DCG detection method based on K-band continuous wave FDDBF

Abstract

The invention discloses a DCG detection method based on a K-band continuous wave FDDBF, and the method comprises the steps: transmitting electromagnetic waves to a detection object through a K-band continuous wave millimeter wave radar, receiving an echo signal, carrying out the data reconstruction of an intermediate frequency signal obtained through frequency mixing, obtaining a virtual frequency domain array, carrying out the linear extraction of waveforms corresponding to different frequencies through the FDDBF, and carrying out the detection of the waveform. And a pure DCG signal, a differential DCG signal and a chest movement signal are obtained through summation filtering. According to the invention, linear and undistorted DCG and differential DCG signal extraction is realized through K-band continuous wave millimeter waves, and the quality and precision of the DCG signals are significantly improved.

Description

Technical Field

[0001] The present invention relates to a technology in the field of medical detection, in particular to a Doppler cardiogram (DCG) detection method based on K-band continuous wave frequency domain digital beamforming (FDDBF). Background Art

[0002] Existing cardiac activity is mainly detected by electrocardiogram (ECG), but due to its contact characteristics, it is easy to cause discomfort during long-term monitoring and is not suitable for routine use. Doppler cardiography (DCG) is a non-contact detection method that reflects the mechanical movement of the heart and has broad medical application prospects. However, traditional technology has the problem of respiratory interference when extracting DCG. The respiratory signal interferes with the heart signal. Existing methods mostly use digital filtering technology to eliminate it, but it is difficult to avoid nonlinear distortion. Summary of the invention

[0003] In view of the problems of respiratory interference and signal distortion in the existing DCG detection technology, the present invention proposes a DCG detection method based on K-band continuous wave FDDBF, which realizes linear and distortion-free DCG and differential DCG signal extraction through K-band continuous wave millimeter wave, and significantly improves the quality and accuracy of DCG signals.

[0004] The present invention is achieved through the following technical solutions:

[0005] The present invention relates to a DCG detection method based on K-band continuous wave FDDBF. After a K-band continuous wave millimeter wave radar transmits electromagnetic waves to a detection object and receives an echo signal, data reconstruction is performed on an intermediate frequency signal obtained by mixing to obtain a virtual frequency domain array. Then, waveforms corresponding to different frequencies are linearly extracted through FDDBF, and pure DCG signals, differential DCG signals and chest motion signals are obtained through summation and filtering.

[0006] The present invention relates to a system for implementing the above method, comprising: a displacement demodulation unit, a frequency domain array reconstruction unit, a frequency domain array processing unit, and a DCG extraction unit, wherein: the displacement demodulation unit performs phase demodulation processing according to a CW intermediate frequency signal to obtain chest movement, the frequency domain array reconstruction unit performs grouping reconstruction according to the phase characteristics of the intermediate frequency signal to obtain an equivalent frequency domain array, the frequency domain array processing unit performs frequency domain beamforming and frequency domain digital beamforming processing according to the phase relationship between frequency domain array elements to obtain the spectrum of chest movement and extract the time domain waveforms corresponding to the respiratory fundamental wave and harmonic waves, and the DCG extraction unit removes the acquired respiratory waveforms of different frequencies from the total movement to obtain a pure Doppler electrocardiogram (DCG). Technical Effects

[0007] The present invention utilizes the motion phase characteristics obtained by demodulating CW intermediate frequency signals to group and reconstruct displacement data, obtain an equivalent frequency domain array, and obtain the frequency distribution of motion through frequency domain beamforming technology. Then, the time domain waveforms corresponding to the fundamental and harmonic waves of breathing are linearly obtained using frequency domain digital beamforming technology, and the pure heartbeat motion is obtained by removing the breathing component from the total motion. Compared with the prior art, the present invention can linearly and efficiently utilize millimeter wave radar to non-contactly obtain the heartbeat displacement of human targets, avoiding the waveform distortion and subjective selection of cutoff frequency caused by the filter technology in the prior art, and avoiding the problems of large computational complexity and subjective decomposition parameter selection of technologies such as VMD and CEEMDAN. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 It is a flow chart of the present invention;

[0009] Figure 2 It is a flow chart of continuous wave frequency domain digital beamforming;

[0010] Figure 3 It is a simulation diagram of the present invention;

[0011] In the figure: (a) is a simulation scene, (b) is the original motion spectrum obtained by the radar and the spectrum after processing by the present invention, (c) is the original motion time domain waveform obtained by the radar, the breathing waveform obtained by the present invention and the DCG time domain waveform obtained by extraction, (d) is the DCG signal obtained after simulation processing and the simulated reference motion, (e) is a comparison of DCG extraction results of different existing technologies;

[0012] Figure 4 This is an experimental diagram of the present invention;

[0013] In the figure: (a) is the experimental scene and system settings. (b) is the original motion time domain waveform obtained by the radar, the breathing waveform obtained by the present invention, and the DCG time domain waveform obtained by extraction. (c) is the comparison between the differential DCG signal obtained after processing by the present invention and the reference ECG signal. DETAILED DESCRIPTION

[0014] like Figure 1 As shown, this embodiment involves a DCG detection method based on K-band continuous wave FDDBF. After the K-band continuous wave millimeter wave radar transmits electromagnetic waves to the detection object and receives the echo signal, the intermediate frequency signal obtained by mixing is reconstructed to obtain a virtual frequency domain array, and then the waveforms corresponding to different frequencies are linearly extracted through FDDBF, and the pure DCG signal, differential DCG signal and chest motion signal are obtained through summation and filtering.

[0015] The K-band continuous wave millimeter wave radar comprises: a monolithic microwave integrated unit MMIC and a microcontroller unit MCU, a transmitting antenna TX and a receiving antenna RX respectively connected thereto.

[0016] The micro control unit MCU comprises: a digital-to-analog converter DAC, an analog-to-digital converter ADC and a direct current adjustment unit.

[0017] like Figure 2 As shown, the intermediate frequency signal Where: A is the amplitude, λ is the wavelength, x(t) is the motion signal of the chest cavity, d 0 is the distance between the radar and the target, and Δθ(t) is the residual phase. The chest motion signal is obtained by demodulating the intermediate frequency signal K is the total number of harmonics, and A k 、f k and are the amplitude, frequency and initial phase of each harmonic, respectively.

[0018] The data reconstruction means that during the detection duration T, the sampling rate f s After collecting a total of N+M-1 sampling points, they are reorganized into a virtual frequency domain array of M channels, each containing N snapshots. Where: A(f) = [a(f 1 ),a(f 2 ),…,a(f K )] is the array manifold vector, a(f k )=[1,exp(j2πf k t s ),…,exp(j2πf k (M-1)t s )] T is the steering vector, X(t)=[x 1 (t),x 2 (t),…,x K (t)] T is the corresponding signal vector, To reconstruct the motion signal at different frequencies after grouping, N(t) is additive white Gaussian noise.

[0019] The spectrum distribution S obtained by frequency domain spatial beam scanning of Sarray CBF (f i )=l(f i )S[m,n]S H [m,n]l H (fi ),in: i is the frequency index, S and S Array It is the same matrix, and the superscript H indicates the conjugate transpose of this matrix.

[0020] The fundamental breathing frequency f is identified by analyzing the frequency domain of the signal k and its harmonics. It is worth noting that respiratory harmonics with frequencies higher than heartbeat are usually weaker and have less impact on DCG, so they can be ignored. In addition, the FDDBF technique is used to construct the steering vector l(pf) corresponding to the respiratory frequency. k ). These steering vectors are then directed to the corresponding frequency regions to extract the time domain waveform associated with each frequency.

[0021] The frequency domain array processing unit obtains the time domain signals of different respiratory frequency components, and the frequency domain digital beamforming (FDDBF) is performed from the frequency domain array S Array Extract each frequency-related time domain waveform from the waveform generator, specifically: Among them: the time domain displacement and time domain waveform corresponding to different respiratory frequencies p is the index of the basic respiratory frequency and its harmonics, p=1,2,…,P.

[0022] The pure DCG signal refers to: Wherein: x(t) is the chest motion signal; D-DCG(t)=diff(DCG(t)), wherein: D represents the difference of DCG, DCG is the heartbeat motion, i.e. Doppler cardiogram, x(t) is the collected chest motion signal, and D-DCG is the differential DCG.

[0023] Through specific practical experiments, Figure 3 The high-precision slide shown in (a) is used to simulate the breathing and heartbeat movements of the human body. The signal is received by the radar unit and processed by the present invention to obtain high-linearity, high-precision DCG and differential DCG signals.

[0024] The slide is set up to reproduce breathing and heart motion with high accuracy (1 μm). Breathing is modeled by a combination of multiple sine waves, ranging from the fundamental frequency to the third harmonic of breathing. Since an accurate model of heart motion is not yet available, a DCG signal pre-recorded from a subject during breath holding is used. This signal represents one cardiac cycle and is periodically stretched to simulate the motion caused by the heartbeat.

[0025] like Figure 3 The dotted line in (b) shows the identified breathing frequency. FDDBF is used to extract the waveforms corresponding to these frequency components. By summing these components, the overall breathing signal is reconstructed with high linearity, as shown in Figure 3As shown by the dotted line in (c), the overall motion signal is processed to remove the respiratory component and obtain a pure DCG, as shown in Figure 3 (c) and Figure 3 As shown by the short dashed line in (d). Figure 3 The solid line in (b) is the spectrum of the signal, indicating that the respiratory component is effectively removed, leaving only the heart frequency, some of which is shown in Table 1.

[0026] Table 1

[0027] like Figure 3 (e) shows a horizontal comparison between the present invention and the prior art, wherein the correlation coefficient of the present invention is significantly better than the prior average level.

[0028] like Figure 4 As shown in Figure 1, it is a human body measurement experiment, that is, it is carried out on a person sitting in front of the radar. The radar measurement is carried out simultaneously with the synchronized ECG, and the ECG is used as a reference signal; Figure 4 (b) is the demodulated motion signal obtained from the radar; the present invention extracts Figure 4 The respiratory component shown by the dotted line in (b) is obtained as Figure 4 (b) The pure DCG signal is shown by the short dashed line. In addition, the D-DCG signal is extracted and compared with the ECG data, as shown in Figure 4 (c), some of its data are shown in Table 2.

[0029] Table 2

[0030] The comparison results show remarkable consistency in the RR gap with a root mean square error (RMSE) of 6.82ms. These results verify that the technique is capable of highly accurate, linear, and distortion-free pure DCG and differential DCG reconstruction, highlighting its significant potential in radar-based medical applications.

[0031] Compared with the prior art, the present invention utilizes the phase characteristics of the intermediate frequency signal to reconstruct the intermediate frequency signal to construct a virtual frequency domain array, utilizes the traditional beamforming technology to perform spectrum estimation on the frequency domain array formed by the motion to obtain the frequency corresponding to the motion component, utilizes the frequency domain digital beamforming technology to guide the frequency space of the fundamental and harmonic waves of the breathing, and extracts the time domain waveform corresponding to different frequency spaces, and finally realizes the extraction of DCG signals and differential DCG signals. Compared with the traditional technology, this technology has the characteristics of high efficiency, robustness, no distortion, linearity, etc., and is of great significance in the application of non-contact vital signs monitoring.

[0032] The above-mentioned specific implementation can be partially adjusted in different ways by those skilled in the art without departing from the principle and purpose of the present invention. The protection scope of the present invention shall be based on the claims and shall not be limited by the above-mentioned specific implementation. Each implementation scheme within its scope shall be subject to the constraints of the present invention.

Claims

1. A DCG detection method based on K-band continuous wave FDDBF, characterized in that: After the K-band continuous wave millimeter wave radar transmits electromagnetic waves to the detection object and receives the echo signal, the intermediate frequency signal obtained by mixing is reconstructed to obtain a virtual frequency domain array, and then the waveforms corresponding to different frequencies are linearly extracted through FDDBF. After summation and filtering, the pure DCG signal, differential DCG signal and chest motion signal are obtained.

2. The DCG detection method based on K-band continuous wave FDDBF according to claim 1 is characterized in that: The K-band continuous wave millimeter wave radar comprises: a monolithic microwave integrated unit MMIC and a microcontroller unit MCU, a transmitting antenna TX and a receiving antenna RX respectively connected thereto; The micro control unit MCU comprises: a digital-to-analog converter DAC, an analog-to-digital converter ADC and a direct current adjustment unit.

3. The DCG detection method based on K-band continuous wave FDDBF according to claim 1 is characterized in that: The intermediate frequency signal Where: A is the amplitude, λ is the wavelength, x(t) is the collected chest motion signal, d0 is the distance between the radar and the target, Δθ(t) is the residual phase, and the chest motion signal is obtained by demodulating the intermediate frequency signal K is the total number of harmonics, and A k 、f k and are the amplitude, frequency and initial phase of each harmonic, respectively.

4. The DCG detection method based on K-band continuous wave FDDBF according to claim 1 is characterized in that: The data reconstruction means that during the detection duration T, the sampling rate f s After collecting a total of N+M-1 sampling points, they are reorganized into a virtual frequency domain array of M channels, each containing N snapshots. Among them: A(f)=[a(f1),a(f2),…,a(f K )] is the array manifold vector, a(f k )=[1,exp(j2πf k t s ),…,exp(j2πf k (M-1)t s )] T is the guiding vector, X(t)=[x1(t),x2(t),…,x K (t)] T is the corresponding signal vector, To reconstruct the motion signal at different frequencies after grouping, N(t) is additive white Gaussian noise; The spectrum distribution S obtained by frequency domain spatial beam scanning of Sarray CBF (f i )=l(f i )S[m,n]S H [m,n]l H (f i ),in: i is the frequency index, S and S Array It is the same matrix, and the superscript H indicates the conjugate transpose of this matrix.

5. The DCG detection method based on K-band continuous wave FDDBF according to claim 1 is characterized in that: The frequency domain array processing unit obtains the time domain signals of different respiratory frequency components, and the frequency domain digital beamforming (FDDBF) is performed from the frequency domain array S Array Extract each frequency-related time domain waveform from the waveform generator, specifically: Among them: the time domain displacement and time domain waveform corresponding to different respiratory frequencies p is the index of the basic respiratory frequency and its harmonics, p=1,2,…,P.

6. The DCG detection method based on K-band continuous wave FDDBF according to claim 1 is characterized in that: The pure DCG signal refers to: Wherein: x(t) is the chest motion signal; D-DCG(t)=diff(DCG(t)), wherein: D represents the difference of DCG, DCG is the heartbeat motion, i.e. Doppler cardiogram, x(t) is the collected chest motion signal, and D-DCG is the differential DCG.

7. A DCG detection system for implementing the method described in any one of claims 1 to 6, characterized in that: include: A displacement demodulation unit, a frequency domain array reconstruction unit, a frequency domain array processing unit, and a DCG extraction unit, wherein: the displacement demodulation unit performs phase demodulation processing according to the CW intermediate frequency signal to obtain the movement of the chest cavity; the frequency domain array reconstruction unit performs grouping reconstruction according to the phase characteristics of the intermediate frequency signal to obtain an equivalent frequency domain array; the frequency domain array processing unit performs frequency domain beamforming and frequency domain digital beamforming processing according to the phase relationship between frequency domain array elements to obtain the spectrum of the chest cavity movement and extract the time domain waveforms corresponding to the respiratory fundamental wave and harmonic waves; the DCG extraction unit removes the acquired respiratory waveforms of different frequencies from the total movement to obtain a pure Doppler echocardiogram (DCG).

Images