CN112587149A - Atrial premature beat target detection device based on convolutional neural network - Google Patents
Atrial premature beat target detection device based on convolutional neural network Download PDFInfo
- Publication number
- CN112587149A CN112587149A CN202011252291.5A CN202011252291A CN112587149A CN 112587149 A CN112587149 A CN 112587149A CN 202011252291 A CN202011252291 A CN 202011252291A CN 112587149 A CN112587149 A CN 112587149A
- Authority
- CN
- China
- Prior art keywords
- layer
- sampling
- output
- convolution
- neural network
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 230000001746 atrial effect Effects 0.000 title claims abstract description 39
- 238000013527 convolutional neural network Methods 0.000 title claims abstract description 38
- 238000001514 detection method Methods 0.000 title claims abstract description 32
- 208000000418 Premature Cardiac Complexes Diseases 0.000 title claims abstract description 31
- 238000005070 sampling Methods 0.000 claims abstract description 70
- 238000007781 pre-processing Methods 0.000 claims abstract description 11
- 230000002159 abnormal effect Effects 0.000 claims abstract description 10
- 238000004590 computer program Methods 0.000 claims description 14
- 238000000034 method Methods 0.000 claims description 10
- 230000002028 premature Effects 0.000 claims description 10
- 238000001914 filtration Methods 0.000 claims description 9
- 206010015856 Extrasystoles Diseases 0.000 claims description 8
- 230000000747 cardiac effect Effects 0.000 claims description 7
- 230000008569 process Effects 0.000 claims description 6
- 238000002372 labelling Methods 0.000 claims description 5
- 238000013528 artificial neural network Methods 0.000 claims description 4
- 238000013480 data collection Methods 0.000 claims 1
- 230000006870 function Effects 0.000 description 13
- 238000010586 diagram Methods 0.000 description 10
- 206010003119 arrhythmia Diseases 0.000 description 9
- 230000006793 arrhythmia Effects 0.000 description 9
- 238000004422 calculation algorithm Methods 0.000 description 9
- 230000000875 corresponding effect Effects 0.000 description 3
- 238000003745 diagnosis Methods 0.000 description 3
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 3
- 230000005284 excitation Effects 0.000 description 3
- 230000008602 contraction Effects 0.000 description 2
- 238000011176 pooling Methods 0.000 description 2
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 description 1
- 208000002102 Atrial Premature Complexes Diseases 0.000 description 1
- 230000005856 abnormality Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 230000008034 disappearance Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 210000002837 heart atrium Anatomy 0.000 description 1
- 238000012886 linear function Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000000284 resting effect Effects 0.000 description 1
- 230000007704 transition Effects 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
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
Landscapes
- Health & Medical Sciences (AREA)
- Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Physics & Mathematics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physiology (AREA)
- Psychiatry (AREA)
- Signal Processing (AREA)
- Fuzzy Systems (AREA)
- Evolutionary Computation (AREA)
- Pathology (AREA)
- Mathematical Physics (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Molecular Biology (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Veterinary Medicine (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The application relates to an atrial premature beat target detection device based on a convolutional neural network, which is characterized in that each collected electrocardiosignal is labeled with a data label through data preprocessing, the data labels comprise heart beat types and heart beat positions at different sampling points of the electrocardiosignals, a training set is formed, a convolutional neural network model is trained, the real-time electrocardio data to be detected is input into the trained convolutional neural network model, the type of abnormal heart beat and the position of the abnormal heart beat are predicted, and therefore target detection of the atrial premature beat is achieved.
Description
Technical Field
The application belongs to the technical field of electrocardiographic detection, and particularly relates to an atrial premature beat target detection device based on a convolutional neural network.
Background
Various signs caused by abnormalities in the cardiac electrical conduction system are called arrhythmias. Atrial premature beat is common in clinical arrhythmia diseases, and refers to abnormal contraction of ventricles or atria caused by the premature beat of the heart, and various complications can be caused once the premature contraction occurs frequently, so that the target detection of atrial premature arrhythmia is of great significance.
The electrocardiosignal is an important tool for detecting arrhythmia diseases and is an electric signal generated when the heart of a human body moves. Diagnosis of arrhythmia by a physician via an ecg signal is often time consuming, labor intensive and highly correlated to the physician level. With the wide application of artificial intelligence technology in arrhythmia detection, doctors are assisted in diagnosis, and diagnosis efficiency is greatly improved, but most algorithms can only provide type identification based on heart beats or segments, and cannot simultaneously judge arrhythmia types and provide position information.
Disclosure of Invention
The technical problem to be solved by the invention is as follows: in order to solve the problem that the algorithm in the prior art cannot judge the arrhythmia types and give position information at the same time for arrhythmia diseases.
In order to solve the technical problems, the invention provides an atrial premature beat target detection device based on a convolutional neural network, the convolutional neural network adopts the convolutional neural network which is improved based on a discrete wavelet thought and has a symmetrical frame, the acquired electrocardiosignal is marked with a data label containing the heart beat type and the heart beat position at different sampling points of the electrocardiosignal, and the type and the position of the abnormal heart beat are predicted by taking real-time electrocardio data to be detected as input through training a convolutional neural network model, so that the target detection of the atrial premature beat is realized.
The technical scheme adopted by the invention for solving the technical problems is as follows:
the first aspect of the present invention provides an atrial premature beat target detection device based on a convolutional neural network, comprising:
the data acquisition module is used for collecting a plurality of electrocardiosignals which contain atrial premature beats and are marked with R waves;
the preprocessing module is used for labeling each electrocardiosignal with a data label through data preprocessing, and forming a training set by the electrocardiosignal data labeled with the data label, wherein the data label comprises the heart beat type and the heart beat position at different sampling points of the electrocardiosignal;
the neural network training module is used for taking the electrocardiosignal data in the training set as input, taking the heart beat types and positions at different sampling points of the electrocardiosignals as output, and training a convolutional neural network model;
and the target detection module is used for acquiring real-time electrocardio data to be detected, inputting the real-time electrocardio data to be detected into the trained convolutional neural network model, and predicting the type and the position of the abnormal heart beat so as to realize the target detection of the atrial premature beat.
A second aspect of the present invention provides a computer storage medium having a computer program stored thereon, the computer program, when executed by a processor, is configured to perform the processes of the detection apparatus of the first aspect of the present invention.
The invention has the beneficial effects that: according to the method, the acquired electrocardiosignal labels comprise data labels of the heart beat types and the heart beat positions at different sampling points of the electrocardiosignals, the electrocardiosignal data to be detected in real time are used as input through training a convolutional neural network model, the type of abnormal heart beat and the position of the abnormal heart beat can be predicted at the same time, and the target detection of atrial premature beat is realized.
Drawings
The technical solution of the present application is further explained below with reference to the drawings and the embodiments.
FIG. 1 is a schematic diagram of a detection device according to an embodiment of the present invention;
FIG. 2 is a block diagram of a convolutional neural network architecture of an embodiment of the present invention.
Detailed Description
It should be noted that the embodiments and features of the embodiments in the present application may be combined with each other without conflict.
The technical solutions of the present application will be described in detail below with reference to the accompanying drawings in conjunction with embodiments.
Example 1
The embodiment provides an atrial premature target detection device based on a convolutional neural network, as shown in fig. 1, including:
the data acquisition module is used for collecting a plurality of electrocardiosignals which contain atrial premature beats and are marked with R wave crests;
the preprocessing module is used for labeling each electrocardiosignal with a data label through data preprocessing, and forming a training set by the electrocardiosignal data labeled with the data label, wherein the data label comprises the heart beat type and the heart beat position at different sampling points of the electrocardiosignal;
the neural network training module is used for taking the electrocardiosignal data in the training set as input, taking the heart beat types and positions at different sampling points of the electrocardiosignals as output, and training a convolutional neural network model;
and the target detection module is used for acquiring real-time electrocardio data to be detected, inputting the real-time electrocardio data to be detected into the trained convolutional neural network model, and predicting the type and the position of the abnormal heart beat so as to realize the target detection of the atrial premature beat.
In this embodiment, 15000 clinical resting 10s twelve-lead electrocardiosignals marked with R waves are collected, and the sampling frequency of the electrocardiosignals is 500 Hz.
The 10s cardiac signal collected in this embodiment includes: comprises the electrocardiosignals of the left bundle branch and the atrial premature beat, comprises the electrocardiosignals of the right bundle branch and the atrial premature beat and comprises the electrocardiosignals of the normal atrial premature beat and the atrial premature beat.
Since most of atrial premature beats appear together with the left and right fascicles and normal conditions, the 10s electrocardiosignals collected in the embodiment are the electrocardiosignals under the three conditions. The collected electrocardiographic signal data in this embodiment are all II-lead electrocardiographic signals, wherein 5000 normal atrial premature beats of 10s signals, 5000 left atrial premature beats of 10s electrocardiographic signals, and 5000 right atrial premature beats of 10s electrocardiographic signals form an electrocardiographic data set, and the electrocardiographic data set is preprocessed to form a training set of a convolutional neural network.
And labeling a data label for each electrocardiosignal in the electrocardio data set through a preprocessing process, wherein the data label comprises the heart beat type and the heart beat position at different sampling points of the electrocardiosignal. Training a convolutional neural network according to the electrocardiographic data preprocessed in the training set, inputting the electrocardiographic data to be detected in real time into a trained convolutional neural network model, and predicting the type and the position of the abnormal heart beat, thereby realizing the target detection of atrial premature beat.
Optionally, the preprocessing module in this embodiment includes:
the filtering unit is used for performing wavelet filtering and Fir windowing filtering on the collected electrocardiosignals;
and the type marking unit is used for marking a plurality of sampling points of each collected electrocardiosignal, intercepting the heart beats from left to right according to the marked R wave crest and marking the heart beat type.
The preprocessing module of this embodiment needs to preprocess each collected 10s electrocardiographic signal, and first performs wavelet filtering and Fir windowing filtering on each 10s electrocardiographic signal through the filtering unit; secondly, marking 5000 sampling points in each electrocardiosignal through a type marking unit, performing left-right intercepting heartbeat according to an R wave label given by a doctor, and marking heartbeat types for the 5000 sampling points.
In order to meet AAMI marking, the normal, left and right cardiac beats are all defaulted to be normal cardiac beats, namely the data labels are consistent. The type marking unit in this embodiment includes:
the first marking element is used for marking the sampling points in the signal section from the R wave peak position to the left 120 to the right 130 as 0.5 if the marked R wave peak is normal, a left beam branch and a right beam branch;
the second marking element is used for marking the sampling points in the signal section from the left 198 to the right 175 of the R wave peak position as 1 if the marked R wave peak is atrial premature;
and the third marking element is used for defaulting the residual sampling points marked by the first marking unit and the second marking unit as background points and marking the sampling points as 0.
For example, 5000 sampling points of an electrocardiosignal are marked, a data tag is composed of 5000 numbers, and if the data tag is: the [ 0000000.50.50.50.50.50.50.5.. 00000.50.50.5.. 1111111111.. 000000.50.5. ] section indicates that the signal section has no heartbeat type, and may be a transition section between two heartbeats; the 0.5, 0.5, 0.5.. section indicates that the signal section is normal heartbeat; the 1, 1, 1.. section indicates that the signal section is a premature atrial beat.
In order to facilitate training of the convolutional neural network and perform an independent thermalization operation on the data labels, a data label with a length of 5000 corresponding to each electrocardiographic signal is represented in a form of 5000 rows and 3 columns, each row represents a type of a point, and for example, the type can be represented as:
optionally, the convolutional neural network model of this embodiment is a convolutional neural network model based on the idea of discrete wavelet, and includes a coding unit, a coefficient processing unit, and a decoding unit, where the input electrocardiographic signal is output after being sequentially subjected to coding, coefficient processing, and decoding, so as to implement automatic coding and decoding, and directly output the heartbeat type of each sampling point;
as shown in fig. 2, the encoding unit is divided into three layers of convolutions, wherein a first layer of convolution includes a first convolutional layer and a second convolutional layer, the first convolutional layer connects the first downsampling layer, and the second convolutional layer connects the second downsampling layer and the third downsampling layer; the second convolution layer comprises a third convolution layer and a fourth convolution layer, wherein the input of the third convolution layer is the output of the second down-sampling layer, the input of the fourth convolution layer is the output of the third down-sampling layer, the third convolution layer is connected with the fourth down-sampling layer, and the fourth convolution layer is connected with the fifth down-sampling layer and the sixth down-sampling layer; the third layer convolution comprises a fifth convolution layer and a sixth convolution layer, wherein the input of the fifth convolution layer is the output of a fifth down-sampling layer, the input of the sixth convolution layer is the output of a sixth down-sampling layer, the fifth convolution layer is connected with a seventh down-sampling layer, and the sixth convolution layer is connected with an eighth down-sampling layer;
the coefficient processing unit comprises a first coefficient processing layer, a second coefficient processing layer, a third coefficient processing layer and a fourth coefficient processing layer, wherein the input of the first coefficient processing layer is the output of the first downsampling layer, the input of the second coefficient processing layer is the output of the fourth downsampling layer, the input of the third coefficient processing layer is the output of the seventh downsampling layer, and the fourth coefficient processing layer is the output of the eighth downsampling layer;
the decoding unit includes a first upsampling layer, a second upsampling layer, a third upsampling layer, a fourth upsampling layer, a fifth upsampling layer, and a sixth upsampling layer, the input of the first upsampling layer is the output of the fourth coefficient processing layer, the input of the second upsampling layer is the output of the third coefficient processing layer, the first up-sampling layer and the second up-sampling layer are connected in characteristic and then output to the third up-sampling layer, the input of the fourth up-sampling layer is the output of the second coefficient processing layer, the third up-sampling layer and the fourth up-sampling layer are connected in characteristic and then output to the fifth up-sampling layer, the input of the sixth up-sampling layer is the output of the first coefficient processing layer, and the characteristics of the fifth upsampling layer and the sixth upsampling layer are connected and then output to the seventh convolutional layer, the output of the seventh convolutional layer is used as the input of the eighth convolutional layer, and the eighth convolutional layer outputs a detection result.
The convolutional neural network model of the embodiment is a convolutional neural network model improved based on the idea of discrete wavelet reconstruction, an electrocardiosignal of 5000 × 1 is input, a vector of 5000 × 1 is output, the convolutional neural network model is capable of automatically encoding and decoding, and a heartbeat type of each sampling point can be directly output.
The number of the filters in the first convolution layer, the second convolution layer, the third convolution layer, the fourth convolution layer, the fifth convolution layer, the sixth convolution layer, the seventh convolution layer and the eighth convolution layer is sixteen, thirty-two, sixty-four, sixteen and one in sequence, the sizes of the corresponding convolution kernels are (32 x 1), the step lengths are all one, and the filling modes are all 'same'. The excitation functions of the first, second, third, fourth, fifth, sixth, and seventh convolutional layers are all Relu functions, and the excitation function of the eighth convolutional layer is a softmax function.
The first downsampling layer, the second downsampling layer, the third downsampling layer, the fourth downsampling layer, the fifth downsampling layer, the sixth downsampling layer, the seventh downsampling layer and the eighth downsampling layer adopt a maximum pooling method, and the size of a pooling window is set to be two.
The first coefficient processing layer, the second coefficient processing layer, the third coefficient processing layer and the fourth coefficient processing layer are sequentially provided with learning parameters w4, b4, w3, b3, w2, b2, w1 and b1, corresponding processing functions are linear functions, and the calculation process is yi＝wi*xi+bi(i＝1,2,3,4)。
The first upper sampling layer, the second upper sampling layer, the third upper sampling layer, the fourth upper sampling layer, the fifth upper sampling layer and the sixth upper sampling layer are subjected to convolution operation after linear interpolation is carried out, the number of the filters in the convolution layer is sixty four, thirty two, sixteen and sixteen in sequence, the convolution kernel size of the convolution layer is (32 x 1), the step length of the convolution layer is one, the filling mode is 'same', and the excitation function of the convolution layer is a Relu function.
The loss functions all use the categorical _ crosssentryp. The training algorithm may be: a random gradient descent algorithm, an Adam algorithm, a RMSProp algorithm, an adagard algorithm, an adapelta algorithm, an Adamax algorithm, and the like.
The benefits of using the convolutional neural network model of the present embodiment are: on one hand, the convolutional layer can be continuously decomposed downwards to obtain a network with a deep structure, and on the other hand, a loop network formed by the first convolutional layer, the first lower sampling layer, the first coefficient processing layer, the fifth upper sampling layer, the seventh convolutional layer and the eighth convolutional layer is shallow, so that the deep characteristics can be obtained by mining, and the gradient disappearance can be avoided.
The convolutional neural network model of the embodiment is based on the idea of discrete wavelet, if the learned coefficients are appropriate, all information of the signal can be restored after encoding and decoding, and the middle coefficient processing layer can play a role in highlighting features, so that the convolutional neural network model of the embodiment has a wider application range.
The embodiment provides a convolution neural network with a symmetrical framework improved based on a discrete wavelet thought, and the target detection of atrial premature beats is realized. The model reconstructs signals through functions in the up-sampling and convolution operations after the convolution operation and the down-sampling operation, ensures that the input and the output are equal, and thus outputs the position and the type of the central beat of the clinical 10s electrocardiosignal segment.
Example 2:
embodiment 2 of the present invention provides a computer storage medium, on which a computer program is stored, where the computer program is used to implement an execution process of the detection apparatus according to embodiment 1 of the present invention when the computer program is executed by a processor.
In light of the foregoing description of the preferred embodiments according to the present application, it is to be understood that various changes and modifications may be made without departing from the spirit and scope of the invention. The technical scope of the present application is not limited to the contents of the specification, and must be determined according to the scope of the claims.
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.
Claims (6)
1. An atrial premature target detection device based on a convolutional neural network, comprising:
the data acquisition module is used for collecting a plurality of electrocardiosignals which contain atrial premature beats and are marked with R wave crests;
the preprocessing module is used for labeling each electrocardiosignal with a data label through data preprocessing, and forming a training set by the electrocardiosignal data labeled with the data label, wherein the data label comprises the heart beat type and the heart beat position at different sampling points of the electrocardiosignal;
the neural network training module is used for taking the electrocardiosignal data in the training set as input, taking the heart beat types and positions at different sampling points of the electrocardiosignals as output, and training a convolutional neural network model;
and the target detection module is used for acquiring real-time electrocardio data to be detected, inputting the real-time electrocardio data to be detected into the trained convolutional neural network model, and predicting the type and the position of the abnormal heart beat so as to realize the target detection of the atrial premature beat.
2. The convolutional neural network-based atrial premature target detection device as recited in claim 1, wherein the cardiac electrical signals collected by the data collection module are all derived from II-lead cardiac electrical signals, and the collected cardiac electrical signals comprise: comprises the electrocardiosignals of the left bundle branch and the atrial premature beat, comprises the electrocardiosignals of the right bundle branch and the atrial premature beat and comprises the electrocardiosignals of the normal atrial premature beat and the atrial premature beat.
3. The convolutional neural network-based atrial premature target detection device as claimed in claim 2, wherein the preprocessing module comprises:
the filtering unit is used for performing wavelet filtering and Fir windowing filtering on the collected electrocardiosignals;
and the type marking unit is used for marking a plurality of sampling points of each collected electrocardiosignal, intercepting the heart beats from left to right according to the marked R wave crest and marking the heart beat type.
4. The convolutional neural network based atrial premature target detection device as claimed in claim 3, wherein the type labeling unit further comprises:
the first marking element is used for marking the sampling points in the signal section from the R wave peak position to the left 120 to the right 130 as 0.5 if the marked R wave peak is normal, a left beam branch and a right beam branch;
the second marking element is used for marking the sampling points in the signal section from the left 198 to the right 175 of the R wave peak position as 1 if the marked R wave peak is atrial premature;
and the third marking element is used for defaulting the residual sampling points marked by the first marking unit and the second marking unit as background points and marking the sampling points as 0.
5. The atrial premature target detection device based on the convolutional neural network as claimed in claim 1, wherein the convolutional neural network model comprises an encoding unit, a coefficient processing unit and a decoding unit; the coding unit is divided into three layers of convolution, wherein the first layer of convolution comprises a first convolution layer and a second convolution layer, the first convolution layer is connected with a first downsampling layer, and the second convolution layer is connected with a second downsampling layer and a third downsampling layer; the second convolution layer comprises a third convolution layer and a fourth convolution layer, wherein the input of the third convolution layer is the output of the second down-sampling layer, the input of the fourth convolution layer is the output of the third down-sampling layer, the third convolution layer is connected with the fourth down-sampling layer, and the fourth convolution layer is connected with the fifth down-sampling layer and the sixth down-sampling layer; the third layer convolution comprises a fifth convolution layer and a sixth convolution layer, wherein the input of the fifth convolution layer is the output of a fifth down-sampling layer, the input of the sixth convolution layer is the output of a sixth down-sampling layer, the fifth convolution layer is connected with a seventh down-sampling layer, and the sixth convolution layer is connected with an eighth down-sampling layer;
the coefficient processing unit comprises a first coefficient processing layer, a second coefficient processing layer, a third coefficient processing layer and a fourth coefficient processing layer, wherein the input of the first coefficient processing layer is the output of the first downsampling layer, the input of the second coefficient processing layer is the output of the fourth downsampling layer, the input of the third coefficient processing layer is the output of the seventh downsampling layer, and the input of the fourth coefficient processing layer is the output of the eighth downsampling layer;
the decoding unit includes a first upsampling layer, a second upsampling layer, a third upsampling layer, a fourth upsampling layer, a fifth upsampling layer, and a sixth upsampling layer, the input of the first upsampling layer is the output of the fourth coefficient processing layer, the input of the second upsampling layer is the output of the third coefficient processing layer, the first up-sampling layer and the second up-sampling layer are connected in characteristic and then output to the third up-sampling layer, the input of the fourth up-sampling layer is the output of the second coefficient processing layer, the third up-sampling layer and the fourth up-sampling layer are connected in characteristic and then output to the fifth up-sampling layer, the input of the sixth up-sampling layer is the output of the first coefficient processing layer, and the characteristics of the fifth upsampling layer and the sixth upsampling layer are connected and then output to the seventh convolutional layer, the output of the seventh convolutional layer is used as the input of the eighth convolutional layer, and the eighth convolutional layer outputs a detection result.
6. A computer storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, is adapted to perform the process of the detection apparatus of any one of claims 1-5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011252291.5A CN112587149A (en) | 2020-11-11 | 2020-11-11 | Atrial premature beat target detection device based on convolutional neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011252291.5A CN112587149A (en) | 2020-11-11 | 2020-11-11 | Atrial premature beat target detection device based on convolutional neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112587149A true CN112587149A (en) | 2021-04-02 |
Family
ID=75183359
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011252291.5A Pending CN112587149A (en) | 2020-11-11 | 2020-11-11 | Atrial premature beat target detection device based on convolutional neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112587149A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115130509A (en) * | 2022-06-29 | 2022-09-30 | 哈尔滨工业大学（威海） | Electrocardiosignal generation method based on conditional variational self-encoder |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20020183633A1 (en) * | 2001-03-30 | 2002-12-05 | Cardiac Pacemakers Inc. | Method and apparatus for determining the coronary sinus vein branch accessed by a coronary sinus lead |
| CN107184198A (en) * | 2017-06-01 | 2017-09-22 | 广州城市职业学院 | A kind of electrocardiosignal classifying identification method |
| CN109284752A (en) * | 2018-08-06 | 2019-01-29 | 中国科学院声学研究所 | A kind of rapid detection method of vehicle |
| EP3446628A1 (en) * | 2017-08-24 | 2019-02-27 | Karlsruher Institut für Technologie | Method and system for determining ventricular far field contribution in atrial electrograms |
| CN109480824A (en) * | 2018-12-11 | 2019-03-19 | 武汉中旗生物医疗电子有限公司 | Processing method, device and the server of EGC waveform data |
| US20190374166A1 (en) * | 2018-06-08 | 2019-12-12 | Cambridge Heartwear Limited | Detecting abnormalities in ecg signals |
| CN111772619A (en) * | 2020-06-19 | 2020-10-16 | 厦门纳龙科技有限公司 | Electrocardiogram heart beat identification method, terminal device and storage medium |
-
2020
- 2020-11-11 CN CN202011252291.5A patent/CN112587149A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20020183633A1 (en) * | 2001-03-30 | 2002-12-05 | Cardiac Pacemakers Inc. | Method and apparatus for determining the coronary sinus vein branch accessed by a coronary sinus lead |
| CN107184198A (en) * | 2017-06-01 | 2017-09-22 | 广州城市职业学院 | A kind of electrocardiosignal classifying identification method |
| EP3446628A1 (en) * | 2017-08-24 | 2019-02-27 | Karlsruher Institut für Technologie | Method and system for determining ventricular far field contribution in atrial electrograms |
| US20190374166A1 (en) * | 2018-06-08 | 2019-12-12 | Cambridge Heartwear Limited | Detecting abnormalities in ecg signals |
| CN109284752A (en) * | 2018-08-06 | 2019-01-29 | 中国科学院声学研究所 | A kind of rapid detection method of vehicle |
| CN109480824A (en) * | 2018-12-11 | 2019-03-19 | 武汉中旗生物医疗电子有限公司 | Processing method, device and the server of EGC waveform data |
| CN111772619A (en) * | 2020-06-19 | 2020-10-16 | 厦门纳龙科技有限公司 | Electrocardiogram heart beat identification method, terminal device and storage medium |
Non-Patent Citations (1)
| Title |
|---|
| 吉川隼人: "《基于Google 云平台的机器学习和深度学习入门》", pages: 162 - 164 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115130509A (en) * | 2022-06-29 | 2022-09-30 | 哈尔滨工业大学（威海） | Electrocardiosignal generation method based on conditional variational self-encoder |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Liu et al. | Arrhythmia classification of LSTM autoencoder based on time series anomaly detection | |
| CN106725428B (en) | Electrocardiosignal classification method and device | |
| WO2021057328A1 (en) | St segment classification convolutional neural network based on feature selection and method for using same | |
| CN110717416B (en) | Neural network training method for ST segment classification recognition based on feature selection | |
| CN109259756B (en) | ECG signal processing method based on secondary neural network of unbalanced training | |
| Zhang et al. | Global hybrid multi-scale convolutional network for accurate and robust detection of atrial fibrillation using single-lead ECG recordings | |
| CN115470828A (en) | Multi-lead electrocardiogram classification and identification method based on convolution and self-attention mechanism | |
| JP7487965B2 (en) | Prediction method of electrocardiogram heart rate multi-type based on graph convolution | |
| CN110974211A (en) | ST-segment classification neural network of high-order polynomial activation function and application thereof | |
| CN115153581A (en) | Electrocardiosignal characteristic waveform detection method and device and storage medium | |
| CN111956208A (en) | ECG signal classification method based on ultra-lightweight convolutional neural network | |
| CN112587149A (en) | Atrial premature beat target detection device based on convolutional neural network | |
| Jia et al. | Automatic detection and classification of 12-lead ECGs using a deep neural network | |
| Gao et al. | ECG-CL: A comprehensive electrocardiogram interpretation method based on continual learning | |
| CN112587147A (en) | Atrial premature beat target detection method based on convolutional neural network | |
| CN112587152B (en) | 12-lead T wave extraction method fusing U-net network and filtering method | |
| CN111084621B (en) | QRS wave group form identification method and device based on depth self-encoder | |
| CN114451898B (en) | Electrocardiosignal classification method for short-time training convolutional neural network | |
| CN112587151B (en) | 12-lead T wave extraction device fusing U-net network and filtering method | |
| Pomprapa et al. | Deep learning of arrhythmia analysis based on convolutional neural network | |
| CN114287950A (en) | Heart disease computer-assisted classification method based on continuous coherence and Fourier transform | |
| CN112603324A (en) | Neural network training method based on improved loss function | |
| RU2823433C1 (en) | Method of processing and analysing electrocardiogram (ecg) data | |
| Odugoudar et al. | Ecg classification system for arrhythmia detection using convolutional neural networks | |
| Barišić et al. | Cardiac Arrhythmia Classification from 12-lead Electrocardiogram Using a Combination of Deep Learning Approaches |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210402 |