KR20100065084A - Apparatus for measuring motion noise robust pulse wave and method thereof - Google Patents

Abstract

PURPOSE: A pulse wave measuring device robust to movement noises and a measuring method thereof are provided to eliminate movement noises by using a statistical model, an acceleration sensor, and adaptive signal processing technology in a distorted pulse wave signal. CONSTITUTION: A PPG(photoplethysmogram) sensor(130) is arranged on the back of a wrist of an examinee. The PPG sensor detects a pulse wave signal from the signal emitted to the back of the wrist. An acceleration sensor(150) detects the movement of the examinee in the mounted area of the PPG sensor. A signal preprocessor(111) predicts the movement of the examinee based on an acceleration signal detected through the acceleration sensor. The signal preprocessor estimates the movement noise model which corresponds to the detected movement of the examinee. The signal preprocessor eliminates the movement noises in response to the pulse wave signal from the movement noise model.

Description

Apparatus for measuring motion noise robust pulse wave and method

The present invention relates to a pulse wave measuring apparatus that is robust to motion noise, and a method thereof, comprising: an apparatus and method for estimating a motion component using an acceleration sensor and measuring the pulse wave robust to motion noise in a region such as the upper part of a wrist through an adaptive signal processing technique Is about.

The present invention is derived from a study conducted as part of the IT source technology development project of the Ministry of Knowledge Economy [Task Management Number: 2008-F-039-01, Task name: Development of human-robot interaction mediated technology].

Pulse wave measurement technology is an important field that has been studied in the field of bio signal processing technology and is a technique used to extract weak pulse wave bio signals buried in noise.

Pulse wave measurement technology is largely invasive type and non-invasive type. Invasive type is a method of directly measuring inside the human body. Non-invasive type is a method of measuring biological signals inside the human body indirectly from outside the human body. This is mostly.

The non-invasive type is largely measured using a piezoelectric element and an optical sensor. In the case of the method using the optical sensor, pulse wave is measured by the change of absorption rate using the near infrared emitted by the optical sensor, Infra-red LED. do.

The principle of measuring by using optical sensor is the contraction / relaxation of the heart rhythm by applying the modified principle when Beer-Lambert's law, which calculates the absorption rate of light to a medium, is applied to human tissue. The volume change rate of the blood flowing inside the blood vessel is represented as the change in the absorption rate of the near infrared rays, and this is the principle that the near-infrared photo detector as a receiver acquires the near infrared rays that have penetrated or reflected the blood vessels.

In measuring the pulse wave in the u-healthcare field based on the ubiquitous environment, it is most important to obtain accurate biosignal information of the user. However, to measure the pulse wave in a portable environment, there are various problems depending on the area and the characteristic to be measured.

For the portable environment, it should be worn on the body measuring part. In the portable environment, such as a patient in a hospital, the pulse wave signal is distorted due to the movement of the measuring part, and thus it cannot be used as a health indicator data.

Recently, in order to solve such motion noise, researches for minimizing the motion noise by adopting techniques such as adaptive signal processing, wavelet signal processing, and morphological signal processing have also been conducted in existing medical equipment.

An object of the present invention for solving the above problems is to provide a pulse wave measuring apparatus and method that is robust to the movement noise to enable the pulse wave measurement in the upper part of the wrist, such as difficult to measure the pulse wave so as not to inconvenience the user in a mobile environment. In providing.

In addition, another object of the present invention is to provide a pulse wave measuring apparatus and a method which is robust to the motion noise to be robust to the motion noise due to wrist movement by restoring the original signal with a relatively minimal error.

The pulse wave measuring apparatus robust to the movement noise according to the present invention for achieving the above object is disposed in the upper part of the wrist of the subject, the PPG sensor unit for detecting the pulse wave signal from the signal emitted to the upper lamp, the PPG sensor unit An acceleration sensor unit that detects the movement of the subject at the mounted site, and the movement of the subject is predicted from the acceleration signal detected by the acceleration sensor unit to estimate a motion noise model corresponding to the detected movement of the subject; And a signal preprocessor which removes the motion noise of the pulse wave signal from the motion noise model.

The signal preprocessor may include a signal acquirer configured to sample the pulse wave signal and the acceleration signal by a predetermined unit, and a noise remover to remove noise of the pulse wave signal sampled by the signal acquirer using a high pass filter. do.

The motion noise model is a unique motion noise model, and the signal preprocessor obtains a highly correlated motion noise based on the unique motion noise model.

The signal preprocessor estimates the inherent motion noise model using an autoregressive (AR) model estimator.

The signal preprocessor may filter the acceleration signal detected by the inherent motion noise model and the acceleration sensor unit and output a motion noise signal having a high correlation with the inherent motion noise of the examinee.

The signal preprocessor is characterized by using a finite impulse response (FIR) filter.

The signal preprocessor may remove the motion noise of the pulse wave signal using a motion noise signal having a high correlation with the motion noise of the examinee.

The apparatus may further include a communication unit configured to transmit the pulse wave signal from which the motion noise is removed to an external host server.

The PPG sensor is characterized in that it comprises a light emitting unit for outputting an optical signal to the upper part of the wrist of the subject, and a light receiving unit for receiving the optical signal emitted by the light emitting unit.

The acceleration sensor unit includes a three-axis acceleration sensor, characterized in that for detecting the acceleration component of each axis according to the movement of the subject using the three-axis acceleration sensor.

On the other hand, the pulse wave measurement method robust to the motion noise according to the present invention for achieving the above object, detecting the pulse wave signal from the PPG sensor disposed on the upper back portion of the subject, the PPG sensor is mounted using an acceleration sensor Estimating a motion noise model corresponding to the movement of the examinee based on the movement information of the examinee and the acceleration signal detected from the acceleration sensor, and the movement noise model And removing the motion noise for the pulse wave signal.

And based on the estimated motion noise model and the acceleration signal detected by the acceleration sensor, acquiring a motion noise having a high correlation with the inherent motion noise of the examinee.

The removing of the motion noise may include removing the motion noise of the pulse wave signal using a motion noise signal having a high correlation with the inherent motion noise of the examinee.

The removing of the motion noise may include repeatedly removing the motion noise of the pulse wave signal using an adaptive filter.

The method may further include transmitting the pulse wave signal from which the motion noise is removed to an external host server.

According to the present invention, the above two problems, that is, to avoid the inconvenience to the user in the mobile environment, to be able to measure the pulse wave in the upper part of the wrist difficult to measure pulse wave and to be robust to the movement noise due to wrist movement By restoring the original signal with a relatively minimal error, there is an advantage that the pulse wave can be detected in a wrist region or a wrist region where the body restraint is relatively weak.

In addition, there is an advantage in that the motion noise can be removed using a statistical model, an acceleration sensor, and an adaptive signal processing technique in the distorted pulse wave signal measured by adding the motion noise.

On the other hand, the present invention has the advantage that can be used in a portable environment because it provides a wrist-wearing device in the form of a watch does not inconvenience the user.

In addition, the present invention has the advantage of identifying and managing personal health indicators using pulse waves in PCs, laptops, mobile phones, PDAs, and other mobile devices that support it with the support of Bluetooth / Zigbee.

In addition, recently, as interest in health has increased, it may be used as a health measurement tool and emergency prognostic situation detection function of an adult or the elderly, and if contents for stress, concentration, sleep quality, etc. through HRV analysis are provided to the host device, It can be used for more diverse purposes, such as games, education, and well-being contents. In addition, wireless protocols such as Zigbee can be used to collect and process data from multiple wearable devices in groups such as hospitals, nursing homes and silver towns. It can be used as a basic device in the field of ubiquitous healthcare.

Hereinafter, with reference to the accompanying drawings will be described an embodiment of the present invention.

1 illustrates a configuration to which a pulse wave measuring apparatus robust to motion noise according to the present invention is applied.

The pulse wave measuring apparatus 100 according to the present invention includes a communication protocol for using Bluetooth or Zigbee. When the pulse wave measuring apparatus 100 measures pulse wave from a subject, the pulse wave measuring apparatus 100 may transmit the measurement data to a PC, a robot, or other medical equipment, which is the host apparatus 200, using a wireless communication technology.

At this time, the pulse wave measuring apparatus 100 according to the present invention is composed of a main body 100a for measuring the pulse wave of the subject, and wearing means 100b for wearing the main body 100a on the wrist. The wearing means 100b may be implemented in the form of a band, or may be implemented in the form of a wrist watch or bracelet.

Thus, with reference to Figure 2 will be described in more detail the configuration of the pulse wave measuring apparatus according to the present invention.

2 is a block diagram showing a detailed configuration of a pulse wave measuring apparatus robust to motion noise according to the present invention. 2, the pulse wave measuring apparatus 100 according to the present invention includes the MCU platform 110, the sensor low power driver 120, the PPG sensor 130, the signal amplifier 140, and the acceleration sensor 150. , A communication unit 160, and a power supply unit 170.

The MCU platform 110 includes a signal preprocessor 111, an A / D converter 114, a PWM 115, and a power manager 116.

The sensor low power driver 120 outputs a driving signal for driving the PPG sensor to the PPG sensor unit 130.

The PPG sensor unit 130 is disposed on the upper back of the subject's wrist. At this time, the PPG sensor unit 130 is a light emitting unit (not shown) for outputting an optical signal to the upper part of the wrist for pulse wave measurement, and a receiving unit (not shown) for receiving the optical signal reflected to the wrist, blood vessels, etc. of the subject Include.

In this case, the PPG sensor unit 130 performs the pulse wave measuring operation according to the driving signal from the sensor low power driver 120. The PPG sensor unit 130 measures the pulse wave signal from the subject and transmits the measured pulse wave signal to the signal amplifier 140.

In addition, the signal amplifier 140 amplifies the pulse wave signal measured by the PPG sensor into a signal having a predetermined level, and then transfers it to the A / D converter 114 of the MCU platform 110.

On the other hand, the acceleration sensor unit 150 is provided with a sensor for measuring the movement of the subject's wrist or hand. In this case, the sensor provided in the acceleration sensor unit 150 may correspond to a gravity acceleration sensor, an angular velocity sensor, and the like. The acceleration sensor unit 150 transmits the measured motion data to the A / D converter 114 of the MCU platform 110.

The motion data measured by the acceleration sensor unit 150 is later used to correct motion noise in the pulse wave signal.

The A / D converter 114 converts the analog pulse wave signal input from the signal amplifier 140 into a digital signal and outputs the digital signal to the signal preprocessor 111. Similarly, the A / D converter 114 converts the analog signal transmitted from the acceleration sensor unit 150 into a digital signal and outputs the digital signal to the signal preprocessor 111.

The signal preprocessor 111 performs a function of removing noise by preprocessing signals measured from the sensors of the PPG sensor 130 and the acceleration sensor 150. In this case, the signal preprocessor 111 includes a signal acquirer 112 and a noise remover 113.

The signal acquisition unit 112 quantizes in a range of about 10 bits to about 12 bits, and the sampling rate samples pulse waves at a speed in a range of 100 to 1000 pulses per second, and an acceleration signal at 100 speeds per second.

The noise removing unit 113 removes optical noise, electrical noise, and the like of the pulse wave signal sampled by the signal obtaining unit 112 using a low pass filter. On the other hand, the noise removing unit 113 removes the respiratory noise or DC component of the pulse wave signal sampled by the signal acquisition unit 112 using a high pass filter.

Here, the order of each filter is 4th order, the cutoff frequency is 1.5Hz and 0.5Hz, and a Butterworth type Infinite Impulse Response (IIR) filter is used.

Meanwhile, the noise removing unit 113 removes the offset voltage of the acceleration signal and removes high frequency noise of the acceleration signal by using the smoothing filter.

Finally, the signal preprocessor 111 serializes the pulse wave signal from which the noise is removed and the acceleration signal, and outputs the serialized pulse wave signal to the communication unit 160.

For a detailed operation description of the signal preprocessor 111, refer to the description of FIG. 6.

3 is a block diagram showing the configuration of a host device according to the present invention.

Referring to FIG. 3, the host device 200 includes a communication unit 210, an active noise canceller 220, and a signal generator 230.

The communication unit 210 receives a signal transmitted from the pulse wave measuring apparatus 100.

The active noise canceller 220 removes the motion noise by the active noise canceling method using the pulse wave signal and the acceleration signal received through the communication unit 210.

The signal generator 230 stores the final reconstructed pulse wave signal by removing the motion noise by the active noise canceller 220, and generates a PP signal, which is an interval between peak points of the pulses, from the reconstructed pulse wave signal. Provide basic data.

4 is a view showing the structure of the pulse wave measuring apparatus according to the present invention.

Referring to FIG. 4, in the pulse wave measuring apparatus 100 according to the present invention, the PPG sensor unit 130 includes two light emitting devices for emitting near infrared rays to the upper part of the wrist of a subject. In the embodiment of FIG. 4, two light emitting devices are infrared ray (IR) LEDs 131 and 132.

The two IR LEDs 131 and 132 are driven using a modulated pulse that is not driven between time intervals that are not sampled, and thus do not drive continuously.

In addition, the PPG sensor unit 130 further includes a light detecting element for detecting near infrared rays emitted by the two IR LEDs 131 and 132. In the embodiment of FIG. 4, the photodetecting device is an IR detector 133.

The IR detector 133 adjusts the duty cycle appropriately in conjunction with the period for sampling the input signal to sufficiently respond to the sampling rate.

Here, two IR LEDs 131 and 132 are disposed at both sides of the IR detector 133. This is because the wrist has a complex structure compared to various body tissues, especially the wrist bone, and the arteries are located deep inside the wrist, so that the wrist supports a wider range of wrist structures than the fingers and finds deep artery flow.

In this case, when the two IR LEDs 131 and 132 and the IR detector 133 are close to each other, the IR detector 133 absorbs the light emitted from the IR LEDs 131 and 132 into the wrist, as well as the reflected light. The light emitted from the IR LEDs 131 and 132 may be directly absorbed. Therefore, the two IR LEDs 131 and 132 and the IR detector 133 are arranged to maintain a constant interval. Preferably it is arranged so as to maintain the interval of about 7 to 10mm.

Also, when the two IR LEDs 131 and 132 and the IR detector 133 are mounted inside the main body 100a of the pulse wave measuring apparatus 100, the two IR LEDs 131 and 132 are mounted to be inserted inward from the surface of the main body 100a. Preferably it is mounted to be inserted in about 1.5 to 2mm inward.

In this case, it is possible to reduce the direct absorption of the optical signal emitted from the two IR LEDs (131, 132) to the IR detector 133, the body contact surface, the two IR LEDs (131, 132) and the IR detector (133) The distance between them can be reduced to reduce the noise caused by physical contact.

In addition, in the pulse wave measuring apparatus 100 according to the present invention, the lower structure of the PPG sensor unit 130, the sensor low power driver 120 for applying a drive signal to the PPG sensor unit 130, and to measure the movement of the subject Acceleration sensor is disposed, MCU platform 110 is disposed.

In addition, a power supply unit 170 including a battery is disposed in the lower structure of the sensor low power driver 120, the acceleration sensor, and the MCU platform 110 to supply power to the pulse wave measuring apparatus 100. At this time, the power supply unit 170 is provided with a charging circuit, can supply power from a lithium ion rechargeable battery (3.3v), etc., it is possible to drive with low power is supported by the standby mode.

FIG. 5 is a diagram illustrating characteristic curves of the IR LED and the IR detector shown in FIG. 4.

First, Figure 5 (a) is a graph showing the output characteristics of the IR LED.

As shown in Fig. 5A, the IR LEDs 131 and 132 emit near infrared light, which is light of a wavelength of 940 nm. In this case, the IR LEDs 131 and 132 should have an output intensity such as 'P' in FIG.

In particular, in the pulse wave measuring apparatus 100 according to the present invention, the IR LEDs 131 and 132 are limited within the range of 900 nm to 1000 nm to emit near infrared rays so that pulse wave measurement can be performed on the upper back of the subject's wrist.

5B is a graph showing the response characteristics of the IR detector.

When the IR detector 133 has a wide spread of the response curve, light of different bands is absorbed by the IR detector 133 to cause noise, and as shown in (b) of FIG. To a sensor that responds well in the 1000 nm wavelength range.

Equation 1 below shows that the Beer-Lambert law is applied to body tissues.

Here, I _i (t) is the intensity of the light input, I _o (t) is the intensity of the light transmitted through the body tissue. Ε _{λ, k} is the absorption coefficient for each medium, C _k (t) is the concentration, and d _k (t) represents the distance of each medium.

As shown in [Equation 1], the intensity of light transmitted from the body tissue I _o (t) changes the intensity of the input light according to three change factors: absorption coefficient, concentration, and distance between the media. do.

In other words, the light intensity of the wavelength varies depending on the distance, concentration, and absorption coefficient between the media, so that the intensity of the light may be changed in a region having a complex body tissue such as a wrist region.

On the other hand, in the case of using sensors with spread curves of output characteristics and response characteristics, the Beer-Lambert law of [Equation 1] should consider the light of various wavelengths, and should be modified as shown in [Equation 2] below. .

In other words, since the absorption of light is different for each wavelength band, light incident to different wavelength bands is mixed to affect the output. Therefore, it is important to have a relatively narrow wavelength band in a wrist tissue having a complicated medium.

FIG. 6 is an exemplary view referred to for explaining an operation of removing motion noise in a signal preprocessor according to the present invention. FIG. 6 is an adaptive filter structure for an active noise canceller for removing motion noise from a distorted pulse wave signal measured by adding motion noise. It is shown.

Referring to FIG. 6, if there is no motion noise, a pure pulse wave signal P is generated. If a motion is generated by a subject, a signal in which n, which is a unique component of motion noise, is added to the pure pulse wave signal P is generated.

Therefore, as a signal measured by the PPG sensor unit 130, a signal d = p + n, in which n, which is a motion noise intrinsic component, is added to the pure pulse wave signal P is measured.

At this time, the noise-native component n cannot know exactly its value. If a signal having a high correlation with n can be obtained, it is possible to attenuate the natural noise n included in the original signal P with the minimum mean square error by using an active noise canceller (ANC) structure in the adaptive filter.

를 구하여 상관도가 높은 잡음을 얻는다. In the pulse wave measuring apparatus according to the present invention, a model of inherent motion noise is estimated using an AR model estimator, and a transfer function of the estimated model is estimated.

To obtain a highly correlated noise.

On the other hand, the movement of the wrist portion obtains the acceleration component of each axis, not the pulse wave component, from the three-axis acceleration sensor of the acceleration sensor unit 150. At this time, the acceleration sensor a transmits to the finite impulse response (FIR) filter so that the acceleration signal a correlates with the signal distorting the pulse wave signal.

As another example, when the cuff of the blood pressure monitor is used, the AR model estimator estimates the value of the pulse wave component due to the approximate intrinsic motion noise generated through the movement of the wrist after the pulse wave component is temporarily removed from the wrist.

와 가속도 신호 a를 필터링하여 고유 움직임 잡음 n과 상관이 있는 신호

를 출력한다. 이때, 신호

는 능동잡음 제거기의 입력신호로 작용한다. Thus, the FIR filter has a transfer function

And the acceleration signal a to be filtered to correlate with the natural motion noise n

. Signal

Acts as an input signal to the active noise canceller.

Thereafter, the signal preprocessor 111 obtains the final recovered signal from which the final motion noise is removed to some extent according to the inherent function of the adaptive filter.

As described above, the pulse wave measuring apparatus and the method which are robust to the motion noise according to the present invention have been described with reference to the illustrated drawings, but the present invention is not limited by the embodiments and drawings disclosed herein, and the scope of the technical idea is protected. It can be applied within.

1 is a system configuration to which the pulse wave measuring apparatus according to the present invention is applied.

2 is a block diagram referred to for explaining the configuration of the pulse wave measuring apparatus according to the present invention.

3 is a block diagram referred to to explain the configuration of the host apparatus according to the present invention.

4 is a structural diagram showing the configuration of a pulse wave measuring apparatus according to the present invention.

5 is an exemplary view referred to for describing the operation of the pulse wave measuring apparatus according to the present invention.

FIG. 6 is a view referred to for describing a motion noise removing operation in the pulse wave measuring apparatus according to the present invention.

Claims (15)

A PPG sensor unit disposed on an upper part of a wrist of the examinee and detecting a pulse wave signal from the signal emitted to the upper part of the wrist; An acceleration sensor unit detecting a movement of the subject at a portion where the PPG sensor unit is mounted; And Predicting a motion of the subject from the acceleration signal detected by the acceleration sensor unit, estimating a motion noise model corresponding to the detected motion of the subject, and removing motion noise for the pulse wave signal from the motion noise model Signal preprocessing unit; Pulse wave measurement apparatus robust to the movement noise, characterized in that it comprises a. The method according to claim 1, The signal preprocessor, A signal acquisition unit sampling the pulse wave signal and the acceleration signal in a predetermined unit; And And a noise removing unit for removing noise of the pulse wave signal sampled by the signal obtaining unit by using a high pass filter. The method according to claim 1, The motion noise model is an inherent motion noise model, and the signal preprocessor acquires a highly correlated motion noise based on the inherent motion noise model. The method according to claim 3, The signal preprocessor, A pulse wave measuring apparatus robust to motion noise, characterized by estimating the inherent motion noise model using an autoregressive (AR) model estimator. The method according to claim 3, The signal preprocessor, The pulse wave measuring apparatus robust to the motion noise, characterized by filtering the acceleration signal detected by the inherent motion noise model and the acceleration sensor unit and outputting a motion noise signal having a high correlation with the inherent motion noise of the subject. . The method according to claim 5, The signal preprocessor, A pulse wave measuring apparatus robust to motion noise, characterized by using a finite impulse response (FIR) filter. The method according to claim 5, The signal preprocessor, The pulse wave measuring apparatus robust to the motion noise, characterized in that to remove the motion noise of the pulse wave signal using a motion noise signal having a high correlation with the inherent motion noise of the subject. The method according to claim 1, And a communication unit which transmits the pulse wave signal from which the motion noise has been removed to an external host server. The method according to claim 1, The PPG sensor, A light emitting unit for outputting an optical signal to the upper back of the wrist of the subject; And And a light receiving unit receiving an optical signal emitted by the light emitting unit. The method according to claim 1, The acceleration sensor unit, And a three-axis acceleration sensor, wherein the three-axis acceleration sensor detects the acceleration component of each axis according to the movement of the subject. Detecting a pulse wave signal from a PPG sensor disposed on an upper back portion of a subject's wrist; Detecting a movement of the subject at the site where the PPG sensor is mounted using an acceleration sensor; Estimating a motion noise model corresponding to the movement of the examinee based on the movement information of the examinee and the acceleration signal detected by the acceleration sensor; Removing the motion noise for the pulse wave signal from the motion noise model. The method of claim 11, Acquiring motion noise having a high correlation with the inherent motion noise of the examinee based on the estimated motion noise model and the acceleration signal detected by the acceleration sensor. Robust pulse wave measurement method. The method according to claim 12, Removing the motion noise, A method for measuring pulse wave robust to motion noise, characterized by removing the motion noise of the pulse wave signal using a motion noise signal having a high correlation with the motion noise of the examinee. The method of claim 11, Removing the motion noise, A method of measuring pulse wave robust to motion noise, characterized by repeatedly removing motion noise of the pulse wave signal using an adaptive filter. The method of claim 11, And transmitting the pulse wave signal from which the motion noise has been removed to an external host server.

Images