JP2012157423A - Pulse wave signal measurement apparatus, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable body motion noise to be suitably removed from pulse wave signals considering influence of gravity on blood perfusion at a measurement site when performing pulse wave measurement by photoelectric conversion.SOLUTION: A pulse wave signal measurement apparatus is caused to perform a first process and a second process. In the first process, according to any of a plurality of kinds of prescribed algorithm, a pulse wave is measured by removing noise included in a fluctuation component signal in which pulsation of arterial vessel of a subject is reflected, the fluctuation component signal being included in a pulse wave signal representing light reception intensity of reflection light out of light with which a measurement site of the subject is irradiated. In the second process, an algorithm used in the first process is selected according to: good or bad of blood perfusion of the subject determined based on a change in bottom-peak interval of the fluctuation component signal; and the presence or absence of body motion determined based on a magnitude of acceleration acting on the measurement site.

Description

  The present invention relates to a technique for optically measuring a pulse wave.

  Conventionally, a pulse wave measuring method by photoelectric conversion has been used as a general method of measuring a pulse wave in a living body, particularly a human body. In this type of pulse wave measurement method, light having a wavelength that is easily absorbed by blood is emitted from a light emitting element such as a light emitting diode toward a living body, and is reflected by light transmitted through the living body or scattered after entering the living body. Light is received by a light receiving element such as a photodiode or a phototransistor and converted into an electrical signal (hereinafter referred to as a pulse wave signal) to detect a pulse wave. By analyzing the pulse wave signal, the pulse wave is analyzed. Measurement (measurement of pulse interval and pulse rate per unit time) is realized. Here, the pulse interval is a time length corresponding to the RR interval in electrocardiogram. This is because the absorption of light that has entered the living body is greater when the artery is dilated than when the artery is contracted, and the signal level of the pulse wave signal changes according to the pulsation of the artery (see, for example, Patent Documents 1 to 3). ). Patent Document 4 discloses a technology that makes it possible to maintain the S / N ratio even when there is a change in the external environment such as external light. Further, Patent Document 5 discloses a technique that makes it possible to maintain detection accuracy without being affected by the constitution of the subject (specifically, pulsation is small and the AC component of the pulse wave signal is small). Proposed.

JP-A-10-118038 JP 2003-169780 A JP 2007-117591 A JP 2010-004972 A JP 2010-233908 A

  The pulse wave signal includes various noise components such as those caused by the breathing of the subject. In particular, when the subject is performing body movement such as walking, the pulse wave signal includes noise (hereinafter referred to as body movement noise) due to the body movement. Needless to say, when pulse wave measurement is performed by analyzing the pulse wave signal, it is desirable that these noises be removed from the pulse wave signal. The blood flow (hereinafter referred to as blood perfusion) varies depending on the subject according to the constitution of the subject, and local blood perfusion at the measurement site is also affected by gravity. For example, when measuring the pulse wave with the subject's left finger as the measurement site, the pulse wave detected between the posture with the left arm extended horizontally to the same height as the shoulder and the posture with the left arm suspended vertically It is known that a difference occurs in the amplitude of the signal and the S / N ratio also changes. Therefore, when the subject is performing some kind of body movement, it is important to properly remove noise (particularly body movement noise) from the pulse wave signal while taking into account the effect of gravity on blood perfusion at the measurement site. In the techniques disclosed in Documents 1 to 5, no such countermeasure is taken.

  The present invention has been made in view of the above problems, and in the case of performing pulse wave measurement by photoelectric conversion, the pulse wave is measured in consideration of the presence or absence of body movement of the subject and the state of blood perfusion at the measurement site. The purpose is to provide a technology that makes it possible.

  In order to solve the above problems, the present invention is directed to an acceleration measuring means for detecting an acceleration generated according to an external force and outputting an acceleration signal representing the acceleration, and a portion defined as a measurement portion in the body of the subject. A light emitting means for irradiating light of a predetermined wavelength, light reflected from the light emitting means toward the measurement site, light reflected at the measurement site, or light transmitted through the measurement site, Light receiving means for receiving a signal and outputting a signal having an amplitude corresponding to the received light intensity, a fluctuation component signal reflecting the pulsation of the arterial blood vessel of the subject from the output signal of the light receiving means, and the activity of living tissue other than the arterial blood vessel One of a plurality of kinds of algorithms prepared in advance as a signal generating means for extracting the non-variable component signal and for removing noise included in the variable component signal, Selection is made according to the quality of blood perfusion at the measurement site determined by analyzing at least one of the dynamic component signal and the non-variable component signal, and the presence or absence of body movement of the subject determined from the acceleration signal And a pulse wave measuring means for measuring a pulse wave while removing noise contained in the fluctuation component signal in accordance with the algorithm selected by the algorithm selecting means. A wave signal measuring device is provided.

  According to such a pulse wave signal measuring apparatus, an algorithm for removing noise from the fluctuation component signal is selected according to the presence or absence of the subject's body movement and the quality of blood perfusion at the measurement site, and the noise is reduced according to the algorithm. A process of measuring the pulse wave while performing the removal is executed. As a result, an effect is obtained that the pulse wave can be measured while appropriately removing noise according to the presence or absence of body movement of the subject and the state of blood perfusion at the measurement site.

  As the acceleration measuring means, it is conceivable to use a three-axis acceleration sensor that detects the acceleration acting on the measurement site by decomposing it into three axially orthogonal components. In this case, the algorithm selection means may determine that there is body movement when an acceleration signal indicating that acceleration other than gravitational acceleration is working is received. The algorithm for removing noise includes a combination of a filter processing algorithm that smoothes the waveform of the fluctuation component signal and an adaptive algorithm that optimizes the filter coefficient used in the filter processing. If at least two types are prepared, it is determined that there is body movement, and blood perfusion at the measurement site is determined to be poor, the latter algorithm is selected by the algorithm selection means, and in other cases, the former algorithm It is conceivable to let the user select. An example of the filter processing algorithm for smoothing the waveform of the fluctuation component signal is an SG filter processing algorithm, and an example of the adaptive algorithm is an LMS (Least Mean Square) algorithm. Details will be clarified in the detailed description of the invention, but when the subject is performing some body movement and blood perfusion at the measurement site is poor, the SG filter processing algorithm and the filter coefficient used in the filter processing are as follows. This is because body motion noise can be satisfactorily removed by a combination of LMS algorithms to be optimized.

  As a more preferable aspect, the pulse wave signal measuring device includes a data storage circuit, and a fluctuation range of the bottom-peak interval of the fluctuation component signal when blood perfusion at the measurement site changes from a good state to a bad state. And whether or not the perfusion state indicating the ease of change in blood perfusion of the subject is determined in advance from at least one of the fluctuation range of the signal intensity of the non-variable component signal, and determination result data indicating the determination result is stored in the data storage Perfusion state determination means for writing to the circuit, and the pulse wave measurement means switches a control parameter used for waveform smoothing in accordance with the determination result indicated by the determination result data in the filter processing algorithm. For example, when SG filter processing is used as the filter processing, the frame size, which is one of the control parameters in SG filter processing, is used to determine whether or not the perfusion state of the subject is good (the degree of influence of gravity on blood perfusion, as will be described later). It switches according to. Since the magnitude of the effect of gravity on blood perfusion at the measurement site can vary depending on the subject's constitution, etc., according to such an aspect, it is possible to remove noise smoothly in an aspect that conforms to the subject's constitution, etc. It becomes possible.

  As another preferred mode, the acceleration measuring means is mounted in the measurement part or in the vicinity of the measurement part, and the algorithm selection means analyzes the acceleration signal for the direction of gravity acting on the measurement part. Thus, it is possible to determine the quality of blood perfusion at the measurement site in consideration of the identification result. As described above, since blood perfusion is affected by gravity, according to such an aspect, the quality of blood perfusion is determined based on only one (or both) of the fluctuation component signal and the non-fluctuation component signal. The effect that the quality of blood perfusion can be determined more accurately than in the embodiment is obtained.

  In another preferred aspect, the signal generating means amplifies and extracts the signal strength of the fluctuation component signal at a predetermined amplification factor when extracting the fluctuation component signal from the output signal of the receiving means, and the measurement There can be considered a mode in which the pulse wave signal measuring device is provided with an optimization means for adjusting the intensity of the light emitted from the light emitting means toward the measurement site or the amplification factor depending on the quality of blood perfusion at the site. Specifically, if the blood perfusion at the measurement site is in a poor state, depending on the degree of deterioration of the blood perfusion, or so that the bottom-peak interval of the fluctuation component signal becomes the same value as the state of good blood perfusion, A process for increasing the light emission intensity of the light emitting means or the amplification factor is executed by the optimization means. According to such an aspect, there is an effect that a decrease in the signal intensity of the fluctuation component signal (that is, a decrease in the S / N ratio) accompanying the deterioration of blood perfusion at the measurement site is avoided.

  In order to solve the above-mentioned problem, the present invention provides a computer that reflects reflected light from the measurement site out of light emitted toward a predetermined site as a pulse wave measurement site in the body of the subject. The signal component included in the pulse wave signal representing the received light intensity or the received light intensity of the transmitted light that has passed through the measurement site, and the fluctuation component signal reflecting the pulsation of the arterial blood vessel of the subject and the activity of biological tissue other than the artery Depending on whether the subject's blood perfusion is good or bad as determined from at least one of the non-variable component signals reflecting the above and the presence or absence of body movement of the subject as determined from the magnitude of acceleration acting on the measurement site An algorithm selection process for selecting one of a plurality of types of algorithms prepared in advance for removing noise from the signal, and a noise component included in the fluctuation component signal. It provides a program, characterized in that to execute a pulse wave measurement process to measure the pulse wave while removing according to the selected algorithm's in the algorithm selection process, the. According to such a program, it is possible to cause a computer including the light emitting element, the light receiving element, and the acceleration measuring means to function as the pulse wave signal measuring apparatus of the present invention. In addition, as a specific provision mode of the program, a mode in which the program is written and distributed on a computer-readable recording medium such as a CD-ROM (Compact Disk-Read Only Memory) or a memory stick, or via an electric communication line such as the Internet It is conceivable to distribute by downloading.

It is a figure which shows the external appearance of the pulse-wave signal measuring device 1 of one Embodiment of this invention. It is a figure which shows an example of the mounting | wearing aspect of the sensor part 30 of the pulse wave signal measuring device. It is a figure which shows the electrical structural example of the same pulse wave signal measuring device. It is a figure which shows an example of a posture with good blood perfusion, and a posture with bad blood perfusion. It is a figure showing the relationship between the quality of blood perfusion and the signal strength of a fluctuation component signal. It is a figure which shows an example of the individual difference of the influence of gravity with respect to blood perfusion. It is a figure for demonstrating the processing content of an optimization process. It is a figure which shows the example of noise removal by SG filter process. It is a figure which shows the example of noise removal by SG filter process. It is a figure which shows the example of noise removal by the combination of SG filter process and LMS.

(A: Configuration)
FIG. 1 is a diagram showing an appearance of a pulse wave signal measuring apparatus 1 according to an embodiment of the present invention.
As shown in FIG. 1, the pulse wave signal measuring device 1 has a wristwatch structure, a device main body 10 to be worn on the wrist of a subject, and a sensor unit connected to the device main body 10 via a cable 20. 30. As shown in FIG. 1, a wristband 12 is attached to the apparatus main body 10. The pulse wave signal measuring device 1 is attached to the subject's body by winding the wristband 12 around the subject's wrist (left wrist in the example shown in FIG. 1). A display unit 14 such as a liquid crystal display is provided on the surface of the apparatus body 10. The display unit 14 displays the pulse interval calculated from the pulse wave signal detected by the sensor unit 30, the pulse rate per unit time, and the like. A button switch 16 is provided on the outer periphery of the apparatus body 10. The button switch 16 is used for inputting various instructions such as pulse wave measurement start and measurement end, and measurement result reset.

FIG. 2 is a diagram illustrating an example of how the sensor unit 30 is attached to the body of the subject.
As shown in FIG. 2, the sensor unit 30 includes a pulse wave sensor 32 and a sensor fixing band 34. The sensor unit 30 is attached to the subject's body by, for example, winding a sensor fixing band 34 around a portion (hereinafter, a measurement site) between the base of the left index finger of the subject and the second finger joint. In a state where the sensor unit 30 is mounted on the body of the subject, the pulse wave sensor 32 is shielded from external light by the sensor fixing band 34. This is to eliminate noise caused by external light. In the present embodiment, the case where the sensor unit 30 is attached to the left index finger of the subject will be described, but it may of course be attached to another finger such as the middle finger or the ring finger of the left hand. Further, as shown in FIG. 1, in this embodiment, the apparatus main body 10 is attached to the left arm of the subject. However, the apparatus main body 10 may be attached to the right arm. In this case, the sensor unit 30 is attached to the finger of the right hand. It should be installed.

FIG. 3 is a block diagram illustrating an electrical configuration example of the pulse wave signal measuring apparatus 1.
As shown in FIG. 3, the pulse wave signal measurement device 1 includes an analog circuit unit 110, an arithmetic processing circuit 120, a data storage circuit 130, an analog control circuit 140, and an acceleration measurement circuit 150 in addition to the pulse wave sensor 32. It is out. Among the components shown in FIG. 3, components other than the pulse wave sensor 32 are built in the apparatus main body 10. Although not shown in detail in FIG. 3, the display unit 14 and the button switch 16 are connected to the arithmetic processing circuit 120. In addition to the components shown in FIG. 3, the pulse wave signal measuring device 1 also has an audio output unit connected to the arithmetic processing circuit 120 and housed in the device body 10.

  As shown in FIG. 3, the pulse wave sensor 32 includes a light emitting element 32a and a light receiving element 32b. The light emitting element 32 a is, for example, a blue LED (Light Emitting Diode), and radiates light having a wavelength that is easily absorbed by blood with an intensity corresponding to the current value of the current supplied from the apparatus body 10 via the cable 20. In a state where the sensor unit 30 is attached to the body of the subject (see FIG. 2), the light is emitted from the light emitting element 32a toward the measurement site. The light emitted from the light emitting element 32a toward the measurement site in this way enters the measurement site, and then is partially absorbed by the blood flowing through the capillaries in the dermis. Of the light emitted from the light emitting element 32a, a part of the light that is not absorbed by the blood passes through the measurement site, and the rest is received by the light receiving element 32b as reflected light after being scattered by the living tissue. . The light receiving element 32 b is, for example, a photodiode, and outputs a signal having a current value corresponding to the intensity of received light to the apparatus main body 10 via the cable 20.

  The analog circuit section 110 generates a signal component 1110 that generates a fluctuation component signal and a non-fluctuation component signal from the output signal of the light receiving element 32b, and supplies the fluctuation component signal to the arithmetic processing circuit 120. Is included. The fluctuation component signal is a signal component reflecting the pulsation of the arterial blood vessel of the subject among the signal components of the output signal of the light receiving element 32b. On the other hand, the non-variable component signal is a signal component reflecting the activity of other living tissue such as a venous blood vessel, and is mainly a direct current component.

  The drive circuit 1120 controls the light emission by controlling the supply of current to the light emitting element 32a. More specifically, the drive circuit 1120 controls the light emission intensity of the light emitting element 32a by increasing / decreasing the current value applied to the light emitting element 32a in accordance with the light emission intensity control signal LC provided from the analog control circuit 140. The signal generation means 1110 includes an IV conversion circuit 1112, filters 1114a and 1114b, and amplification circuits 1116a and 1116b. The IV conversion circuit 1112 is a current / voltage conversion circuit, generates a pulse wave signal having a voltage corresponding to the current value of the signal received from the light receiving element 32b, and outputs the pulse wave signal to the filters 1114a and 1114b. The filter 1114 a serves as a fluctuation component signal extraction unit that extracts the fluctuation component signal Z from the output signal of the IV conversion circuit 1112. On the other hand, the filter 1114b serves as non-variable component signal extraction means for extracting the non-variable component signal X from the pulse wave signal.

  More specifically, the filter 1114a is a band-pass filter having a frequency band of 0.1 [Hz] to 5 [Hz], and the filter 1114b has a cutoff frequency of 0.1 [Hz]. It is a low-pass filter. That is, in the pulse wave signal measuring apparatus 1 of the present embodiment, the frequency component of 0.1 [Hz] to 5 [Hz] among the various frequency components included in the pulse wave signal is extracted as the fluctuation component signal Z, and 0 A frequency component having a frequency lower than 1 [Hz] is extracted as a non-variable component signal X. The reason why the pass band of the filter 1114a is set to 0.1 [Hz] to 5 [Hz] is that the pulse frequency of a person is generally 1 [Hz] if it is a healthy person, even in the case of a person who has a history of heart disease or the like. This is because it is generally about 3 [Hz]. The reason why the cutoff frequency of the filter 1114b is set to 0.1 [Hz] is to extract a signal component other than the fluctuation component signal from the output signal of the light receiving element 32b as a non-fluctuation component signal. In this embodiment, the pass band of the filter 1114a is set to a frequency band of 0.1 [Hz] to 5 [Hz], and the cutoff frequency of the filter 1114b is set to 0.1 [Hz]. The cut-off frequency is not limited to the above values, and may be set to a suitable value by experiment or the like.

  The amplifier circuit 1116a amplifies the signal level of the fluctuation component signal Z output from the filter 1114a to a level suitable for execution of various arithmetic processes in the arithmetic processing circuit 120, and outputs the amplified signal. Similarly, the amplifier circuit 1116b amplifies the signal level of the non-variable component signal X output from the filter 1114b to a level suitable for execution of various operations in the arithmetic processing circuit 120, and outputs the amplified signal. In the present embodiment, a fixed gain (amplification factor) is used for the amplifier circuit 1116b, but the gain of the amplifier circuit 1116a is changed according to the gain control signal GC given from the analog control circuit 140. Is used. The reason why the gain of the amplification circuit 1116a for amplifying the fluctuation component signal Z is made variable will be clarified later.

  The acceleration measurement circuit 150 detects the acceleration generated according to the external force by decomposing it into components in directions of three axes (x, y and z axes) orthogonal to each other, and outputs an acceleration signal AS indicating the magnitude of each component. To do. The acceleration measuring circuit 150 is built in the sensor unit 30 so that the x-axis coincides with the extending direction of the sensor fixing band 34 in FIG. 1 and the y-axis coincides with the direction indicated by the index finger in FIG. . For this reason, when the subject wearing the pulse wave signal measuring device 1 on the left arm in a manner as shown in FIG. 1 is stationary with the left arm vertically suspended, the acceleration measurement circuit 150 moves in the y-axis direction. An acceleration signal AS indicating that gravity is applied is output. In addition, when the subject is stationary with the left thumb pointing up and the posture where the left arm is horizontally extended to the same height as the shoulder, the acceleration measuring circuit 150 is subject to gravity in the x-axis direction. Is output. That is, in the present embodiment, by analyzing the acceleration signal AS, the direction of gravity acceleration with respect to the three axes of xyz can be specified. Then, when a subject wearing the pulse wave signal measuring device 1 performs body movement such as walking, the acceleration measurement circuit 150 outputs an acceleration signal AS indicating a combined acceleration of the gravitational acceleration and the acceleration generated by the body movement. To do.

  The arithmetic processing circuit 120 includes an A / D conversion circuit, a CPU (Central Processing Unit), a volatile memory such as a RAM (Random Access Memory), and a non-volatile memory such as a ROM (Read Only Memory) (all not shown). Contains. The A / D conversion circuit samples the fluctuation component signal Z and the non-fluctuation component signal X output from the analog circuit unit 110 at a predetermined sampling period, and represents a sample sequence (hereinafter referred to as a digital fluctuation component) representing the waveform of each signal. Signal and non-varying component signal). The non-volatile memory of the arithmetic processing circuit 120 stores a control program for analyzing the fluctuation component signal Z, measuring the pulse wave, and causing the CPU to execute processing for notifying the subject of the measurement result. The volatile memory is used as a work area when executing the control program.

  As described above, the pulse wave signal measuring device 1 according to the present embodiment is a device that measures the pulse wave using the photoelectric conversion described above, and uses the vicinity of the second finger joint from the base of the left index finger of the subject as a measurement site. Yes. In the posture where the left arm is horizontally extended to the same height as the shoulder and the posture where the left arm is vertically suspended, the blood perfusion at the measurement site changes due to the effect of gravity, resulting in a difference in the amplitude of the pulse wave signal, blood As described above, there is an individual difference in the magnitude of the influence of gravity on perfusion. When the removal of body motion noise from the pulse wave signal is not considered at all while considering the influence of gravity on blood perfusion at the measurement site, the S / N ratio is reduced due to the change in blood perfusion. This will interfere with pulse wave measurement. In the present embodiment, this problem is solved by causing a CPU that operates according to the control program to execute processing that significantly shows the features of the present embodiment. Hereinafter, the process executed by the CPU of the arithmetic processing circuit 120 in accordance with the control program will be described in detail.

  The processing executed by the CPU of the arithmetic processing circuit 120 in accordance with the control program includes four processes: a perfusion state determination process, a pulse wave measurement process, an optimization process, and an algorithm selection process. In FIG. 3, the perfusion state determination process is abbreviated as “determination process”, and the pulse wave measurement process is abbreviated as “measurement process”. The contents of these four processes are as follows.

  The perfusion state determination process is a process for determining whether the subject who is the user of the pulse wave signal measuring apparatus 1 is a person with a good perfusion state or a person with a poor perfusion state by analyzing the fluctuation component signal Z. is there. Here, the perfusion state means the ease of changing the blood perfusion of the subject. A person with a good perfusion state means a person who has a large influence of gravity on blood perfusion, and a person with a poor perfusion state means a person who has a small influence of gravity on blood perfusion. In the following, prior to detailed description of the processing content of the perfusion state determination processing, the influence of gravity on blood perfusion will be described with reference to FIG.

  FIG. 4 shows a posture with good blood perfusion at the measurement site (in the example shown in FIG. 4, the posture in which the left arm is horizontally extended to the same height as the shoulder) and a posture with poor blood perfusion (in the example shown in FIG. 4). FIG. 3 is a view showing an example of a posture in which the left arm is hung vertically. When the posture of the subject changes from a posture with good blood perfusion to a posture with poor blood perfusion, blood accumulates in the left hand peripheral portion (that is, the left fingertip) due to the influence of gravity, thereby changing blood perfusion in the peripheral portion. In other words, “poor blood perfusion” refers to a state where blood accumulates in the peripheral portion and blood flow is inhibited by congestion of living tissue, etc., and “the state where blood perfusion is good” refers to the peripheral portion. It is a state where blood is not collected and blood is flowing smoothly.

  FIG. 5 shows a photoelectric conversion system pulse when the posture is changed from a posture in which the left index finger is a measurement site and the left arm is horizontally extended to the same height as the shoulder to a posture in which the left arm is vertically hung. It is a figure which shows an example of the change of the pulse strength measured by wave measurement (the height difference of the trough and the peak of the waveform of the fluctuation component signal Z: hereinafter, the bottom-peak interval). As shown in FIG. 5, when the posture is changed from the posture in which the left arm is horizontally extended to the same height as the shoulder to the posture in which the arm is hung vertically, the bottom-peak interval of the fluctuation component signal Z is gradually reduced. When it falls to a certain value, it will no longer change. The peak-to-bottom interval of the fluctuation component signal Z gradually decreases after the posture is changed to the posture in which the arm is hung vertically. The absorption of light irradiated from the light emitting element 32a increases as the blood staying at the measurement site increases. This is because the amount increases and the reflected light decreases. The reason why the bottom-peak interval of the fluctuation component signal Z does not change after it has decreased to a certain value is that there is an upper limit on the amount of blood that stays at the measurement site.

  As shown in FIG. 5, the influence of gravity on blood perfusion appears as a change in the bottom-peak interval of the fluctuation component signal, but the magnitude of the influence of gravity on blood perfusion varies among individuals. Specifically, as shown in FIG. 6 (A), as shown in FIG. 6 (B), there is a person who has a large fluctuation width of the peak / bottom interval of the fluctuation component signal in a state where blood perfusion is good and bad. Some of them have a small fluctuation range. In general, blood perfusion is better and worse than healthy people in people with strong muscles such as those with cold muscles such as those with cold or those with advanced arteriosclerosis In many cases, the fluctuation range of the pulse strength is small.

  In the perfusion state determination processing according to the present embodiment, the CPU of the arithmetic processing circuit 120 first prompts the subject to move to a posture in which the blood perfusion is poor after the subject has taken a posture in which blood perfusion is good. And CPU collects the data used as the parameter | index for determining the quality of a test subject's perfusion state by analyzing the fluctuation component signal Z in both attitude | positions, and determines a test subject's perfusion state based on the said data. As an index for determining whether the perfusion state of the subject is good or bad, the ratio y / Y of the bottom-peak interval Y of the fluctuation component signal Z in the state of good blood perfusion and the bottom-peak interval y in the state of poor blood perfusion And the slope α of the envelope of the fluctuation component signal Z after the blood perfusion is changed from a good state to a bad state.

  Various aspects can be considered as to how to determine the quality of blood perfusion of a subject using the above-described indices extracted from the fluctuation component signal Z. For example, the average value of the ratio y / Y of the bottom-peak interval of the fluctuation component signal for a healthy person is obtained statistically, and the ratio y / Y of the bottom-peak interval measured for the subject is the average value. If it exceeds, it is determined that the perfusion state of the subject is bad, and if the measured value is not more than the average value, conversely, the perfusion state of the subject is good. Similarly, when determining the quality of the perfusion state based on the slope α of the envelope of the fluctuation component signal Z, an average value is obtained for a healthy person, and the average value and the measured value for the subject are determined. And the quality of the perfusion state may be determined. Also, instead of calculating the average y of the peak-to-peak interval ratio y / Y of the fluctuation component signal or the slope α of the envelope for a healthy person, the perfusion state such as a person with advanced coldness or arteriosclerosis The average value of the bottom peak-to-peak interval ratio y / Y of the fluctuation component signal or the slope of the envelope α for a poor person is obtained, and the perfusion state of the subject is determined by comparing these average values with the measured values of the subject. Of course, it may be judged.

Then, the CPU of the arithmetic processing circuit 120 uses the data indicating each index extracted from the fluctuation component signal Z and the determination result data indicating the determination result based on the index (that is, the subject is a person with good or poor perfusion state). For example) is written in the data storage circuit 130 such as an EEPROM, and the determination process is terminated. In this embodiment, 1 or 0 is used as the value of the determination result data. When the CPU determines that the person is in a good perfusion state, the determination result data with a value of 1 is used as a bad perfusion state. If it is determined that the user is a user, the determination result data having a value of 0 is written in the data storage circuit 130. In this example, whether the perfusion state is good or bad is expressed in two stages. However, by preparing a plurality of threshold values as a criterion for determination and comparing them with these, the degree of the perfusion state may be classified in three or more stages. .
The above is the content of the perfusion state determination process.

  The pulse wave measurement process is a process executed subsequent to the perfusion state determination process, and this pulse wave measurement process is executed in parallel with an optimization process and an algorithm selection process described later. This pulse wave measurement process is a process of counting the pulse interval or pulse rate per unit time while removing noise from the fluctuation component signal Z that has undergone A / D conversion. In the present embodiment, two types of algorithms, an SG filter processing algorithm and a combination of an LMS algorithm and an SG filter processing algorithm, are prepared in advance as algorithms for removing noise from the fluctuation component signal Z. Hereinafter, an outline of the SG filter processing algorithm and the LMS algorithm will be described.

The LMS algorithm is a kind of adaptive algorithm, and the adaptive filter coefficient sequence w is set to the optimum filter coefficient sequence so that the mean square error e ² is minimized based on the input signal x and the error signal e observed at each time. This is a process that approaches w ^* . Here, the error signal e is a difference signal between the desired signal d and the output signal y of the adaptive filter processing (ie, y = w ^T x, where w ^T is a transposed matrix of the adaptive filter coefficient sequence w) (ie, e = d−y). In the combination of the LMS algorithm and the SG filter processing algorithm, the filter coefficient sequence used in the SG filter processing is optimized according to the LMS algorithm. On the other hand, the SG filter is a digital smoothing polynomial filter, and the SG filter processing is to smooth the input signal waveform by approximating it with a polynomial of a predetermined order for each frame size of a predetermined number of samples. This is a process for removing noise components. In this SG filter processing, the order of the approximate polynomial and the frame size that is the approximate range are control parameters, and the degree of smoothing increases as the order of the approximate polynomial is lower or the frame size is larger. It is known that (close to a low-pass filter).

  In the present embodiment, among the above two types of control parameters for SG filter processing, the order of the approximate polynomial is fixed to a predetermined value (for example, 3), and the frame size is determined according to the perfusion state of the subject. A suitable value is set. Specifically, the CPU of the arithmetic processing circuit 120 refers to the determination result data stored in the data storage circuit 130. If the value is 1, the frame size is set to 100, and the value is 0. For example, the frame size is set to 50. This is because, in the case of a person with a poor perfusion state, the fluctuation range of the fluctuation component signal Z is smaller than that of a person with a good perfusion state, and it is not necessary to shorten the waveform section for smoothing. In the present embodiment, the frame sizes are different between those with poor perfusion and those with good perfusion. However, the frame size may be fixed regardless of whether the perfusion state of the subject is good or bad.

  The reason why the SG filter processing algorithm is used as the algorithm for smoothing the fluctuation component signal Z in the present embodiment is as follows. In order to obtain the pulse wave RR interval from the fluctuation component signal Z, the pulse wave bottom peak corresponding to R of the electrocardiogram (that is, the position where the inclination of the fluctuation component signal for each beat appearing in a convex shape becomes maximum) is detected. For this purpose, it is necessary to calculate the differential signal of the fluctuation component signal Z and detect the amplitude peak of the differential signal. As described above, in the SG filter process, since the waveform of the input signal is approximated by a polynomial of a predetermined order for each frame size, the differential signal can be easily obtained by differentiation of this approximate polynomial. This is the reason why the SG filter processing algorithm is used as an algorithm for smoothing the fluctuation component signal Z.

  The optimization process is a process of adjusting the light emission intensity of the light emitting element 32a and the gain of the amplification circuit 1116a so that the pulse wave is measured in a manner that takes into account the influence of gravity on blood perfusion. More specifically, in this optimization process, the CPU of the arithmetic processing circuit 120 analyzes the envelope of the fluctuation component signal Z from which noise has been removed to determine whether or not a change has occurred in blood perfusion at the measurement site. And a control signal instructing adjustment of the light emission intensity of the light emitting element 32a and the gain of the amplifier circuit 1116a is given to the analog control circuit 140 according to the determination result. In response to this control signal, the analog control circuit 140 outputs the light emission intensity control signal LC and the gain control signal GC, whereby the light emission intensity of the light emitting element 32a and the gain of the amplifier circuit 1116a are adjusted. In the present embodiment, a mode in which both the light emission intensity of the light emitting element 32a and the gain of the amplifier circuit 1116a are adjusted will be described, but it is of course possible to adjust only one of them.

In the optimization processing of this embodiment, the CPU of the arithmetic processing circuit 120 calculates the difference between the same bottom-peak interval Y in a state where blood perfusion is good and the bottom-peak interval of the fluctuation component signal from which noise has been removed. The analog control circuit 140 is supplied with a control signal indicating the difference value. The analog control circuit 140 outputs a light emission intensity control signal LC indicating a strong light emission intensity and a gain control signal GC indicating a high amplification factor as the value indicated by the control signal increases. If adjustment of the light emission intensity of the light emitting element 32a and the gain of the amplifier circuit 1116a is not performed, the bottom of the fluctuation component signal as shown by the solid line in FIG.・ Peak interval is narrowed. However, in the present embodiment, the light emission intensity of the light emitting element 32a and the gain of the amplifier circuit 1116a are increased according to the decrease width of the bottom-peak interval of the fluctuation component signal, as shown by the broken line in FIG. In addition, the bottom-peak interval of the fluctuation component signal can be kept substantially constant. In the present embodiment, the case has been described where auto gain controller-like control is performed in which the bottom-peak interval of the fluctuation component signal is kept substantially constant in the optimization process, but the bottom-peak interval of the fluctuation component signal is a predetermined value (for example, Then, triggered by the decrease in the bottom-peak interval y) in a state where blood perfusion is poor, thereafter, the emission intensity of the light-emitting element 32a and the gain of the amplification circuit 1116a are gradually increased, and the bottom-peak interval of the fluctuation component signal gradually increases. Of course, it may be an aspect (see FIG. 7B) that is made larger.
The above is the content of the optimization process.

  The algorithm selection process is a process of selecting an algorithm used when removing noise from the fluctuation component signal Z according to the presence or absence of the subject's body movement and the quality of blood perfusion at the measurement site. As described above, in the present embodiment, as an algorithm for removing noise from the fluctuation component signal Z, a combination of an SG filter processing algorithm and an LMS algorithm and an SG filter processing algorithm are prepared. In this algorithm selection process, the CPU of the arithmetic processing circuit 120 selects an algorithm for removing noise from the fluctuation component signal Z in the following manner.

  The CPU of the arithmetic processing circuit 120 selects the SG filter processing algorithm at the start of execution of the measurement process. Thereafter, the CPU determines the presence / absence of the subject's body movement based on the acceleration signal AS, and further determines the quality of blood perfusion at the measurement site as in the optimization process. There are various modes for determining whether or not a subject is moving based on the acceleration signal AS. In the present embodiment, when the magnitude of acceleration indicated by the acceleration signal AS exceeds the magnitude of gravitational acceleration. It is determined that there is body movement. This is because when the subject is performing body movement such as walking, the magnitude of the combined acceleration of the acceleration caused by the body movement and the gravitational acceleration is generally larger than the magnitude of the gravitational acceleration. Thus, the reason for selecting an algorithm for removing noise from the fluctuation component signal Z in accordance with the presence or absence of body movement of the subject and the quality of blood perfusion at the measurement site is as follows.

  FIG. 8 shows the variation component signal Z of the subject walking with a posture in which the arm to which the pulse wave signal measuring device 1 is attached (in this embodiment, the left arm) is horizontally extended to the same height as the shoulder. It is the figure which superimposed the graph pls1 which plotted the pulse interval calculated by giving the noise removal by a filter processing algorithm, and the graph egc1 which plotted the RR interval measured from the subject's electrocardiogram. As is apparent from FIG. 8, both graphs are substantially coincident, and it can be seen that body motion noise associated with walking is removed by the SG filter processing, and the RR interval is accurately measured. On the other hand, FIG. 9 is a graph pls2 plotting pulse wave intervals calculated by removing noise by the SG filter processing algorithm on the fluctuation component signal Z of the subject walking in a posture in which the left arm is vertically suspended. And a graph egc2 in which the RR interval measured from the electrocardiogram of the subject is plotted. As can be seen from FIG. 9, both graphs do not match, and it can be seen that body movement noise accompanying walking is not successfully removed.

  As described above, in the posture in which the arm on which the pulse wave signal measuring device 1 is mounted is horizontally extended to the same height as the shoulder, blood perfusion at the pulse wave measurement site is in a good state, and the arm is hung vertically. In the posture, blood perfusion at the same measurement site is in a bad state. In a state where blood perfusion at the measurement site is poor, the received light intensity of the light receiving element 32b is reduced due to the influence of blood accumulated in the measurement site and the S / N ratio is lowered. The fluctuation component signal is superimposed on the fluctuation component signal. That is, when the blood perfusion at the measurement site is poor, the body motion noise included in the fluctuation component signal Z is larger than that when the blood perfusion is good, and the body motion noise when the blood perfusion is poor only by the SG filter processing. It is thought that it cannot be removed.

FIG. 10 shows the noise removal by the combination of the SG filter processing algorithm and the LMS algorithm on the fluctuation component signal Z of the subject who is walking in a posture in which the arm on which the pulse wave signal measuring device 1 is mounted is vertically hung. 6 is a diagram in which a graph pls3 plotting the RR interval calculated in the above and a graph egc3 plotting the RR interval measured from the electrocardiogram of the subject are superimposed. As can be seen from FIG. 10, both graphs are substantially the same, and it can be seen that body motion noise accompanying walking is appropriately removed and the RR interval is accurately calculated. Thus, when blood perfusion at the measurement site is poor and the subject is performing body movements such as walking, the SG filter processing algorithm cannot remove the body movement noise, but the SG filter processing algorithm and the LMS By using a combination of algorithms, body motion noise can be appropriately removed and the RR interval and the like can be accurately measured. This is the reason why the algorithm for removing noise from the fluctuation component signal is selected according to the presence or absence of the subject's body movement and the quality of blood perfusion.
The above is the processing content of the algorithm selection processing.

Next, the operation of the pulse wave signal measuring apparatus 1 of the present embodiment will be described.
When the subject wears the pulse wave signal measuring device 1 on his / her left arm and turns on the power of the pulse wave signal measuring device 1, the arithmetic processing circuit 120 first executes a perfusion state determination process. In this perfusion state determination process, the arithmetic processing circuit 120 sends a voice message instructing to take a posture in which the arm on which the pulse wave signal measuring device 1 is worn is horizontally extended as much as the height of the shoulder. And the data indicating the peak / bottom interval Y of the fluctuation signal component Z in that state is stored in the data storage circuit 130 as data indicating a good state of blood perfusion.

  Next, the arithmetic processing circuit 120 causes the voice output unit to emit a voice message instructing to take a posture in which the arm on which the pulse wave signal measuring device 1 is mounted is vertically hung. When the subject changes posture from the posture in which the arm wearing the pulse wave signal measuring device 1 is horizontally extended to the same height as the shoulder to the posture in which the arm is vertically hung, the subject gradually moves toward the left fingertip of the subject. Blood accumulates. Along with this, the bottom-peak interval of the fluctuation component signal Z is gradually narrowed, and when the blood accumulated in the left fingertip of the subject reaches a certain amount, the bottom-peak interval of the fluctuation component signal Z becomes substantially constant. The arithmetic processing circuit 120 receives the data indicating the bottom-peak interval y as data indicating a poor blood perfusion in the data storage circuit 130 when the bottom-peak interval of the fluctuation component signal becomes substantially constant. Remember me.

  Next, the arithmetic processing circuit 120 calculates a value y / Y indicating the perfusion state of the subject, compares the calculated result with a predetermined threshold value, determines whether the perfusion state of the subject is good, and indicates the determination result. The determination result data is written in the data storage circuit 130, and the determination result is notified to the subject by voice, and the perfusion state determination process is terminated.

  When the perfusion state determination process is completed, the CPU of the arithmetic processing circuit 120 selects the SG filter processing algorithm by the algorithm selection process and starts executing the pulse wave measurement process, and further performs the optimization process and the algorithm selection process. Run in parallel with the wave measurement process. In the pulse wave measurement process, the CPU measures the pulse wave while removing noise from the fluctuation component signal Z according to the algorithm selected in the algorithm selection process, and displays the measurement result on the display unit 14. As described above, in the optimization process of the present embodiment, control like an auto gain controller is performed so that the bottom-peak interval of the fluctuation component signal Z is constant. As a result, even when blood perfusion at the measurement site deteriorates with a change in the posture of the subject, the signal intensity of the fluctuation component signal Z does not decrease and the S / N ratio does not decrease.

  On the other hand, in the algorithm selection processing, the CPU of the arithmetic processing circuit 120 first analyzes the acceleration signal AS to determine the presence or absence of the subject's body movement, and when the determination result that there is body movement is obtained, The fluctuation component signal is analyzed to determine the quality of blood perfusion at the measurement site. If the CPU determines that there is body movement and determines that blood perfusion at the measurement site is poor, the CPU selects a combination of the SG filter processing algorithm and the LMS algorithm as an algorithm for noise removal. It is.

  For example, when a subject wearing the pulse wave signal measuring device 1 on the left arm starts walking in a posture in which the left arm is horizontally extended to the same height as the shoulder, the fluctuation component signal is generated by the SG filter processing algorithm at the start time. Noise removal is performed. This is because blood perfusion at the measurement site is in a good state when the left arm is horizontally extended to the same height as the shoulder. When the subject changes his / her posture to the posture in which his / her left arm is hung vertically while walking, blood perfusion deteriorates in the measurement site with the change in the posture, and the CPU of the arithmetic processing circuit 120 removes noise. A combination of the SG filter processing algorithm and the LMS algorithm is selected as the algorithm. Thereafter, in the measurement processing, noise is removed from the fluctuation component signal and pulse wave measurement is performed by a combination of the SG filter processing algorithm and the LMS algorithm. According to the combination of the SG filter processing algorithm and the LMS algorithm, the body motion noise can be appropriately removed from the fluctuation component signal even when the subject is performing some body motion and the blood perfusion at the measurement site is poor. Is as described above.

  As described above, according to the pulse wave signal measuring apparatus 1 of the present embodiment, it is possible to measure a pulse wave while appropriately removing noise according to the presence / absence of body movement of the subject and the quality of blood perfusion at the measurement site. It becomes possible.

(B: Deformation)
Although the embodiment of the present invention has been described above, it is needless to say that the following modifications may be added to this embodiment.
(1) In the above-described embodiment, the fluctuation component signal is analyzed to determine whether the perfusion state of the subject is good or not and whether the blood perfusion is good at the measurement site. However, it is also possible to determine these based on the non-variation component signal. It is. The non-variable component signal is a DC component signal that is unrelated to the pulsation of the arterial blood vessel, but the signal strength changes between a good state and a bad state of blood perfusion (specifically, This is because the signal intensity decreases when blood perfusion deteriorates. Also, instead of determining the quality of the perfusion state of the subject and the quality of blood perfusion at the measurement site based on either the fluctuation component signal or the non-variation component signal, both the fluctuation component signal and the non-variation component signal are used. Of course, the determination may be made based on this. In the above-described embodiment, the reflected light of the light emitted from the light emitting element 32a toward the measurement site is analyzed to determine pulse wave measurement and blood perfusion at the measurement site. Of course, the light may be analyzed to measure the pulse wave and determine the quality of blood perfusion at the measurement site.

(2) In the embodiment described above, the acceleration measurement circuit 150 is built in the apparatus main body 10, but the acceleration measurement circuit 150 may be built in the sensor unit 30. Further, the output signal of the acceleration measurement circuit 150 may be analyzed to detect the relative direction of gravity with respect to the measurement site, and the quality of blood perfusion at the measurement site may be determined by taking into account the direction of gravity acting on the measurement site. . This is because blood perfusion is considered to deteriorate when the direction of gravity acting on the measurement site matches the direction in which arterial blood flows in the measurement site.

(3) In the above-described embodiment, as the algorithm for removing noise from the fluctuation component signal, a filter processing algorithm that smoothes the waveform of the fluctuation component signal, or a combination of the filter processing algorithm and the adaptive algorithm is used. SG filtering algorithm was used as the algorithm for smoothing and LMS was used as the adaptive algorithm. However, the algorithm for smoothing the waveform of the fluctuation component signal is not limited to the SG filter processing algorithm. For example, low-pass filter processing for removing frequency components of a predetermined cutoff frequency or higher may be used. Processing to calculate may be used. The adaptive algorithm is not limited to the LMS, and a Normalized LMS algorithm, a projection algorithm, an RLS (Recursive Least Square) algorithm, or the like may be used.

(4) In the above-described embodiment, the perfusion state determination process, the pulse wave measurement process, the optimization process, and the algorithm selection process are realized by software. However, each means of the perfusion state determination means for executing the perfusion state determination processing, the pulse wave measurement means for executing the pulse wave measurement processing, the optimization means for executing the optimization processing, and the algorithm selection means for executing the algorithm selection processing are hardened. Of course, the arithmetic processing circuit 120 may be configured by combining these circuits.

(5) In the above-described embodiment, the control parameter of the SG filter process is adjusted according to the quality of the perfusion state of the subject (in other words, the magnitude of the influence of gravity on the blood perfusion of the subject). However, it is not essential to adjust the parameters according to the quality of the perfusion state of the subject. For example, when the body motion noise can be sufficiently removed only by switching the algorithm for removing noise from the fluctuation component signal, the control parameter need not be adjusted. And in the aspect which does not adjust the control parameter according to the quality of a test subject's perfusion state, you may abbreviate | omit the said perfusion state determination process. In addition, if the body motion noise can be sufficiently removed by switching the algorithm for removing noise from the fluctuation component signal according to the presence or absence of the subject's body motion and the quality of blood perfusion at the measurement site, the optimization process May be omitted.

  DESCRIPTION OF SYMBOLS 1 ... Pulse wave signal measuring device, 10 ... Apparatus main body, 12 ... Wristband, 14 ... Display part, 16 ... Button switch, 20 ... Cable, 30 ... Sensor part, 32 ... Pulse wave sensor, 32a ... Light emitting element, 32b ... Light receiving element, 34 ... Sensor fixing band, 110 ... Analog circuit unit, 1110 ... Signal generating means, 1112 ... IV conversion circuit, 1114a, 1114b ... Filter, 1116a, 1116b ... Amplifying circuit, 120 ... Arithmetic processing circuit, 130 ... Data Memory circuit, 140 ... analog control circuit, 150 ... acceleration measurement circuit.

Claims (7)

An acceleration measuring means for detecting an acceleration generated according to an external force and outputting an acceleration signal representing the acceleration;
A light emitting means for irradiating light of a predetermined wavelength toward a portion determined as a measurement portion of the body of the subject;
Of the light emitted from the light emitting means toward the measurement site, the light reflected by the measurement site or the light transmitted through the measurement site is received and a signal having an amplitude corresponding to the received light intensity is output. A light receiving means;
A signal generation means for extracting a fluctuation component signal reflecting the pulsation of the arterial blood vessel of the subject and a non-fluctuation component signal reflecting the activity of biological tissue other than the arterial blood vessel from the output signal of the light receiving means;
One of a plurality of algorithms prepared in advance for removing noise included in the fluctuation component signal is determined by analyzing at least one of the fluctuation component signal and the non-fluctuation component signal. Algorithm selection means for selecting according to the quality of blood perfusion at the measurement site and the presence or absence of body movement of the subject determined from the acceleration signal;
Pulse wave measuring means for measuring pulse waves while removing noise included in the fluctuation component signal according to the algorithm selected by the algorithm selecting means;
A pulse wave signal measuring device characterized by comprising: As the plurality of types of algorithms, at least two types of a filter processing algorithm for smoothing the waveform of the fluctuation component signal and a combination of the filter processing algorithm and an adaptive algorithm are prepared,
The algorithm selection means includes
2. The pulse wave according to claim 1, wherein when it is determined that there is body movement and blood perfusion at the measurement site is determined to be poor, a combination of the filtering algorithm and the adaptive algorithm is selected. Signal measuring device. A data storage circuit;
When blood perfusion at the measurement site changes from a good state to a bad state, at least one of the fluctuation width of the bottom-peak interval of the fluctuation component signal and the fluctuation width of the signal intensity of the non-variation component signal Perfusion state determination means for preliminarily determining the quality of the perfusion state indicating the ease of change of blood perfusion, and writing determination result data indicating the determination result in the data storage circuit,
3. The pulse wave according to claim 1, wherein the pulse wave measurement unit switches a control parameter used for smoothing a waveform in the filter processing algorithm according to a determination result indicated by the determination result data. Signal measuring device. The acceleration measuring means is a three-axis acceleration sensor that detects acceleration generated according to an external force by decomposing the acceleration into three axial components orthogonal to each other,
The algorithm selection means determines that the subject has body movement when receiving an acceleration signal indicating that acceleration has occurred in addition to gravitational acceleration from the acceleration measurement means. The pulse wave signal measuring apparatus according to any one of the above. The acceleration measuring means is attached to the measurement site or the vicinity of the measurement site,
The algorithm selection means identifies the direction of gravity acting on the measurement site by analyzing the acceleration signal, and determines the quality of blood perfusion at the measurement site by taking the identification result into consideration. The pulse wave signal measuring device according to claim 4. The signal generating means, when extracting the fluctuation component signal from the output signal of the receiving means, amplifies and extracts the signal intensity of the fluctuation component signal at a predetermined amplification rate,
The apparatus further comprises an optimization unit that adjusts the intensity of light irradiated from the light emitting unit toward the measurement site or the amplification factor according to the quality of blood perfusion at the measurement site. The pulse wave signal measuring device according to any one of 5. On the computer,
The light reception intensity of reflected light from the measurement site or the light reception intensity of transmitted light that has passed through the measurement site out of light irradiated toward a predetermined site as a pulse wave measurement site in the subject's body, The subject included in the representing pulse wave signal, wherein the subject is determined from at least one of a fluctuation component signal reflecting pulsation of an arterial blood vessel of the subject and a non-fluctuation component signal reflecting the activity of biological tissue other than the artery A plurality of types prepared in advance for removing noise from the fluctuation component signal according to the presence or absence of body movement of the subject determined from the quality of blood perfusion of the subject and the magnitude of acceleration acting on the measurement site An algorithm selection process for selecting one of the algorithms of:
A pulse wave measurement process for measuring a pulse wave while removing noise included in the fluctuation component signal according to the algorithm selected in the algorithm selection process;
A program characterized by having executed.

Images