CN115956889A - Blood pressure monitoring method and device and electronic equipment - Google Patents
Blood pressure monitoring method and device and electronic equipment Download PDFInfo
- Publication number
- CN115956889A CN115956889A CN202211554662.4A CN202211554662A CN115956889A CN 115956889 A CN115956889 A CN 115956889A CN 202211554662 A CN202211554662 A CN 202211554662A CN 115956889 A CN115956889 A CN 115956889A
- Authority
- CN
- China
- Prior art keywords
- array
- blood pressure
- value
- ppg signal
- pressure
- 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
- 230000036772 blood pressure Effects 0.000 title claims abstract description 199
- 238000000034 method Methods 0.000 title claims abstract description 70
- 238000012544 monitoring process Methods 0.000 title claims abstract description 70
- 230000035488 systolic blood pressure Effects 0.000 claims abstract description 102
- 230000035487 diastolic blood pressure Effects 0.000 claims abstract description 98
- 239000011159 matrix material Substances 0.000 claims abstract description 50
- 230000002123 temporal effect Effects 0.000 claims abstract description 38
- 238000007781 pre-processing Methods 0.000 claims abstract description 31
- 230000004927 fusion Effects 0.000 claims abstract description 23
- 238000012806 monitoring device Methods 0.000 claims abstract description 10
- 238000013186 photoplethysmography Methods 0.000 claims description 156
- 238000012549 training Methods 0.000 claims description 30
- 238000012545 processing Methods 0.000 claims description 29
- 238000013528 artificial neural network Methods 0.000 claims description 15
- 230000015654 memory Effects 0.000 claims description 15
- 238000004422 calculation algorithm Methods 0.000 claims description 10
- 238000012417 linear regression Methods 0.000 claims description 9
- 238000003860 storage Methods 0.000 claims description 9
- 238000010586 diagram Methods 0.000 description 18
- 238000000605 extraction Methods 0.000 description 18
- 238000004590 computer program Methods 0.000 description 11
- 230000008569 process Effects 0.000 description 8
- 238000004891 communication Methods 0.000 description 6
- 230000006870 function Effects 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 5
- 238000012360 testing method Methods 0.000 description 5
- 238000013527 convolutional neural network Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 230000003205 diastolic effect Effects 0.000 description 3
- 108010076504 Protein Sorting Signals Proteins 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 2
- 239000013256 coordination polymer Substances 0.000 description 2
- 238000013136 deep learning model Methods 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000006403 short-term memory Effects 0.000 description 2
- 240000006829 Ficus sundaica Species 0.000 description 1
- 208000001953 Hypotension Diseases 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000009530 blood pressure measurement Methods 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 238000005056 compaction Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 230000007787 long-term memory Effects 0.000 description 1
- 208000012866 low blood pressure Diseases 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 230000035479 physiological effects, processes and functions Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Landscapes
- Measuring Pulse, Heart Rate, Blood Pressure Or Blood Flow (AREA)
Abstract
The application provides a blood pressure monitoring method, a blood pressure monitoring device and electronic equipment; the method comprises the following steps: preprocessing an array photoelectric pulse wave (PPG) signal to obtain a corresponding standardized array PPG signal and an adjacent matrix of a graph structure corresponding to the array PPG signal; determining a spatial signature and a temporal signature corresponding to the array PPG signal based on the normalized array PPG signal and the adjacency matrix; performing feature fusion on the spatial features and the temporal features to obtain space-time features; determining a blood pressure prediction value based on the space-time feature; and determining a systolic blood pressure interval and a diastolic blood pressure interval corresponding to the array PPG signal based on the blood pressure predicted value. The blood pressure monitoring method has high accuracy.
Description
Technical Field
The present application relates to the field of blood pressure monitoring technologies, and in particular, to a blood pressure monitoring method and apparatus, and an electronic device.
Background
Currently, in the blood pressure monitoring technology, CNN (Convolutional Neural Networks) can automatically extract discriminative features and perform regression of systolic pressure and diastolic pressure on the extracted features, so as to realize end-to-end blood pressure measurement based on regression prediction. However, when monitoring systolic pressure and diastolic compaction of blood pressure by using an array PPG (photoplethysmography) signal, since the correlation of the PPG signal in the array is irregular, the conventional euclidean distance cannot be used for accurate description, and the CNN model cannot be used for sufficiently extracting feature information with discrimination, and also causes a problem of low blood pressure monitoring accuracy.
Therefore, an algorithm for monitoring the array PPG signals at different systolic pressures and diastolic pressures needs to be provided, which can solve the problem that the real-time monitoring accuracy of the array PPG signals at different systolic pressure and diastolic pressure values is not high in the conventional deep learning method.
Disclosure of Invention
The embodiment of the application provides a blood pressure monitoring method, a blood pressure monitoring device and electronic equipment, which can improve the accuracy of blood pressure monitoring.
The technical scheme of the embodiment of the application is realized as follows:
in a first aspect, an embodiment of the present application provides a blood pressure monitoring method, including:
preprocessing the array PPG signal to obtain a standardized array PPG signal and an adjacent matrix of a graph structure corresponding to the array PPG signal;
determining a spatial feature and a temporal feature corresponding to the array PPG signal based on the normalized array PPG signal and the adjacency matrix;
performing feature fusion on the spatial features and the temporal features to obtain space-time features;
determining a blood pressure prediction value based on the space-time feature;
and determining a systolic blood pressure interval and a diastolic blood pressure interval corresponding to the array PPG signal based on the blood pressure predicted value.
In the above scheme, the preprocessing the array PPG signal to obtain a normalized array PPG signal and an adjacency matrix of a graph structure corresponding to the array PPG signal includes:
and carrying out standardization processing on the array PPG signal to obtain the standardized array PPG signal.
In the above scheme, the preprocessing the array PPG signal to obtain a normalized array PPG signal and an adjacency matrix of a graph structure corresponding to the array PPG signal includes:
selecting an array PPG signal of a first length based on a sliding window;
and constructing the adjacency matrix aiming at the array PPG signal with the first length selected each time based on the average value of the sum of the maximum mutual information coefficients between each path of PPG signal and the adjacent PPG signals.
In the above scheme, the determining spatial and temporal features corresponding to the array PPG signals based on the normalized array PPG signals and the adjacency matrix includes:
obtaining a spatial feature corresponding to the array PPG signal based on a graph convolution neural network;
and obtaining the corresponding time characteristics of the array PPG signals based on a long-short term memory neural network.
In the above scheme, the performing feature fusion on the spatial feature and the temporal feature to obtain a space-time feature includes:
and performing feature fusion on the spatial features and the temporal features through a full connection layer to obtain the space-time features.
In the above scheme, the determining a predicted blood pressure value based on the space-time characteristic includes:
and predicting the space-time characteristics based on a linear regression algorithm to obtain a predicted value of systolic pressure and a predicted value of diastolic pressure.
In the above scheme, the determining, based on the predicted blood pressure value, a blood pressure interval of a systolic pressure and a blood pressure interval of a diastolic pressure corresponding to the array PPG signal includes:
determining a first upper boundary value and a first lower boundary value corresponding to the systolic pressure based on the predicted value of the systolic pressure and the predicted value of the diastolic pressure;
determining a blood pressure interval of the systolic pressure based on the first upper threshold value and the first lower threshold value;
determining a second upper boundary value and a second lower boundary value corresponding to the diastolic pressure based on the predicted value of the systolic pressure and the predicted value of the diastolic pressure;
and determining the blood pressure interval of the diastolic pressure based on the second upper threshold value and the second lower threshold value.
In a second aspect, an embodiment of the present application provides a training method for a blood pressure monitoring model, where the method includes:
preprocessing an array PPG signal in a sample data set to obtain a standardized array PPG signal and an adjacent matrix of a graph structure corresponding to the array PPG signal, wherein the sample data set comprises the array PPG signal and a labeled value of systolic pressure and a labeled value of diastolic pressure corresponding to the array PPG signal;
taking the standardized array PPG signal and the adjacency matrix as the input of a first sub-model in a blood pressure monitoring model to obtain a blood pressure predicted value, wherein the blood pressure predicted value comprises a predicted value of systolic pressure and a predicted value of diastolic pressure;
and taking the predicted value of the systolic pressure and the predicted value of the diastolic pressure as the input of a second sub-model to obtain the range of the blood pressure interval of the systolic pressure and the range of the blood pressure interval of the diastolic pressure.
In the above scheme, the obtaining a predicted blood pressure value by using the normalized array PPG signal and the adjacency matrix as input of a first sub-model in a blood pressure monitoring model includes:
obtaining a spatial feature corresponding to the array PPG signal based on a graph convolution neural network;
obtaining a time characteristic corresponding to the array PPG signal based on a long-short term memory neural network;
performing feature fusion on the spatial features and the time features through a full connection layer, and predicting the fused space-time features based on a linear regression algorithm to obtain a predicted value of systolic pressure and a predicted value of diastolic pressure;
and the first sub-model adjusts the parameters of the first sub-model based on the difference between the marked value of the systolic pressure and the predicted value of the systolic pressure and the difference between the marked value of the diastolic pressure and the predicted value of the diastolic pressure, and performs model training.
In the above aspect, the obtaining the blood pressure interval range of the systolic pressure and the blood pressure interval range of the diastolic pressure by using the predicted value of the systolic pressure and the predicted value of the diastolic pressure as input of the second submodel includes:
the second sub-model performs Gaussian regression processing based on the predicted value of the systolic pressure, determines a first upper threshold and a first lower threshold corresponding to the systolic pressure, and determines a blood pressure interval of the systolic pressure based on the first upper threshold and the first lower threshold;
and the second sub-model performs Gaussian regression processing on the basis of the predicted value of the diastolic pressure, determines a second upper bound value and a second lower bound value corresponding to the diastolic pressure, and determines a blood pressure interval of the diastolic pressure on the basis of the second upper bound value and the second lower bound value.
In a third aspect, an embodiment of the present application provides a blood pressure monitoring device, including:
the preprocessing module is used for preprocessing the array PPG signal to obtain a standardized array PPG signal and an adjacent matrix of a graph structure corresponding to the array PPG signal;
a blood pressure value determination module, configured to obtain a spatial feature and a temporal feature corresponding to the array PPG signal based on the normalized array PPG signal and the adjacency matrix; performing feature fusion on the spatial feature and the temporal feature, and determining a predicted value of systolic pressure and a predicted value of diastolic pressure based on the space-time feature obtained after fusion;
and the blood pressure interval determining module is used for determining a systolic blood pressure interval and a diastolic blood pressure interval corresponding to the array PPG signal based on the predicted value of the systolic pressure and the predicted value of the diastolic pressure.
In a fourth aspect, an embodiment of the present application provides a blood pressure monitoring model training device, where the device includes:
the system comprises a preprocessing module, a data acquisition module and a data processing module, wherein the preprocessing module is used for preprocessing array PPG signals in a sample data set to obtain standardized array PPG signals and an adjacent matrix of a graph structure corresponding to the array PPG signals, and the sample data set comprises the array PPG signals and labeled values of systolic pressure and diastolic pressure corresponding to the array PPG signals;
the blood pressure value determining module is used for taking the standardized array PPG signal and the adjacency matrix as the input of a first sub-model in a blood pressure monitoring model to obtain a blood pressure predicted value, and the blood pressure predicted value comprises a predicted value of systolic pressure and a predicted value of diastolic pressure;
a blood pressure interval determining module for inputting the predicted value of the systolic pressure and the predicted value of the diastolic pressure as the second submodel to obtain the range of the blood pressure interval of the systolic pressure and the range of the blood pressure interval of the diastolic pressure
In a fifth aspect, an embodiment of the present application provides an electronic device, including: at least one processor; and a memory communicatively coupled to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the blood pressure monitoring method provided by the embodiment of the present application or the blood pressure monitoring model training method provided by the embodiment of the present application.
In a sixth aspect, the present application provides a computer-readable storage medium, where the storage medium includes a set of computer-executable instructions, and when the instructions are executed, the instructions are used to execute the blood pressure monitoring method or the blood pressure monitoring model training method provided by the present application.
According to the blood pressure monitoring method provided by the embodiment of the application, the array PPG signal is preprocessed to obtain a corresponding standardized array PPG signal and an adjacent matrix of a graph structure corresponding to the array PPG signal; determining a spatial feature and a temporal feature corresponding to the array PPG signal based on the normalized array PPG signal and the adjacency matrix; performing feature fusion on the spatial features and the temporal features to obtain space-time features; determining a blood pressure prediction value based on the space-time feature; and determining a systolic blood pressure interval and a diastolic blood pressure interval corresponding to the array PPG signal based on the blood pressure predicted value. According to the blood pressure monitoring method, the spatial features and the temporal features are extracted and fused, so that the accuracy of blood pressure value prediction and the accuracy of blood pressure interval prediction are improved.
Drawings
The drawings are included to provide a better understanding of the present solution and are not intended to limit the present application. Wherein:
FIG. 1 is a schematic view of an alternative processing flow of a blood pressure monitoring method provided by an embodiment of the present application;
fig. 2 is a schematic diagram of a process for constructing map data based on array PPG signals provided by an embodiment of the present application;
FIG. 3 is a schematic structural diagram of a spatial feature extraction model provided in an embodiment of the present application;
FIG. 4 is a schematic structural diagram of a temporal feature extraction model provided in an embodiment of the present application;
FIG. 5 is a schematic flowchart of obtaining a predicted blood pressure value according to an embodiment of the present application;
FIG. 6 is a schematic view of an alternative processing flow of a blood pressure monitoring method provided by an embodiment of the present application;
FIG. 7 is a schematic view of an alternative configuration of a blood pressure monitoring device according to an embodiment of the present application;
FIG. 8 is a schematic view of an alternative processing flow of a blood pressure monitoring model training method provided by an embodiment of the present application;
FIG. 9 is a schematic view of an alternative processing flow of a blood pressure monitoring model training method provided by an embodiment of the present application;
FIG. 10 is a diagram illustrating a predicted result of a blood pressure interval according to an embodiment of the present application;
FIG. 11 is a schematic diagram of an alternative structure of a blood pressure monitoring model training device according to an embodiment of the present application;
fig. 12 is a schematic block diagram of an alternative electronic device provided in an embodiment of the present application.
Detailed Description
In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be described in further detail with reference to the accompanying drawings, the described embodiments should not be considered as limiting the present application, and all other embodiments obtained by a person of ordinary skill in the art without making creative efforts fall within the protection scope of the present application.
In the following description, reference is made to "some embodiments" which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict.
In the following description, references to the terms "first," "second," and the like, are intended only to distinguish similar objects and not to imply a particular order to the objects, it being understood that "first," "second," and the like may be interchanged under appropriate circumstances or a sequential order, such that the embodiments of the application described herein may be practiced in other than those illustrated or described herein.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of the present application only and is not intended to be limiting of the application.
A blood pressure monitoring method provided in the embodiment of the present application will be described below, referring to fig. 1, fig. 1 is a schematic processing flow diagram of an alternative blood pressure monitoring method provided in the embodiment of the present application, and the following description will be given with reference to steps S101 to S105 shown in fig. 1.
Step S101, preprocessing the array PPG signal to obtain a standardized array PPG signal and an adjacent matrix of a graph structure corresponding to the array PPG signal.
In some embodiments, the array PPG signal comprises two dimensions, one dimension being time and the other dimension being the number of arrays. Such as a 3 x 3 array PPG signal of duration 2 s. Fig. 2 shows the PPG signals with an array 3 × 3, including v1, v2, v3, v4, v5, v6, v7, v8, and v9, for a total of 9 PPG signals.
In some embodiments, the sequence of raw array PPG signals is assumed to be, in order: s is 1 ,…,s n Note that, if the average of the signal values of all array PPG signals is μ, and the standard deviation of the signal values of all array PPG signals is σ, the normalized array PPG signals are, in order:
wherein the normalized array PPG signal follows a normal distribution.
In some embodiments, taking fig. 2 as an example, the array PPG signal is 9 lines of 3 × 3 PPG signals, the original array PPG signals acquired in real time may be selected according to a fixed first length by using a sliding window method, and for each selected array PPG signal of the first length, an adjacency matrix of a graph corresponding to the array PPG signal is constructed based on an average value of a sum of maximum mutual information coefficients between each line of the PPG signal and an adjacent PPG signal. The following formula (1) shows a calculation formula of a maximum mutual information coefficient between two adjacent PPG signals in the array PPG signals.
In the formula (1), V i And V j Representing two adjacent PPG signals, M, in the array PPG signal IC (V i, V j ) Represents V i And V j Maximum mutual information coefficient between, D represents V i And V j Two PPG signals, x represents V i Any one of the PPG signal values; y represents V j R (n) represents the maximum value of the product between x and y, and I (D, x, y) represents V i And V j Maximum value of the correlation mutual information.
After the maximum mutual information coefficient of each path of PPG signal in the array PPG signal is obtained based on the formula (1), the average value of the sum of the maximum mutual information coefficients between each path of PPG signal and the adjacent PPG signal is obtained as shown in the formula (2), and the finally constructed adjacency matrix is obtained.
In the formula (2), in
For example, since the PPG signals adjacent to v1 are v2 and v4 in fig. 2, the average value of the sum of the maximum mutual information number between v1 and v2 and the maximum mutual information number between v1 and v4 is determined as the vertex value corresponding to v1 in the adjacency matrix.
And S102, determining a spatial feature and a temporal feature corresponding to the array PPG signal based on the normalized array PPG signal and the adjacency matrix.
In some embodiments, the spatial features may be obtained by a Graph Convolutional Neural network (GCN) model for normalized array PPG signals and adjacency matrices. The temporal features are obtained by a Long Short-term Memory neural network (LSTM) model.
The normalized array PPG signal and the adjacency matrix are input into the spatial feature extraction model shown in fig. 3, so that the spatial feature corresponding to the array PPG signal can be obtained. In fig. 3, the spatial feature extraction model includes two graph convolution layers and one full-link layer.
By inputting the normalized array PPG signal and the adjacency matrix into the temporal feature extraction model shown in fig. 4, the temporal features corresponding to the array PPG signal can be obtained. In fig. 4, the temporal feature extraction model includes two LSTM layers and one fully-connected layer.
And S103, performing feature fusion on the spatial features and the temporal features to obtain space-time features.
In some embodiments, the spatial features and the temporal features may be feature-fused through a full connection layer to obtain a space-time feature of the array PPG signal.
The space-time characteristics are obtained by further linear mapping of the space characteristics and the time characteristics, so that the problem that different characteristics are directly insufficient in extraction of distinguishing characteristics due to different amplitudes is solved, and the characteristics are more accurately extracted.
And step S104, determining a predicted blood pressure value based on the space-time characteristics.
In some embodiments, the fused spatiotemporal features may be predicted based on a linear regression algorithm, and a predicted blood pressure value may be determined, where the predicted blood pressure value includes a predicted systolic pressure value and a predicted diastolic pressure value. As shown in fig. 5, fig. 5 shows a flow diagram for obtaining a predicted value of blood pressure. The structure of fig. 5 includes the spatial feature extraction model of fig. 3 and the temporal feature extraction model of fig. 4, and after the spatial feature is extracted based on the spatial feature extraction model and the temporal feature is extracted based on the temporal feature extraction model, the space-time feature is obtained based on the spatial feature and the temporal feature through the full link layer, and the linear regression calculation is performed on the space-time feature based on the regression layer, so as to determine the predicted value of systolic pressure and the predicted value of diastolic pressure.
And step S105, determining a blood pressure interval range of the systolic pressure and a blood pressure interval range of the diastolic pressure based on the blood pressure predicted value.
In some embodiments, a first upper threshold and a first lower threshold corresponding to the systolic pressure may be determined based on the predicted value of the systolic pressure and the predicted value of the diastolic pressure, and a blood pressure interval of the systolic pressure may be determined based on the first upper threshold and the first lower threshold;
a second upper boundary value and a second lower boundary value corresponding to the diastolic pressure may be determined based on the predicted value of the diastolic pressure and the predicted value of the diastolic pressure, and a blood pressure interval of the diastolic pressure may be determined based on the second upper boundary value and the second lower boundary value.
Another processing flow of the blood pressure monitoring method provided in the embodiment of the present application is described below. Referring to fig. 6, fig. 6 is a schematic view of an alternative processing flow of a blood pressure monitoring method provided in the embodiment of the present application.
In some embodiments, the blood pressure monitoring method may be implemented based on a graph convolution neural network space-time feature fusion deep learning model GT-net model provided by the present application.
Step 1, inputting an original array PPG signal into a preprocessing unit to obtain a standardized array PPG signal and a corresponding adjacent matrix.
And 2, extracting corresponding spatial features of the standardized array PPG signals and the adjacent matrix through a spatial feature extraction unit, and simultaneously extracting corresponding temporal features through a temporal feature extraction unit.
And 3, passing the spatial characteristics and the time characteristics through a characteristic fusion unit to obtain space-time characteristics.
And 4, outputting the blood pressure value by the space-time characteristic through a point output unit, wherein the blood pressure value comprises a predicted value of systolic pressure and a predicted value of diastolic pressure.
And 5, passing the blood pressure value through a Gaussian process regression unit to obtain a first upper threshold value and a first lower threshold value of the systolic pressure, and a second upper threshold value and a second lower threshold value of the diastolic pressure.
And 6, outputting a blood pressure value interval by passing the first upper limit value and the first lower limit value of the systolic pressure and the second upper limit value and the second lower limit value of the diastolic pressure through an interval output unit, wherein the blood pressure value interval comprises a blood pressure range of the systolic pressure and a blood pressure range of the diastolic pressure.
The following describes a blood pressure monitoring device provided in an embodiment of the present application. Fig. 7 is a schematic structural diagram of an optional device of a blood pressure monitoring device according to an embodiment of the present application, where the blood pressure monitoring device 700 includes a preprocessing module 701, a blood pressure value determining module 702, and a blood pressure interval determining module 703. Wherein,
the preprocessing module 701 is configured to preprocess the array photoelectric pulse wave PPG signal to obtain a normalized array PPG signal corresponding to the array PPG signal and an adjacency matrix of a graph structure corresponding to the array PPG signal;
a blood pressure value determining module 702, configured to obtain a spatial feature and a temporal feature corresponding to the array PPG signal based on the normalized array PPG signal and the adjacency matrix; performing feature fusion on the spatial features and the temporal features to obtain space-time features, and determining a blood pressure predicted value based on the space-time features;
a blood pressure interval determining module 703 is configured to determine, based on the predicted blood pressure value, a blood pressure interval of systolic pressure and a blood pressure interval of diastolic pressure corresponding to the array PPG signal.
In some embodiments, the pre-processing module 701 is configured to: and carrying out standardization processing on the array PPG signal to obtain the standardized array PPG signal.
In some embodiments, the preprocessing module 701 is further configured to: selecting an array PPG signal of a first length based on a sliding window; and aiming at the array PPG signal with the first length selected each time, constructing the adjacency matrix based on the average value of the sum of the maximum mutual information coefficients between each path of PPG signal and the adjacent PPG signals.
In some embodiments, the blood pressure value determination module 702 is configured to: obtaining a spatial feature corresponding to the array PPG signal based on a graph convolution neural network; and obtaining the time characteristics corresponding to the PP G signals of the array based on the long-term and short-term memory neural network.
In some embodiments, the blood pressure value determination module 702 is further configured to: and performing feature fusion on the spatial features and the temporal features through a full connection layer to obtain the space-time features.
In some embodiments, the blood pressure value determination module 702 is further configured to: and predicting the space-time characteristics based on a linear regression algorithm to obtain a predicted value of systolic pressure and a predicted value of diastolic pressure.
In some embodiments, the blood pressure interval determination module 703 is configured to determine a first upper threshold and a first lower threshold corresponding to the systolic pressure based on the predicted value of the systolic pressure and the predicted value of the diastolic pressure; determining a blood pressure interval of the systolic pressure based on the first upper threshold value and the first lower threshold value; determining a second upper boundary value and a second lower boundary value corresponding to the diastolic pressure based on the predicted value of the systolic pressure and the predicted value of the diastolic pressure; determining a blood pressure interval of the diastolic pressure based on the second upper threshold value and the second lower threshold value.
It should be noted that the blood pressure monitoring device in the embodiment of the present application is similar to the blood pressure monitoring method in the above description, and has similar beneficial effects to the method in the embodiment, and therefore, the detailed description is omitted. The inexhaustible technical details in the blood pressure monitoring device provided by the embodiment of the application can be understood according to the description of any one of the drawings in fig. 1 to 6.
The following describes a training method of a blood pressure monitoring model provided by an embodiment of the present application, where the blood pressure monitoring model provided by the present application is a GT-GPR model, and is described with reference to fig. 8 to 9.
In some embodiments, array PPG signals may be acquired for a plurality of subjects, and a true systolic and diastolic value corresponding to each subject's array PPG signal is obtained;
in some embodiments, all array PPG signals are normalized, resulting in a normalized array PPG signal.
Wherein the standardization processing process comprises the following steps: suppose that the signal sequences of all array PPG signals are respectively s 1 ,…,s n And recording the mean value and standard deviation value of the PPG signal as mu and sigma respectively, then the normalized array PPG signal sequence is as follows:
wherein the normalized array PPG signal follows a normal distribution.
In some embodiments, all array PPG signals are selected based on a sliding window method, and for each selected array PPG signal of fixed length, a corresponding adjacency matrix is obtained.
The method includes the steps that for array PPG signals with fixed length selected each time, an adjacency matrix of a graph corresponding to the array PPG signals is constructed based on the average value of the sum of maximum mutual information coefficients between each path of PPG signals and adjacent PPG signals.
In some embodiments, the array PPG signals obtained after the pre-processing are used as sample data, and the systolic and diastolic values corresponding to each array PPG signal are used as sample labels. Sample data is divided into a test set and a training set. The test set is used for training the blood pressure monitoring model and is used for testing the trained blood pressure monitoring model.
And step 802, taking the standardized array PPG signal and the adjacent matrix as the input of a first sub-model in a blood pressure monitoring model to obtain a blood pressure predicted value, wherein the blood pressure predicted value comprises a predicted value of systolic pressure and a predicted value of diastolic pressure.
In some embodiments, the first sub-model is the GT-net model in fig. 9, a GCN-LSTM neural network model consisting of a spatial feature extraction network, a temporal feature extraction network, a feature fusion network and a blood pressure prediction value output.
The GT-net model training process comprises the following steps:
obtaining a spatial feature corresponding to the array PPG signal based on a spatial feature extraction network formed by two GCNs and one full connection layer; meanwhile, the time characteristics corresponding to the array PPG signals are obtained based on a time characteristic extraction network formed by two LSTMs and one full connection layer. And performing feature fusion on the spatial features and the time features through a feature fusion network formed by full connection layers, and predicting a blood pressure value of the fused space-time features based on a linear regression algorithm to obtain a predicted value of the blood pressure, wherein the predicted value of the blood pressure comprises a predicted value of systolic pressure and a predicted value of diastolic pressure. And updating the parameters of the GT-net model according to the difference between the marked value of the systolic pressure and the predicted value of the systolic pressure and the difference between the marked value of the diastolic pressure and the predicted value of the diastolic pressure until the training termination condition is met, and finishing the training of the model.
And 803, taking the predicted value of the systolic pressure and the predicted value of the diastolic pressure as input of a second sub-model, and obtaining the range of the blood pressure interval of the systolic pressure and the range of the blood pressure interval of the diastolic pressure.
In some embodiments, the second sub-model is the GPR model of fig. 9, which may be composed of a GPR model predicting the range of the blood pressure interval of the systolic pressure and a GPR model predicting the range of the blood pressure interval of the diastolic pressure.
The GPR model for predicting the range of the blood pressure interval of the systolic pressure performs Gaussian regression processing based on the predicted value of the systolic pressure, determines a first upper bound value and a first lower bound value corresponding to the systolic pressure, and determines the blood pressure interval of the systolic pressure as the blood pressure range between the first upper bound value and the first lower bound value, wherein the blood pressure range comprises the first upper bound value and the first lower bound value.
And on the basis of the second upper boundary value and the second lower boundary value, determining the blood pressure interval of the diastolic pressure to be a blood pressure range between the second upper boundary value and the second lower boundary value which comprise the second upper boundary value and the second lower boundary value.
In summary, the first sub-model and the second sub-model constitute the GT-GPR model proposed in the present application, and the trained GT-GPR model may be subjected to model verification by using data of the test set.
As shown in the following table, the GT-GPR model presented in the present application was used to validate SBP (Systolic Blood Pressure ) using the published data set MIMIC-III developed by the institute of technology, massachusetts, inc. in computational physiology laboratories. The point prediction adopts performance indexes MAE (Mean Absolute Error), ME (Mean Error) and STD; section prediction using CP α (Coverage, MWP) α (Mean Width Percentage, average Percentage of interval Width to observed value) and custom MC α Wherein, MC is used in the present application α Is defined as
MC α Is defined because the ideal prediction interval should have a high CP α And low MWP α Where α is the confidence interval, 95% confidence intervals are used herein. Compared with CNN-GPR and CLSTM-GPR models with similar structures, the GT-GPR model has obvious performance improvement, and the point prediction value is in the AAMI standard range. Wherein the AAMI standard has a ME + -STD of 5 + -8.
In conclusion, in order to fully utilize information on a space-time domain of an array PPG signal and fuse the features extracted by a GCN model and an LSTM model, the application provides a space-time feature fusion deep learning model GT-net based on a graph convolution neural network; the model provided by the application makes full use of the advantages of each network, adopts GCN to extract the spatial domain characteristics of the array PPG signal, adopts LSTM to extract the time domain characteristics of the array PPG signal, and realizes the fusion of the spatial and temporal characteristics, thereby realizing the full extraction of the characteristics of the array PPG signal for measuring the blood pressure value; on the basis of obtaining higher accuracy rate of prediction of systolic pressure and prediction of diastolic pressure by utilizing GT-net provided by the application, the application combines the advantages of a GPR model in interval prediction, provides the GT-GPR model for uncertain measurement of blood pressure values in the application, and improves the accuracy rate of monitoring the blood pressure interval; the model that this application provided has avoided the not enough problem of the differentiation characteristic expression that exists of artificial feature, has promoted the generalization performance of model, and the model that this application provided is more fit for the uncertain monitoring based on the sleeveless area blood pressure value of array PPG signal. Fig. 10 shows the results of prediction of the systolic and diastolic blood pressure intervals in the test set.
The following describes a blood pressure monitoring model training device provided in an embodiment of the present application. Fig. 11 is a schematic structural diagram of an optional device of a blood pressure monitoring model training device provided in an embodiment of the present application, where the blood pressure monitoring model training device 1100 includes a preprocessing module 1101, a blood pressure value determining module 1102, and a blood pressure interval determining module 1103. Wherein,
the preprocessing module 1101 is configured to preprocess an array photoelectric pulse wave (PPG) signal in a sample data set to obtain a normalized array PPG signal and an adjacency matrix of a graph structure corresponding to the array PPG signal, where the sample data set includes the array PPG signal and a labeled value of systolic pressure and a labeled value of diastolic pressure corresponding to the array PPG signal;
a blood pressure value determining module 1102, configured to use the normalized array PPG signal and the adjacency matrix as inputs of a first sub-model in a blood pressure monitoring model to obtain a blood pressure predicted value, where the blood pressure predicted value includes a predicted value of systolic pressure and a predicted value of diastolic pressure;
a blood pressure interval determining module 1103, configured to use the predicted value of the systolic pressure and the predicted value of the diastolic pressure as inputs of the second sub-model, so as to obtain a blood pressure interval range of the systolic pressure and a blood pressure interval range of the diastolic pressure.
In some embodiments, the pre-processing module 1101 is configured to: preprocessing an array photoelectric pulse wave (PPG) signal in a sample data set to obtain a standardized array PPG signal and an adjacency matrix of a graph structure corresponding to the array PPG signal, wherein the sample data set comprises the array PPG signal, and a systolic pressure marked value and a diastolic pressure marked value corresponding to the array PPG signal; taking the standardized array PPG signal and the adjacency matrix as the input of a first sub-model in a blood pressure monitoring model to obtain a blood pressure predicted value, wherein the blood pressure predicted value comprises a predicted value of systolic pressure and a predicted value of diastolic pressure; and taking the predicted value of the systolic pressure and the predicted value of the diastolic pressure as the input of a second sub model to obtain the blood pressure interval range of the systolic pressure and the blood pressure interval range of the diastolic pressure.
In some embodiments, the blood pressure value determination module 1102 is configured to: obtaining a spatial feature corresponding to the array PPG signal based on a graph convolution neural network; obtaining a time characteristic corresponding to the array PPG signal based on a long-short term memory neural network; performing feature fusion on the spatial features and the time features through a full connection layer, and predicting the fused space-time features based on a linear regression algorithm to obtain a predicted value of systolic pressure and a predicted value of diastolic pressure; and the first sub-model adjusts the parameters of the first sub-model based on the difference between the marked value of the systolic pressure and the predicted value of the systolic pressure and the difference between the marked value of the diastolic pressure and the predicted value of the diastolic pressure, and performs model training.
In some embodiments, the blood pressure interval determination module 1103 is configured to: the second sub-model performs Gaussian regression processing based on the predicted value of the systolic pressure, determines a first upper threshold and a first lower threshold corresponding to the systolic pressure, and determines a blood pressure interval of the systolic pressure based on the first upper threshold and the first lower threshold; and the second sub-model performs Gaussian regression processing on the basis of the predicted value of the diastolic pressure, determines a second upper bound value and a second lower bound value corresponding to the diastolic pressure, and determines a blood pressure interval of the diastolic pressure on the basis of the second upper bound value and the second lower bound value.
FIG. 12 shows a schematic block diagram of an example electronic device 1200 that can be used to implement embodiments of the present disclosure. The electronic device 1200 is used for implementing the blood pressure monitoring method or the blood pressure monitoring model training method of the embodiment of the present disclosure. In some optional embodiments, the electronic device 1200 may implement the blood pressure monitoring method or the blood pressure monitoring model training method provided in the embodiments of the present application by running a computer program, for example, the computer program may be a software module in an operating system; may be a local (Native) APP (application), i.e. a program that needs to be installed in the operating system to run; or may be an applet, i.e. a program that can be run only by downloading it to the browser environment; but also an applet that can be embedded into any APP. In general, the computer programs described above may be any form of application, module, or plug-in.
In practical applications, the electronic device 1200 may be an independent physical server, a server cluster or a distributed system formed by a plurality of physical servers, or a Cloud server providing basic Cloud computing services such as a Cloud service, a Cloud database, cloud computing, a Cloud function, cloud storage, a network service, cloud communication, a middleware service, a domain name service, a security service, a CDN, and a big data and artificial intelligence platform, where Cloud Technology (Cloud Technology) refers to a hosting Technology for unifying series resources such as hardware, software, and a network in a wide area network or a local area network to implement computing, storing, processing, and sharing of data. The electronic device 1200 may be, but is not limited to, a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smart television, a smart watch, and the like.
Electronic devices are intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, vehicle terminals, and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be examples only, and are not meant to limit implementations of the present application that are described and/or claimed herein.
As shown in fig. 12, the electronic apparatus 1200 includes a computing unit 1201 that can perform various appropriate actions and processes in accordance with a computer program stored in a Read Only Memory (ROM) 1202 or a computer program loaded from a storage unit 1208 into a Random Access Memory (RAM) 1203. In the RAM 1203, various programs and data necessary for the operation of the electronic apparatus 1200 may also be stored. The computing unit 1201, the ROM 1202, and the RAM 1203 are connected to each other by a bus 1204. An input/output (I/O) interface 1205 is also connected to bus 1204.
Various components in the electronic device 1200 are connected to the I/O interface 1205, including: an input unit 1206 such as a keyboard, a mouse, or the like; an output unit 1207 such as various types of displays, speakers, and the like; a storage unit 1208, such as a magnetic disk, optical disk, or the like; and a communication unit 1209 such as a network card, modem, wireless communication transceiver, etc. The communication unit 1209 allows the electronic device 1200 to exchange information/data with other devices via a computer network such as the internet and/or various telecommunication networks.
The computing unit 1201 may be a variety of general purpose and/or special purpose processing components having processing and computing capabilities. Some examples of the computing unit 1201 include, but are not limited to, a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), various specialized Artificial Intelligence (AI) computing chips, various computing units running machine learning model algorithms, a Digital Signal Processor (DSP), and any suitable processor, controller, microcontroller, and so forth. The calculation unit 1201 performs the various methods and processes described above, such as a blood pressure monitoring method or a training method of a blood pressure monitoring model. For example, in some alternative embodiments, the blood pressure monitoring method or the training method of the blood pressure monitoring model may be implemented as a computer software program tangibly embodied in a machine-readable medium, such as storage unit 1208. In some alternative embodiments, part or all of the computer program may be loaded and/or installed onto the electronic device 1200 via the ROM 1202 and/or the communication unit 1209. When the computer program is loaded into the RAM 1203 and executed by the computing unit 1201, one or more steps of the above described blood pressure monitoring method or training method of the blood pressure monitoring model may be performed. Alternatively, in other embodiments, the computing unit 1201 may be configured in any other suitable way (e.g. by means of firmware) as a blood pressure monitoring method or a training method of a blood pressure monitoring model.
The embodiment of the present application provides a computer-readable storage medium storing executable instructions, wherein the executable instructions, when executed by a processor, cause the processor to execute the blood pressure monitoring method or the training method of the blood pressure monitoring model provided by the embodiment of the present application.
In some embodiments, the computer-readable storage medium may be memory such as FRAM, ROM, PROM, EPROM, EEPROM, flash, magnetic surface memory, optical disk, or CD-ROM; or may be various devices including one or any combination of the above memories.
In some embodiments, the executable instructions may be in the form of a program, software module, script, or code written in any form of programming language, including compiled or interpreted languages, or declarative or procedural languages, and it may be deployed in any form, including as a stand-alone program or as a module, component, subroutine, or other unit suitable for use in a computing environment.
By way of example, executable instructions may be deployed to be executed on one computing device or on multiple computing devices at one site or distributed across multiple sites and interconnected by a communication network.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
It should be understood that, in the various embodiments of the present application, the size of the serial number of each implementation process does not mean the execution sequence, and the execution sequence of each process should be determined by its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.
The above description is only an example of the present application, and is not intended to limit the scope of the present application. Any modification, equivalent replacement, and improvement made within the spirit and scope of the present application are included in the protection scope of the present application.
Claims (14)
1. A method of monitoring blood pressure, the method comprising:
preprocessing an array photoelectric pulse wave (PPG) signal to obtain a standardized array PPG signal and an adjacent matrix of a graph structure corresponding to the PPG signal;
determining a spatial feature and a temporal feature corresponding to the array PPG signal based on the normalized array PPG signal and the adjacency matrix;
performing feature fusion on the spatial features and the temporal features to obtain space-time features;
determining a blood pressure prediction value based on the space-time feature;
and determining a systolic blood pressure interval and a diastolic blood pressure interval corresponding to the array PPG signals based on the blood pressure predicted value.
2. The method of claim 1, wherein the pre-processing of the photoplethysmographic array (PPG) signals to obtain a normalized array PPG signal and a contiguous matrix of graph structures to which the PPG signal corresponds comprises:
and carrying out standardization processing on the array PPG signal to obtain the standardized array PPG signal.
3. The method according to claim 1, wherein the preprocessing the photoplethysmography array (PPG) signal to obtain a normalized array PPG signal and a contiguous matrix of graph structures corresponding to the array PPG signal comprises:
selecting an array PPG signal of a first length based on a sliding window;
and aiming at the array PPG signal with the first length selected each time, constructing the adjacency matrix based on the average value of the sum of the maximum mutual information coefficients between each path of PPG signal and the adjacent PPG signals.
4. The method of claim 1, wherein determining the corresponding spatial and temporal features of the array PPG signal based on the normalized array PP G signal and the adjacency matrix comprises:
obtaining a spatial feature corresponding to the array PPG signal based on a graph convolution neural network;
and obtaining the corresponding time characteristics of the array PPG signals based on a long-short term memory neural network.
5. The method according to claim 1, wherein the feature fusing the spatial feature and the temporal feature to obtain a space-time feature comprises:
and performing feature fusion on the spatial features and the temporal features through a full connection layer to obtain the space-time features.
6. The method of claim 1, wherein determining a predicted blood pressure value based on the space-time feature comprises:
and predicting the space-time characteristics based on a linear regression algorithm to obtain a predicted value of systolic pressure and a predicted value of diastolic pressure.
7. The method according to claim 1, wherein determining a systolic blood pressure interval and a diastolic blood pressure interval corresponding to the array PPG signal based on the predicted blood pressure value comprises:
determining a first upper boundary value and a first lower boundary value corresponding to the systolic pressure based on the predicted value of the systolic pressure and the predicted value of the diastolic pressure;
determining a blood pressure interval of the systolic pressure based on the first upper threshold value and the first lower threshold value;
determining a second upper boundary value and a second lower boundary value corresponding to the diastolic pressure based on the predicted value of the systolic pressure and the predicted value of the diastolic pressure;
determining a blood pressure interval of the diastolic pressure based on the second upper threshold value and the second lower threshold value.
8. A training method of a blood pressure monitoring model is characterized by comprising the following steps:
preprocessing an array PPG signal in a sample data set to obtain a standardized array PPG signal and an adjacent matrix of a graph structure corresponding to the array PPG signal, wherein the sample data set comprises the array PPG signal and a labeled value of systolic pressure and a labeled value of diastolic pressure corresponding to the array PPG signal;
taking the standardized array PPG signal and the adjacent matrix as the input of a first sub-model in a blood pressure monitoring model to obtain a blood pressure predicted value, wherein the blood pressure predicted value comprises a predicted value of systolic pressure and a predicted value of diastolic pressure;
and taking the predicted value of the systolic pressure and the predicted value of the diastolic pressure as the input of a second sub model to obtain the blood pressure interval range of the systolic pressure and the blood pressure interval range of the diastolic pressure.
9. The method of claim 8, wherein said inputting the normalized array PPG signal and the adjacency matrix as a first sub-model in a blood pressure monitoring model to obtain a blood pressure prediction value comprises:
obtaining a spatial feature corresponding to the array PPG signal based on a graph convolution neural network;
obtaining a time characteristic corresponding to the array PPG signal based on a long-short term memory neural network;
performing feature fusion on the spatial features and the time features through a full connection layer, and predicting the fused space-time features based on a linear regression algorithm to obtain a predicted value of systolic pressure and a predicted value of diastolic pressure;
and the first sub-model adjusts the parameters of the first sub-model based on the difference between the marked value of the systolic pressure and the predicted value of the systolic pressure and the difference between the marked value of the diastolic pressure and the predicted value of the diastolic pressure, and performs model training.
10. The model training method according to claim 8, wherein the obtaining of the blood pressure interval range of the systolic pressure and the blood pressure interval range of the diastolic pressure using the predicted values of the systolic pressure and the predicted values of the diastolic pressure as inputs to the second submodel comprises:
the second sub-model performs Gaussian regression processing on the basis of the predicted value of the systolic pressure, determines a first upper bound value and a first lower bound value corresponding to the systolic pressure, and determines a blood pressure interval of the systolic pressure on the basis of the first upper bound value and the first lower bound value;
and the second sub-model performs Gaussian regression processing on the basis of the predicted value of the diastolic pressure, determines a second upper bound value and a second lower bound value corresponding to the diastolic pressure, and determines a blood pressure interval of the diastolic pressure on the basis of the second upper bound value and the second lower bound value.
11. A blood pressure monitoring device, the device comprising:
the preprocessing module is used for preprocessing the array photoelectric pulse wave PPG signal to obtain a standardized array PPG signal and an adjacent matrix of a graph structure corresponding to the array PPG signal;
a blood pressure value determination module, configured to obtain a spatial feature and a temporal feature corresponding to the array PPG signal based on the normalized array PPG signal and the adjacency matrix; performing feature fusion on the spatial features and the temporal features to obtain space-time features, and determining a blood pressure predicted value based on the space-time features;
and the blood pressure interval determining module is used for determining a systolic blood pressure interval and a diastolic blood pressure interval corresponding to the array PPG signal based on the blood pressure predicted value.
12. A blood pressure monitoring model training device, the device comprising:
the system comprises a preprocessing module, a data acquisition module and a data processing module, wherein the preprocessing module is used for preprocessing array PPG signals in a sample data set to obtain standardized array PPG signals and an adjacent matrix of a graph structure corresponding to the array PPG signals, and the sample data set comprises the array PPG signals and labeled values of systolic pressure and diastolic pressure corresponding to the array PPG signals;
the blood pressure value determining module is used for taking the standardized array PPG signal and the adjacent matrix as the input of a first sub-model in a blood pressure monitoring model to obtain a blood pressure predicted value, and the blood pressure predicted value comprises a predicted value of systolic pressure and a predicted value of diastolic pressure;
and the blood pressure interval determining module is used for taking the predicted value of the systolic pressure and the predicted value of the diastolic pressure as the input of a second submodel to obtain the range of the blood pressure interval of the systolic pressure and the range of the blood pressure interval of the diastolic pressure.
13. An electronic device, characterized in that the electronic device comprises:
at least one processor; and a memory communicatively coupled to the at least one processor;
wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-7 or 8-10.
14. A computer-readable storage medium comprising a set of computer-executable instructions for performing the method of monitoring blood pressure of any one of claims 1-7 or the method of training a blood pressure monitoring model of any one of claims 8-10 when the instructions are executed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211554662.4A CN115956889A (en) | 2022-12-05 | 2022-12-05 | Blood pressure monitoring method and device and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211554662.4A CN115956889A (en) | 2022-12-05 | 2022-12-05 | Blood pressure monitoring method and device and electronic equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115956889A true CN115956889A (en) | 2023-04-14 |
Family
ID=87351906
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211554662.4A Pending CN115956889A (en) | 2022-12-05 | 2022-12-05 | Blood pressure monitoring method and device and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115956889A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116803340A (en) * | 2023-07-19 | 2023-09-26 | 北京理工大学 | Noninvasive blood pressure detection method based on multi-source data fusion and graph neural network |
-
2022
- 2022-12-05 CN CN202211554662.4A patent/CN115956889A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116803340A (en) * | 2023-07-19 | 2023-09-26 | 北京理工大学 | Noninvasive blood pressure detection method based on multi-source data fusion and graph neural network |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10902245B2 (en) | Method and apparatus for facial recognition | |
| CN111950638B (en) | Image classification method and device based on model distillation and electronic equipment | |
| CN112396613B (en) | Image segmentation method, device, computer equipment and storage medium | |
| CN111368937A (en) | Image classification method and device, and training method, device, equipment and medium thereof | |
| CN113095370A (en) | Image recognition method and device, electronic equipment and storage medium | |
| CN112270686B (en) | Image segmentation model training method, image segmentation device and electronic equipment | |
| CN114550053A (en) | Traffic accident responsibility determination method, device, computer equipment and storage medium | |
| CN112395979A (en) | Image-based health state identification method, device, equipment and storage medium | |
| WO2022111387A1 (en) | Data processing method and related apparatus | |
| CN112995414B (en) | Behavior quality inspection method, device, equipment and storage medium based on voice call | |
| CN115956889A (en) | Blood pressure monitoring method and device and electronic equipment | |
| CN116935449A (en) | Fingerprint image matching model training method, fingerprint matching method and related medium | |
| CN114970357A (en) | Energy-saving effect evaluation method, system, device and storage medium | |
| CN113379045A (en) | Data enhancement method and device | |
| CN110414586A (en) | Antifalsification label based on deep learning tests fake method, device, equipment and medium | |
| CN113240071A (en) | Graph neural network processing method and device, computer equipment and storage medium | |
| CN112885377A (en) | Voice quality evaluation method and device, computer equipment and storage medium | |
| CN111783688A (en) | Remote sensing image scene classification method based on convolutional neural network | |
| CN111414922B (en) | Feature extraction method, image processing method, model training method and device | |
| CN116712048A (en) | Blood pressure measurement method and device and related equipment | |
| CN111275752A (en) | Water flow velocity measurement method and device, computer equipment and storage medium | |
| CN117037917A (en) | Cell type prediction model training method, cell type prediction method and device | |
| CN115131600A (en) | Detection model training method, detection method, device, equipment and storage medium | |
| CN113139490A (en) | Image feature matching method and device, computer equipment and storage medium | |
| CN116777869A (en) | Image detection method, device, equipment and storage medium based on artificial intelligence |
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 |