CN112998704A - Wearable device blood oxygen saturation calculation method - Google Patents
Wearable device blood oxygen saturation calculation method Download PDFInfo
- Publication number
- CN112998704A CN112998704A CN202110240523.3A CN202110240523A CN112998704A CN 112998704 A CN112998704 A CN 112998704A CN 202110240523 A CN202110240523 A CN 202110240523A CN 112998704 A CN112998704 A CN 112998704A
- Authority
- CN
- China
- Prior art keywords
- signal
- blood oxygen
- value
- wavelet
- oxygen saturation
- 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
Links
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 title claims abstract description 44
- 210000004369 blood Anatomy 0.000 title claims abstract description 44
- 239000008280 blood Substances 0.000 title claims abstract description 44
- 229910052760 oxygen Inorganic materials 0.000 title claims abstract description 44
- 239000001301 oxygen Substances 0.000 title claims abstract description 44
- 238000004364 calculation method Methods 0.000 title claims description 17
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 41
- 239000011159 matrix material Substances 0.000 claims abstract description 33
- 238000000034 method Methods 0.000 claims abstract description 24
- 238000010183 spectrum analysis Methods 0.000 claims abstract description 10
- 238000012360 testing method Methods 0.000 claims abstract description 6
- 238000012216 screening Methods 0.000 claims abstract description 4
- 238000012163 sequencing technique Methods 0.000 claims abstract description 3
- 238000005070 sampling Methods 0.000 claims description 6
- 238000004458 analytical method Methods 0.000 claims description 5
- 239000000126 substance Substances 0.000 claims description 3
- 238000001914 filtration Methods 0.000 description 14
- 238000005259 measurement Methods 0.000 description 8
- 238000004422 calculation algorithm Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 210000000707 wrist Anatomy 0.000 description 4
- 102000001554 Hemoglobins Human genes 0.000 description 3
- 108010054147 Hemoglobins Proteins 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 230000000717 retained effect Effects 0.000 description 3
- 238000001228 spectrum Methods 0.000 description 3
- 238000006073 displacement reaction Methods 0.000 description 2
- 238000000862 absorption spectrum Methods 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 238000001035 drying Methods 0.000 description 1
- 210000003743 erythrocyte Anatomy 0.000 description 1
- 210000001061 forehead Anatomy 0.000 description 1
- 238000012880 independent component analysis Methods 0.000 description 1
- 238000000691 measurement method Methods 0.000 description 1
- 238000002496 oximetry Methods 0.000 description 1
- 238000006213 oxygenation reaction Methods 0.000 description 1
- 238000013186 photoplethysmography Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000029058 respiratory gaseous exchange Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/145—Measuring characteristics of blood in vivo, e.g. gas concentration, pH value; Measuring characteristics of body fluids or tissues, e.g. interstitial fluid, cerebral tissue
- A61B5/1455—Measuring characteristics of blood in vivo, e.g. gas concentration, pH value; Measuring characteristics of body fluids or tissues, e.g. interstitial fluid, cerebral tissue using optical sensors, e.g. spectral photometrical oximeters
- A61B5/14551—Measuring characteristics of blood in vivo, e.g. gas concentration, pH value; Measuring characteristics of body fluids or tissues, e.g. interstitial fluid, cerebral tissue using optical sensors, e.g. spectral photometrical oximeters for measuring blood gases
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/68—Arrangements of detecting, measuring or recording means, e.g. sensors, in relation to patient
- A61B5/6801—Arrangements of detecting, measuring or recording means, e.g. sensors, in relation to patient specially adapted to be attached to or worn on the body surface
- A61B5/6802—Sensor mounted on worn items
-
- 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
-
- 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
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
Abstract
The invention discloses a method for calculating the blood oxygen saturation of wearable equipment, which comprises the following steps of S1), collecting PPG signals of P seconds, including red light, infrared light and green light signals (the value of P is selected according to the requirement of real-time test), and calculating the direct current values of the red light and the infrared light. S2), removing high-frequency and low-frequency interference in the PPG signal by using wavelet denoising. S3), calculating a heart rate value by using the green light signal after wavelet denoising. S4), converting the red light and infrared light signals after wavelet de-noising into a Hankel matrix, then performing singular value decomposition on the Hankel matrix to obtain a characteristic value matrix, and then sequencing characteristic values from large to small. S5), respectively carrying out spectrum analysis on the signal components of the first N characteristic values of the red light and the infrared light, and screening out the signal components meeting the threshold value according to the frequency threshold value in the step 3). By using the technical scheme of the invention, various interferences in the wearable device blood oxygen signal can be effectively eliminated, a 'clean' PPG signal is obtained, and the value of the blood oxygen saturation is accurately calculated.
Description
Technical Field
The invention relates to measurement of pulse blood oxygen saturation, in particular to a real-time measurement method of blood oxygen saturation of wearable equipment.
Background
The blood oxygen saturation refers to the relative oxygen content of the hemoglobin of the red blood cells, and is one of the main indicators of the health degree of human respiration. The measurement mode of the blood oxygen saturation is divided into: invasive measurement and non-invasive measurement. Invasive measurement requires the first acquisition of arterial blood from the body, which has the disadvantages of not being able to measure continuously and causing pain to the tester. The invention relates to a non-invasive measuring method. The non-invasive measurement is to obtain transmitted or reflected red light and infrared light PPG signals by utilizing a photoplethysmography, and calculate the blood oxygen saturation through Lambert-beer law according to the absorption spectrum difference of oxygen, hemoglobin and reduced hemoglobin to the red light and the infrared light.
The common calculation formula of blood oxygen saturation is:
SpO2＝A+B*R+C*R2 (1)
wherein A, B, C is an empirical constant, and only the reflected or transmitted red light and infrared light signals need to be obtained, and the direct current and alternating current components thereof are substituted into the formula (2) to obtain the R value, and then the R value is substituted into the formula (1) to obtain the blood oxygen saturation value.
In the process of collecting the PPG signal, due to internal factors of differences of human tissues and external factor interference (a. influence of movement of the measured part, b. relative displacement of the test device and the test part, c. ambient light, etc.), the obtained PPG signal usually has large noise interference, and these interference signals seriously affect accurate calculation of blood oxygen saturation. Therefore, the de-drying of the raw PPG signal is a very important task. In general, a high-low pass filter, an FFT filter, and an FIR filter all have phase distortion, and there are problems of insufficient accuracy and boundary for filtering of a finite data length in a short time, and when the frequency of an interference signal coincides with the frequency of a pulse wave, it is difficult for a general filtering and denoising method to separate the interference signal.
For the real-time measurement of the blood oxygen saturation of the wearable device, due to the limit of the calculation capability of the mcu of most wearable test devices, the calculation time is seriously affected by directly using a complex denoising algorithm, and even the calculation capability of the device is exceeded, such as adaptive filtering, singular value decomposition, independent component analysis and the like.
Therefore, the calculation of blood oxygen saturation in real time is really a difficult problem, especially the difficulty in extracting "clean" blood oxygen signal.
Disclosure of Invention
The invention aims to effectively remove various interference signals in a PPG signal, extract a 'clean' blood oxygen signal and realize real-time calculation of the blood oxygen saturation of wearable equipment.
In order to solve the technical problems, the invention adopts the following technical method:
s1), collecting PPG signals of P seconds, including red light, infrared light and green light signals (the value of P is selected according to the requirement of real-time test), and calculating the direct current value of red light and infrared light.
S2), removing high-frequency and low-frequency interference in the PPG signal by using wavelet denoising.
S3), calculating a heart rate value by using the green light signal after wavelet de-noising, and providing a reference frequency threshold value for de-noising the singular value of red light and infrared light.
S4), converting the red light and infrared light signals after wavelet de-noising into a Hankel matrix, then performing singular value decomposition on the Hankel matrix to obtain a characteristic value matrix, and then sequencing characteristic values from large to small.
S5), respectively carrying out spectrum analysis on the signal components of the first N characteristic values of the red light and the infrared light, and screening out the signal components meeting the threshold value according to the frequency threshold value in the step 3). (wherein the larger N is, the larger the calculation amount is, and the appropriate N value can be selected according to actual conditions).
S6), reconstructing the decomposed signals according to the screened singular values, and obtaining two paths of clean blood oxygen signals.
S7), calculating the blood oxygen saturation value according to the clean blood oxygen signal by using the formulas (1) and (2).
S8), continuing sampling, and repeating steps S1) -S7), the calculation of the blood oxygen saturation in real time can be realized.
Optionally, in step S2), wavelet denoising is performed, and the specific subdivision steps are as follows:
wavelet basis selection: since the PPG signal approximates a sinusoidal signal, a DB wavelet basis may be selected.
Wavelet decomposition: according to the sampling signal frequency, an appropriate number of decomposition layers (generally 5-8 layers) is selected.
Wavelet denoising and reconstruction: and selecting a proper denoising reconstruction mode according to the heart rate frequency range of the human body and the frequency characteristics of the interference signal to obtain a relatively clean PPG signal.
Optionally, the step S4) is converted into a Hankel matrix, because the PPG signal is a one-dimensional time-amplitude signal, and the singular value decomposition is a decomposition method of matrix eigenvalues, it needs to be converted into a two-dimensional Hankel matrix form first.
Let x (n) be the observed signal, n ═ 1,2, …, L, whose Hankel matrix is as follows:
note: l is m + n-1, A is m x n dimensional matrix
Optionally, the singular value decomposition formula in step 4) is as follows:
wherein, U and V are orthogonal matrixes of m dimension and n dimension, and Σ is an eigenvalue matrix of A, and the value on the diagonal is r eigenvalues.
Optionally, the singular value screening method in step 5) includes:
and reconstructing a signal component represented by each singular value, analyzing the frequency spectrum of the signal component, if the frequency threshold is met, reserving the signal component, and if the frequency threshold is not met, removing the signal component.
Optionally, in step S5), performing spectrum analysis on the signal components of the first N feature values of the red light and the infrared light, respectively: the method is characterized in that Fourier transform is carried out on component signals corresponding to each characteristic value, and then required components are screened out according to a threshold value.
Optionally, the signal reconstructing in step S6) includes: the decomposition of the singular values reconstructs and restores the two-dimensional signal to the one-dimensional signal.
The results of the invention are:
by adopting the technical scheme of the invention, various interferences in the wearable device blood oxygen signal can be effectively eliminated, a 'clean' PPG signal is obtained, and the value of the blood oxygen saturation is accurately calculated. The method is relatively complex in operation, but the operation amount can be greatly reduced by selecting a proper signal length, a proper number of wavelet decomposition layers and a proper singular value decomposition reconstruction algorithm, so that the accurate real-time calculation of the blood oxygen saturation on the wearable device is realized.
Drawings
FIG. 1 is a schematic diagram of the algorithm of the present invention.
Fig. 2 is a diagram of the raw PPG signal collected from the back of the wrist in a first embodiment of the invention.
Figure 3 is a schematic diagram of wavelet decomposition.
Fig. 4 is a wavelet-filtered signal of the original signal in fig. 2.
Fig. 5 is a graph obtained by converting the small-wave filtered red light signals in fig. 4 into a Hankel matrix, then performing singular value decomposition to obtain component values corresponding to the characteristic values, and performing spectrum analysis on each component.
Fig. 6 is a flowchart of singular value decomposition and reconstruction.
Fig. 7 is a diagram that the infrared light signal after the small wave filtering in fig. 4 is converted into a Hankel matrix, then singular value decomposition is performed to obtain component values corresponding to the characteristic values, and the component frequency spectrums are analyzed.
FIG. 8 is the resulting "clean" blood oxygenation signal for red and infrared light.
Detailed Description
In order to make the technical solutions of the present invention better understood, the technical solutions in the embodiments of the present invention will be described below clearly and completely with reference to the accompanying drawings in the embodiments of the present invention.
The specific embodiment provides a method for calculating blood oxygen saturation on wearable equipment, and a flow chart of the method is shown in fig. 1:
first, raw PPG data is collected, which contains green light, red light, infrared light. Usually, data of 6 seconds is selected to realize the requirement of real-time measurement, what this embodiment was gathered is the PPG signal of wrist back, and the characteristics of the PPG signal of wrist back are: the signal is weak in strength and poor in quality compared with the finger tip, the forehead and the like; the interference signals are also more, especially the red light signals, which are easily interfered. As shown in fig. 2.
And (4) solving the direct current components of the acquired infrared signals and the infrared signals according to the acquired infrared signals and the acquired infrared signals, and finally solving the blood oxygen saturation for use.
And then, performing wavelet filtering on the acquired three paths of signals, wherein the wavelet filtering removes high-frequency and low-frequency interference in the PPG signal. In this embodiment, the wavelet basis is Daubechies (db5), the number of wavelet decomposition layers is 6, the high-frequency parts of the 1 st and 2 nd layers are set to zero, and the low-frequency part of the 6 th layer is set to zero, so that not only can the band-pass filtering effect be achieved, but also the boundary problem can be solved through boundary extension. The wavelet decomposition is schematically shown in fig. 3, and the specific post-filtering effect is shown in fig. 4.
Then, the following processing is respectively carried out on the three paths of light:
and selecting green light to perform Fourier transform, and calculating a heart rate value, wherein the heart rate value provides a frequency threshold value for de-noising singular values of red light and infrared light.
Selecting red light to be converted into a Hankel matrix, then decomposing singular values of the Hankel matrix, sorting the singular values from large to small, selecting the first 10 eigenvalue components to carry out spectral analysis respectively, comparing green light reference frequencies, and reserving signal components meeting threshold conditions. Of course, the feature value can be selected to be N, wherein the larger N is, the larger the calculation amount is, and the appropriate N value can be selected according to actual conditions. The component and spectral analysis is shown in fig. 5. The analysis flow is shown in fig. 6.
Infrared light is selected for the same process as red light. The components and spectral analysis are shown in fig. 7.
The remaining characteristic components of the red and infrared light are reconstructed to obtain a "clean" blood oxygen signal, as shown in fig. 8.
And finally, according to the 'clean' blood oxygen signal, the alternating current value of the red light and the infrared light is calculated. And calculating the blood oxygen saturation together with the direct current value obtained by the original signal.
The key points of the invention are two points: one is wavelet filtering and the other is Singular Value Decomposition (SVD) denoising. The following two points are described in detail.
The key step one: wavelet filtering
The wavelet transform has time-frequency multi-resolution analysis characteristics, and is particularly suitable for analyzing and processing pulse wave signals.
The wavelet transform includes a continuous wavelet transform and a discrete wavelet transform. The invention uses a one-dimensional discrete wavelet transform.
Discrete wavelet transform (dwt) (discrete wavelet transform) defines:
discretizing the scale parameters according to a power series, and uniformly discretizing the time (satisfying the Nyquist sampling theorem).
Where ψ (t) is the wavelet basis function, under the framework of discrete wavelets, Meyer, Daubechies et al construct different wavelet basis functions.
In the embodiment, a Daubechies wavelet base (db5) is selected, and the characteristics are as follows:
1) is of limited support in the time domain, i.e.
Limited length, its high-order origin moment ^ tPψ (t) dt is 0, p is 0 to N, the larger N,
the longer the length of (a).
2) There is an N-th order zero at 0 at ψ (ω).
3)
And orthonormal to its integer displacement.
The number of decomposition levels in this example is 6, and the wavelet decomposition tree is shown in fig. 8 below, where cA is the low frequency part (also called approximation component) and cD is the high frequency part (also called detail component).
After wavelet decomposition, zero setting of cA6, cD1 and cD2 can achieve better effect than common band-pass filtering.
Further, the wavelet component reconstruction after denoising is also called Inverse Discrete Wavelet Transform (IDWT), which is a wavelet decomposition Inverse process.
The key step two: singular value decomposition denoising
Singular value definition:
let A be an element of Rm×n，ATThe non-negative square root of the eigenvalues of a is called the singular values of a, and is still true when a is a complex matrix.
Singular value decomposition theorem:
let A be an element of Rm×nThen the orthogonal matrix U ═ U must exist1,...,um]∈Rm×nAnd V ═ V1,...,vn]∈Rn×nSo that the following holds:
therein, sigmar＝diag(σ1,...,σr),σ1≥...≥σr＞0
The decomposition equation (7) is called the singular value decomposition of the matrix a, commonly referred to as SVD.
Singular value decomposition is widely applied, and is applied to signal denoising, image compression, fault analysis, face recognition and the like.
The signal processing method based on singular value decomposition is a nonlinear filtering method, and has a good processing effect on nonlinear and non-stationary signals.
Singular value decomposition calculation algorithm:
the present example uses a hybrid algorithm of conventional QR iteration and zero-shift QR iteration. The method applies a plurality of mature skills of the traditional QR, has higher efficiency and good precision, and is a moderate method.
SVD decomposition is a complex operation requiring a geometric progression of the number of computations as the data length increases. In this example, the sampling rate is 50Hz, the data length is 300, and therefore the SVD operation amount is relatively low.
Singular value decomposition denoising:
obviously, the singular value decomposition is a decomposition method of matrix eigenvalue, and the PPG signal in this embodiment is a one-dimensional time signal, so that a dimension expansion method needs to be adopted to resample the signal.
Let x (i) be a red signal, i ═ 1,2, …,300, whose Hankel matrix is as follows:
wherein A is 150x151 dimensional matrix
Then singular value decomposition
Wherein the content of the first and second substances,
S＝diag(σ1,σ2,...,σr)
this example analyzes only the first 10 singular values (i.e.: σ)1,σ2,...,σ10) Because the first 10 largest singular values after wavelet filtering substantially contain the desired oximetry signal. The component information of red light and infrared light and the spectrum analysis thereof are shown in fig. 5 and 7, respectively.
According to fig. 5, the red y (1) and y (2) components are spectrally close to the reference value, with their components retained and the remaining components removed.
According to fig. 7, the infrared light y (1), y (2) component spectra are close to the reference value, the components thereof are retained, and the remaining components are removed, i.e. the characteristic value is set to zero.
Singular value reconstruction:
1) reconstructing the retained feature components according to equation (9)
For example: singular value sigma2Retention, with reconstructed matrix of
Wherein the content of the first and second substances,
2) converting the two-dimensional matrix into a one-dimensional signal according to equation (8)
For example: obtaining the matrix from step 1)
3) Adding the reserved components to obtain a one-dimensional signal after singular value decomposition and reconstruction
For example: retaining the components y (1) and y (2), wherein
Then, singular value decomposition of the reconstructed one-dimensional signal
This example shows in particular the calculation of the blood oxygen saturation at the wrist. In the original signal, green light and infrared light are not easily interfered, and the signal is good. However, the red light is very poor, the pulse signal can hardly be seen from the original signal, and a relatively clean signal is still separated from the noise mixed signal after wavelet filtering and SVD denoising.
The above examples are further detailed descriptions of the present invention in conjunction with specific scenario implementations, and the present invention is not considered to be limited thereto. For those skilled in the art to which the invention pertains, several simple deductions or substitutions can be made without departing from the spirit of the invention, and all of them belong to the protection scope of the invention.
Claims (7)
1. A wearable device blood oxygen saturation calculation method comprises the following steps
S1), collecting PPG signals of P seconds, including red light, infrared light and green light signals (the value of P is selected according to the requirement of real-time test), and calculating the direct current value of red light and infrared light.
S2), removing high-frequency and low-frequency interference in the PPG signal by using wavelet denoising.
S3), calculating a heart rate value by using the green light signal after wavelet de-noising, and providing a reference frequency threshold value for de-noising the singular value of red light and infrared light.
S4), converting the red light and infrared light signals after wavelet de-noising into a Hankel matrix, then performing singular value decomposition on the Hankel matrix to obtain a characteristic value matrix, and then sequencing characteristic values from large to small.
S5), respectively carrying out spectrum analysis on the signal components of the first N characteristic values of the red light and the infrared light, and screening out the signal components meeting the threshold value according to the frequency threshold value in the step 3).
S6), reconstructing the decomposed signals according to the screened singular values, and obtaining two paths of 'clean' blood oxygen signals.
S7), calculating a value of blood oxygen saturation based on the clean blood oxygen signal.
S8), continuing sampling, and repeating steps S1) -S7), the calculation of the blood oxygen saturation in real time can be realized.
2. The method for calculating blood oxygen saturation level of a wearable device according to claim 1, wherein the step S2) specifically refers to: the noise is removed by a wavelet analysis method, and a relatively 'clean' pulse wave signal is obtained.
3. The wearable device blood oxygen saturation calculation method according to claim 2, characterized in that the wavelet analysis method mainly comprises the following steps:
wavelet basis selection: since the PPG signal approximates a sinusoidal signal, a DB wavelet basis may be selected.
Wavelet decomposition: according to the sampling signal frequency, an appropriate number of decomposition layers (generally 5-8 layers) is selected.
Wavelet denoising and reconstruction: and selecting a proper denoising reconstruction mode according to the heart rate frequency range of the human body and the frequency characteristics of the interference signal to obtain a relatively clean PPG signal.
4. The method for calculating the blood oxygen saturation level of a wearable device according to claim 1, wherein in step S4), the red light signal and the infrared light signal after wavelet denoising are first converted into a Hankel matrix, and the formula is as follows:
let x (i) be the observed signal, i ═ 1,2, …, L, whose Hankel matrix is as follows:
wherein, L is m + n-1, and A is an m × n dimensional matrix.
5. The method for calculating the blood oxygen saturation level of a wearable device as claimed in claim 1, wherein in step S4), the Hankel matrix is subjected to singular value decomposition, and the formula is as follows:
wherein the content of the first and second substances,
S＝diag(σ1,σ2,...,σr) U and V are orthogonal matrixes of m dimension and n dimension, and sigma is an eigenvalue matrix of A, and the value on the diagonal line is r eigenvalues.
6. The method for calculating the blood oxygen saturation level of a wearable device according to claim 1, wherein in step S5), the signal components of the first N characteristic values of the red light and the infrared light are respectively subjected to spectrum analysis: the method is characterized in that Fourier transform is carried out on component signals corresponding to each characteristic value, and then required components are screened out according to a threshold value.
7. The method for calculating the blood oxygen saturation level of a wearable device according to claim 1, wherein the signal reconstruction in the step S6) includes: the decomposition of the singular values reconstructs and restores the two-dimensional signal to the one-dimensional signal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110240523.3A CN112998704A (en) | 2021-03-04 | 2021-03-04 | Wearable device blood oxygen saturation calculation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110240523.3A CN112998704A (en) | 2021-03-04 | 2021-03-04 | Wearable device blood oxygen saturation calculation method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112998704A true CN112998704A (en) | 2021-06-22 |
Family
ID=76405337
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110240523.3A Pending CN112998704A (en) | 2021-03-04 | 2021-03-04 | Wearable device blood oxygen saturation calculation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112998704A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113616203A (en) * | 2021-09-02 | 2021-11-09 | 中船海洋探测技术研究院有限公司 | Diver underwater blood oxygen detection method based on wavelet filtering algorithm |
| CN114403823A (en) * | 2022-01-20 | 2022-04-29 | 杭州纳境科技有限公司 | Heart rate blood oxygen detection method and device and wearable device |
| CN116982952A (en) * | 2023-08-03 | 2023-11-03 | 迈德医疗科技(深圳)有限公司 | Noninvasive blood pressure measurement method and system based on fractional derivative |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103932688A (en) * | 2014-05-15 | 2014-07-23 | 成都天奥电子股份有限公司 | Sudden death early-warning device and wrist watch applying sudden death early-warning device |
| CN109883547A (en) * | 2019-01-10 | 2019-06-14 | 武汉大学 | A kind of wide-band spectrum signal antinoise method based on wavelet threshold difference |
| CN112263249A (en) * | 2019-07-08 | 2021-01-26 | 复旦大学附属中山医院 | ECG-based method and device for enhancing blood oxygen saturation monitoring |
-
2021
- 2021-03-04 CN CN202110240523.3A patent/CN112998704A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103932688A (en) * | 2014-05-15 | 2014-07-23 | 成都天奥电子股份有限公司 | Sudden death early-warning device and wrist watch applying sudden death early-warning device |
| CN109883547A (en) * | 2019-01-10 | 2019-06-14 | 武汉大学 | A kind of wide-band spectrum signal antinoise method based on wavelet threshold difference |
| CN112263249A (en) * | 2019-07-08 | 2021-01-26 | 复旦大学附属中山医院 | ECG-based method and device for enhancing blood oxygen saturation monitoring |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113616203A (en) * | 2021-09-02 | 2021-11-09 | 中船海洋探测技术研究院有限公司 | Diver underwater blood oxygen detection method based on wavelet filtering algorithm |
| CN114403823A (en) * | 2022-01-20 | 2022-04-29 | 杭州纳境科技有限公司 | Heart rate blood oxygen detection method and device and wearable device |
| CN116982952A (en) * | 2023-08-03 | 2023-11-03 | 迈德医疗科技(深圳)有限公司 | Noninvasive blood pressure measurement method and system based on fractional derivative |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112998704A (en) | Wearable device blood oxygen saturation calculation method | |
| Castillo et al. | Noise suppression in ECG signals through efficient one-step wavelet processing techniques | |
| Gupta et al. | A comparison of ECG signal pre-processing using FrFT, FrWT and IPCA for improved analysis | |
| US20080262371A1 (en) | Method for Adaptive Complex Wavelet Based Filtering of Eeg Signals | |
| Bhateja et al. | A composite wavelets and morphology approach for ECG noise filtering | |
| CN104305992B (en) | A kind of interactive fast automatic extracting method of Fetal ECG | |
| CN109359506A (en) | A kind of mcg-signals noise-reduction method based on wavelet transformation | |
| Al-kadi et al. | Compatibility of mother wavelet functions with the electroencephalographic signal | |
| Mir et al. | ECG denoising and feature extraction techniques–a review | |
| Carvalho et al. | A tool for time-frequency analysis of heart rate variability | |
| Adochiei et al. | ECG waves and features extraction using Wavelet Multi-Resolution Analysis | |
| Altay et al. | Signal-to-noise ratio and mean square error improving algorithms based on newton filters for measurement ECG data processing | |
| Ram et al. | Use of multi-scale principal component analysis for motion artifact reduction of PPG signals | |
| Mahadevan et al. | Ballistocardiogram artifact removal in EEG-fMRI signals using discrete Hermite transforms | |
| Fehér | Denoising ECG signals by applying discrete wavelet transform | |
| German-Sallo | ECG signal baseline wander removal using wavelet analysis | |
| Ionescu et al. | Fetal heart rate detection and monitoring from noninvasive abdominal ECG recordings | |
| Elouaham et al. | Combination time-frequency and empirical wavelet transform methods for removal of composite noise in EMG signals | |
| Stolojescu | ECG signals classification using statistical and time-frequency features | |
| Sampath et al. | ECG morphological marking using discrete wavelet transform | |
| Sun et al. | An improved PPG denoising methodology based on EEMD and wavelet threshold | |
| Singh et al. | Virtual instrument based real time ECG monitoring device | |
| Jhang et al. | Integration Design of Portable ECG Signal Acquisition With Deep-Learning Based Electrode Motion Artifact Removal on an Embedded System | |
| El Khadiri et al. | A Comparison of the Denoising Performance Using Capon Time-Frequency and Empirical Wavelet Transform Applied on Biomedical Signal. | |
| Shao et al. | A Photoplethysmograph Signal Preprocess Method Based on Wavelet Transform |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210622 |