CN113080984A - Myocardial infarction identification and positioning method based on CNN and LSTM - Google Patents
Myocardial infarction identification and positioning method based on CNN and LSTM Download PDFInfo
- Publication number
- CN113080984A CN113080984A CN202110317031.XA CN202110317031A CN113080984A CN 113080984 A CN113080984 A CN 113080984A CN 202110317031 A CN202110317031 A CN 202110317031A CN 113080984 A CN113080984 A CN 113080984A
- Authority
- CN
- China
- Prior art keywords
- dimension
- myocardial infarction
- neural network
- layer
- lstm
- 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
- 208000010125 myocardial infarction Diseases 0.000 title claims abstract description 58
- 238000000034 method Methods 0.000 title claims abstract description 33
- 238000013528 artificial neural network Methods 0.000 claims abstract description 38
- 238000012549 training Methods 0.000 claims abstract description 32
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 24
- 238000011176 pooling Methods 0.000 claims description 30
- 230000005284 excitation Effects 0.000 claims description 18
- 238000012360 testing method Methods 0.000 claims description 18
- 230000006870 function Effects 0.000 claims description 15
- 239000013598 vector Substances 0.000 claims description 12
- 238000010586 diagram Methods 0.000 claims description 10
- 239000011159 matrix material Substances 0.000 claims description 9
- 201000009661 posterior myocardial infarction Diseases 0.000 claims description 6
- 238000010977 unit operation Methods 0.000 claims description 6
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 claims description 3
- 208000009785 Inferior Wall Myocardial Infarction Diseases 0.000 claims description 3
- 238000009825 accumulation Methods 0.000 claims description 3
- 230000000052 comparative effect Effects 0.000 claims description 3
- 201000009526 inferior myocardial infarction Diseases 0.000 claims description 3
- 201000009518 lateral myocardial infarction Diseases 0.000 claims description 3
- 238000003062 neural network model Methods 0.000 claims description 3
- 230000000747 cardiac effect Effects 0.000 claims description 2
- 230000004807 localization Effects 0.000 claims 4
- 238000004458 analytical method Methods 0.000 abstract description 4
- 238000003745 diagnosis Methods 0.000 abstract description 2
- 230000009471 action Effects 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 238000001514 detection method Methods 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 208000019622 heart disease Diseases 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000002107 myocardial effect Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention discloses a myocardial infarction identification and positioning method based on CNN and LSTM, which comprises the following steps: 1) acquiring two training databases; 2) generating an electrocardiosignal sample with 12 leads by intercepting a single heart beat, and reading in data of the electrocardiosignal with 12 leads; 3) according to different lead positions caused by different myocardial infarction generating positions, performing second-dimension connection on W points of electrocardiosignals intercepted by leads generating electrocardiogram expression at R wave top points at the same time respectively in the front wall myocardial infarction, the side wall myocardial infarction, the lower wall myocardial infarction and the rear wall myocardial infarction; 4) building a deep neural network; 5) and automatically identifying the sample. The myocardial infarction identification and positioning method based on CNN and LSTM solves the problem that the existing myocardial infarction analysis system is not enough to meet the accuracy requirement of clinical application and the problem of positioning diagnosis.
Description
Technical Field
The invention relates to the technical field of medical signal processing, in particular to a myocardial infarction identification and positioning method based on CNN and LSTM.
Background
The electrocardiogram is a conventional and efficient technical means for clinical doctors to diagnose heart diseases, and each wave of the electrocardiogram has a close relation with a myocardial action potential. At present, on the detection of myocardial infarction, the traditional signal processing methods such as Fourier transform, wavelet transform and the like are mostly adopted, the traditional method has high requirements on waveform quality, and has poor waveform effects on noisy interference and poor forms. At present, students simply use the convolutional neural network to diagnose the myocardial infarction, but the method has good effect of directly extracting the characteristics of the detection accuracy rate of typical myocardial infarction and has larger time complexity.
The myocardial infarction analysis system based on the deep learning technology can effectively improve the identification precision by utilizing data dividend, but the current myocardial infarction analysis system is not enough to meet the accuracy requirement of clinical application.
Disclosure of Invention
Technical problem to be solved
Aiming at the defects of the prior art, the invention provides a myocardial infarction identification and positioning method based on CNN and LSTM, which solves the problem that the existing myocardial infarction analysis system is not enough to meet the accuracy requirement of clinical application and the problem of positioning diagnosis.
(II) technical scheme
In order to achieve the purpose, the invention is realized by the following technical scheme: a myocardial infarction identification and location method based on CNN and LSTM comprises the following steps:
1) acquiring two training databases, wherein the training database 1 is a database known to suffer from myocardial infarction, and the other database is a comparative electrocardiosignal database of healthy people;
2) generating an electrocardiosignal sample with 12 leads by intercepting a single heartbeat, reading data of the electrocardiosignal with 12 leads in the electrocardiosignal sample, intercepting P points forward and Q points backward for each electrocardiosignal with 12 leads according to the position of the peak of an R wave at the same moment, and intercepting data of W = P + Q points for each heartbeat of the leads;
3) according to different lead positions which are changed due to different myocardial infarction generating positions, performing second-dimension connection on W points of electrocardiosignals intercepted by leads which generate electrocardiogram expression when anterior myocardial infarction, lateral myocardial infarction, inferior myocardial infarction and posterior myocardial infarction are generated at the same time by each lead of R wave top points, respectively, performing second-dimension connection on the electrocardiosignals corresponding to the anterior myocardial infarction expressed by V1-V4 leads, performing second-dimension splicing on the four leads, and amplifying the electrocardiosignals from 1W dimension to 4W dimension as input X1 of a convolutional neural network model;
electrocardiosignals corresponding to the myocardial infarction on the side wall are represented by leads I, aVL, V5 and V6, the four leads are spliced in a second dimension, and the electrocardiosignals are amplified from 1X W dimension to 4X W dimension and serve as input X2 of a convolutional neural network model;
the corresponding electrocardiosignals of the lower wall myocardial infarction are represented in II, III and aVF leads, the three leads are spliced in a second dimension, and the electrocardiosignals are amplified from 1W dimension to 3W dimension and serve as input X3 of a convolutional neural network model;
the simple posterior myocardial infarction is rare, the accumulation range can be expanded to the lower wall or the side wall under most conditions, all 12-lead electrocardiosignals are intercepted by the single cardiac beat intercepting method, and the processed X1, X2 and X3 are respectively input into one lead channel, so that the input signal size dimension of the lead channels corresponding to X1 and X2 is 4W (4 is the number of channels), the input signal size of the lead channel corresponding to X3 is 3W (the number of channels is 3), three lead channels are totally adopted, namely the combined lead signals are combined into one lead channel, and the output end of each convolutional layer unit in each lead channel is sequentially connected with an excitation unit operation and a pooling layer operation in series; the number of convolution kernels of the first convolution layer unit is 32, the size of the convolution kernels is 4, the excitation unit behind the first convolution layer unit is a relu function, the size of the pooling kernel of the pooling layer unit is 6, and the pooling step size is 3; the dimension of the characteristic diagram after passing through the first layer of pooling units is 200 x 32; the number of convolution kernels of the second convolution layer unit is 64, the size of the convolution kernels is 5, the excitation unit behind the second convolution layer unit is a relu function, the size of the pooling kernel of the pooling layer unit is 6, and the pooling step size is 3; the characteristic graph dimension after passing through the second layer of pooling units is 67 x 64;
4) building a deep neural network, wherein the deep neural network comprises a plurality of sequentially parallel lead channels, each lead channel consists of convolution layer units connected in series, the output end of each lead channel is provided with a merging layer, and the feature maps of each lead channel are merged along the last dimension, namely the dimension of the depth of the feature map; an attention layer is arranged between the combination layer at the output end of each lead channel and the LSTM layer unit to be used as a connecting unit; each convolutional layer unit comprises a convolutional layer and an excitation unit operation and a pooling layer operation which are sequentially connected with the output end of the convolutional layer in series, and the convolutional layer unit uses one-dimensional convolution and is used for extracting the characteristics of one-dimensional electrocardiosignals; the output of the LSTM layer unit is connected with a full connection layer of which the excitation unit is softmax in series; outputting;
5) and automatically identifying the sample.
Preferably, the deep neural network is composed of a convolutional layer unit and an LSTM layer unit which are sequentially connected in series; the outputs of the two convolution units in each lead channel are combined into the last dimension through a combination layer, namely the dimension of the depth of the combined output characteristic diagram, and the size of the combined characteristic diagram is 67 x 128;
after the layers are combined, an attention unit is connected in series, the attention unit constructs a weight matrix with the same dimension of 67 x 128, and is subjected to dot multiplication with corresponding elements of the convolved feature map, the output dimension of the weighted feature map is 67 x 128, elements of the weight matrix are obtained by training a neural network, and the initial value of the matrix elements is a random number within the range of 0-1; inputting the weighted feature map into an LSTM layer unit, wherein the number of hidden layers of the LSTM layer unit is 128, the dimension of an output feature map of the LSTM layer unit is 128, the output of the LSTM layer unit is connected in series with a full-connected layer of which an excitation unit is softmax, and the output dimension of the full-connected layer is 4, namely the number of categories; and finally, outputting the prediction vector dimension by the deep neural network model.
Preferably, the dimensionality of a prediction vector output by the deep neural network is 4; the method is constructed by using a keras open source framework and a python language, cross entropy is used as a loss function, and an Adam optimizer is used for optimizing the loss function.
Preferably, the parameters of the learning deep neural network are as follows: initializing training parameters of the deep neural network, and dividing the sampled signals into training set samples and test set samples; randomly extracting a part of samples from the total samples to be used as a training set, and regarding other unselected samples as a test set; inputting the multichannel electrocardiosignals X in the training set into the initialized deep neural network, and performing iteration by taking a minimized cost function as a target to generate and store the deep neural network; and updating the training parameters once every iteration until the loss value and the accuracy of the deep neural network are stabilized near a certain value, and stopping training and storing the training parameters and the model structure information of the current network.
Preferably, the automatically identifying the sample is: inputting all the divided test set samples into the stored neural network, operating the deep neural network to obtain 4-dimensional predicted value vector output corresponding to the test set samples, generating 4-dimensional label vectors by using a one-hot coding method for labels of the test set samples, and then comparing the output predicted values with the labels of the test set samples to check whether the classification is correct. .
(III) advantageous effects
The invention provides a myocardial infarction identification and positioning method based on CNN and LSTM. The method has the following beneficial effects:
(1) the method combines the advantages of CNN and LSTM, has unique advantages in the aspects of learning space data structure and time sequence structure, and can improve the learning efficiency of the network and the accuracy of myocardial infarction identification by utilizing multi-lead electrocardiogram data training.
(2) And the signal classification of multiple channels is integrated, so that the position of the myocardial infarction is positioned, and a doctor is assisted to further diagnose and treat.
Drawings
Fig. 1 is a diagram of a deep neural network architecture.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1, the present invention provides a technical solution: a myocardial infarction identification and positioning method based on CNN and LSTM is carried out according to the following steps:
1) acquiring two training databases, wherein the training database 1 is a database (abnormal data) known to suffer from myocardial infarction, and the other database is a comparative electrocardiosignal database (normal data) of healthy people;
2) generating a 12-lead electrocardiosignal sample by intercepting a single heartbeat, reading data of the 12-lead electrocardiosignal, intercepting P points forwards and Q points backwards for each lead electrocardiosignal according to the position of the peak of an R wave at the same moment, and intercepting data of W = P + Q points for each heartbeat of each lead;
3) respectively carrying out second-dimensional connection on W points of electrocardiosignals intercepted by leads generating electrocardiogram expression at the same time R wave vertex points of anterior myocardial infarction, lateral myocardial infarction, inferior myocardial infarction and posterior myocardial infarction according to different lead positions causing change caused by different myocardial infarction occurrence positions, wherein the electrocardiosignals corresponding to the anterior myocardial infarction are expressed in leads from V1 to V4, the four leads are subjected to second-dimensional splicing, and the electrocardiosignals are amplified from 1W dimension to 4W dimension and serve as input X1 of a convolutional neural network model; electrocardiosignals corresponding to the myocardial infarction on the side wall are represented by leads I, aVL, V5 and V6, the four leads are spliced in a second dimension, and the electrocardiosignals are amplified from 1X W dimension to 4X W dimension and serve as input X2 of a convolutional neural network model; the corresponding electrocardiosignals of the lower wall myocardial infarction are represented in II, III and aVF leads, the three leads are spliced in a second dimension, and the electrocardiosignals are amplified from 1W dimension to 3W dimension and serve as input X3 of a convolutional neural network model; simple posterior myocardial infarction is rare, and the accumulation range can be expanded to the lower wall or the lateral wall in most cases, so that the consideration is not separately listed;
intercepting all 12-lead electrocardiosignals by the single heart beat intercepting method, and respectively inputting the processed X1, X2 and X3 serving as input1, input2 and input3 into one lead channel, wherein the input signal size of the lead channel corresponding to X1 and X2 is 4W (4 is the number of channels), the input signal size of the lead channel corresponding to X3 is 3W (the number of channels is 3), and the number of the channels is three; combining the combined lead signals into a channel to be used as a lead channel, wherein the output end of each layer of convolution layer unit in each lead channel is sequentially connected with an excitation unit operation and a pooling layer operation in series; the number of convolution kernels of the first convolution layer unit is 32, the size of the convolution kernels is 4, the excitation unit behind the first convolution layer unit is a relu function, the size of the pooling kernel of the pooling layer unit is 6, and the pooling step size is 3; the dimension of the characteristic diagram after passing through the first layer of pooling units is 200 x 32; the number of convolution kernels of the second convolution layer unit is 64, the size of the convolution kernels is 5, the excitation unit behind the second convolution layer unit is a relu function, the size of the pooling kernel of the pooling layer unit is 6, and the pooling step size is 3; the characteristic graph dimension after passing through the second layer of pooling units is 67 x 64;
4) building a deep neural network, wherein the deep neural network comprises a plurality of sequentially parallel lead channels, each lead channel consists of convolution layer units connected in series, the output end of each lead channel is provided with a merging layer, and the feature maps of each lead channel are merged along the last dimension, namely the dimension of the depth of the feature map; an attention layer is arranged between the combination layer at the output end of each lead channel and the LSTM layer unit to be used as a connecting unit; each convolution layer unit comprises a convolution layer and an excitation unit operation and a pooling layer operation which are sequentially connected with the output end of the convolution layer in series; the convolution layer unit uses one-dimensional convolution and is used for extracting the characteristics of one-dimensional electrocardiosignals;
the output of the LSTM layer unit is connected with a full connection layer of which the excitation unit is softmax in series; outputting;
5) and automatically identifying the sample.
The deep neural network comprises a convolutional layer unit and an LSTM layer unit which are sequentially connected in series; the outputs of the two convolution units in each lead channel are combined into the last dimension through a combination layer, namely the dimension of the depth of the combined output characteristic diagram, and the size of the combined characteristic diagram is 67 x 128;
after the layers are combined, an attention unit is connected in series, the attention unit constructs a weight matrix with the same dimension of 67 x 128, and is subjected to dot multiplication with corresponding elements of the convolved feature map, the output dimension of the weighted feature map is 67 x 128, elements of the weight matrix are obtained by training a neural network, and the initial value of the matrix elements is a random number within the range of 0-1; inputting the weighted feature map into an LSTM layer unit, wherein the number of hidden layers of the LSTM layer unit is 128, the dimension of an output feature map of the LSTM layer unit is 128, the output of the LSTM layer unit is connected in series with a full-connected layer of which an excitation unit is softmax, and the output dimension of the full-connected layer is 4, namely the number of categories; and finally, outputting the prediction vector dimension by the deep neural network model.
The dimensionality of a prediction vector output by the deep neural network is 4; the method is constructed by using a keras open source framework and a python language, cross entropy is used as a loss function, and an Adam optimizer is used for optimizing the loss function.
The parameters of the learning deep neural network are as follows: initializing training parameters of the deep neural network, and dividing the sampled signals into training set samples and test set samples; randomly extracting a part of samples from the total samples to be used as a training set, and regarding other unselected samples as a test set; inputting the multichannel electrocardiosignals X in the training set into the initialized deep neural network, and performing iteration by taking a minimized cost function as a target to generate and store the deep neural network; and updating the training parameters once every iteration until the loss value and the accuracy of the deep neural network are stabilized near a certain value, and stopping training and storing the training parameters and the model structure information of the current network.
The automatic identification of the sample is: inputting all the divided test set samples into the stored neural network, operating the deep neural network to obtain 4-dimensional predicted value vector output corresponding to the test set samples, generating 4-dimensional label vectors by using a one-hot coding method for labels of the test set samples, and then comparing the output predicted values with the labels of the test set samples to check whether the classification is correct.
It is noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, 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. The term "comprising", without further limitation, means that the element so defined is not excluded from the group consisting of additional identical elements in the process, method, article, or apparatus that comprises the element.
Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.
Claims (5)
1. A myocardial infarction identification and positioning method based on CNN and LSTM is characterized in that: comprises the following steps;
1) acquiring two training databases, wherein the training database 1 is a database known to suffer from myocardial infarction, and the other database is a comparative electrocardiosignal database of healthy people;
2) generating an electrocardiosignal sample with 12 leads by intercepting a single heartbeat, reading data of the electrocardiosignal with 12 leads in the electrocardiosignal sample, intercepting P points forward and Q points backward for each electrocardiosignal with 12 leads according to the position of the peak of an R wave at the same moment, and intercepting data of W (P + Q) points for each heartbeat of the leads;
3) according to different lead positions which are changed due to different myocardial infarction generating positions, performing second-dimension connection on W points of electrocardiosignals intercepted by leads which generate electrocardiogram expression when anterior myocardial infarction, lateral myocardial infarction, inferior myocardial infarction and posterior myocardial infarction are generated at the same time by each lead of R wave top points, respectively, performing second-dimension connection on the electrocardiosignals corresponding to the anterior myocardial infarction expressed by V1-V4 leads, performing second-dimension splicing on the four leads, and amplifying the electrocardiosignals from 1W dimension to 4W dimension as input X1 of a convolutional neural network model;
electrocardiosignals corresponding to the myocardial infarction on the side wall are represented by leads I, aVL, V5 and V6, the four leads are spliced in a second dimension, and the electrocardiosignals are amplified from 1X W dimension to 4X W dimension and serve as input X2 of a convolutional neural network model;
the corresponding electrocardiosignals of the lower wall myocardial infarction are represented in II, III and aVF leads, the three leads are spliced in a second dimension, and the electrocardiosignals are amplified from 1W dimension to 3W dimension and serve as input X3 of a convolutional neural network model;
the simple posterior myocardial infarction is rare, the accumulation range can be expanded to the lower wall or the side wall under most conditions, all 12-lead electrocardiosignals are intercepted by the single cardiac beat intercepting method, and the processed X1, X2 and X3 are respectively input into one lead channel, so that the input signal size dimension of the lead channels corresponding to X1 and X2 is 4W (4 is the number of channels), the input signal size of the lead channel corresponding to X3 is 3W (the number of channels is 3), three lead channels are totally adopted, namely the combined lead signals are combined into one lead channel, and the output end of each convolutional layer unit in each lead channel is sequentially connected with an excitation unit operation and a pooling layer operation in series; the number of convolution kernels of the first convolution layer unit is 32, the size of the convolution kernels is 4, the excitation unit behind the first convolution layer unit is a relu function, the size of the pooling kernel of the pooling layer unit is 6, and the pooling step size is 3; the dimension of the characteristic diagram after passing through the first layer of pooling units is 200 x 32; the number of convolution kernels of the second convolution layer unit is 64, the size of the convolution kernels is 5, the excitation unit behind the second convolution layer unit is a relu function, the size of the pooling kernel of the pooling layer unit is 6, and the pooling step size is 3; the characteristic graph dimension after passing through the second layer of pooling units is 67 x 64;
4) building a deep neural network, wherein the deep neural network comprises a plurality of sequentially parallel lead channels, each lead channel consists of convolution layer units connected in series, the output end of each lead channel is provided with a merging layer, and the feature maps of each lead channel are merged along the last dimension, namely the dimension of the depth of the feature map; an attention layer is arranged between the combination layer at the output end of each lead channel and the LSTM layer unit to be used as a connecting unit; each convolutional layer unit comprises a convolutional layer and an excitation unit operation and a pooling layer operation which are sequentially connected with the output end of the convolutional layer in series, and the convolutional layer unit uses one-dimensional convolution and is used for extracting the characteristics of one-dimensional electrocardiosignals; the output of the LSTM layer unit is connected with a full connection layer of which the excitation unit is softmax in series; outputting;
5) and automatically identifying the sample.
2. The method of claim 1 for myocardial infarction identification and localization based on CNN and LSTM, wherein: the deep neural network comprises a convolutional layer unit and an LSTM layer unit which are sequentially connected in series; the outputs of the two convolution units in each lead channel are combined into the last dimension through a combination layer, namely the dimension of the depth of the combined output characteristic diagram, and the size of the combined characteristic diagram is 67 x 128; after the layers are combined, an attention unit is connected in series, the attention unit constructs a weight matrix with the same dimension of 67 x 128, and is subjected to dot multiplication with corresponding elements of the convolved feature map, the output dimension of the weighted feature map is 67 x 128, elements of the weight matrix are obtained by training a neural network, and the initial value of the matrix elements is a random number within the range of 0-1; inputting the weighted feature map into an LSTM layer unit, wherein the number of hidden layers of the LSTM layer unit is 128, the dimension of an output feature map of the LSTM layer unit is 128, the output of the LSTM layer unit is connected in series with a full-connected layer of which an excitation unit is softmax, and the output dimension of the full-connected layer is 4, namely the number of categories; and finally, outputting the prediction vector dimension by the deep neural network model.
3. The method of claim 2 for myocardial infarction identification and localization based on CNN and LSTM, wherein: the dimensionality of a prediction vector output by the deep neural network is 4; the method is constructed by using a keras open source framework and a python language, cross entropy is used as a loss function, and an Adam optimizer is used for optimizing the loss function.
4. A method of CNN and LSTM based myocardial infarction identification and localisation according to claim 1, 2 or 3, characterized in that: the parameters of the learning deep neural network are as follows: initializing training parameters of the deep neural network, and dividing the sampled signals into training set samples and test set samples; randomly extracting a part of samples from the total samples to be used as a training set, and regarding other unselected samples as a test set; inputting the multichannel electrocardiosignals X in the training set into the initialized deep neural network, and performing iteration by taking a minimized cost function as a target to generate and store the deep neural network; and updating the training parameters once every iteration until the loss value and the accuracy of the deep neural network are stabilized near a certain value, and stopping training and storing the training parameters and the model structure information of the current network.
5. The method of claim 1 for myocardial infarction identification and localization based on CNN and LSTM, wherein: the automatic identification of the sample is as follows: inputting all the divided test set samples into the stored neural network, operating the deep neural network to obtain 4-dimensional predicted value vector output corresponding to the test set samples, generating 4-dimensional label vectors by using a one-hot coding method for labels of the test set samples, and then comparing the output predicted values with the labels of the test set samples to check whether the classification is correct.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110317031.XA CN113080984B (en) | 2021-03-25 | 2021-03-25 | Myocardial infarction identification and positioning method based on CNN and LSTM |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110317031.XA CN113080984B (en) | 2021-03-25 | 2021-03-25 | Myocardial infarction identification and positioning method based on CNN and LSTM |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113080984A true CN113080984A (en) | 2021-07-09 |
| CN113080984B CN113080984B (en) | 2023-01-10 |
Family
ID=76669436
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110317031.XA Active CN113080984B (en) | 2021-03-25 | 2021-03-25 | Myocardial infarction identification and positioning method based on CNN and LSTM |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113080984B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114246591A (en) * | 2021-12-30 | 2022-03-29 | 清华大学 | Intelligent myocardial infarction auxiliary verification method and system based on knowledge map |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5456261A (en) * | 1993-12-16 | 1995-10-10 | Marquette Electronics, Inc. | Cardiac monitoring and diagnostic system |
| US20050197586A1 (en) * | 2000-01-31 | 2005-09-08 | Pearlman Justin D. | Method of and system for signal separation during multivariate physiological monitoring |
| CN109846471A (en) * | 2019-01-30 | 2019-06-07 | 郑州大学 | A kind of myocardial infarction detection method based on BiGRU deep neural network |
| CN109998532A (en) * | 2019-06-04 | 2019-07-12 | 广州视源电子科技股份有限公司 | Electrocardiosignal identification method and device based on multi-lead multi-structure aggregation network |
| CN110141219A (en) * | 2019-06-20 | 2019-08-20 | 鲁东大学 | Myocardial infarction automatic testing method based on lead fusion deep neural network |
| CN110226920A (en) * | 2019-06-26 | 2019-09-13 | 广州视源电子科技股份有限公司 | Electrocardiosignal identification method and device, computer equipment and storage medium |
| CN110464334A (en) * | 2019-07-23 | 2019-11-19 | 苏州国科视清医疗科技有限公司 | A kind of electrocardiographic abnormality detection method based on composite depth learning network |
| CN110495877A (en) * | 2019-08-21 | 2019-11-26 | 中国科学院深圳先进技术研究院 | A kind of Multi resolution feature extraction method and device based on ECG |
| CN111657925A (en) * | 2020-07-08 | 2020-09-15 | 中国科学院苏州生物医学工程技术研究所 | Electrocardiosignal classification method, system, terminal and storage medium based on machine learning |
-
2021
- 2021-03-25 CN CN202110317031.XA patent/CN113080984B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5456261A (en) * | 1993-12-16 | 1995-10-10 | Marquette Electronics, Inc. | Cardiac monitoring and diagnostic system |
| US20050197586A1 (en) * | 2000-01-31 | 2005-09-08 | Pearlman Justin D. | Method of and system for signal separation during multivariate physiological monitoring |
| CN109846471A (en) * | 2019-01-30 | 2019-06-07 | 郑州大学 | A kind of myocardial infarction detection method based on BiGRU deep neural network |
| CN109998532A (en) * | 2019-06-04 | 2019-07-12 | 广州视源电子科技股份有限公司 | Electrocardiosignal identification method and device based on multi-lead multi-structure aggregation network |
| CN110141219A (en) * | 2019-06-20 | 2019-08-20 | 鲁东大学 | Myocardial infarction automatic testing method based on lead fusion deep neural network |
| CN110226920A (en) * | 2019-06-26 | 2019-09-13 | 广州视源电子科技股份有限公司 | Electrocardiosignal identification method and device, computer equipment and storage medium |
| CN110464334A (en) * | 2019-07-23 | 2019-11-19 | 苏州国科视清医疗科技有限公司 | A kind of electrocardiographic abnormality detection method based on composite depth learning network |
| CN110495877A (en) * | 2019-08-21 | 2019-11-26 | 中国科学院深圳先进技术研究院 | A kind of Multi resolution feature extraction method and device based on ECG |
| CN111657925A (en) * | 2020-07-08 | 2020-09-15 | 中国科学院苏州生物医学工程技术研究所 | Electrocardiosignal classification method, system, terminal and storage medium based on machine learning |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114246591A (en) * | 2021-12-30 | 2022-03-29 | 清华大学 | Intelligent myocardial infarction auxiliary verification method and system based on knowledge map |
| CN114246591B (en) * | 2021-12-30 | 2024-04-09 | 清华大学 | Intelligent auxiliary myocardial infarction verification method and system based on knowledge graph |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113080984B (en) | 2023-01-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108875575B (en) | Automatic arrhythmia analysis method based on multi-channel signal fusion deep neural network | |
| CN108720831B (en) | Automatic arrhythmia analysis method based on lead deep neural network | |
| CN108937912B (en) | Automatic arrhythmia analysis method based on deep neural network | |
| CN108766557B (en) | Automatic arrhythmia analysis method based on channel signal fusion neural network | |
| CN111956212B (en) | Inter-group atrial fibrillation recognition method based on frequency domain filtering-multi-mode deep neural network | |
| CN110141220B (en) | Myocardial infarction automatic detection system based on multi-mode fusion neural network | |
| EP3366203A1 (en) | Method and system for cuffless blood pressure estimation using photoplethysmogram features and pulse transit time | |
| CN111700608B (en) | Electrocardiosignal multi-classification method and device | |
| CN115470828A (en) | Multi-lead electrocardiogram classification and identification method based on convolution and self-attention mechanism | |
| CN115281688B (en) | Multi-mode deep learning-based cardiac hypertrophy multi-label detection system | |
| CN113180688B (en) | Coronary heart disease electrocardiogram screening system and method based on residual error neural network | |
| CN113440149A (en) | ECG signal classification method based on twelve-lead electrocardiogram data two-dimensional multi-input residual error neural network | |
| CN113080984B (en) | Myocardial infarction identification and positioning method based on CNN and LSTM | |
| CN112270240A (en) | Signal processing method and device, electronic equipment and storage medium | |
| Tong et al. | Locating abnormal heartbeats in ECG segments based on deep weakly supervised learning | |
| Liu et al. | Precise and efficient heartbeat classification using a novel lightweight-modified method | |
| Berger et al. | Generative adversarial networks in electrocardiogram synthesis: Recent developments and challenges | |
| RU2704913C1 (en) | Neural network analysis method of cardiac condition | |
| CN117582235A (en) | Electrocardiosignal classification method based on CNN-LSTM model | |
| CN117204860A (en) | Electrocardiosignal characteristic wave positioning method | |
| CN115429284B (en) | Electrocardiosignal classification method, system, computer device and readable storage medium | |
| CN115886833A (en) | Electrocardiosignal classification method and device, computer readable medium and electronic equipment | |
| WO2023004403A1 (en) | Translating ai algorithms from 12-lead clinical ecgs to portable and consumer ecgs with fewer leads | |
| Moqurrab et al. | HRIDM: Hybrid Residual/Inception-Based Deeper Model for Arrhythmia Detection from Large Sets of 12-Lead ECG Recordings | |
| CN115836353A (en) | Method and system for searching ECG database |
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 |
Effective date of registration: 20230901 Address after: 318000 Medical Equipment Town Incubation Park, Xianju County Economic Development Zone, Taizhou City, Zhejiang Province (No. 8 Yong'an Avenue, Baita Block, Xianju Economic Development Zone, Baita Town, Xianju County), Zone A, West Side of Building 1, 2nd Floor Patentee after: Zhejiang Maomao Technology Co.,Ltd. Address before: Room 309, 3rd floor, building 2, Guorui building, 359 Jiangdong Middle Road, Jianye District, Nanjing City, Jiangsu Province, 210000 Patentee before: Nanjing diegu Health Technology Co.,Ltd. |
|
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20231128 Address after: Room 309, 3rd floor, building 2, Guorui building, 359 Jiangdong Middle Road, Jianye District, Nanjing City, Jiangsu Province, 210000 Patentee after: Nanjing diegu Health Technology Co.,Ltd. Address before: 318000 Medical Equipment Town Incubation Park, Xianju County Economic Development Zone, Taizhou City, Zhejiang Province (No. 8 Yong'an Avenue, Baita Block, Xianju Economic Development Zone, Baita Town, Xianju County), Zone A, West Side of Building 1, 2nd Floor Patentee before: Zhejiang Maomao Technology Co.,Ltd. |
|
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20240102 Address after: 318000 Medical Equipment Town Incubation Park, Xianju County Economic Development Zone, Taizhou City, Zhejiang Province (No. 8 Yong'an Avenue, Baita Block, Xianju Economic Development Zone, Baita Town, Xianju County), Zone A, West Side of Building 1, 2nd Floor Patentee after: Zhejiang Maomao Technology Co.,Ltd. Address before: Room 309, 3rd floor, building 2, Guorui building, 359 Jiangdong Middle Road, Jianye District, Nanjing City, Jiangsu Province, 210000 Patentee before: Nanjing diegu Health Technology Co.,Ltd. |
|
| TR01 | Transfer of patent right |