CN112336326A - Volume pulse wave signal processing method, blood pressure measuring device, and storage medium - Google Patents
Volume pulse wave signal processing method, blood pressure measuring device, and storage medium Download PDFInfo
- Publication number
- CN112336326A CN112336326A CN202011129244.1A CN202011129244A CN112336326A CN 112336326 A CN112336326 A CN 112336326A CN 202011129244 A CN202011129244 A CN 202011129244A CN 112336326 A CN112336326 A CN 112336326A
- Authority
- CN
- China
- Prior art keywords
- pulse wave
- volume pulse
- network
- wave signal
- blood 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.)
- 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
- A61B5/7203—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal
-
- 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
- A61B5/7225—Details of analog processing, e.g. isolation amplifier, gain or sensitivity adjustment, filtering, baseline or drift compensation
-
- 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
- A61B5/7246—Details of waveform analysis using correlation, e.g. template matching or determination of similarity
-
- 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
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- Molecular Biology (AREA)
- Public Health (AREA)
- Pathology (AREA)
- Physiology (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Artificial Intelligence (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- General Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Veterinary Medicine (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Psychiatry (AREA)
- Cardiology (AREA)
- Vascular Medicine (AREA)
- Evolutionary Computation (AREA)
- Fuzzy Systems (AREA)
- Mathematical Physics (AREA)
- Power Engineering (AREA)
- Measuring Pulse, Heart Rate, Blood Pressure Or Blood Flow (AREA)
Abstract
The invention discloses a volume pulse wave signal processing method, a blood pressure measuring device and a storage medium, wherein the volume pulse wave signal processing method comprises the steps of extracting single-period waveform morphological characteristics from a volume pulse wave signal, acquiring space-time correlation characteristics of the single-period waveform morphological characteristics by using a bidirectional circulation network, inputting the space-time correlation characteristics into a shared depth convolution network to acquire a characteristic map, inputting the characteristic map into a full-connection network to acquire a blood pressure fitting value and the like. By processing the volume pulse wave signals by using a bidirectional circulation network, a shared deep convolution network, a full-connection network and the like, high-dimensional information can be extracted from the volume pulse wave signals, so that more available information is provided for the application of a pulse wave analysis method, and the precision of the pulse wave analysis method is improved. The invention is widely applied to the technical field of signal processing.
Description
Technical Field
The invention relates to the technical field of signal processing, in particular to a volume pulse wave signal processing method, a blood pressure measuring device and a storage medium.
Background
The blood pressure is an important physiological index, and the blood pressure of a human body can be measured by an arterial tension method, a volume compensation method, a pulse wave transmission method, a pulse wave analysis method and the like. The sum pulse wave analysis method is used for processing the measured pulse wave signals to obtain the blood pressure value, the pulse wave conduction method needs to collect and process multi-path volume pulse signals, and the pulse wave analysis method only needs to collect and process single-path volume pulse signals, so that the pulse wave analysis method has more advantages than the pulse wave conduction method. The existing pulse wave analysis technology mainly has the defects that high-dimensional information of volume pulse wave signals cannot be fully mined and the like.
Disclosure of Invention
In view of at least one of the above-mentioned problems, it is an object of the present invention to provide a volume pulse wave signal processing method, a blood pressure measuring device, and a storage medium.
In one aspect, an embodiment of the present invention includes a method for processing a volume pulse wave signal, including:
extracting monocycle waveform morphological characteristics from the volume pulse wave signal; the monocycle waveform morphology features describe one of the first cycles of the volume pulse wave signal;
acquiring space-time correlation characteristics of the single-period waveform morphological characteristics by using a bidirectional circulation network;
inputting the space-time correlation characteristics into a shared deep convolutional network to obtain a characteristic diagram output by the shared deep convolutional network;
and inputting the characteristic diagram into a fully-connected network, and acquiring a blood pressure fitting value output by the fully-connected network.
Further, the extracting the monocycle waveform morphology features from the volume pulse wave signal comprises:
determining peak characteristic points and trough characteristic points in the volume pulse wave signals;
and extracting partial volume pulse wave signals between the peak characteristic points and the trough characteristic points to serve as the form characteristics of the monocycle waveform.
Further, the obtaining the space-time correlation characteristics of the morphological characteristics of the monocycle waveform by using a bidirectional circulation network comprises:
extracting a second periodic waveform morphology feature and a third periodic waveform morphology feature from the volume pulse wave signal; the second periodic waveform morphology feature describes a second period of the volume pulse wave signal, the third periodic waveform morphology feature describes a third period of the volume pulse wave signal, the second period is one or more periods before the first period, the third period is one or more periods after the first period, the second period and the third period are both adjacent to the first period;
extracting a first potential morphological feature from the monocycle waveform morphological feature, a second potential morphological feature from the second periodic waveform morphological feature, a third potential morphological feature from the third periodic waveform morphological feature using a convolutional neural network;
and inputting the first potential morphological characteristics, the second potential morphological characteristics and the third potential morphological characteristics into the bidirectional circulation network, and acquiring the space-time correlation characteristics output by the bidirectional circulation network.
Further, the inputting the spatio-temporal correlation features into a shared deep convolutional network to obtain a feature map output by the shared deep convolutional network includes:
setting a blood pressure fitting subtask for the shared deep convolutional network;
inputting the spatiotemporal correlation features into a shared deep convolutional network;
acquiring a feature map output by the shared deep convolutional network according to the time-space correlation feature to execute the blood pressure fitting subtask;
and setting a channel-by-channel attention mechanism, and taking a multiplication result of the attention diagram and the feature diagram as a result of the blood pressure fitting subtask.
Further, the volume pulse wave signal processing method further includes:
removing baseline drift noise and high frequency noise in the volume pulse wave signal using a butterworth low-pass and high-pass filter bank;
removing the burr noise in the volume pulse wave signal by using a median filter.
Further, the volume pulse wave signal processing method further includes:
acquiring a waveform form template;
determining similarity between the monocycle waveform morphology features and the waveform morphology template;
and tracking the time sequence change of the single-period waveform morphological characteristics based on a Kalman model with constraints, and determining the effectiveness rating of the single-period waveform morphological characteristics according to the similarity.
Further, the volume pulse wave signal processing method further includes:
inputting the space-time correlation characteristics into a deconvolution network to obtain a one-dimensional waveform signal output by the deconvolution network; the one-dimensional waveform signal is used as a reference value in a training process of the bidirectional cyclic network, the shared deep convolutional network and the fully-connected network.
Further, in the training process of the bidirectional cyclic network, the shared deep convolutional network and the fully-connected network, the loss function used is that
Wherein SBP is systolic pressure value, DBP is diastolic pressure value, MAP is mean arterial pressure value,
for the fitted value, Y is the reference value, L (-) is the absolute error loss function, wiThe weight values of subtask i are fitted to the blood pressure.
On the other hand, the embodiment of the invention also comprises a blood pressure measuring device, which comprises:
the signal acquisition module is used for acquiring volume pulse signals;
a signal processing module for performing the method of an embodiment to obtain the blood pressure fit value;
and the signal output module is used for outputting the blood pressure fitting value.
In another aspect, an embodiment of the present invention further includes a computer apparatus, including a memory and a processor, where the memory is used to store at least one program, and the processor is used to load the at least one program to perform the method of the embodiment.
In another aspect, the present embodiment further includes a storage medium in which a processor-executable program is stored, which, when executed by a processor, is used to perform the volume pulse wave signal processing method in the embodiment.
The invention has the beneficial effects that: by processing the volume pulse wave signals by using a bidirectional circulation network, a shared deep convolution network, a full-connection network and the like, high-dimensional information can be extracted from the volume pulse wave signals, so that more available information is provided for the application of a pulse wave analysis method, and the precision of the pulse wave analysis method is improved.
Drawings
FIG. 1 is a flow chart of a method for processing a volume pulse wave signal according to an embodiment;
fig. 2 is a schematic structural diagram of a blood pressure measuring device in an embodiment.
Detailed Description
In this embodiment, referring to fig. 1, the method for processing the volume pulse wave signal includes the following steps:
s1, extracting single-period waveform morphological characteristics from volume pulse wave signals, wherein the single-period waveform morphological characteristics are morphological characteristics of one single period of the volume pulse wave signals, the single period is called a first period in the embodiment, and the single-period waveform morphological characteristics can describe the property of the first period;
s2, acquiring space-time correlation characteristics of the morphological characteristics of the monocycle waveform by using a bidirectional circulation network;
s3, inputting the space-time correlation characteristics into a shared deep convolutional network to obtain a characteristic diagram output by the shared deep convolutional network;
and S4, inputting the characteristic graph into the full-connection network to obtain a blood pressure fitting value output by the full-connection network.
In this embodiment, when step S1 is executed, the following steps are specifically executed:
s101, determining peak characteristic points and trough characteristic points in volume pulse wave signals;
s102, extracting partial volume pulse wave signals between the peak characteristic points and the trough characteristic points to serve as single-cycle waveform morphological characteristics.
Since the peak feature points and the trough feature points are relatively less interfered by noise, the first period in the volume pulse wave signal can be accurately divided by the peak feature points and the trough feature points when steps S101 to S102 are executed in this embodiment.
In this embodiment, in step S2, the step of obtaining the spatiotemporal correlation characteristic of the morphological characteristic of the monocycle waveform by using the bidirectional loop network includes:
s201, extracting second periodic waveform morphological characteristics and third periodic waveform morphological characteristics from the volume pulse wave signal; wherein the second periodic waveform morphology feature describes a second period of the volume pulse wave signal, and the third periodic waveform morphology feature describes a third period of the volume pulse wave signal, specifically, the second period is one or more periods before the first period, the third period is one or more periods after the first period, and both the second period and the third period are adjacent to the first period;
s202, extracting a first potential morphological feature from the single-period waveform morphological feature, a second potential morphological feature from the second period waveform morphological feature, and a third potential morphological feature from the third period waveform morphological feature by using a convolutional neural network;
s203, inputting the first potential morphological characteristics, the second potential morphological characteristics and the third potential morphological characteristics into a bidirectional circulation network, and obtaining space-time correlation characteristics output by the bidirectional circulation network.
Because the effectiveness and robustness of the pulse wave analysis method depend on the signal quality and the waveform form, if the signal quality and the waveform form of the pulse wave in the current cardiac cycle cannot meet the requirements of the pulse wave analysis method, the accuracy and the stability of the blood pressure measurement cannot be ensured. In this embodiment, steps S201 to S203 are executed to extract the morphological features of the monocycle waveform through the convolutional neural network, and first, the positioning of the peak and trough feature points of the pulse wave signal is less interfered by noise, and can be accurately segmented into the monocycle waveform signal. Secondly, a deep convolutional network is adopted to extract the potential morphological characteristics of the monocycle waveform, so that the uncertainty of extracting morphological characteristic parameters by artificial characteristic engineering is avoided.
The blood pressure value of the current cardiac cycle is determined by the waveform morphological characteristics of the volume pulse wave signal of the current cycle and the morphological characteristic variability of a plurality of continuous cardiac cycle signals before and after the current cycle. In the embodiment, a bidirectional circulation network is adopted to extract the space-time correlation morphological characteristics of the pulse wave signals of the current cardiac cycle. The method mainly comprises the following steps: firstly, selecting a plurality of continuous periodic waveforms before and after the current periodic waveform, and extracting potential morphological characteristics of each periodic waveform by adopting a convolution network; secondly, the extracted potential waveform morphological characteristics are input to a bidirectional circulation network in sequence, and the space-time correlation characteristics of the waveform in the current period are obtained through calculation.
In this embodiment, the step S3, that is, the step of inputting the spatio-temporal correlation features into the shared deep convolutional network to obtain the feature map output by the shared deep convolutional network, includes:
s301, setting a blood pressure fitting subtask for the shared depth convolution network;
s302, inputting the space-time correlation characteristics into a shared depth convolution network;
s303, acquiring a feature map output by the shared depth convolution network according to the time-space correlation features to execute a blood pressure fitting subtask;
and S304, setting a channel-by-channel attention mechanism, and taking the result of multiplying the attention diagram and the feature diagram as the result of the blood pressure fitting subtask.
In step S301, a blood pressure fitting subtask may be set for each measurement of blood pressure values. In this embodiment, a blood pressure fitting subtask may be set for measurement of the systolic blood pressure value SBP, the diastolic blood pressure value DBP, and the average arterial pressure value MAP, respectively.
In this embodiment, a multitask deep network model is used for joint training learning to fit the blood pressure value. Aiming at the difference between different tasks of blood pressure fitting, an attention mechanism is introduced between a blood pressure fitting sub-network and a shared network module, and the high resolution of the blood pressure fitting sub-task is ensured. The specific idea for constructing the continuous blood pressure model is as follows: and inputting the blood pressure related characteristics into a shared deep convolutional network to generate a series of characteristic maps. And introducing a channel-by-channel attention mechanism, and multiplying the attention diagram by the feature diagram to obtain feature adaptive learning of different blood pressure fitting subtasks. And (3) inputting the characteristic diagram obtained by self-adaptive learning into a full-connection sub-network aiming at different blood pressure fitting subtasks, and outputting the fitted systolic pressure value SBP, diastolic pressure value DBP and mean arterial pressure value MAP.
In this embodiment, the following steps may be further performed when the method for processing the volume pulse wave signal is performed:
and S5, inputting the space-time correlation characteristics into a deconvolution network, and acquiring a one-dimensional waveform signal output by the deconvolution network, wherein the one-dimensional waveform signal is used as a reference value in the training process of the bidirectional cyclic network, the shared deep convolution network and the full-connection network.
In this embodiment, the dimension of the one-dimensional waveform signal is the same as the dimension of the original signal input into the bidirectional loop network. According to the signal reconstruction theory, the waveform signal output by the deconvolution network is equal to the input pulse wave period waveform signal. Therefore, the network parameters of the morphological feature extraction model can be optimized by using the input signal data as the real reference value. In the training process, a heuristic intelligent optimization technology is adopted to optimize the network structure parameters and the model hyper-parameters.
In this embodiment, in the training process of the bidirectional cyclic network, the shared deep convolutional network and the fully-connected network, the loss function used is
Wherein SBP is systolic pressure value, DBP is diastolic pressure value, MAP is mean arterial pressure value,
for the fitted value, Y is a reference value, specifically,
is a fit to the systolic blood pressure value,
is a fitting value of the diastolic blood pressure value,
is flatFitting value of mean arterial pressure value, YSBPAs a reference or actual value of the systolic pressure value, YDBPIs a reference or actual value of the diastolic blood pressure value, YMAPIs a reference value or an actual value of the mean arterial pressure value. L (-) is an absolute error loss function, wiThe weight values of subtask i are fitted to the blood pressure.
According to the AAMI standard, the difference between the real blood pressure value and the fitting blood pressure value is within +/-5 mmHg, and the clinical blood pressure monitoring requirement is considered to be met. Therefore, in this embodiment, the blood pressure fitting error loss function is improved, and in the iterative optimization process of the model, if a certain fitting blood pressure error loss meets the AAMI standard, the model stops optimizing the subtask. If all of the fitting blood pressure subtasks meet the AAMI standard, the model optimizes all of the subtasks. Specifically, the loss function may be expressed as:
in addition, how to determine the blood pressure fitting subtask loss function weight is crucial to the final blood pressure fitting effect. Due to the huge calculation amount of the method based on the grid search, the training time is too long. Therefore, in this embodiment, the blood pressure fitting subtask loss function weight may be determined based on the training set error loss trend and the check set error loss trend in the model training process.
In this embodiment, before executing steps S1-S5, the following steps may also be executed:
p1, removing baseline drift noise and high-frequency noise in the volume pulse wave signal by using a Butterworth low-pass filter bank and a Butterworth high-pass filter bank;
p2. use median filter to remove the glitch noise in the volume pulse wave signal.
P3, obtaining a waveform form template;
p4, determining the similarity between the single-period waveform morphological characteristics and the waveform morphological template;
p5., tracking the time sequence change of the single-period waveform morphological characteristics based on a Kalman model with constraints, and determining the effectiveness rating of the single-period waveform morphological characteristics according to the similarity.
Steps P1-P5 are pre-processing of the volume pulse wave signals. Through the steps P1-P2, the interference of the outside and the individual limb actions can be relieved, and the quality of the volume pulse wave signals is improved. Through the steps P3-P5, the artifact signals which are difficult to remove in the volume pulse wave signals can be effectively dealt with.
In this embodiment, the volume pulse wave signals are processed by using a bidirectional cyclic network, a shared deep convolutional network, a full-connection network, and the like, and high-dimensional information can be extracted from the volume pulse wave signals, so that more available information is provided for the application of the pulse wave analysis method, and the precision of the pulse wave analysis method is improved.
In this embodiment, referring to fig. 2, the blood pressure measuring apparatus includes:
the signal acquisition module is used for acquiring volume pulse signals;
the signal processing module is used for executing the volume pulse wave signal processing method in the embodiment to obtain a blood pressure fitting value;
and the signal output module is used for outputting the blood pressure fitting value.
In this embodiment, the signal acquisition module may be a sensor or other data interface to obtain the volume pulse signal directly from the human body or from the data storage. The signal processing module is a hardware module with a data processing function and can run a computer program to execute the volume pulse wave signal processing method so as to obtain a blood pressure fitting value. In this embodiment, signal output module can be display screen, printer, data record appearance or internet upload module, and it can directly show, save in local or high in the clouds blood pressure fitting value. The blood pressure measuring device in the embodiment may perform the volume pulse wave signal processing method in the embodiment, and since the volume pulse wave signal processing method in the embodiment processes the volume pulse wave signal by using the bidirectional loop network, the shared deep convolution network, the full connection network, and the like, high dimensional information may be extracted from the volume pulse wave signal, and the accuracy of the pulse wave analysis method may be improved, the blood pressure measuring device in the embodiment has a high accuracy of the blood pressure value measurement performance.
In the present embodiment, a storage medium in which a program executable by a processor is stored, the program executable by the processor being for executing the volume pulse wave signal processing method in the embodiment, achieves the same technical effects as described in the embodiment.
It should be noted that, unless otherwise specified, when a feature is referred to as being "fixed" or "connected" to another feature, it may be directly fixed or connected to the other feature or indirectly fixed or connected to the other feature. Furthermore, the descriptions of upper, lower, left, right, etc. used in the present disclosure are only relative to the mutual positional relationship of the constituent parts of the present disclosure in the drawings. As used in this disclosure, the singular forms "a", "an", and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. In addition, unless defined otherwise, all technical and scientific terms used in this example have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in the description of the embodiments herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used in this embodiment, the term "and/or" includes any combination of one or more of the associated listed items.
It will be understood that, although the terms first, second, third, etc. may be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element of the same type from another. For example, a first element could be termed a second element, and, similarly, a second element could be termed a first element, without departing from the scope of the present disclosure. The use of any and all examples, or exemplary language ("e.g.," such as "or the like") provided with this embodiment is intended merely to better illuminate embodiments of the invention and does not pose a limitation on the scope of the invention unless otherwise claimed.
It should be recognized that embodiments of the present invention can be realized and implemented by computer hardware, a combination of hardware and software, or by computer instructions stored in a non-transitory computer readable memory. The methods may be implemented in a computer program using standard programming techniques, including a non-transitory computer-readable storage medium configured with the computer program, where the storage medium so configured causes a computer to operate in a specific and predefined manner, according to the methods and figures described in the detailed description. Each program may be implemented in a high level procedural or object oriented programming language to communicate with a computer system. However, the program(s) can be implemented in assembly or machine language, if desired. In any case, the language may be a compiled or interpreted language. Furthermore, the program can be run on a programmed application specific integrated circuit for this purpose.
Further, operations of processes described in this embodiment can be performed in any suitable order unless otherwise indicated herein or otherwise clearly contradicted by context. The processes described in this embodiment (or variations and/or combinations thereof) may be performed under the control of one or more computer systems configured with executable instructions, and may be implemented as code (e.g., executable instructions, one or more computer programs, or one or more applications) collectively executed on one or more processors, by hardware, or combinations thereof. The computer program includes a plurality of instructions executable by one or more processors.
Further, the method may be implemented in any type of computing platform operatively connected to a suitable interface, including but not limited to a personal computer, mini computer, mainframe, workstation, networked or distributed computing environment, separate or integrated computer platform, or in communication with a charged particle tool or other imaging device, and the like. Aspects of the invention may be embodied in machine-readable code stored on a non-transitory storage medium or device, whether removable or integrated into a computing platform, such as a hard disk, optically read and/or write storage medium, RAM, ROM, or the like, such that it may be read by a programmable computer, which when read by the storage medium or device, is operative to configure and operate the computer to perform the procedures described herein. Further, the machine-readable code, or portions thereof, may be transmitted over a wired or wireless network. The invention described in this embodiment includes these and other different types of non-transitory computer-readable storage media when such media include instructions or programs that implement the steps described above in conjunction with a microprocessor or other data processor. The invention also includes the computer itself when programmed according to the methods and techniques described herein.
A computer program can be applied to input data to perform the functions described in the present embodiment to convert the input data to generate output data that is stored to a non-volatile memory. The output information may also be applied to one or more output devices, such as a display. In a preferred embodiment of the invention, the transformed data represents physical and tangible objects, including particular visual depictions of physical and tangible objects produced on a display.
The above description is only a preferred embodiment of the present invention, and the present invention is not limited to the above embodiment, and any modifications, equivalent substitutions, improvements, etc. within the spirit and principle of the present invention should be included in the protection scope of the present invention as long as the technical effects of the present invention are achieved by the same means. The invention is capable of other modifications and variations in its technical solution and/or its implementation, within the scope of protection of the invention.
Claims (10)
1. A method for processing a volume pulse wave signal, comprising:
extracting monocycle waveform morphological characteristics from the volume pulse wave signal; the monocycle waveform morphology features describe one of the first cycles of the volume pulse wave signal;
acquiring space-time correlation characteristics of the single-period waveform morphological characteristics by using a bidirectional circulation network;
inputting the space-time correlation characteristics into a shared deep convolutional network to obtain a characteristic diagram output by the shared deep convolutional network;
and inputting the characteristic diagram into a fully-connected network, and acquiring a blood pressure fitting value output by the fully-connected network.
2. The method for processing a volume pulse wave signal according to claim 1, wherein the extracting a monocycle waveform morphological feature from the volume pulse wave signal comprises:
determining peak characteristic points and trough characteristic points in the volume pulse wave signals;
and extracting partial volume pulse wave signals between the peak characteristic points and the trough characteristic points to serve as the form characteristics of the monocycle waveform.
3. The method for processing the volume pulse wave signal according to claim 1, wherein the obtaining the spatiotemporal correlation characteristic of the monocycle waveform morphology characteristic by using a bidirectional circulation network comprises:
extracting a second periodic waveform morphology feature and a third periodic waveform morphology feature from the volume pulse wave signal; the second periodic waveform morphology feature describes a second period of the volume pulse wave signal, the third periodic waveform morphology feature describes a third period of the volume pulse wave signal, the second period is one or more periods before the first period, the third period is one or more periods after the first period, the second period and the third period are both adjacent to the first period;
extracting a first potential morphological feature from the monocycle waveform morphological feature, a second potential morphological feature from the second periodic waveform morphological feature, a third potential morphological feature from the third periodic waveform morphological feature using a convolutional neural network;
and inputting the first potential morphological characteristics, the second potential morphological characteristics and the third potential morphological characteristics into the bidirectional circulation network, and acquiring the space-time correlation characteristics output by the bidirectional circulation network.
4. The method for processing the volume pulse wave signal according to claim 1, wherein the inputting the spatio-temporal correlation features into a shared deep convolutional network to obtain a feature map output by the shared deep convolutional network comprises:
setting a blood pressure fitting subtask for the shared deep convolutional network;
inputting the spatiotemporal correlation features into a shared deep convolutional network;
acquiring a feature map output by the shared deep convolutional network according to the time-space correlation feature to execute the blood pressure fitting subtask;
and setting a channel-by-channel attention mechanism, and taking a multiplication result of the attention diagram and the feature diagram as a result of the blood pressure fitting subtask.
5. The volume pulse wave signal processing method according to any one of claims 1 to 4, further comprising:
removing baseline drift noise and high frequency noise in the volume pulse wave signal using a butterworth low-pass and high-pass filter bank;
removing the burr noise in the volume pulse wave signal by using a median filter.
6. The volume pulse wave signal processing method according to claim 5, further comprising:
acquiring a waveform form template;
determining similarity between the monocycle waveform morphology features and the waveform morphology template;
and tracking the time sequence change of the single-period waveform morphological characteristics based on a Kalman model with constraints, and determining the effectiveness rating of the single-period waveform morphological characteristics according to the similarity.
7. The volume pulse wave signal processing method according to any one of claims 1 to 4, further comprising:
inputting the space-time correlation characteristics into a deconvolution network to obtain a one-dimensional waveform signal output by the deconvolution network; the one-dimensional waveform signal is used as a reference value in a training process of the bidirectional cyclic network, the shared deep convolutional network and the fully-connected network.
8. The method for processing a plethysmographic signal according to claim 7, wherein in the training of the bi-directional cyclic network, the shared deep convolutional network and the fully connected network, the loss function used is
Wherein SBP is systolic pressure value, DBP is diastolic pressure value, MAP is mean arterial pressure value,
for the fitted value, Y is the reference value, L (-) is the absolute error loss function, wiThe weight values of subtask i are fitted to the blood pressure.
9. A blood pressure measuring device, comprising:
the signal acquisition module is used for acquiring volume pulse signals;
a signal processing module for performing the method of any one of claims 1-8 to obtain the blood pressure fit values;
and the signal output module is used for outputting the blood pressure fitting value.
10. A storage medium having stored therein a program executable by a processor, wherein the program executable by the processor is adapted to perform the method of any one of claims 1-8 when executed by the processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011129244.1A CN112336326B (en) | 2020-10-21 | 2020-10-21 | Volume pulse wave signal processing method, blood pressure measuring device, and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011129244.1A CN112336326B (en) | 2020-10-21 | 2020-10-21 | Volume pulse wave signal processing method, blood pressure measuring device, and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112336326A true CN112336326A (en) | 2021-02-09 |
| CN112336326B CN112336326B (en) | 2023-09-05 |
Family
ID=74359385
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011129244.1A Active CN112336326B (en) | 2020-10-21 | 2020-10-21 | Volume pulse wave signal processing method, blood pressure measuring device, and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112336326B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116383617A (en) * | 2023-04-21 | 2023-07-04 | 复旦大学 | Intelligent blood pressure detection method and system based on pulse wave waveform characteristics |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106821356A (en) * | 2017-02-23 | 2017-06-13 | 吉林大学 | High in the clouds continuous BP measurement method and system based on Elman neutral nets |
| CN107157461A (en) * | 2017-07-06 | 2017-09-15 | 中国科学院合肥物质科学研究院 | Noninvasive continuous BP measurement method based on photoplethysmographic |
| CN107669256A (en) * | 2017-11-13 | 2018-02-09 | 深圳先进技术研究院 | Measuring method, the device and system of continuous blood pressure |
| CN109833034A (en) * | 2017-11-24 | 2019-06-04 | 深圳市岩尚科技有限公司 | The method and device of blood pressure data is extracted in a kind of pulse wave signal |
| CN110322962A (en) * | 2019-07-03 | 2019-10-11 | 重庆邮电大学 | A kind of method automatically generating diagnostic result, system and computer equipment |
| CN110840427A (en) * | 2019-10-25 | 2020-02-28 | 广州视源电子科技股份有限公司 | Continuous blood pressure measuring method, device and equipment based on volume pulse wave signals |
| CN111160139A (en) * | 2019-12-13 | 2020-05-15 | 中国科学院深圳先进技术研究院 | Electrocardiosignal processing method and device and terminal equipment |
| US20200221960A1 (en) * | 2019-01-11 | 2020-07-16 | Samsung Electronics Co., Ltd. | Apparatus and method for estimating blood pressure |
| CN111493850A (en) * | 2020-04-13 | 2020-08-07 | 中国科学院深圳先进技术研究院 | Blood pressure measuring method and device based on deep neural network |
| CN111528814A (en) * | 2020-04-29 | 2020-08-14 | 浙江工业大学 | Method for monitoring blood pressure through machine learning based on LSTM neural network |
-
2020
- 2020-10-21 CN CN202011129244.1A patent/CN112336326B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106821356A (en) * | 2017-02-23 | 2017-06-13 | 吉林大学 | High in the clouds continuous BP measurement method and system based on Elman neutral nets |
| CN107157461A (en) * | 2017-07-06 | 2017-09-15 | 中国科学院合肥物质科学研究院 | Noninvasive continuous BP measurement method based on photoplethysmographic |
| CN107669256A (en) * | 2017-11-13 | 2018-02-09 | 深圳先进技术研究院 | Measuring method, the device and system of continuous blood pressure |
| CN109833034A (en) * | 2017-11-24 | 2019-06-04 | 深圳市岩尚科技有限公司 | The method and device of blood pressure data is extracted in a kind of pulse wave signal |
| US20200221960A1 (en) * | 2019-01-11 | 2020-07-16 | Samsung Electronics Co., Ltd. | Apparatus and method for estimating blood pressure |
| CN110322962A (en) * | 2019-07-03 | 2019-10-11 | 重庆邮电大学 | A kind of method automatically generating diagnostic result, system and computer equipment |
| CN110840427A (en) * | 2019-10-25 | 2020-02-28 | 广州视源电子科技股份有限公司 | Continuous blood pressure measuring method, device and equipment based on volume pulse wave signals |
| CN111160139A (en) * | 2019-12-13 | 2020-05-15 | 中国科学院深圳先进技术研究院 | Electrocardiosignal processing method and device and terminal equipment |
| CN111493850A (en) * | 2020-04-13 | 2020-08-07 | 中国科学院深圳先进技术研究院 | Blood pressure measuring method and device based on deep neural network |
| CN111528814A (en) * | 2020-04-29 | 2020-08-14 | 浙江工业大学 | Method for monitoring blood pressure through machine learning based on LSTM neural network |
Non-Patent Citations (1)
| Title |
|---|
| 张佳骕 等: "基于卷积递归神经网络的血压测量模型", 《北京生物医学工程》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116383617A (en) * | 2023-04-21 | 2023-07-04 | 复旦大学 | Intelligent blood pressure detection method and system based on pulse wave waveform characteristics |
| CN116383617B (en) * | 2023-04-21 | 2023-09-22 | 复旦大学 | Intelligent blood pressure detection method and system based on pulse wave waveform characteristics |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112336326B (en) | 2023-09-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110141215B (en) | Training method of noise reduction self-encoder, noise reduction method of electrocardiosignal, related device and equipment | |
| Morettin et al. | Wavelets in functional data analysis | |
| EP3752048A1 (en) | Monitoring device for monitoring a physiological parameter and methods thereof | |
| CN109512395B (en) | Method, device and equipment for analyzing and processing biological signals | |
| CN106798552A (en) | Heart rate with status switch optimization estimates equipment | |
| CN111588367A (en) | Heart rate detection method and device and computer readable storage medium | |
| US9729693B1 (en) | Determining measurement confidence for data collected from sensors of a wearable device | |
| Kong et al. | Medical image fusion using guided filter random walks and spatial frequency in framelet domain | |
| CN112336326B (en) | Volume pulse wave signal processing method, blood pressure measuring device, and storage medium | |
| CN112836546A (en) | Method and device for detecting quality of physiological signal and electronic equipment | |
| Sundar et al. | A comprehensive assessment of the performance of modern algorithms for enhancement of digital volume pulse signals | |
| CN114052693A (en) | Heart rate analysis method, device and equipment | |
| CN113647905A (en) | Training method for predicting blood pressure by deep neural network, computer device and storage medium | |
| CN111860102A (en) | Apparatus and method for analyzing state of system in noisy environment | |
| EP1665160B1 (en) | System and method for spatio-temporal modeling using guidepoints | |
| CN116712056B (en) | Characteristic image generation and identification method, equipment and storage medium for electrocardiogram data | |
| Lee et al. | Lstc-rppg: Long short-term convolutional network for remote photoplethysmography | |
| CN111278353A (en) | Method and system for detecting vital sign signal noise | |
| CN117357080A (en) | Near infrared spectrum signal denoising method and device, terminal equipment and storage medium | |
| EP3770921A1 (en) | Method and system for pressure autoregulation based synthesizing of photoplethysmogram signal | |
| Schibon et al. | Saving energy on wrist-mounted inertial sensors by motion-adaptive duty-cycling in free-living | |
| van Gent | Python Heart Rate Analysis Toolkit Documentation | |
| Todo et al. | Dimension Reduction for time series with Variational AutoEncoders | |
| Zugal et al. | Cheetah experimental platform web 1.0: cleaning pupillary data | |
| Banerjee et al. | A New Moving Horizon Estimation Based Real-Time Motion Artifact Removal from Wavelet Subbands of ECG Signal Using Particle Filter |
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 |