly before being detected by the sensor, and it is considered that disturbances associated with body movement and biological information that does not include disturbances are mixed, use the probability density function, Separation matrix calculation means for calculating two or more separation matrices for separating disturbance from biological information including disturbance by using statistical independence between biological information not including disturbance and disturbance, A By the action of two or more separate matrix converted digital signal by the D converting means, and a signal separation calculation means for separating the body motion signal indicating the biological signal and disturbance without the disturbance of the digital signal.
[0011]
In Claim 5, in any one of the biological signal detection apparatuses described in Claims 1-4,
Signal output means for outputting the biological signal separated by the signal separation calculation means to the outside is provided.
[0012]
In claim 6, in any one of the biological signal detection devices according to claims 1 to 5,
A body motion determining means for determining whether or not disturbance due to body motion is included in the biological information input from the sensor, and when it is determined by the body motion determining means that a disturbance due to body motion is not included, A And a signal bypass means for outputting the digital signal converted by the / D conversion means without passing through the separation matrix calculation means and the signal separation calculation means.
[0013]
In Claim 7, In any one of the biological signal detection apparatuses described in Claims 3-5,
When the biological information input from the sensor includes body motion determination means for determining whether or not disturbance due to body motion is included, the probability density function estimation means is configured so that the body motion determination means does not include disturbance due to body motion. When the determination is made, the probability density function of the biological information input from the sensor is estimated.
[0014]
In Claim 8, In the biological signal detection apparatus according to Claim 7,
Signal bypass for outputting the digital signal converted by the A / D conversion means without passing through the separation matrix calculation means and the signal separation calculation means when it is determined by the body movement determination means that no disturbance due to body movement is included. Means.
[0015]
In the ninth aspect, the biological signal detection apparatus according to any one of the first to eighth aspects detects a biological signal that does not include disturbance from the pulse wave (biological information) of the subject.
[0016]
Moreover, the noninvasive blood pressure monitor of this invention has the following structures as a technical means to solve said subject.
In claim 10, at least two or more sensors that are held on the body surface of the subject and detect a pulsation component obtained in synchronization with a heartbeat and output as an analog signal, an analog signal is input from the sensor, and the analog signal The pulsation component detected by the A / D conversion means that converts the signal into a digital signal includes disturbance due to body movement, and the frequency and phase are detected until the pulsation component including the disturbance is detected by the sensor. Statistical analysis of pulsation components that do not include disturbances and disturbances when it is assumed that the pulsation components synchronized with the heartbeat that do not include disturbance are mixed with disturbances due to body movement that change linearly Separation matrix calculation means for calculating a plurality of separation matrices for separating the disturbance from the pulsation component including the disturbance using independence, and a plurality of digital signals converted by the A / D conversion means. A separation matrix is applied to separate the digital signal into a pulsation component synchronized with a heartbeat that does not include disturbances and a body movement component that indicates disturbances, and does not include disturbances captured from the signal separation calculation unit. Blood pressure value calculating means for continuously calculating the blood pressure value based on the pulsation component synchronized with the heartbeat.
[0017]
In claim 11, in the non-invasive blood pressure monitor according to claim 10,
Input / storage means for inputting and storing the blood pressure value of the subject measured separately as a reference blood pressure, and a coefficient for associating the pulsation component synchronized with the heartbeat without the disturbance obtained by the signal separation calculation means and the reference blood pressure The blood pressure value calculating means continuously calculates the blood pressure value from the pulsation component synchronized with the heartbeat that does not include a disturbance taken in from the signal separation calculating means and the coefficient calculated by the coefficient calculating means. It is characterized by calculating.
[0018]
In claim 12, in the non-invasive blood pressure monitor according to claim 10,
Input / storage means for inputting and storing the blood pressure value of the subject measured separately and the propagation time of the pulsation component synchronized with the heartbeat not including disturbance as the reference blood pressure and the reference pulse wave propagation time, and the reference blood pressure and the reference pulse wave propagation Coefficient calculation means for obtaining a coefficient for matching the time, pulse wave propagation time calculation means for calculating the propagation time of the pulsation component synchronized with the heartbeat without disturbance obtained by the signal separation calculation means, and blood pressure value calculation The means is characterized in that the blood pressure value is continuously calculated from the propagation time of the pulsation component taken in from the pulse wave propagation time calculation means and the coefficient calculated by the coefficient calculation means.
[0019]
In claim 13, in the non-invasive blood pressure monitor according to claim 10-12,
A cuff attached to the body of the subject is provided, and the sensor is held at a pressure lower than the diastolic pressure of the subject by the cuff.
[0020]
In claim 14, in the non-invasive blood pressure monitor according to claim 10-13,
Two or more sensors are arranged at a position where a pulsation component obtained in synchronization with a heartbeat can be detected from the same vascular system.
[0021]
In claim 15, in the non-invasive blood pressure monitor according to claim 14,
Two or more sensors are arranged linearly along the same vasculature.
[0022]
In claim 16, in the non-invasive blood pressure monitor according to claim 14,
Two or more sensors are arranged in a circulation direction orthogonal to the same vascular system.
[0023]
[Action]
When the biological information detected by the sensor includes disturbance due to body movement, only the amplitude changes until the biological information including the disturbance is detected by the sensor, and the disturbance and disturbance caused by the body movement are reduced. It is assumed that biological information that does not contain is mixed, and the frequency and phase change by passing through a linear system, and the biological information that does not include disturbance due to body motion and disturbance It can be assumed that it is.
[0024]
When such statistically independent signals were additively mixed with only amplitude changes, and additively mixed through a linear system, they were detected by sensors from an algorithm that estimated the inverse matrix It is possible in principle to recover the original signal by estimating the inverse matrix of biological information and applying this inverse matrix to the biological information (for details of the algorithm, see, for example, “An information-maximization approach to blind separation and blind deconvolution, "Anthony J. Bell and Terrence J. Sejnowsky, Technical Report no. INC-9501, February 1995, Institute for Neural Computation, UCSD, San Diego, CA 92093-0523).
[0025]
Therefore, when there are two sensors, biological information (biological signal) without body motion is s1 (n), disturbance due to body motion (body motion signal) is s2 (n), and sensor signal is x1 (n ), X2 (n) (n: digital signal index, n = 0,1,2,3 ...), when only the amplitude is changed,
x1 (n) = a₁₁× s1 (n) + a₁₂× s2 (n) ………………………… (1)
x2 (n) = a_{twenty one}× s1 (n) + a_{twenty two}× s2 (n) ………………………… (2)
Can be described.
[0026]
Also, when deformed and added through a linear system,
x1 (n) = a_0,11× s1 (n) + a_0,12× s2 (n)
+ A_1,11× s1 (n-1) + a_1,12× s2 (n-1) + ...
+ A_{L, 11}× s1 (n-L) + a_{L, 12}× s2 (n-L) …………… (3)
x2 (n) = a_0,21× s1 (n) + a_0,22× s2 (n)
+ A_1,21× s1 (n-1) + a_1,22× s2 (n-1) + ...
+ A_{L, 21}× s1 (n-L) + a_{L, 22}× s2 (n-L) ………… (4)
Can be described. L is a positive integer of 1 or more.
[0027]
In the case of (1) and (2) above, matrix A (A₁₁= A₁₁, A₁₂= A₁₂, A_{twenty one}= A_{twenty one}, A_{twenty two}= A_{twenty two}) Is the inverse matrix of {x1 (n), x2 (n)}^tWhen acting on (t: transposition), a biological signal and a body motion signal without body motion can be obtained.
In the case of (3) and (4), matrix A₀(A₀₁₁= a_1,11, A₀₁₂= a_0,12, A₀₂₁= a_0,21, A₀₂₂= a_0,22), A₁(A₁₁₁= a_1,11, A₁₁₂= a_1,12, A₁₂₁= a_1,21, A₁₂₂= a_1,22), ..., A_L(A_L11= a_{L, 11}, A_L12= a_{L, 12}, A_L21= a_{L, 21}, A_L22= a_{L, 22}The inverse filter matrix of the filter matrix determined by The inverse filter matrix consists of two or more matrices.
[0028]
When estimating the inverse matrix, a nonlinear function such as g (x) = 1 / {1 + exp (−x)} is used. The function obtained by differentiating the nonlinear function is a probability density function of a biological signal to be obtained. Is most desirable from the viewpoint of signal separation.
However, signal separation can also be achieved with nonlinear functions such as g (x) = 1 / {1 + exp (−x)}.
[0029]
If a pulse wave signal synchronized with a heartbeat that does not include a disturbance component due to body motion is obtained by the above action, the blood pressure can be continuously estimated from the characteristic change of the pulse wave by, for example, the following method.
a) A method of calculating the blood pressure from the change in the waveform of the pulse wave by utilizing the fact that the waveform of the pulse wave in the measured part itself depends on the blood pressure value based on the heartbeat.
b) On the peripheral blood vessel side, the propagation time of the pulse wave is delayed compared to the pulsation of the heart (pulse wave propagation time). The pulse wave velocity with respect to the pulse wave propagation time of a certain distance is expressed as a function of the volume elasticity of the blood vessel. As the blood pressure increases, the volume elasticity of the blood vessel increases, the blood vessel wall becomes harder, and the propagation speed becomes faster. Therefore, if a pulse wave signal synchronized with a heartbeat that does not include a disturbance component due to body motion is obtained, blood pressure fluctuation can be obtained from the pulse wave propagation time.
[0030]
DETAILED DESCRIPTION OF THE INVENTION
Next, an embodiment of the biological signal detection apparatus of the present invention will be described based on the drawings.
(First embodiment)
FIG. 1 is a block diagram showing the configuration of the biological signal detection apparatus.
This biological signal detection apparatus 1 is an apparatus that separates disturbance caused by body movement from biological information detected by two sensors 2 and outputs a biological signal that does not include disturbance. An A / D converter described in detail below 3, a separation matrix calculation unit 4, a signal separation calculation unit 5, and an external output unit 6.
[0031]
The sensor 2 measures a human body's pulse wave as biological information, and is, for example, a well-known optical pulse wave sensor including a light emitting element and a light receiving element. The sensor 2 is attached to, for example, a finger of a human body, and when light is emitted from the light emitting element toward the finger, a part of the light hits a capillary artery passing through the inside of the finger and is applied to hemoglobin in blood flowing through the capillary artery. After being absorbed, the remaining light is reflected and scattered by the capillary artery, and a part of the light enters the light receiving element. At this time, since the amount of hemoglobin changes in a wave manner due to blood pulsation, the amount of light absorbed in the hemoglobin also changes in a wave manner. As a result, the amount of received light reflected by the capillary artery and detected by the light receiving element changes, and the change in the amount of received light is output to the A / D converter 3 as pulse wave information (for example, a voltage signal).
[0032]
The A / D converter 3 receives signals from the two sensors 2 and converts analog signals into digital signals x1 (n), x2 (n), (n = 0, 1, 2,. )) And sent to the separation matrix calculation means 4.
The separation matrix calculation means 4 has a CPU and a memory, and uses a digital signal x1 (n), x2 (n) at the present time sent from the A / D converter 3 to separate 2 × 2 at the present time in the memory. The matrix W (n) is corrected by ΔW (n), and the value of the memory is updated as W (n + 1) = W (n) + ΔW (n). Note that when the apparatus is activated, W (0) of the memory is set in a 2 × 2 unit matrix.
The correction amount ΔW (n) of the 2 × 2 separation matrix is calculated by the following equation.
ΔW (n) = [W (n)^t ]^-1+ [I-2 × g {W (n) × x (n)}] × x (n)^t
g (x) = 1 / {1 + exp (-x)}
However, x (n) = {x1 (n), x2 (n)}^t, (T: transpose)
I is a two-dimensional vector having 1 as a component. I = {1, 1}^t
[0033]
Matrix W (n)^tIs obtained by a standard numerical calculation method such as the LU decomposition method ("Numerical calculation method" by Hideyo Nagashima: see Tsubaki Shoten). The value of the function g is also obtained by a numerical calculation method. The separation matrix W (n + 1) at a certain time corrected in this way is stored in the memory instead of the separation matrix W (n) immediately before, and the original separation matrix W (n) and the A / D converter are stored. The digital signal x (n) converted at 3 is sent to the signal separation calculation means 5.
The signal separation calculation means 5 has a CPU and a memory, and adds the product of the separation matrix W (n) and the digital signal x (n) sent from the separation matrix calculation means 4 backward in the past. Thus, a vector y (n) = {y1 (n), y2 (n)} consisting of a biological signal separated from the body motion and a body motion signal indicating disturbance due to the body motion^tIs sent to the external output means 6.
[0034]
The external output means 6 is a monitor device that outputs a biological signal and a body motion signal from which disturbance due to body motion is separated in synchronization with a sampling frequency of 100 Hz.
As described above, even when a disturbance due to body movement is added to the biological information detected by the sensor 2, only the amplitude changes until the biological information including the disturbance is detected by the sensor 2, and the disturbance due to the body movement. Can be separated from the biological information by using a single separation matrix calculated by the separation matrix calculation means 4.
It should be noted that the frequency and phase change linearly until the biological information including disturbance is detected by the sensor 2, and it is considered that the disturbance accompanying the body motion and the biological information not including the disturbance are mixed. In this case, it is possible to calculate two or more separation matrices by the separation matrix calculation means 4 and to separate the disturbance from the biological information using the two or more separation matrices.
[0035]
(Second embodiment)
FIG. 2 is a block diagram showing a configuration of the biological signal detection apparatus 1.
In the present embodiment, in addition to the configuration of the first embodiment, mode control means 7 for switching the output of the A / D converter 3 according to the setting mode described below, and biological information that does not include disturbance due to body movement A signal probability density function estimating means 8 for estimating a probability density function (histogram) is provided.
The setting mode includes, for example, a training mode and a separation mode that can be selected by operating an external button. When one of the modes is selected, a mode control signal (a training mode signal and a separation mode signal) corresponding to that mode is output. . However, the training mode is performed in a state where the subject is stationary (no body movement).
[0036]
The mode control means 7 is also a body movement determination means of the present invention. When receiving the training mode signal, the mode control means 7 sends the output of the A / D converter 3 to the signal probability density function estimation means 8 and receives the separation mode signal. The output of the / D converter 3 is sent to the separation matrix calculation means 4.
The signal probability density function estimating means 8 has a CPU and a memory, and creates a histogram of the amplitudes of digital signals, for example, x1 (n) sequentially sent from the A / D converter 3 in the training mode. The histogram is created by dividing, for example, 100 between the maximum possible amplitude M and the minimum possible amplitude m of x1 (n), and when entering the divided section with the amplitude of x1 (n), the counter of the divided section is set. Increase by “1”. The creation of the histogram is performed until the mode control signal is changed to the separation mode signal. When the training mode is changed to the separation mode, the estimated histogram (function approximated by a broken line) is sent to the separation matrix calculation means 4.
[0037]
In the separation matrix calculation means 4, a function obtained by integrating the histogram (function approximated by a broken line) sent from the signal probability density function estimation means 8 is used as the function g in the first embodiment.
Other operations are the same as those in the first embodiment.
In this embodiment, a histogram is created from biological information detected when the subject is stationary, and a separation matrix is calculated using this histogram. Therefore, disturbance due to body movement can be accurately separated.
In this embodiment, only the amplitude changes until the biological information including the disturbance is detected by the sensor 2, and the disturbance accompanying the body motion and the biological information not including the disturbance are mixed. When the biological information including disturbance is detected by the sensor 2, the frequency and the phase change linearly, and the disturbance accompanying the body motion and the biological information not including the disturbance are mixed. It goes without saying that it can be applied to both cases of being tricked.
[0038]
(Third embodiment)
FIG. 3 is a block diagram showing a configuration of the biological signal detection apparatus 1.
In this embodiment, in the configuration described in the second embodiment, when the training mode is selected (when the training mode signal is input to the mode control means 7), the subject is in a stationary state. There is basically no disturbance due to body movement. Therefore, in the training mode, as shown in FIG. 3, a bypass signal circuit 9 that sends the output of the A / D converter 3 directly from the mode control means 7 to the external output means 6 may be provided.
[0039]
(Fourth embodiment)
FIG. 4 is a block diagram showing a configuration of the biological signal detection apparatus 1.
In this embodiment, in addition to the configuration of the first embodiment shown in FIG. 1, body movement determination means 10 is provided for determining whether or not the biological information detected by the sensor 2 includes disturbance due to body movement. When the body motion determining means 10 determines that disturbance due to body motion is not included, the bypass signal circuit 9 sends the output of the A / D converter 3 directly from the body motion determining means 10 to the external output means 6. May be provided.
In this case, for example, an acceleration sensor that can detect the presence or absence of body movement can be used as the body movement determination means 10. Alternatively, as described in the second embodiment, a signal detection unit that can detect a signal (a training mode signal in the second embodiment) output when the subject is in a stationary state may be provided.
[0040]
Next, an embodiment of the noninvasive blood pressure monitor of the present invention will be described based on the drawings.
(First embodiment)
FIG. 5 is a block diagram showing the configuration of the noninvasive blood pressure monitor.
(Configuration of non-invasive blood pressure monitor)
As shown in FIG. 5, the non-invasive blood pressure monitor 11 of the present embodiment includes two sensors 2 that detect a subject's pulse wave, a cuff 12 that holds the two sensors 2 at a constant pressure, and a sensor signal (analog signal). An A / D converter 3 for converting the signal into a digital signal, a disturbance separating device 13 for separating disturbance due to body movement (described later), an input / storage means 14 for inputting / storing a blood pressure value of a subject measured separately as a reference blood pressure, Coefficient calculation means 15 for obtaining a coefficient for associating a pulse wave signal synchronized with a heartbeat not including disturbance and a reference blood pressure, a blood pressure value continuously from the calculated coefficient and a pulse wave signal synchronized with a heartbeat not including disturbance Blood pressure value calculating means 16, a recording device 17 for recording the calculated continuous blood pressure, and an output device 18 for outputting the calculated continuous blood pressure to the outside.
[0041]
The sensor 2 is, for example, the optical pulse wave sensor described in the biological signal detection device 1 and is held in close contact with the arm of the subject by the cuff 12. The sensor output may be amplified by an amplifier, and both the sensor 2 and the amplifier have good linearity. In addition, it is desirable that the sensor outputs have a unified requirement for performing the disturbance separation process. That is, if the positions of the plurality of sensors 2 are separated too much, different disturbance factors appear dominantly only in some sensor outputs, and the blood pressure calculation accuracy decreases. As an extreme example, when two sensors 2 are placed on the right arm tip and left arm tip, respectively, and only the right arm is shaken lightly, the sensor 2 attached to the right arm is greatly affected by body movement, but attached to the left arm. The sensor 2 is less susceptible to body movement. Therefore, it is desirable to arrange the individual sensors 2 so that pulse waves can be measured in the same blood vessel system in which the same disturbance factor is easily superimposed on the signal.
[0042]
For example, if two sensors 2 are placed 3cm apart from each other in the length direction of the arm and the arm is shaken lightly, the phase will be out of phase with each other, but a waveform that perfectly matches the disturbance due to body movement is obtained. It is done. That is, uniformity can be obtained with respect to the waveform shape. In addition, when the two sensors 2 are arranged 3 cm apart from each other in the circumferential direction of the arm and the arm is shaken lightly, the phases of each other are in good agreement, but the amplitude of the disturbance is slightly different. In this case, uniformity can be obtained with respect to each other's phase.
The cuff 12 can hold the two sensors 2 at a constant pressure (for example, 20 mmHg) lower than the diastolic pressure of the subject.
[0043]
The A / D converter 3 receives signals from the two sensors 2 and converts analog signals into digital signals x1 (n), x2 (n), (n = 0, 1, 2,. )) And sent to the disturbance separator 13.
The disturbance separation device 13 separates the disturbance y2 (n) {≈s2 (n)} due to body motion from the signals x1 (n) and x2 (n) digitally converted by the A / D converter 3 to thereby remove the disturbance. An apparatus for recovering a pulse wave signal y1 (n) {≈s1 (n)} based on the beat of the heart not included, and is composed of a separation matrix calculation means 4 and a signal separation calculation means 5 as shown in FIG. . The functions and operations of the separation matrix calculation means 4 and the signal separation calculation means 5 are the same as those in the embodiment of the biological signal detection apparatus 1 described above, and a description thereof will be omitted.
[0044]
(Operation of the non-invasive blood pressure monitor 11)
Based on the pulse wave signal y1 (n) based on the heart beat recovered by the disturbance separation device 13, the continuous blood pressure value is calculated as follows.
First, the coefficient calculation means 15 samples, for example, 10 beats of the separated pulse wave signal y1 (n), adds and averages, and synthesizes the reference pulse wave A. An average value Amean is obtained by averaging the peak value Apeak, the bottom value Abbtm, and the DC component from the reference pulse wave A (see FIG. 7).
Separately, a subject's systolic pressure SYS, diastolic pressure DIA, and mean arterial pressure MEANP are measured, and input and stored as a reference value for calibration (reference blood pressure value).
[0045]
Next, a coefficient C that correlates the difference between the peak value Apeak and the bottom value Abtm and the difference between the systolic pressure SYS and the diastolic pressure DIA is obtained.
Thereafter, the continuous blood pressure value BP can be calculated by the blood pressure value calculating means 16 from the digital value y1 (t) of the pulse wave at the current time point where the body motion is obtained from time to time using the following equation.
BP = {y1 (t) -Amean} * C + MEANP
The calculated continuous blood pressure value BP is recorded in the recording device 17 and is output to the outside from the output device 18 as necessary.
[0046]
(Second embodiment)
FIG. 8 is a block diagram showing the configuration of the noninvasive blood pressure monitor 11.
(Configuration of non-invasive blood pressure monitor 11)
As shown in FIG. 8, the non-invasive blood pressure monitor 11 of the present embodiment includes two sensors 2 that detect a subject's pulse wave, a cuff 12 that holds the two sensors 2 at a constant pressure, and a sensor signal (analog signal). A / D converter 3 that converts the signal into a digital signal, a disturbance separation device 13 that separates disturbance due to body movement, a blood pressure value (systolic pressure and diastolic pressure) of the subject measured separately, and a pulse wave propagation time (including disturbance) Input / storage means 19 for inputting / storing a pulsation component propagation time synchronized with a non-beating heart) as a reference value for calibration, and calculating a coefficient corresponding to both from the reference value (blood pressure value and pulse wave propagation time). Coefficient calculation means 20, pulse wave propagation time calculation means 21 for obtaining a pulse wave propagation time from a pulse wave signal based on the beat of the heart from which disturbance has been separated, blood pressure continuously from the calculated pulse wave propagation time and the coefficient Blood pressure value calculating means 22 for calculating a value; Recording device 17 for recording a continuous blood pressure issued, and an output unit 18 for outputting the calculated continuous blood pressure externally.
[0047]
(Operation of the non-invasive blood pressure monitor 11)
The operation up to the disturbance separator 13 is the same as that of the first embodiment. Based on the pulse wave signal y1 (n) based on the heart beat recovered by the disturbance separation device 13, the continuous blood pressure value is calculated as follows.
First, separately, for example, at rest and during exercise, the pulse wave propagation time, systolic pressure, and diastolic pressure are measured, and these measured values are input to the input / storage means 19 as reference values for calibration.
[0048]
Next, the coefficient calculation means 20 obtains a coefficient necessary for calculating the blood pressure value from the pulse wave propagation time. That is, if the blood pressure value and the pulse wave propagation time when each reference value is measured are P1, P2, T1, T2, and the specific constants that are different depending on the subject are α and β, the following equation is established.
P1 = α x T1 + β
P2 = α x T2 + β
[0049]
Once α and β are obtained from the given equation, the pulse wave propagation time T is obtained by the pulse wave propagation time calculation means 21 from the pulse wave waveform of the body motion that is obtained from time to time, and the blood pressure value calculation means 22 is obtained. Thus, the continuous blood pressure value BP can be calculated.
BP = α × T + β
The calculated continuous blood pressure value BP is recorded in the recording device 17 and is output to the outside from the output device 18 as necessary.
In order to obtain the pulse wave propagation time, for example, the delay time of the singular point of the pulse wave may be obtained based on the R wave of the electrocardiographic waveform measured by the electrocardiographic electrode (see FIG. 9).
[0050]
(Effect of Example)
In the non-invasive blood pressure monitor 11 described in the first embodiment and the second embodiment, disturbance due to body motion can be separated from pulse wave information detected by the sensor 2 and a pulsation component synchronized with the heartbeat can be extracted. The blood pressure can be continuously measured even under free action where body movement frequently occurs. In addition, since it is not necessary to repeat tightening with the cuff 12, it is possible to perform continuous blood pressure measurement non-invasively without giving a large burden to the subject.
Furthermore, it does not require mechanical equipment such as a compressor necessary for controlling the cuff pressure and large heavy parts such as a large-capacity battery for driving the cuff pressure. It can also be pumped up as a total.
[0051]
Needless to say, when the blood pressure value of the subject is separately measured as the reference value for calibration, it can be measured using the cuff 12 holding the sensor 2.
However, in the above embodiment, an example is shown in which the sensor 2 is held by the cuff 12, but the cuff 12 is not necessarily used, and the sensor 2 can be held at a constant pressure below the diastolic pressure of the subject. For example, a rubber band or a belt capable of adjusting the holding force may be used.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a configuration of a biological signal detection device (first embodiment).
FIG. 2 is a block diagram showing a configuration of a biological signal detection apparatus (second embodiment).
FIG. 3 is a block diagram showing a configuration of a biological signal detection apparatus (third embodiment).
FIG. 4 is a block diagram showing a configuration of a biological signal detection device (fourth embodiment).
FIG. 5 is a block diagram showing a configuration of a noninvasive blood pressure monitor (first embodiment).
FIG. 6 is a block diagram showing a configuration of a disturbance separation device.
FIG. 7 is an explanatory diagram showing a blood pressure measurement method in the first embodiment.
FIG. 8 is a block diagram showing a configuration of a noninvasive blood pressure monitor (second embodiment).
FIG. 9 is a graph showing a correlation between an electrocardiogram waveform and a pulse waveform.
[Explanation of symbols]
1 Biological signal detector
2 sensors
3 A / D converter (A / D conversion means)
4 Separation matrix calculation means
5 Signal separation calculation means
6 External output means (signal output means)
7 Mode control means (body movement judgment means)
8 Signal probability density function estimation means (probability density function estimation means)
9 Bypass signal circuit (signal bypass means)
10 Body movement judging means
11 Noninvasive blood pressure monitor
12 cuffs
14 Input / storage means (first embodiment)
15 Coefficient calculation means (first embodiment)
16 Blood pressure value calculation means (first embodiment)
19 Input / storage means (second embodiment)
20 Coefficient calculation means (second embodiment)
21 Pulse wave propagation time calculation means (second embodiment)
22 Blood pressure value calculation means (second embodiment)

Claims (16)

A / D conversion means for inputting the biological information from two or more sensors that detect biological information as analog signals and converting the analog signals into digital signals;
The biological information detected by the sensor includes disturbance due to body movement, and the biological information including the disturbance changes only in amplitude until the biological information including the disturbance is detected by the sensor, and includes disturbance and disturbance due to body movement. When it is considered that biological information that is not present is mixed, the disturbance is separated from the biological information that includes the disturbance by using statistical independence between the biological information that does not include the disturbance and the disturbance. Separation matrix calculation means for calculating one separation matrix for
Signal separation calculation means for separating the digital signal into a biological signal not including disturbance and a body motion signal indicating disturbance by applying the separation matrix to the digital signal converted by the A / D conversion means. Biological signal detection device. A / D conversion means for inputting the biological information from two or more sensors that detect biological information as analog signals and converting the analog signals into digital signals;
The biological information detected by the sensor includes disturbance due to body movement, and the frequency and phase change linearly until the biological information including the disturbance is detected by the sensor. When it is considered that the disturbance and the biological information not including the disturbance are mixed, the biological information including the disturbance is obtained using statistical independence between the biological information not including the disturbance and the disturbance. Separation matrix calculation means for calculating two or more separation matrices for separating disturbances from information;
Signal separation calculation for separating the digital signal into a biological signal not including disturbance and a body motion signal indicating disturbance by applying the two or more separation matrices to the digital signal converted by the A / D conversion means. And a biological signal detection device. A / D conversion means for inputting the biological information from two or more sensors that detect biological information as analog signals and converting the analog signals into digital signals;
Probability density function estimation means for inputting biological information not including disturbance due to body movement from the sensor and estimating a probability density function of the biological information;
The biological information detected by the sensor includes disturbance due to body movement, and the biological information including the disturbance changes only in amplitude until the biological information including the disturbance is detected by the sensor, and includes disturbance and disturbance due to body movement. If the biological information that is not included is considered to be mixed, the probability density function is used, and the disturbance is included using the statistical independence between the biological information not including the disturbance and the disturbance. Separation matrix calculation means for calculating one separation matrix for separating disturbance from biological information;
Signal separation calculation means for separating the digital signal into a biological signal not including disturbance and a body motion signal indicating disturbance by applying the separation matrix to the digital signal converted by the A / D conversion means. Biological signal detection device. A / D conversion means for inputting the biological information from two or more sensors that detect biological information as analog signals and converting the analog signals into digital signals;
Probability density function estimation means for inputting biological information not including disturbance due to body movement from the sensor and estimating a probability density function of the biological information;
The biological information detected by the sensor includes disturbance due to body movement, and the frequency and phase change linearly until the biological information including the disturbance is detected by the sensor. When it is considered that disturbance and biological information not including disturbance are mixed, the probability density function is used, and statistical independence between the biological information not including the disturbance and the disturbance is used. Separation matrix calculation means for calculating two or more separation matrices for separating the disturbance from the biological information including the disturbance;
Signal separation calculation for separating the digital signal into a biological signal not including disturbance and a body motion signal indicating disturbance by applying the two or more separation matrices to the digital signal converted by the A / D conversion means. And a biological signal detection device. The biological signal detection device according to any one of claims 1 to 4,
A biological signal detection apparatus comprising signal output means for outputting the biological signal separated by the signal separation calculation means to the outside. The biological signal detection device according to any one of claims 1 to 5,
Body movement determination means for determining whether disturbance due to body movement is included in the biological information input from the sensor;
When it is determined by the body motion determination means that disturbance due to body motion is not included, the digital signal converted by the A / D conversion means is not passed through the separation matrix calculation means and the signal separation calculation means. A biological signal detection device comprising a signal bypass means for outputting. The biological signal detection device according to any one of claims 3 to 5,
Body movement determining means for determining whether or not disturbance due to body movement is included in the biological information input from the sensor,
The probability density function estimating unit estimates a probability density function of biological information input from the sensor when the body motion determining unit determines that a disturbance due to body motion is not included. Signal detection device. The biological signal detection device according to claim 7,
When it is determined by the body movement determination means that disturbance due to body movement is not included, the digital signal converted by the A / D conversion means is not passed through the separation matrix calculation means and the signal separation calculation means. A biological signal detection device comprising signal bypass means for outputting. The biological signal detection apparatus according to claim 1,
The biological signal detection apparatus, wherein the biological information is a pulse wave of a subject. At least two sensors that are held on the body surface of the subject and detect a pulsating component obtained in synchronization with the heartbeat and output as an analog signal;
A / D conversion means for inputting an analog signal from the sensor and converting the analog signal into a digital signal;
The pulsation component detected by the sensor includes disturbance due to body movement, and the frequency and phase change linearly until the pulsation component including the disturbance is detected by the sensor, resulting in body movement. When it is considered that a disturbance and a pulsation component synchronized with a heartbeat not including the disturbance are mixed, the disturbance is statistically independent of the pulsation component not including the disturbance and the disturbance. Separation matrix calculation means for calculating a plurality of separation matrices for separating disturbances from pulsation components including
A signal that separates the digital signal into a pulsation component synchronized with a heartbeat that does not include disturbance and a body movement component that indicates disturbance by applying the plurality of separation matrices to the digital signal converted by the A / D conversion means. Separation calculation means;
A non-invasive sphygmomanometer, comprising: a blood pressure value calculating means for continuously calculating a blood pressure value based on a pulsation component synchronized with a heartbeat that does not include a disturbance taken in from the signal separation calculating means. The non-invasive blood pressure monitor according to claim 10,
Input / storage means for inputting and storing the blood pressure value of the subject separately measured as a reference blood pressure;
Coefficient calculation means for obtaining a coefficient for associating the pulsation component synchronized with the heartbeat that does not include disturbance obtained by the signal separation calculation means and the reference blood pressure,
The blood pressure value calculating means continuously calculates a blood pressure value from a pulsation component synchronized with a heartbeat that does not include a disturbance taken in from the signal separation calculating means and a coefficient calculated by the coefficient calculating means. Non-invasive blood pressure monitor. The non-invasive blood pressure monitor according to claim 10,
Input / storage means for inputting and storing the blood pressure value of the subject separately measured and the propagation time of the pulsation component synchronized with the heartbeat not including disturbance as the reference blood pressure and the reference pulse wave propagation time, respectively;
Coefficient calculating means for obtaining a coefficient for making the reference blood pressure correspond to the reference pulse wave transit time;
A pulse wave propagation time calculation means for calculating a propagation time of a pulsation component synchronized with a heartbeat not including a disturbance obtained by the signal separation calculation means;
The blood pressure value calculating means continuously calculates a blood pressure value from the propagation time of a pulsating component taken in from the pulse wave propagation time calculating means and the coefficient calculated by the coefficient calculating means. Sphygmomanometer. The non-invasive blood pressure monitor according to claim 10,
Having a cuff to be worn on the subject's body;
A non-invasive blood pressure monitor, wherein the cuff holds the sensor at a pressure equal to or lower than the diastolic pressure of the subject. The noninvasive blood pressure monitor according to claim 10,
The two or more sensors are arranged at a position where a pulsation component obtained in synchronization with a heartbeat can be detected from the same vascular system. The noninvasive blood pressure monitor according to claim 14,
The non-invasive blood pressure monitor, wherein the two or more sensors are arranged linearly along the same vascular system. The noninvasive blood pressure monitor according to claim 14,
The non-invasive blood pressure monitor, wherein the two or more sensors are arranged in a circumferential direction orthogonal to the same vascular system.

Images