TW201440725A - Denoising method and apparatus of pulse wave signal and pulse oximetry - Google Patents

Abstract

The disclosure is a denoising method and an apparatus of a pulse wave signal and a pulse oximetry. The denoising method includes steps of a. normalizing a clock signal of a pulse wave obtained by transmitting lights through a living body tissue, b. using the average gradient of the normalized clock signal of the pulse wave to estimate the heart rate, and c. denoising the clock signal of the pulse wave according to the estimated heart rate. Under this circumstance, the denoising method and the apparatus of the pulse wave signal and the pulse oximetry of the present invention achieve the advantages of simple algorithm, such that the accurate measurement is implemented under low perfusion.

Description

Method and device for denoising pulse wave signal and pulse oximeter

The invention relates to the field of medical instruments, in particular to a method and a device for denoising a pulse wave signal in a low perfusion condition and a pulse oximeter.

At present, techniques for non-invasive detection of blood oxygen saturation and heart rate by optical methods have been widely used in the medical field. With the continuous advancement and development of technology, this device for detecting blood oxygen saturation has been continuously miniaturized, and portable oximeters are becoming more and more widely used.

There are many calculation methods used in the existing pulse oximeter, such as infrared spectroscopy photoelectric method. But no matter what calculation method is used to obtain blood oxygen saturation and heart rate, the most basic problem is the noise removal of the pulse wave signal.

Regarding the noise removal of the pulse wave signal, the following two methods are mainly used: 1. Simple noise reduction processing is performed only in the time domain, that is, the pulse wave signal is filtered by a conventional filtering method such as low pass, band pass, and high pass. Noise, this kind of method is relatively simple to process, the amount of calculation is small, and the requirements on the hardware platform are low. 2. Denoising processing is performed by using relatively complex algorithms such as Fourier transform, wavelet transform and adaptive filtering. Such methods have a large amount of computation and relatively high requirements on the hardware platform. At present, due to the miniaturization of oximeters and the need to maintain a high level of performance specifications, high demands are placed on the algorithm, ie, the algorithm is required to be simple and at the same time achieve higher performance specifications. For the first type of method described above, since it does not perform targeted filtering and denoising for different heart rates, it is difficult to achieve accurate measurement of blood oxygen saturation and heart rate under low perfusion (weak pulse beat). In addition, for the above-mentioned Type 2 method, the algorithm is relatively complicated. Even if the blood oxygen saturation and heart rate are accurately measured under low perfusion, the oximeter is miniaturized because of its high requirements on the hardware platform. Very difficult.

Therefore, there is currently no need for a small oximeter (such as a finger-clip oximeter) that is simple in algorithm and can achieve accurate measurement under low perfusion conditions.

An object of the present application is to provide a denoising processing method and apparatus for a pulse wave signal which is simple in algorithm and capable of accurately measuring under low perfusion condition, and a pulse oximeter.

In order to achieve the above object, the present application provides a method for denoising a pulse wave signal, including the following steps: a. Normalizing the pulse wave time domain signal obtained by transmitting light to the living tissue; b. Estimating the heart rate using the average gradient of the normalized pulse wave time domain signal; c. The pulse wave time domain signal is denoised according to the estimated heart rate.

Preferably, in step b, a plurality of pulse wave data points may be arbitrarily sampled within the waveform of the normalized pulse wave time domain signal, and the heart rate PR may be estimated using the following formula:

Where t is the sampling period of the pulse wave data point in milliseconds, and N1 and N2 are the minimum and maximum values of the amplitude of the normalized pulse wave time domain waveform, respectively. Is the average gradient of the pulse wave data points.

Preferably, the average gradient can be calculated using the following formula :

Where M is the number of sampled pulse wave data points, Xi is the value of the i-th pulse wave data point, and i is an integer greater than zero and less than or equal to M.

Preferably, the step of performing low-pass or band-pass filtering processing and baseline drift processing on the pulse wave time domain signal may also be included before step a.

Preferably, the denoising process in step c may comprise low pass or band pass filtering of the pulse wave time domain signal.

The present application also provides a denoising processing device for a pulse wave signal, comprising: a normalization module that normalizes a pulse wave time domain signal obtained by transmitting light to a living tissue; a heart rate estimation module, using normalization The average gradient of the pulse wave time domain signal is used to estimate the heart rate; the denoising processing module performs denoising processing on the pulse wave time domain signal according to the estimated heart rate.

Preferably, the heart rate estimation module can arbitrarily sample a plurality of pulse wave data points within the waveform of the normalized pulse wave time domain signal, and estimate the heart rate PR by using the following formula:

Where t is the sampling period of the pulse wave data point in milliseconds, and N1 and N2 are the minimum and maximum values of the amplitude of the normalized pulse wave time domain waveform, respectively. Is the average gradient of the pulse wave data points.

Preferably, the average gradient It can be calculated using the following formula:

Where M is the number of sampled pulse wave data points, Xi is the value of the i-th pulse wave data point, and i is an integer greater than zero and less than or equal to M.

Preferably, the denoising processing apparatus may further comprise a pre-processing module for performing low-pass or band-pass filtering processing and baseline drift processing on the pulse wave time domain signal prior to normalization.

Preferably, the denoising processing module can perform low pass or band pass filtering on the pulse wave time domain signal.

The present application also provides a pulse oximeter, comprising: a photoelectric sensor that emits light to a living tissue, receives light intensity transmitted through the living tissue, and converts the received light intensity into an electrical signal; A/D conversion Transducing the electrical signal into a digitized pulse wave time domain signal; and a data processing module receiving the pulse wave time domain signal from the A/D converter and processing the pulse wave time domain signal to obtain blood oxygenation Saturation and heart rate, wherein the data processing module includes a denoising processing device for the pulse wave signal as described above.

The application has the following advantages: the algorithm is simple, and at the same time, accurate measurement can be achieved under low perfusion conditions.

1. . . Pulse oximeter

11. . . photoelectric sensor

12. . . A/D converter

13. . . Data processing module

131. . . Normalization module

132. . . Heart rate estimation module

133. . . Denoising processing module

S01, S02, S03. . . step

1 is a flow chart of pulse wave time domain signal denoising processing.
2 is a normalized pulse wave time domain waveform diagram.
Figures 3a and 3b show waveform diagrams of the original pulse wave time domain signal and the pulse wave time domain signal processed in accordance with an embodiment of the present application, respectively.
4 is a configuration diagram showing a pulse oximeter according to an embodiment of the present application.
FIG. 5 is a configuration diagram showing a data processing module according to an embodiment of the present application.

Some exemplary embodiments embodying the features and advantages of the present invention are described in detail in the following description. It is to be understood that the present invention is capable of various modifications in various aspects, and is not to be construed as a limitation.

Embodiments of the present application will be described in detail below. It should be noted that the embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the application.

First, a specific flow of denoising processing of a pulse wave time domain signal will be described with reference to FIG.

In step S01, the pulse wave time domain signal obtained by transmitting light to the living tissue (for example, a human finger) is normalized;

Next, in step S02, the heart rate is estimated by using the average gradient of the normalized pulse wave time domain signal in step S01;

Finally, in step S03, the pulse wave time domain signal is subjected to denoising processing based on the heart rate estimated in step S02.

Next, the configuration of a pulse oximeter according to an embodiment of the present application will be described in detail with reference to FIG.

As shown in FIG. 4, the pulse oximeter 1 includes a photosensor 11 that emits light to a living tissue, receives light intensity transmitted through the living tissue, and converts the received light intensity into an electrical signal; A/ D converter 12 converts the electrical signal into a digitized pulse wave time domain signal; and data processing module 13 receives the pulse wave time domain signal from the A/D converter and processes the pulse wave time domain signal To get blood oxygen saturation and heart rate.

Then, the configuration of the data processing module 13 according to an embodiment of the present application will be described in detail with reference to FIG.

As shown in FIG. 5, the data processing module 13 includes a normalization module 131 that normalizes the pulse wave time domain signal obtained by transmitting light to the living tissue; the heart rate estimation module 132 uses the normalized The average gradient of the pulse wave time domain signal is used to estimate the heart rate; and the denoising processing module 133 performs denoising processing on the pulse wave time domain signal based on the estimated heart rate.

The process of denoising the pulse wave time domain signal according to an embodiment of the present application will be described in detail below.

First, the first low-pass or band-pass filtering process (ie, high-frequency denoising) and baseline drift processing are performed by the pulse wave time domain signal obtained by transmitting light to the living tissue. The low-pass or band-pass filtering described above can eliminate high-frequency noise in the pulse wave time domain waveform, that is, eliminate glitch in the pulse wave time domain waveform; and the baseline drift processing can eliminate the DC fluctuation of the pulse wave time domain signal. With the first low-pass or band-pass filtering and baseline drift processing, the subsequent rough evaluation of the heart rate step can be more accurate.

Then, the pulse wave time domain signal processed by the high frequency denoising and baseline drift described above is normalized within a range [N1, N2], and the normalized pulse wave time domain signal is as shown in FIG. 2. In Fig. 2, the horizontal axis is the sampling point of the waveform of the normalized pulse wave time domain signal, and the vertical axis is the normalized amplitude (the maximum value is N2 and the minimum value is N1).

Then, arbitrarily select M pulse wave data points from the sampling points, the value of the M data points is Xi (0<i ≤ M), and sequentially calculate the gradient between each two adjacent data points. Thus, the average gradient of the M pulse wave data points can be calculated using the following formula (1) as follows:

(1)

Let the period of the pulse wave data sampling point be t milliseconds, and the number of data points included in one minute is 60000/t. As shown in Fig. 2, in the sampling points included in a complete pulse wave period, the sum of the gradients between adjacent sampling points is about 2 (N2-N1), so the sampling points included in one complete pulse wave period. The number is approximately: . Therefore, heart rate PR and pulse wave data point average gradient There is a relationship of the following formula (2):

(2)

Finally, targeted low pass or band pass filtering is performed based on the above rough estimated heart rate to eliminate as much noise as possible from the current pulse wave data.

The effect of the pulse wave time domain signal denoising process using the present application will be described below with reference to Figs. 3a and 3b. Fig. 3a shows a pulse wave time domain signal with a perfusion degree of 0.1% and a heart rate of 60. It can be seen that the pulse wave time domain signal has many burrs and large fluctuations. FIG. 3b shows the pulse wave time domain waveform after the above-mentioned targeted pulse wave time domain signal denoising process. It can be seen that the pulse wave time domain waveform processed by the present application has been smoother, such that It is relatively easy to accurately calculate the oxygen saturation and heart rate in the pulse wave time domain waveform. It can be seen that, after the pulse wave time domain signal denoising process of the present application described above, it is feasible to achieve an accurate measurement of the perfusion degree of 0.1% in a miniaturized oximeter (such as a finger-clip oximeter).

The data processing module, the denoising processing device, the normalization module and the heart rate estimation module in the above embodiments may be designed in a single chip microcomputer or integrated in other semiconductor chips.

While the present invention has been described in detail by way of example embodiments, the scope of the present application is not limited to the embodiments described above, and various modifications and changes can be made by those skilled in the art without departing from the scope of the application. Conception.

S01, S02, S03. . . step

Claims (11)

A method for denoising a pulse wave signal, comprising the steps of:
a. Normalizing the pulse wave time domain signal obtained by transmitting light to the living tissue;
b. Estimating the heart rate using the average gradient of the normalized pulse wave time domain signal; and c. The pulse wave time domain signal is denoised according to the estimated heart rate. The denoising processing method according to claim 1, wherein in step b, a plurality of pulse wave data points are arbitrarily sampled in a waveform of the normalized pulse wave time domain signal, and utilized as follows Formula to estimate heart rate PR:

Where t is the sampling period of the pulse wave data point in milliseconds, and N1 and N2 are the minimum and maximum values of the amplitude of the normalized pulse wave time domain waveform, respectively. Is the average gradient of the pulse wave data points. The denoising processing method according to claim 2, characterized in that the average gradient is calculated by using the following formula :


Where M is the number of sampled pulse wave data points, Xi is the value of the i-th pulse wave data point, and i is an integer greater than zero and less than or equal to M. The denoising processing method according to claim 1, wherein before the step a, the step of performing low-pass or band-pass filtering processing and baseline drift processing on the pulse wave time domain signal is further included. The denoising processing method according to claim 1, wherein the denoising processing in step c comprises performing low-pass or band-pass filtering on the pulse wave time domain signal. A denoising processing device for a pulse wave signal, comprising:
a normalization module that normalizes a pulse wave time domain signal obtained by transmitting light to a living tissue;
The heart rate estimation module estimates the heart rate using the average gradient of the normalized pulse wave time domain signal; and the denoising processing module performs denoising on the pulse wave time domain signal according to the estimated heart rate. The denoising processing apparatus according to claim 6, wherein the heart rate estimating module arbitrarily samples a plurality of pulse wave data points in a waveform of the normalized pulse wave time domain signal, and uses the following formula To estimate heart rate PR:

Where t is the sampling period of the pulse wave data point in milliseconds, and N1 and N2 are the minimum and maximum values of the amplitude of the normalized pulse wave time domain waveform, respectively. Is the average gradient of the pulse wave data points. A denoising processing apparatus according to claim 7 of the patent application, characterized in that the average gradient Use the following formula to calculate:

Where M is the number of sampled pulse wave data points, Xi is the value of the i-th pulse wave data point, and i is an integer greater than zero and less than or equal to M. The denoising processing apparatus according to claim 6 is characterized in that it further comprises a preprocessing module for performing low-pass or band-pass filtering processing and baseline drift processing on the pulse wave time domain signal prior to normalization. The denoising processing apparatus according to claim 6 is characterized in that the denoising processing module performs low-pass or band-pass filtering processing on the pulse wave time domain signal. A pulse oximeter characterized by comprising:
a photoelectric sensor that emits light to a living tissue, receives light intensity transmitted through the living tissue, and converts the received light intensity into an electrical signal;
An A/D converter that converts the electrical signal into a digitized pulse wave time domain signal; and a data processing module that receives the pulse wave time domain signal from the A/D converter and processes the pulse wave time domain signal To get blood oxygen saturation and heart rate,
The data processing module includes: a denoising processing device for a pulse wave signal according to any one of claims 6 to 10.