WO2018152675A1 - Early screening system and method for peripheral arteriosclerosis - Google Patents

Abstract

An early screening system and method for peripheral arteriosclerosis. The method comprises the following steps: S1: acquire pulse wave signals from different locations on the body of a patient; S2: perform data pre-processing on a pulse wave signal; S3: perform real-time frequency-domain processing on the pulse wave signal which has undergone data pre-processing, and obtain a continuous three-dimensional waterfall spectrogram and/or a three-dimensional contour grayscale map, and determine whether arteriosclerosis is present in the patient by means of the three-dimensional waterfall spectrogram and/or the three-dimensional contour grayscale map. The pulse wave signal which has undergone data pre-processing is processed by using a cross-correlation processing method to obtain the location of a thrombus in a limb artery of the patient. The early screening method and system for peripheral arteriosclerosis may be used by clinicians to evaluate arterial lesions in the human body, and may be especially used by clinicians to provide a reliable basis for evaluating whether early lesions of arteriosclerosis are present in the human body and whether a thrombus is present in an artery.

Description

Early screening system and method for peripheral arteriosclerosis Technical field

The invention relates to the technical field of medical instruments, in particular to an early screening system and method for peripheral arteriosclerosis.

Background technique

Arteriosclerosis is a primary disease that seriously affects human health. It is an effective, simple, low-risk, low-cost measure to improve human quality of life through early accurate, non-invasive screening and prevention. At present, the pulse wave signal is processed to obtain an index of arteriosclerosis. However, in the field of pulse wave signal processing, both existing products and related theories are based on time domain processing technology. Although the time domain treatment effectively solves the cases in the early screening of some peripheral arteriosclerosis in clinical applications, its clinical use is still limited and contingent. For example, the location of a thrombus cannot be accurately performed, the development trend of the disease cannot be accurately evaluated, and the early stage of thrombosis and the early degeneration of the arterial smooth muscle are difficult to find. Therefore, there is a need for an early screening system and method for arteriosclerosis that will provide clinicians with a reliable basis for disease assessment.

Summary of the invention

It is an object of the present invention to provide an early screening method and system for peripheral arteriosclerosis in view of current deficiencies.

The present invention provides an early screening method for peripheral arteriosclerosis, the method comprising the steps of:

S1: acquiring a pulse wave signal at different positions of the patient's body;

S2: performing data preprocessing on the pulse wave signal, where the data preprocessing comprises performing data preprocessing on the pulse wave signal by using FIR filtering;

S3: performing real-time frequency domain processing on the pulse wave signal preprocessed by the data, and obtaining a continuous three-dimensional waterfall spectrogram and/or a three-dimensional contour grayscale image, by using the three-dimensional waterfall spectrogram and/or three-dimensional The contour grayscale map determines whether the patient has arteriosclerosis.

Preferably, in the early screening method of the peripheral arteriosclerosis, the method further comprises: S4: processing the pulse wave signal preprocessed by the data to obtain a thrombus in the artery of the patient by using a cross-correlation method position.

The present invention also provides an early screening system for peripheral arteriosclerosis, the system comprising: an infrared photoelectric data sensor for acquiring a pulse wave signal; a data preprocessing module for preprocessing the pulse wave signal, the data The pre-processing module includes an FIR filtering module, configured to perform real-time frequency domain processing on the pulse wave signal preprocessed by the data, and obtain an amplitude spectrum for generating a three-dimensional waterfall spectrogram and a three-dimensional contour grayscale image. An FFT processing module of the power spectrum; optionally including a cross-correlation processing module for processing the pulse wave signal preprocessed by the data and acquiring a thrombus position in the artery of the patient by using a cross-correlation processing method; a main control unit for control and data calculation; and for displaying the three-dimensional waterfall spectrogram, the three-dimensional contour grayscale map, and the position of the thrombus a display module, wherein the infrared photoelectric data sensor is electrically connected to the data pre-processing module, and the data pre-processing module is electrically connected to an FFT processing module and a cross-correlation processing module, respectively, and the main control unit and the infrared photoelectric data respectively The sensor, the data pre-processing module, the FFT processing module, and the cross-correlation processing module are electrically connected.

Advantageous Effects of the Invention: The early screening method and system for peripheral arteriosclerosis provided by the present invention can be used by clinicians to evaluate human arterial lesions, especially for clinicians to have arteriosclerosis in the human body, and further The presence or absence of thrombus in the arteries provides a reliable basis for assessment.

DRAWINGS

The present invention will be further described below in conjunction with the accompanying drawings and embodiments, in which:

1 is a flow chart of an early screening method for peripheral arteriosclerosis according to a preferred embodiment of the present invention;

2 is a waveform diagram of a pulse wave signal collected by an early screening method for peripheral arteriosclerosis of the present invention;

Figure 3 is a waveform diagram of another pulse wave signal collected by the early screening method of peripheral arteriosclerosis of the present invention;

4a, 4b are three-dimensional waterfall spectrograms of Example 1 obtained by testing a normal person by the early screening method of peripheral arteriosclerosis of the present invention;

5a and 5b are three-dimensional waterfall spectrograms of Example 2 obtained by testing an arteriosclerotic patient by an early screening method for peripheral arteriosclerosis of the present invention;

Figure 6 is an arteriosclerosis by an early screening method for peripheral arteriosclerosis of the present invention. The three-dimensional contour line grayscale image of Example 3 obtained by the patient;

Figure 7 is a cross-correlation curve showing the presence of a thrombus in an artery obtained by testing an arteriosclerotic patient by an early screening method for peripheral arteriosclerosis of the present invention;

Figure 8 is a cross-correlation curve when there is no thrombus in an artery obtained by testing a normal person by the early screening method of peripheral arteriosclerosis of the present invention;

Figure 9 is a functional block diagram of an early arteriosclerosis screening system in accordance with a preferred embodiment of the present invention.

detailed description

The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It is understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

Referring to Figure 1, a preferred embodiment of the present invention provides an early screening method for peripheral arteriosclerosis comprising the following steps:

S1: acquiring a pulse wave signal at different positions of the patient's body; in the present invention, preferably, the pulse wave signal is collected on the limbs of the human body by infrared photoelectric sensing technology. Specifically, a plurality of infrared photoelectric data sensors are respectively fixed on the skin of the limbs of the human body, and infrared light is sent to the skin surface of the human body corresponding to the artery as the probe light through the infrared photoelectric data sensor, and the reflected light of the probe light is received, and the reflected light is transmitted through the reflected light. The pulse wave signal is acquired to collect pulse signals of the arterial pulse around the human body. The infrared photoelectric sensing technology may employ infrared photoelectric sensing technology known to those skilled in the art, and other prior art techniques capable of acquiring pulse wave signals may also be used in the present invention.

S2: Perform data preprocessing on the acquired pulse wave signal, where the data preprocessing includes performing noise reduction and filtering on the acquired pulse wave signal. Specifically, the obtained pulse wave signal is subjected to noise reduction processing by using a filtering method to obtain a signal with high signal to noise ratio and then digitally filtered, thereby retaining a pulse wave signal having a high signal to noise ratio in the frequency band for subsequent processing; The data pre-processing can be performed using correlation filtering known to those skilled in the art, for example, using a filter. Of course, other prior art techniques capable of optimizing the pulse wave signal can also be used in the present invention. In this step, preferably, the data pre-processing further comprises: performing time domain data pre-processing on the pulse wave signal to obtain an associated vascular elasticity index, for example, PWV (pulse wave wave velocity), etc., and the time domain data preprocessing may be Relevant time domain pretreatment methods known to those skilled in the art are used to obtain PWV and other related vascular elasticity indices for subsequent processing.

In this step, preferably, the minimum mean square error linear FIR filtering is performed on the pulse wave signal before the frequency domain processing, and the FIR adopts low-pass filtering, the pulse wave signal sampling frequency is 1 to 3 KHz, and the filter cutoff frequency is 0.5 to 1.5 KHz; In order to obtain a better pulse wave signal, that is, to clean the noise, to avoid affecting subsequent data processing.

More preferably, for the pulse wave signal frequency characteristic, the FIR adopts low-pass filtering, the pulse wave signal sampling frequency is 2 KHz, and the filter cutoff frequency is 1 KHz. There is only one sensor when the actual pulse wave signal is acquired, so the delayed digital pulse wave signal is used as a time sample value for linear prediction of the expected value. The sample time sample value for the pulse wave signal. N is the sequence length of the digital fetal pulse wave signal. Considering the real-time performance of the system and the convenience of calculation, M=5, M is called the linear mean square error estimator order.

y(n)=x(n) 0≤n≤N (a)

x _K (n)=x(n+1-k) 1≤k≤M (b)

Since linear sampling is expected to be the time sample value of the same signal, the equation (a) is the sampling time series of the pulse wave signal; and the equation (b) is the delayed sampling time series.

In the present invention, the linear mean square error estimator, that is, using the formula (c), is a method which is operability and easy to calculate mathematically.

By determining the coefficient c _k (n), the equation (e) is minimized, and ε(n) in the equation (d) is an error signal.

ε(n)=y(n)-Y(n) (d)

P(n)=E{|ε(n)| ² } (e)

The value of c _k (n) can be calculated by the positive definite matrix theory.

Apply the above results to the optimal FIR (finite impulse response filter) design, reference pattern (f), (g).

The output form of the FIR filter is determined by the following equation (h).

The coefficient h _k (n) represents the impulse response of the FIR filter. It is solved by a second-order rectangular equation system to obtain the best coefficient c _k (n). The specific process is to solve the linear equations; after solving, f(n) is the time domain complex data.

This is an adaptive filtering. The adaptive filter is different from the traditional filter in that it needs to know the spectral distribution of the signal, but only needs to know the sample value of the sampled signal. This adaptive filtering is based on the sample values of the sampled signal and is based on the optimization of the statistical properties of the sample values. Using the above FIR filtered pulse wave signal data, the frequency characteristics of the pulse wave signal are preserved, and the random interference generated by the system is purified; the optimal signal under the minimum mean square error is provided for the frequency domain data processing. .

S3: Perform real-time frequency domain processing on the pulse wave signal preprocessed by the data to obtain a continuous three-dimensional waterfall spectrogram and/or a three-dimensional contour grayscale image, and pass the three-dimensional waterfall spectrogram and/or three-dimensional, etc. High-line grayscale map to determine whether the patient has arteriosclerosis.

Specifically, in this step S3, the FFT of the real-time time extraction method is performed on the pulse wave signal preprocessed by the data, and the FFT (Fast Fourier Transform) complex operation process is performed according to the formula (1):

Where x(n) is time domain complex data, and x(n) is preferably f(n) output through the FIR filter. The real and imaginary parts are respectively composed of two pulse signals (time domain signals) collected, that is, X(n)=S ₁ (n)+jS ₂ (n); as shown in FIG. 2 and FIG. Waveform of the two pulse wave signals. e is an exponential function, n=0, 1, 2..., 511; m=0, 1, 2..., N-1; N can be adjusted according to the processing capability of the FFT processing module, for example, the value of N can be 512. 1024, 2048, etc. In this embodiment, the value of N is 512.

Through the above FFT (Fast Fourier Transform) complex operation process, X(n) is obtained, and X(n) is a two-way complex function, that is, X ₁ (n)=R ₁ (n)+jI ₁ (n), X ₂ (n)=R ₂ (n)+jI ₂ (n), where R ₁ (n) and I ₁ (n) are real and imaginary parts corresponding to S ₁ (n), R ₂ (n), I ₂ (n) is the real part and the imaginary part corresponding to S ₂ (n). Since the result of the FFT operation is mirror-symmetric, the processing result of the 256-point frequency domain complex of the two-way pulse wave signal is obtained, that is, R ₁ (n), I ₁ (n), and R ₂ (n), I ₂ (n) Where n = 0, 1, 2, ..., 255.

The processing results R ₁ (n) and I ₁ (n) of the frequency domain complex of one of the pulse wave signals are calculated according to formulas (2) and (3) to obtain corresponding amplitude spectrum and power spectrum:

P(n)=R ² (n)+I ² (n) (2)

Then, the processing results R ₂ (n) and I ₂ (n) of the frequency domain complex of the other pulse wave signal are subjected to the same operation according to the equations (2) and (3) to obtain a corresponding amplitude spectrum and power spectrum.

In actual use, P(n) or A(n) can be logarithmically processed according to the requirements of the three-dimensional waterfall spectrogram and the three-dimensional contour grayscale image. Preferably, in order to reduce the frequency leakage effect caused by the truncation of the time domain data during the FFT processing, the pulse wave signal (time domain signal) can be windowed according to the actual situation, and the method of windowing processing is commonly used by those skilled in the art. Window processing methods, such as: triangular window, Hanning window, Hamming window, Gaussian window, and the like.

In this step S3, further, according to the calculated amplitude spectrum and power spectrum Line 3D dynamic processing.

First, the three-dimensional waterfall spectrum map, carried out

with

Calculate and plot the calculation results into a three-dimensional waterfall spectrum. The three-dimensional waterfall spectrum map is also called the spectrum array diagram. It is mainly applied to the three-dimensional spectrogram in which the signal power spectrum or the amplitude spectrum of the engineering vibration field is superposed with the vibration change, and the harmonic components in the vibration signal are changed with the vibration. Case. This technique is used in early screening for arteriosclerosis because it truly reflects the vibration of the artery as it pulsates. The test proves that this vibration is closely related to the degree of hardening of the arterial wall. The application of the three-dimensional waterfall spectrogram overcomes the follow-up caused by the human body's mental factors, resulting in instability and uncertainty in the time domain analysis; and, through the three-dimensional display, better depicts the arterial pulsation. With complete information on changes in blood pressure (systolic and diastolic), this is information that is difficult to detect in the time domain. It is the frequency distribution (i = 0, 1, 2., 255, j = 0, 1, 2., 255) of the harmonic components of the pulse wave signal as the heartbeat is pulsating (F _ij) Equal to A(n) corresponding to a value on the three-dimensional waterfall spectrogram), can clearly show the obvious difference between normal and abnormal heart beats, the conduction velocity between normal and abnormal pulse waves, blood in the artery The frequency distribution characteristics of reflux and blood flow through the thrombus provide a reliable basis for clinicians to evaluate early peripheral arteriosclerosis.

The following two comparisons of the completed Examples 1 and 2 verify that the three-dimensional waterfall spectrogram can provide a basis for evaluating early peripheral arteriosclerosis. The experimental procedures of Examples 1 and 2 are as follows:

Experiments were performed on clinically confirmed normal and arteriosclerotic patients. The subject took a flat posture and stood for 10 minutes to start the test. The infrared photoelectric data sensor is placed inside the ankle of the subject, and data acquisition is performed at intervals of 6 ms to obtain a time domain signal sequence x(n) of the pulse wave, and time domain data preprocessing and real time FFT complex operation are performed thereon. deal with. In formula (1), the length of the point is N=512, and this experiment is not windowed. For the FFT processing result X ₁ (n)=R ₁ (n)+jI ₁ (n), X ₂ (n)=R ₂ (n)+jI ₂ (n), (n=0,1,2...255 )get on

with

Calculate and plot the results of the 3D waterfall spectrum. 4 and 5 are three-dimensional waterfall spectrograms of Example 1 and Example 2 obtained by testing normal humans and arteriosclerotic patients; and comparing the two three-dimensional waterfall spectrograms of Figures 4 and 5, from Figure 5 It can be clearly seen that when arteriosclerosis occurs (Fig. 5a, 5b), the higher harmonic components (yellow) are relatively abundant due to the decrease in the elasticity of the blood vessel wall and the vibration of the blood vessel wall caused by the beat of the heart. Under normal conditions (Fig. 4a, 4b), there is no high frequency harmonic component (blue) except the fundamental frequency. In the present invention, it is to be noted that Figures 4a, 5a are provided in color for ease of reference by the examiner in the examination, and Figures 4b, 5b of the black and white in accordance with the patent law are additionally provided.

It can be seen from the above that since the mechanism of arteriosclerosis is due to the decrease in elasticity of the arterial smooth muscle wall and the increase in rigidity, the frequency of the vibration of the blood vessel wall due to the beat of the heart is relatively high. Because it is a three-dimensional waterfall spectrogram, it can observe the small dynamic changes of the arterial blood vessels during pulse conduction in a continuous and large amount of data, and the small vibration changes of the arterial vessel wall which cannot be obtained in the early time hardening. Provides a reliable basis for identifying certain uncertainties in early arteriosclerosis screening.

Second, the three-dimensional contour line gray map: P1 (n) = R1 ² (n) + I1 ² (n) and P2 (n) = R2 ² (n) + I2 ² (n) calculation, the calculation results are drawn A three-dimensional contour line grayscale image.

The three-dimensional contour line gray map is a contour line dynamic spectrum distribution 16-level gray scale three-dimensional display, it The information of the arterial pulsation as it changes with blood pressure (systolic and diastolic) is shown from another angle. The time-frequency dynamic distribution of the arterial pulsation can be visually displayed as the gradation changes, which is based on the nonlinear time-frequency representation of the energy distribution Eij (i=0, 1, 2....255, j=0, 1, 2) ....255) (Eij is equal to P(n) corresponding to a value on the contour map), and the method of time-frequency texture analysis of the pulse wave signal can clearly show the energy distribution on the time-frequency plane. Characteristics, from a new perspective, better reflect some of the essential features of human arterial pulsation, to assess the extent and development of arteriosclerosis. This is another main inventive aspect of the present invention.

Experiments were performed on patients with clinically confirmed mild arteriosclerosis. The subject took a flat posture and stood for 10 minutes to start the test. The infrared photoelectric data sensor is placed inside the ankle of the subject, and data acquisition is performed at intervals of 6 ms to obtain a time domain signal sequence x(n) of the pulse wave, and time domain data preprocessing and real time FFT complex operation are performed thereon. deal with. In formula (1), the length of the point is N=512, and this experiment is not windowed. For the FFT processing result X ₁ (n)=R ₁ (n)+jI ₁ (n), X ₂ (n)=R ₂ (n)+jI ₂ (n), (n=0,1,2...255 )get on

with

Calculate and plot the 3D contour grayscale image. 6 is a three-dimensional contour grayscale image of Example 3 obtained by testing a mild arteriosclerosis patient; as can be clearly seen from FIG. 6, the energy of the three-dimensional contour grayscale image of Embodiment 3 is at HZ50. There is a strip distribution between ~60 (the greater the energy, the darker the black). Under normal circumstances (testing for normal people), the energy distribution of the three-dimensional contour grayscale image is mainly concentrated in the lower part of the three-dimensional contour grayscale image.

The three-dimensional contour grayscale map reflects the energy characteristics of arteriosclerosis under time-frequency distribution. Because during the formation of arteriosclerosis, the elasticity of the arterial wall gradually decreases, along with the heart In addition to the increase in high-frequency components, the vibration of the tube wall during the pulsation also increases the energy generated by the tube wall during the vibration. When the arterial wall is elastic, it has an absorption effect on the energy generated by the vibration. Through the three-dimensional contour grayscale map, it is very clear that different energy distributions occur during arteriosclerosis, and it can be concluded whether the subject has arteriosclerosis, and the pathological characteristics of arteriosclerosis can be accurately determined according to different energy distribution characteristics. For example, for the degenerative changes of the smooth muscle of the wall, the identification of atherosclerosis, etc., to determine the patient's future treatment plan.

S4: The pulse wave signal preprocessed by the data is processed by a cross-correlation processing method to obtain a position of a thrombus in the artery of the patient.

Specifically, baseline processing is performed on the pulse wave signal preprocessed by the data: after extracting the pulse wave morphological feature, performing spline interpolation to fit the baseline to obtain a digital sequence function b of the fitted baseline (n), then perform x(n)=x(n)-b(n), y(n)=y(n)-b(n) operations on the sequence function to suppress baseline drift; (4) calculates the pulse wave signal and outputs the cross-correlation graph (Figure 8), to determine when the R _xy (n) corresponding to n at a maximum value _T n, then the resulting n _t in equation (5), determine the position of the artery thrombosis:

Wherein, in formula (4), x(n), y(n) are pulse wave signals collected at different positions of the limb of the patient, for example: pulse wave signals at the left arm and the right arm, R _xy (n) ) is a cross-correlation function that represents the degree of correlation between two different time series at a certain time. N can be any positive integer, where 0 ≤ n ≤ N-1, 0 ≤ m ≤ N-1 . In this embodiment, the value of N is 512.

Wherein, in formula (5), v is PWV, K is the correction coefficient of PWV, K is in the range of 1 to 10, and d is the center line of the two infrared photoelectric data sensors at the position of the blood flow to the thrombus The distance from which the position of the thrombus can be determined.

When there is no thrombus in the artery of the subject, more specifically, there is no thrombus formation in the artery between the two infrared photoelectric data sensors, the pulse wave signal is calculated by using formula (4) and the cross-correlation curve is output as shown in the figure. As shown in Figure 8, the cross-correlation curve has no distinct peak characteristics.

It can be seen from the above that the pulse wave signal preprocessed by the data is processed by the method of cross-correlation processing, and it is possible to test whether there is a thrombus in the artery of the patient, and further obtain the position of the thrombus in the artery of the patient, and can accurately perform the thrombus The positioning provides a reliable basis for clinicians to locate the location of the thrombus.

As shown in Fig. 9, the present invention also provides an early screening system for peripheral arteriosclerosis, comprising an infrared photoelectric data sensor 11 for acquiring a pulse wave signal; and a data preprocessing module 12 for preprocessing the pulse wave signal. An FFT processing module 13 for performing real-time frequency domain processing on the pulse wave signal and obtaining a continuous three-dimensional waterfall spectrogram and a three-dimensional contour grayscale image; preferably including pre-correlation processing The processed pulse wave signal is processed, and a cross-correlation processing module 14 for acquiring the position of the thrombus in the artery of the patient; a main control unit 15 for control and data management of the entire system; and a display of the three-dimensional waterfall spectrogram, three-dimensional A display module 16 of the contour grayscale map and the location of the thrombus. The infrared photoelectric data sensor 11 is electrically connected to the data preprocessing module 12, and the data preprocessing module 12 and the FFT processing module 13 are respectively The cross-correlation processing module 14 is electrically connected, and the main control unit 15 is electrically connected to the infrared photoelectric data sensor 11, the data pre-processing module 12, the FFT processing module 13, the cross-correlation processing module 14, and the display module 16, respectively.

1. The infrared photoelectric data sensor 11 is configured to acquire pulse wave signals at different positions of the patient's body. Specifically, a plurality of infrared photoelectric data sensors 11 are respectively fixed on the skin surface of the human limbs, and infrared light is sent to the skin surface of the human body corresponding to the artery as the probe light through the infrared photoelectric data sensor 11, and the reflected light of the probe light is received. The reflected light acquires a pulse wave signal to collect an arterial pulse wave signal around the human body. The infrared photoelectric sensor may be an infrared photoelectric sensor known to those skilled in the art, and may be purchased or developed by itself.

The data preprocessing module 12 is configured to preprocess the pulse wave signal, and the infrared photoelectric data sensor 11 is electrically connected to the data preprocessing module 12 for using the pulse wave acquired by the infrared photoelectric data sensor 11. The signal is pre-processed by the data pre-processing module 12. Specifically, the data pre-processing module 12 includes a filter for performing noise reduction and filtering processing on the acquired pulse wave signal, and a time domain data pre-processing of the acquired pulse wave signal to obtain an associated blood vessel elasticity index ( For example: PWV, pulse wave velocity) time domain data preprocessing module. The obtained pulse wave signal is denoised by a filter to obtain a signal with high signal to noise ratio and then digitally filtered, thereby preserving the pulse wave signal with high signal to noise ratio in the frequency band for subsequent processing. The time domain data preprocessing module performs time domain data preprocessing on the pulse wave signal to obtain an associated vascular elasticity index, such as PWV (pulse wave wave velocity), etc., and the time domain data preprocessing can be known by those skilled in the art. Correlated time domain preprocessing method to obtain PWV and other related vascular elasticity fingers Number for subsequent processing. Preferably, the FIR filtering is used to preserve the frequency characteristics of the pulse wave signal, and the random interference generated by the system is purified; and the optimal signal under the minimum mean square error is provided for the frequency domain data processing.

The FFT processing module 13 is configured to perform real-time frequency domain processing on the pulse wave signal, and obtain an amplitude spectrum and a power spectrum for generating a three-dimensional waterfall spectrogram and a three-dimensional contour grayscale image. The data pre-processing module 12 is electrically connected to the FFT processing module 13 for sending the pulse wave signal processed by the data pre-processing module 12 to the FFT processing module 13, and the FFT processing module 13 and the main control unit 15 The electrical connection is used to send the amplitude spectrum and the power spectrum obtained by the FFT processing module 13 to the main control unit 15 to obtain a three-dimensional waterfall spectrogram, a three-dimensional contour grayscale image, and is electrically connected to the main control unit 15. Display module 16 is displayed.

Specifically, the FFT processing module 13 performs a real-time time extraction method FFT on the pulse wave signal preprocessed by the data, and performs an FFT (Fast Fourier Transform) complex operation process according to the formula (1):

Where x(n) is the time domain complex data, and the real part and the imaginary part are respectively composed of two acquired pulse wave signals (time domain signals), that is, X(n)=S ₁ (n)+jS ₂ (n) 2 and FIG. 3 are waveform diagrams of the acquired two-way pulse wave signals. e is an exponential function, n=0, 1, 2..., 511; m=0, 1, 2..., N-1; N can be adjusted according to the processing capability of the FFT processing module, for example, the value of N can be 512. 1024, 2048, etc. In this embodiment, the value of N is 512.

Through the above FFT (Fast Fourier Transform) complex operation processing, X(n), X(n) are obtained. Is a two-way complex function, that is, X1(n)=R1(n)+jI1(n), X2(n)=R2(n)+jI2(n), where R1(n), I1(n) is S1 (n) Corresponding real and imaginary parts, R2(n) and I2(n) are real and imaginary parts corresponding to S2(n). Since the result of the FFT operation is mirror-symmetric, the processing results of the 256-point frequency domain complex of the two-way pulse wave signals are obtained, that is, R1(n), I1(n), and R2(n), I2(n), where n =0, 1, 2..., 255.

The processing results R ₁ (n) and I ₁ (n) of the frequency domain complex of one of the pulse wave signals are calculated according to formulas (2) and (3) to obtain corresponding amplitude spectrum and power spectrum:

P(n)=R ² (n)+I ² (n) (2)

Then, the processing results R ₂ (n) and I ₂ (n) of the frequency domain complex of the other pulse wave signal are subjected to the same operation according to the equations (2) and (3) to obtain a corresponding amplitude spectrum and power spectrum.

4. The main control unit 15 is electrically connected to the infrared photoelectric data sensor 11, the data preprocessing module 12, the FFT processing module 13, the cross correlation processing module 14, and the display module 16, respectively, for controlling the functions of these modules, and will pass the FFT. The amplitude spectrum and the power spectrum obtained by the processing module 13 are drawn into a three-dimensional waterfall spectrum map and a three-dimensional contour gray scale image. The main control unit 15 is used to implement control and data calculation of the entire system, and has functions such as a three-dimensional waterfall spectrogram and a three-dimensional contour gray scale drawing, and the device having the above functions can be used in the present invention, for example, it can be Digital signal processors, computer mainframes, and the like, which are commonly used in the field, and software programs, data operation programs, and drawing programs installed thereon can be implemented by those skilled in the art according to the prior art.

1, three-dimensional waterfall spectrum map, carried out

with

Calculate and plot the calculation results into a three-dimensional waterfall spectrum. The three-dimensional waterfall spectrum map is also called the spectrum array diagram. It is mainly applied to the three-dimensional spectrogram in which the signal power spectrum or the amplitude spectrum of the engineering vibration field is superposed with the vibration change, and the harmonic components in the vibration signal are changed with the vibration. Case. This technique is used in early screening for arteriosclerosis because it truly reflects the vibration of the artery as it pulsates. The test proves that this vibration is closely related to the degree of hardening of the arterial wall. The application of the three-dimensional waterfall spectrogram overcomes the follow-up caused by the mental factors of the human body, resulting in instability and uncertainty in the time domain analysis; it displays the information that is difficult to detect in the time domain through three-dimensional display. Better information on the arterial pulsation as it changes with blood pressure (systolic and diastolic). It is the frequency distribution (i = 0, 1, 2., 255, j = 0, 1, 2., 255) of the harmonic components of the pulse wave signal as the heartbeat is pulsating (F _ij) Equal to A(n) corresponding to a value on the three-dimensional waterfall spectrogram), can clearly show the obvious difference between normal and abnormal heart beats, the conduction velocity between normal and abnormal pulse waves, blood in the artery The frequency distribution characteristics of reflux and blood flow through the thrombus, this method provides a reliable basis for clinicians to evaluate early peripheral arteriosclerosis.

2. Three-dimensional contour grayscale map: P1(n)=R1 ² (n)+I1 ² (n) and P2(n)=R2 ² (n)+I2 ² (n) are calculated by the main control unit 15 , the calculation result is drawn into a three-dimensional contour grayscale image. The three-dimensional contour grayscale map is a 16-level grayscale three-dimensional display of the contour line dynamic spectrum distribution, which shows the information of the arterial pulsation as it changes with blood pressure (systolic and diastolic blood pressure) from another angle. The time-frequency dynamic distribution of the arterial pulsation can be visually displayed as the gradation changes, which is based on the nonlinear time-frequency representation of the energy distribution Eij (i=0, 1, 2....255, j=0, 1, 2) ....255) (Eij is equal to P(n) corresponding to a value on the contour map), and the method of time-frequency texture analysis of the pulse wave signal can clearly show the energy distribution on the time-frequency plane. Characteristics, from a new perspective, better reflect some of the essential features of human arterial pulsation, to assess the extent and development of arteriosclerosis.

5. In the present invention, preferably, a cross-correlation processing module 14 is further included, and the patient's artery can be further located by the cross-correlation processing module 14 on the basis of testing the patient's presence of arteriosclerosis by the FFT processing module 14 and the main control unit 15. Whether there is a thrombus or a location of a thrombus. Thus, the cross-correlation processing module 14 is configured to process the pulse wave signal preprocessed by the data and acquire the position of the thrombus in the artery of the patient. The data pre-processing module 12 is electrically connected to the cross-correlation processing module 14 for sending the pulse wave signal processed by the data pre-processing module 12 to the cross-correlation processing module 14, the cross-correlation processing module 14 and the main The control unit 15 is electrically coupled for controlling the operation of the cross-correlation processing module 14 by the main control unit 15 and displaying the position of the thrombus in the patient's artery through the display module 16 electrically coupled to the main control unit 15.

The cross-correlation processing module 14 processes the pulse wave signal preprocessed by the data to acquire the position of the thrombus in the artery of the patient. Specifically, baseline processing is performed on the pulse wave signal preprocessed by the data: after extracting the pulse wave morphological feature, performing spline interpolation to fit the baseline to obtain a digital sequence function b of the fitted baseline (n), then perform x(n)=x(n)-b(n), y(n)=y(n)-b(n) operations on the sequence function to suppress baseline drift; (4) Calculate the pulse wave signal and output a cross-correlation curve (as shown in Fig. 8), thereby determining that when R _xy (n) is at the maximum value, the corresponding value of n is nt, and then substituting the obtained nt Equation (5), determine the location of the thrombus in the artery:

Wherein, in formula (4), x(n), y(n) are pulse wave signals collected at different positions of the patient's body, for example, pulse wave signals at the left arm and the right arm, R _xy (n) ) is a cross-correlation function that represents the degree of correlation between two different time series at a certain time, and N can be any integer. In the present embodiment, the value of N is 512, where 0 ≤ n ≤ N-1, and 0 ≤ m ≤ N-1.

In formula (5), v is PWV, K is the correction coefficient of PWV, and K is in the range of 1 to 10, where d is the distance from the center line of the two infrared photoelectric data sensors to the direction of blood flow to the thrombus. Thereby the position of the thrombus can be determined and displayed by the display module 16.

In this embodiment, it can be understood that the display module 14 can be a display screen, such as a liquid crystal touch screen or the like, for displaying the three-dimensional waterfall spectrogram, the three-dimensional contour grayscale image, and the position of the thrombus, so that As seen by the clinician, an intuitive display of the condition of the arteriosclerosis and the location of the thrombus is provided.

In summary, the early screening system and method for peripheral arteriosclerosis according to a preferred embodiment of the present invention have at least the following beneficial effects: the FFT complex module performs FFT complex operation on the pulse wave signal to obtain a corresponding amplitude spectrum and power spectrum, and then The three-dimensional waterfall spectrogram and/or three-dimensional contour grayscale map obtained by the main control unit can provide the clinician with the possible pathological mechanism, degree and development trend of early arteriosclerosis, and can clearly reflect the blood with the change of blood pressure. Flow velocity in the arteries, reflux, and response to thrombosis provide an effective basis for physicians' clinical risk assessment and future treatment options. Further, by placing the sensors in different positions on the limbs of the human body, and then performing the cross-correlation processing module Cross-correlation processing can accurately determine the position and size of the embolus in the artery, especially in the early stage of emboli formation. When it is difficult to find through imaging medical equipment, it can provide the doctor with the basis for early effective treatment.

It is to be understood that those skilled in the art can devise modifications and variations in the above description, and all such modifications and changes are intended to be included within the scope of the appended claims.

Claims (10)

1. An early screening method for peripheral arteriosclerosis, characterized in that the method comprises the following steps:

   S1: acquiring a pulse wave signal at different positions of the patient's body;

   S2: performing data preprocessing on the pulse wave signal, where the data preprocessing comprises performing data preprocessing on the pulse wave signal by using FIR filtering;

   S3: performing real-time frequency domain processing on the pulse wave signal preprocessed by the data, and obtaining a continuous three-dimensional waterfall spectrogram and/or a three-dimensional contour grayscale image, by using the three-dimensional waterfall spectrogram and/or three-dimensional The contour grayscale map determines whether the patient has arteriosclerosis.
2. The method for early screening of peripheral arteriosclerosis according to claim 1, wherein the method further comprises: S4: processing the pulse wave signal preprocessed by the data to obtain a patient by using a cross-correlation method; The location of the thrombus in the artery.
3. The early screening method for peripheral arteriosclerosis according to claim 1, wherein in step S2, the minimum mean square error linear FIR filtering is performed on the acquired pulse wave signal, and the FIR adopts low-pass filtering and pulse wave signal. The sampling frequency is 1 to 3 KHz, and the filter cutoff frequency is 0.5 to 1.5 KHz.
4. The early screening method for peripheral arteriosclerosis according to claim 1, wherein in the step S3, the pulse wave signal is subjected to a real-time time extraction method FFT, and the FFT complex arithmetic operation is performed according to the formula (1):

   

   Where x(n) is the time domain complex data, and the real part and the imaginary part are respectively composed of two acquired pulse wave signals, that is, X(n)=S ₁ (n)+jS ₂ (n), where e is Exponential function, n = 0, 1, 2, ..., 511; m = 0, 1, 2, ..., N-1;

   Through the above FFT complex arithmetic processing, X(n) is obtained, and X(n) is a two-way complex function including X ₁ (n)=R ₁ (n)+jI ₁ (n), X ₂ (n)=R ₂ (n) + jI ₂ (n), R ₁ (n), I ₁ (n) are real and imaginary parts corresponding to S ₁ (n), and R ₂ (n) and I ₂ (n) are S ₂ ( n) corresponding real part, imaginary part, wherein n=0, 1, 2..., 255;

   The processing results R ₁ (n) and I ₁ (n) of the frequency domain complex of one of the pulse wave signals are calculated according to formulas (2) and (3) to obtain corresponding amplitude spectrum and power spectrum:

   P(n)=R ² (n)+I ² (n) (2)

   

   Then, the processing results R ₂ (n) and I ₂ (n) of the frequency domain complex of the other pulse wave signal are subjected to the same operation according to the equations (2) and (3) to obtain a corresponding amplitude spectrum and power spectrum.
5. The early screening method for peripheral arteriosclerosis according to claim 1, wherein in step S3,

   

   with

   

   

   To calculate the result of the operation into a three-dimensional waterfall spectrum; and/or to perform P1(n)=R1 ² (n)+I1 ² (n) and P2(n)=R2 ² (n)+I2 ² (n) , the calculation result is drawn into a three-dimensional contour grayscale image.
6. The early screening method for peripheral arteriosclerosis according to claim 2, wherein in step S4, the cross-correlation processing is:

   Calculate the pre-processed two-way pulse wave signal using equation (4) to determine the value of n corresponding to n _t when R _xy (n) is at the maximum value, and then substitute n _t into formula (5) to determine The location of the thrombus in the artery:

   

   

   Where x(n) and y(n) are the pulse wave signals at different positions of the collected human body, and R _xy (n) is a cross-correlation function, which means that two different time series are taken at a certain time. The degree of correlation of values; N can be any positive integer, where 0 ≤ n ≤ N-1, 0 ≤ m ≤ N-1;

   v is PWV, K is the correction coefficient of PWV, and K is in the range of 1 to 10. d is the distance from the center line of the two infrared photoelectric data sensors to the blood flow direction to the thrombus.
7. An early screening system for peripheral arteriosclerosis, characterized in that the system comprises:

   An infrared photoelectric data sensor for acquiring a pulse wave signal;

   a data preprocessing module for preprocessing a pulse wave signal, the data preprocessing module comprising an FIR filtering module;

   The method is used for performing real-time frequency domain processing on the pulse wave signal preprocessed by the data, and obtaining an FFT processing module for generating a magnitude spectrum and a power spectrum of the three-dimensional waterfall spectrogram and the three-dimensional contour grayscale image;

   Optionally, a cross-correlation processing module that processes the pulse wave signal preprocessed by the data and acquires a thrombus position in the artery of the patient by using a method of cross-correlation processing;

   Main control unit for control and data calculation of the entire system;

   And a display module for displaying the three-dimensional waterfall spectrogram, the three-dimensional contour grayscale image, and the position of the thrombus;

   The infrared photoelectric data sensor is electrically connected to the data preprocessing module, and the data preprocessing module is electrically connected to the FFT processing module and the cross correlation processing module respectively, and the main control unit and the infrared photoelectric data sensor and the data respectively The pre-processing module, the FFT processing module, and the cross-correlation processing module are electrically connected.
8. The early screening system for peripheral arteriosclerosis according to claim 7, wherein the FFT processing module performs a real-time time extraction method FFT on the pulse wave signal, and performs FFT complex operation processing according to formula (1):

   

   Where x(n) is the time domain complex data, and the real part and the imaginary part are respectively composed of two acquired pulse wave signals, that is, X(n)=S ₁ (n)+jS ₂ (n), where e is Exponential function, n = 0, 1, 2, ..., 511; m = 0, 1, 2, ..., N-1;

   Through the above FFT complex arithmetic processing, X(n) is obtained, and X(n) is a two-way complex function including X ₁ (n)=R ₁ (n)+jI ₁ (n), X ₂ (n)=R ₂ (n) + jI ₂ (n), R ₁ (n), I ₁ (n) are real and imaginary parts corresponding to S ₁ (n), and R ₂ (n) and I ₂ (n) are S ₂ ( n) corresponding real part, imaginary part, wherein n=0, 1, 2..., 255;

   The processing result R ₁ (n), I ₁ (n) of the frequency domain complex of one of the pulse wave signals is subjected to the processing result of the frequency domain complex of one of the pulse wave signals according to the formula R ₁ (n), I ₁ (n) ), according to the formula (2) and (3) operation to obtain the corresponding amplitude spectrum and power spectrum:

   P(n)=R ² (n)+I ² (n) (2)

   

   Then, the processing results R ₂ (n) and I ₂ (n) of the frequency domain complex of the other pulse wave signal are subjected to the same operation according to the equations (2) and (3) to obtain a corresponding amplitude spectrum and power spectrum.
9. An early screening system for peripheral arteriosclerosis according to claim 8, wherein said main control unit performs

   

   with

   

   

   To calculate the result of the operation into a three-dimensional waterfall spectrum; and/or to perform P1(n)=R1 ² (n)+I1 ² (n) and P2(n)=R2 ² (n)+I2 ² (n) , the calculation result is drawn into a three-dimensional contour grayscale image.
10. The early screening system for peripheral arteriosclerosis according to claim 7, wherein said cross-correlation processing module calculates the pre-processed two-way pulse wave signals using equation (4) to determine when R _xy ( n) The value of n corresponding to the maximum value is nt, and then the obtained nt is substituted into the formula (5) to determine the position of the thrombus in the artery:

    

    

    Where x(n) and y(n) are the pulse wave signals at different positions of the collected human body, and R _xy (n) is a cross-correlation function, which means that two different time series are taken at a certain time. The degree of correlation of values; N can be any positive integer, where 0 ≤ n ≤ N-1, 0 ≤ m ≤ N-1;

    v is PWV, K is the correction coefficient of PWV, K is in the range of 1 to 10, and d is the distance from the center line of the two infrared photoelectric data sensors to the blood flow direction to the thrombus;

    Moreover, the cross-correlation processing module further includes performing baseline processing on the data pre-processed pulse wave signal: extracting the pulse wave morphological feature, and fitting the baseline by using spline interpolation method to obtain a number fitting the baseline The sequence function b(n) performs x(n)=x(n)-b(n), y(n)=y(n)-b(n) operations on the digital sequence function b(n), shifting the baseline The suppression is performed, and the pulse wave signal is calculated by using the formula (4).

Images