CN111803041A

CN111803041A - Human body pulse peak prediction method and device

Info

Publication number: CN111803041A
Application number: CN202010568815.5A
Authority: CN
Inventors: 周淼淼; 聂荣志; 彭波; 储涛; 熊泉; 赵子健; 张广维
Original assignee: Zhejiang Shuyuan Intelligent Technology Co ltd
Current assignee: Zhejiang Shuyuan Intelligent Technology Co ltd
Priority date: 2020-06-19
Filing date: 2020-06-19
Publication date: 2020-10-23
Anticipated expiration: 2040-06-19
Also published as: CN111803041B

Abstract

The embodiment of the invention provides a human body pulse peak prediction method and a human body pulse peak prediction device. The method comprises the following steps: acquiring a human body pulse signal; and performing first-order derivation, positive value screening, continuity screening, average value threshold value screening and peak searching judgment processing on the human body pulse signals, and predicting the human body pulse peaks to obtain a prediction result. Meanwhile, the prediction method can be carried out for a long time, and on the basis, the prediction method can be applied to predicting the spectral data of the pulse peak-valley value for imaging, qualitative and quantitative analysis, so that the practicability is good.

Description

Human body pulse peak prediction method and device

Technical Field

The invention relates to the technical field of human pulse prediction, in particular to a human pulse peak prediction method and a human pulse peak prediction device.

Background

Arteries and capillaries in the human body contract and relax under the influence of heartbeat pulses, so that the components of human tissues repeatedly change: when the heart contracts, blood vessels expand and blood components increase; during diastole, blood vessels contract and blood components decrease. This phenomenon is equivalent to the change of the proportion of equivalent sample components in the spectrum measurement, which causes the change of the measurement result, so that the pulse information can be obtained by collecting the spectrum information of the human body. For example, the photoplethysmography is to measure the absorption spectrum of human tissue along with the change of pulse, thereby reflecting the pulse condition in the human body and the component information of partial human blood.

In the photoplethysmography, the most widely used method is to measure various information by using an absorption spectroscopy, and when the absorption spectroscopy is performed, a single-point detector is generally used and matched with a high-speed analog-to-digital conversion module, so that the measurement of 100-1000Hz can be realized. The human body pulse period is generally 500-1000 ms, correspondingly, as long as the acquisition frequency exceeds 50 Hz, the basic pulse signal can be acquired, and the signal can be very clear when the acquisition frequency exceeds 200 Hz. However, when some work is performed, the corresponding detectors cannot acquire at high speed, or the signal-to-noise ratio in a high-speed acquisition mode is not good, for example, in human body infrared imaging, if infrared images of a human body during pulse contraction and pulse relaxation are measured respectively, the high-speed acquisition mode cannot be used. At the moment, a photoplethysmography device is needed to capture human body pulse signals, judge the human body pulse conditions and drive the infrared camera to acquire the signals at a proper time. Such a need can be classified as a need to predict the pulse condition in advance.

Because the human body pulse is not stable periodic change and is influenced by various factors such as mood change, health condition change, motion state change and the like of people, the human body pulse is not stable, and lines at the time points of pulse contraction and relaxation can float, so that people can not analyze the pulse by using a mode of processing periodic behaviors so as to prejudge the pulse contraction and relaxation.

Disclosure of Invention

The invention aims to provide a human body pulse peak prediction method and a human body pulse peak prediction device, so as to solve the problem that pulse analysis is influenced due to unstable human body pulse.

In order to solve the technical problems, the technical scheme of the invention is as follows:

according to an aspect of the present invention, there is provided a method for predicting a peak of a human pulse, including:

acquiring a human body pulse signal;

carrying out first-order derivation, positive value screening, continuity screening, average value threshold value screening and peak searching judgment processing on the human pulse signals;

and predicting the pulse peak of the human body to obtain a prediction result.

Further, acquiring a human pulse signal, comprising:

collecting human body pulse analog voltage signals;

and obtaining a digital signal of pulse intensity according to the human pulse analog voltage signal.

Further, the positive value screening processing is carried out on the human pulse signals, and the positive value screening processing comprises the following steps:

and setting all negative derivatives in the human pulse signals to be zero.

Further, the human pulse signals are subjected to continuous screening processing, which comprises the following steps:

and (4) continuously setting data points which are positive values in the first derivative data of all the data points after the first derivative, and setting the data points with insufficient continuity as zero according to a continuity positive value quantity threshold.

Further, the average value threshold value screening processing is carried out on the human pulse signals, and the average value threshold value screening processing comprises the following steps:

and averaging all non-zero data points subjected to continuity screening, and setting all data points smaller than the product of the average value multiplied by the average value proportion threshold value to zero according to the average value proportion threshold value.

And further, outputting the prediction result to the terminal.

Further, still include: the terminal comprises but is not limited to an oscilloscope, a computer, a triggerable human body infrared imaging system and a triggerable human body spectrum acquisition system.

In another aspect of the present invention, there is provided a human pulse peak prediction apparatus, including:

the acquisition module is used for acquiring a human body pulse signal;

the processing module is used for carrying out first-order derivation, positive value screening, continuity screening, average value threshold value screening and peak searching judgment processing on the human pulse signals;

and the prediction module is used for predicting the human pulse peak to obtain a prediction result.

Further, the obtaining module is specifically configured to:

collecting human body pulse analog voltage signals;

Further, the processing module is specifically configured to:

and setting all negative derivatives in the human pulse signals to be zero.

The scheme of the invention at least comprises the following beneficial effects:

according to the scheme, the human body pulse signal is obtained, and a pulse peak prediction algorithm is utilized to carry out a series of processing on the human body pulse signal, so that the human body pulse peak is predicted, and the human body pulse peak prediction method has the advantages of high stability, high universality and high accuracy. Meanwhile, the prediction method can be carried out for a long time, and on the basis, the prediction method can be applied to predicting the spectral data of the pulse peak-valley value for imaging, qualitative and quantitative analysis, so that the practicability is good.

Drawings

Fig. 1 is a flowchart illustrating a method for predicting a peak of a human pulse according to a first embodiment of the present invention.

Fig. 2 is a device connection diagram of a human pulse peak predicting apparatus according to a first embodiment of the present invention.

Fig. 3 is a schematic structural diagram of a human pulse peak prediction device according to a second embodiment of the present invention.

Fig. 4 is a circuit diagram of a human pulse peak predicting device according to a second embodiment of the invention.

FIG. 5 is a flowchart illustrating a method for predicting a peak of a human pulse according to a second embodiment of the present invention.

Fig. 6 is a schematic diagram of a pulse peak prediction algorithm.

Fig. 7 is a diagram showing the effect of the pulse signal and the pulse peak prediction signal output by the USB _ MiniB interface.

Detailed Description

Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

Example one

As shown in fig. 1, an embodiment of the invention provides a method for predicting a peak of a human pulse, including:

s1, obtaining a human pulse signal;

s2, carrying out first-order derivation, positive value screening, continuity screening, average value threshold value screening and peak searching judgment processing on the human pulse signals;

and S3, predicting the pulse peak of the human body to obtain a prediction result.

The human body pulse peak prediction method has the advantages of high stability, high universality and high accuracy. Meanwhile, the prediction method can be carried out for a long time, and on the basis, the prediction method can be applied to predicting the spectral data of the pulse peak-valley value for imaging, qualitative and quantitative analysis, so that the practicability is good.

In an optional embodiment of the present invention, the acquiring a human pulse signal includes:

collecting human body pulse analog voltage signals;

The pulse acquisition sensor can be used for acquiring human pulse analog voltage signals, and the pulse analog voltage signals are converted into digital signals of pulse intensity by analog-to-digital conversion (AD), so that subsequent processing is facilitated.

In an optional embodiment of the present invention, the positive value screening processing of the human pulse signal includes:

and setting all negative derivatives in the human pulse signals to be zero.

In an optional embodiment of the present invention, the continuously filtering the human pulse signal includes:

In an optional embodiment of the present invention, the average threshold value screening process of the human pulse signals includes:

In an optional embodiment of the present invention, the method further comprises:

and outputting the prediction result to a display terminal.

The display terminal comprises but is not limited to an oscilloscope, a computer, a triggerable human body infrared imaging system, a triggerable human body spectrum acquisition system and the like, and is convenient for people to check or use.

As shown in fig. 2, an embodiment of the present invention provides a human pulse peak predicting device, including:

the acquisition module is used for acquiring a human body pulse signal;

The human body pulse peak prediction method has the advantages of high stability, high universality and high accuracy. Meanwhile, the prediction device can predict for a long time, and on the basis, the prediction device can be applied to predict the spectral data of the pulse peak-valley value for imaging, qualitative and quantitative analysis, so that the practicability is good.

In an optional embodiment of the present invention, the obtaining module is specifically configured to:

collecting human body pulse analog voltage signals;

In an optional embodiment of the present invention, the processing module is specifically configured to:

and setting all negative derivatives in the human pulse signals to be zero.

In an optional embodiment of the invention, the apparatus further comprises:

and the output module is used for outputting the prediction result to the display terminal.

It should be noted that the apparatus is an apparatus corresponding to the method described in fig. 1, and all the implementations of the illustrated method are applicable to the embodiment of the apparatus, and the same technical effects can be achieved.

Example two

The human body pulse peak prediction system provided by the embodiment of the invention comprises the human body pulse peak prediction device described in the embodiment, for example, an STM32 single chip microcomputer and a pulse sensor, as shown in fig. 3 and 4, and the system mainly comprises a pulse acquisition sensor and a single chip microcomputer with a pulse peak prediction function. The singlechip has two external interfaces: the single chip microcomputer is connected with a computer through the USB _ MiniB interface by using a USB _ MiniB connecting line and is used for power supply of the single chip microcomputer and communication between the single chip microcomputer and the computer; the singlechip is connected with equipment such as an oscilloscope and the like through a BNC interface by using a BNC connecting line and is used for outputting a pulse peak prediction signal. Carry out the 5V power supply for singlechip development board through +5V pin and GND pin in the USB _ MiniB interface, carry out the 3.3V power supply for pulse acquisition sensor through the Vcc output pin and the GND pin of singlechip development board.

The pulse acquisition sensor outputs a pulse analog voltage signal to an AD conversion (analog-to-digital conversion) unit of a single chip microcomputer (a processing module in the embodiment) through a pulse pin; the singlechip converts the pulse analog voltage signal into a digital signal of pulse intensity through the AD conversion unit and then transmits the digital signal to the pulse peak prediction algorithm unit; the pulse peak prediction algorithm unit outputs a pulse peak prediction signal after calculation and then transmits the pulse peak prediction signal to the USB unit and the GPIO unit; the single chip microcomputer is connected with a computer through a USB _ MiniB interface of a USB (Universal serial bus) unit and outputs pulse information and a pulse peak prediction signal; meanwhile, the pulse peak prediction signal is output in the form of a pulse signal through a BNC interface of the GPIO unit.

The pulse acquisition sensor adopted by the system is an electrocardio pulse HRV heart rate detection induction biological simulation sensor. The power supply voltage is 3.3-5V, and the working current is 10 mA; the light source used was an LED light source with a peak wavelength of 515 nm; the photoreceptors used are photodiodes; the filter circuit uses a low-pass filter circuit; the amplifier circuit uses a transimpedance amplifier circuit; the pulse analog voltage signal is output to be 0-3.3V, and the output current is 0-4 mA.

The singlechip that this system adopted is STM32F103C8T6, and under normal operating condition, operating frequency 72MHz, 64KFlash memory, 20 kbyte's SRAM, 2.0-3.6V power supplies, 2 12bit AD analog to digital converters, 3 USART hardware serial ports. The kernel is a CPU of Cortex-M3 with ARM32 bit, and the debugging mode is SWD debugging.

The AD conversion unit adopts a 12bit successive approximation type AD conversion unit in an STM32F103C8T 6. The AD conversion unit mainly functions to convert the pulse analog voltage signal output by the pulse sensor into a digital signal of pulse intensity. The AD conversion unit uses a single-channel single-sampling mode, the sampling channel is channel 1, and the sampling time is 1.5 clock cycles. The sampling conversion in the AD conversion is performed in such a manner that the sampling conversion is interrupted in the AD conversion. The total sampling conversion time is 1us, the precision of the digital signal converted into the pulse intensity is 1mv, and the data alignment mode of the digital signal is right alignment. Sampling is carried out once every 5ms, and a digital signal of the pulse intensity is output to a pulse peak prediction algorithm unit. The AD conversion unit output frequency should be greater than 200 Hz.

The pulse peak prediction algorithm unit is an algorithm based on a first derivative, and the contents of the algorithm comprise processes of smoothing, first derivative finding, positive value screening, continuity screening, average value threshold screening, peak searching judgment and the like. The calculation needs the data quantity provided by the AD conversion unit to be more than 500, and the time width corresponding to the data point is more than 2.5 seconds.

In this embodiment, the data frequency input to the pulse peak prediction algorithm unit (abbreviated as peak finding algorithm) is 200Hz, and the corresponding data point time interval is 5 ms. The number of data points stored is N = 400. Let the input data be y (i), i = 1-N, and the peak finding algorithm (pseudo code) is as follows:

//1, median filtering

y1=smooth(y,3);

//2, derivative, data length becomes N-1

Y2=diff(y1);

N=N-1;

V/3, Positive value screening, is the process of setting all negative derivatives to zero

for i=1:N

if y2(i)>0

y3(i)=y2(i);

else

y3(i)=0;

And 4, continuity screening, namely counting the number of continuous positive values in the first-order derivative data of all the data points after the previous process, and setting the data points with insufficient continuity as zero according to a threshold value of the number of continuous positive values. Here continuity positive number threshold =12

for i=1:N

If y3(i) number of consecutive positive values left and right >12

y4(i)=y3(i);

else

y4(i)=0;

And/5, mean screening, namely averaging all non-zero data points after the previous process, and setting all data points smaller than (mean value-mean value ratio threshold value) to be zero according to the mean value ratio threshold value. Here mean ratio threshold =1.5

avg=average(y4);

for i=1:N

if y4(i)>avg*1.5

y5(i)=y4(i);

else

y5(i)=0;

V/6, peak finding

a=8,b=3;

peak=[],n_peak=0

for (i=N-1-b;i>a;i--)

if (y5(i)>y5(i-1:i-a)&&y5(i)>y5(i+1:i+b))

n_peak++,peak(n_peak)=i

V/initialization Return value

re=False;

T=0;

error=0;

//7, determination of the number of peaks

if n_peak>=2

T=(peak(1)-peak(2))* 5;

else

error=1;

return(error, re, T)

//8, period determination

if T<500 || T>1100

error=2;

return(error, re, T)

//9, determination of Peak position

if peak (1) = N-3-k// k is a prediction advance time option, generally taking a value of 0-6, and correspondingly advancing by 30-0 ms.

re=True;

else

re=False;

return(error, re, T)；

The algorithm ends.

In the above algorithm, re is the prediction judgment value of the pulse peak prediction algorithm, T is the predicted pulse period, and error is the error information output. The remaining variable arrays are intermediate arrays. smooth (), diff (), average () are median filter functions, first derivative functions, array mean functions.

The algorithm 1-5 is a core step, can screen out the first derivative peak value corresponding to the rising process before the pulse peak appears, and can identify the first derivative peak value before the peak appears and when the first derivative signal is not complete. The process is shown in fig. 6.

When re = True, the singlechip outputs a pulse peak prediction signal through a USB _ miniB interface of the USB unit; meanwhile, the pulse peak prediction signal is output in the form of a pulse signal through a BNC interface of the GPIO unit. The USB unit uses a self-contained USB _ VCP unit in STM32F103C8T6, and can realize USB _ VCP communication with other devices.

When the single chip microcomputer in the human body pulse peak prediction system is connected with a computer through a USB (universal serial bus) line. There are two communication modes: in the first mode, the computer can send a pulse inquiry command 0x 000 x060x 720 x 650 x 610 x680xA6 to the single chip microcomputer, and the single chip microcomputer returns a pulse inquiry response command 0x 000 x06 XX (pulse intensity data high byte) XX (pulse intensity data low byte) XX (pulse period T data high byte) XX (pulse period T data low byte) XX ("01" indicates re = True, "00" indicates re = False). The second way is to actively send the instruction "0 x 000 x060x 700 x 650 x 610 x680xA 4" to the computer when re =1. And when the single chip receives the pulse inquiry command, stopping actively sending the command.

The GPIO unit is also a self-contained module in the single chip microcomputer. And selecting the BNC as the GPIO unit output interface. And outputting a low level by default, wherein the output mode of the GPIO unit is push-pull output. The output high level is 3.3V, and the output speed of the GPIO unit is 20 MHz. The operating mode of the GPIO unit is to output a 1ms high level signal, namely a pulse peak prediction signal when re = True.

As shown in fig. 7, the pulse signals obtained by the system are real pulse signals, and the pulse peak prediction signals exist stably before each pulse peak appears, because the human pulse signals and the pulse peak prediction signals obtained from the single chip microcomputer are respectively connected with the USB _ MiniB and the BNC interface by using a computer and an oscilloscope to observe product output. And an oscillograph can be used for obtaining a pulse peak prediction signal oscillogram from the single chip microcomputer through a BNC interface. The form of the signal is a square wave signal with the width of 1ms, the signal has no stable period, the high level is 3.3V, and the low level is 0V.

Each device of the system can continuously run, the calculation time is not more than 2ms, when the device runs stably, the pulse peak capturing rate is more than 80%, and the misjudgment rate is less than 3%.

The invention uses the pulse sensor to collect the pulse signal of the human body, and uses a specific algorithm to process, thereby predicting the occurrence of the pulse peak in about 30ms before the pulse peak occurs and giving the cycle time of the previous cycle of the pulse. By applying the invention, other equipment can be conveniently driven to acquire spectral information or other information of the upcoming pulse peak, and the pulse peak ending time is estimated according to the period information so as to acquire the information of the pulse valley. The method and the device have the advantages of high stability, high universality and high accuracy, the collection can be carried out for a long time, and on the basis, the device can be applied to measuring the spectral data of the peak-valley value of the pulse for imaging and qualitative and quantitative analysis.

While the foregoing is directed to the preferred embodiment of the present invention, it will be understood by those skilled in the art that various changes and modifications may be made without departing from the spirit and scope of the invention as defined in the appended claims.

Claims

1. A method for predicting a human pulse peak is characterized by comprising the following steps:

acquiring a human body pulse signal;

and predicting the pulse peak of the human body to obtain a prediction result.

2. The method for predicting human pulse peak according to claim 1, wherein obtaining the human pulse signal comprises:

collecting human body pulse analog voltage signals;

3. The method for predicting human pulse peak according to claim 2, wherein the step of subjecting the human pulse signal to positive value filtering comprises:

and setting all negative derivatives in the human pulse signals to be zero.

4. The method of claim 3, wherein the step of continuously filtering the pulse signal comprises:

5. The method for predicting human pulse peak according to claim 4, wherein the step of filtering the human pulse signal by the mean threshold comprises:

6. The method of predicting human pulse peak according to claim 5, wherein the prediction result is outputted to a terminal.

7. The method of predicting human pulse peak according to claim 6, further comprising: the terminal comprises but is not limited to an oscilloscope, a computer, a triggerable human body infrared imaging system and a triggerable human body spectrum acquisition system.

8. A human pulse peak prediction device, comprising:

the acquisition module is used for acquiring a human body pulse signal;

9. The human pulse peak prediction device of claim 8, wherein the obtaining module is specifically configured to:

collecting human body pulse analog voltage signals;

10. The human pulse peak prediction device of claim 9, wherein the processing module is specifically configured to:

and setting all negative derivatives in the human pulse signals to be zero.