KR101514151B1 - Noise robust algorithm for ppg signal measurement - Google Patents

Abstract

The present invention relates to a PPG signal measuring method which is strong against motion noise and is able to solve problems caused by high frequency noise and motion noise generated when measuring a PPG signal which is one of biological signals. The PPG signal measuring method includes: a PPG signal measuring step through multiple channels receiving two or more channels of a PPG signal from an end part of a human body; a high frequency noise removing step of removing the signal noise of a high frequency band using a low-pass filter from the PPG signal measured in the PPG signal measuring step through the multiple channels; a reference signal setting step to determine whether motion noise exists in the PPG signal from which a high frequency is removed through the high frequency noise removing step; a time correcting step for a comparison between multi-channel signals since a time difference exists between a signal measured through multiple channels and a signal detected by a distance difference between each measurement position and a heart; and a motion noise determining and signal selecting step to output a normal signal by comparing the cycle time of the signal of multiple channels corrected in the time correcting step with the cycle time of a reference signal.

Description

TECHNICAL FIELD [0001] The present invention relates to a method of measuring a PPG signal,

The present invention relates to a PPG signal measuring method robust against motion noise, and more particularly, to a motion-robust PPG signal measuring method capable of eliminating problems caused by high frequency noise and motion noise occurring in PPG signal measurement, ≪ / RTI >

Recently, interest in health and body has increased, and as the incidence of diseases such as hypertension and diabetes increases, there is a growing interest in techniques for confirming health in real time. In addition to the increased interest, research on biological signal processing techniques such as extracting characteristics of bio-signals, developing algorithms for recognizing and extracting bio-signals, and so on are conducted in various fields such as medical field, psychology and psychology have.

Electroencephalography (EEG), Electrocardiography (ECG), Electromyography (EMG), Photoplethysmography (PPG) and Electrooculography (EOG) PPG is a signal that is easy to monitor in real time in various biological signal processing. In the case of EEG and ECG, the electrodes should be attached to the scalp and chest, and movement is very limited. However, since PPG uses one sensor attached to the body end region such as finger, earlobe, and toe, It is advantageous in that the user's motion can be secured relative to the EEG. However, because of the problem that the signal is distorted due to the noise of the measurement environment itself and the operation of the user, a process of removing such noise is required during signal processing.

In recent years, there is a limitation to remove motion noise, though there is a method to remove motion noise using hardware techniques such as acceleration sensor and gyro sensor to remove motion noise. The dynamic noise can occur not only at high frequency but also at all frequency bands. Among the dynamic noise of various bands, noise occurring in the high frequency band can be removed, but noise occurs in the frequency band similar to the bio signal. It can not be removed.

Various new methods for eliminating such motion artifacts have been proposed. For example, in "Dispersion of motion noise in an improved PMAF-based PPG signal for a U-healthcare system" published in a well-known paper, We have developed a monitoring system with a clock type and proposed a new algorithm. The proposed method uses a periodic moving average filter (PMAF), which is a cyclic moving average filter using adaptive filters or pseudo periodicity of PPG signals, among the methods used in the past. It is confirmed that noise cancellation is more effective than the method.

However, this method is also effective in removing motion noise at frequencies higher than the cycle, but it is not easy to detect the cycle at low frequencies such as respiration and arm movement, and there is a problem that the motion noise can not be completely removed.

Korean Patent Publication No. 10-2010-0008239 (Jan. 25, 2010)

The present invention provides a robust PPG signal measuring method capable of eliminating problems caused by high frequency noise and motion noise generated in PPG signal measurement, which is one of living body signals, It has its purpose.

In order to achieve the above object, the present invention provides a robust motion-robust PPG signal measuring method capable of eliminating problems caused by high-frequency noise and motion noise occurring in PPG signal measurement, which is one of biological signals, Measuring a PPG signal through a first antenna; Removing a signal noise in a high frequency band using a low pass filter (LPF) in the multiple received PPG signals; Selecting a reference signal; Correcting a temporal difference between multiplexed signals by applying a cross-correlation technique to remove motion noise from the multi-channel PPG signal based on the set reference signal; And comparing the magnitude of the signal with a reference signal to output a steady signal.

In the PPG signal measurement method robust to such motion noise, the low-pass filter uses an IIR filter, and the reference signal has a PPG sensor attached at the body end, and when a normal signal is measured, The normal signal is compared with the reference signal to adjust the magnitude of the multiple signal to be similar to the reference signal and the delayed signal is shifted by the delay time difference through the time correction method, And the PPI is measured and output through comparison with the reference signal period.

Also, when comparing the period of the reference signal with the period of the reference signal, if the PPI interval exists within the reference signal period and the error range, it is determined that the signal is normal. If there is the PPI interval above the error range, And if the PPI of the other signal is within the normal range, it is determined that the PPI is a normal signal and is outputted. If the normal signal is more than one signal, a signal most similar to the reference signal is selected .

According to the method for measuring the PPG signal robust to motion noise according to the present invention, the PPG signal occupied by the noise for 20% of the time is measured by using the two multiple measurement signals, The distortion time can be reduced to 9.56%, and the time difference between the signals measured by the two or more-channel sensor is analyzed to effectively remove motion noise.

Brief Description of the Drawings Fig. 1 is a block diagram showing steps of a robust PPG signal measuring method according to the present invention
FIG. 2 is a graph showing the characteristics of the IIR-Butterworth filter in the high-frequency noise canceling step of the PPG signal measuring method robust against motion noise according to the present invention
FIG. 3 is a graph showing the relationship between two reference signals of two different users in the reference signal setting step of the PPG signal measuring method robust to motion noise according to the present invention
FIG. 4 is a graph showing the results of two types of graphs using the cross correlation method of the time correction step of the PPG signal measuring method robust to motion noise according to the present invention
5 is a detailed configuration diagram of the motion noise determination and signal selection of the PPG signal measurement method robust to the motion noise according to the present invention

The present invention may have various modifications and various embodiments, and specific embodiments are illustrated in the drawings and described in detail in the detailed description. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the present invention will be described in detail with reference to the accompanying drawings.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings.

FIG. 1 is a block diagram showing steps of a robust PPG signal measuring method according to the present invention.

As shown in FIG. 1, the method for measuring a PPG signal robust to motion noise according to the present invention includes steps of measuring a PPG signal through a plurality of channels for receiving two or more PPG signals at a body end such as a finger, a toe, and an earlobe, A high frequency noise removing step (S120) for removing a signal noise in a high frequency band from a PPG signal measured in a PPG signal measuring step (S110) using a low pass filter, and a high frequency noise removing step (S120) A reference signal setting step (S130) for determining whether the signal is a signal in which motion noise is present in the removed PPG signal (S130), and a signal measured through multiple channels is a signal detected due to a difference in distance between each measurement position and the heart Since there is a temporal difference, the cycle time of the signals of the plurality of channels corrected through the time correction step (S140) and the time correction step (S140) And a dynamic noise determination and signal selection step (S150) for outputting a normal signal in comparison with the time.

In step (S110) of measuring a PPG signal over multiple channels, it is important to determine from which of the various body parts where the PPG signal can be measured to measure the signal. In general, PPG signals are measured at the extremities of the body, such as the fingers, toes, and earlobes. When measured at the end of the body, motion noise is likely to occur due to user's movement. Thus, when measured at the end of the body at the same site, the received signal is affected by the same motion, and all of the multiple received signals contain motion artifacts. It is difficult to remove the dynamic noise if the received signals contain dynamic noise.

On the other hand, when the signal is measured at the end of the body that can have independent motion, motion noise may occur in different sections. Therefore, by selecting a signal that does not include motion noise among the measured signals, Can be removed.

However, there is a problem that occurs when the measurement is performed at different sites. Because they are not measured at the same end of the body, the magnitude and characteristics of the measured signals are slightly different. Therefore, when the signal is measured through the multiple reception at the body end where the motion occurs independently, the size of each signal should be made similar.

In the high frequency noise removing step (S120) of the present invention, the PPG signal measured in the step of measuring the PPG signal through the multiple channels (S110) has high frequency band signal noise due to mechanical and physical characteristics. The signal noise of this high frequency band is an obstacle to the detection of the period of the PPG signal and should be removed. A low-pass filter is used to eliminate high-frequency noise. Due to the characteristics of the PPG signal, high frequency noise appears near the spectrum of the PPG signal. In order to remove the noise, a high-dividing filter must be used to eliminate high-frequency noise. Therefore, an IIR filter which does not exhibit ripple phenomenon in the band of interest and has excellent filter splitting ability is used. Various types of IIR filters are available, such as Butterworth, Chebyshev, and Elliptic, but commonly used Butterworth filters. The IIR filter used is a 4th order Butterworth filter and the cutoff frequency is set to 7Hz. The system function of the filter is shown in Equation (1) below.

The characteristic of the fourth IRR-Butterworth filter of the high-frequency noise removing step of the PPG signal measuring method robust against motion noise according to the present invention is shown in FIG.

In the reference signal setting step (S130), a reference signal is introduced to determine motion noise in a signal from which high-frequency noise has been removed. Generally, when a signal is measured by attaching a PPG sensor to a body end portion, an abnormal signal whose amplitude is initially uneven is measured based on the characteristics of the sensor, and a normal signal with an amplitude is measured after about 2 seconds. Therefore, when a normal signal is measured, the reference signal is determined by comparing the first period and the second period of the PPG signal. Since the PPG signal has a period of about 0.6 to 1.2 seconds, the data of about 3 seconds of the measured PPG signal is analyzed and the first period is set as the reference signal. Since the characteristic of the measured PPG signal varies from person to person, they have different reference signals.

Figure 3 is a reference signal of any two persons. This is because the reference signal of one cycle obtained by analyzing data for each person has different characteristics. Therefore, it is easy to judge the motion noise by using the reference signal that reflects the characteristics of the user's pulse wave signal as it is.

In the time correction step S140, in order to remove the motion noise based on the set reference signal, the time difference between the multiplexed signals should be corrected. Any two signals are assumed to illustrate how to calibrate. The distance from the first receiver to the heart is different from the distance from the second receiver to the heart. Therefore, a time difference due to the distance difference occurs. In the process of replacing the noise-generated signal with the normal signal, the phase difference does not coincide due to the time difference between the two signals. Therefore, in order to solve this problem, the time difference between the two signals must be corrected. The time difference between the two signals is calculated using the following Equation 2 using the cross-correlation technique.

은 첫 번째 측정부위의 신호 x와 두 번째 측정부위의 신호 y를 상호상관기법을 적용한 결과이다. 신호 x와 신호 y가 N의 길이를 가진 데이터라고 할 때

는 -(N-1)에서 N-1의 길이를 가진 데이터가 된다.

의 m번째 데이터는 부호에 따라 달리 계산된다. m이 양수일 경우

와 같이 계산되며 m이 음수일 경우에는 m이 양수일 경우의 데이터가 시간이 반전된 형태이다.here

Is the result of applying the cross-correlation technique between the signal x of the first measurement site and the signal y of the second measurement site. Assuming that the signal x and the signal y are data having a length of N

Becomes data having a length of N-1 in - (N-1).

M < / RTI > of data is calculated differently depending on the sign. If m is positive

If m is a negative number, the data in the case where m is a positive number is inverted in time.

When the cross - correlation technique is applied, the signal has two types. Two types are the form in which no time error occurs and the form in which time error occurs. A graphical representation of these two forms is shown in FIG.

4 (A) shows a state in which a time error does not occur. This indicates that the cross-correlation technique is applied to the two signals when the PPG signal is measured at the same site. It can be seen that the y-axis is the largest at the zero point of the x-axis of the graph. In this way, in the form of the time error, the distance between the x-axis and the y-axis at the large point where the size of the y-axis increases is reflected in the time error. In this way, we extract the data of about 3 seconds from the two signals, find the time difference between the two signals by the cross-correlation technique, and correct the time by moving the data of one signal to the time difference.

In the motion noise determination and signal selection step (S150), high frequency noises are removed, motion compensation is performed between the multiplexed signals, and motion noise is removed. Refer to the selective coupling technique that is mainly used in the field of communication to remove motion noise. The selective combining method is a method of outputting the most ideal signal by comparing the signals received from each branch by combining signals obtained from multiple receivers and outputting the best signal for each section. The overall flow of motion noise determination and signal selection is shown in FIG.

When measuring the PPG signal at different body end regions, since the PPG sensor detects the transmitted light by detecting the pulse wave, the intensity of the received signal is different due to factors such as thickness of the skin. The size of such a signal is compared with a reference signal, and the magnitude of each of the multiple signals is adjusted through the interchannel size adjustment step S151 so as to be similar to the reference signal.

Each of the signals that have corrected the intensity of the signal can know the time difference between the signals using the time correction method. Using this time difference, phase adjustment is performed by shifting the time-delayed signal by a time difference through an inter-channel phase adjustment step (S152).

The PPI is measured in step S153 to determine whether each of the signals matching the intensity and phase of the signal is a normal signal or a signal including motion noise. Peak to Peak Interval (PPI) means a time interval between a maximum value of one periodic signal and a maximum value of the next periodic signal.
If the period of the reference signal is compared with the measured PPI, if there is a PPI interval within the period of the reference signal and within a slight error range, it is judged as a normal signal. If a sudden change in PPI is detected, the interval is determined as the interval in which the noise is generated, and the PPI of the other signal is detected. If the PPI of the other signal is within the normal range, it is judged as a normal signal and outputted. If the normal signal of the received signal is more than one signal, the signal which is the most similar to the reference signal And signal selection step S154.

The embodiments of the present invention described in the present specification and the configurations shown in the drawings relate to the most preferred embodiments of the present invention and are not intended to encompass all of the technical ideas of the present invention so that various equivalents It should be understood that water and variations may be present. Therefore, it is to be understood that the present invention is not limited to the above-described embodiments, and that various modifications may be made without departing from the spirit and scope of the invention as defined in the appended claims. , Such changes shall be within the scope of the claims set forth in the claims.

Claims (7)

a) measuring the PPG signal through multiple channels at the body end;
b) removing signal noise in the high frequency band using a low pass filter (LPF) in the multiple received PPG signals;
c) selecting a reference signal;
d) correcting a temporal difference between the multiplexed signals by applying a cross-correlation technique to remove motion noise in the multi-channel PPG signal based on the set reference signal;
e) comparing the magnitude of the signal with a reference signal to output a steady signal;
/ RTI > The method of claim 1, further comprising the steps of:
The method according to claim 1,
Wherein the low-pass filter uses an IIR filter.
The method according to claim 1,
Wherein the reference signal is selected by comparing a first period and a second period of the PPG signal when a normal signal is measured by attaching a PPG sensor to a body end region.
The method according to claim 1,
The normal signal compares the magnitude of the signal with the reference signal to adjust the magnitude of each of the multiple signals so that there is no intensity difference between the reference signal and the signal and adjusts the phase by moving the delayed signal by the delay time difference through the time correction method Measuring the PPI, and comparing the measured signal with the reference signal period to output the PPG signal.
5. The method of claim 4,
And comparing the period of the reference signal with the period of the measured PPI, if the period of the reference signal and the interval of the PPI exist within the error range, the signal is determined to be a normal signal.
5. The method of claim 4,
When comparing the period of the reference signal with the period of the reference signal, if there is a PPI interval above the period of the reference signal and the error range, the PPI of the other signal is determined as the period in which the motion noise occurs. If the PPI of the other signal is within the normal range And if it is present, determines that the signal is a normal signal and outputs the signal as a normal signal.
5. The method of claim 4,
And selecting a signal having a minimum PPI interval from the reference signal, which is determined as a normal signal, when the normal signal is more than one signal.

Images