Human heart rate variability and respiratory rate measurement method based on variational mode decomposition and constrained independent component analysis

Abstract

The present invention discloses a human heart rate variability (HRV) and respiratory rate (RR) measurement method based on variational mode decomposition (VMD) and constrained independent component analysis (cICA). The method performs pixel coherent averaging operation on human facial video data, converts it into RGB observation signals, and performs preprocessing operation on the RGB observation signals to achieve standardization. Then, the G channel signal is decomposed into 4 channels using the VMD algorithm, and the reference signal of the blood flow pulse (BVP) is calculated based on the component with the largest spectrum peak in the decomposed 4-channel components. Based on the reference signal, the BVP source signal is separated from the RGB observation signal using the cICA algorithm, and the BVP source signal is decomposed into 4 channels using the VMD algorithm, and the high-quality pulse wave component is extracted from the separated 4-channel components, and the HRV parameters and RR are further calculated. The present invention can avoid the inherent source sorting fuzzy problem in the traditional blind source separation / independent component analysis algorithm, and has good noise interference resistance, and has good application prospects in this field.

Description

Technical Field

[0001] The present invention relates to the research field of facial video physiological parameter extraction based on imaging photoplethysmography (IPPG) technology, including the extraction of heart rate variability (HRV) parameters and respiratory rate (RR), and in particular to a variational mode decomposition (VMD) algorithm and a constrained independent component analysis (cICA) algorithm. Background Art

[0002] Heart rate variability (HRV) and respiratory rate (RR) are important clinical physiological parameters of the human body. Exploring non-contact measurement methods for HRV and RR has become one of the hot topics in the current research field of biomedical engineering and instrumentation.

[0003] Imaging photoplethysmography (IPPG) is a biomedical signal processing technology that analyzes video data of sensitive areas on the human body surface through intelligent information processing algorithms to extract physiological parameters such as heart rate (HR), HRV, RR, etc. It has many advantages such as non-contact, easy operation, and broad application prospects.

[0004] Since the video data in IPPG technology usually depends on the visible light environment of the human body surface, the generated RGB observation signal and the obtained BVP component are usually interfered by complex background noise. This defect will further affect the extraction of various physiological parameters.

[0005] Most of the published patents or documents avoid or use different algorithms to overcome this problem. For example, the Chinese invention patent with publication number CN113657345A published on 2021-11-16, the Chinese invention patent with publication number CN112237421A published on 2021-01-19, and the Chinese invention patent with publication number CN107616795A published on 2018-01-23.

[0006] These video-based physiological parameter extraction schemes have tried different pattern recognition algorithms to extract target parameters, such as using adaptive threshold skin detection, LSTM convolutional network and other algorithms to extract HRV feature parameters, and extracting respiratory signal sequences from chest videos through video motion magnification technology.

[0007] In the existing IPPG application research, the independent component analysis (ICA) method has become an important idea due to its advantages in extracting blood flow pulse (BVP) source signals. The BVP source signal extracted based on ICA can be used to accurately extract HR data. This research method appears in a large number of published related literature and patents.

[0008] However, the residual noise of different intensities in the BVP source signal makes it ineffective for the extraction of physiological parameters such as HRV and RR, which have high requirements for the quality of BVP waveform. In recent years, a new adaptive signal processing method, variational mode decomposition, has begun to attract attention for its good application effect in non-stationary and nonlinear signal processing.

[0009] Through a lot of analysis, research and testing, the inventor team of the present invention found that the VMD method can decompose more ideal pulse wave components from RGB observation signals and BVP source signals under different noise interference environments, and further apply it to the extraction of target physiological parameters.

[0010] Secondly, the ICA algorithm has the problem of fuzzy source sorting when separating BVP source signals. In practical applications, it is necessary to design an additional BVP source signal recognition algorithm. In the published related literature, the BVP source discrimination is basically based on spectrum analysis and other methods, and its recognition accuracy is difficult to achieve ideal results in a complex noise interference environment. For example, the Chinese invention patent with a publication date of 2012-06-20 and a publication number of 102499664A, and the Chinese invention patent with a publication date of 2013-09-04 and a publication number of 103271734A.

[0011] The constrained independent component analysis (cICA) method can effectively avoid this problem. The algorithm introduces a reference signal that carries some key features of the target source signal, constrains the convergence direction of the spatial filter during the iterative calculation process, and directly obtains the BVP source signal, thereby discarding the additional BVP source identification algorithm. Summary of the invention

[0012] The purpose of the present invention is to provide a HRV and RR measurement method based on variational mode decomposition (VMD) and constrained independent component analysis (cICA), which synchronously extracts HRV parameters and RR from human facial videos, avoids the BVP source sorting ambiguity problem inherent in traditional methods, and overcomes the more difficult BVP source signal noise residual problem in physiological parameter extraction.

[0013] To this end, the present invention provides a method for measuring human HRV and RR based on variational mode decomposition (VMD) and constrained independent component analysis (cICA), including: S100, performing pixel coherent averaging operation on human facial video data to convert it into RGB observation signals; S101, performing preprocessing operations on the RGB observation signals to obtain standardized observation signals for subsequent analysis; S102, performing 4-channel decomposition on the G channel signal using the VMD algorithm, and generating a BVP reference signal based on the component with the largest spectrum peak in the decomposed 4-channel components; S103, based on the BVP reference signal, using the cICA algorithm to separate the BVP source signal from the RGB observation signal; S104, using the VMD algorithm to perform 4-channel decomposition on the BVP source signal, and extracting high-quality pulse wave components from the decomposed 4-channel components; and S105, obtaining HRV parameters based on the high-quality pulse wave components: low-frequency component power (LF), high-frequency component power (HF), power ratio of low-frequency component to high-frequency component (LF / HF), and RR.

[0014] According to another aspect of the present invention, a human HRV and RR measurement device based on variational mode decomposition (VMD) and constrained independent component analysis (cICA) is provided, comprising: S100, a program module 1 for performing pixel coherent averaging operation on human face video data to convert it into RGB observation signals; S101, a program module 2 for performing preprocessing operation on the RGB observation signals to obtain standardized observation signals for subsequent analysis; S102, a program module 2 for performing 4-channel decomposition on the G channel signal using the VMD algorithm, and obtaining the component with the largest spectrum peak value among the decomposed 4 channel components as the basis; S104, a program module for performing 4-channel decomposition of the BVP source signal using the VMD algorithm, and extracting high-quality pulse wave components from the decomposed 4-channel components; and S105, a program module for obtaining HRV parameters and RR based on the high-quality pulse wave components, wherein the HRV parameters include: low-frequency component power (LF), high-frequency component power (HF), and power ratio of low-frequency component to high-frequency component (LF / HF).

[0015] The present invention also provides a computer-readable storage medium storing a program, which, when executed, implements the steps of the human HRV and RR measurement method based on variational mode decomposition and constrained independent component analysis described above.

[0016] The present invention also provides a computer device, comprising a processor and a memory, characterized in that a program is stored on the memory, and when the program is executed on the processor, the steps of the human HRV and RR measurement method based on variational mode decomposition and constrained independent component analysis described above are implemented.

[0017] Compared with the prior art, the human heart rate variability (HRV) and respiratory rate (RR) measurement method based on variational mode decomposition (VMD) and constrained independent component analysis (cICA) of the present invention has the following characteristics.

[0018] 1. The present invention overcomes the thorny problem of BVP source signal noise residue in traditional research methods.

[0019] Based on a large amount of previous analysis and testing, the present invention applies VMD, a new adaptive signal processing method, to the processing of RGB observation signals and BVP source signals, which can decompose relatively ideal pulse wave components and further apply them to the extraction of HRV parameters and RR. Compared with the existing technology, the use of the VMD method can obtain pulse wave components with high-quality waveforms from the BVP source signal, overcoming the problem of residual noise in the BVP source signal that has always plagued the extraction of physiological parameters. In addition, the BVP reference signal obtained based on the G channel and the VMD method can carry accurate frequency and phase information, which helps to improve the robustness of the cICA algorithm. There is no related research published on these schemes in the present invention.

[0020] 2. The present invention effectively solves the BVP source sorting ambiguity problem in the solution using traditional independent component analysis.

[0021] Compared with the existing technology, the present invention adopts the constrained independent component analysis (cICA) method to effectively avoid the ambiguity of BVP source signal sorting. By introducing a reference signal carrying some key features of the BVP source signal, the convergence direction of the spatial domain filter is constrained during the iterative calculation process, and the BVP source signal is directly obtained, thereby eliminating the need for an additional BVP source identification algorithm. In the case of noise interference, the BVP source signal extraction method based on cICA can greatly improve the robustness of the physiological parameter detection algorithm.

[0022] 3. The present invention has huge application potential.

[0023] The non-contact physiological parameter measurement method based on facial video has gradually become a hot topic in the research and application of the field of physiological parameter detection. However, in addition to traditional heart rate detection, the extraction, research and application of other physiological parameters are still in the exploratory stage. The relevant method proposed in the present invention has better overcome the problems that have been troubled by published patents and literature, and realized the effective measurement of HRV parameters and RR. This method has achieved an accuracy rate of more than 90% in multiple comparative tests with traditional contact physiological parameter detectors. The extracted high-quality pulse wave components can be further expanded and applied in the field of non-contact physiological parameter detection, and has great application potential.

[0024] Therefore, the HRV parameter and RR measurement method based on variational mode decomposition (VMD) and constrained independent component analysis (cICA) provided in the present invention has the advantages of strong anti-noise interference ability, high measurement accuracy, and great application potential.

[0025] In addition to the above-described purposes, features and advantages, the present invention has other purposes, features and advantages. The present invention will be further described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The drawings constituting a part of the present application are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0027] Figure 1 A brief flow chart of the heart rate variability (HRV) and respiratory rate (RR) measurement method of the present invention;

[0028] Figure 2 A detailed implementation block diagram of the method of the present invention;

[0029] Figure 3 It is an effect diagram of the RGB observation signal generated based on the selected video data sample after preprocessing;

[0030] Figure 4 This is the 4-channel VMD decomposition effect diagram of the G channel observation signal;

[0031] Figure 5 The BVP reference signal generated based on the VMD-1 component of the G channel and the BVP source signal effect diagram output by the cICA algorithm;

[0032] Figure 6 This is the 4-channel VMD decomposition effect diagram of the BVP source signal;

[0033] Figure 7 This is a diagram showing the HRV parameters and RR extraction effects based on the high-quality pulse wave components in the BVP source signal.

[0034] In order to more easily understand the content of the present invention, the present invention is further described below through specific implementation methods in conjunction with the accompanying drawings. DETAILED DESCRIPTION

[0035] The human HRV and RR measurement method of the present invention performs pixel coherent averaging operation on human facial video data to convert it into RGB observation signals. The RGB observation signal is then preprocessed to obtain a standardized observation signal, and the G channel signal is decomposed into 4 channels using the VMD algorithm, and the BVP reference signal is obtained based on the component with the largest spectrum peak in the decomposed 4-channel components. Further, based on the reference signal, the BVP source signal is separated from the RGB observation signal using the cICA algorithm, and the BVP source signal is decomposed into 4 channels using the VMD algorithm to extract high-quality pulse wave components from the decomposed 4-channel components. Finally, based on the high-quality pulse wave components, the HRV parameters are obtained: low-frequency component power (LF), high-frequency component power (HF), power ratio of low-frequency component to high-frequency component (LF / HF), and RR.

[0036] Compared with the non-contact HRV parameter or RR extraction scheme proposed in the disclosed method, the present invention can avoid the BVP source sorting ambiguity problem in the scheme based on the traditional ICA algorithm, and has better resistance to noise interference, and has good application prospects in this field.

[0037] The following is a detailed description of the method for measuring human heart rate variability (HRV) and respiratory rate (RR) of the present invention.

[0038] See also Figure 1 The HRV and RR measurement method provided by the present invention comprises the following steps:

[0039] S100, performing pixel coherent averaging operation on the human face video data to convert it into RGB observation signals;

[0040] S101, performing a preprocessing operation on the RGB observation signal to obtain a standardized observation signal for subsequent analysis;

[0041] S102, performing 4-channel decomposition on the G channel signal using the VMD algorithm, and generating a BVP reference signal based on the component with the largest spectrum peak among the decomposed 4 channel components;

[0042] S103, based on the BVP reference signal, using the cICA algorithm to separate the BVP source signal from the RGB observation signal;

[0043] S104, using VMD algorithm to perform 4-channel decomposition on the BVP source signal, and extracting high-quality pulse wave components from the decomposed 4-channel components;

[0044] S105. Obtain HRV parameters based on the high-quality pulse wave components: low-frequency component power (LF), high-frequency component power (HF), power ratio of low-frequency component to high-frequency component (LF / HF), and RR.

[0045] See also Figure 2 The detailed implementation steps of the HRV and RR measurement method provided by the present invention are as follows:

[0046] 1. Collect the subject's facial video data. The device used is an ordinary RGB camera (sampling rate is 30Hz), and the subject is about 0.3-1 meter away from the camera.

[0047] 2. Use the coherent averaging method to operate on the facial video data, perform spatial coherent averaging calculation on the RGB pixel values ​​of the selected sensitive area in each frame image in the video, and convert the frame sequence into an RGB observation signal sequence, where the selected sensitive area is the forehead area.

[0048] 3. Perform cubic spline interpolation on the generated RGB observation signal and increase the sampling rate of the signal to 300 Hz to improve the time accuracy of the peak in the subsequent interbeat interval (IBI) statistics.

[0049] 4. Perform averaging, normalization, and bandpass filtering to remove noise on the RGB observation signal to obtain a standardized observation signal for subsequent analysis. The cutoff frequency of the bandpass filter is 0.5-4Hz;

[0050] 5. Use the VMD algorithm to perform 4-channel decomposition on the G channel signal, and extract the component with the largest spectrum peak from the decomposed 4-channel components.

[0051] 6. Extract the Fourier series corresponding to the maximum spectral peak in the component, perform inverse Fourier transform on it using Euler's formula, and take out the real part of the complex signal obtained by the inverse transform as the BVP source signal reference signal (the reference signal carries the frequency and phase information of the BVP source signal).

[0052] 7. Based on the BVP reference signal, the cICA algorithm is used to separate the BVP source signal from the RGB observation signal. The cICA algorithm uses an objective function based on the maximum negative entropy. During the iterative operation, the threshold of the difference between the estimated signal of the BVP source signal and the reference signal is set to 1.5, the learning step is set to 0.15, and the maximum number of iterations is 400.

[0053] 8. Use the VMD algorithm to decompose the BVP source signal into 4 channels, and take the component with the largest spectrum peak among the 4-channel components as the high-quality pulse wave component.

[0054] 9. Based on the high-quality pulse wave components, the time interval between each heartbeat peak, i.e., IBI, is calculated, and the middle value of the time points at which adjacent peaks appear is used as the time coordinate of the IBI.

[0055] 10. The obtained IBI sequence and the corresponding time coordinate sequence are subjected to uneven sampling spectrum analysis using the LS spectrum analysis method.

[0056] 11. Based on the LS spectrum analysis results, calculate the sum of the power in the 0.04-0.15 Hz frequency band (LF), the sum of the power in the 0.15-0.4 Hz frequency band (HF), and the power ratio of the low-frequency component to the high-frequency component (LF / HF) in the spectrum.

[0057] 12. Based on the LS spectrum analysis results, the frequency point corresponding to the maximum peak in the 0.15-0.4 Hz frequency band is extracted, namely the respiratory rate (RR).

[0058] See also Figure 3 Without loss of generality, in this embodiment, a video data of an ordinary subject is selected, based on the forehead area, a RGB observation signal is generated by the pixel coherent averaging method, and the above-mentioned preprocessing operation is performed.

[0059] See also Figure 4 , which shows the Figure 3 The RGB observation signal is subjected to 4-channel VMD decomposition. From the results, the component containing more obvious pulse wave information is well decomposed, as shown in the VMD-1 component. This component carries more complete key information such as BVP source signal frequency and phase, and can be used to generate BVP reference signal.

[0060] See also Figure 5 , showing that in this embodiment Figure 4 The VMD-1 component generates a BVP reference signal, and based on the reference signal, a constrained independent component analysis (cICA) operation is performed on the RGB observation signal to successfully separate the BVP source signal without the need to design an additional BVP source identification method.

[0061] See also Figure 6 , which shows the Figure 5 The BVP source signal extracted by cICA is subjected to 4-channel VMD decomposition as described in . From the decomposition results, the components carrying high-quality pulse wave components are well decomposed, such as the VMD-1 component, which carries relatively ideal human pulse wave information for the extraction of HRV parameters and RR.

[0062] See also Figure 7 , showing the example based on Figure 6The high-quality pulse wave components described in the figure provide details on the extraction of HRV parameters and RR and examples of measurement results.

[0063] First, the time interval between each heartbeat peak, i.e., IBI, is calculated from the pulse wave component, and the middle value of the time points at which adjacent peaks appear is used as the time coordinate of the IBI. Furthermore, the obtained IBI sequence and the corresponding time coordinate sequence are subjected to uneven sampling spectrum analysis using the LS spectrum analysis method. Finally, the low-frequency component power (LF), high-frequency component power (HF), power ratio of low-frequency component to high-frequency component (LF / HF) in the spectrum, and the frequency point corresponding to the maximum peak in the high-frequency band, i.e., RR, are calculated.

[0064] The present invention also provides a human HRV and RR measurement device based on variational mode decomposition (VMD) and constrained independent component analysis (cICA), including program modules one to six.

[0065] Program module one is used to perform pixel coherent averaging operation on human facial video data to convert it into RGB observation signals; program module two is used to perform preprocessing operations on RGB observation signals to obtain standardized observation signals for subsequent analysis; program module three is used to perform 4-channel decomposition of G channel signals using VMD algorithm, and obtain BVP reference signal based on the component with the largest spectrum peak in the decomposed 4-channel components; program module four is used to separate BVP source signal from RGB observation signal based on BVP reference signal using cICA algorithm; program module five is used to perform 4-channel decomposition of BVP source signal using VMD algorithm, and extract high-quality pulse wave component from the decomposed 4-channel components; and program module six is ​​used to obtain HRV parameters and RR based on high-quality pulse wave components, where HRV parameters include: low-frequency component power (LF), high-frequency component power (HF), and power ratio of low-frequency component to high-frequency component (LF / HF).

[0066] The human body HRV and RR measurement method according to the present invention is preferably integrated into an electronic device in the form of a computer processing program, and the electronic device may be a server, or a terminal or other device.

[0067] Among them, the server can be an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, network acceleration services (Content Delivery Network, CDN), as well as big data and artificial intelligence platforms.

[0068] The terminal may be a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smart watch, etc. The terminal and the server may be directly or indirectly connected via wired or wireless communication.

[0069] The computer-readable storage medium may include: a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk, etc., which stores a human HRV and RR measurement program, which is used to implement the steps of the above-mentioned human HRV and RR measurement method when executed.

[0070] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A human HRV and RR measurement method based on variational mode decomposition (VMD) and constrained independent component analysis (cICA), characterized in that: include: S100, performing pixel coherent averaging operation on the human face video data to convert it into RGB observation signals; S101, performing a preprocessing operation on the RGB observation signal to obtain a standardized observation signal for subsequent analysis; S102, performing 4-channel decomposition on the G channel signal using the VMD algorithm, and obtaining a BVP reference signal based on the component with the largest spectrum peak among the decomposed 4-channel components; S103, based on the reference signal, using the cICA algorithm to separate the BVP source signal from the RGB observation signal; S104, using VMD algorithm to perform 4-channel decomposition on the BVP source signal, and extracting high-quality pulse wave components from the decomposed 4-channel components; S105, based on the high-quality pulse wave components, obtain HRV parameters: low-frequency component power (LF), high-frequency component power (HF), power ratio of low-frequency component to high-frequency component (LF / HF), and respiratory rate (RR), Obtaining the BVP reference signal in step S102 includes: Step 21: Perform fast Fourier transform and spectrum analysis on the 4-channel components decomposed by the VMD algorithm; Step 22: extract the component with the largest spectrum peak among the 4-channel components; Step 23: extracting the Fourier series corresponding to the maximum spectrum peak in the component; Step 24: Perform inverse Fourier transform on the Fourier series using the Euler formula, and extract the real part of the complex signal obtained by the inverse transform as the BVP reference signal. The reference signal carries the frequency and phase information of the BVP source signal. The cICA algorithm used in step S103 adopts an objective function based on the maximum negative entropy; in addition, during the iterative operation of the cICA algorithm to extract the BVP source signal, the threshold of the gap between the estimated signal and the reference signal of the BVP source signal is set to 1.5, the learning step size is set to 0.15, and the maximum number of iterations is 400. The high-quality pulse wave component extracted in step S104 is the component with the largest spectrum peak among the four-channel components decomposed from the BVP source signal.

2. The method for measuring human HRV and RR based on variational mode decomposition (VMD) and constrained independent component analysis (cICA) according to claim 1, characterized in that: Generating the RGB observation signal in step S100 includes: calculating the RGB pixel values ​​of the selected sensitive area in each frame image of the facial video using the coherent averaging method, and then converting the frame sequence into an RGB observation signal sequence, wherein the selected sensitive area is the forehead area.

3. The method for measuring human HRV and RR based on variational mode decomposition (VMD) and constrained independent component analysis (cICA) according to claim 1, characterized in that: The preprocessing operation of the RGB observation signal in step S101 includes: Step 11: Perform cubic spline interpolation on the RGB observation signal to increase the sampling rate from the common approximately 30 Hz to 300 Hz; Step 12: De-mean and normalize the RGB observation signal; Step 13: Bandpass filter the RGB observation signal to denoise it, with the filter cutoff frequency being 0.5-4 Hz.

4. The method for measuring human HRV and RR based on variational mode decomposition (VMD) and constrained independent component analysis (cICA) according to claim 1, characterized in that: In step S105, the HRV parameters and RR are obtained based on the high-quality pulse wave components, including: Step 31: Count the time intervals between the peaks in the high-quality pulse wave component, i.e., the interbeat interval (IBI), and use the middle value of the time points at which adjacent peaks appear as the time coordinate of the IBI; Step 32: Perform uneven sampling spectrum analysis on the obtained IBI sequence and the corresponding time coordinate sequence using the S (Lomb-Scargle) spectrum analysis method; Step 33: Calculate the sum of the powers in the frequency band of 0.04-0.15 Hz in the spectrum as the low frequency component power (LF), and the sum of the powers in the frequency band of 0.15-0.4 Hz as the high frequency component power (HF); then calculate the power ratio of the low frequency component to the high frequency component (LF / HF); Step 34: Find the frequency point corresponding to the maximum peak in the 0.15-0.4 Hz frequency band, that is, the respiratory rate (RR).

5. A human HRV and RR measurement device based on variational mode decomposition (VMD) and constrained independent component analysis (cICA), characterized in that: include: Program module 1 is used to perform pixel coherent averaging operation on human face video data to convert it into RGB observation signal; Program module 2 is used to perform preprocessing operations on RGB observation signals to obtain standardized observation signals for subsequent analysis; Program module three is used to perform 4-channel decomposition on the G channel signal using the VMD algorithm, and obtain the BVP reference signal based on the component with the largest spectrum peak among the decomposed 4-channel components; Program module 4 is used to separate the BVP source signal from the RGB observation signal using the cICA algorithm based on the BVP reference signal; Program module 5 is used to perform 4-channel decomposition of BVP source signal using VMD algorithm, and extract high-quality pulse wave components from the decomposed 4-channel components; Program module six is ​​used to obtain HRV parameters and RR based on high-quality pulse wave components, wherein the HRV parameters include: low-frequency component power (LF), high-frequency component power (HF), and power ratio of low-frequency component to high-frequency component (LF / HF).

6. A computer-readable storage medium storing a program, characterized in that: When the program is executed, each step of the method for measuring human HRV and RR based on variational mode decomposition and constrained independent component analysis according to any one of claims 1 to 5 is implemented.

7. A computer device comprising a processor and a memory, characterized in that: The memory stores a program, which, when executed on the processor, implements the steps of the method for measuring human HRV and RR based on variational mode decomposition and constrained independent component analysis according to any one of claims 1 to 5.

Images