CN108403111A - A kind of epileptic electroencephalogram (eeg) recognition methods and system based on convolutional neural networks - Google Patents
A kind of epileptic electroencephalogram (eeg) recognition methods and system based on convolutional neural networks Download PDFInfo
- Publication number
- CN108403111A CN108403111A CN201810101447.6A CN201810101447A CN108403111A CN 108403111 A CN108403111 A CN 108403111A CN 201810101447 A CN201810101447 A CN 201810101447A CN 108403111 A CN108403111 A CN 108403111A
- Authority
- CN
- China
- Prior art keywords
- neural networks
- convolutional neural
- electroencephalogram
- piece
- sample
- 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
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/369—Electroencephalography [EEG]
-
- 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
-
- 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
- A61B5/7271—Specific aspects of physiological measurement analysis
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Public Health (AREA)
- Surgery (AREA)
- Veterinary Medicine (AREA)
- Artificial Intelligence (AREA)
- General Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Molecular Biology (AREA)
- Psychiatry (AREA)
- Animal Behavior & Ethology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physiology (AREA)
- Signal Processing (AREA)
- Evolutionary Computation (AREA)
- Fuzzy Systems (AREA)
- Mathematical Physics (AREA)
- Psychology (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The epileptic electroencephalogram (eeg) recognition methods and system, wherein method that the invention discloses a kind of based on convolutional neural networks include:EEG signals are acquired by electrode, analyzing processing is carried out to EEG signals, obtains electroencephalogram piece;Electroencephalogram piece is inputted into trained convolutional neural networks, obtains whether electroencephalogram piece has epileptic discharge label;The training process of the convolutional neural networks includes:Collecting sample electroencephalogram piece, the sample electroencephalogram piece include that the sample electroencephalogram piece of epileptic discharge label and the sample electroencephalogram piece marked without epileptic discharge obtain trained convolutional neural networks using sample electroencephalogram piece training convolutional neural networks.The present invention has process simple, and calculation amount is small, and discrimination is high, the advantages of being not easily susceptible to environmental influence.
Description
Technical field
The invention belongs to the area of pattern recognition of image, more particularly, to a kind of epilepsy based on convolutional neural networks
Electroencephalogramrecognition recognition method and system.
Background technology
With the continuous development of medical technology and being constantly progressive for intelligent algorithm, medical technology is to more intelligentized technology
Demand bigger, intelligent medical system has become the hot issue of social life.
Existing epilepsy knows method for distinguishing, the method for mostly using feature extraction and wavelet transformation greatly, since epilepsy waveform is more
Sample and be easy interfered by electrocardio, myoelectricity, such methods be easy influenced by environment and cause to identify.There is one
Divide epileptic electroencephalogram (eeg) to know method for distinguishing to be suggested, especially uses more use similar to the morther wavelet mould of epileptic discharge waveform at present
Plate carries out wavelet transformation, to which prominent sharp spike shape inhibits other waveforms, obtains doubtful epileptic discharge waveform.But due to brain
Electrograph waveform is bigger by environmental disturbances, and is susceptible to electrocardio crosstalk, and since ecg wave form is more regular sharp spine
Wave adopts this method indistinguishable epileptic electroencephalogram (eeg) and crosstalk electrocardio, therefore is susceptible to erroneous judgement.Moreover, existing big
Partial epilepsy electroencephalogramrecognition recognition method, process is complicated, and calculation amount is too big, and discrimination is not high, is easy to be influenced by environmental condition, institute
To need the proposition of new research method.
It can be seen that the prior art is there are process complexity, calculation amount is too big, and discrimination is not high, is easy by environmental condition
The technical issues of influence.
Invention content
For the disadvantages described above or Improvement requirement of the prior art, the present invention provides a kind of based on the insane of convolutional neural networks
Thus epilepsy electroencephalogramrecognition recognition method and system solves the prior art there are process complexity, calculation amount is too big, and discrimination is not high, is easy
The technical issues of by environmental influence.
To achieve the above object, according to one aspect of the present invention, a kind of epilepsy based on convolutional neural networks is provided
Electroencephalogramrecognition recognition method, including:
(1) EEG signals are acquired by electrode, analyzing processing is carried out to EEG signals, obtains electroencephalogram piece;
(2) electroencephalogram piece is inputted into trained convolutional neural networks, obtains whether electroencephalogram piece has epileptic discharge label;
The training process of the convolutional neural networks includes:Collecting sample electroencephalogram piece, the sample electroencephalogram piece include
There are the sample electroencephalogram piece that epileptic discharge marks and the sample electroencephalogram piece marked without epileptic discharge, is instructed using sample electroencephalogram piece
Practice convolutional neural networks, obtains trained convolutional neural networks.
Further, the specific implementation of the training process of convolutional neural networks is:
(S1) collecting sample electroencephalogram piece, the sample electroencephalogram piece include the sample electroencephalogram piece of epileptic discharge label
The sample electroencephalogram piece marked with no epileptic discharge;
(S2) convolutional neural networks are built, the convolutional neural networks include input layer, convolutional layer Conv1, pond layer
Pool2, convolutional layer Conv3, pond layer Pool4, full articulamentum Fc5, full articulamentum Fc6, classification layer Softmax7 and output layer;
(S3) utilize sample electroencephalogram piece training convolutional neural networks, until output layer loss function J (θ)≤
0.0001, obtain the trained convolutional neural networks identified for epilepsy.
Further, step (S3) includes:
(S31) sample is taken from sample electroencephalogram piece, by sample input convolutional neural networks input layer, by by
The transformation of grade, the training output for obtaining convolutional neural networks are transmitted to the output layer of convolutional neural networks;
(S32) error for calculating the training output ideal output corresponding with sample label of convolutional neural networks, by minimum
The output layer of the training output input convolutional neural networks of convolutional neural networks is carried out backpropagation, adjusted by the method for changing error
The weights of whole convolutional neural networks;
(S33) step (S31) and step (S32) are repeated, it is defeated until convolutional neural networks classify layer Softmax7
The loss function J (θ)≤0.0001 for going out layer, obtains the trained convolutional neural networks identified for epilepsy.
Further, the learning rate of convolutional neural networks is 0.1.
It is another aspect of this invention to provide that providing a kind of epileptic electroencephalogram (eeg) identifying system based on convolutional neural networks, wrap
It includes:
Training module, is used for collecting sample electroencephalogram piece, and the sample electroencephalogram piece includes the sample of epileptic discharge label
This electroencephalogram piece and the sample electroencephalogram piece marked without epileptic discharge are obtained using sample electroencephalogram piece training convolutional neural networks
To trained convolutional neural networks;
Picture processing module carries out analyzing processing for acquiring EEG signals by electrode to EEG signals, obtains brain electricity
Picture;
Epileptic electroencephalogram (eeg) identification module obtains electroencephalogram piece for electroencephalogram piece to be inputted trained convolutional neural networks
Whether epileptic discharge label is had.
Further, training module includes:
Collecting sample submodule is used for collecting sample electroencephalogram piece, and the sample electroencephalogram piece includes epileptic discharge mark
The sample electroencephalogram piece of note and the sample electroencephalogram piece marked without epileptic discharge;
Convolutional neural networks submodule is built, for building convolutional neural networks, the convolutional neural networks include input
Pond layer Pool2, layer, convolutional layer Conv1 convolutional layer Conv3, pond layer Pool4, full articulamentum Fc5, full articulamentum Fc6, divide
Class layer Softmax7 and output layer;
Training submodule, for utilizing sample electroencephalogram piece training convolutional neural networks, until the loss function J of output layer
(θ)≤0.0001 obtains the trained convolutional neural networks identified for epilepsy.
Further, training submodule includes:
Sample is inputted convolutional neural networks by forward-propagating module for taking a sample from sample electroencephalogram piece
Input layer, by transformation step by step, the training output for obtaining convolutional neural networks is transmitted to the output layer of convolutional neural networks;
Backpropagation module, what the training output ideal corresponding with sample label for calculating convolutional neural networks exported
Error is carried out the output layer of the training output input convolutional neural networks of convolutional neural networks by the method for minimization error
Backpropagation adjusts the weights of convolutional neural networks;
Repetition training module, for executing repeated forward propagation module and backpropagation module, until convolutional neural networks
The loss function J (0)≤0.0001 of output layer after classification layer Softmax7, obtains the trained volume identified for epilepsy
Product neural network.
Further, the learning rate of convolutional neural networks is 0.1.
In general, through the invention it is contemplated above technical scheme is compared with the prior art, can obtain down and show
Beneficial effect:
(1) the method for the present invention has process simple, and calculation amount is small, and discrimination is high, is not easily susceptible to the excellent of environmental influence
Point.Present system is participated in the overall process inspection without doctor, automatically quick identification epileptic electroencephalogram (eeg), save doctor's energy have it is very strong
Practice value.
(2) present invention structure convolutional neural networks, using forward-propagating and backpropagation repetition training convolutional neural networks,
Make trained convolutional neural networks, realization accurately identifies epileptic electroencephalogram (eeg), have the characteristics that at a high speed, it is accurate.
Description of the drawings
Fig. 1 is overview flow chart provided in an embodiment of the present invention;
Fig. 2 is the structure chart of convolutional neural networks provided in an embodiment of the present invention;
Fig. 3 is the schematic diagram provided in an embodiment of the present invention for having epileptic electroencephalogram (eeg).
Specific implementation mode
In order to make the purpose , technical scheme and advantage of the present invention be clearer, with reference to the accompanying drawings and embodiments, right
The present invention is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, and
It is not used in the restriction present invention.As long as in addition, technical characteristic involved in the various embodiments of the present invention described below
It does not constitute a conflict with each other and can be combined with each other.
As shown in Figure 1, a kind of epileptic electroencephalogram (eeg) recognition methods based on convolutional neural networks, including:
(1) EEG signals are acquired by electrode, analyzing processing is carried out to EEG signals, obtains electroencephalogram piece;
(2) electroencephalogram piece is inputted into trained convolutional neural networks, obtains whether electroencephalogram piece has epileptic discharge label;
The training process of the convolutional neural networks includes:Collecting sample electroencephalogram piece, the sample electroencephalogram piece include
There are the sample electroencephalogram piece that epileptic discharge marks and the sample electroencephalogram piece marked without epileptic discharge, is instructed using sample electroencephalogram piece
Practice convolutional neural networks, obtains trained convolutional neural networks.
Preferably, the specific implementation of the training process of convolutional neural networks is the embodiment of the present invention:
(S1) collecting sample electroencephalogram piece, the sample electroencephalogram piece include the sample electroencephalogram piece of epileptic discharge label
The sample electroencephalogram piece marked with no epileptic discharge;
(S2) convolutional neural networks are built, as shown in Fig. 2, the convolutional neural networks include input layer, convolutional layer
Conv1, pond layer Pool2, convolutional layer Conv3, pond layer Pool4, full articulamentum Fc5, full articulamentum Fc6, classification layer
Softmax7 and output layer;The convolution kernel that wherein convolutional layer Conv1 is used when carrying out convolution can just for the convolution kernel of 11*11
What enough covering 16 was led in sample electroencephalogram image 1 leads the thick detailed information that can preferably extract waveform in picture;By convolution
After Conv1, pond layer Pool1 extracts the Eigen Structure after convolution, merges and highlights useful information;Followed by
Convolution Conv2 in, the convolution kernel for using 5*5 carries out convolution, and smaller convolution kernel is used to extract more subtle details spy
Sign;Pond layer Pool2 completes work identical with pond Pool1, and finally all characteristic images connect entirely by Fc5, Fc6 two
It connects layer and is connected with last Softmax classification layer and carry out classification comparison.Specifically:
The sample electroencephalogram of 861 × 601 pixel sizes is inputted convolutional layer Conv1 by the 1st step, and carrying out block size to it is
11 × 11 pixels and step-length are the convolution operation of 4 pixels, in total with 32 convolution kernels, obtain 32 114 × 149 pixel sizes
Characteristic pattern;
32 characteristic patterns that convolutional layer Conv1 is exported are input to pond layer Pool2, maximum pond are carried out to it by the 2nd step
The size of operation, pond block is 3 × 3 pixels, and step-length is 2 pixels, obtains the characteristic pattern that 32 resolution ratio are 57 × 74 pixels;
32 characteristic patterns of pond layer Pool2 outputs are inputted convolutional layer Conv3 by the 3rd step, and it is 5 that block size is carried out to it
× 5 pixels and the convolution operation that step-length is 4 pixels, in total with 64 convolution kernels, it is 15 × 19 pixels to obtain 64 resolution ratio
Characteristic pattern;
4th step, 64 characteristic patterns that convolutional layer Conv3 is exported input pond layer Pool4, and maximum pondization behaviour is carried out to it
Make, the size of pond block is 3 × 3 pixels, and step-length is 2 pixels, obtains the characteristic pattern that 64 resolution ratio are 7 × 9 pixels;
64 characteristic patterns of pond layer Pool4 outputs are inputted full articulamentum Fc5 by the 5th step, according to the following formula, to wherein every
One pixel is into line activating, the value of the pixel of the characteristic pattern after being activated, the sequence by the characteristic pattern after activation to arrange
It is arranged in 1 dimensional vector, obtains the feature vector of 1 × 3072 dimension.
The feature vector of full articulamentum Fc5 outputs is inputted full articulamentum Fc6, constitutes general neural network by the 6th step, defeated
Go out for the feature vector of 1 × 512 dimension;
The feature vector input classification layer Softmax7 of full articulamentum Fc6 outputs is obtained epileptic electroencephalogram (eeg) figure by the 7th step
Tag along sort, this layer of meeting calculates the probability of each tag along sort, and the label of maximum probability is exported.
(S3) utilize sample electroencephalogram piece training convolutional neural networks, until output layer loss function J (θ)≤
0.0001, obtain the trained convolutional neural networks identified for epilepsy.
Preferably, step (S3) includes the embodiment of the present invention:
(S31) sample is taken from sample electroencephalogram piece, by sample input convolutional neural networks input layer, by by
The transformation of grade, the training output for obtaining convolutional neural networks are transmitted to the output layer of convolutional neural networks;
(S32) error for calculating the training output ideal output corresponding with sample label of convolutional neural networks, by minimum
The output layer of the training output input convolutional neural networks of convolutional neural networks is carried out backpropagation, adjusted by the method for changing error
The weights of whole convolutional neural networks;
(S33) step (S31) and step (S32) are repeated, it is defeated until convolutional neural networks classify layer Softmax7
The loss function J (θ)≤0.0001 for going out layer, obtains the trained convolutional neural networks identified for epilepsy.
Preferably, the learning rate of convolutional neural networks is 0.1 to the embodiment of the present invention.
EEG signals are acquired by electrode, analyzing processing is carried out to EEG signals, obtains electroencephalogram piece;To electroencephalogram piece into
Row gray processing operates, and is contracted to input trained convolution god after 861 × 601 pixel sizes by electroencephalogram administrative division map resolution ratio
Through network, obtain whether electroencephalogram piece has epileptic discharge label;As shown in figure 3, what is obtained in the embodiment of the present invention is to have epilepsy
Discharge labelling electroencephalogram piece.
It is another aspect of this invention to provide that providing a kind of epileptic electroencephalogram (eeg) identifying system based on convolutional neural networks, wrap
It includes:
Training module, is used for collecting sample electroencephalogram piece, and the sample electroencephalogram piece includes the sample of epileptic discharge label
This electroencephalogram piece and the sample electroencephalogram piece marked without epileptic discharge are obtained using sample electroencephalogram piece training convolutional neural networks
To trained convolutional neural networks;
Picture processing module carries out analyzing processing for acquiring EEG signals by electrode to EEG signals, obtains brain electricity
Picture;
Epileptic electroencephalogram (eeg) identification module obtains electroencephalogram piece for electroencephalogram piece to be inputted trained convolutional neural networks
Whether epileptic discharge label is had.
Preferably, training module includes the embodiment of the present invention:
Collecting sample submodule is used for collecting sample electroencephalogram piece, and the sample electroencephalogram piece includes epileptic discharge mark
The sample electroencephalogram piece of note and the sample electroencephalogram piece marked without epileptic discharge;
Convolutional neural networks submodule is built, for building convolutional neural networks, the convolutional neural networks include input
Pond layer Pool2, layer, convolutional layer Conv1 convolutional layer Conv3, pond layer Pool4, full articulamentum Fc5, full articulamentum Fc6, divide
Class layer Softmax7 and output layer;
Training submodule, for utilizing sample electroencephalogram piece training convolutional neural networks, until the loss function J of output layer
(θ)≤0.0001 obtains the trained convolutional neural networks identified for epilepsy.
Preferably, training submodule includes the embodiment of the present invention:
Sample is inputted convolutional neural networks by forward-propagating module for taking a sample from sample electroencephalogram piece
Input layer, by transformation step by step, the training output for obtaining convolutional neural networks is transmitted to the output layer of convolutional neural networks;
Backpropagation module, what the training output ideal corresponding with sample label for calculating convolutional neural networks exported
Error is carried out the output layer of the training output input convolutional neural networks of convolutional neural networks by the method for minimization error
Backpropagation adjusts the weights of convolutional neural networks;
Repetition training module, for executing repeated forward propagation module and backpropagation module, until convolutional neural networks
The loss function J (θ)≤0.0001 of output layer after classification layer Softmax7, obtains the trained volume identified for epilepsy
Product neural network.
As it will be easily appreciated by one skilled in the art that the foregoing is merely illustrative of the preferred embodiments of the present invention, not to
The limitation present invention, all within the spirits and principles of the present invention made by all any modification, equivalent and improvement etc., should all include
Within protection scope of the present invention.
Claims (8)
1. a kind of epileptic electroencephalogram (eeg) recognition methods based on convolutional neural networks, which is characterized in that including:
(1) EEG signals are acquired by electrode, analyzing processing is carried out to EEG signals, obtains electroencephalogram piece;
(2) electroencephalogram piece is inputted into trained convolutional neural networks, obtains whether electroencephalogram piece has epileptic discharge label;
The training process of the convolutional neural networks includes:Collecting sample electroencephalogram piece, the sample electroencephalogram piece includes insane
The sample electroencephalogram piece of epilepsy discharge labelling and the sample electroencephalogram piece marked without epileptic discharge utilize sample electroencephalogram piece to train and roll up
Product neural network, obtains trained convolutional neural networks.
2. a kind of epileptic electroencephalogram (eeg) recognition methods based on convolutional neural networks as described in claim 1, which is characterized in that described
The specific implementation of the training process of convolutional neural networks is:
(S1) collecting sample electroencephalogram piece, the sample electroencephalogram piece include the sample electroencephalogram piece and nothing of epileptic discharge label
The sample electroencephalogram piece of epileptic discharge label;
(S2) convolutional neural networks are built, the convolutional neural networks include input layer, convolutional layer Conv1, pond layer Pool2,
Convolutional layer Conv3, pond layer Pool4, full articulamentum Fc5, full articulamentum Fc6, classification layer Softmax7 and output layer;
(S3) it is obtained until the loss function J (θ)≤0.0001 of output layer using sample electroencephalogram piece training convolutional neural networks
To the trained convolutional neural networks identified for epilepsy.
3. a kind of epileptic electroencephalogram (eeg) recognition methods based on convolutional neural networks as claimed in claim 2, which is characterized in that described
Step (S3) includes:
(S31) it takes a sample from sample electroencephalogram piece, sample is inputted to the input layer of convolutional neural networks, by step by step
Transformation, the training output for obtaining convolutional neural networks are transmitted to the output layer of convolutional neural networks;
(S32) error for calculating the training output ideal output corresponding with sample label of convolutional neural networks, is missed by minimization
The output layer of the training output input convolutional neural networks of convolutional neural networks is carried out backpropagation, adjustment volume by the method for difference
The weights of product neural network;
(S33) step (S31) and step (S32) are repeated, output layer until convolutional neural networks classify layer Softmax7
Loss function J (θ)≤0.0001, obtain the trained convolutional neural networks identified for epilepsy.
4. a kind of epileptic electroencephalogram (eeg) recognition methods based on convolutional neural networks as described in any one of claims 1-3, feature exist
In the learning rate of the convolutional neural networks is 0.1.
5. a kind of epileptic electroencephalogram (eeg) identifying system based on convolutional neural networks, which is characterized in that including:
Training module, is used for collecting sample electroencephalogram piece, and the sample electroencephalogram piece includes the sample brain of epileptic discharge label
Electrograph piece and the sample electroencephalogram piece marked without epileptic discharge are instructed using sample electroencephalogram piece training convolutional neural networks
The convolutional neural networks perfected;
Picture processing module carries out analyzing processing to EEG signals, obtains electroencephalogram for acquiring EEG signals by electrode
Piece;
Whether epileptic electroencephalogram (eeg) identification module obtains electroencephalogram piece for electroencephalogram piece to be inputted trained convolutional neural networks
There is epileptic discharge label.
6. a kind of epileptic electroencephalogram (eeg) identifying system based on convolutional neural networks as claimed in claim 5, which is characterized in that described
Training module includes:
Collecting sample submodule is used for collecting sample electroencephalogram piece, and the sample electroencephalogram piece includes what epileptic discharge marked
Sample electroencephalogram piece and the sample electroencephalogram piece marked without epileptic discharge;
Convolutional neural networks submodule is built, for building convolutional neural networks, the convolutional neural networks include input layer, volume
Lamination Conv1, pond layer Pool2, convolutional layer Conv3, pond layer Pool4, full articulamentum Fc5, full articulamentum Fc6, classification layer
Softmax7 and output layer;
Training submodule, for utilizing sample electroencephalogram piece training convolutional neural networks, until the loss function J (θ) of output layer
≤ 0.0001, obtain the trained convolutional neural networks identified for epilepsy.
7. a kind of epileptic electroencephalogram (eeg) identifying system based on convolutional neural networks as claimed in claim 6, which is characterized in that described
Training submodule include:
Sample is inputted the input of convolutional neural networks by forward-propagating module for taking a sample from sample electroencephalogram piece
Layer, by transformation step by step, the training output for obtaining convolutional neural networks is transmitted to the output layer of convolutional neural networks;
Backpropagation module, the mistake of the training output ideal output corresponding with sample label for calculating convolutional neural networks
Difference is carried out the output layer of the training output input convolutional neural networks of convolutional neural networks anti-by the method for minimization error
To propagation, the weights of convolutional neural networks are adjusted;
Repetition training module, for executing repeated forward propagation module and backpropagation module, until convolutional neural networks are classified
The loss function J (θ)≤0.0001 of output layer after layer Softmax7 obtains the trained convolution god identified for epilepsy
Through network.
8. a kind of epileptic electroencephalogram (eeg) identifying system based on convolutional neural networks as described in claim 5-7 is any, feature exist
In the learning rate of the convolutional neural networks is 0.1.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810101447.6A CN108403111A (en) | 2018-02-01 | 2018-02-01 | A kind of epileptic electroencephalogram (eeg) recognition methods and system based on convolutional neural networks |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810101447.6A CN108403111A (en) | 2018-02-01 | 2018-02-01 | A kind of epileptic electroencephalogram (eeg) recognition methods and system based on convolutional neural networks |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108403111A true CN108403111A (en) | 2018-08-17 |
Family
ID=63127549
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810101447.6A Pending CN108403111A (en) | 2018-02-01 | 2018-02-01 | A kind of epileptic electroencephalogram (eeg) recognition methods and system based on convolutional neural networks |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108403111A (en) |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110236533A (en) * | 2019-05-10 | 2019-09-17 | 杭州电子科技大学 | Epileptic seizure prediction method based on the study of more deep neural network migration features |
JP2019162237A (en) * | 2018-03-19 | 2019-09-26 | 国立大学法人 東京大学 | Epilepsy determination device, epilepsy determination system, epilepsy determination method, and epilepsy determination program |
CN110292379A (en) * | 2019-07-02 | 2019-10-01 | 重庆大学 | The recognizer of epilepsy non-invasive diagnosis based on convolutional neural networks |
CN110321783A (en) * | 2019-05-07 | 2019-10-11 | 北京慧脑云计算有限公司 | A kind of MEG spike detection method and system based on 1D convolutional neural networks |
CN110916653A (en) * | 2019-11-01 | 2020-03-27 | 天津大学 | Early warning device for reminding epileptic patient in head-wearing manner |
CN111166327A (en) * | 2020-01-06 | 2020-05-19 | 天津大学 | Epilepsy diagnosis device based on single-channel electroencephalogram signal and convolutional neural network |
CN111657935A (en) * | 2020-05-11 | 2020-09-15 | 浙江大学 | Epilepsia electroencephalogram recognition system based on hierarchical graph convolutional neural network, terminal and storage medium |
CN112294337A (en) * | 2020-09-29 | 2021-02-02 | 山东师范大学 | Epilepsy detection system based on linear graph convolution neural network |
CN113598791A (en) * | 2021-07-13 | 2021-11-05 | 杭州电子科技大学 | Consciousness disturbance classification method using space-time convolution neural network based on resting electroencephalogram |
CN113974655A (en) * | 2021-07-17 | 2022-01-28 | 北京工业大学 | Epileptic seizure prediction method based on electroencephalogram signals |
JP2022045870A (en) * | 2020-09-09 | 2022-03-22 | 生物島実験室 | Electrocardiographic abnormality detection network training method, and electrocardiographic abnormality early warning method and device |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105354568A (en) * | 2015-08-24 | 2016-02-24 | 西安电子科技大学 | Convolutional neural network based vehicle logo identification method |
CN106691378A (en) * | 2016-12-16 | 2017-05-24 | 深圳市唯特视科技有限公司 | Deep learning vision classifying method based on electroencephalogram data |
CN106821376A (en) * | 2017-03-28 | 2017-06-13 | 南京医科大学 | A kind of epileptic attack early warning system and method based on deep learning algorithm |
-
2018
- 2018-02-01 CN CN201810101447.6A patent/CN108403111A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105354568A (en) * | 2015-08-24 | 2016-02-24 | 西安电子科技大学 | Convolutional neural network based vehicle logo identification method |
CN106691378A (en) * | 2016-12-16 | 2017-05-24 | 深圳市唯特视科技有限公司 | Deep learning vision classifying method based on electroencephalogram data |
CN106821376A (en) * | 2017-03-28 | 2017-06-13 | 南京医科大学 | A kind of epileptic attack early warning system and method based on deep learning algorithm |
Non-Patent Citations (1)
Title |
---|
丁木涵等: "《基于卷积神经网络的癫痈脑电自动分类》", 《电脑知识与技术》 * |
Cited By (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2019162237A (en) * | 2018-03-19 | 2019-09-26 | 国立大学法人 東京大学 | Epilepsy determination device, epilepsy determination system, epilepsy determination method, and epilepsy determination program |
JP7072771B2 (en) | 2018-03-19 | 2022-05-23 | 国立大学法人 東京大学 | Epilepsy determination device, epilepsy determination system, epilepsy determination method and epilepsy determination program |
CN110321783A (en) * | 2019-05-07 | 2019-10-11 | 北京慧脑云计算有限公司 | A kind of MEG spike detection method and system based on 1D convolutional neural networks |
CN110236533A (en) * | 2019-05-10 | 2019-09-17 | 杭州电子科技大学 | Epileptic seizure prediction method based on the study of more deep neural network migration features |
CN110292379A (en) * | 2019-07-02 | 2019-10-01 | 重庆大学 | The recognizer of epilepsy non-invasive diagnosis based on convolutional neural networks |
CN110916653B (en) * | 2019-11-01 | 2022-03-15 | 天津大学 | Early warning device for reminding epileptic patient in head-wearing manner |
CN110916653A (en) * | 2019-11-01 | 2020-03-27 | 天津大学 | Early warning device for reminding epileptic patient in head-wearing manner |
CN111166327A (en) * | 2020-01-06 | 2020-05-19 | 天津大学 | Epilepsy diagnosis device based on single-channel electroencephalogram signal and convolutional neural network |
CN111657935A (en) * | 2020-05-11 | 2020-09-15 | 浙江大学 | Epilepsia electroencephalogram recognition system based on hierarchical graph convolutional neural network, terminal and storage medium |
CN111657935B (en) * | 2020-05-11 | 2021-10-01 | 浙江大学 | Epilepsia electroencephalogram recognition system based on hierarchical graph convolutional neural network, terminal and storage medium |
JP2022045870A (en) * | 2020-09-09 | 2022-03-22 | 生物島実験室 | Electrocardiographic abnormality detection network training method, and electrocardiographic abnormality early warning method and device |
CN112294337A (en) * | 2020-09-29 | 2021-02-02 | 山东师范大学 | Epilepsy detection system based on linear graph convolution neural network |
CN113598791A (en) * | 2021-07-13 | 2021-11-05 | 杭州电子科技大学 | Consciousness disturbance classification method using space-time convolution neural network based on resting electroencephalogram |
CN113598791B (en) * | 2021-07-13 | 2024-04-02 | 杭州电子科技大学 | Consciousness disturbance classification method based on time-space convolution neural network used by resting state electroencephalogram |
CN113974655A (en) * | 2021-07-17 | 2022-01-28 | 北京工业大学 | Epileptic seizure prediction method based on electroencephalogram signals |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108403111A (en) | A kind of epileptic electroencephalogram (eeg) recognition methods and system based on convolutional neural networks | |
CN109886273B (en) | CMR image segmentation and classification system | |
CN110516571B (en) | Cross-library micro-expression recognition method and device based on optical flow attention neural network | |
CN103996018B (en) | Face identification method based on 4DLBP | |
CN103793702B (en) | Pedestrian re-identifying method based on coordination scale learning | |
CN109584251A (en) | A kind of tongue body image partition method based on single goal region segmentation | |
CN109614869B (en) | Pathological image classification method based on multi-scale compression reward and punishment network | |
CN106097335A (en) | Digestive tract focus image identification system and recognition methods | |
CN108647614A (en) | The recognition methods of electrocardiogram beat classification and system | |
CN107451565B (en) | Semi-supervised small sample deep learning image mode classification and identification method | |
CN110287805A (en) | Micro- expression recognition method and system based on three stream convolutional neural networks | |
CN105956524A (en) | Method and device for identifying traffic signs | |
CN108038519A (en) | A kind of uterine neck image processing method and device based on dense feature pyramid network | |
CN106780434A (en) | Underwater picture visual quality evaluation method | |
CN107330393A (en) | A kind of neonatal pain expression recognition method based on video analysis | |
CN106709528A (en) | Method and device of vehicle reidentification based on multiple objective function deep learning | |
Ghosh et al. | Automatic bleeding detection in wireless capsule endoscopy based on RGB pixel intensity ratio | |
CN107767379A (en) | Pcb board marks print quality inspection method | |
CN109009102A (en) | A kind of aided diagnosis method and system based on electroencephalogram deep learning | |
CN109492668A (en) | MRI based on multichannel convolutive neural network not same period multi-mode image characterizing method | |
CN106778444A (en) | A kind of expression recognition method based on multi views convolutional neural networks | |
CN113011386B (en) | Expression recognition method and system based on equally divided characteristic graphs | |
CN113010013A (en) | Wasserstein distance-based motor imagery electroencephalogram migration learning method | |
CN113486752A (en) | Emotion identification method and system based on electrocardiosignals | |
CN111798408A (en) | Endoscope interference image detection and grading system and method |
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: 20180817 |