Heart mechanical motion signal extraction method and device based on multi-domain joint modulation

Through multi-domain joint modulation and signal reconstruction technology, the problem of cardiac mechanical motion signals being annihilated in the prior art is solved, and the accurate extraction and cleaning of cardiac mechanical motion signals is achieved, and the electrocardiogram waveform acquisition and health monitoring is supported.

CN118924276APending Publication Date: 2024-11-12XI AN JIAOTONG UNIV

Patent Information

Application Number
CN202411024476.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-29
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

The prior art is difficult to effectively extract cardiac mechanical motion signals because weak signals are annihilated by breathing and limb movements, and the adaptive filtering method is sensitive to noise, resulting in signal distortion.

Method used

Using multi-domain joint modulation method, data acquisition is collected by millimeter wave radar, spatial reconstruction and decoupling are performed, and the cardiac beating signals, chest ups and downs and interference signals caused by breathing are separated, using Hanning Window and Fourier. Transform noise reduction and further extract cardiac mechanical motion signals through differential and filters.

Benefits of technology

The weak signals of cardiac mechanical movement were successfully extracted, which improved the accuracy and cleanliness of the signal, and provided technical support for contactless physiological signal acquisition for electrocardiogram waveform acquisition and heart health monitoring.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118924276A_ABST
    Figure CN118924276A_ABST
Patent Text Reader

Abstract

The invention discloses a heart mechanical motion signal extraction method and device based on multi-domain joint modulation, and the method comprises the steps: carrying out the data collection of a chest cavity of a to-be-detected person through a millimeter-wave radar, and obtaining a chest wall vibration signal from a radar echo signal; performing data reconstruction on the chest wall vibration signal, and separating data of each channel to obtain a signal generated by heart beating, a signal generated by chest fluctuation caused by respiration and an interference signal; and according to a data reconstruction result, decoupling the chest wall vibration signal, filtering out an interference signal, and separating and extracting a signal generated by cardiac beating from the signal generated by cardiac beating and a signal generated by thoracic cavity fluctuation caused by respiration. According to the method based on multi-domain joint modulation, weak signals of heart mechanical motion can be successfully extracted, and technical support is provided for applications such as electrocardiogram waveform acquisition, heart health real-time monitoring and alarming, motion health monitoring and other non-contact physiological signal acquisition.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of radar signal processing, and in particular to a method and device for extracting cardiac mechanical motion signals based on multi-domain joint modulation. Background Art

[0002] The weak mechanical motion displacement caused by heart beating ranges from 0.2 to 0.5 mm. Although this magnitude of weak motion can be sensed by millimeter-wave radar, the weak cardiac mechanical motion signal is often obliterated by factors such as chest fluctuations caused by breathing and the subject's limb movements. This makes it extremely difficult to directly extract the cardiac mechanical motion signal from the millimeter-wave radar signal.

[0003] In addition, since electric pulses cannot directly modulate millimeter-wave radar waves, existing radar signal processing methods based on adaptive filtering, wavelet transform and Kalman filtering are sensitive to noise and interference signals, which makes the extracted cardiac mechanical motion signals easily distorted. Summary of the invention

[0004] The purpose of the present invention is to address the problems in the above-mentioned prior art and provide a method and device for extracting cardiac mechanical motion signals based on multi-domain joint modulation. After using a millimeter-wave radar to collect data on the chest of a subject, the radar echo signal is spatially reconstructed and decoupled to accurately obtain the weak signal of cardiac mechanical motion.

[0005] In order to achieve the above object, the present invention has the following technical solutions:

[0006] In a first aspect, a method for extracting cardiac mechanical motion signals based on multi-domain joint modulation is provided, comprising:

[0007] Use millimeter wave radar to collect data from the chest cavity of the subject, and obtain the chest wall vibration signal S from the radar echo signal;

[0008] Reconstruct the chest wall vibration signal S, separate the data of each channel, and obtain the signal S generated by the heartbeat H , the signal S generated by the chest rise and fall caused by breathing B And the interference signal S O ;

[0009] According to the results of data reconstruction, the chest wall vibration signal S is decoupled and the interference signal S is filtered out. O , and the signal S generated from the heart beat H and the signal S generated by the chest rise and fall caused by breathing B In the process, the heartbeat signal S is separated and extracted. H .

[0010] As a preferred solution, in the step of using the millimeter wave radar to collect data from the chest cavity of the subject to be tested and obtaining the chest wall vibration signal S from the radar echo signal, when the radar wave contacts the human body, part of the energy is absorbed by the human body, and the other part is reflected and received by the antenna, thereby obtaining the echo signal S raw ; Radar T n The transmitting antennas take turns to transmit signals at a rate of 200 frames per second, and the echo signals are simultaneously detected by R n The receiving antennas receive, forming c = T n ×R n channels; each frame of signal contains d chirps, each chirp corresponds to r sampling points; each frame of data is a baseband complex signal with three dimensions [c, d, r].

[0011] As a preferred solution, in the chest wall vibration signal S, data is reconstructed, and the data of each channel is separated to obtain the signal S generated by the heart beat. H , the signal S generated by the chest rise and fall caused by breathing B And the interference signal S O When the echo signal S raw Expand according to the total number of chirps to get the expanded data S res_1 :

[0012] S res_1 =restructure(S raw ,R n ×r,D),

[0013]

[0014] Separate the data of each receiving antenna: along the data S res_1 The D dimension, each interval of r data is the data of a receiving antenna, traversing the data S res_1 Separate the data to obtain data S res_2 :

[0015] S res_2 =restructure(S res_1 ,R n ,r×C n )

[0016]

[0017] Separate the data of each channel: According to the signal composition, in the data of each receiving antenna, the data of one channel is obtained by d×r data length interval, and the data of one channel is obtained by T interval. nThe channel is all the data of the corresponding receiving antenna in the f frame. Based on this, the data of each receiving antenna is traversed, the data of each channel is separated, and the reconstructed data S is obtained. res_3 :

[0018] S res_3 =restructure(S res_2 ,c,r,d,F)

[0019] S res_3 ∈C c×r×d×F

[0020] Where F is the number of signal frames.

[0021] As a preferred solution, the interference signal S O These include signals caused by ground echoes, intermodulation products, discrete reflectors, vibrations of components inside the receiving device, and inadvertent movements of the human body.

[0022] As a preferred solution, according to the result of data reconstruction, the chest wall vibration signal S is decoupled and the interference signal S is filtered out. O In the step, the weight parameters W of the two dimensions of k channel data r and d are generated through the Hanning window. r (n), W d (n):

[0023] w(n)=hanning(L,H)

[0024] Where L and H are the length and height of the window function respectively;

[0025] When generating W r (n) When L = 1, H = n = r, W is generated d (n) when L = n = d, H = 1;

[0026] The data S(c k ) and W r (n) to perform the first smoothing of the data:

[0027] S sth_1 (c k )=S(c k )×w r (n)

[0028] Convert the signal from the time domain to the frequency domain along the r dimension, multiply the frequency domain signal by the d dimension weight, and perform a second smoothing:

[0029] S fr_1 (c k )=FFT(S sth_1 (c k ))

[0030] S sth_2 (c k )=S fr_1 (c k )×w d (n)

[0031] Then, along the d dimension, the data is Fourier transformed to obtain S fr_2 (c k ), the noise reduction and two-dimensional Fourier transform of the k-channel data are completed; then the zero-frequency component in the signal is moved to the center of the spectrum:

[0032] S fr (c k )=FFTShift(S fr_2 (c k ),γ)

[0033] Among them, S fr (c k ) represents the signal after the zero-frequency component is moved to the center of the spectrum, FFTShift() represents the signal after the zero-frequency component is moved to the center of the spectrum, and γ represents the dimension in which the operation is performed;

[0034] Repeat the above process until the data of each channel is processed;

[0035] By extracting the signal S generated by the heart beat H The signal S generated by the chest rise and fall caused by breathing B The combined signal S H +S B , filter out the interference signal S O .

[0036] As a preferred solution, in the step of extracting the signal S generated by the heart beat H The signal S generated by the chest rise and fall caused by breathing B The combined signal S H +S B , filter out the interference signal S O In the steps, the following formula is used to achieve signal amplification and data dimension reduction at the same time:

[0037]

[0038] S acc ∈C r×d

[0039] Where N represents the total number of channels;

[0040] Take out segments of the same length from the back end and front end of the entire data and put them in S accThe front and rear ends of the test data S are obtained det :

[0041] S det =com(S acc ,up,down)

[0042] In the formula, up means that the acc The front-end data, down is the back-end data, and com means to combine the data; det Traverse and calculate the background clutter power level at each location to identify possible targets:

[0043] T(k)=identify(S det (k),z(k),ref)

[0044]

[0045] Among them, T(k) represents the result of target recognition, which can be True or Fales. True means that there is a target at the corresponding position, and Fales means that there is no target at the corresponding position. det (k) represents the value of the kth reference cell, ref represents the value of the reference threshold, Z(k) represents the background clutter power level of the kth reference cell, UpCell and DownCell represent the background clutter levels before and after the reference cell respectively; the background clutter level of each position is calculated, and the difference between the data at the corresponding position and Z(k) is recorded as snr. When snr is greater than the reference threshold, it is considered that there is a target, otherwise, it is considered that there is no target;

[0046] After obtaining all possible targets, perform peak search processing on the target detection results to filter out all peaks:

[0047] tar=Peak(T)

[0048] In the formula, tar represents all peaks, that is, the set of all candidate targets;

[0049] Traverse tar, retain the target data within a specific threshold, and locate the signal S at the location of the heart tar :

[0050] S tar =select(tar,L1,L2)

[0051] In the formula, L1 represents the lower limit of the distance between the radar and the chest cavity, and L2 represents the upper limit between the radar and the chest cavity. The data within a specific threshold range is retained by select(), thereby filtering out the interference signal S O .

[0052] As a preferred solution, the signal S generated from the heart beat is H and the signal S generated by the chest rise and fall caused by breathing B In the process, the heartbeat signal S is separated and extracted. H The steps include:

[0053] The signal S generated by the chest rise and fall caused by breathing is obtained by performing a second-order difference B To filter:

[0054] S diff =diff([S tar (2)-S tar (1),

[0055] S tar (3)-S tar (2)......

[0056] S tar (m)-S tar (m-1)])

[0057] Among them, S diff YesS tar The second-order difference of , diff(.) represents the first-order difference;

[0058] Design FIR filter fir1 to filter the data, the passband boundary frequencies of the filter are w1 and w2 respectively:

[0059] fir1=fir(n,[w1w2],window)

[0060]

[0061] Where n represents the order of the filter, [w1w2] represents the cutoff frequency of the filter, window is the window function of the filter, M Prescend is the data sampling frequency, fc1 and fc2 are the upper and lower limits of the heartbeat amplitude respectively;

[0062] Use the filter to filter the target data:

[0063]

[0064] Among them, S H A signal representing the mechanical movement of the heart.

[0065] In a second aspect, a cardiac mechanical motion signal extraction system based on multi-domain joint modulation is provided, comprising:

[0066] A signal acquisition module, used to collect data from the chest cavity of the subject using a millimeter-wave radar, and obtain a chest wall vibration signal S from the radar echo signal;

[0067] The data reconstruction module is used to reconstruct the chest wall vibration signal S, separate the data of each channel, and obtain the signal S generated by the heart beat H , the signal S generated by the chest rise and fall caused by breathing B And the interference signal S O ;

[0068] The decoupling module is used to decouple the chest wall vibration signal S according to the data reconstruction result and filter out the interference signal S O , and the signal S generated from the heart beat H and the signal S generated by the chest rise and fall caused by breathing B In the process, the heartbeat signal S is separated and extracted. H .

[0069] According to a third aspect, an electronic device is provided, including:

[0070] A memory storing at least one instruction; and a processor executing the instruction stored in the memory to implement the cardiac mechanical motion signal extraction method based on multi-domain joint modulation.

[0071] In a fourth aspect, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for extracting cardiac mechanical motion signals based on multi-domain joint modulation is implemented.

[0072] Compared with the prior art, the present invention has at least the following beneficial effects:

[0073] By detecting the chest wall of the subject with millimeter wave radar, the chest wall vibration signal of the subject can be obtained. The chest wall vibration signal contains the signal generated by the heart beat, the signal generated by the chest rise and fall caused by breathing, and other interference signals. The mixed data of the receiving antenna is separated by data reconstruction, and the interference signal is further filtered out by decoupling, and the signal generated by the heart beat is separated and extracted from the signal generated by the heart beat and the signal generated by the chest rise and fall caused by breathing. The method based on multi-domain joint modulation of the present invention can successfully extract the weak signal of the mechanical movement of the heart, and provides technical support for the acquisition of non-contact physiological signals such as electrocardiogram waveform acquisition, real-time monitoring and alarm of heart health, and sports health monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0075] Figure 1 This is a flow chart of a cardiac mechanical motion signal extraction method based on multi-domain joint modulation according to the present invention;

[0076] Figure 2 This is a schematic diagram of an application scenario in which the millimeter wave radar is aimed at the chest of a subject for data collection in the present invention;

[0077] Figure 3 Schematic diagram of the reconstruction of the original radar signal. (a) is the signal before reconstruction, (b) is the signal after reconstruction. The blue color represents the imaginary part of the signal, and the red color represents the real part of the signal. The horizontal axis represents the number of sampling points, and the vertical axis represents the amplitude.

[0078] Figure 4 Schematic diagram of data smoothing and Fourier transform processing: (a) is the morphology before processing; (b) is the morphology after processing;

[0079] Figure 5 Schematic diagram of Fourier transform and channel superposition processing:

[0080] (a) is the result of Fourier transform of signal S, and (b) is the result after channel superposition;

[0081] Figure 6 Schematic diagram of target detection and PeakSearch processing:

[0082] (a) is the result of target detection, (b) is the result of PeakSearch;

[0083] Figure 7 Schematic diagram of high-order difference and time-frequency filtering for signals:

[0084] (a) is the signal before differentiation, (b) is the result after differentiation, and (c) is the cardiac mechanical motion signal obtained after filtering. DETAILED DESCRIPTION

[0085] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, ordinary technicians in this field can also obtain other embodiments without making creative work.

[0086] See also Figure 1 The cardiac mechanical motion signal extraction method based on multi-domain joint modulation in an embodiment of the present invention includes:

[0087] Step 1: spatially reconstruct the radar echo signal, including:

[0088] Step 1: Use the AWR1843 millimeter-wave radar to collect data from the chest of the person being tested. When the radar wave contacts the human body, part of the energy is absorbed by the human body, and the other part is reflected and received by the antenna, thereby obtaining the echo signal S. raw ∈C 1×T . T in radar systems n The transmitting antennas take turns to transmit signals at a rate of 200 frames per second, and the echo signals are simultaneously detected by R n The receiving antennas receive, forming c = T n ×R n channels. Each frame of signal contains d chirps, and each chirp corresponds to r sampling points. Therefore, each frame of data is actually a baseband complex signal with three dimensions [c, d, r]. In order to further process the signal, the signal S raw Expand according to the total number of chirps to get a new signal, expressed as S res_1 :

[0089] S res_1 =restructure(S raw ,R n ×r,D),

[0090]

[0091] Step 2: Separate the data from each receiving antenna along the signal S res_1 The D dimension of S is such that each interval of r data is the data of a receiving antenna. res_1 Separate the data to get S res_2 :

[0092] S res_2 =restructure(S res_1 ,R n ,r×C n )

[0093]

[0094] Step 3: Separate the data of each channel. According to the signal structure, in the data of each receiving antenna, the interval d×r data length is the data of one channel, and the interval T nThe channel is all the data of the receiving antenna in the fth frame. Based on this, the data of each receiving antenna is traversed, the data of each channel is separated, and the final reconstructed data is obtained, where F is the number of signal frames:

[0095] S res_3 =restructure(S res_2 ,c,r,d,F)

[0096] S res_3 ∈C c×r×d×F

[0097] Step 2: decoupling the radar echo signal, including:

[0098] Step 1: Since the radar antenna is facing the chest during data collection, the radar echo data S is decomposed into the signal S generated by the heart beat. H , the signal S generated by the chest rise and fall caused by breathing B , as well as interference signals S caused by ground echoes, intermodulation products, discrete reflectors, vibrations of components inside the receiving device, and inadvertent movements of the human body O :

[0099] It is described by the following formula:

[0100] S=S H +S B +S O

[0101] Step 2, due to the interference information S O There is a lot of information contained in it, so the interference information S is filtered by the Hanning window. O The high-frequency noise and low-amplitude signals in the ground echo are filtered out to improve the signal-to-noise ratio.

[0102] Through the Hanning window, the weight parameters W of the two dimensions of k channel data r and d are generated. r (n), W d (n):

[0103] w(n)=hanning(L,H)

[0104] Where L and H are the length and height of the window function respectively.

[0105] When generating W r (n) When L = 1, H = n = r, W is generated d (n) L = n = d, H = 1. The data S(c k ) and W r (n) and perform the first smoothing of the data to obtain S sth_1 :

[0106] S sth_1 (c k )=S(c k )×w r (n)

[0107] In order to filter out these high-frequency and low-amplitude noises to the greatest extent possible, the embodiment of the present invention converts the signal S from the time domain along the r dimension to the frequency domain, and then multiplies the frequency domain signal by the d-dimensional weight, and performs a second smoothing to obtain S sth_2 :

[0108] S fr_1 (c k )=FFT(S sth_1 (c k ))

[0109] S sth_2 (c k )=S fr_1 (c k )×w d (n)

[0110] Then, along the d dimension, the data is Fourier transformed to obtain S fr_2 (c k ), thus completing the noise reduction and two-dimensional Fourier transform of the k-channel data. In order to prevent the spectrum from leaking or drifting during the Fourier transform process, which affects the subsequent processing process, the zero-frequency component in the signal is moved to the center of the spectrum;

[0111] S fr (c k )=FFTShift(S fr_2 (c k ),γ)

[0112] Among them, S fr (c k ) represents the signal after the zero-frequency component is moved to the center of the spectrum, FFTShift() represents moving the zero-frequency component to the center of the data; γ represents the dimension in which the operation is performed.

[0113] After the k-channel data is processed as above, the k+1-channel data is smoothed and Fourier transformed as above until the data of each channel is processed, that is, the preliminary noise reduction and time-frequency domain conversion of the data are completed. O There are also noises caused by human motion and discrete reflectors in the image. The efficiency of extracting and separating these noises one by one is very low. Therefore, the embodiment of the present invention extracts S H +S B , so as to filter out the interference signal S O effect.

[0114] Step 3: To successfully extract S H +S B , it is necessary to find all possible targets from the radar signal. In order to improve the accuracy of target recognition, the embodiment of the present invention converts the signal S fr After taking the absolute value of the data of all channels, the data is accumulated to achieve signal amplification and data dimension reduction at the same time:

[0115]

[0116] S acc ∈C r×d

[0117] Where N represents the total number of channels.

[0118] Step 4: Data S after signal amplification acc The targets contained in S are easier to be identified by the algorithm. In order to further reduce the false alarm rate in the target recognition process, the embodiment of the present invention takes out segments of the same length from the back end and the front end of the entire data and puts them in S acc The front and rear ends of the test data S are obtained det :

[0119] S det =com(S acc ,up,down)

[0120] Among them, up means to put it in S acc The front-end data, down is the back-end data, and com means to combine the data. det Traverse and calculate the background clutter power level at each location to identify possible targets.

[0121] T(k)=identify(S det (k),z(k),ref)

[0122]

[0123] Wherein, T(k) represents the result of target recognition performed by the embodiment of the present invention, and the value is True and False, which respectively represent whether there is a target at the location. det (k) represents the value of the kth reference cell, ref represents the value of the reference threshold. Z(k) represents the background clutter power level of the kth reference cell. UpCell and DownCell represent the background clutter levels before and after the reference cell, respectively.

[0124] Table 1 Identify processing flow

[0125]

[0126] Table 1 calculates the background clutter level at each position, and uses the data at that position and z(k) as the difference, which is recorded as snr. When snr is greater than the reference threshold, it is considered that there is a target, otherwise it is considered that there is no target. After obtaining all possible targets, the results of target detection are processed by peaksearch to filter out all peaks:

[0127] tar=Peak(T)

[0128] Among them, tar represents all peaks, that is, the set of all candidate targets.

[0129] During the data collection process, the subjects are required to remain still, and the chest distance from the radar will be within a controllable range. Therefore, traversing tar, the target data within a specific threshold is retained, and finally the signal S at the location of the heart is located. tar , described by the following formula:

[0130] S tar =select(tar,L1,L2)

[0131] Among them, L1 represents the lower limit of the distance between the radar and the chest, and L2 represents the upper limit between the radar and the chest. Through select(), the data within a specific threshold is retained, and the interference signal S is successfully filtered out. O .

[0132] Step 5: For S obtained in step 4 H +S B , further decoupling is needed. The differential filtering method can effectively retain the key features of the signal. Through the first-order or second-order difference, the rapid changes in the signal can be quickly detected, and the slope or curvature information of the signal can be extracted. At the same time, the calculation is relatively simple and efficient. Therefore, in order to better B Filtering, the embodiment of the present invention performs a second-order difference on the signal to filter the signal fluctuations caused by breathing:

[0133] S diff =diff([S tar (2)-S tar (1),

[0134] S tar (3)-S tar (2)......

[0135] S tar (m)-S tar (m-1)])

[0136] Among them, S diff YesS tarThe second-order difference of , and diff(.) represents the first-order difference.

[0137] In order to further filter the signal, only the relatively clean cardiac mechanical motion signal S is left. H , the embodiment of the present invention designs an FIR filter fir1 to filter the data. The passband boundary frequencies of the filter are w1 and w2 respectively:

[0138] fir1=fir(n,[w1w2],window)

[0139]

[0140] Where n represents the order of the filter, [w1w2] is the cutoff frequency of the filter, window is the window function of the filter, and the embodiment of the present invention uses a Kaiser window with a length of n and a parameter of β, M Prescend is the data sampling frequency, and fc1 and fc2 are the upper and lower limits of the heartbeat amplitude, respectively. Then let the signal use the filter to filter the target data:

[0141]

[0142] Where S H A signal representing the mechanical movement of the heart.

[0143] At this point, the embodiment of the present invention completes S H and S B The decoupling between them successfully extracted the signal of cardiac mechanical motion.

[0144] See also Figure 2 , use the millimeter wave radar to collect data on the chest of the subject, and then you can get the chest wall vibration signal S of the subject, S includes the signal S generated by the heartbeat H , the signal S generated by the chest rise and fall caused by breathing B , interference signals S caused by ground echo, intermodulation products, discrete reflectors, vibration of components inside the receiving device, and inadvertent human movements O . See Figure 3 , extract 256 sampling points from the original radar signal and draw the waveform as Figure 3 As shown in (a), it is reconstructed into the data of the baseband complex signal, where the red part is the real part and the blue part is the imaginary part, as shown in Figure 3 As shown in (b), compared with Figure 3 From (a) and (b) in the figure, we can see that before the data is reconstructed, the data from multiple receiving antennas are mixed, and the imaginary and real parts of the signal are not clearly distinguished. However, the signal after data reconstruction solves this problem.

[0145] Continue to smooth and Fourier transform the reconstructed data. Use the data of one channel to show the effect before and after smoothing, such as Figure 4 As shown in the experimental results, Figure 4 (b) retains the original Figure 4 In (a), the part with higher peak value is filtered out, and the part with lower peak value is filtered out, thereby effectively filtering out the signal noise caused by environmental factors.

[0146] The smoothed signal is subjected to a two-dimensional Fourier transform, and then the data is amplified in the frequency domain. The present invention visualizes the data of one channel, and the processing result is as follows: Figure 5 As shown in the figure. During the data collection process, the human body is in a quasi-stationary state, so almost all the wave peaks are on the plane with a velocity of 0. At the same time, the higher the wave peak, the greater the energy of the radar echo, and the greater the possibility that there is an object at that location. After the spatial channel accumulation processing, the dimensionality reduction and amplification of the signal are achieved. By comparing the experimental results before and after signal amplification, it can be seen that the echo energy becomes larger after signal amplification at the location where the target is more likely to exist, making the target easier to identify.

[0147] See also Figure 6 ,from Figure 6 From (a), the target detection results include multiple adjacent targets, and there are multiple secondary peaks near the main peak. The energy of the secondary peaks is relatively weak, and the energy of the main peak is relatively strong. In order to avoid the secondary peaks interfering with the experimental results, PeakSearch is performed on the target set T to retain the main peak information and filter out the secondary peak information.

[0148] See also Figure 7 , in order to obtain S H +S B And filter out the noise S O , the results of PeakSearch are filtered again to retain the targets whose distance from the chest cavity is within a specific threshold. Figure 7 (a) is the result of filtering PeakSearch to get S H +S B In order to decouple the signal and further filter out the noise caused by breathing to avoid negative impact on the final ECG signal, the present invention performs a H +S B A second-order difference is performed to filter out large signal fluctuations, and then a FIR filter is performed to further filter out the signal SB, ultimately obtaining a signal of cardiac mechanical movement.

[0149] Another embodiment of the present invention further provides a cardiac mechanical motion signal extraction system based on multi-domain joint modulation, comprising:

[0150] A signal acquisition module, used to collect data from the chest cavity of the subject using a millimeter-wave radar, and obtain a chest wall vibration signal S from the radar echo signal;

[0151] The data reconstruction module is used to reconstruct the chest wall vibration signal S, separate the data of each channel, and obtain the signal S generated by the heart beat H , the signal S generated by the chest rise and fall caused by breathing B And the interference signal S O ;

[0152] The decoupling module is used to decouple the chest wall vibration signal S according to the data reconstruction result and filter out the interference signal S O , and the signal S generated from the heart beat H and the signal S generated by the chest rise and fall caused by breathing B In the process, the heartbeat signal S is separated and extracted. H .

[0153] Another embodiment of the present invention further provides an electronic device, comprising:

[0154] A memory storing at least one instruction; and a processor executing the instruction stored in the memory to implement the cardiac mechanical motion signal extraction method based on multi-domain joint modulation.

[0155] Another embodiment of the present invention further proposes a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for extracting cardiac mechanical motion signals based on multi-domain joint modulation is implemented.

[0156] Exemplarily, the instructions stored in the memory may be divided into one or more modules / units, which are stored in a computer-readable storage medium and executed by the processor to complete the cardiac mechanical motion signal extraction method based on multi-domain joint modulation of the present invention. The one or more modules / units may be a series of computer-readable instruction segments capable of completing specific functions, which are used to describe the execution process of the computer program in the server.

[0157] The electronic device may be a computing device such as a smart phone, a notebook, a PDA, and a cloud server. The electronic device may include, but is not limited to, a processor and a memory. Those skilled in the art will appreciate that the electronic device may also include more or fewer components, or a combination of certain components, or different components, for example, the electronic device may also include an input / output device, a network access device, a bus, etc.

[0158] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0159] The memory may be an internal storage unit of the server, such as a hard disk or memory of the server. The memory may also be an external storage device of the server, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the server. Furthermore, the memory may include both an internal storage unit of the server and an external storage device. The memory is used to store the computer-readable instructions and other programs and data required by the server. The memory may also be used to temporarily store data that has been output or is to be output.

[0160] It should be noted that the information interaction, execution process and other contents between the above-mentioned module units are based on the same concept as the method embodiment. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0161] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0162] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may at least include: any entity or device that can carry the computer program code to the camera device / terminal device, recording medium, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electrical carrier signal, telecommunication signal and software distribution medium. For example, a USB flash drive, a mobile hard disk, a disk or an optical disk.

[0163] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0164] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A cardiac mechanical motion signal extraction method based on multi-domain joint modulation, characterized in that: include: Use millimeter wave radar to collect data from the chest cavity of the subject, and obtain the chest wall vibration signal S from the radar echo signal; Reconstruct the chest wall vibration signal S, separate the data of each channel, and obtain the signal S generated by the heartbeat H , the signal S generated by the chest rise and fall caused by breathing B And the interference signal S O ; According to the results of data reconstruction, the chest wall vibration signal S is decoupled and the interference signal S is filtered out. O , and the signal S generated from the heart beat H and the signal S generated by the chest rise and fall caused by breathing B In the process, the heartbeat signal S is separated and extracted. H .

2. The cardiac mechanical motion signal extraction method based on multi-domain joint modulation according to claim 1 is characterized in that: In the step of using the millimeter wave radar to collect data from the chest cavity of the subject to be tested and obtaining the chest wall vibration signal S from the radar echo signal, when the radar wave contacts the human body, part of the energy is absorbed by the human body, and the other part is reflected and received by the antenna, thereby obtaining the echo signal S raw ; Radar T n The transmitting antennas transmit signals in turn at a rate of 200 frames per second, and the echo signals are simultaneously detected by R n The receiving antennas receive, forming c = T n ×R n channels; each frame of signal contains d chirps, each chirp corresponds to r sampling points; each frame of data is a baseband complex signal with three dimensions [c, d, r].

3. The cardiac mechanical motion signal extraction method based on multi-domain joint modulation according to claim 2 is characterized in that: In the above, the chest wall vibration signal S is reconstructed, and the data of each channel is separated to obtain the signal S generated by the heart beat. H , the signal S generated by the chest rise and fall caused by breathing B And the interference signal S O When the echo signal S raw Expand according to the total number of chirps to get the expanded data S res_1 : Separate the data of each receiving antenna: along the data S res_1 The D dimension, each interval of r data is the data of a receiving antenna, traversing the data S res_1 Separate the data to obtain data S res_2 : S res_2 =restructure(S res_1 ,R n ,r×C n ) Separate the data of each channel: According to the signal composition, in the data of each receiving antenna, the data of one channel is obtained by d×r data length interval, and the data of one channel is obtained by T interval. n The channel is all the data of the corresponding receiving antenna in the f frame. Based on this, the data of each receiving antenna is traversed, the data of each channel is separated, and the reconstructed data S is obtained. res_3 : S res_3 =restructure(S res_2 ,c,r,d,F) S res_3 ∈C c×r×d×F Where F is the number of signal frames.

4. The cardiac mechanical motion signal extraction method based on multi-domain joint modulation according to claim 1 is characterized in that: The interference signal S O These include signals caused by ground echoes, intermodulation products, discrete reflectors, vibrations of components inside the receiving device, and inadvertent movements of the human body.

5. The cardiac mechanical motion signal extraction method based on multi-domain joint modulation according to claim 1, characterized in that: According to the result of data reconstruction, the chest wall vibration signal S is decoupled and the interference signal S is filtered out. O In the step, the weight parameters W of the two dimensions of k channel data r and d are generated through the Hanning window. r (n), W d (n): w(n)=hanning(L,H) Where L and H are the length and height of the window function respectively; When generating W r (n) When L = 1, H = n = r, W is generated d (n) when L = n = d, H = 1; The data S(c k ) and W r (n) to perform the first smoothing of the data: S sth_1 (c k )=S(c k )×w r (n) Convert the signal from the time domain to the frequency domain along the r dimension, multiply the frequency domain signal by the d dimension weight, and perform a second smoothing: S fr_1 (c k )=FFT(S sth_1 (c k )) S sth_2 (c k )=S fr_1 (c k )×w d (n) Then, along the d dimension, the data is Fourier transformed to obtain S fr_2 (c k ), the noise reduction and two-dimensional Fourier transform of the k-channel data are completed; then the zero-frequency component in the signal is moved to the center of the spectrum: S fr (c k )=FFTShift(S fr_2 (c k ),γ) Among them, S fr (c k ) represents the signal after the zero-frequency component is moved to the center of the spectrum, FFTShift() represents the signal after the zero-frequency component is moved to the center of the spectrum, and γ represents the dimension in which the operation is performed; Repeat the above process until the data of each channel is processed; By extracting the signal S generated by the heart beat H The signal S generated by the chest rise and fall caused by breathing B The combined signal S H +S B , filter out the interference signal S O .

6. The cardiac mechanical motion signal extraction method based on multi-domain joint modulation according to claim 5, characterized in that: In the above method, the signal S generated by extracting the heart beat is H The signal S generated by the chest rise and fall caused by breathing B The combined signal S H +S B , filter out the interference signal S O In the steps, the following formula is used to achieve signal amplification and data dimension reduction at the same time: S acc ∈C r×d Where N represents the total number of channels; Take out fragments of the same length from the back end and front end of the entire data and put them in S acc The front and rear ends of the test data S are obtained det : S det =com(S acc ,up,down) In the formula, up means that the acc The front-end data, down is the back-end data, and com means to combine the data; det Traverse and calculate the background clutter power level at each location to identify possible targets: T(k)=identify(S det (k),z(k),ref) Among them, T(k) represents the result of target recognition, which can be True or Fales. True means that there is a target at the corresponding position, and Fales means that there is no target at the corresponding position. det (k) represents the value of the kth reference cell, ref represents the value of the reference threshold, Z(k) represents the background clutter power level of the kth reference cell, UpCell and DownCell represent the background clutter levels before and after the reference cell respectively; the background clutter level of each position is calculated, and the difference between the data at the corresponding position and Z(k) is recorded as snr. When snr is greater than the reference threshold, it is considered that there is a target, otherwise, it is considered that there is no target; After obtaining all possible targets, perform peak search processing on the target detection results to filter out all peaks: tar=Peak(T) In the formula, tar represents all peaks, that is, the set of all candidate targets; Traverse tar, retain the target data within a specific threshold, and locate the signal S at the location of the heart tar : S tar =select(tar,L1,L2) In the formula, L1 represents the lower limit of the distance between the radar and the chest cavity, and L2 represents the upper limit between the radar and the chest cavity. The data within a specific threshold range is retained by select(), thereby filtering out the interference signal S O .

7. The cardiac mechanical motion signal extraction method based on multi-domain joint modulation according to claim 6, characterized in that: The signal S generated from the heart beat H and the signal S generated by the chest rise and fall caused by breathing B In the process, the heartbeat signal S is separated and extracted. H The steps include: The signal S generated by the chest rise and fall caused by breathing is obtained by performing a second-order difference B To filter: S diff =diff([S tar (2)-S tar (1), S tar (3)-S tar (2)...... S tar (m)-S tar (m-1)]) Among them, S diff YesS tar The second-order difference of , diff(.) represents the first-order difference; Design FIR filter fir1 to filter the data, the passband boundary frequencies of the filter are w1 and w2 respectively: fir1=fir(n,[w1w2],window) Where n represents the order of the filter, [w1w2] represents the cutoff frequency of the filter, window is the window function of the filter, MPrescend is the data sampling frequency, fc1 and fc2 are the upper and lower limits of the heartbeat amplitude respectively; Use the filter to filter the target data: Among them, S H A signal representing the mechanical movement of the heart.

8. A cardiac mechanical motion signal extraction system based on multi-domain joint modulation, characterized in that: include: A signal acquisition module, used to collect data from the chest cavity of the subject using a millimeter-wave radar, and obtain a chest wall vibration signal S from the radar echo signal; The data reconstruction module is used to reconstruct the chest wall vibration signal S, separate the data of each channel, and obtain the signal S generated by the heart beat H , the signal S generated by the chest rise and fall caused by breathing B And the interference signal S O ; The decoupling module is used to decouple the chest wall vibration signal S according to the data reconstruction result and filter out the interference signal S O , and the signal S generated from the heart beat H and the signal S generated by the chest rise and fall caused by breathing B In the process, the heartbeat signal S is separated and extracted. H .

9. An electronic device, characterized in that: include: A memory storing at least one instruction; and a processor, executing instructions stored in the memory to implement the cardiac mechanical motion signal extraction method based on multi-domain joint modulation as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method for extracting cardiac mechanical motion signals based on multi-domain joint modulation as described in any one of claims 1 to 7 is implemented.

Citation Information

Patent Citations

  • Life signal monitoring method based on millimeter wave radar

    CN113116342A

  • Personnel respiration and heartbeat detection method based on millimeter wave radar

    CN113440120A

  • Millimeter wave radar vital sign signal extraction and measurement method

    CN115089143A

  • Three-dimensional imaging material level measuring method, device and system of millimeter wave radar

    CN116105830A

  • Human body vital sign detection method based on MIMO radar and beam forming

    CN116840805A

Cited By

  • Non-contact electrocardiosignal measuring device and measuring method thereof

    CN119837523A