CN114680848A - Millimeter wave vital sign detection method introducing compressed sensing theory - Google Patents

Millimeter wave vital sign detection method introducing compressed sensing theory Download PDF

Info

Publication number
CN114680848A
CN114680848A CN202210219024.0A CN202210219024A CN114680848A CN 114680848 A CN114680848 A CN 114680848A CN 202210219024 A CN202210219024 A CN 202210219024A CN 114680848 A CN114680848 A CN 114680848A
Authority
CN
China
Prior art keywords
vital sign
millimeter wave
detection method
sign detection
compressed sensing
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202210219024.0A
Other languages
Chinese (zh)
Inventor
曹欣远
张忠祥
陈兵兵
齐琦
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hefei Normal University
Original Assignee
Hefei Normal University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Hefei Normal University filed Critical Hefei Normal University
Priority to CN202210219024.0A priority Critical patent/CN114680848A/en
Publication of CN114680848A publication Critical patent/CN114680848A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/02—Detecting, measuring or recording pulse, heart rate, blood pressure or blood flow; Combined pulse/heart-rate/blood pressure determination; Evaluating a cardiovascular condition not otherwise provided for, e.g. using combinations of techniques provided for in this group with electrocardiography or electroauscultation; Heart catheters for measuring blood pressure
    • A61B5/0205—Simultaneously evaluating both cardiovascular conditions and different types of body conditions, e.g. heart and respiratory condition
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7235—Details of waveform analysis
    • A61B5/7253—Details of waveform analysis characterised by using transforms
    • A61B5/7257—Details of waveform analysis characterised by using transforms using Fourier transforms

Abstract

The invention is suitable for the field of millimeter wave vital sign detection, provides a millimeter wave vital sign detection method introducing a compressive sensing theory, and can effectively realize the non-contact real-time monitoring of human respiration and heart rate by a Frequency Modulated Continuous Wave (FMCW) radar system while greatly reducing the transmission frame rate of a radio frequency front end. The algorithm is used for constructing an underdetermined equation by introducing a compressed sensing theory aiming at the phase information of a human body target, and can accurately capture the position of a spectral peak by directly adopting a small amount of inner product operation under the condition of not calculating a phase signal frequency spectrum, thereby outputting vital sign data. Compared with the existing FMCW radar sign monitoring system, the method has obvious advantages in the aspects of Chirp number required to be transmitted by the radio frequency module, calculation cost of the DSP and the like, and the algorithm can be fused with fast Fourier transform, wavelet transform and various spectrum thinning technologies, and can be used in the fields of nursing monitoring, sleep analysis, family medical treatment, earthquake rescue and the like.

Description

Millimeter wave vital sign detection method introducing compressed sensing theory
Technical Field
The invention belongs to the field of millimeter wave vital sign detection, and particularly relates to a millimeter wave vital sign detection method introducing a compressive sensing theory.
Background
With the rapid development of technologies such as radio frequency, DSP chips, integrated circuits, sensors, and internet of things, research on millimeter wave radar products and technologies is receiving increasing attention. Especially in world abuse of new crown diseases, the demand for contactless technologies is increasing. In recent years, a series of domestic and foreign enterprises, colleges and scientific research institutes represented by Texas Instruments (TI) have conducted research, design and development around millimeter wave radar sign monitoring systems. At present, the mainstream algorithm for measuring and calculating the respiration and heart rate of a human body by applying an FMCW radar is still based on FFT (fast Fourier transform and continuous phase-shift keying), and spectral peak search is carried out on a respiratory frequency band (0.1-0.5Hz) and a heartbeat frequency band (0.8-2Hz) respectively by unwrapping, differentiating and filtering phase signals of a human body target, so as to further convert a corresponding BPM (BPM) value.
Compressed Sensing (CS) is a new signal processing theory that has attracted much attention in recent years, and mainly includes three major elements, namely compressed sampling, sparse representation and recovery algorithm, and can achieve accurate reconstruction of an original signal under a condition much smaller than a Nyquist sampling rate. The invention provides a novel low-complexity FMCW radar rear-end intermediate frequency signal processing algorithm by introducing a compressed sensing idea for the first time, and the method can directly and accurately capture the spectral peak position of the FMCW radar without acquiring human phase spectrum information so as to output vital sign data.
Disclosure of Invention
The invention aims to provide a millimeter wave vital sign detection method introducing a compressed sensing theory, and aims to construct a novel low-complexity basic algorithm which can be effectively applied to a millimeter wave radar vital sign monitoring system. By applying the method, the transmission frame rate of the radio frequency front end of the FMCW radar sign monitoring system and the operation amount of the DSP chip can be obviously reduced.
The invention is realized by the following steps:
1) constructing an underdetermined equation; recording the phase signal as
Figure BDA0003536138450000021
Having a frequency spectrum of
Figure BDA0003536138450000022
The respiration and heart rate values of the measured human body are respectively determined by
Figure BDA0003536138450000023
Determining the peak positions at 0.1-0.5Hz and 0.8-2 Hz; the model is based on
Figure BDA0003536138450000024
And
Figure BDA0003536138450000025
the following matrix equation is satisfied:
Figure BDA0003536138450000026
wherein A is an inverse Fourier transform base, and partial lines of A are selected and marked as AEXTSimultaneously to each other
Figure BDA0003536138450000027
The same drawing is also performed
Figure BDA0003536138450000028
Then there is
Figure BDA0003536138450000029
2) Determining the position of a spectral peak; considering AEXTThe module values of all the columns are different, firstly, every column is normalized to form partial Fourier matrix and said partial Fourier matrix is recorded as AEXT1The ith column is marked as Ai EXT1And then the following internal product comparison and screening are carried out:
Figure BDA00035361384500000210
3) outputting sign data; suppose Ab EXT1And Ah EXT1Corresponding spectral peaks are respectively located at
Figure BDA00035361384500000211
The respiration and heart rate values of the measured human body at the b-th and h-th positions under the frequency spectrum resolution delta f are the
fb_BPM＝60(b-1)·Δf (4)
fh_BPM＝60(h-1)·Δf (5)。
In the further technical scheme, in the construction of the underdetermined equation in the step 1), from the perspective of CS, the equation (2) can be regarded as a pair
Figure BDA00035361384500000212
Several observations are carried out with AEXTAs an observation matrix, a matrix is used,
Figure BDA00035361384500000213
namely the observation result.
In the further technical scheme, in the step 1) of constructing the underdetermined equation, A is one of CZT base and DWT base.
In the further technical scheme, in the step 1) of constructing the underdetermined equation, for the equation (2), finding out
Figure BDA0003536138450000031
The peak position of (c).
In the further technical scheme, in the step 1) of constructing the underdetermined equation, according to the principle of maximum inner product in the matching pursuit algorithm, screening out the AEXTNeutralization of
Figure BDA0003536138450000032
The most similar column.
In a further technical scheme, in the step 2) of determining the spectral peak position, the inner product comparison is actually performed only by respectively limiting two frequency bands, namely a respiratory frequency band and a heartbeat frequency band.
The further technical proposal is that the steps are2) In the determination of the position of the spectral peak, the maximum inner array A is determined along with the maximum inner array A in the two frequency bands of the respiration frequency band and the heartbeat frequency bandb EXT1、Ah EXT1Is selected, corresponding to
Figure BDA0003536138450000033
Two peak positions in (a) are determined.
Compared with the prior art, the invention has the following beneficial effects:
according to the millimeter wave vital sign detection method introduced with the compressive sensing theory, the number of Chirp to be transmitted by a radio frequency module and the operation times of corresponding modules for data acquisition, processing and the like are directly reduced from the aspect of an algorithm by constructing an underdetermined equation, the hardware load is reduced, the solution of an underdetermined equation is skipped in the process of selecting an in-band product comparison and screening, the spectrum peak position is directly captured with low complexity, and the operation cost is saved; by applying the method, the transmission frame rate of the radio frequency front end of the FMCW radar sign monitoring system and the operation amount of the DSP chip can be obviously reduced.
Drawings
FIG. 1 is a comparison schematic diagram of a Chirp transmitting situation of a radar of the invention;
FIG. 2 is a flowchart of the main process of the present invention;
FIG. 3 is a schematic diagram of the system for algorithm verification and the location of the human target under test according to the present invention;
FIG. 4 is a graph of the peak frequency calculation error for a Chirp number of 0.1-0.5Hz (the "respiratory" band) of the present invention;
FIG. 5 is a graph of the calculated error variation of the peak frequency for a Chirp number of 0.8-2Hz ("heartbeat" frequency band) in accordance with the present invention;
FIG. 6 is a graph comparing the calculated peak frequency of the spectrum with time;
FIG. 7 is a graph comparing the peak positions obtained for the novel method of the present invention at 0.1-0.5Hz ("respiratory" frequency band) and CZT;
FIG. 8 is a graph comparing the peak positions obtained for the novel method of the present invention at 0.8-2Hz ("heartbeat" frequency band) and CZT.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Specific implementations of the present invention are described in detail below with reference to specific embodiments.
As shown in fig. 1 to 8, the millimeter wave vital sign detection method introduced with the compressive sensing theory provided by the present invention includes constructing a novel low-complexity basic algorithm which can be effectively applied to a millimeter wave radar vital sign monitoring system, and the algorithm is based on the compressive sensing concept, and realizes direct capture of the spectral peak position without calculating the phase spectrum of the human body by constructing an underdetermined equation and combining a small amount of inner product operation, and further outputs vital sign data of human respiration, heart rate, etc., and further includes the following steps:
1) constructing an underdetermined equation; in order to obtain beat frequency and phase change information generated by a human body target, the FFT is generally carried out on the echo intermediate frequency signal of each Chirp, and phase extraction, unwrapping and differentiation are carried out along with the determination of a distance unit where the human body target is located, so as to obtain a phase signal for vital sign measurement. Noting the phase signal as
Figure BDA0003536138450000041
Having a frequency spectrum of
Figure BDA0003536138450000042
The respiration and heart rate values of the measured human body are respectively determined by
Figure BDA0003536138450000043
Determining the peak positions at 0.1-0.5Hz and 0.8-2 Hz; to capture the peak position of the spectrum, the conventional method always needs to firstly obtain
Figure BDA0003536138450000044
And traversing the amplitude spectrum. To pair
Figure BDA0003536138450000045
The solution of (2) occupies most of the calculation amount thereof, and
Figure BDA0003536138450000046
the calculation results of the elements in (a) are not directly related to the respiratory or heartbeat measurement values. If can skip the pair
Figure BDA0003536138450000047
The position of the spectrum peak is directly found by solving, so that the calculation resource is saved better; in consideration, the invention constructs an underdetermined equation solving model by using the concept of compressed sensing; the model is based on
Figure BDA0003536138450000051
And
Figure BDA0003536138450000052
the following matrix equation is satisfied:
Figure BDA0003536138450000053
wherein A is an inverse Fourier transform base. Selecting partial rows of A as AEXTSimultaneously to each other
Figure BDA0003536138450000054
The same drawing is also performed
Figure BDA0003536138450000055
Then there is
Figure BDA0003536138450000056
2) Determining the position of a spectral peak; considering AEXTThe module values of all the columns are different, firstly, every column is normalized to form partial Fourier matrix and said partial Fourier matrix is recorded as AEXT1The ith column is marked as Ai EXT1And then the following internal product comparison and screening are carried out:
Figure BDA0003536138450000057
3) outputting sign data; suppose Ab EXT1And Ah EXT1Corresponding spectral peaks are respectively located at
Figure BDA0003536138450000058
The values of the respiration and heart rate of the measured human body at the b-th and h-th positions under the frequency spectrum resolution delta f are
fb_BPM＝60(b-1)·Δf (4)
fh_BPM＝60(h-1)·Δf (5)。
In the embodiment of the present invention, as shown in fig. 1, as a preferred embodiment of the present invention, in the construction of the underdetermined equation, in the step 1) of the construction of the underdetermined equation, the equation (2) can be regarded as a pair from the perspective of CS
Figure BDA0003536138450000059
Several observations are carried out with AEXTAs an observation matrix, a matrix is used,
Figure BDA00035361384500000510
namely the observation result.
In the embodiment of the present invention, as shown in fig. 1, as a preferred embodiment of the present invention, in the construction of the underdetermined equation, frequency spectrum refinement or other time-frequency analysis methods are considered, and a may also be a CZT base, a DWT base, or the like.
Due to the fact that
Figure BDA00035361384500000511
Each element in the FMCW radar signal corresponds to a Chirp signal, and the extraction of the elements means the reduction of the number of Chirp signals, so that if equation (2) is solved, the transmission frame rate of the FMCW radar radio frequency front end and the corresponding transmission frame rate are predictedThe processing times of a series of links such as mixing, sampling, FFT, phase extraction, unwrapping, difference, filtering and the like are expected to be reduced.
In the embodiment of the present invention, as shown in fig. 3, as a preferred embodiment of the present invention, in the construction of the underdetermined equation, unlike the standard solution process of CS, for the equation (2), we aim to find out
Figure BDA0003536138450000061
Rather than an exact reconstruction of the peak position
Figure BDA0003536138450000062
Therefore, it is not necessary to apply a complete recovery algorithm, and simultaneously
Figure BDA0003536138450000063
The sparsity of the signal is not necessarily critical.
In the embodiment of the present invention, as shown in fig. 3, as a preferred embodiment of the present invention, in the construction of the underdetermined equation, a is screened out according to the inner product maximum principle in the matching pursuit algorithmEXTNeutralization of
Figure BDA0003536138450000064
The most similar column.
In the embodiment of the present invention, as shown in fig. 4-5, as a preferred embodiment of the present invention, in the output of the vital sign data, in the actual calculation process, since the "respiration" frequency band and the "heartbeat" frequency band only correspond to the a frequency bandEXT1A small number of columns, so that the inner product comparison actually only needs to be performed in the two frequency bands respectively.
In the embodiment of the present invention, as shown in fig. 6 to 8, as a preferred embodiment of the present invention, the maximum inner product in two frequency bands (assumed to be respectively denoted as a) is followedb EXT1、Ah EXT1) Is selected, corresponding to
Figure BDA0003536138450000065
Two peak positions inAnd is thus determined.
Firstly, a two-dimensional array is constructed and used for storing ADC sampling data of radar beat signals, and a radio frequency module is made to transmit Chirp signals at a low frame rate (the existing FMCW radar sign monitoring system generally transmits at least 20 frames per second, the invention only needs to select a small number of frames from the existing FMCW radar sign monitoring system to transmit, as shown in figure 1), and the distance unit where a human body target is located is obtained by adopting conventional processing means such as Range-FFT and the like. Next, to increase the phase signal at the right end of the underdetermined equation (i.e., in equation (2))
Figure BDA0003536138450000066
) The accuracy of the method and the method avoid errors of a common unwrapping method, a binary iteration idea can be adopted to perform a small amount of frequency interpolation on a distance unit where a human body is located, and therefore the actual frequency of each beat signal is better approximated; extracting corresponding phase information based on the interpolation frequencies to form a phase signal for sign detection
Figure BDA0003536138450000067
Then, the phase signal is subjected to difference and band-pass filtering (the pass-band ranges are 0.1-0.5Hz and 0.8-2Hz, respectively, and it is assumed that the two filtered signals are respectively recorded as
Figure BDA0003536138450000071
And
Figure BDA0003536138450000072
) And meanwhile, 2 column vector groups corresponding to a respiratory frequency band and a heartbeat frequency band in the partial Fourier matrix are respectively constructed. Finally, by inner product comparison, two vector groups are screened out and respectively compared with
Figure BDA0003536138450000073
Figure BDA0003536138450000074
And (5) calculating the respiration and heart rate values of the tested human body by applying the column with the maximum inner product to obtain the corresponding spectral peak position. A program flow diagram of the algorithm is shown in fig. 2.
In order to verify the effectiveness of the algorithm, an FMCW radar sign monitoring system is constructed based on a TI-IWR 1443 millimeter wave sensor chip, an algorithm program is written and burned, and the test is performed by taking two human body targets located at different distance units as an example. In the experiment, in order to form comparison with the calculation result of the existing common method, the Chirp signal is still set to be transmitted at intervals of 50ms (namely, the frame rate is 20 Hz). When the underdetermined equation is adopted for solving, parts are randomly selected from the transmitted Chirp to participate in calculation. Firstly, taking a plurality of continuous frames (1024 frames are taken as an example), carrying out phase extraction, unwrapping, differentiating and filtering on the target 1 on the basis of Range-FFT, comparing the peak frequency of a 'respiratory' frequency band (0.1-0.5Hz) and a 'heartbeat' frequency band (0.8-2Hz) obtained when different Chirp numbers are taken by an underdetermined equation solution model, and obtaining a calculation result through 1024-point FFT, wherein relative error curves of the two are shown in fig. 4 and 5. As can be seen from fig. 4 and 5, when the Chirp number reaches about 300 (i.e., the decimation ratio 300/1024 ≈ 30%), the peak positions of the two-band spectrum captured based on the underdetermined equation are stably consistent with the result obtained by the conventional FFT. Next, still for the human target 1, the frame decimation ratio is fixed to 30% (i.e. corresponding to the emission frame rate of 6Hz), the underdetermined equation is examined to solve the calculation effect of the model in a continuous time (51 s as an example), and the result is compared with the peak frequency variation curve obtained by the conventional FFT method at the frame rate of 20Hz, and is shown in fig. 6. As can be seen from FIG. 6, the respiratory rate curve and the heartbeat rate curve obtained by both methods are consistent with each other. Further, to improve the measurement accuracy, frequency interpolation is performed on a distance unit where a human body is located on the basis of Range-FFT, meanwhile, a coefficient matrix in an underdetermined equation solution model is adjusted from a partial fourier matrix to a partial CZT matrix (i.e., a partial row is arbitrarily extracted on the basis of a general CZT matrix, and then column normalization is performed), a target 2 is taken as a detection object, then, 1024 frames are continuously taken, spectral peak positions respectively determined by a new method (the decimation ratio is still set to be 30%, i.e., 1024 × 30% ≈ 307 frames are selected) and a traditional CZT are compared, and the result is shown in fig. 7 and fig. 8. As can be seen from fig. 7 and 8, after the introduction of the spectrum refinement, the spectral peak position determined by the new method is still accurate.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents and improvements made within the spirit and principle of the present invention are intended to be included within the scope of the present invention.
Furthermore, it should be understood that although the present description refers to embodiments, not every embodiment may contain only a single embodiment, and such description is for clarity only, and those skilled in the art should integrate the description, and the embodiments may be combined as appropriate to form other embodiments understood by those skilled in the art.

Claims (7)

1. A millimeter wave vital sign detection method introducing a compressed sensing theory is characterized by comprising the following steps:
1) constructing an underdetermined equation; recording the phase signal as
Figure FDA0003536138440000011
Having a frequency spectrum of
Figure FDA0003536138440000012
The respiration and heart rate values of the measured human body are respectively determined by
Figure FDA0003536138440000013
Determining the peak positions at 0.1-0.5Hz and 0.8-2 Hz; the model is based on
Figure FDA0003536138440000014
And
Figure FDA0003536138440000015
the following matrix equation is satisfied:
Figure FDA0003536138440000016
wherein A is an inverse Fourier transform base, and part of the lines of A is selected and is marked as AEXTSimultaneously to each other
Figure FDA0003536138440000017
The same drawing is also performed
Figure FDA0003536138440000018
Then there is
Figure FDA0003536138440000019
2) Determining the position of a spectral peak; considering AEXTThe module values of all the columns are different, firstly, every column is normalized to form partial Fourier matrix and said partial Fourier matrix is recorded as AEXT1The ith column is marked as Ai EXT1And then the following inner product comparison and screening are carried out:
Figure FDA00035361384400000110
3) outputting sign data; suppose Ab EXT1And Ah EXT1Corresponding spectral peaks are respectively located
Figure FDA00035361384400000113
The values of the respiration and heart rate of the measured human body at the b-th and h-th positions under the frequency spectrum resolution delta f are
fb_BPM＝60(b-1)·Δf (4)
fh_BPM＝60(h-1)·Δf (5)。
2. The millimeter wave vital sign detection method introducing compressed sensing theory according to claim 1, wherein in the step 1) of constructing the underdetermined equation, from the perspective of CS, the equation (2) can be regarded as a pair
Figure FDA00035361384400000111
Several observations are carried out with AEXTAs an observation matrix, a matrix is used,
Figure FDA00035361384400000112
namely the observation result.
3. The millimeter wave vital sign detection method introducing compressed sensing theory according to claim 1, wherein in the step 1) of constructing the underdetermined equation, A is one of CZT basis and DWT basis.
4. The millimeter wave vital sign detection method introducing compressed sensing theory according to claim 1, wherein in the step 1) of constructing the underdetermined equation, for the equation (2), finding out the unknown vital sign is performed
Figure FDA0003536138440000021
The peak position of (c).
5. The millimeter wave vital sign detection method introduced with the compressed sensing theory according to claim 4, wherein in the step 1), according to the principle of maximum inner product in the matching pursuit algorithm, the uncertain equation A is screened outEXTNeutralization of
Figure FDA0003536138440000022
The most similar column.
6. The millimeter wave vital sign detection method based on introduction of compressed sensing theory according to claim 1, wherein in the step 2) of determining the spectral peak position, the inner product comparison is actually performed only by being limited to two frequency bands, namely a "respiration" frequency band and a "heartbeat" frequency band.
7. The millimeter wave vital sign detection method based on compressed sensing theory according to claim 6, wherein the method is applied to the millimeter wave vital sign detection method based on compressed sensing theoryStep 2) in the determination of the spectrum peak position, the maximum inner array A in two frequency bands of a respiration frequency band and a heartbeat frequency bandb EXT1、Ah EXT1Of a selection of, corresponding to
Figure FDA0003536138440000023
Two peak positions in (a) are determined.
CN202210219024.0A 2022-03-08 2022-03-08 Millimeter wave vital sign detection method introducing compressed sensing theory Pending CN114680848A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210219024.0A CN114680848A (en) 2022-03-08 2022-03-08 Millimeter wave vital sign detection method introducing compressed sensing theory

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210219024.0A CN114680848A (en) 2022-03-08 2022-03-08 Millimeter wave vital sign detection method introducing compressed sensing theory

Publications (1)

Publication Number Publication Date
CN114680848A true CN114680848A (en) 2022-07-01

Family

ID=82137862

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210219024.0A Pending CN114680848A (en) 2022-03-08 2022-03-08 Millimeter wave vital sign detection method introducing compressed sensing theory

Country Status (1)

Country Link
CN (1) CN114680848A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116831540A (en) * 2023-07-10 2023-10-03 康力元(天津)医疗科技有限公司 Millimeter wave-based non-contact vital sign monitoring method and system

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116831540A (en) * 2023-07-10 2023-10-03 康力元(天津)医疗科技有限公司 Millimeter wave-based non-contact vital sign monitoring method and system

Similar Documents

Publication Publication Date Title
Hamilton et al. Machine learning for rapid magnetic resonance fingerprinting tissue property quantification
Chang et al. DeepHeart: a deep learning approach for accurate heart rate estimation from PPG signals
CN107468250B (en) Biological tissue terahertz imaging method, system and equipment based on multi-scale entropy
Güler et al. A modified mixture of experts network structure for ECG beats classification with diverse features
Saxena et al. QRS detection using new wavelets
Yin et al. Pattern identification of biomedical images with time series: Contrasting THz pulse imaging with DCE-MRIs
Yılmaz et al. Analysis of the Doppler signals using largest Lyapunov exponent and correlation dimension in healthy and stenosed internal carotid artery patients
Humeau-Heurtier et al. Refined multiscale Hilbert–Huang spectral entropy and its application to central and peripheral cardiovascular data
Mohsen et al. AI aided noise processing of spintronic based IoT sensor for magnetocardiography application
Aravind Kumar et al. Efficient FPGA‐based VLSI architecture for detecting R‐peaks in electrocardiogram signal by combining Shannon energy with Hilbert transform
CN110554384A (en) imaging method based on microwave signal
CN114680848A (en) Millimeter wave vital sign detection method introducing compressed sensing theory
Zhuang et al. Accurate contactless sleep apnea detection framework with signal processing and machine learning methods
Guo et al. Wrist pulse signal acquisition and analysis for disease diagnosis: A review
Suhail et al. Cardiac disease detection from ECG signal using discrete wavelet transform with machine learning method
Liu et al. Precise and efficient heartbeat classification using a novel lightweight-modified method
Kupka et al. New method for beat-to-beat fetal heart rate measurement using Doppler ultrasound signal
Shin et al. Lightweight Ensemble Network for detecting heart disease using ECG signals
Oh et al. A neural framework for multi-variable lesion quantification through b-mode style transfer
Quanyu Pulse signal analysis based on deep learning network
Idaji et al. Nonlinear interaction decomposition (NID): A method for separation of cross-frequency coupled sources in human brain
El Gueddari et al. Calibration-less multi-coil compressed sensing magnetic resonance image reconstruction based on OSCAR regularization
Darmawahyuni et al. Accurate Fetal QRS-Complex Classification from Abdominal Electrocardiogram Using Deep Learning
Pan et al. A spectrum estimation approach for accurate heartbeat detection using Doppler radar based on combination of FTPR and TWV
CN115500817A (en) Non-contact type heart-shake signal detection method and device integrating millimeter wave radar and deep learning model

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination