Vital sign detection signal denoising method and apparatus

Abstract

The present application provides a kind of vital signs detection signal denoising method and device, it is related to physiological perception and signal processing technical field, including: based on millimeter wave radar's Doppler technique, obtains the original phase signal containing physiological signal;Original phase signal is carried out variational mode decomposition based on the search algorithm of variational mode decomposition super parameter optimization algorithm optimized by particle swarm optimization algorithm, obtain the vibration modal function information corresponding to the original phase signal;Wherein, the particle swarm optimization algorithm uses permutation entropy and fuzzy entropy as fitness function;After denoising processing is carried out to the vibration modal function, the vibration modal function after denoising processing is recombined, and high-precision denoising processing is obtained after physiological signal;From the high-precision denoising processing physiological signal, extract respiratory and heartbeat physiological signal.

Description

Technical Field

[0001] This invention relates to the field of physiological sensing and signal processing technology, and in particular to a method and apparatus for denoising vital sign detection signals. Background Technology

[0002] In recent years, biomedical radar technology has been widely used in many fields. It can sense the number, location, and state of living organisms and obtain physiological signals that reflect vital signs. It has great development potential in medical diagnosis, health monitoring, psychological assessment, scientific research and other fields.

[0003] As a novel tool for monitoring vital signs, millimeter-wave radar can monitor minute physiological activities in the human body, such as respiration and heartbeat, in a non-contact manner. It offers advantages such as real-time monitoring of physiological states without interfering with the monitored subject, comfort, avoidance of cross-infection, and suitability for patients with skin diseases, infants, and other applications. However, considering the weak nature of human respiratory and heartbeat signals and the complexity of environmental factors, bio-radio frequency signals face difficulties in data processing, including low signal-to-noise ratio and susceptibility to interference. Background noise reflections, interference from electronic devices, and the movement of organisms can all interfere with the detection signals. Therefore, effectively denoising vital sign detection signals has become a pressing issue for the industry. Summary of the Invention

[0004] This invention provides a method and apparatus for denoising vital sign detection signals, which addresses the urgent problem in the industry of how to effectively denoise vital sign detection signals.

[0005] This invention provides a method for denoising vital sign detection signals, comprising:

[0006] Based on Doppler technology of millimeter-wave radar, the raw phase signal containing physiological signals is obtained;

[0007] A variational mode decomposition hyperparameter search algorithm based on particle swarm optimization is used to perform variational mode decomposition on the original phase signal to obtain the vibration mode function information corresponding to the original phase signal; wherein, the particle swarm optimization algorithm uses permutation entropy and fuzzy entropy as fitness functions; after denoising the vibration mode function, the denoised vibration mode function is recombined to obtain a high-precision denoised physiological signal.

[0008] Respiratory and heartbeat physiological signals are extracted from the physiological signals after high-precision noise reduction processing.

[0009] According to the present invention, a method for denoising vital sign detection signals, wherein the Doppler technology based on millimeter-wave radar is used to acquire the original phase signal containing physiological signals, comprising:

[0010] The raw phase data containing physiological signals was acquired using millimeter-wave radar to obtain two-dimensional intermediate frequency signals along the fast time dimension and along the slow time dimension.

[0011] A one-dimensional Fourier transform is performed on the intermediate frequency signal along the fast time dimension to obtain a distance-amplitude map. The distance-amplitude map is then averaged along the slow time dimension to obtain static background information.

[0012] Based on the static background information, the distance-amplitude map is subjected to background filtering to obtain a background-filtered distance-amplitude map;

[0013] Phase unwrapping processing is performed on the phase signals with energy exceeding a preset threshold in the distance-amplitude map after background filtering to obtain the original phase signals containing physiological signals.

[0014] According to the present invention, a method for denoising vital sign detection signals employs a variational mode decomposition hyperparameter search algorithm based on particle swarm optimization to perform variational mode decomposition on the original phase signal, thereby obtaining vibration mode function information corresponding to the original phase signal, including:

[0015] Initialize hyperparameters, with each particle representing a set of variational mode decomposition parameters;

[0016] The original phase signal is decomposed using the variational mode decomposition parameters represented by each particle, and the complexity of the decomposed original phase signal is calculated using permutation entropy and fuzzy entropy. The complexity is then used as the fitness score of the particle.

[0017] Based on the fitness score, the variational mode decomposition parameters corresponding to each particle are updated according to the particle swarm parameter iterative algorithm. The original phase signal is then decomposed based on the updated particle swarm until the preset conditions are met, and the optimal variational mode decomposition parameters are obtained.

[0018] Based on the optimal variational mode decomposition parameters, the original phase signal is decomposed to obtain the vibration mode function information corresponding to the original phase signal.

[0019] According to a method for denoising vital sign detection signals provided by the present invention, after denoising the vibration mode functions, the denoised vibration mode functions are recombined to obtain a high-precision denoised physiological signal, including:

[0020] The complexity of each vibration mode function is calculated using permutation entropy and fuzzy entropy;

[0021] Vibration mode functions with complexity in the first preset range are considered as primary information, and vibration mode functions with complexity in the second preset range are considered as noise vibration mode functions;

[0022] The noise vibration mode function is denoised to obtain the denoised noise vibration mode function;

[0023] The vibration mode functions corresponding to the main information are combined in their original order, and the noise vibration mode functions after denoising are also combined in their original order. The two combinations are then superimposed to obtain the physiological signal after high-precision denoising.

[0024] According to the present invention, a method for denoising vital sign detection signals extracts respiratory and cardiac physiological signals from the high-precision denoised physiological signals, including:

[0025] The signal whose center frequency after high-precision noise reduction is within the first frequency range is selected as respiratory physiological data.

[0026] The center frequency of the physiological signal after high-precision denoising processing is selected as the heartbeat physiological data.

[0027] The present invention also provides a denoising device for vital sign detection signals, comprising:

[0028] The acquisition module is used to acquire raw phase signals containing physiological signals based on Doppler technology of millimeter-wave radar;

[0029] The decomposition module is used to perform variational mode decomposition on the original phase signal using a variational mode decomposition hyperparameter search algorithm optimized by particle swarm optimization, so as to obtain the vibration mode function information corresponding to the original phase signal; wherein, the particle swarm optimization algorithm uses permutation entropy and fuzzy entropy as fitness functions;

[0030] The denoising module is used to denoise the vibration mode functions and then recombine the denoised vibration mode functions to obtain a high-precision denoised physiological signal.

[0031] The detection module is used to extract respiratory and heartbeat physiological signals from the physiological signals after high-precision noise reduction processing.

[0032] According to the present invention, a denoising device for vital sign detection signals is provided, the device is further used for:

[0033] The raw phase data containing physiological signals was acquired using millimeter-wave radar to obtain two-dimensional intermediate frequency signals along the fast time dimension and along the slow time dimension.

[0034] A one-dimensional Fourier transform is performed on the intermediate frequency signal along the fast time dimension to obtain a distance-amplitude map. The distance-amplitude map is then averaged along the slow time dimension to obtain static background information.

[0035] Based on the static background information, the distance-amplitude map is subjected to background filtering to obtain a background-filtered distance-amplitude map;

[0036] Phase unwrapping processing is performed on the phase signals with energy exceeding a preset threshold in the distance-amplitude map after background filtering to obtain the original phase signals containing physiological signals.

[0037] According to the present invention, a denoising device for vital sign detection signals is provided, the device is further used for:

[0038] Initialize hyperparameters, with each particle representing a set of variational mode decomposition parameters;

[0039] The original phase signal is decomposed using the variational mode decomposition parameters represented by each particle, and the complexity of the decomposed original phase signal is calculated using permutation entropy and fuzzy entropy. The complexity is then used as the fitness score of the particle.

[0040] Based on the fitness score, the variational mode decomposition parameters corresponding to each particle are updated according to the particle swarm parameter iterative algorithm. The original phase signal is then decomposed based on the updated particle swarm until the optimization process converges, and the optimal variational mode decomposition parameters are obtained.

[0041] Based on the optimal variational mode decomposition parameters, the original phase signal is decomposed to obtain the vibration mode function information corresponding to the original phase signal.

[0042] According to the present invention, a denoising device for vital sign detection signals is provided, the device is further used for:

[0043] The complexity of each vibration mode function is calculated using permutation entropy and fuzzy entropy;

[0044] Vibration mode functions with complexity in the first preset range are considered as primary information, and vibration mode functions with complexity in the second preset range are considered as noise vibration mode functions;

[0045] The noise vibration mode function is denoised to obtain the denoised noise vibration mode function;

[0046] The vibration mode functions corresponding to the main information are combined in their original order, and the noise vibration mode functions after denoising are also combined in their original order. The two combinations are then superimposed to obtain the physiological signal after high-precision denoising.

[0047] According to the present invention, a denoising device for vital sign detection signals is provided, the device is further used for:

[0048] The signal whose center frequency after high-precision noise reduction is within the first frequency range is selected as respiratory physiological data.

[0049] The center frequency of the physiological signal after high-precision denoising processing is selected as the heartbeat physiological data.

[0050] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the denoising method for vital sign detection signals as described above.

[0051] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the denoising method for vital sign detection signals as described above.

[0052] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the denoising method for vital sign detection signals as described above.

[0053] This invention provides a method and apparatus for denoising vital sign detection signals. It employs the VMD algorithm to decompose and denoise physiological signals acquired by millimeter-wave radar. Since the chest wall displacement signal acquired by millimeter-wave radar is composed of multiple superimposed components, VMD can effectively decompose each mode, avoiding mode aliasing and thus efficiently restoring the true physiological signal information. Specifically, a particle swarm optimization algorithm is used to optimize hyperparameters such as the number of modes and bandwidth penalty factor in VMD to improve the denoising effect. Fuzzy entropy and permutation entropy are used as fitness functions of the particle swarm optimization algorithm to measure the complexity and physiological information content of different modes, thereby optimizing the performance of variational mode decomposition. The heuristic search of the PSO algorithm avoids the computational overhead of brute-force search for optimal hyperparameters, accelerating the convergence speed of the algorithm. Each decomposed mode is smoothed and denoised, and modes of interest in different frequency bands of respiration and heartbeat are selected and combined to accurately restore physiological signals, effectively achieving denoising of vital sign detection signals. Attached Figure Description

[0054] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0055] Figure 1 This is a schematic flowchart of a method for denoising vital sign detection signals provided in an embodiment of this application;

[0056] Figure 2 This is a denoising diagram provided for an embodiment of this application;

[0057] Figure 3 This is a schematic diagram of the structure of the vital signs detection signal denoising device provided in the embodiments of this application;

[0058] Figure 4 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0059] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0060] Figure 1 This is a schematic flowchart of the vital sign detection signal denoising method provided in the embodiments of this application, as shown below. Figure 1 As shown, it includes:

[0061] Step 110: Based on the Doppler technology of millimeter-wave radar, acquire the raw phase signal containing physiological signals;

[0062] In this embodiment, a millimeter-wave radar device is first configured to ensure its ability to transmit and receive electromagnetic waves of a specific frequency. The radar transmitter emits millimeter-wave signals, which are linearly frequency modulated, with the frequency increasing linearly over time.

[0063] Radar receivers capture signals reflected from the human body. Due to the Doppler effect, the frequency of the reflected signal changes with minute movements of the human body (such as breathing and heartbeat).

[0064] The received signal is converted into a digital format for further analysis and processing. A Fourier transform is performed on the digital signal, converting it from the time domain to the frequency domain to facilitate analysis of the signal's frequency components. Analyzing the signal in the frequency domain of the linear frequency modulated radar enables the localization of the human body. Subsequently, phase information is extracted from the signals across various frequency ranges, which is crucial for monitoring physiological activities. The raw phase data will include phase changes caused by respiration and heartbeat.

[0065] Optionally, the Doppler technology based on millimeter-wave radar acquires the raw phase signal containing physiological signals, including:

[0066] The raw phase data containing physiological signals was acquired using millimeter-wave radar to obtain two-dimensional intermediate frequency signals along the fast time dimension and along the slow time dimension.

[0067] A one-dimensional Fourier transform is performed on the intermediate frequency signal along the fast time dimension to obtain a distance-amplitude map. The static background information is obtained by averaging the distance-amplitude map along the slow time dimension.

[0068] Based on the static background information, the distance-amplitude map is filtered to remove background interference, resulting in an intermediate frequency signal with reduced background reflection interference.

[0069] Phase unwrapping processing is performed on the phase signals with energy exceeding a preset threshold in the distance-amplitude map after background filtering to obtain the original phase signals containing physiological signals.

[0070] In this embodiment, the radar signal, after processing, yields a two-dimensional intermediate frequency signal along two time dimensions. The fast time dimension corresponds to each sampling point within a single pulse of the radar signal, while the slow time dimension corresponds to the sampling time between radar pulse signals.

[0071] Performing a one-dimensional Fourier transform (Range-FFT) on the intermediate frequency signal in the fast time dimension transforms the signal into the distance-frequency domain, resulting in a distance-amplitude plot. This plot illustrates the amplitude information of the signal at different distances.

[0072] A one-dimensional Fourier transform is performed on the intermediate frequency signal in the slow time dimension, and the average is calculated to obtain static background information. This helps to identify and remove non-physiological variations in the signal, such as environmental noise or stationary noise sources from equipment.

[0073] Background filtering is performed on the distance-amplitude map using static background information. This step removes signal components that do not change over time by subtracting static background information from the distance-amplitude map, thereby highlighting dynamic signals related to human physiological activities.

[0074] In the distance-amplitude map after background filtering, phase signals with energy exceeding a preset threshold are identified. These signals contain most of the physiological information, such as respiration and heartbeat.

[0075] These signals undergo phase unwrapping. Since phase signals may experience discontinuous transitions (e.g., from -π to π), unwrapping adjusts these transitions to make the phase signal continuous, facilitating subsequent algorithm processing. After phase unwrapping, the original phase signal containing the physiological signal is finally obtained.

[0076] In an optional embodiment, the phase of the complex domain signal is calculated and demodulated using the arctan function. The resulting phase signal ranges from -π to π, with abrupt changes occurring at the boundaries. Accordingly, a phase unwinding algorithm is used for continuity processing; that is, for adjacent phase values, if an amplitude abrupt change exceeding π occurs, ±2π compensation is applied until the abrupt change disappears. The formula is as follows:

[0077]

[0078]

[0079] Step 120: The original phase signal is subjected to variational mode decomposition by a variational mode decomposition hyperparameter search algorithm optimized by particle swarm optimization, to obtain the vibration mode function information corresponding to the original phase signal; wherein, the particle swarm optimization algorithm uses permutation entropy and fuzzy entropy as fitness functions;

[0080] In the embodiments of this application, a set of particles is randomly generated, each particle representing a set of hyperparameters in the VMD algorithm, including the number of mode decompositions (K) and the frequency band penalty parameter (α).

[0081] Variational Mode Decomposition (VMD) decomposes the original signal into a series of Intrinsic Mode Functions (IMFs) through iterative optimization, minimizing the reconstruction error. The center frequency and frequency domain representation of each IMF are iteratively optimized, as shown in the following formula.

[0082]

[0083]

[0084] More specifically, the original phase signal is decomposed using hyperparameters represented by each particle using VMD decomposition. The VMD algorithm decomposes the original signal into a series of vibrational mode functions (IMFs) through iterative optimization, with each IMF representing a specific frequency component of the signal.

[0085] Calculate the fitness value for each particle. The fitness function consists of permutation entropy (PE) and fuzzy entropy (FE), used to evaluate the quality of the decomposition results.

[0086] Permutation entropy (PE) measures the randomness and complexity of a signal and helps identify noise components in the signal. Fuzzy entropy (FE) considers the uncertainty and ambiguity of a signal and helps distinguish between information and noise in the signal.

[0087] In an optional embodiment, fuzzy entropy is a method for measuring the complexity of a time series, quantifying the uncertainty or randomness of a signal, and taking into account the fuzziness of the signal, making it more applicable to complex and noisy signals. The calculation method is as follows:

[0088] Given a time series x(i), construct a fuzzy vector X. i =[x(i),x(i+1),…,x(i+m-1)];

[0089] For each fuzzy vector X i The Euclidean distance d between them ij =||X i -X j ||2;

[0090] For a given threshold r, define the fuzzy function C. r (i) indicates that X i The number of distances less than the threshold; and the normalized fuzzy probability.

[0091] The formula for fuzzy entropy is:

[0092] Permutation entropy is a method used to analyze the complexity and irregularity of time series. It measures the uncertainty of the series by converting the time series into a permutation sequence and calculating the frequency of different permutations. It is commonly used to explore patterns and structures in time series. The calculation method is as follows:

[0093] Given a time series x(i), use a sliding window of length m to generate a subsequence;

[0094] Sort the elements in each subsequence by size and record the sorting method;

[0095] Calculate the frequency p of each permutation i And it uses information entropy to measure the complexity of the sequence.

[0096] The formula for calculating permutation entropy is:

[0097] The fitness of each particle is evaluated based on the values ​​of PE and FE. The fitness of a particle reflects the contribution of the hyperparameter it represents to the signal decomposition effect. The position and velocity of each particle are updated based on the fitness evaluation results. The particle position update is based on historical information of the individual best (pbest) and global best (gbest), as well as the particle's current velocity.

[0098] In an optional embodiment, the particles are updated by updating the position and velocity of each particle based on its fitness score, i.e., adjusting the VMD parameters. The update formula is as follows:

[0099] v i (t+1)=wv i (t)+c1r1(pbest-x i (t))+c2r2(gbest-x i (t))

[0100] x i (t+1)=x i (t+1)+v i (t+1);

[0101] Repeat the VMD decomposition to particle update steps described above until the predetermined number of iterations is reached or the convergence condition is met. In each iteration, the particle swarm continuously explores the parameter space to find the optimal combination of hyperparameters.

[0102] After the iteration, the hyperparameter represented by the particle with the highest fitness value is selected as the optimal solution. These optimal hyperparameters are then used to perform a final VMD decomposition on the original phase signal to obtain the vibrational mode function information corresponding to the original phase signal.

[0103] Step 130: After denoising the vibration mode function, the denoised vibration mode function is recombined to obtain a high-precision denoised physiological signal.

[0104] Optionally, after denoising the vibration mode functions, the denoised vibration mode functions are recombined to obtain a high-precision denoised physiological signal, including:

[0105] The complexity of each vibration mode function is calculated using permutation entropy and fuzzy entropy;

[0106] Vibration mode functions with complexity in the first preset range are considered as primary information, and vibration mode functions with complexity in the second preset range are considered as noise vibration mode functions;

[0107] The noise vibration mode function is denoised to obtain the denoised noise vibration mode function;

[0108] The vibration mode functions corresponding to the main information are combined in their original order, and the noise vibration mode functions after denoising are also combined in their original order. The two combinations are then superimposed to obtain the physiological signal after high-precision denoising.

[0109] In this embodiment, permutation entropy (PE) and fuzzy entropy (FE) are used as metrics to calculate the complexity of each vibration mode function. These entropy values ​​reflect the randomness, complexity, and information content of the IMFs.

[0110] IMFs with complexity values ​​within a first preset range are identified as primary information IMFs, containing key information about the physiological signal. IMFs with complexity values ​​within a second preset range are identified as noisy IMFs, primarily composed of noise.

[0111] Denoising is performed on IMFs identified as noise. This may include applying filters, smoothing, thresholding, or other signal processing techniques to reduce noise components. IMFs containing key physiological information are also combined in the order they appear in the original signal.

[0112] In this embodiment, the main information IMFs combined after denoising and the noise IMFs combined are superimposed. This step merges the information from the two parts to reconstruct the original signal.

[0113] By superimposing the signals, a highly accurate denoised physiological signal is obtained. This signal should have a low noise level while retaining important physiological information.

[0114] Step 140: Extract respiratory and heartbeat physiological signals from the physiological signals after high-precision noise reduction processing.

[0115] Optionally, respiratory and cardiac physiological signals are extracted from the high-precision denoised physiological signals, including:

[0116] The signal whose center frequency after high-precision noise reduction is within the first frequency range is selected as respiratory physiological data.

[0117] The center frequency of the physiological signal after high-precision denoising processing is selected as the heartbeat physiological data.

[0118] In this embodiment of the application, frequency analysis is performed on the physiological signal after high-precision denoising processing to determine the distribution of different frequency components in the signal.

[0119] Two frequency ranges are defined: the first frequency range corresponds to the respiratory rate (typically between 0.1 Hz and 0.5 Hz, corresponding to the normal adult respiratory rate), and the second frequency range corresponds to the heart rate (typically between 0.8 Hz and 2.5 Hz, corresponding to the normal adult heart rate). Using bandpass filters, the respiratory and heart rate-related signal components are filtered out from the denoised physiological signal for both the first and second frequency ranges.

[0120] Signals with center frequencies within the first frequency range are selected, as this portion reflects respiratory activity. Further analysis of this signal, such as peak detection and periodicity analysis, is performed to extract physiological parameters such as respiratory rate and respiratory depth.

[0121] Signals with a center frequency within the second frequency range are selected, as this portion of the signal reflects heartbeat activity. Peak detection is performed on this portion of the signal to identify the peak points of the heartbeats, thereby calculating the heart rate (number of heartbeats per minute).

[0122] The filtered signal is smoothed to reduce residual noise and interference. Statistical analysis or time-frequency analysis methods are applied to extract features from the respiratory and heartbeat signals, such as respiratory rhythm and heart rate variability.

[0123] In this embodiment, the VMD algorithm is used to decompose and denoise the physiological signals acquired by millimeter-wave radar. Since the chest wall displacement signal acquired by millimeter-wave radar is composed of multiple superimposed components, VMD can effectively decompose each mode, avoiding mode aliasing problems, thereby efficiently restoring the true physiological signal information. Specifically, the particle swarm optimization algorithm is used to optimize the hyperparameters of VMD, such as the number of modes and the bandwidth penalty factor, to improve the denoising effect. The heuristic search of the PSO algorithm avoids the computational overhead of brute-force search for optimal hyperparameters, accelerating the convergence speed of the algorithm. The decomposed modes are smoothed and denoised, and modes of different frequency bands of interest, such as respiration and heartbeat, are selected and combined to restore the physiological signals with high precision, effectively realizing the denoising of vital sign detection signals.

[0124] Figure 2 This is a denoising diagram provided for an embodiment of this application, such as... Figure 2 As shown, it includes:

[0125] The raw phase data containing physiological signals is acquired using Doppler technology based on millimeter-wave radar. Then, the VMD hyperparameter particle swarm is initialized, and VMD iterative optimization is performed to calculate FE and PE. After the particle swarm is updated and the iteration is completed, the entropy of each IMF sample is calculated using the optimal VMD decomposition parameters, the noise IMF threshold is filtered, and finally the signal is reconstructed.

[0126] Figure 3 This is a schematic diagram of the structure of the vital sign detection signal denoising device provided in the embodiments of this application, as shown below. Figure 3 As shown, it includes:

[0127] The acquisition module 310 is used for Doppler technology based on millimeter-wave radar to acquire raw phase signals containing physiological signals;

[0128] The decomposition module 320 is used to perform variational mode decomposition on the original phase signal using a variational mode decomposition hyperparameter search algorithm optimized by particle swarm optimization, to obtain the vibration mode function information corresponding to the original phase signal; wherein, the particle swarm optimization algorithm uses permutation entropy and fuzzy entropy as fitness functions;

[0129] The denoising module 330 is used to denoise the vibration mode function and then recombine the denoised vibration mode function to obtain a high-precision denoised physiological signal.

[0130] The detection module 340 is used to extract respiratory and heartbeat physiological signals from the physiological signals after high-precision noise reduction processing.

[0131] The device is also used for:

[0132] The raw phase data containing physiological signals was acquired using millimeter-wave radar to obtain two-dimensional intermediate frequency signals along the fast time dimension and along the slow time dimension.

[0133] A one-dimensional Fourier transform is performed on the intermediate frequency signal along the fast time dimension to obtain a distance-amplitude map. The average of the intermediate frequency signal along the slow time dimension is then calculated to obtain static background information.

[0134] Based on the static background information, the distance-amplitude map is subjected to background filtering to obtain a background-filtered distance-amplitude map;

[0135] Phase unwrapping processing is performed on the phase signals with energy exceeding a preset threshold in the distance-amplitude map after background filtering to obtain the original phase signals containing physiological signals.

[0136] In this embodiment, the VMD algorithm is used to decompose and denoise the physiological signals acquired by millimeter-wave radar. Since the chest wall displacement signal acquired by millimeter-wave radar is composed of multiple superimposed components, VMD can effectively decompose each mode, avoiding mode aliasing problems, thereby efficiently restoring the true physiological signal information. Specifically, the particle swarm optimization algorithm is used to optimize the hyperparameters of VMD, such as the number of modes and the bandwidth penalty factor, to improve the denoising effect. The heuristic search of the PSO algorithm avoids the computational overhead of brute-force search for optimal hyperparameters, accelerates the convergence speed of the algorithm, and effectively achieves denoising of vital sign detection signals.

[0137] Figure 4 This is a schematic diagram of the structure of the electronic device provided by the present invention, such as... Figure 4 As shown, the electronic device may include a processor 410, a communications interface 420, a memory 430, and a communication bus 440. The processor 410, communications interface 420, and memory 430 communicate with each other via the communication bus 440. The processor 410 can call logical instructions from the memory 430 to execute a vital sign detection signal denoising method. This method includes: acquiring the original phase signal containing physiological signals based on Doppler technology from millimeter-wave radar.

[0138] A variational mode decomposition hyperparameter search algorithm based on particle swarm optimization is used to perform variational mode decomposition on the original phase signal to obtain the vibration mode function information corresponding to the original phase signal; wherein, the particle swarm optimization algorithm uses permutation entropy and fuzzy entropy as fitness functions;

[0139] After denoising the vibration mode functions, the denoised vibration mode functions are recombined to obtain a high-precision denoised physiological signal.

[0140] Respiratory and heartbeat physiological signals are extracted from the physiological signals after high-precision noise reduction processing.

[0141] Furthermore, the logical instructions in the aforementioned memory 430 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0142] On the other hand, the present invention also provides a computer program product, the computer program product including a computer program, the computer program being able to be stored on a non-transitory computer-readable storage medium, the computer program being executed by a processor, the computer being able to execute the vital sign detection signal denoising method provided by the above methods, the method including: acquiring the original phase signal containing physiological signals based on Doppler technology of millimeter-wave radar;

[0143] A variational mode decomposition hyperparameter search algorithm based on particle swarm optimization is used to perform variational mode decomposition on the original phase signal to obtain the vibration mode function information corresponding to the original phase signal; wherein, the particle swarm optimization algorithm uses permutation entropy and fuzzy entropy as fitness functions;

[0144] After denoising the vibration mode functions, the denoised vibration mode functions are recombined to obtain a high-precision denoised physiological signal.

[0145] Respiratory and heartbeat physiological signals are extracted from the physiological signals after high-precision noise reduction processing.

[0146] In another aspect, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a method for denoising vital sign detection signals provided by the methods described above, the method comprising: acquiring an original phase signal containing physiological signals based on Doppler technology of millimeter-wave radar;

[0147] A variational mode decomposition hyperparameter search algorithm based on particle swarm optimization is used to perform variational mode decomposition on the original phase signal to obtain the vibration mode function information corresponding to the original phase signal; wherein, the particle swarm optimization algorithm uses permutation entropy and fuzzy entropy as fitness functions;

[0148] After denoising the vibration mode functions, the denoised vibration mode functions are recombined to obtain a high-precision denoised physiological signal.

[0149] Respiratory and heartbeat physiological signals are extracted from the physiological signals after high-precision noise reduction processing.

[0150] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0151] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0152] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for denoising vital sign detection signals, characterized in that, include: Based on Doppler technology of millimeter-wave radar, the raw phase signal containing physiological signals is obtained; A variational mode decomposition hyperparameter search algorithm based on particle swarm optimization is used to perform variational mode decomposition on the original phase signal to obtain the vibration mode function information corresponding to the original phase signal; wherein, the particle swarm optimization algorithm uses permutation entropy and fuzzy entropy as fitness functions; After denoising the vibration mode functions, the denoised vibration mode functions are recombined to obtain a high-precision denoised physiological signal. Respiratory and cardiac physiological signals are extracted from the physiological signals after high-precision noise reduction processing; The variational mode decomposition hyperparameter search algorithm based on particle swarm optimization performs variational mode decomposition on the original phase signal to obtain the vibration mode function information corresponding to the original phase signal, including: Initialize hyperparameters, with each particle representing a set of variational mode decomposition parameters; The original phase signal is decomposed using the variational mode decomposition parameters represented by each particle, and the complexity of the decomposed original phase signal is calculated using permutation entropy and fuzzy entropy. The complexity is then used as the fitness score of the particle. Based on the fitness score, the variational mode decomposition parameters corresponding to each particle are updated according to the particle swarm parameter iterative algorithm. The original phase signal is then decomposed based on the updated particle swarm until the preset conditions are met, and the optimal variational mode decomposition parameters are obtained. Based on the optimal variational mode decomposition parameters, the original phase signal is decomposed to obtain the vibration mode function information corresponding to the original phase signal; The vibration mode functions are denoised, and then the denoised vibration mode functions are recombined to obtain a high-precision denoised physiological signal, including: The complexity of each vibration mode function is calculated using permutation entropy and fuzzy entropy; Vibration mode functions with complexity in the first preset range are considered as primary information, and vibration mode functions with complexity in the second preset range are considered as noise vibration mode functions; The noise vibration mode function is denoised to obtain the denoised noise vibration mode function; The vibration mode functions corresponding to the main information are combined in their original order, and the noise vibration mode functions after denoising are also combined in their original order. The two combinations are then superimposed to obtain the physiological signal after high-precision denoising.

2. The method for denoising vital sign detection signals according to claim 1, characterized in that, The Doppler technology based on millimeter-wave radar acquires raw phase signals containing physiological signals, including: The raw phase data containing physiological signals was acquired using millimeter-wave radar to obtain two-dimensional intermediate frequency signals along the fast time dimension and along the slow time dimension. A one-dimensional Fourier transform is performed on the intermediate frequency signal along the fast time dimension to obtain a distance-amplitude map. The distance-amplitude map is then averaged along the slow time dimension to obtain static background information. Based on the static background information, the distance-amplitude map is subjected to background filtering to obtain a background-filtered distance-amplitude map; Phase unwrapping processing is performed on the phase signals with energy exceeding a preset threshold in the distance-amplitude map after background filtering to obtain the original phase signals containing physiological signals.

3. The method for denoising vital sign detection signals according to claim 1, characterized in that, From the physiological signals after high-precision denoising processing, respiratory and cardiac physiological signals are extracted, including: The signal whose center frequency after high-precision noise reduction is within the first frequency range is selected as respiratory physiological data. The center frequency of the physiological signal after high-precision denoising processing is selected as the heartbeat physiological data.

4. A noise reduction device for vital sign detection signals, characterized in that, include: The acquisition module is used to acquire raw phase signals containing physiological signals based on Doppler technology of millimeter-wave radar; The decomposition module is used to perform variational mode decomposition on the original phase signal using a variational mode decomposition hyperparameter search algorithm optimized by particle swarm optimization, so as to obtain the vibration mode function information corresponding to the original phase signal; wherein, the particle swarm optimization algorithm uses permutation entropy and fuzzy entropy as fitness functions; The denoising module is used to denoise the vibration mode functions and then recombine the denoised vibration mode functions to obtain a high-precision denoised physiological signal. The detection module is used to extract respiratory and heartbeat physiological signals from the high-precision denoising processed physiological signals; The device is also used for: Initialize hyperparameters, with each particle representing a set of variational mode decomposition parameters; The original phase signal is decomposed using the variational mode decomposition parameters represented by each particle, and the complexity of the decomposed original phase signal is calculated using permutation entropy and fuzzy entropy. The complexity is then used as the fitness score of the particle. Based on the fitness score, the variational mode decomposition parameters corresponding to each particle are updated according to the particle swarm parameter iterative algorithm. The original phase signal is then decomposed based on the updated particle swarm until the preset conditions are met, and the optimal variational mode decomposition parameters are obtained. Based on the optimal variational mode decomposition parameters, the original phase signal is decomposed to obtain the vibration mode function information corresponding to the original phase signal; The device is also used for: The complexity of each vibration mode function is calculated using permutation entropy and fuzzy entropy; Vibration mode functions with complexity in the first preset range are considered as primary information, and vibration mode functions with complexity in the second preset range are considered as noise vibration mode functions; The noise vibration mode function is denoised to obtain the denoised noise vibration mode function; The vibration mode functions corresponding to the main information are combined in their original order, and the noise vibration mode functions after denoising are also combined in their original order. The two combinations are then superimposed to obtain the physiological signal after high-precision denoising.

5. The vital sign detection signal denoising device according to claim 4, characterized in that, The device is also used for: The raw phase data containing physiological signals was acquired using millimeter-wave radar to obtain two-dimensional intermediate frequency signals along the fast time dimension and along the slow time dimension. A one-dimensional Fourier transform is performed on the intermediate frequency signal along the fast time dimension to obtain a distance-amplitude map. The distance-amplitude map is then averaged along the slow time dimension to obtain static background information. Based on the static background information, the distance-amplitude map is subjected to background filtering to obtain a background-filtered distance-amplitude map; Phase unwrapping processing is performed on the phase signals with energy exceeding a preset threshold in the distance-amplitude map after background filtering to obtain the original phase signals containing physiological signals.

6. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the vital sign detection signal denoising method as described in any one of claims 1 to 3.

7. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the vital sign detection signal denoising method as described in any one of claims 1 to 3.

8. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the vital sign detection signal denoising method as described in any one of claims 1 to 3.

Images