CN111460951A - Electrocardiosignal automatic analysis method based on deep learning - Google Patents
Electrocardiosignal automatic analysis method based on deep learning Download PDFInfo
- Publication number
- CN111460951A CN111460951A CN202010219375.2A CN202010219375A CN111460951A CN 111460951 A CN111460951 A CN 111460951A CN 202010219375 A CN202010219375 A CN 202010219375A CN 111460951 A CN111460951 A CN 111460951A
- Authority
- CN
- China
- Prior art keywords
- data
- deep learning
- electrocardio
- learning model
- dimensional
- 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
- 238000013135 deep learning Methods 0.000 title claims abstract description 14
- 238000004458 analytical method Methods 0.000 title claims abstract description 12
- 238000013136 deep learning model Methods 0.000 claims abstract description 31
- 238000012549 training Methods 0.000 claims abstract description 20
- 238000012545 processing Methods 0.000 claims abstract description 15
- 238000012360 testing method Methods 0.000 claims abstract description 11
- 238000012795 verification Methods 0.000 claims abstract description 10
- 230000000694 effects Effects 0.000 claims abstract description 8
- 238000000034 method Methods 0.000 claims description 33
- 238000005070 sampling Methods 0.000 claims description 15
- 238000010606 normalization Methods 0.000 claims description 12
- 239000013598 vector Substances 0.000 claims description 12
- 238000004422 calculation algorithm Methods 0.000 claims description 10
- 230000008569 process Effects 0.000 claims description 10
- 230000002776 aggregation Effects 0.000 claims description 7
- 238000004220 aggregation Methods 0.000 claims description 7
- 238000006243 chemical reaction Methods 0.000 claims description 4
- 238000011478 gradient descent method Methods 0.000 claims description 3
- 238000010200 validation analysis Methods 0.000 claims 1
- 238000013527 convolutional neural network Methods 0.000 description 14
- 238000001514 detection method Methods 0.000 description 4
- 230000006872 improvement Effects 0.000 description 4
- 238000013528 artificial neural network Methods 0.000 description 3
- 230000007547 defect Effects 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 238000007781 pre-processing Methods 0.000 description 3
- 238000007635 classification algorithm Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 230000015654 memory Effects 0.000 description 2
- 238000000513 principal component analysis Methods 0.000 description 2
- 238000001228 spectrum Methods 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 208000019622 heart disease Diseases 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012806 monitoring device Methods 0.000 description 1
- 208000010125 myocardial infarction Diseases 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 238000007637 random forest analysis Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 238000011426 transformation method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Biophysics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Computational Linguistics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Signal Processing (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
An automatic analysis method of electrocardiosignals based on deep learning includes downloading marked electrocardio data from a public data set, processing the electrocardio data to obtain a data set, dividing the data set into a training set, a verification set and a test set, constructing a deep learning model according to a D L A structure, training to obtain a trained deep learning model, adjusting hyper-parameters, selecting a model with the best classification effect on the verification set and the test set, processing the electrocardio data of 12 leads to be classified to obtain the data set, inputting the data of the data set into the model with the best classification effect to obtain the classification of the electrocardio signals of the electrocardio data.
Description
Technical Field
The invention relates to an electrocardiosignal automatic analysis method based on deep learning.
Background
The electrocardiogram is the most direct reaction of the electrical activity of the heart of a human body and is one of the important bases for doctors to diagnose and treat the heart diseases, and usually, the acquisition and classification of electrocardiogram waveform data are carried out in hospitals or physical examination centers, so that the defects of inconvenient detection, low detection frequency and the like exist. In recent years, with the popularization of networks and mobile smart phones, the portable electrocardiograph monitor and the home personal electrocardiograph wave monitor can be released, so that the automatic identification and classification method of the electrocardiograph signals has high practical significance.
The traditional electrocardio measurement classification method mainly comprises the following steps of signal preprocessing, waveform detection, feature extraction and classification, wherein the signal preprocessing is used for removing noises such as baseline drift, electromyographic disturbance and the like, common waveform detection algorithms comprise a wavelet transformation method, a template matching method, a graph recognition method, an energy threshold value method and the like, common electrocardio features comprise features based on waveform morphology, such as amplitude, wave width, slope and the like of each wave, and statistical features such as power spectrum features, a high-order statistics method, principal component analysis and the like, the traditional features are extracted based on the methods, machine learning methods such as random forests, support vector machines and the like are applied to classify the electrocardio beats, such as Likunyang and the like, the QRS wave groups are detected by combining the wavelet transformation and the morphology, 4 classification is carried out on the electrocardio beats, 2016L i and the like, wavelet packet decomposition, wavelet packet information entropy and interval RR are utilized, the random algorithms are used for carrying out five classification on the electrocardio signals, Khazaee and the like, and the power spectrum of the electrocardio.
The traditional electrocardiogram classification algorithm has the defects that a feature extraction method needs to be designed to obtain useful information, the selection of characteristics is usually carried out through repeated experiments or experience, the dependence on the selected features is high, and when the designed features cannot reflect the internal attributes in the data, the precision of the classification algorithm is greatly reduced. In the process, the nonlinear fitting capacity of methods such as principal component analysis, wavelet transformation and the like is limited, partial information can be lost, misclassification is easy to occur in practical application, and the method has the defects of easy generation of false positives and the like.
Disclosure of Invention
In order to overcome the problems in the prior art, the invention aims to provide an automatic electrocardiosignal analysis method based on deep learning.
In order to achieve the purpose, the invention adopts the following technical scheme:
an electrocardiosignal automatic analysis method based on deep learning comprises the following steps:
step 1, downloading electrocardio data to which electrocardio signals belong according to a public data set, wherein the electrocardio data comprise one-dimensional time sequences with different sampling frequencies and electrocardiogram image data;
step 2, processing the electrocardiogram data to obtain a data set, and dividing the data set into a training set, a verification set and a test set;
step 3, model training:
a. according to the D L A structure, a deep learning model is constructed, the input of the deep learning model is the data set obtained in the step 2, and the output layer of the deep learning model is a softmax layer;
b. putting the data of the training set obtained in the step (2) into the deep learning model constructed in the step (a) for training, continuously updating and iterating until the deep learning model converges, and obtaining a well-trained deep learning model;
c. b, adjusting the hyper-parameters of the deep learning model obtained in the step b, repeating the step b for multiple times to obtain multiple models, and selecting the model with the best classification effect on the verification set and the test set;
and 4, repeating the step 2 on the 12-lead electrocardio data to be classified to obtain a data set, and inputting the data of the data set into the deep learning model trained in the step 3 to obtain the classification of the electrocardio signals of the electrocardio data.
The further improvement of the invention is that in the step 2, the electrocardio data is processed to obtain a data set, and the specific process is as follows:
a. converting the one-dimensional time sequences with different sampling frequencies into one-dimensional electrocardio time sequences with the sampling frequency of 125Hz by a linear interpolation method;
b. performing digital-to-analog conversion on electrocardiogram image data, and converting into a one-dimensional electrocardiogram time sequence with the sampling frequency of 125Hz by using a linear interpolation method;
c. and c, cutting the 12-lead data in the one-dimensional electrocardio time sequence with the sampling frequency of 125Hz obtained in the step a or the step b into 10s time sequences at the same time interval to obtain 12 x 1250 vectors, then carrying out normalization processing on the 12 x 1250 vectors, and forming a data set by the normalization processing result.
The further improvement of the invention is that the specific process of the normalization treatment is as follows: vector values of 12 x 1250 are mapped to [0,1 ].
The invention is further improved in that in the step 2, the ratio of the data volumes of the training set, the verification set and the test set is 8:1: 1.
The invention has the further improvement that the D L A structure comprises a basic operation block and an aggregation node, wherein the basic operation block and the aggregation node are respectively composed of a plurality of one-dimensional CNN layers with convolution kernels of 1 and step length of 1, a Batch Normalization layer and a relu layer.
A further improvement of the invention consists in continuously updating the iterations using a back-propagation algorithm and a random gradient descent method.
Compared with the prior art, the invention has the following beneficial effects:
the Deep learning model used in the invention takes a one-dimensional Convolutional Neural Network (CNN) as a basic Network, selects D L A (Deep L ayerlarge Neural Network, D L A) as an integral framework, has good solving effect on the classification problem, extracts low-level waveform structural features through one-dimensional convolution, aggregates shallow and Deep layers to obtain space and features of electrocardiosignals, completes semantic analysis, obtains mutual correlation of semantic forms, and obtains good compatibility of the electrocardio images and time series of the electrocardio images.
Drawings
FIG. 1 is a flow chart of the present invention.
FIG. 2 is a view showing the structure of D L A.
Detailed Description
The present invention will be described in detail with reference to the accompanying drawings.
The deep learning model is used for processing 12-lead electrocardio data, a one-dimensional convolutional neural network is used as a basis, L AD network architecture is applied to train the deep learning model, the deep learning model is used as a black box, classification is carried out by using certain electrocardio signal characteristics without artificial designation, and a computer learns the classification process.
The invention directly processes the original electrocardio data by using the one-dimensional convolutional neural network CNN, and can eliminate the step of preprocessing the electrocardio data to remove signals, namely removing noises such as baseline drift, electromyographic disturbance and the like. This is due to the relatively low sensitivity of CNN to noise, and the use of CNN by charya et al investigated the effect of detecting myocardial infarction in the presence or absence of noise in an ECG (electrocardiogram) signal. The results are shown to be significant on both sets of data, with a slight decrease in accuracy in noisy ecg signals.
The method comprises the steps that an ECG signal is used as a time sequence, when the ECG is modeled by using a Deep learning method, waveform characteristics and connection of each waveform are usually considered, namely local correlation of the ECG and dependency relationship of front and back information are considered, a CNN algorithm and a long and Short time Memory neural network L STM (L ong Short-Term Memory, L STM) algorithm are commonly used for processing bioelectricity data, the CNN algorithm and the long and Short time Memory neural network have the advantages that the CNN can be used for extracting and identifying local waveform characteristics, and the L STM algorithm can capture the front and back information dependency relationship of time sequence, but when the L STM algorithm processes multi-lead electrocardio signals, a L STM model needs to be established for each lead, and the related information among the leads is difficult to obtain, so the method uses the CNN as a base network, and uses a D L A (Deep L a layer Aggregation, D L A) structure to replace L to obtain front and back information, and simultaneously uses 12 lead information as a whole input to extract more implicit relationships.
Referring to fig. 1, the method for automatically analyzing electrocardiographic signals based on deep learning of the present invention can realize classification of different electrocardiographic data according to a data set used for training, and the specific process is as follows:
step 1, data acquisition: and collecting a large amount of marked electrocardio data for model training. According to the method, electrocardio data are downloaded from a public data set, the part of data are marked with classification to which electrocardio signals belong, and the electrocardio data comprise one-dimensional time sequences with different sampling frequencies and electrocardiogram image data.
Step 2, data processing: the method supports one-dimensional time series or two-dimensional picture data as input.
a. Converting the one-dimensional time sequences with different sampling frequencies into one-dimensional electrocardio time sequences with the sampling frequency of 125Hz by a linear interpolation method;
b. performing digital-to-analog conversion on electrocardiogram image data, and converting into a one-dimensional electrocardiogram time sequence with the sampling frequency of 125Hz by using a linear interpolation method;
c. and c, cutting the 12-lead data in the one-dimensional electrocardio time sequence with the sampling frequency of 125Hz obtained in the step a or the step b into 10s time sequences at the same time interval to obtain 12 x 1250 vectors, then carrying out normalization processing on the 12 x 1250 vectors, and forming a data set by the normalization processing result.
The specific process of normalization treatment is as follows: vector values of 12 x 1250 are mapped to [0,1 ].
d. The data set is divided into a training set, a verification set and a test set, and the data volume ratio of the training set, the verification set and the test set is 8:1: 1.
Step 3, model training, namely constructing and training a deep learning model of the CNN + D L A architecture;
a. referring to fig. 2, a Deep learning model is constructed, according to a D L a (Deep L a Aggregation, D L a) structure, a basic operation block and an Aggregation node are composed of a plurality of one-dimensional CNN layers with convolution kernels of 1 and step length of 1, a Batch Normalization layer and a relu layer, an input of the Deep learning model is a data set obtained in step 2, an output layer of the model is a softmax layer, a model loss function L is defined as follows:
L=-∑kyklog f(xk) (1)
wherein x iskIs the vector of 12 x 1250 obtained in step 2, as input to the deep learning model, function f (x)k) Vector x representing 12 x 1250kVia a neural network model, ykIs a vector with dimension n, wherein n is the category number of the electrocardio data to be classified and represents the number input data xkClass of (a), yk=(y0,y1,...,yi,...,yn),yi∈{0,1}。ykElement (ii) of (iii), xkY corresponding to the categoryi1 and the remainder 0. the overall network is trained to aim at the parameters that minimize the loss function value L.
b. And (c) putting the data of the training set obtained by the processing in the step (2) into the deep learning model constructed in the step (a) in batches for training. And continuously updating iteration by adopting a back propagation algorithm and a random gradient descent method, updating the iteration until the deep learning model converges, and storing the deep learning model at the moment.
c. And c, adjusting the hyper-parameters of the deep learning model obtained in the step b, repeating the step b for multiple times to obtain multiple models, and selecting the model with the best classification effect on the verification set and the test set in the step 2.
And 4, processing the 12-lead electrocardio data to be classified according to the step 2, inputting the data of the obtained data set into the model trained in the step 3, and obtaining the classification of the electrocardio signal of the user through model operation.
The 12-lead electrocardiogram data to be classified can be acquired by an electrocardiogram monitoring device.
The input in the method can be one-dimensional time series electrocardiogram data or two-dimensional electrocardiogram data, the electrocardiogram data to be classified is input into a trained deep learning model after being normalized, and a classification result is obtained after calculation.
Claims (6)
1. An electrocardiosignal automatic analysis method based on deep learning is characterized by comprising the following steps:
step 1, downloading electrocardio data to which electrocardio signals belong according to a public data set, wherein the electrocardio data comprise one-dimensional time sequences with different sampling frequencies and electrocardiogram image data;
step 2, processing the electrocardiogram data to obtain a data set, and dividing the data set into a training set, a verification set and a test set;
step 3, model training:
a. according to the D L A structure, a deep learning model is constructed, the input of the deep learning model is the data set obtained in the step 2, and the output layer of the deep learning model is a softmax layer;
b. putting the data of the training set obtained in the step (2) into the deep learning model constructed in the step (a) for training, continuously updating and iterating until the deep learning model converges, and obtaining a well-trained deep learning model;
c. b, adjusting the hyper-parameters of the deep learning model obtained in the step b, repeating the step b for multiple times to obtain multiple models, and selecting the model with the best classification effect on the verification set and the test set;
and 4, repeating the step 2 on the 12-lead electrocardio data to be classified to obtain a data set, and inputting the data of the data set into the deep learning model trained in the step 3 to obtain the classification of the electrocardio signals of the electrocardio data.
2. The method for automatically analyzing the electrocardiosignals based on the deep learning as claimed in claim 1, wherein in the step 2, the electrocardio data are processed to obtain a data set by the specific process as follows:
a. converting the one-dimensional time sequences with different sampling frequencies into one-dimensional electrocardio time sequences with the sampling frequency of 125Hz by a linear interpolation method;
b. performing digital-to-analog conversion on electrocardiogram image data, and converting into a one-dimensional electrocardiogram time sequence with the sampling frequency of 125Hz by using a linear interpolation method;
c. and c, cutting the 12-lead data in the one-dimensional electrocardio time sequence with the sampling frequency of 125Hz obtained in the step a or the step b into 10s time sequences at the same time interval to obtain 12 x 1250 vectors, then carrying out normalization processing on the 12 x 1250 vectors, and forming a data set by the normalization processing result.
3. The deep learning-based automatic electrocardiosignal analysis method according to claim 2, wherein the normalization process comprises the following specific steps: vector values of 12 x 1250 are mapped to [0,1 ].
4. The automatic electrocardiosignal analysis method based on deep learning of claim 1, wherein in step 2, the ratio of the data volume of the training set, the validation set and the test set is 8:1: 1.
5. The deep learning-based automatic electrocardiosignal analysis method according to claim 1, wherein the D L A structure comprises a basic operation block and an aggregation node, and the basic operation block and the aggregation node are respectively composed of a plurality of one-dimensional CNN layers with convolution kernels of 1 and step length of 1, a Batch Normalization layer and a relu layer.
6. The automatic electrocardiosignal analysis method based on deep learning of claim 1, wherein iteration is continuously updated by adopting a back propagation algorithm and a random gradient descent method.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010219375.2A CN111460951A (en) | 2020-03-25 | 2020-03-25 | Electrocardiosignal automatic analysis method based on deep learning |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010219375.2A CN111460951A (en) | 2020-03-25 | 2020-03-25 | Electrocardiosignal automatic analysis method based on deep learning |
Publications (1)
Publication Number | Publication Date |
---|---|
CN111460951A true CN111460951A (en) | 2020-07-28 |
Family
ID=71679812
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202010219375.2A Pending CN111460951A (en) | 2020-03-25 | 2020-03-25 | Electrocardiosignal automatic analysis method based on deep learning |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111460951A (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112580360A (en) * | 2020-11-11 | 2021-03-30 | 上海数创医疗科技有限公司 | Electrocardio term semantic matching device |
CN112633005A (en) * | 2020-11-11 | 2021-04-09 | 上海数创医疗科技有限公司 | Electrocardio term semantic matching method |
CN112932433A (en) * | 2021-02-05 | 2021-06-11 | 青岛大学 | Arrhythmia automatic classification method of electrocardiogram |
CN113143286A (en) * | 2021-04-30 | 2021-07-23 | 广州大学 | Electrocardiosignal identification method, system, device and medium based on distributed learning |
CN113901893A (en) * | 2021-09-22 | 2022-01-07 | 西安交通大学 | Electrocardiosignal identification and classification method based on multiple cascade deep neural network |
CN114648049A (en) * | 2022-05-20 | 2022-06-21 | 合肥心之声健康科技有限公司 | Method, device and system for constructing and classifying electrocardio image classification model |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106901723A (en) * | 2017-04-20 | 2017-06-30 | 济南浪潮高新科技投资发展有限公司 | A kind of electrocardiographic abnormality automatic diagnosis method |
CN107657318A (en) * | 2017-11-13 | 2018-02-02 | 成都蓝景信息技术有限公司 | A kind of electrocardiogram sorting technique based on deep learning model |
CN107822622A (en) * | 2017-09-22 | 2018-03-23 | 成都比特律动科技有限责任公司 | Electrocardiographic diagnosis method and system based on depth convolutional neural networks |
CN108647565A (en) * | 2018-03-28 | 2018-10-12 | 浙江工业大学 | A kind of data preprocessing method classified to electrocardiosignal based on deep learning model |
CN110269605A (en) * | 2019-05-28 | 2019-09-24 | 安徽心之声医疗科技有限公司 | A kind of electrocardiosignal noise recognizing method based on deep neural network |
-
2020
- 2020-03-25 CN CN202010219375.2A patent/CN111460951A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106901723A (en) * | 2017-04-20 | 2017-06-30 | 济南浪潮高新科技投资发展有限公司 | A kind of electrocardiographic abnormality automatic diagnosis method |
CN107822622A (en) * | 2017-09-22 | 2018-03-23 | 成都比特律动科技有限责任公司 | Electrocardiographic diagnosis method and system based on depth convolutional neural networks |
CN107657318A (en) * | 2017-11-13 | 2018-02-02 | 成都蓝景信息技术有限公司 | A kind of electrocardiogram sorting technique based on deep learning model |
CN108647565A (en) * | 2018-03-28 | 2018-10-12 | 浙江工业大学 | A kind of data preprocessing method classified to electrocardiosignal based on deep learning model |
CN110269605A (en) * | 2019-05-28 | 2019-09-24 | 安徽心之声医疗科技有限公司 | A kind of electrocardiosignal noise recognizing method based on deep neural network |
Non-Patent Citations (2)
Title |
---|
李凌等: "心率变异信号的非均匀采样问题研究", 《北京生物医学工程》 * |
芦振寰: "基于深度学习的端到端心电自动分类方法研究与实现", 《中国优秀硕士学位论文全文数据(电子期刊)医药卫生科技辑》 * |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112580360A (en) * | 2020-11-11 | 2021-03-30 | 上海数创医疗科技有限公司 | Electrocardio term semantic matching device |
CN112633005A (en) * | 2020-11-11 | 2021-04-09 | 上海数创医疗科技有限公司 | Electrocardio term semantic matching method |
CN112633005B (en) * | 2020-11-11 | 2024-06-21 | 上海数创医疗科技有限公司 | Electrocardiogram term semantic matching method |
CN112932433A (en) * | 2021-02-05 | 2021-06-11 | 青岛大学 | Arrhythmia automatic classification method of electrocardiogram |
CN112932433B (en) * | 2021-02-05 | 2022-11-08 | 青岛大学 | Arrhythmia automatic classification method of electrocardiogram |
CN113143286A (en) * | 2021-04-30 | 2021-07-23 | 广州大学 | Electrocardiosignal identification method, system, device and medium based on distributed learning |
CN113143286B (en) * | 2021-04-30 | 2023-06-20 | 广州大学 | Electrocardiosignal identification method, system, device and medium based on distributed learning |
CN113901893A (en) * | 2021-09-22 | 2022-01-07 | 西安交通大学 | Electrocardiosignal identification and classification method based on multiple cascade deep neural network |
CN113901893B (en) * | 2021-09-22 | 2023-09-15 | 西安交通大学 | Electrocardiosignal identification and classification method based on multi-cascade deep neural network |
CN114648049A (en) * | 2022-05-20 | 2022-06-21 | 合肥心之声健康科技有限公司 | Method, device and system for constructing and classifying electrocardio image classification model |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111460951A (en) | Electrocardiosignal automatic analysis method based on deep learning | |
US11564612B2 (en) | Automatic recognition and classification method for electrocardiogram heartbeat based on artificial intelligence | |
CN107822622B (en) | Electrocardiogram diagnosis method and system based on deep convolutional neural network | |
CN111990989A (en) | Electrocardiosignal identification method based on generation countermeasure and convolution cyclic network | |
CN108596142B (en) | PCANet-based electrocardiogram feature extraction method | |
Zhao et al. | ECG classification using deep CNN improved by wavelet transform | |
CN109480824B (en) | Method and device for processing electrocardio waveform data and server | |
CN109009102B (en) | Electroencephalogram deep learning-based auxiliary diagnosis method and system | |
CN112426160A (en) | Electrocardiosignal type identification method and device | |
CN110751131B (en) | Arrhythmia detection device | |
CN110558975B (en) | Electrocardiosignal classification method and system | |
CN113116300A (en) | Physiological signal classification method based on model fusion | |
CN111053552B (en) | QRS wave detection method based on deep learning | |
CN110897634A (en) | Electrocardiosignal generation method based on generation countermeasure network | |
CN115530788A (en) | Arrhythmia classification method based on self-attention mechanism | |
CN112336369B (en) | Coronary heart disease risk index evaluation system of multichannel heart sound signals | |
CN117582235A (en) | Electrocardiosignal classification method based on CNN-LSTM model | |
CN115281676A (en) | Fatigue detection method based on GRU neural network and ECG signal | |
CN113712525A (en) | Physiological parameter processing method and device and medical equipment | |
Rajesh et al. | A Silent Cardiac Atrial Fibrillation Detection and Classification using Deep Learning Approach | |
Salah et al. | Cardiac arrhythmia classification by wavelet transform | |
Gupta et al. | ECG signal analysis using emerging tools in current scenario of health informatics | |
CN116570295B (en) | Electrocardiogram low-voltage T-wave end point positioning method | |
Verma et al. | Review On Classification And Prediction Of Ecg Morphology And Intervals Features | |
Veerappan et al. | A Review on ECG-Signal Classification of Scalogram Snap shots the use of Convolutional Neural Network and Continuous Wavelet Transform |
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 |
Application publication date: 20200728 |
|
RJ01 | Rejection of invention patent application after publication |