CN115381438A - Method and device for reconstructing vital sign signals, computer equipment and storage medium - Google Patents
Method and device for reconstructing vital sign signals, computer equipment and storage medium Download PDFInfo
- Publication number
- CN115381438A CN115381438A CN202211016863.9A CN202211016863A CN115381438A CN 115381438 A CN115381438 A CN 115381438A CN 202211016863 A CN202211016863 A CN 202211016863A CN 115381438 A CN115381438 A CN 115381438A
- Authority
- CN
- China
- Prior art keywords
- signal
- body motion
- motion artifact
- normal
- reconstruction
- 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
- 238000000034 method Methods 0.000 title claims abstract description 58
- 238000003860 storage Methods 0.000 title claims abstract description 19
- 230000033001 locomotion Effects 0.000 claims abstract description 298
- 238000004590 computer program Methods 0.000 claims description 14
- 230000000877 morphologic effect Effects 0.000 claims description 9
- 238000013528 artificial neural network Methods 0.000 claims description 8
- 238000005070 sampling Methods 0.000 claims description 6
- 238000012935 Averaging Methods 0.000 claims description 4
- 238000012549 training Methods 0.000 claims description 4
- 238000001914 filtration Methods 0.000 claims description 2
- 230000011218 segmentation Effects 0.000 claims description 2
- 230000006870 function Effects 0.000 description 18
- 230000008569 process Effects 0.000 description 12
- 238000012545 processing Methods 0.000 description 11
- 238000010586 diagram Methods 0.000 description 10
- 238000004891 communication Methods 0.000 description 5
- 230000004913 activation Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000000605 extraction Methods 0.000 description 3
- 230000002452 interceptive effect Effects 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 230000002457 bidirectional effect Effects 0.000 description 2
- 239000008280 blood Substances 0.000 description 2
- 210000004369 blood Anatomy 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000013527 convolutional neural network Methods 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 239000012634 fragment Substances 0.000 description 2
- 230000004217 heart function Effects 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000036961 partial effect Effects 0.000 description 2
- 230000000306 recurrent effect Effects 0.000 description 2
- 230000002829 reductive effect Effects 0.000 description 2
- 238000002759 z-score normalization Methods 0.000 description 2
- 241001522296 Erithacus rubecula Species 0.000 description 1
- 108010076504 Protein Sorting Signals Proteins 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 239000000654 additive Substances 0.000 description 1
- 230000000996 additive effect Effects 0.000 description 1
- 108010038083 amyloid fibril protein AS-SAM Proteins 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 206010003119 arrhythmia Diseases 0.000 description 1
- 230000006793 arrhythmia Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 210000004204 blood vessel Anatomy 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 125000004122 cyclic group Chemical group 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 238000005315 distribution function Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000005484 gravity Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000000670 limiting effect Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000002503 metabolic effect Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 210000000056 organ Anatomy 0.000 description 1
- 230000004962 physiological condition Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000009877 rendering Methods 0.000 description 1
- 230000000241 respiratory effect Effects 0.000 description 1
- 238000013215 result calculation Methods 0.000 description 1
- 230000007958 sleep Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/103—Detecting, measuring or recording devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
- A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb
- A61B5/1102—Ballistocardiography
-
- 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
- A61B5/7207—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal of noise induced by motion artifacts
- A61B5/7214—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal of noise induced by motion artifacts using signal cancellation, e.g. based on input of two identical physiological sensors spaced apart, or based on two signals derived from the same sensor, for different optical wavelengths
Abstract
The application relates to a vital sign signal reconstruction method, a vital sign signal reconstruction device, computer equipment and a storage medium, wherein the method comprises the following steps: acquiring a vital sign signal; the vital sign signals comprise at least one body motion artifact signal, a first normal signal and a second normal signal; judging whether the body motion artifact signal can be reconstructed or not according to the first normal signal and the second normal signal; if the body motion artifact signal can be reconstructed, the first normal signal and the second normal signal are input into a preset body motion artifact signal reconstruction model to obtain a reconstruction signal corresponding to the body motion artifact signal, the reconstruction signal replaces the corresponding body motion artifact signal in the vital sign signal to obtain the reconstructed vital sign signal, and the accuracy of reconstructing the vital sign signal is improved.
Description
Technical Field
The present application relates to the field of signal processing technologies, and in particular, to a method and an apparatus for reconstructing a vital sign signal, a computer device, and a storage medium.
Background
The Ballistocardiogram (BCG) signal is a signal which can reflect the heart function of a human body and is acquired by undisturbed sensor signal acquisition equipment, and the generation mechanism of the BCG signal is that the heart pump blood contracts and blood quickly impacts blood vessels to enable the human body to generate slight shiver. The BCG signal is a weak force signal, and can be converted into an electric signal through the pressure sensor under the condition of non-direct contact, so that the interference-free collection of the vital sign signals of the user is realized.
As can be seen from the generation mechanism, the BCG signal is closely related to the human heart function, but the following characteristics are also inherent: (1) poor robustness: because the micro-vibration generated by the metabolic motion of each organ of the human body can interfere the BCG signal, the acquired BCG is unstable in form; (2) individual differences were significant: the BCG signal form can cause different heart activity conditions along with different physiological conditions of people; (3) environmental factors have a great influence: since the BCG signal is acquired by the pressure sensor, the distance between the heart of the patient and the sensor affects the acquisition quality of the acquired signal.
By combining the characteristics, the BCG signals are easily interfered by body movement artifacts collected by non-contact equipment in the collection process, partial segment waveforms are damaged, and subsequent further processing and analysis such as heart beat positioning, HRV and the like are seriously influenced. However, in the prior art, the signal damaged by the body motion artifact is reconstructed to obtain the original normal signal, and the accuracy is not high.
Disclosure of Invention
Based on this, it is an object of the present invention to provide a method, an apparatus, a computer device and a storage medium for reconstructing a vital sign signal, which have the advantage of improving the accuracy of the vital sign signal reconstruction.
According to a first aspect of embodiments of the present application, a method for reconstructing a vital sign signal is provided, which includes the following steps:
acquiring a vital sign signal; the vital sign signals comprise at least one body motion artifact signal, a first normal signal and a second normal signal; the first normal signal is a continuous normal signal which is in front of the body movement artifact signal and adjacent to the body movement artifact signal, and the second normal signal is a continuous normal signal which is behind the body movement artifact signal and adjacent to the body movement artifact signal;
judging whether the body motion artifact signal can be reconstructed or not according to the first normal signal and the second normal signal;
and if the body motion artifact signal can be reconstructed, inputting the first normal signal and the second normal signal into a preset body motion artifact signal reconstruction model to obtain a reconstruction signal corresponding to the body motion artifact signal.
According to a second aspect of embodiments of the present application, there is provided a vital sign signal reconstruction apparatus, including:
the signal acquisition module is used for acquiring vital sign signals; the vital sign signals comprise at least one body movement artifact signal, a first normal signal and a second normal signal; the first normal signal is a continuous normal signal which is before the body motion artifact signal and is adjacent to the body motion artifact signal, and the second normal signal is a continuous normal signal which is after the body motion artifact signal and is adjacent to the body motion artifact signal;
the signal judgment module is used for judging whether the body motion artifact signal can be reconstructed or not according to the first normal signal and the second normal signal;
the signal reconstruction module is used for inputting the first normal signal and the second normal signal into a preset body motion artifact signal reconstruction model if the body motion artifact signal can be reconstructed, and acquiring a reconstruction signal corresponding to the body motion artifact signal;
and the signal replacement module is used for replacing the corresponding body motion artifact signals in the vital sign signals with the reconstructed signals to obtain the reconstructed vital sign signals.
According to a third aspect of embodiments of the present application, there is provided a computer apparatus comprising: a processor and a memory; wherein the memory stores a computer program adapted to be loaded by the processor and to perform the method of vital sign signal reconstruction as defined in any of the above.
According to a fourth aspect of embodiments of the present application, there is provided a computer-readable storage medium, on which a computer program is stored, which computer program, when being executed by a processor, implements the method for reconstructing a vital sign signal as defined in any one of the above.
The embodiment of the application acquires the vital sign signals; the vital sign signals comprise at least one body motion artifact signal, a first normal signal and a second normal signal; the first normal signal is a continuous normal signal which is before the body motion artifact signal and is adjacent to the body motion artifact signal, and the second normal signal is a continuous normal signal which is after the body motion artifact signal and is adjacent to the body motion artifact signal; judging whether the body motion artifact signal can be reconstructed or not according to the first normal signal and the second normal signal; if the body motion artifact signal can be reconstructed, inputting the first normal signal and the second normal signal into a preset body motion artifact signal reconstruction model to obtain a reconstruction signal corresponding to the body motion artifact signal, replacing the corresponding body motion artifact signal in the vital sign signal with the reconstruction signal to obtain a reconstructed vital sign signal, and improving the accuracy of reconstruction of the vital sign signal.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the application.
For a better understanding and practice, the invention is described in detail below with reference to the accompanying drawings.
Drawings
Fig. 1 is a schematic flowchart of a vital sign signal reconstruction method according to an embodiment of the present application;
fig. 2 is a schematic flowchart of a step S20 in a vital sign signal reconstruction method according to an embodiment of the present application;
fig. 3 is a schematic flowchart of step S20 in a vital sign signal reconstruction method according to another embodiment of the present application;
fig. 4 is a schematic flowchart of step S1 in a vital sign signal reconstruction method according to an embodiment of the present application;
fig. 5 is a schematic flowchart of step S30 in a vital sign signal reconstruction method according to an embodiment of the present application;
fig. 6 is a schematic flowchart of step S303 in a vital sign signal reconstruction method according to an embodiment of the present application;
fig. 7 is a block diagram of a vital sign signal reconstruction apparatus according to an embodiment of the present application;
fig. 8 is a block diagram illustrating a structure of an electronic device according to an embodiment of the present application.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more clear, embodiments of the present application will be described in further detail below with reference to the accompanying drawings.
It should be understood that the embodiments described are only a few embodiments of the present application, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
The terminology used in the embodiments of the present application is for the purpose of describing particular embodiments only and is not intended to be limiting of the embodiments of the present application. As used in the examples of this application and the appended claims, the singular forms "a", "an", and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and/or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items.
When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the application, as detailed in the claims that follow. In the description of the present application, it is to be understood that the terms "first," "second," "third," and the like are used solely to distinguish one from another and are not necessarily used to describe a particular order or sequence, nor are they to be construed as indicating or implying relative importance. The specific meaning of the above terms in the present application can be understood by those of ordinary skill in the art as appropriate.
Further, in the description of the present application, "a plurality" means two or more unless otherwise specified. "and/or" describes the association relationship of the associated objects, meaning that there may be three relationships, e.g., a and/or B, which may mean: a exists alone, A and B exist simultaneously, and B exists alone. The character "/" generally indicates that the former and latter associated objects are in an "or" relationship.
Example 1
Referring to fig. 1, an embodiment of the present invention provides a method for reconstructing a vital sign signal, including the following steps:
s10: acquiring a vital sign signal; the vital sign signals comprise at least one body motion artifact signal, a first normal signal and a second normal signal; the first normal signal is a continuous normal signal which is in front of the body movement artifact signal and adjacent to the body movement artifact signal, and the second normal signal is a continuous normal signal which is behind the body movement artifact signal and adjacent to the body movement artifact signal.
In the embodiment of the present application, an execution subject of the reconstruction method for vital sign signals is reconstruction equipment for vital sign signals (hereinafter referred to as reconstruction equipment for short), where the reconstruction equipment may be one computer equipment, a server, or a server cluster formed by combining multiple computer equipments.
The reconstruction device may obtain the vital sign signal of the user by querying in a preset database. The reconstruction device may also employ an undisturbed sensor to acquire vital sign signals of the user in real time. The undisturbed sensor comprises a signal acquisition module and a data storage module.
Specifically, the signal acquisition module is placed below a user pillow, the user sleeps on the pillow, and the user generates slight vibration due to heart activity, respiratory activity and the like, so that the center of gravity shifts, and a force signal is generated. The signal acquisition module can convert the force signal into an analog electric signal, and then the analog electric signal is filtered, amplified and A/D converted into a digital signal with the sampling rate of 1000Hz, namely the vital sign signal of the user, through a built-in filter circuit, an amplifying circuit and an A/D conversion circuit. The vital sign signals comprise a plurality of normal signals and at least one body motion artifact signal which are continuous in time sequence.
Taking a certain body motion artifact signal as an example, a continuous normal signal adjacent in time sequence before the body motion artifact signal, that is, a first normal signal, is acquired, and for example, the signal length of the first normal signal is 6s. The signal length of a continuous normal signal adjacent in time sequence after the body motion artifact signal, that is, a second normal signal, for example, a second normal signal is acquired as 7s. And recording the first normal signal as sig _ left, and recording the second normal signal as sig _ right.
S20: and judging whether the body motion artifact signal can be rebuilt or not according to the first normal signal and the second normal signal.
In the embodiment of the present application, the feasibility of reconstructing the body motion artifact signal may be determined according to the signal lengths and the signal qualities of the first normal signal and the second normal signal. If the signal length is too short, the body motion artifact signal cannot be effectively restored to the original normal signal. If the signal quality is poor, even if the body motion artifact signal is accurately recovered, the signal processing method does not play any role in subsequent signal processing.
S30: and if the body motion artifact signal can be reconstructed, inputting the first normal signal and the second normal signal into a preset body motion artifact signal reconstruction model to obtain a reconstruction signal corresponding to the body motion artifact signal.
In the embodiment of the present application, the preset body motion artifact signal reconstruction model is a deep learning model, and specifically, is a Generative Adaptive Network (GAN), and includes a generator and a discriminator. The Generator (Generator) reconstructs the body motion artifact signal by receiving the normal signals at the front edge and the rear edge of the body motion artifact and restores the body motion artifact signal to the original normal signals.
The Discriminator (Discriminator) receives the real original signal and the signal generated by the generator, and is used for discriminating whether the two signals are consistent. During the training process, the goal of the generator is to try to generate a true signal to fool the arbiter. The goal of the discriminator is to try to distinguish the signal generated by the generating network from the true signal. Thus, the generator and the arbiter form a dynamic "gaming process". The result of the final game is: under optimal conditions the generator can generate enough signals to be "spurious". It is difficult for the arbiter to determine whether the signal generated by the generator is real or not.
After the body motion artifact signal can be reconstructed, inputting the first normal signal into a preset body motion artifact signal reconstruction model to obtain a forward predicted reconstruction signal; and inputting the second normal signal into a preset body motion artifact signal reconstruction model to obtain a backward prediction reconstruction signal. And averaging the forward predicted reconstruction signal and the backward predicted reconstruction signal to obtain a reconstruction signal corresponding to the body motion artifact signal. And after the body motion artifact signal is judged not to be rebuilt, skipping the body motion artifact signal, and judging whether the next body motion artifact signal can be rebuilt or not.
S40: and replacing the corresponding body motion artifact signals in the vital sign signals with the reconstructed signals to obtain the reconstructed vital sign signals.
In the embodiment of the application, all the reconstructable body motion artifact signals in the vital sign signals are replaced by corresponding reconstruction signals, and the reconstructed vital sign signals are obtained.
By applying the embodiment of the application, the vital sign signals are obtained; the vital sign signals comprise at least one body motion artifact signal, a first normal signal and a second normal signal; the first normal signal is a continuous normal signal which is before the body motion artifact signal and is adjacent to the body motion artifact signal, and the second normal signal is a continuous normal signal which is after the body motion artifact signal and is adjacent to the body motion artifact signal; judging whether the body motion artifact signal can be reconstructed or not according to the first normal signal and the second normal signal; if the body movement artifact signal can be reconstructed, inputting the first normal signal and the second normal signal into a preset body movement artifact signal reconstruction model to obtain a reconstruction signal corresponding to the body movement artifact signal, and replacing the corresponding body movement artifact signal in the vital sign signal with the reconstruction signal to obtain a reconstructed vital sign signal. According to the method and the device, the time sequence characteristics and the morphological characteristics of the front and back normal signals of the body movement artifact are extracted through the preset body movement artifact signal reconstruction model, the front and back bidirectional reconstruction is carried out on the body movement artifact signals according to the front and back normal signals, the reconstruction signals corresponding to the body movement artifact signals are obtained, and the reconstruction accuracy of the vital sign signals is improved.
In an alternative embodiment, referring to fig. 2, the step S20 includes steps S201 to S202, which are as follows:
s201: acquiring a first signal length of the first normal signal and a second signal length of the second normal signal;
s202: and if the first signal length and the second signal length are both greater than or equal to a first preset signal length, determining that the body motion artifact signal can be reconstructed.
In this embodiment of the present application, the first preset signal length is 5s, and if the first signal length of the first normal signal or the second signal length of the second normal signal is greater than or equal to 5s, it is determined that the body motion artifact signal can be reconstructed. And if the first signal length of the first normal signal or the second signal length of the second normal signal is less than 5s, skipping the reconstruction of the body motion artifact signal. Whether the body motion artifact signal has the feasibility of reconstruction is judged according to the signal length of the normal signal before and after the body motion artifact signal, and the reconstructable body motion artifact signal can be automatically and quickly screened.
In an alternative embodiment, referring to fig. 3, the step S20 includes steps S203 to S204, which are as follows:
s203: respectively calculating sample entropies, first-order standard deviations and second-order standard deviations of the first normal signal and the second normal signal;
s204: and if the sample entropies are all smaller than a preset first sample entropy threshold value and the sample entropies are all larger than a preset second sample entropy threshold value, the first-order standard deviations are all smaller than a preset first-order standard deviation threshold value, and the second-order standard deviations are all smaller than a preset second-order standard deviation threshold value, determining that the body motion artifact signal can be reconstructed.
Wherein, the Sample Entropy (Sample Entropy) measures the complexity of the time sequence by measuring the probability of generating a new pattern in the signal, and the greater the probability of generating the new pattern, the greater the complexity of the sequence. The method for calculating the sample entropy is the prior art, and is not described herein again.
In the embodiment of the present application, taking the first signal length of the first normal signal as 6s and the second signal length of the second normal signal as 7s as an example, the sample entropy is calculated for the first normal signal and the second normal signal respectively, and is denoted as sam _ left and sam _ right. The first normal signal is divided into 6 1s signal segments, and first-order standard deviations of the first normal signal are calculated as [ std _ left1, std _ left2, \8230 ], std _ left6], respectively. And calculating the standard deviation of the 6 first-order standard deviations to obtain a second-order standard deviation of the first normal signal, which is recorded as std2_ left. The second normal signal is divided into 7 1s signal segments, and the first-order standard deviations of the second normal signal are calculated as [ std _ right1, std _ right2, \ 8230;, std _ right7], respectively. The standard deviation is calculated from the 7 first-order standard deviations to obtain a second-order standard deviation of the second normal signal, which is denoted as std2_ right.
And setting an experience threshold value for signal quality division, wherein the preset first sample entropy threshold value is a sample entropy upper limit threshold value and is recorded as thr _ max _ sam, the preset second sample entropy threshold value is a sample entropy lower limit threshold value and is recorded as thr _ min _ sam, a preset first-order standard deviation threshold value thr _ std1, and a preset second-order standard deviation threshold value is recorded as thr _ std2. The sample entropy describes the time sequence complexity of the signal, the worse the signal quality is, the more messy and complicated the signal time sequence is, the larger the sample entropy is, so the sample entropy is smaller than the sample entropy upper threshold, the signal quality is judged to be better, and in addition, the sample entropy lower threshold is set to prevent the condition from being classified as a normal signal because the waveform is a straight line due to the fact that a part of the signal leaves the bed by a subject and the sample entropy is extremely small. And secondly, comparing the maximum value of the first-order standard deviation with thr _ std1, and ensuring that the discrete degree of each 1s signal segment is smaller than the first-order standard deviation threshold thr _ std1, so that each signal segment can be considered as a normal signal with better signal quality. The second-order standard deviation is used for measuring the discrete degree of the first-order standard deviation of each 1s signal segment, if the value is larger, the situation that signal fluctuation is serious, and the abnormal conditions of front and back signal forms or time sequences caused by arrhythmia and the like exist possibly is shown, so that the body motion artifact signal can be reconstructed only when the second-order standard deviation is smaller than a second-order standard deviation threshold thr _ std2.
In the embodiment of the present application, if the sample entropies of the first normal signal and the second normal signal are both smaller than a preset first sample entropy threshold, and the sample entropies are both larger than a preset second sample entropy threshold, and the first-order standard deviation is both smaller than a preset first-order standard deviation threshold, and the second-order standard deviation is both smaller than a preset second-order standard deviation threshold, it is determined that the body motion artifact signal can be reconstructed; and if one of the sample entropies, the first-order standard deviations and the second-order standard deviations of the first normal signal and the second normal signal does not meet a preset threshold, judging that the body motion artifact signal cannot be reconstructed.
Optionally, the signal lengths of the first normal signal and the second normal signal are determined first, and if the signal lengths both satisfy the first preset signal length, the signal qualities of the first normal signal and the second normal signal are determined, and whether the body motion artifact signal can be reconstructed is determined by combining the signal lengths and the signal qualities.
Whether the body motion artifact signal has the reconstruction necessity is judged through the sample entropy, the first-order standard deviation and the second-order standard deviation, and the reconstruction of a segment with poor signal quality is prevented, so that the time sequence of an original signal sequence is further damaged.
In an optional embodiment, before the step S30, steps S1 to S7 are included, which are specifically as follows:
s1: and acquiring normal signal sample data of a plurality of second preset signal lengths.
In the embodiment of the present application, the second preset signal length is 6s.
S2: dividing the normal signal sample data into a first sample data and a second sample data which are continuous in time sequence; the signal length of the first sample data is a first preset signal length, and the signal length of the second sample data is a difference between the second preset signal length and the first preset signal length.
In the embodiment of the present application, the 6s normal signal sample data is divided into the first 5s signal segment (first sample data) and the last 1s signal segment (second sample data).
S3: and taking the first sample data as the input of a generator in the generated confrontation network to obtain a first prediction result of the generator, and taking the first prediction result and the second sample data as the input of a discriminator in the generated confrontation network to obtain a first discrimination result of the discriminator.
In the embodiment of the application, 5s of first sample data is input to the generator to obtain 1s of first prediction results, and 1s of first prediction results and 1s of second sample data are input to the discriminator to obtain first discrimination results.
S4: dividing the normal signal sample data into third sample data and fourth sample data which are continuous in time sequence; the signal length of the third sample data is a third preset signal length, and the signal length of the fourth sample data is a difference between the second preset signal length and the third preset signal length.
In the embodiment of the present application, 6s of normal signal sample data is divided into a first 1s signal segment (third sample data) and a second 5s signal segment (fourth sample data).
S5: and taking the fourth sample data as the input of the generator to obtain a second prediction result of the generator, and taking both the second prediction result and the third sample data as the input of the discriminator to obtain a second discrimination result of the discriminator.
In the embodiment of the present application, 5s of fourth sample data is input to the generator, a second prediction result of 1s is obtained, and the second prediction result of 1s and the third sample data of 1s are input to the discriminator, so as to obtain a second discrimination result.
S6: continuously training the generator and the discriminator until the first discrimination result and the second discrimination result reach a preset threshold value, and obtaining a trained generated countermeasure network;
s7: and taking the trained generated confrontation network as a preset body motion artifact signal reconstruction model.
In the embodiment of the present application, the generator includes a gated round robin unit and a full convolution network, and the arbiter adopts a Multi Layer Perceptron (MLP) model including an input layer, a hidden layer, and an output layer. The discrimination result calculation formula of the discriminator is as follows:
ReLU(x)＝max(0，x)
h[n]＝ReLU(w[0]·x[0]+w[1]·x[1]+...+w[p]·x[p]+b)
where ReLU is the activation function, w and v are the weight parameters, b is the bias parameter, h [ n ]]As a result of the hidden layer(s),
is the result of the discrimination output by the output layer.
Generating an objective function of the countermeasure network, which is as follows:
wherein the content of the first and second substances,
expressing the expected value, p, of the distribution function data Representing the distribution of real data, p z (x) Representing the distribution of model input data.
The objective function of the discriminator D is specifically as follows:
when the true data label is 1 and the generated data label is 0, we expect that the true data is closer to 1, that is, D (x) is 1, and the generated data is closer to 0, D (G (x)) =0, so the objective function increases.
The objective function of the generator G is specifically as follows:
the generator is a desire to generate data, we expect D (G (x)) to tend to 1, so that the objective function is reduced because the true label is 1. During training of the generator, the arbiter needs to be fixed. The generator and the discriminator are trained respectively and alternately. The generator and the discriminator game with each other, the discriminator hopefully can discriminate the generated data, and the generator continuously optimizes the network to enable the data to be more and more vivid, so that the trained generator and the discriminator, namely the trained generation countermeasure network, are obtained, and the trained generation countermeasure network is used as a preset body motion artifact signal reconstruction model.
In an alternative embodiment, referring to fig. 4, the step S1 includes steps S101 to S104, which are as follows:
s101: acquiring vital sign signal sample data of a user all night;
s102: and performing down-sampling on the sample data of the vital sign signals, removing the down-sampled vital sign signals, and filtering body movement artifact signals, power frequency interference and high-frequency noise to obtain normal signal sample data.
In the embodiment of the present application, the vital sign signal sample data is down-sampled, specifically, the vital sign signal sample data is sampled at intervals, for example, every 10 signal points, and since the sampling rate of the vital sign signal sample data is 1000Hz, the vital sign signal sample data with the sampling rate of 100Hz after down-sampling is obtained.
Because the body motion artifact, the power frequency interference, the high-frequency noise and other interferences exist in the vital sign signal sample data, after the body motion artifact is manually eliminated, the power frequency interference and the Gaussian additive noise are removed through a six-order Butterworth low-pass filter with the cut-off frequency of 10Hz, and the normal signal sample data is obtained.
S103: and according to a second preset signal length, performing sliding window segmentation on the normal signal sample data to obtain the segmented normal signal sample data.
In the embodiment of the present application, the preset length of the second signal is 6s, the normal signal sample data is sequentially stepped by 1s, and the sliding window is divided into 6s segments (including 600 signal points).
S104: and carrying out Z-Score standardization operation on the segmented normal signal sample data to obtain a plurality of normal signal sample data with second preset signal length.
In the examples of the present application, several normal signal sample fragments were obtained by performing a Z-Score normalization operation of removing the mean and dividing by the standard deviation for each 6s fragment. Wherein, the calculation formula of the Z-Score standardization operation is as follows:
wherein Z is the signal amplitude of each signal point after the Z-Score normalization operation, x is the signal amplitude of each signal point, μ is the average of the signal amplitudes of all signal points within the 6s segment, and σ is the standard deviation of the signal amplitudes of all signal points within the 6s segment.
Through the Z-Score standardization operation, the signal amplitude of each signal point fluctuates around the 0 range, and the fluctuation range is uniform, so that the difference of the amplitude range of the ballistocardiographic signals among different users can be reduced, and the amplitude of each signal segment is controlled to be changed in the same range as far as possible.
In an alternative embodiment, referring to fig. 5, the step S30 includes steps S301 to S312, which are as follows:
s301: and dividing the body motion artifact signal into a plurality of body motion artifact signal units with continuous time sequence and third preset signal length.
Wherein the third preset signal length is 1s. If the body motion artifact signal is 1s, the body motion artifact signal is a body motion artifact signal unit. And if the body movement artifact signal is more than 1s, dividing the body movement artifact signal into a plurality of 1s body movement artifact signal units. For example, if the body motion artifact signal is 2s, the body motion artifact signal is divided into two 1s body motion artifact signal units, and if the body motion artifact signal is 2.5s, the body motion artifact signal is divided into three 1s body motion artifact signal units. In the embodiment of the present application, the body motion artifact signal is 3s, and the body motion artifact signal is divided into three 1s body motion artifact signal units.
S302: selecting a third normal signal from the first normal signals; and the third normal signal is a normal signal with a first preset signal length which is continuous with the body motion artifact signal in time sequence.
In the embodiment of the present application, the first predetermined signal length is 5s, and the last 5s signal segment is selected from the 6s first normal signal as the third normal signal.
S303: and inputting the third normal signal into a preset body motion artifact signal reconstruction model to obtain a first reconstruction signal of a first body motion artifact signal unit.
In the embodiment of the application, the third normal signal of 5s is input to the preset body motion artifact signal reconstruction model, and the 1 st signal segment of the body motion artifact signal is reconstructed, that is, the first reconstructed signal of the first body motion artifact signal unit is obtained.
S304: selecting a fourth normal signal from the third normal signals, splicing the fourth normal signal with the first reconstruction signal of the first body motion artifact signal unit, and inputting the fourth normal signal into a preset body motion artifact signal reconstruction model to obtain a first reconstruction signal of a second body motion artifact signal unit; the fourth normal signal is a normal signal which is continuous with the first body motion artifact signal unit in time sequence, and the signal length of the fourth normal signal is the difference between the first preset signal length and the third preset signal length.
In the embodiment of the present application, the last 4s signal segment is selected from the 5s third normal signal as the fourth normal signal. Splicing the fourth normal signal of 4s and the first reconstruction signal of the first body motion artifact signal unit of 1s to obtain a signal segment of 5s, inputting the signal segment of 5s into a preset body motion artifact signal reconstruction model, and reconstructing the signal segment of 2s of the body motion artifact signal, namely obtaining the first reconstruction signal of the second body motion artifact signal unit.
S305: selecting a fifth normal signal from the fourth normal signals, splicing the fifth normal signal with the first reconstruction signal of the first body movement artifact signal unit and the first reconstruction signal of the second body movement artifact signal unit, and inputting the spliced fifth normal signal into a preset body movement artifact signal reconstruction model to obtain a first reconstruction signal of the next body movement artifact signal unit; the fifth normal signal is a normal signal which is continuous with the first body motion artifact signal unit in time sequence, and the signal length of the fifth normal signal is the difference between the first preset signal length and twice the third preset signal length;
s306: until the first reconstructed signal of all body motion artifact signal units is determined.
Similarly, the last 3s signal segment is selected from the fourth normal signal of 4s as the fifth normal signal. Splicing the fifth normal signal of 3s with the first reconstruction signal of the first body motion artifact signal unit of 1s and the first reconstruction signal of the second body motion artifact signal unit of 1s to obtain a signal segment of 5s, inputting the signal segment of 5s into a preset body motion artifact signal reconstruction model, and reconstructing the signal segment of 3s of the body motion artifact signal, namely, obtaining the first reconstruction signal of the third body motion artifact signal unit.
S307: selecting a sixth normal signal from the second normal signals; and the sixth normal signal is a normal signal which is continuous with the body motion artifact signal in time sequence and has a first preset signal length.
In the embodiment of the present application, the first 5s signal segment is selected from the 7s second normal signal as the sixth normal signal.
S308: and inputting the sixth normal signal into a preset body motion artifact signal reconstruction model to obtain a second reconstruction signal of a penultimate body motion artifact signal unit.
In the embodiment of the application, a sixth normal signal of 5s is input into a preset body motion artifact signal reconstruction model, a last 1s signal segment of a body motion artifact signal is reconstructed, and a second reconstruction signal of a last body motion artifact signal unit is obtained.
S309: selecting a seventh normal signal from the sixth normal signal, splicing the seventh normal signal and a second reconstruction signal of the last-but-one body motion artifact signal unit, and inputting the signals into a preset body motion artifact signal reconstruction model to obtain a second reconstruction signal of the last-but-one body motion artifact signal unit; the seventh normal signal is a normal signal which is continuous with the last-but-one body motion artifact signal unit in time sequence, and the signal length of the seventh normal signal is the difference between the first preset signal length and the third preset signal length.
In the embodiment of the present application, the first 4s signal segment is selected from the sixth normal signal of 5s as the seventh normal signal. And splicing the seventh normal signal of 4s and the second reconstruction signal of the last-but-one body motion artifact signal unit of 1s to obtain a signal segment of 5s, inputting the signal segment of 5s into a preset body motion artifact signal reconstruction model, and obtaining the second reconstruction signal of the last-but-one body motion artifact signal unit.
S310: selecting an eighth normal signal from the seventh normal signal, splicing the eighth normal signal with a second reconstruction signal of the last-but-one body motion artifact signal unit and a second reconstruction signal of the last-but-one body motion artifact signal unit, and inputting the spliced eighth normal signal into a preset body motion artifact signal reconstruction model to obtain a second reconstruction signal of the last body motion artifact signal unit; the eighth normal signal is a normal signal which is continuous with the last-but-one body motion artifact signal unit in time sequence, and the signal length of the eighth normal signal is the difference between the first preset signal length and twice the third preset signal length;
s311: until the second reconstructed signal of all body motion artifact signal elements is determined.
Similarly, the first 3s signal segment is selected from the seventh normal signal of 4s as the eighth normal signal. Splicing the eighth normal signal of 3s with the second reconstruction signal of the last-but-one body motion artifact signal unit of 1s and the second reconstruction signal of the last-but-one body motion artifact signal unit of 1s to obtain a 5s signal segment, inputting the 5s signal segment into a preset body motion artifact signal reconstruction model, and obtaining the first reconstruction signal of the last-but-one body motion artifact signal unit.
S312: averaging the first reconstruction signals and the second reconstruction signals of all the body motion artifact signal units, and taking an average result as a reconstruction signal corresponding to the body motion artifact signal.
In the embodiment of the application, after the first reconstruction signal and the second reconstruction signal of three body motion artifact signal units are obtained, the first reconstruction signal and the second reconstruction signal of each body motion artifact signal unit are averaged, so that the reconstruction signal corresponding to the body motion artifact signal is obtained.
In an alternative embodiment, the preset body motion artifact signal reconstruction model includes a generator, the generator includes a gating cycle unit and a full convolution neural network, please refer to fig. 6, the step S303 includes steps S3031 to S3032, which are as follows:
s3031: inputting the third normal signal into the gating circulating unit to obtain a time sequence characteristic vector;
a Gated Recurrent Unit (GRU) is a modified form of a common Recurrent Neural Network (RNN), and is a cyclic unit based on a gate, which includes an update gate and a reset gate, where the update gate is used to control the previous memory information to be able to continuously keep the data amount at the current time, or determine how much information of the previous time step and the current time step are to be continuously transmitted to the future. The reset gate is used for determining how much past information is forgotten, controlling the data volume of the current information and the memory information, and generating new memory information to be continuously transmitted.
Inputting the third normal signal of 5s into the full convolution neural network for morphological feature extraction, and obtaining a time sequence feature vector corresponding to the third normal signal, wherein the specific formula is as follows:
z t ＝σ(W z ·[h t-1 ，x t ])
r t ＝σ(W r ·[h t-1 ，x t ])
wherein z is t Is the output of the current time step t of the update gate, r t Is the output of the reset gate at the current time step t, x t Is a one-dimensional input vector h corresponding to the vital sign signal segment input at the current time step t t-1 Is the timing feature vector of the output of the previous time step t-1,
is an intermediate result of the current time step t output, h t Is output at the current time step tTime-series feature vector, W z Is to update the weight parameter, W, of the gate r Is the weight parameter of the reset gate, W is the weight parameter corresponding to the calculated intermediate result, σ is the sigmoid activation function, tanh is the tanh activation function, [,]is a vector stitching operation.
S3032: and inputting the third normal signal into the full convolution neural network to obtain a morphological feature vector.
Convolutional Neural Network (CNN) is commonly used for feature extraction of objects such as images and time series. Full Convolutional Networks (FCNs) are composed of multiple layers of Convolutional blocks, which may have different or the same Convolutional kernel size. Inputting the third normal signal of 5s into the full convolution neural network for morphological feature extraction to obtain a morphological feature vector corresponding to the third normal signal, wherein the specific formula is as follows:
y＝W′*x+b
z＝BN(y)
Out＝Relu(z)
wherein, x represents a one-dimensional input vector corresponding to the third normal signal, W' is a weight parameter of a one-dimensional convolution kernel, b is a bias parameter, y is an output vector of the convolution kernel, and z is an intermediate result after a Batch Normalization operation is applied to the convolution kernel, and then z is transmitted to a linear rectification function ReLU to calculate the output of the convolution kernel Out, so as to obtain a morphological feature vector.
S3033: and splicing the time sequence characteristic vector and the form characteristic vector, inputting a splicing result to a full-connection layer, and obtaining a first reconstruction signal corresponding to a first body motion artifact signal unit.
In the embodiment of the present application, the time sequence feature vector and the morphological feature vector are spliced, a splicing result is input to a full connection layer, and a 1s segment (100 data points) is output, that is, a first reconstruction signal corresponding to a first body motion artifact signal unit.
In this embodiment of the application, with respect to the processes of steps S304 to S306, and steps S308 to S311, refer to steps S3031 to S3033 for obtaining the first reconstruction signal and the second reconstruction signal corresponding to each body motion artifact signal unit, which are not described herein again.
Example 2
Referring to fig. 7, an embodiment of the present invention provides a vital sign signal reconstruction apparatus 4, which includes:
a signal obtaining module 41, configured to obtain a vital sign signal; the vital sign signals comprise at least one body motion artifact signal, a first normal signal and a second normal signal; the first normal signal is a continuous normal signal which is before the body motion artifact signal and is adjacent to the body motion artifact signal, and the second normal signal is a continuous normal signal which is after the body motion artifact signal and is adjacent to the body motion artifact signal;
a signal determining module 42, configured to determine whether the body motion artifact signal can be reconstructed according to the first normal signal and the second normal signal;
a signal reconstruction module 43, configured to input the first normal signal and the second normal signal into a preset body motion artifact signal reconstruction model if the body motion artifact signal can be reconstructed, and obtain a reconstructed signal corresponding to the body motion artifact signal;
and a signal replacement module 44, configured to replace the corresponding body motion artifact signal in the vital sign signal with the reconstructed signal, so as to obtain the reconstructed vital sign signal.
Optionally, the signal determining module includes:
a signal length acquiring unit configured to acquire a first signal length of the first normal signal and a second signal length of the second normal signal;
and the first signal reconstruction determining unit is used for determining that the body motion artifact signal can be reconstructed if the first signal length and the second signal length are both greater than or equal to a first preset signal length.
Optionally, the signal determining module includes:
a sample entropy calculation unit for calculating sample entropies, first order standard deviations, and second order standard deviations of the first normal signal and the second normal signal, respectively;
and a second signal reconstruction determining unit, configured to determine that the body motion artifact signal can be reconstructed if the sample entropies are all smaller than a preset first sample entropy threshold and are all larger than a preset second sample entropy threshold, the first-order standard deviations are all smaller than a preset first-order standard deviation threshold, and the second-order standard deviations are all smaller than a preset second-order standard deviation threshold.
Optionally, the signal reconstructing module includes:
the signal dividing unit is used for dividing the body motion artifact signals into a plurality of body motion artifact signal units with continuous time sequence and third preset signal length;
a first signal selection unit for selecting a third normal signal from the first normal signals; the third normal signal is a normal signal with a first preset signal length which is continuous with the body motion artifact signal in time sequence;
the first reconstruction signal obtaining unit is used for inputting the third normal signal into a preset body motion artifact signal reconstruction model to obtain a first reconstruction signal of a first body motion artifact signal unit;
the second signal selection unit is used for selecting a fourth normal signal from the third normal signals, splicing the fourth normal signal with the first reconstruction signal of the first body motion artifact signal unit, and inputting the spliced fourth normal signal and the first reconstruction signal of the first body motion artifact signal unit into a preset body motion artifact signal reconstruction model to obtain a first reconstruction signal of a second body motion artifact signal unit; the fourth normal signal is a normal signal which is continuous with the first body motion artifact signal unit in time sequence, and the signal length of the fourth normal signal is the difference between the first preset signal length and the third preset signal length;
a third signal selection unit, configured to select a fifth normal signal from the fourth normal signals, splice the fifth normal signal with the first reconstructed signal of the first body motion artifact signal unit and the first reconstructed signal of the second body motion artifact signal unit, and input the spliced fifth normal signal to a preset body motion artifact signal reconstruction model to obtain a first reconstructed signal of a next body motion artifact signal unit; the fifth normal signal is a normal signal which is continuous with the first body motion artifact signal unit in time sequence, and the signal length of the fifth normal signal is the difference between the first preset signal length and twice the third preset signal length;
a first reconstruction signal determination unit for determining the first reconstruction signals of all the body motion artifact signal units until;
a fourth signal selecting unit, configured to select a sixth normal signal from the second normal signals; the sixth normal signal is a normal signal with a first preset signal length which is continuous with the body motion artifact signal in time sequence;
the second reconstruction signal obtaining unit is used for inputting the sixth normal signal into a preset body motion artifact signal reconstruction model to obtain a second reconstruction signal of a penultimate body motion artifact signal unit;
the fifth signal selection unit is used for selecting a seventh normal signal from the sixth normal signals, splicing the seventh normal signal with the second reconstruction signal of the last-but-one body motion artifact signal unit, and inputting the spliced seventh normal signal and the second reconstruction signal of the last-but-one body motion artifact signal unit into a preset body motion artifact signal reconstruction model to obtain a second reconstruction signal of the last-but-one body motion artifact signal unit; the seventh normal signal is a normal signal which is continuous with the last-but-one body motion artifact signal unit in time sequence, and the signal length of the seventh normal signal is the difference between the first preset signal length and the third preset signal length;
a sixth signal selection unit, configured to select an eighth normal signal from the seventh normal signals, splice the eighth normal signal with the second reconstruction signal of the penultimate body motion artifact signal unit and the second reconstruction signal of the penultimate body motion artifact signal unit, and input the spliced eighth normal signal to a preset body motion artifact signal reconstruction model to obtain the second reconstruction signal of the last body motion artifact signal unit; the eighth normal signal is a normal signal which is continuous with the last-but-one body motion artifact signal unit in time sequence, and the signal length of the eighth normal signal is the difference between the first preset signal length and twice the third preset signal length;
a second reconstruction signal determination unit for determining a second reconstruction signal of all body motion artifact signal units up to;
and the signal reconstruction unit is used for averaging the first reconstruction signals and the second reconstruction signals of all the body motion artifact signal units, and taking an average result as a reconstruction signal corresponding to the body motion artifact signal.
Optionally, the first reconstructed signal obtaining unit includes:
a time sequence vector obtaining unit, configured to input the third normal signal to the gated loop unit, and obtain a time sequence feature vector;
a form vector obtaining unit, configured to input the third normal signal to the full convolution neural network, so as to obtain a form feature vector;
and the vector splicing unit is used for splicing the time sequence characteristic vector and the form characteristic vector, inputting a splicing result to the full-connection layer and obtaining a first reconstruction signal corresponding to the first body motion artifact signal unit.
By applying the embodiment of the application, the vital sign signals are obtained; the vital sign signals comprise at least one body movement artifact signal, a first normal signal and a second normal signal; the first normal signal is a continuous normal signal which is before the body motion artifact signal and is adjacent to the body motion artifact signal, and the second normal signal is a continuous normal signal which is after the body motion artifact signal and is adjacent to the body motion artifact signal; judging whether the body motion artifact signal can be reconstructed or not according to the first normal signal and the second normal signal; if the body motion artifact signal can be reconstructed, inputting the first normal signal and the second normal signal into a preset body motion artifact signal reconstruction model to obtain a reconstruction signal corresponding to the body motion artifact signal, and replacing the corresponding body motion artifact signal in the vital sign signal with the reconstruction signal to obtain a reconstructed vital sign signal. According to the method and the device, the time sequence characteristics and the morphological characteristics of the front and back normal signals of the body motion artifact are extracted through the preset body motion artifact signal reconstruction model, the front and back bidirectional reconstruction is carried out on the body motion artifact signals according to the front and back normal signals, the reconstruction signals corresponding to the body motion artifact signals are obtained, and the accuracy of vital sign signal reconstruction is improved.
Example 3
The following is an embodiment of the apparatus of the present application, which may be used to perform the method of embodiment 1 of the present application. For details which are not disclosed in the device example of the present application, reference is made to the content of the method in example 1 of the present application.
Referring to fig. 7, the present application further provides an electronic device 300, where the electronic device may be embodied as a computer, a mobile phone, a tablet computer, an interactive tablet, and the like, and in an exemplary embodiment of the present application, the electronic device 300 is an interactive tablet, and the interactive tablet may include: at least one processor 301, at least one memory 302, at least one display, at least one network interface 303, a user interface 304, and at least one communication bus 305.
The user interface 304 is mainly used for providing an input interface for a user to obtain data input by the user. Optionally, the user interface may also include a standard wired interface, a wireless interface.
The network interface 303 may optionally include a standard wired interface or a wireless interface (e.g., WI-FI interface).
Wherein a communication bus 305 is used to enable the connection communication between these components.
The Memory 302 may include a Random Access Memory (RAM) or a Read-Only Memory (Read-Only Memory). Optionally, the memory includes a non-transitory computer-readable medium. The memory may be used to store an instruction, a program, code, a set of codes, or a set of instructions. The memory may include a stored program area and a stored data area, wherein the stored program area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playing function, an image playing function, etc.), instructions for implementing the various method embodiments described above, and the like; the storage data area may store data and the like referred to in the above respective method embodiments. The memory may optionally be at least one memory device located remotely from the processor. The memory, which is a type of computer storage medium, may include an operating system, a network communication module, a user interface module, and an operating application.
The processor may be configured to invoke an application program of the video resolution adjustment method stored in the memory, and specifically execute the method steps in embodiment 1 shown above, and the specific execution process may refer to the specific description shown in embodiment 1, which is not described herein again.
Example 4
The present application further provides a computer-readable storage medium, on which a computer program is stored, where the instructions are suitable for being loaded by a processor and executing the method steps of embodiment 1, and specific execution processes may refer to specific descriptions shown in the embodiments and are not described herein again. The device where the storage medium is located can be an electronic device such as a personal computer, a notebook computer, a smart phone and a tablet computer.
For the apparatus embodiment, since it substantially corresponds to the method embodiment, reference may be made to the partial description of the method embodiment for relevant points. The above-described device embodiments are merely illustrative, and components illustrated as separate components may or may not be physically separate, and components shown as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the scheme of the application. One of ordinary skill in the art can understand and implement it without inventive effort.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
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.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
In a typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.
The memory may include forms of volatile memory in a computer readable medium, random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). The memory is an example of a computer-readable medium.
Computer-readable media, including both permanent and non-permanent, removable and non-removable media, may implement the information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Disks (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium, which can be used to store information that can be accessed by a computing device. As defined herein, a computer readable medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.
It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrases "comprising a," "8230," "8230," or "comprising" does not exclude the presence of additional identical elements in the process, method, article, or apparatus comprising the element.
The above are merely examples of the present application and are not intended to limit the present application. Various modifications and changes may occur to those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.
Claims (10)
1. A method for reconstructing a vital sign signal, the method comprising the steps of:
acquiring a vital sign signal; the vital sign signals comprise at least one body motion artifact signal, a first normal signal and a second normal signal; the first normal signal is a continuous normal signal which is before the body motion artifact signal and is adjacent to the body motion artifact signal, and the second normal signal is a continuous normal signal which is after the body motion artifact signal and is adjacent to the body motion artifact signal;
judging whether the body motion artifact signal can be reconstructed or not according to the first normal signal and the second normal signal;
if the body motion artifact signal can be reconstructed, inputting the first normal signal and the second normal signal into a preset body motion artifact signal reconstruction model to obtain a reconstruction signal corresponding to the body motion artifact signal;
and replacing the corresponding body motion artifact signals in the vital sign signals with the reconstructed signals to obtain the reconstructed vital sign signals.
2. Method for reconstruction of vital sign signals according to claim 1, characterized in that:
the step of judging whether the body motion artifact signal can be reconstructed according to the first normal signal and the second normal signal comprises the following steps:
acquiring a first signal length of the first normal signal and a second signal length of the second normal signal;
and if the first signal length and the second signal length are both greater than or equal to a first preset signal length, determining that the body motion artifact signal can be reconstructed.
3. Method for reconstruction of vital sign signals according to claim 1, characterized in that:
the step of judging whether the body motion artifact signal can be reconstructed according to the first normal signal and the second normal signal comprises the following steps:
respectively calculating sample entropies, first-order standard deviations and second-order standard deviations of the first normal signal and the second normal signal;
and if the sample entropies are all smaller than a preset first sample entropy threshold value and the sample entropies are all larger than a preset second sample entropy threshold value, the first-order standard deviations are all smaller than a preset first-order standard deviation threshold value, and the second-order standard deviations are all smaller than a preset second-order standard deviation threshold value, determining that the body motion artifact signal can be reconstructed.
4. Method for reconstruction of vital sign signals according to claim 1, characterized in that:
if the body motion artifact signal can be reconstructed, inputting the first normal signal and the second normal signal into a preset body motion artifact signal reconstruction model to obtain a reconstruction signal corresponding to the body motion artifact signal, including:
dividing the body movement artifact signal into a plurality of body movement artifact signal units with continuous time sequence and third preset signal length;
selecting a third normal signal from the first normal signals; the third normal signal is a normal signal with a first preset signal length which is continuous with the body motion artifact signal in time sequence;
inputting the third normal signal into a preset body motion artifact signal reconstruction model to obtain a first reconstruction signal of a first body motion artifact signal unit;
selecting a fourth normal signal from the third normal signals, splicing the fourth normal signal with the first reconstruction signal of the first body motion artifact signal unit, and inputting the fourth normal signal into a preset body motion artifact signal reconstruction model to obtain a first reconstruction signal of a second body motion artifact signal unit; the fourth normal signal is a normal signal which is continuous with the first body motion artifact signal unit in time sequence, and the signal length of the fourth normal signal is the difference between the first preset signal length and the third preset signal length;
selecting a fifth normal signal from the fourth normal signals, splicing the fifth normal signal with the first reconstruction signal of the first body motion artifact signal unit and the first reconstruction signal of the second body motion artifact signal unit, and inputting the spliced fifth normal signal into a preset body motion artifact signal reconstruction model to obtain a first reconstruction signal of the next body motion artifact signal unit; the fifth normal signal is a normal signal which is continuous with the first body motion artifact signal unit in time sequence, and the signal length of the fifth normal signal is the difference between the first preset signal length and twice the third preset signal length;
determining first reconstruction signals of all body motion artifact signal units;
selecting a sixth normal signal from the second normal signals; the sixth normal signal is a normal signal with a first preset signal length which is continuous with the body motion artifact signal in time sequence;
inputting the sixth normal signal into a preset body motion artifact signal reconstruction model to obtain a second reconstruction signal of a penultimate body motion artifact signal unit;
selecting a seventh normal signal from the sixth normal signals, splicing the seventh normal signal with a second reconstruction signal of the penultimate body motion artifact signal unit, and inputting the spliced seventh normal signal and the second reconstruction signal of the penultimate body motion artifact signal unit into a preset body motion artifact signal reconstruction model to obtain a second reconstruction signal of the penultimate body motion artifact signal unit; the seventh normal signal is a normal signal which is continuous with the last-but-one body motion artifact signal unit in time sequence, and the signal length of the seventh normal signal is the difference between the length of the first preset signal and the length of the third preset signal;
selecting an eighth normal signal from the seventh normal signal, splicing the eighth normal signal with the second reconstruction signal of the penultimate body motion artifact signal unit and the second reconstruction signal of the penultimate body motion artifact signal unit, and inputting the spliced eighth normal signal into a preset body motion artifact signal reconstruction model to obtain a second reconstruction signal of the last body motion artifact signal unit; the eighth normal signal is a normal signal which is continuous with the penultimate individual motion artifact signal unit in time sequence, and the signal length of the eighth normal signal is the difference between the length of the first preset signal and twice the length of the third preset signal;
until determining second reconstruction signals of all body motion artifact signal units;
averaging the first reconstruction signals and the second reconstruction signals of all the body motion artifact signal units, and taking an average result as a reconstruction signal corresponding to the body motion artifact signal.
5. Method for reconstruction of vital sign signals according to claim 4, characterized in that: the preset body motion artifact signal reconstruction model comprises a generator, wherein the generator comprises a gating cycle unit and a full convolution neural network;
the step of inputting the third normal signal into a preset body motion artifact signal reconstruction model to obtain a first reconstruction signal of a first body motion artifact signal unit includes:
inputting the third normal signal into the gating circulating unit to obtain a time sequence characteristic vector;
inputting the third normal signal into the full convolution neural network to obtain a morphological feature vector;
and splicing the time sequence characteristic vector and the form characteristic vector, and inputting a splicing result into a full-connection layer to obtain a first reconstruction signal of a first body motion artifact signal unit.
6. Method for reconstruction of vital sign signals according to any one of claims 1 to 5, wherein:
if the body motion artifact signal can be reconstructed, inputting the first normal signal and the second normal signal into a preset body motion artifact signal reconstruction model, and obtaining a reconstruction signal corresponding to the body motion artifact signal, including:
acquiring normal signal sample data of a plurality of second preset signal lengths;
dividing the normal signal sample data into a first sample data and a second sample data which are continuous in time sequence; the signal length of the first sample data is a first preset signal length, and the signal length of the second sample data is the difference between the second preset signal length and the first preset signal length;
the first sample data is used as the input of a generator in a generated confrontation network to obtain a first prediction result of the generator, and the first prediction result and the second sample data are both used as the input of a discriminator in the generated confrontation network to obtain a first discrimination result of the discriminator;
dividing the normal signal sample data into third sample data and fourth sample data which are continuous in time sequence; the signal length of the third sample data is a third preset signal length, and the signal length of the fourth sample data is the difference between the second preset signal length and the third preset signal length;
taking the fourth sample data as the input of the generator, obtaining a second prediction result of the generator, taking the second prediction result and the third sample data as the input of the discriminator, and obtaining a second discrimination result of the discriminator;
continuously training the generator and the discriminator until the first discrimination result and the second discrimination result reach a preset threshold value, and obtaining a trained generated countermeasure network;
and taking the trained generated confrontation network as a preset body motion artifact signal reconstruction model.
7. Method for reconstruction of vital sign signals according to claim 6, characterized in that:
the step of obtaining the normal signal sample data of a plurality of second preset signal lengths includes:
acquiring vital sign signal sample data of a user all night;
down-sampling the vital sign signal sample data, removing the down-sampled vital sign signal, and filtering a body movement artifact signal, power frequency interference and high-frequency noise to obtain normal signal sample data;
according to a second preset signal length, performing sliding window segmentation on the normal signal sample data to obtain segmented normal signal sample data;
and performing Z-Score standardization operation on the segmented normal signal sample data to obtain a plurality of normal signal sample data with second preset signal length.
8. A vital sign signal reconstruction device, comprising:
the signal acquisition module is used for acquiring vital sign signals; the vital sign signals comprise at least one body motion artifact signal, a first normal signal and a second normal signal; the first normal signal is a continuous normal signal which is before the body motion artifact signal and is adjacent to the body motion artifact signal, and the second normal signal is a continuous normal signal which is after the body motion artifact signal and is adjacent to the body motion artifact signal;
the signal judgment module is used for judging whether the body motion artifact signal can be reconstructed or not according to the first normal signal and the second normal signal;
the signal reconstruction module is used for inputting the first normal signal and the second normal signal into a preset body motion artifact signal reconstruction model if the body motion artifact signal can be reconstructed, and acquiring a reconstruction signal corresponding to the body motion artifact signal;
and the signal replacement module is used for replacing the corresponding body motion artifact signals in the vital sign signals with the reconstructed signals to obtain the reconstructed vital sign signals.
9. A computer device, comprising: processor, memory and computer program stored in the memory and executable on the processor, characterized in that the steps of the method according to any of claims 1 to 7 are implemented when the processor executes the computer program.
10. A computer-readable storage medium, in which a computer program is stored which, when being executed by a processor, carries out the steps of the method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211016863.9A CN115381438B (en) | 2022-08-24 | 2022-08-24 | Method and device for reconstructing vital sign signals, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211016863.9A CN115381438B (en) | 2022-08-24 | 2022-08-24 | Method and device for reconstructing vital sign signals, computer equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115381438A true CN115381438A (en) | 2022-11-25 |
| CN115381438B CN115381438B (en) | 2023-04-18 |
Family
ID=84121567
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211016863.9A Active CN115381438B (en) | 2022-08-24 | 2022-08-24 | Method and device for reconstructing vital sign signals, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115381438B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117438024A (en) * | 2023-12-15 | 2024-01-23 | 吉林大学 | Intelligent acquisition and analysis system and method for acute diagnosis patient sign data |
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110077484A1 (en) * | 2009-09-30 | 2011-03-31 | Nellcor Puritan Bennett Ireland | Systems And Methods For Identifying Non-Corrupted Signal Segments For Use In Determining Physiological Parameters |
| CN109645979A (en) * | 2017-10-10 | 2019-04-19 | 深圳市理邦精密仪器股份有限公司 | Ambulatory ecg signal artifact identification method and device |
| CN109816095A (en) * | 2019-01-14 | 2019-05-28 | 湖南大学 | Based on the network flow prediction method for improving gating cycle neural network |
| CN111046824A (en) * | 2019-12-19 | 2020-04-21 | 上海交通大学 | Time series signal efficient denoising and high-precision reconstruction modeling method and system |
| CN111035367A (en) * | 2019-12-31 | 2020-04-21 | 华南师范大学 | Signal detection method and system for judging sleep apnea |
| CN111401553A (en) * | 2020-03-12 | 2020-07-10 | 南京航空航天大学 | Missing data filling method and system based on neural network |
| CN111984637A (en) * | 2020-07-06 | 2020-11-24 | 苏州研数信息科技有限公司 | Missing value processing method and device in data modeling, equipment and storage medium |
| US20200405269A1 (en) * | 2018-02-27 | 2020-12-31 | Koninklijke Philips N.V. | Ultrasound system with a neural network for producing images from undersampled ultrasound data |
| CN112836180A (en) * | 2021-04-19 | 2021-05-25 | 北京瑞莱智慧科技有限公司 | Sensing data determination method and device, electronic equipment and readable storage medium |
| CN112884237A (en) * | 2021-03-11 | 2021-06-01 | 山东科技大学 | Power distribution network prediction auxiliary state estimation method and system |
| CN114004258A (en) * | 2021-11-03 | 2022-02-01 | 大连大学 | Semi-supervised electrocardio abnormality detection method |
| CN114041786A (en) * | 2022-01-11 | 2022-02-15 | 华南师范大学 | Ballistocardiogram signal detection method, ballistocardiogram signal detection device and ballistocardiogram signal detection equipment |
| CN114680879A (en) * | 2022-06-01 | 2022-07-01 | 华南师范大学 | Method and device for detecting body motion artifact signal |
-
2022
- 2022-08-24 CN CN202211016863.9A patent/CN115381438B/en active Active
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110077484A1 (en) * | 2009-09-30 | 2011-03-31 | Nellcor Puritan Bennett Ireland | Systems And Methods For Identifying Non-Corrupted Signal Segments For Use In Determining Physiological Parameters |
| CN109645979A (en) * | 2017-10-10 | 2019-04-19 | 深圳市理邦精密仪器股份有限公司 | Ambulatory ecg signal artifact identification method and device |
| US20200405269A1 (en) * | 2018-02-27 | 2020-12-31 | Koninklijke Philips N.V. | Ultrasound system with a neural network for producing images from undersampled ultrasound data |
| CN109816095A (en) * | 2019-01-14 | 2019-05-28 | 湖南大学 | Based on the network flow prediction method for improving gating cycle neural network |
| CN111046824A (en) * | 2019-12-19 | 2020-04-21 | 上海交通大学 | Time series signal efficient denoising and high-precision reconstruction modeling method and system |
| CN111035367A (en) * | 2019-12-31 | 2020-04-21 | 华南师范大学 | Signal detection method and system for judging sleep apnea |
| CN111401553A (en) * | 2020-03-12 | 2020-07-10 | 南京航空航天大学 | Missing data filling method and system based on neural network |
| CN111984637A (en) * | 2020-07-06 | 2020-11-24 | 苏州研数信息科技有限公司 | Missing value processing method and device in data modeling, equipment and storage medium |
| CN112884237A (en) * | 2021-03-11 | 2021-06-01 | 山东科技大学 | Power distribution network prediction auxiliary state estimation method and system |
| CN112836180A (en) * | 2021-04-19 | 2021-05-25 | 北京瑞莱智慧科技有限公司 | Sensing data determination method and device, electronic equipment and readable storage medium |
| CN114004258A (en) * | 2021-11-03 | 2022-02-01 | 大连大学 | Semi-supervised electrocardio abnormality detection method |
| CN114041786A (en) * | 2022-01-11 | 2022-02-15 | 华南师范大学 | Ballistocardiogram signal detection method, ballistocardiogram signal detection device and ballistocardiogram signal detection equipment |
| CN114680879A (en) * | 2022-06-01 | 2022-07-01 | 华南师范大学 | Method and device for detecting body motion artifact signal |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117438024A (en) * | 2023-12-15 | 2024-01-23 | 吉林大学 | Intelligent acquisition and analysis system and method for acute diagnosis patient sign data |
| CN117438024B (en) * | 2023-12-15 | 2024-03-08 | 吉林大学 | Intelligent acquisition and analysis system and method for acute diagnosis patient sign data |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115381438B (en) | 2023-04-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20230131876A1 (en) | Systems and methods of identity analysis of electrocardiograms | |
| CN109247923B (en) | Non-contact type pulse real-time estimation method and device based on video | |
| US20120071780A1 (en) | Method and a System for Classifying Neural Signals, and a Method of Selecting Electrodes for Direct Neural Control | |
| JP2022163153A (en) | Method for determining focus of visual attention of subject from electroencephalographic signals, computer program, computer system, and system for determining focus of visual attention of subject from electroencephalographic signals | |
| CN108363487B (en) | Construction method of dream reproduction model, and dream reproduction method and device | |
| TW201446216A (en) | Optical heartrate tracking | |
| CN115381438B (en) | Method and device for reconstructing vital sign signals, computer equipment and storage medium | |
| US20210076953A1 (en) | Methods and systems for pulse transit time determination | |
| CN110532999B (en) | Smoking behavior identification method, device and equipment | |
| CN113554726B (en) | Image reconstruction method and device based on pulse array, storage medium and terminal | |
| CN112313705A (en) | Analysis and visualization of subtle motion in video | |
| Xu et al. | Classification of motor imagery tasks for electrocorticogram based brain-computer interface | |
| JP7339676B2 (en) | Computer-implemented method and system for direct photoplethysmography (PPG) with multiple sensors | |
| CN114203300A (en) | Health state evaluation method and system, server and storage medium | |
| CN102204820A (en) | Recording apparatus and recording method thereof | |
| EP2769668B1 (en) | System for the adaptive diagnosis of chronic heart failure using classifying means and a boolean decision tree | |
| CN110008881B (en) | Method and device for identifying cow behaviors of multiple moving targets | |
| CN114680879B (en) | Method and device for detecting body motion artifact signal | |
| WO2023112384A1 (en) | Computer system and emotion estimation method | |
| KR101413853B1 (en) | Method and apparatus for measuring physiological signal usuing infrared image | |
| CN115486833A (en) | Respiration state detection method and device, computer equipment and storage medium | |
| EP4140392A1 (en) | Noise removal in physiological signals | |
| CN113850160A (en) | Method and device for counting repeated actions | |
| KR102257295B1 (en) | Diagnostic system for diagnosing vulnerable atheromatous plaque through ultrasound image-based machine learning and the diagnostic method thereof | |
| KR20120072291A (en) | Method for imputating of missing values using incremental expectation maximization principal component analysis |
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 | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20240221 Address after: Building 2, No. 38 Junfeng Road, Huangpu District, Guangzhou City, Guangdong Province, 510000 Patentee after: GUANGDONG JUNFENG BFS INDUSTRY CO.,LTD. Country or region after: China Address before: School of physics and telecommunication engineering, South China Normal University, No. 378, Waihuan West Road, Panyu District, Guangzhou City, Guangdong Province, 510006 Patentee before: SOUTH CHINA NORMAL University Country or region before: China |