CN111248880A - Blood pressure prediction method and device based on photoplethysmography signals - Google Patents
Blood pressure prediction method and device based on photoplethysmography signals Download PDFInfo
- Publication number
- CN111248880A CN111248880A CN202010110279.4A CN202010110279A CN111248880A CN 111248880 A CN111248880 A CN 111248880A CN 202010110279 A CN202010110279 A CN 202010110279A CN 111248880 A CN111248880 A CN 111248880A
- Authority
- CN
- China
- Prior art keywords
- ppg
- matrix
- generate
- wavelet
- factor
- 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.)
- Granted
Links
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/02—Detecting, measuring or recording pulse, heart rate, blood pressure or blood flow; Combined pulse/heart-rate/blood pressure determination; Evaluating a cardiovascular condition not otherwise provided for, e.g. using combinations of techniques provided for in this group with electrocardiography or electroauscultation; Heart catheters for measuring blood pressure
- A61B5/021—Measuring pressure in heart or blood vessels
- A61B5/02108—Measuring pressure in heart or blood vessels from analysis of pulse wave characteristics
-
- 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
- A61B5/726—Details of waveform analysis characterised by using transforms using Wavelet transforms
-
- 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
- A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
- A61B5/7267—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems involving training the classification device
-
- 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
- A61B5/7271—Specific aspects of physiological measurement analysis
- A61B5/7275—Determining trends in physiological measurement data; Predicting development of a medical condition based on physiological measurements, e.g. determining a risk factor
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Public Health (AREA)
- Animal Behavior & Ethology (AREA)
- Veterinary Medicine (AREA)
- Physiology (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Molecular Biology (AREA)
- Surgery (AREA)
- General Health & Medical Sciences (AREA)
- Signal Processing (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Psychiatry (AREA)
- Cardiology (AREA)
- Vascular Medicine (AREA)
- Evolutionary Computation (AREA)
- Fuzzy Systems (AREA)
- Mathematical Physics (AREA)
- Image Processing (AREA)
- Measuring Pulse, Heart Rate, Blood Pressure Or Blood Flow (AREA)
Abstract
The embodiment of the invention relates to a blood pressure prediction method and a blood pressure prediction device based on a photoplethysmography signal, wherein the method comprises the following steps: generating a PPG signal segment; performing signal decomposition on the PPG signal segment by using a continuous wavelet transform mode according to the acquired wavelet base type, the expansion factor and the movement factor to generate a PPG wavelet coefficient matrix; converting a PPG wavelet coefficient matrix by taking a modulus of matrix elements and carrying out normalization processing on the converted matrix to generate a PPG normalization matrix; acquiring an RGB color wheel matrix, and carrying out tensor conversion on a PPG normalization matrix according to the RGB color wheel matrix to generate a PPG time-frequency three-dimensional tensor; performing tensor shape reconstruction on the PPG time-frequency three-dimensional tensor by using a bicubic interpolation algorithm according to the input width threshold of the convolutional network to generate a PPG convolutional three-dimensional tensor; and performing classification regression calculation on the PPG convolution three-dimensional tensor by using a convolution neural network classification regression model to generate a PPG prediction blood pressure data pair.
Description
Technical Field
The invention relates to the technical field of electrophysiological signal processing, in particular to a blood pressure prediction method and device based on a photoplethysmography signal.
Background
The heart is the center of human blood circulation, and the heart produces blood pressure through regular pulsation, and then supplies blood to the whole body to complete the metabolism of the human body, and blood pressure is one of the very important physiological signals of the human body. Human blood pressure contains two important values: systolic pressure and diastolic pressure, and whether the blood pressure of a human body is normal or not is judged by the two quantities medically. The long-term continuous observation of the two parameters of the blood pressure can help people to have clear understanding on the health state of the heart. However, most of the conventional blood pressure measuring methods adopt an external force pressing detection method such as a pressure gauge, which is not only cumbersome to operate, but also easily causes discomfort to the person to be measured, and thus cannot be used for multiple times to achieve the purpose of continuous monitoring.
Disclosure of Invention
The invention aims to provide a blood pressure prediction method and a blood pressure prediction device based on a Photoplethysmography signal, aiming at the defects of the prior art, wherein a Photoplethysmography (PPG) device is used for carrying out non-invasive data acquisition on a tester, so that the problem that the tester cannot be continuously observed in conventional monitoring is solved; in order to fully obtain effective signal data from the PPG signal, the embodiment of the invention adopts a continuous wavelet transform mode to perform signal decomposition on the PPG signal; in order to realize the automatic learning and predicting capability, the embodiment of the invention uses a convolutional neural network model with a classification regression function to predict the decomposed signals to obtain the blood pressure data (diastolic pressure and systolic pressure) of the tester; the embodiment of the invention not only avoids the complexity and the uncomfortable feeling of the conventional testing means, but also generates an automatic intelligent data analysis method, thereby leading an application party to conveniently and continuously monitor the tested object for many times.
In order to achieve the above object, a first aspect of an embodiment of the present invention provides a method for predicting blood pressure based on a photoplethysmography signal, the method comprising:
acquiring a PPG signal by a photoplethysmography, and segmenting the PPG signal to generate a PPG signal segment;
acquiring a wavelet base type, a scaling factor array and a shifting factor array; the array of scale factors comprises M scale factors; the array of shifting factors comprises N shifting factors; both M and N are integers;
performing signal decomposition processing on the PPG signal segment by using a continuous wavelet transform mode according to the expansion factor of the expansion factor array, the movement factor of the movement factor array and the wavelet base type to generate a PPG wavelet coefficient matrix [ M, N ];
performing real number matrix conversion on the PPG wavelet coefficient matrix [ M, N ] in a mode of taking a modulus of matrix elements, and performing matrix element value normalization processing on the converted matrix to generate a PPG normalization matrix [ M, N ];
acquiring an RGB color wheel matrix, and performing PPG time-frequency tensor conversion on the PPG normalization matrix [ M, N ] according to the RGB color wheel matrix to generate a PPG time-frequency three-dimensional tensor [ M, N, 3 ];
carrying out tensor shape reconstruction operation on the PPG time-frequency three-dimensional tensor [ M, N, 3] by using a bicubic interpolation algorithm according to a preset convolution network input width threshold value to generate a PPG convolution three-dimensional tensor [ Y, Y, 3 ]; y is the convolutional network input width threshold;
and performing classification regression calculation on the PPG convolution three-dimensional tensor [ Y, Y, 3] by using a convolution neural network classification regression model to generate a PPG predicted blood pressure data pair.
Preferably, the first and second liquid crystal materials are,
matrix elements of the PPG wavelet coefficient matrix [ M, N ] are wavelet coefficients in a complex form;
the value range of matrix elements of the PPG normalization matrix [ M, N ] is from 0 to 1;
the convolutional neural network classification regression model comprises: the device comprises a two-dimensional convolution layer, a maximum pooling layer, a batch homogenization layer, an activation layer, an addition layer, a global average pooling layer, a random discarding layer and a full connection layer;
the PPG predicted blood pressure data pair comprises diastolic blood pressure data and systolic blood pressure data.
Preferably, the acquiring a photoplethysmography PPG signal and segmenting the photoplethysmography PPG signal to generate PPG signal segments specifically includes:
performing signal acquisition on a tester by using PPG signal acquisition equipment according to a preset sampling frequency to generate a PPG signal; and segmenting the PPG signal according to a preset segment duration threshold value to generate a plurality of PPG signal segments.
Preferably, the performing, according to the scaling factor of the scaling factor array, the motion factor of the motion factor array, and the wavelet base type, signal decomposition processing on the PPG signal segment by using a continuous wavelet transform manner to generate a PPG wavelet coefficient matrix [ M, N ], specifically includes:
step 41, constructing a matrix by using the number of rows as the M and the number of columns as the N, generating a temporary PPG wavelet coefficient matrix [ M, N ], and initializing all matrix elements of the temporary PPG wavelet coefficient matrix [ M, N ] to be null;
step 42, initializing the value of the first index to 1;
step 43, initializing the value of the second index to 1;
step 44, extracting a scaling factor generation factor a corresponding to the first index from the scaling factor array, and extracting a shifting factor generation factor b corresponding to the second index from the shifting factor array;
step 45, using the factor a and the factor b as transformation parameters, and using a continuous wavelet transformation formula corresponding to the wavelet base type to perform continuous wavelet transformation calculation on the PPG signal segment, so as to generate a wavelet coefficient WTf(a, b); the wavelet coefficient WTf(a, b) are plural;
step 46, performing a data item addition operation on the wavelet coefficients WT (a, b) to the provisional PPG wavelet coefficient matrix [ M, N ];
step 47, adding 1 to the second index;
step 48, determining whether the second index is greater than N, if the second index is greater than N, proceeding to step 49, and if the second index is less than or equal to N, proceeding to step 44;
step 49, adding 1 to the first index;
step 50, judging whether the first index is greater than M, if so, turning to step 51, and if not, turning to step 43;
and step 51, setting the PPG wavelet coefficient matrix [ M, N ] as the temporary PPG wavelet coefficient matrix [ M, N ].
Further, the factor a and the factor b are used as transformation parameters, and a continuous wavelet transformation formula corresponding to the wavelet base type is used to perform continuous wavelet transformation calculation on the PPG signal segment, so as to generate a wavelet coefficient WTf(a, b), specifically including:
when the wavelet base type is the generalized Morse wavelet, selecting the wavelet base expansion translation function as
a, b > 0; wherein a is the factor a; the b is the factor b; the above-mentioned
Is a standard constant; e is the Euler number; said H (t) is a unit step function; the t is a time variable;
according to the wavelet basis expansion and translation function psia，b(t) using a formula for the PPG signal segment
Performing continuous wavelet transform calculation to generate the wavelet coefficient WTf(a, b); wherein R is a real number; the f (t) is the PPG signal segment.
Preferably, the performing real number matrix conversion on the PPG wavelet coefficient matrix [ M, N ] in a mode of modulo matrix elements, and performing matrix element value normalization processing on the converted matrix to generate a PPG normalization matrix [ M, N ], specifically includes:
constructing a matrix by using the number of rows as M and the number of columns as N, generating a PPG real matrix [ M, N ], and initializing all matrix elements of the PPG real matrix [ M, N ] to be null;
sequentially extracting matrix elements of the PPG wavelet coefficient matrix [ M, N ] to generate temporary wavelet coefficients, performing complex modulus calculation on the temporary wavelet coefficients to generate a wavelet coefficient modulus calculation result, and performing data item addition operation on the wavelet coefficient modulus calculation result to the PPG real matrix [ M, N ]; the wavelet coefficient modulus calculation result is a real number;
and normalizing the numerical values of all matrix elements of the PPG real number matrix [ M, N ] to generate the PPG normalization matrix [ M, N ].
Preferably, the RGB color wheel matrix is obtained; performing PPG time-frequency tensor conversion on the PPG normalization matrix [ M, N ] according to the RGB color wheel matrix to generate a PPG time-frequency three-dimensional tensor [ M, N, 3], and specifically comprising:
acquiring an RGB color disc matrix; the RGB color disc matrix is [ X, 3 ]; the RGB color wheel matrix includes the X color vectors [3 ]; x is an integer;
constructing a matrix by using the number of rows as M and the number of columns as N, generating a temporary level matrix [ M, N ], and initializing all matrix elements of the temporary level matrix [ M, N ] to be null;
equally dividing the space between 0 and 1 into the X data segments by taking the X as a quantization series; the data segment comprises a data segment index and a data segment threshold range; the value of the data segment index is from 1 to X;
sequentially extracting matrix elements of the PPG normalization matrix [ M, N ] to generate first current elements, performing polling comparison on data segment threshold ranges of all data segments by using values of the first current elements, and extracting currently-compared data segment indexes to perform data item adding operation on the temporary level matrix [ M, N ] when the values of the first current elements are within the compared data segment threshold ranges;
initializing all matrix elements of the PPG time-frequency three-dimensional tensor [ M, N, 3] to be null;
sequentially extracting matrix elements of the temporary level matrix [ M, N ] to generate a second current element, extracting a corresponding color vector [3] from the RGB color wheel matrix by taking the value of the second current element as an index to generate a current color vector [3], and performing data item adding operation on the current color vector [3] to the PPG time-frequency three-dimensional tensor [ M, N, 3 ].
Preferably, the method further comprises:
and after the PPG time-frequency three-dimensional tensor [ M, N, 3] is generated, image conversion is carried out on the PPG time-frequency three-dimensional tensor [ M, N, 3] to generate PPG time-frequency graph data.
According to the blood pressure prediction method based on the photoplethysmography signals, provided by the embodiment of the invention, the problem that a monitor cannot be continuously observed in conventional monitoring is solved by using PPG (photoplethysmography) acquisition equipment to carry out non-invasive data acquisition on the tester; in order to fully obtain effective signal data from the PPG signal, the embodiment of the invention adopts a continuous wavelet transform mode to perform signal decomposition on the PPG signal; in order to realize the automatic learning and predicting capability, the embodiment of the invention predicts the decomposition signal by using a convolutional neural network model with a classification regression function to obtain the blood pressure data (diastolic pressure and systolic pressure) of the tester.
A second aspect of an embodiment of the present invention provides an apparatus, which includes a memory and a processor, where the memory is used to store a program, and the processor is used to execute the first aspect and the method in each implementation manner of the first aspect.
A third aspect of embodiments of the present invention provides a computer program product including instructions, which, when run on a computer, cause the computer to perform the first aspect and the method in each implementation manner of the first aspect.
A fourth aspect of the embodiments of the present invention provides a computer-readable storage medium, where a computer program is stored, and when the computer program is executed by a processor, the computer program implements the first aspect and the methods in the implementation manners of the first aspect.
Drawings
FIG. 1 is a schematic diagram of a method for predicting blood pressure based on a photoplethysmography signal according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a wavelet transform time-frequency diagram generation method for photoplethysmography signals according to a second embodiment of the present invention;
fig. 3 is a schematic device structure diagram of an apparatus for blood pressure prediction based on a photoplethysmography signal according to a third embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention clearer, the present invention will be described in further detail with reference to the accompanying drawings, and it is apparent that the described embodiments are only a part of the embodiments of the present invention, not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Before the present invention is explained in further detail by examples, some technical means mentioned in the text will be briefly explained.
The PPG signal is a set of signals that uses a light-sensitive sensor to record the light intensity changes for light intensity identification of a particular light source. When the heart beats, the blood flow per unit area in the blood vessel changes periodically, and the corresponding blood volume also changes, so that the PPG signal, which reflects the amount of light absorbed by the blood, also shows a periodic change tendency. A cardiac cycle comprises two time periods: systolic and diastolic phases; when the heart contracts, the heart does work on the whole body, so that the pressure in the blood vessel and the volume of blood flow generate continuous periodic change, and the blood in the blood vessel absorbs the light most; when the heart is in diastole, the pressure to the blood vessel is relatively small, and the blood pushed out to the whole body by the last systole strikes a heart valve through circulation so as to generate certain reflection and refraction effects on light, so that the absorption of the blood in the blood vessel to the light energy is reduced during the diastole period. Therefore, the waveform of the PPG signal, which reflects the absorption energy of blood in the blood vessel, is formed by the superposition of two signals: a systolic phase signal and a diastolic phase signal; a common PPG signal has two peaks of magnitude, the first belonging to the systolic phase and the second to the diastolic phase.
After the PPG signal is obtained, a blood pressure prediction value of the current PPG signal can be obtained by utilizing a characteristic calculation and regression model which is trained by batch PPG signals and corresponding measured blood pressure data. When blood pressure is predicted, effective data extraction, so-called feature extraction or feature calculation, needs to be performed on pulse wave data in advance. And then obtaining regression data as a prediction result by using a blood pressure regression calculation mode on the obtained characteristic data. The regression data in the blood pressure regression calculation are two: diastolic pressure data and systolic pressure data, wherein the systolic pressure data is greater than the diastolic pressure data.
Here, regarding the extraction of valid data, on one hand, we can extract the signal amplitude from the signal time domain as a feature value, and on the other hand, can also extract the frequency of the change from the signal frequency domain by performing time-frequency conversion on the signal as a feature value. In the latter case, it is necessary to perform time-frequency conversion on the signals first, and then extract eigenvalues according to the conversion result to form an eigenvalue matrix. Conventionally, the time-frequency conversion method for signals is through fourier transform. But the fourier transform, because its time-frequency analysis window is a fixed size, is prone to loss of feature data for non-stationary signals. Electrophysiological signals like PPG signals, as mentioned herein, are among the non-stationary signals that are susceptible to interference.
The wavelet transform is a time-frequency analysis method, inherits the idea of Fourier transform, and can highlight local characteristics of signals simultaneously in principle. The embodiment of the invention adopts one of the wavelet transformation: and decomposing the PPG signal by a continuous wavelet transform mode. Compared with the short-time Fourier transform, the continuous wavelet transform has the window adjustable property and has higher analysis capability on non-stationary signals; the signal is subjected to multi-scale refinement through wavelet expansion translation operation, so that higher time resolution can be achieved in high-frequency components of the signal, and higher frequency resolution is achieved in low-frequency components of the signal. The continuous wavelet transform has three core parameters: wavelet basis, scaling factor, and shifting factor. The wavelet basis is a wavelet function specifically used for wavelet transformation, the scaling factor is a scale parameter which can be transformed by itself in the wavelet transformation process, and the shifting factor is a shifting time parameter which can be transformed by itself in the wavelet transformation process.
As shown in fig. 1, which is a schematic diagram of a blood pressure prediction method based on a photoplethysmography signal according to an embodiment of the present invention, the method mainly includes the following steps:
step 1, acquiring a PPG signal of a photoplethysmography, and segmenting the PPG signal to generate PPG signal segments;
the method specifically comprises the following steps: performing signal acquisition on a tester by using PPG signal acquisition equipment according to a preset sampling frequency to generate a PPG signal; and segmenting the PPG signal according to a preset segment duration threshold value to generate a plurality of PPG signal segments.
wherein the scaling factor array comprises M scaling factors; the moving factor array comprises N moving factors; m and N are integers.
wherein, the matrix elements of PPG wavelet coefficient matrix [ M, N ] are complex wavelet coefficients;
the method specifically comprises the following steps: step 31, constructing a matrix by using the number of rows as the M and the number of columns as the N, generating a temporary PPG wavelet coefficient matrix [ M, N ], and initializing all matrix elements of the temporary PPG wavelet coefficient matrix [ M, N ] to be null;
step 32, initializing the value of the first index to be 1;
step 33, initializing the value of the second index to 1;
step 34, extracting a scaling factor generation factor a corresponding to the first index from the scaling factor array, and extracting a shifting factor generation factor b corresponding to the second index from the shifting factor array;
step 35, using the factor a and the factor b as transformation parameters, and using a continuous wavelet transformation formula corresponding to the wavelet base type to perform continuous wavelet transformation calculation on the PPG signal segment, so as to generate a wavelet coefficient WTf(a，b)；
Wherein the wavelet coefficients WTf(a, b) are plural;
the method specifically comprises the following steps: step 351, when the wavelet base type is the generalized Morse wavelet, selecting the wavelet base expansion translation function as
a，b＞0；
Wherein a is the factor a; the b is the factor b; the above-mentioned
Is a standard constant; e is the Euler number; said H (t) is a unit step function; the t is a time variable;
step 352, stretch panning function psi according to said wavelet basisa，b(t) using a formula for the PPG signal segment
Performing continuous wavelet transform calculation to generate the wavelet coefficient WTf(a，b)；
Wherein R is a real number; said f (t) is said PPG signal segment;
step 36, performing a data item addition operation on the wavelet coefficients WT (a, b) to the provisional PPG wavelet coefficient matrix [ M, N ];
step 37, adding 1 to the second index;
step 38, determining whether the second index is greater than N, if the second index is greater than N, proceeding to step 39, and if the second index is less than or equal to N, proceeding to step 34;
step 39, adding 1 to the first index;
step 40, determining whether the first index is greater than M, if the first index is greater than M, proceeding to step 41, and if the first index is less than or equal to M, proceeding to step 33;
and step 41, setting the PPG wavelet coefficient matrix [ M, N ] as the temporary PPG wavelet coefficient matrix [ M, N ].
wherein, the value range of matrix elements of the PPG normalization matrix [ M, N ] is from 0 to 1;
the method specifically comprises the following steps: step 42, constructing a matrix by using the number of rows as the M and the number of columns as the N, generating a PPG real number matrix [ M, N ], and initializing all matrix elements of the PPG real number matrix [ M, N ] to be null;
step 43, sequentially extracting matrix elements of the PPG wavelet coefficient matrix [ M, N ] to generate temporary wavelet coefficients, performing complex modulo calculation on the temporary wavelet coefficients to generate wavelet coefficient modulo calculation results, and performing data item addition operation on the wavelet coefficient modulo calculation results to the PPG real matrix [ M, N ];
wherein the wavelet coefficient modulus calculation result is a real number;
and 44, normalizing the numerical values of all matrix elements of the PPG real number matrix [ M, N ] to generate the PPG normalization matrix [ M, N ].
the method specifically comprises the following steps: step 51, acquiring an RGB color disc matrix;
wherein, the RGB color wheel matrix is [ X, 3 ]; the RGB color wheel matrix includes the X color vectors [3 ]; x is an integer;
the RGB color scheme is a color standard in the industry, which obtains various colors by changing three color channels of red (R), green (G) and blue (B) and superimposing them with each other, and the standard almost includes all colors that can be perceived by human vision, and is one of the most widely used color systems; assuming that the RGB color wheel matrix includes 256 color vectors, the length of each color vector is 3, and includes the values of three primary colors respectively; assuming that X is equal to 256, the RGB color wheel matrix comprises 256 colors;
step 52, constructing a matrix by using the number of rows as the M and the number of columns as the N, generating a temporary level matrix [ M, N ], and initializing all matrix elements of the temporary level matrix [ M, N ] to be null;
step 53, equally dividing the space between 0 and 1 into the X data segments by taking the X as a quantization level; the data segment comprises a data segment index and a data segment threshold range; the value of the data segment index is from 1 to X;
step 54, sequentially extracting matrix elements of the PPG normalization matrix [ M, N ] to generate a first current element, performing polling comparison on threshold ranges of data segments of all data segments by using values of the first current element, and extracting indexes of the currently compared data segments to perform data item adding operation on the temporary level matrix [ M, N ] when the values of the first current element are within the threshold ranges of the compared data segments;
here, suppose X is 256, divided into 256 pieces of data on average from 0 to 1, 0 to 1/256 are first pieces, 1/256 to 2/256 are second pieces, and so on, 255/256 to 1 are 256-th pieces; comparing all elements in the PPG normalization matrix [ M, N ] with the data section according to the sampling value, if the value of a certain element is 1/257, the certain element belongs to the first section, the level of the element is 1, namely the value of the element corresponding to the element in the temporary level matrix [ M, N ] is 1;
step 55, initializing all matrix elements of the PPG time-frequency three-dimensional tensor [ M, N, 3] to be null;
and 56, sequentially extracting matrix elements of the temporary level matrix [ M, N ] to generate a second current element, extracting a corresponding color vector [3] from the RGB color wheel matrix by taking the value of the second current element as an index to generate a current color vector [3], and performing data item adding operation on the current color vector [3] to the PPG time-frequency three-dimensional tensor [ M, N, 3 ].
All element values in the temporary level matrix [ M, N ] are integers of 1-256, a corresponding color point can be extracted from the RGB color wheel matrix according to the value, the color point is used as a one-dimensional vector with dimension supplement to perform matrix dimension-increasing processing on the temporary level matrix [ M, N ], and a PPG time-frequency three-dimensional tensor [ M, N, 3] is generated, wherein the actual PPG time-frequency three-dimensional tensor [ M, N, 3] is a three-dimensional tensor composed of M x N color points.
After continuous wavelet transform is performed by using generalized Morse wavelet, an original signal is decomposed into a two-dimensional complex matrix containing wavelet coefficients, wherein each row corresponds to a single scale factor (scale factor), namely a frequency band obtained by dividing a specified octave; then, quantizing the wavelet coefficient, wherein the specific process is to perform modulus operation on each element of the complex matrix, normalize the real matrix obtained by modulus operation, and finally obtain a matrix with an element value range; and then mapping the matrix elements into a two-dimensional plane, mapping the matrix elements into three-dimensional RGB color values through a specified color space, and carrying out size adjustment on the picture to adapt to the input of the convolutional neural network.
Step 6, carrying out tensor shape reconstruction operation on the PPG time-frequency three-dimensional tensor [ M, N, 3] by using a bicubic interpolation algorithm according to a preset convolution network input width threshold value to generate a PPG convolution three-dimensional tensor [ Y, Y, 3 ];
where Y is the convolutional network input width threshold.
Here, it is possible that the size of the PPG time-frequency three-dimensional tensor [ M, N, 3] is deviated from the input size requirement of the convolutional neural network classification regression model, and when the size of the PPG time-frequency three-dimensional tensor [ M, N, 3] is smaller, the intermediate value points are increased by using a bicubic interpolation algorithm to achieve the effect of changing the shape of the three-dimensional tensor, and finally, the PPG convolutional three-dimensional tensor [ Y, 3] meeting the requirement is generated.
Step 7, carrying out classification regression calculation on PPG convolution three-dimensional tensors [ Y, Y and 3] by using a convolution neural network classification regression model to generate PPG prediction blood pressure data pairs;
wherein the PPG predicted blood pressure data pair comprises diastolic blood pressure data and systolic blood pressure data.
Here, the convolutional network used is a custom convolutional network structure, and the convolutional neural network classification regression model includes: the two-dimensional convolution layer, the maximum pooling layer, the batch homogenization layer, the activation layer, the addition layer, the global averaging pooling layer, the random discarding layer and the full-connection layer can finally realize the simultaneous output of the regression model of the diastolic pressure and the systolic pressure by modifying the network structure.
As shown in fig. 2, which is a schematic diagram of a wavelet transform time-frequency diagram generation method for photoplethysmography signals provided by the second embodiment of the present invention, the method mainly includes the following steps:
step 101, collecting a signal by a PPG signal collecting device of a tester according to a preset sampling frequency by using a photoplethysmography to generate a PPG signal; and segmenting the PPG signal according to a preset segment duration threshold value to generate a plurality of PPG signal segments.
102, acquiring a wavelet base type, a telescopic factor array and a moving factor array;
wherein the scaling factor array comprises M scaling factors; the moving factor array comprises N moving factors; m and N are integers.
103, performing signal decomposition processing on the PPG signal segment by using a continuous wavelet transform mode according to the expansion factor of the expansion factor array, the movement factor of the movement factor array and the wavelet base type to generate a PPG wavelet coefficient matrix [ M, N ];
wherein, the matrix elements of PPG wavelet coefficient matrix [ M, N ] are complex wavelet coefficients;
the method specifically comprises the following steps: step 1031, constructing a matrix according to the number of rows M and the number of columns N, generating a temporary PPG wavelet coefficient matrix [ M, N ], and initializing all matrix elements of the temporary PPG wavelet coefficient matrix [ M, N ] to be null;
step 1032, initializing the value of the first index to 1;
step 1033, initializing a value of the second index to 1;
step 1034, extracting a scaling factor generation factor a corresponding to the first index from the scaling factor array, and extracting a moving factor generation factor b corresponding to the second index from the moving factor array;
step 1035, taking the factor a and the factor b as transformation parameters, and using a continuous wavelet transformation formula corresponding to the wavelet base type to perform continuous wavelet transformation calculation on the PPG signal segment, so as to generate a wavelet coefficient WTf(a, b); wavelet coefficient WTf(a, b) are plural;
here, specifically: when the wavelet base type is the generalized Morse wavelet, selecting the wavelet base expansion translation function as
a, b > 0; wherein a is a factor a; b is a factor b;
is a standard constant; e is the Euler number; h (t) is a unit step function; t is a time variable;
according to wavelet basis expansion and translation function psia，b(t) using the formula for PPG signal segments
Performing continuous wavelet transform to calculate wavelet coefficient WTf(a, b); wherein R is a real number; f (t) is a PPG signal segment;
step 1036, performing a data item addition operation on the wavelet coefficients WT (a, b) to the provisional PPG wavelet coefficient matrix [ M, N ];
step 1037, add 1 to the second index;
step 1038, determine whether the second index is greater than N, go to step 1039 if the second index is greater than N, go to step 1034 if the second index is less than or equal to N;
step 1039, add 1 to the first index;
step 1040, determining whether the first index is greater than M, if the first index is greater than M, going to step 1041, and if the first index is less than or equal to M, going to step 1033;
step 1041, setting PPG wavelet coefficient matrix [ M, N ] as temporary PPG wavelet coefficient matrix [ M, N ].
104, performing real number matrix conversion on the PPG wavelet coefficient matrix [ M, N ] in a mode of taking a modulus of matrix elements, and performing numerical value normalization processing on the converted matrix to generate a PPG normalization matrix [ M, N ];
wherein, the value range of the matrix elements of the PPG normalization matrix [ M, N ] is from 0 to 1.
And 105, acquiring an RGB color wheel matrix, and performing PPG time-frequency tensor conversion on the PPG normalization matrix [ M, N ] according to the RGB color wheel matrix to generate a PPG time-frequency three-dimensional tensor [ M, N, 3 ].
The RGB color scheme is a color standard in the industry, which obtains various colors by changing three color channels of red (R), green (G) and blue (B) and superimposing them with each other, and the standard almost includes all colors that can be perceived by human vision, and is one of the most widely used color systems; assuming that the RGB color wheel matrix includes 256 color vectors, the length of each color vector is 3, and includes the values of three primary colors respectively; assuming that X is equal to 256, the RGB color wheel matrix comprises 256 colors;
all the element values in the temporary level matrix [ M, N ] are integers of 1-256, a corresponding color point can be extracted from the RGB color wheel matrix according to the value, the color point is used as a one-dimensional vector for dimension supplement to perform matrix dimension-increasing processing on the temporary level matrix [ M, N ], and a PPG time-frequency three-dimensional tensor [ M, N, 3] is generated, wherein the actual PPG time-frequency three-dimensional tensor [ M, N, 3] is a three-dimensional tensor formed by M × N color points.
After continuous wavelet transform is performed by using generalized Morse wavelet, an original signal is decomposed into a two-dimensional complex matrix containing wavelet coefficients, wherein each row corresponds to a single scale factor (scale factor), namely a frequency band obtained by dividing a specified octave; then, quantizing the wavelet coefficient, wherein the specific process is to perform modulus operation on each element of the complex matrix, normalize the real matrix obtained by modulus operation, and finally obtain a matrix with an element value range; and then mapping the matrix elements into a two-dimensional plane, mapping the matrix elements into three-dimensional RGB color values through a specified color space, and carrying out size adjustment on the picture to adapt to the input of the convolutional neural network.
And step 106, carrying out image conversion on the PPG time-frequency three-dimensional tensor [ M, N, 3] to generate PPG time-frequency graph data.
If the PPG time-frequency three-dimensional tensor [ M, N, 3] is not large enough as image data, a bicubic interpolation algorithm can be used for adding pixel points between points so as to achieve the effect of amplifying the image. Here, the bicubic interpolation method is to expand 4 × 4 pixels around a certain original pixel on the basis of the original pixel. Similarly, if the PPG time-frequency three-dimensional tensor [ M, N, 3] is large enough as image data and needs to be reduced, the abbreviation can be performed by using a bicubic interpolation algorithm. For example, the PPG time-frequency three-dimensional tensor [ M, N, 3] is [224, 128, 3], indicating that the original image is a bitmap of 224 × 128 size, and we can enlarge or adjust the graph to 448 × 256 or 224 × 224 size by bicubic interpolation.
As shown in fig. 3, which is a schematic structural diagram of an apparatus for blood pressure prediction based on a photoplethysmography signal according to a third embodiment of the present invention, the apparatus includes: a processor and a memory. The memory may be connected to the processor by a bus. The memory may be a non-volatile memory such as a hard disk drive and a flash memory, in which a software program and a device driver are stored. The software program is capable of performing various functions of the above-described methods provided by embodiments of the present invention; the device drivers may be network and interface drivers. The processor is used for executing a software program, and the software program can realize the method provided by the embodiment of the invention when being executed.
It should be noted that the embodiment of the present invention also provides a computer-readable storage medium. The computer readable storage medium has a computer program stored thereon, and the computer program can realize the method provided by the embodiment of the invention when being executed by a processor.
The embodiment of the invention also provides a computer program product containing the instruction. The computer program product causes a processor to perform the above-mentioned method when run on a computer.
According to the blood pressure prediction method and device based on the photoplethysmography signals, provided by the embodiment of the invention, the non-invasive data acquisition is carried out on a tester by using PPG acquisition equipment, so that the problem that the monitor cannot be continuously observed in conventional monitoring is solved; in order to fully obtain effective signal data from the PPG signal, the embodiment of the invention adopts a continuous wavelet transform mode to perform signal decomposition on the PPG signal; in order to realize the automatic learning and predicting capability, the embodiment of the invention uses a convolutional neural network model with a classification regression function to predict the decomposed signals to obtain the blood pressure data (diastolic pressure and systolic pressure) of the tester; the embodiment of the invention not only avoids the complexity and the uncomfortable feeling of the conventional testing means, but also generates an automatic intelligent data analysis method, thereby leading an application party to conveniently and continuously monitor the tested object for many times.
Those of skill would further appreciate that the various illustrative components and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both, and that the various illustrative components and steps have been described above generally in terms of their functionality in order to clearly illustrate this interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
The steps of a method or algorithm described in connection with the embodiments disclosed herein may be embodied in hardware, a software module executed by a processor, or a combination of the two. A software module may reside in Random Access Memory (RAM), memory, Read Only Memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.
The above-mentioned embodiments are intended to illustrate the objects, technical solutions and advantages of the present invention in further detail, and it should be understood that the above-mentioned embodiments are merely exemplary embodiments of the present invention, and are not intended to limit the scope of the present invention, and any modifications, equivalent substitutions, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (11)
1. A method of blood pressure prediction based on a photoplethysmography signal, the method comprising:
acquiring a PPG signal by a photoplethysmography, and segmenting the PPG signal to generate a PPG signal segment;
acquiring a wavelet base type, a scaling factor array and a shifting factor array; the array of scale factors comprises M scale factors; the array of shifting factors comprises N shifting factors; both M and N are integers;
performing signal decomposition processing on the PPG signal segment by using a continuous wavelet transform mode according to the expansion factor of the expansion factor array, the movement factor of the movement factor array and the wavelet base type to generate a PPG wavelet coefficient matrix [ M, N ];
performing real number matrix conversion on the PPG wavelet coefficient matrix [ M, N ] in a mode of taking a modulus of matrix elements, and performing matrix element value normalization processing on the converted matrix to generate a PPG normalization matrix [ M, N ];
acquiring an RGB color wheel matrix, and performing PPG time-frequency tensor conversion on the PPG normalization matrix [ M, N ] according to the RGB color wheel matrix to generate a PPG time-frequency three-dimensional tensor [ M, N, 3 ];
carrying out tensor shape reconstruction operation on the PPG time-frequency three-dimensional tensor [ M, N, 3] by using a bicubic interpolation algorithm according to a preset convolution network input width threshold value to generate a PPG convolution three-dimensional tensor [ Y, Y, 3 ]; y is the convolutional network input width threshold;
and performing classification regression calculation on the PPG convolution three-dimensional tensor [ Y, Y, 3] by using a convolution neural network classification regression model to generate a PPG predicted blood pressure data pair.
2. A method of blood pressure prediction based on a photoplethysmography signal according to claim 1,
matrix elements of the PPG wavelet coefficient matrix [ M, N ] are wavelet coefficients in a complex form;
the value range of matrix elements of the PPG normalization matrix [ M, N ] is from 0 to 1;
the convolutional neural network classification regression model comprises: the device comprises a two-dimensional convolution layer, a maximum pooling layer, a batch homogenization layer, an activation layer, an addition layer, a global average pooling layer, a random discarding layer and a full connection layer;
the PPG predicted blood pressure data pair comprises diastolic blood pressure data and systolic blood pressure data.
3. The method for blood pressure prediction based on photoplethysmography signals according to claim 1, wherein the obtaining and fragmenting photoplethysmography PPG signals generates PPG signal fragments, in particular comprising:
performing signal acquisition on a tester by using PPG signal acquisition equipment according to a preset sampling frequency to generate a PPG signal; and segmenting the PPG signal according to a preset segment duration threshold value to generate a plurality of PPG signal segments.
4. The method for predicting blood pressure based on photoplethysmography signals according to claim 1, wherein the signal decomposition processing is performed on the PPG signal segments by using a continuous wavelet transform method according to the scale factors of the scale factor array, the motion factors of the motion factor array and the wavelet base type to generate a PPG wavelet coefficient matrix [ M, N ], which specifically includes:
step 41, constructing a matrix by using the number of rows as the M and the number of columns as the N, generating a temporary PPG wavelet coefficient matrix [ M, N ], and initializing all matrix elements of the temporary PPG wavelet coefficient matrix [ M, N ] to be null;
step 42, initializing the value of the first index to 1;
step 43, initializing the value of the second index to 1;
step 44, extracting a scaling factor generation factor a corresponding to the first index from the scaling factor array, and extracting a shifting factor generation factor b corresponding to the second index from the shifting factor array;
step 45, using the factor a and the factor b as transformation parameters, and using the small valuesThe continuous wavelet transform formula corresponding to the wave base type is used for carrying out continuous wavelet transform calculation on the PPG signal segment to generate a wavelet coefficient WTf(a, b); the wavelet coefficient WTf(a, b) are plural;
step 46, performing a data item addition operation on the wavelet coefficients WT (a, b) to the provisional PPG wavelet coefficient matrix [ M, N ];
step 47, adding 1 to the second index;
step 48, determining whether the second index is greater than N, if the second index is greater than N, proceeding to step 49, and if the second index is less than or equal to N, proceeding to step 44;
step 49, adding 1 to the first index;
step 50, judging whether the first index is greater than M, if so, turning to step 51, and if not, turning to step 43;
and step 51, setting the PPG wavelet coefficient matrix [ M, N ] as the temporary PPG wavelet coefficient matrix [ M, N ].
5. The method of claim 4, wherein the factor a and the factor b are used as transformation parameters to perform continuous wavelet transform calculation on the PPG signal slice by using a continuous wavelet transform formula corresponding to the wavelet base type to generate a wavelet coefficient WTf(a, b), specifically including:
when the wavelet base type is the generalized Morse wavelet, selecting the wavelet base expansion translation function as
Wherein a is the factor a; the b is the factor b; the above-mentioned
Is a standard constant; e is the Euler number; said H (t) is a unit step function; t is timeA variable;
according to the wavelet basis expansion and translation function psia，b(t) using a formula for the PPG signal segment
Performing continuous wavelet transform calculation to generate the wavelet coefficient WTf(a, b); wherein R is a real number; the f (t) is the PPG signal segment.
6. The method for predicting blood pressure based on photoplethysmography signals according to claim 1, wherein the real matrix conversion is performed on the PPG wavelet coefficient matrix [ M, N ] by taking a modulus of matrix elements, and the converted matrix is subjected to matrix element value normalization processing to generate a PPG normalization matrix [ M, N ], which specifically includes:
constructing a matrix by using the number of rows as M and the number of columns as N, generating a PPG real matrix [ M, N ], and initializing all matrix elements of the PPG real matrix [ M, N ] to be null;
sequentially extracting matrix elements of the PPG wavelet coefficient matrix [ M, N ] to generate temporary wavelet coefficients, performing complex modulus calculation on the temporary wavelet coefficients to generate a wavelet coefficient modulus calculation result, and performing data item addition operation on the wavelet coefficient modulus calculation result to the PPG real matrix [ M, N ]; the wavelet coefficient modulus calculation result is a real number;
and normalizing the numerical values of all matrix elements of the PPG real number matrix [ M, N ] to generate the PPG normalization matrix [ M, N ].
7. A method for blood pressure prediction based on photoplethysmography signals according to claim 1, wherein said obtaining an RGB color wheel matrix; performing PPG time-frequency tensor conversion on the PPG normalization matrix [ M, N ] according to the RGB color wheel matrix to generate a PPG time-frequency three-dimensional tensor [ M, N, 3], and specifically comprising:
acquiring an RGB color disc matrix; the RGB color disc matrix is [ X, 3 ]; the RGB color wheel matrix includes the X color vectors [3 ]; x is an integer;
constructing a matrix by using the number of rows as M and the number of columns as N, generating a temporary level matrix [ M, N ], and initializing all matrix elements of the temporary level matrix [ M, N ] to be null;
equally dividing the space between 0 and 1 into the X data segments by taking the X as a quantization series; the data segment comprises a data segment index and a data segment threshold range; the value of the data segment index is from 1 to X;
sequentially extracting matrix elements of the PPG normalization matrix [ M, N ] to generate first current elements, performing polling comparison on data segment threshold ranges of all data segments by using values of the first current elements, and extracting currently-compared data segment indexes to perform data item adding operation on the temporary level matrix [ M, N ] when the values of the first current elements are within the compared data segment threshold ranges;
initializing all matrix elements of the PPG time-frequency three-dimensional tensor [ M, N, 3] to be null;
sequentially extracting matrix elements of the temporary level matrix [ M, N ] to generate a second current element, extracting a corresponding color vector [3] from the RGB color wheel matrix by taking the value of the second current element as an index to generate a current color vector [3], and performing data item adding operation on the current color vector [3] to the PPG time-frequency three-dimensional tensor [ M, N, 3 ].
8. A method of blood pressure prediction based on a photoplethysmography signal according to claim 1, the method further comprising:
and after the PPG time-frequency three-dimensional tensor [ M, N, 3] is generated, image conversion is carried out on the PPG time-frequency three-dimensional tensor [ M, N, 3] to generate PPG time-frequency graph data.
9. An apparatus comprising a memory for storing a program and a processor for performing the method of any one of claims 1 to 8.
10. A computer program product comprising instructions which, when run on a computer, cause the computer to perform the method of any one of claims 1 to 8.
11. A computer-readable storage medium comprising instructions that, when executed on a computer, cause the computer to perform the method of any of claims 1 to 8.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010110279.4A CN111248880B (en) | 2020-02-21 | 2020-02-21 | Blood pressure prediction method and device based on photoplethysmography signals |
| PCT/CN2020/129636 WO2021164349A1 (en) | 2020-02-21 | 2020-11-18 | Blood pressure prediction method and apparatus based on photoplethysmography signal |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010110279.4A CN111248880B (en) | 2020-02-21 | 2020-02-21 | Blood pressure prediction method and device based on photoplethysmography signals |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111248880A true CN111248880A (en) | 2020-06-09 |
| CN111248880B CN111248880B (en) | 2022-07-29 |
Family
ID=70949401
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010110279.4A Active CN111248880B (en) | 2020-02-21 | 2020-02-21 | Blood pressure prediction method and device based on photoplethysmography signals |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN111248880B (en) |
| WO (1) | WO2021164349A1 (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112022125A (en) * | 2020-09-28 | 2020-12-04 | 无锡博智芯科技有限公司 | Intelligent blood pressure prediction method based on CNN-BiGRU model and PPG |
| CN112022126A (en) * | 2020-09-28 | 2020-12-04 | 无锡博智芯科技有限公司 | Intelligent blood pressure prediction method based on CNN-BilSTM model and PPG |
| WO2021164349A1 (en) * | 2020-02-21 | 2021-08-26 | 乐普（北京）医疗器械股份有限公司 | Blood pressure prediction method and apparatus based on photoplethysmography signal |
| WO2022077888A1 (en) * | 2020-10-12 | 2022-04-21 | 乐普（北京）医疗器械股份有限公司 | Scatter diagram classification method and apparatus for photoplethysmography signal |
| CN115444377A (en) * | 2022-10-19 | 2022-12-09 | 北京工业大学 | Method for extracting pulse wave three-dimensional parameters |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118242569B (en) * | 2024-05-27 | 2024-08-06 | 四川旷想科技有限公司 | Wireless intelligent security monitoring method for high-consequence area of pipeline |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107341769A (en) * | 2016-05-03 | 2017-11-10 | 中国科学院微电子研究所 | Electrocardiosignal denoising method and system |
| CN108491863A (en) * | 2018-02-27 | 2018-09-04 | 南京信息工程大学 | Color image processing method based on Non-negative Matrix Factorization and convolutional neural networks |
| CN109589101A (en) * | 2019-01-16 | 2019-04-09 | 四川大学 | A kind of contactless physiological parameter acquisition methods and device based on video |
| CN109872817A (en) * | 2019-02-28 | 2019-06-11 | 泉州师范学院 | A kind of blood pressure prediction technique based on multifactor clue network |
| CN110135510A (en) * | 2019-05-22 | 2019-08-16 | 电子科技大学中山学院 | Dynamic domain self-adaptive method, equipment and computer readable storage medium |
| CN110367950A (en) * | 2019-07-22 | 2019-10-25 | 西安爱特眼动信息科技有限公司 | Contactless physiologic information detection method and system |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110077484A1 (en) * | 2009-09-30 | 2011-03-31 | Nellcor Puritan Bennett Ireland | Systems And Methods For Identifying Non-Corrupted Signal Segments For Use In Determining Physiological Parameters |
| EP3558101B1 (en) * | 2016-12-21 | 2022-06-08 | Emory University | Methods and systems for determining abnormal cardiac activity |
| CN109009033A (en) * | 2018-07-10 | 2018-12-18 | 华南理工大学 | A kind of blood pressure prediction technique based on wavelet analysis and echo state network |
| CN109726751B (en) * | 2018-12-21 | 2020-11-27 | 北京工业大学 | Method for recognizing electroencephalogram based on deep convolutional neural network |
| CN111248880B (en) * | 2020-02-21 | 2022-07-29 | 乐普（北京）医疗器械股份有限公司 | Blood pressure prediction method and device based on photoplethysmography signals |
| CN111358455B (en) * | 2020-03-17 | 2022-07-29 | 乐普（北京）医疗器械股份有限公司 | Blood pressure prediction method and device with multiple data sources |
-
2020
- 2020-02-21 CN CN202010110279.4A patent/CN111248880B/en active Active
- 2020-11-18 WO PCT/CN2020/129636 patent/WO2021164349A1/en active Application Filing
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107341769A (en) * | 2016-05-03 | 2017-11-10 | 中国科学院微电子研究所 | Electrocardiosignal denoising method and system |
| CN108491863A (en) * | 2018-02-27 | 2018-09-04 | 南京信息工程大学 | Color image processing method based on Non-negative Matrix Factorization and convolutional neural networks |
| CN109589101A (en) * | 2019-01-16 | 2019-04-09 | 四川大学 | A kind of contactless physiological parameter acquisition methods and device based on video |
| CN109872817A (en) * | 2019-02-28 | 2019-06-11 | 泉州师范学院 | A kind of blood pressure prediction technique based on multifactor clue network |
| CN110135510A (en) * | 2019-05-22 | 2019-08-16 | 电子科技大学中山学院 | Dynamic domain self-adaptive method, equipment and computer readable storage medium |
| CN110367950A (en) * | 2019-07-22 | 2019-10-25 | 西安爱特眼动信息科技有限公司 | Contactless physiologic information detection method and system |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2021164349A1 (en) * | 2020-02-21 | 2021-08-26 | 乐普（北京）医疗器械股份有限公司 | Blood pressure prediction method and apparatus based on photoplethysmography signal |
| CN112022125A (en) * | 2020-09-28 | 2020-12-04 | 无锡博智芯科技有限公司 | Intelligent blood pressure prediction method based on CNN-BiGRU model and PPG |
| CN112022126A (en) * | 2020-09-28 | 2020-12-04 | 无锡博智芯科技有限公司 | Intelligent blood pressure prediction method based on CNN-BilSTM model and PPG |
| WO2022077888A1 (en) * | 2020-10-12 | 2022-04-21 | 乐普（北京）医疗器械股份有限公司 | Scatter diagram classification method and apparatus for photoplethysmography signal |
| CN115444377A (en) * | 2022-10-19 | 2022-12-09 | 北京工业大学 | Method for extracting pulse wave three-dimensional parameters |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111248880B (en) | 2022-07-29 |
| WO2021164349A1 (en) | 2021-08-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111248880B (en) | Blood pressure prediction method and device based on photoplethysmography signals | |
| Lahoud et al. | Zero-learning fast medical image fusion | |
| Du et al. | An overview of multi-modal medical image fusion | |
| EP4122382A1 (en) | Blood pressure prediction method and device using multiple data sources | |
| CN110269600B (en) | Non-contact video heart rate detection method based on multivariate empirical mode decomposition and combined blind source separation | |
| Montalt-Tordera et al. | Machine learning in magnetic resonance imaging: image reconstruction | |
| US8014616B2 (en) | System and method for fixed point continuation for total variation based compressed sensing imaging | |
| CN111248882B (en) | Method and device for predicting blood pressure | |
| TWI742408B (en) | Method and electronic apparatus for image processing | |
| CN112914527B (en) | Arterial blood pressure signal acquisition method based on pulse wave photoplethysmography | |
| US20080045847A1 (en) | Non-contact and passive measurement of arterial pulse through thermal IR imaging, and analysis of thermal IR imagery | |
| CN105147274A (en) | Method for extracting heart rate from visible spectrum section face video signal | |
| CN111248883B (en) | Blood pressure prediction method and device | |
| CN112806977B (en) | Physiological parameter measuring method based on multi-scale fusion network | |
| CN112788200B (en) | Method and device for determining frequency spectrum information, storage medium and electronic device | |
| Chekmenev et al. | Multiresolution approach for noncontact measurements of arterial pulse using thermal imaging | |
| JP6201520B2 (en) | Gaze analysis system and method using physiological indices | |
| CN112971770B (en) | Method and system for controlling and processing quality of cardiac shock signal | |
| CN115905819B (en) | rPPG signal generation method and device based on generation countermeasure network | |
| US20150182140A1 (en) | Arterial pulse analysis method and system thereof | |
| CN114145725B (en) | PPG sampling rate estimation method based on noninvasive continuous blood pressure measurement | |
| KR102493242B1 (en) | Method and System for Judging Aortic Valve Stenosis Risk and Other Cardiovascular Diseases Risk from Photoplethysmography through Artificial Intelligence Learning | |
| Mu et al. | A new pansharp based method for PET/CT image fusion | |
| CN103020910B (en) | Based on anisotropy and isotropic retinal images unique point Enhancement Method | |
| Martinez et al. | Strategic attention learning for modality translation |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |