CN110353672A - Eye artefact removal system and minimizing technology in a kind of EEG signals - Google Patents
Eye artefact removal system and minimizing technology in a kind of EEG signals Download PDFInfo
- Publication number
- CN110353672A CN110353672A CN201910635391.7A CN201910635391A CN110353672A CN 110353672 A CN110353672 A CN 110353672A CN 201910635391 A CN201910635391 A CN 201910635391A CN 110353672 A CN110353672 A CN 110353672A
- Authority
- CN
- China
- Prior art keywords
- eye
- artefact
- module
- eye artefact
- eeg signals
- 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
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
-
- 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
- A61B5/7203—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal
-
- 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
- A61B5/725—Details of waveform analysis using specific filters therefor, e.g. Kalman or adaptive filters
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Medical Informatics (AREA)
- Surgery (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Veterinary Medicine (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Public Health (AREA)
- Molecular Biology (AREA)
- Physics & Mathematics (AREA)
- Animal Behavior & Ethology (AREA)
- General Health & Medical Sciences (AREA)
- Psychiatry (AREA)
- Signal Processing (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physiology (AREA)
- Psychology (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The present invention relates to eye artefact removal system and minimizing technologies in a kind of EEG signals.Provided system includes signal interception module, eye artefact identification module and eye artefact removal module;One section of original EEG signals of subject are sent into eye artefact identification module and eye artefact removal module simultaneously after the processing of signal interception module, if this section of EEG signals include eye artefact, the eye artefact that eye artefact identification module is identified is sent into eye artefact as reference signal and removes module;Otherwise, by the reference signal zero setting based on recurrence least square sef-adapting filter, the online removal of eye artefact is finally realized using the auto-adaptive filtering technique of eye artefact removal module.The present invention can not only remove eye artefact online, and remain a large amount of useful information in original brain electricity, do not require the port number of input brain electricity, be further reduced the root-mean-square error of signal, time-consuming is few, is more suitable for real-time BCI scene.
Description
Technical field
The present invention relates to technical field of biological information, the in particular to acquisition of EEG signals and preconditioning technique, specifically relate to
And eye artefact removal system and minimizing technology in a kind of EEG signals.
Background technique
Electroencephalogram is most widely used noninvasive brain imaging technique in real-time brain-computer interface, since it has the good time
The price of resolution ratio, ease for use and relative moderate, therefore furtherd investigate in fields such as clinical medicine, affection computations.However,
Since electroencephalogram is multichannel, with the random signal of non-stationary property, its very faint (usually only microvolt magnitude), pole
Pollution vulnerable to physiologic factors such as blink, eye movement, heartbeat, muscle activity, perspirations.Especially with eye movement and blink
Relevant electrical interference (also referred to as eye artefact), it can change eye circumference and corticocerebral field distribution, to point of EEG signals
Analysis brings great difficulty.Since eye artefact is inevitable in the collection process of EEG signals, it effectively removes
Eye artefact in EEG signals has very important significance to obtain clean EEG signals.
In recent years, for the eye artefact in EEG signals, it has been proposed that many minimizing technologies, these methods are main
It is divided into following four classes:
Linear filtering: this is one of the most common method for removing eye artefact in EEG signals.Such method is according to eye
Portion's artefact this different characteristic with EEG signals frequency range, the eye filtered out in EEG signal using high-pass filter are pseudo-
Mark.But since eye artefact usually overlaps with EEG signals, linear filter is probably pseudo- in removal
Useful EEG signals are eliminated during mark.Adaptive filter method can be regarded as a kind of improvement of linear filter method, should
Method is using contaminated EEG signals as input signal, using electro-ocular signal as reference signal, subtracts ginseng by input signal
Signal is examined to obtain pure EEG signals.Compared with linear filter, sef-adapting filter can with automatically adjusting parameter, and
In experiment without any priori statistical knowledge about signal and artefact.But it needs additional eye electric channel as ginseng
It examines.
Blind source separating: wherein independent component analysis (ICA) is one of most popular blind source separation method, it assumes that collecting
EEG signals be spontaneous brain electricity signal Yu various artefacts linear hybrid, original multi-channel EEG signal is decomposed into identical number
The independent element of amount abandons independent element related with eye artefact, and clean EEG letter is rebuild using other independent elements
Number.But this method not only results in the loss of useful information, can also weaken the independence between different brain electric channels, remove
Except this, it also needs a large amount of brain electric channel.
Wavelet transformation: such methods (wavelet analysis for being based especially on threshold technology) have been widely used for artifact removal.
This method passes through wavelet transformation first and the EEG signals comprising eye artefact is decomposed into a series of subband, then uses threshold value
Function automatically corrects the coefficient of subband relevant to eye artefact, finally rebuilds pure brain electricity according to the coefficient after correction
Signal.Such methods can be used for online eye artefact removal, and not need electro-ocular signal as reference.But wavelet transformation
Threshold value is often difficult to determine, the generalization ability that inappropriate threshold value may cause brain machine interface system reduces.
Neural network: in recent years, neural network (such as wavelet neural network etc.) has been applied to the removal of eye artefact.This
Kind method replaces the threshold function table in wavelet analysis using neural network, avoids disadvantage brought by artificial selection threshold function table
End.In addition to this, waiting people quietly recklessly also proposed the fuzzy reasoning based on functional Connection Neural Network and adaptive neural network
System (FLNN-ANFIS) removes eye artefact and electro photoluminescence artefact.But these methods require electro-ocular signal as ginseng
Examine signal.
Summary of the invention
The present invention proposes eye artefact removal system and minimizing technology in a kind of EEG signals, to overcome the prior art to exist
The additional eye electric channel of needs as reference, to the number of brain electric channel have it is fixed require, the technology that generalization ability is not high is asked
Topic.
In order to reach the purpose of the present invention, present invention provide the technical scheme that in a kind of EEG signals eye artefact from
Dynamic removal system, signal interception module, eye artefact identification module and eye artefact including successively connecting remove module, original
EEG signals are sent into eye artefact identification module and eye artefact removal module simultaneously after the processing of signal interception module, and eye is pseudo-
Artefact recognition result is sent into eye artefact again and removes module by mark identification module;
The signal of input is carried out sampling and random interception isoplith processing by the signal interception module, obtains single channel
The sample set of original EEG signals;The eye artefact identification module is constructed based on convolution self-encoding encoder, which will be original
The sample set of brain electricity projects in the subspace by the extension of eye electricity sample, identifies the eye artefact wherein adulterated;The eye
Artefact removal module is constructed based on recurrence least square sef-adapting filter, and the module is for removing artefact.
Eye artefact automatically removes the processing method of system in a kind of EEG signals, is extracted according to random adding window and single channel
Mode establish the original brain electricity sample set of single channel, after obtaining sample set, sequence the following steps are included:
One section of original EEG signals of subject are obtained to the original brain electricity of single channel after the processing of signal interception module first
Signal;
Then the original EEG signals of single channel are sent into eye artefact identification module simultaneously and eye artefact removes module, if
This section of EEG signals include eye artefact, then the eye artefact identified eye artefact identification module is sent into as reference signal
Eye artefact removes module;Otherwise, the reference signal zero setting of the sef-adapting filter in eye artefact identification module;
Last eye artefact remove module using the original EEG signals of single channel and eye artefact as input signal and
Reference signal removes artefact in such a way that input subtracts reference.
The processing method that eye artefact automatically removes in above-mentioned EEG signals, the method for obtaining sample set is: first
The electro-ocular signal of four-way is first sent into signal interception module, obtains eye electricity sample set;Then using eye electricity sample set training volume
Product self-encoding encoder makes peak value and local characteristics of mean of its study to eye artefact, obtains optimized parameter collection, be stored in eye puppet
In mark identification module.
The processing method that eye artefact automatically removes in above-mentioned EEG signals, after obtaining sample set, specific steps are such as
Under:
Step 1: the single channel EEG signals of given segment being sent into signal interception module, obtain brain electricity sample;
Step 2: this section of brain electricity sample being sent into eye artefact as input signal and removes module;
Step 3: this section of brain electricity sample being sent into eye artefact identification module, is projected into the son extended by eye electricity sample
In space, judge whether the segment brain electricity includes eye artefact, if comprising, then follow the steps 4, it is no to then follow the steps 5;
Step 4: the eye artefact come out by step 5 automatic identification being sent into eye artefact as reference signal and removes module
In;
Step 5: by the reference signal zero setting of eye artefact removal module, executing step 6;
Step 6: in eye artefact removal module, input signal subtracts reference signal, final to remove eye artefact, executes
Step 7;
Step 7: end operation or the channel for replacing EEG signals repeat step 1 to step 6.
Compared with prior art, the invention has the beneficial effects that:
1, the present invention identifies artefact using convolution self-encoding encoder, it can preferably retain the spatial information being originally inputted, and
And it is not necessarily to any feature of manual extraction and any prior information about original brain electricity, convolution self-encoding encoder is completed once training,
Then subsequent eye artefact automatic identification and online removal process be all not necessarily to synchronous acquisition electro-ocular signal as reference, and have compared with
High recognition effect and faster recognition speed, this method not only can efficiently and online remove the eye in original EEG signals
Artefact, and remain a large amount of useful information in original EEG signals;
2, the number of input brain electric channel is not required, can be not only used for single channel brain electric equipment, it can also be used to multichannel
Brain electric equipment.
3, whole process of the present invention can systematically be rapidly completed in MATLAB, go eye artefact effect compared to only
Vertical constituent analysis (Independent component analysis, ICA) method is more preferable, can be further reduced the equal of signal
Square error, time-consuming is few, and generalization ability is strong, is more suitable for real-time BCI scene.
Detailed description of the invention
Fig. 1 is system diagram of the invention.
Fig. 2 is the positioning figure of the electro-ocular signal of four-way.
Fig. 3 is eye artefact waveform diagram.
Fig. 4 is tested using the method for the present invention, to the effect picture of original EEG signals removal eye artefact.
Fig. 5 is the effect contrast figure using the method for the present invention and ICA method removal eye artefact.
Fig. 6 is PSD curve graph corresponding with Fig. 5.
Fig. 7 is the time frequency analysis figure using the method for the present invention and ICA method reconstruct brain electricity.
Specific embodiment
Below in conjunction with attached drawing 1, the present invention is described in detail.
Data introduction:
Eeg data derives from the multi-modal data collection DEAP for sentiment analysis, sees respectively including 32 subjects
See collected EEG signals and other peripheral physiological signals when a length of one minute music video at 40, totally 48 lead to
Road, wherein 1~32 channel is EEG signals, 33~36 channels are electro-ocular signals.This experiment uses data for the letter in preceding 36 channel
Number.
Referring to Fig. 1, eye artefact automatically removes system in a kind of EEG signals provided by the invention, including what is successively connected
Signal interception module, eye artefact identification module and eye artefact remove module, original EEG signals through signal interception module at
Eye artefact identification module and eye artefact removal module are sent into after reason simultaneously, artefact is identified tie again by eye artefact identification module
Fruit is sent into eye artefact and removes module;
The signal of input is carried out sampling and random interception isoplith processing by the signal interception module, obtains single channel
The sample set of original EEG signals;The eye artefact identification module is constructed based on convolution self-encoding encoder, which will be original
The sample set of brain electricity projects in the subspace by the extension of eye electricity sample, identifies the eye artefact wherein adulterated;The eye
Artefact removal module is constructed based on recurrence least square sef-adapting filter, and the module is for removing artefact.
Described signal interception module is that the signal of input is carried out to sampling and random interception isoplith processing, obtains sample
This collection.In order to obtain enough training samples, to subject 1 collected four-way electro-ocular signal carry out sample cut
It takes, the positioning figure of the electro-ocular signal of four-way used in the present invention is as shown in Fig. 2, including two horizontal eye electricity
(vEOG) and two vertical eyes are electric (hEOG).Come to intercept 10000 eye electricity samples at random by way of adding rectangular window herein
{x(1), x(2)..., x(10000), wherein each training sample x(i), i≤10000 are s1× 4 matrix, s1Refer to each training
The sampling number of sample, while being also the input layer unit number of convolution self-encoding encoder.Here s1=1000.
Described eye artefact identification module is that the sample set of original brain electricity is projected to the subspace extended by eye electricity sample
In, automatically identify the eye artefact wherein adulterated.Since eye artefact is the interference caused by EEG signals of blink or eye movement,
Therefore the feature of eye artefact is actually identical with electro-ocular signal, and waveform is as shown in Figure 3.Therefore, using being truncated to
Eye electric sample trains the convolution self-encoding encoder being made of one layer of convolution, one layer of pond and one layer of deconvolution, changes by 500 times
After generation, the objective function L (θ) of convolution self-encoding encoder restrains, and it is optimal to indicate that convolution self-encoding encoder is obtained by 500 training
Parameter sets, i.e. unsupervised learning are to the features such as the peak value of electro-ocular signal and local mean value.Once optimized parameter collection is obtained, by sample
This collection is stored in eye artefact identification module, then it represents that convolution self-encoding encoder model has been built up, subsequent progress eye electricity artefact
Automatic identification and removal can be completed online in the case where default electro-ocular signal.By one section of single pass original EEG signals
Y (n) is sent into the module, can project to y (n) in the subspace by electro-ocular signal extension, wherein mix to automatically identify
Miscellaneous eye artefact r (n).
Eye artefact, which removes module, to be obtained by the original EEG signals y (n) of single channel and through eye artefact identification module
Eye artefact r (n) removes artefact in such a way that input subtracts reference respectively as input signal and reference signal.H (m) table
Show that length is the coefficient of finite impulse response (FIR) (FIR) filter of M.In order to obtain optimal artefact removal effect, recurrence is used
Least-squares algorithm (RLS) adjusts the coefficient of FIR filter, filters out the eye artefact in input signal, exports clean brain
Electric signal e (n):
Eye artefact automatically removes the processing method of system in a kind of EEG signals provided by the invention, according to random adding window
The mode extracted with single channel establishes the original brain electricity sample set of single channel, and the method for obtaining sample set is: first by four-way
Electro-ocular signal is sent into signal interception module, obtains eye electricity sample set;Then eye electricity sample set training convolutional self-encoding encoder is used, is made
It learns peak value and local characteristics of mean to eye artefact, obtains optimized parameter collection, is stored in eye artefact identification module.
After obtaining sample set, sequence the following steps are included:
One section of original EEG signals of subject are obtained to the original brain electricity of single channel after the processing of signal interception module first
Signal;
Then the original EEG signals of single channel are sent into eye artefact identification module simultaneously and eye artefact removes module, if
This section of EEG signals include eye artefact, then the eye artefact identified eye artefact identification module is sent into as reference signal
Eye artefact removes module;Otherwise, the reference signal zero setting of the sef-adapting filter in eye artefact identification module;
Last eye artefact remove module using the original EEG signals of single channel and eye artefact as input signal and
Reference signal removes artefact in such a way that input subtracts reference.
In particular: method of the invention is after obtaining sample set, and steps are as follows:
Step 1: the single channel EEG signals of given segment being sent into signal interception module, obtain brain electricity sample;
Step 2: this section of brain electricity sample being sent into eye artefact as input signal and removes module;
Step 3: this section of brain electricity sample being sent into eye artefact identification module, is projected into the son extended by eye electricity sample
In space, judge whether the segment brain electricity includes eye artefact, if comprising, then follow the steps 4, it is no to then follow the steps 5;
Step 4: the eye artefact come out by step 5 automatic identification being sent into eye artefact as reference signal and removes module
In;
Step 5: by the reference signal zero setting of eye artefact removal module, executing step 6;
Step 6: in eye artefact removal module, input signal subtracts reference signal, final to remove eye artefact, executes
Step 7;
Step 7: end operation or the channel for replacing EEG signals repeat step 1 to step 6.
It is handled using F7 channel of the method for the present invention to subject 2, obtains effect picture before and after the processing, such as Fig. 4 institute
Show.As seen from Figure 4, the method for the present invention can be effectively removed by blink and eye movement bring eye artefact, and thus
The EEG signals of method reconstruct are consistent with the trend of original EEG signals.
Beneficial effect in order to further illustrate the present invention will use the present invention treated result and the popular side ICA
Method is compared.Use power spectral density plot (PSD), root-mean-square error (RMSE) and time frequency analysis figure (JTFA) as measurement
Index.
Fig. 5 is the Contrast on effect for removing eye artefact to the channel FP of subject 3 using the method for the present invention and ICA method
Figure, it can be seen that although ICA method also can preferably reconstruct EEG signals, the effect that it removes eye artefact does not have
The method of the present invention is good.Fig. 6 is PSD curve graph corresponding with Fig. 5, and PSD curve illustrates the power with frequency variation of signal
Situation.Theoretically, preferably the PSD value of reconstruct EEG signals should be in the frequency range that eye artefact is concentrated as far as possible
Ground is small, and is consistent in PSD curvilinear trend of other frequency ranges then as much as possible with original brain electricity.As seen from Figure 6,
In the frequency range that eye artefact is concentrated, the PSD value of the method for the present invention CAE-RLS is less than ICA, and in other frequency ranges
Interior, for ICA method, the PSD curve of CAE-RLS method is closer to original EEG signals.
RMSE is used to assess the accuracy between measured value and true value, carrys out appraisal procedure reconstruct brain electricity used here as RMSE
The ability of signal, formula are as follows:
One section is chosen from original EEG signals and does not include eye artefact and relatively stable waveform, is counted according to formula (2)
The RMSE value of the reconstruct EEG signals obtained using CAE-RLS method and ICA method is calculated, the results are shown in Table 1.
Table 1 uses the RMSE value of CAE-RLS and ICA reconstruction brain electricity
As it can be seen from table 1 using the method for the present invention lose brain electric information more less.
Fig. 7 gives the time-frequency gone using the method for the present invention and ICA method to the channel the F7 brain electricity of subject 2 before and after artefact
Analysis chart.As shown in fig. 7, original brain electricity included high-energy information at 0-1 seconds and 2-4 seconds, gone using CAE-RLS of the present invention
After artefact, although the amplitude of the high-energy information on corresponding position is lowered by, but the video properties in original brain electricity are but
It is completely remained, this is more advantageous to emotion recognition or other brain electricity applications.Although and also can be compared with by ICA method
EEG signals are reconstructed well, but some useful informations that it may result in original brain electricity disappear, such as the middle section Fig. 7 institute
Show.Therefore, the method for the present invention is on frequency performance also superior to ICA method.
Claims (4)
1. eye artefact automatically removes system in a kind of EEG signals, it is characterised in that: the signal including successively connecting intercepts mould
Block, eye artefact identification module and eye artefact remove module, and original EEG signals are sent simultaneously after the processing of signal interception module
Enter eye artefact identification module and eye artefact removal module, artefact recognition result is sent into eye again by eye artefact identification module
Artefact removes module;
The signal of input is carried out sampling and random interception isoplith processing by the signal interception module, and it is original to obtain single channel
The sample set of EEG signals;The eye artefact identification module is constructed based on convolution self-encoding encoder, and the module is electric by original brain
Sample set project to by eye electricity sample extension subspace in, identify the eye artefact wherein adulterated;The eye artefact
Removal module is constructed based on recurrence least square sef-adapting filter, and the module is for removing artefact.
2. eye artefact automatically removes the processing method of system, feature in a kind of EEG signals according to claim 1
It is: the original brain electricity sample set of single channel is established according to the mode that random adding window and single channel extract, after obtaining sample set,
Sequence the following steps are included:
One section of original EEG signals of subject are obtained into the original EEG signals of single channel after the processing of signal interception module first;
Then the original EEG signals of single channel are sent into eye artefact identification module simultaneously and eye artefact removes module, if the section
EEG signals include eye artefact, then the eye artefact identified eye artefact identification module is sent into eye as reference signal
Artefact removes module;Otherwise, the reference signal zero setting of the sef-adapting filter in eye artefact identification module;
Last eye artefact removes module using the original EEG signals of single channel and eye artefact as input signal and reference
Signal removes artefact in such a way that input subtracts reference.
3. eye artefact automatically removes the processing method of system, feature in a kind of EEG signals according to claim 2
It is: after obtaining sample set, the specific steps are as follows:
Step 1: the single channel EEG signals of given segment being sent into signal interception module, obtain brain electricity sample;
Step 2: this section of brain electricity sample being sent into eye artefact as input signal and removes module;
Step 3: this section of brain electricity sample being sent into eye artefact identification module, is projected into the subspace extended by eye electricity sample
In, judge whether the segment brain electricity includes eye artefact, if comprising, then follow the steps 4, it is no to then follow the steps 5;
Step 4: being sent into the eye artefact come out by step 5 automatic identification as reference signal in eye artefact removal module;
Step 5: by the reference signal zero setting of eye artefact removal module, executing step 6;
Step 6: in eye artefact removal module, input signal subtracts reference signal, final to remove eye artefact, executes step
7;
Step 7: end operation or the channel for replacing EEG signals repeat step 1 to step 6.
4. eye artefact automatically removes the processing method of system in a kind of EEG signals according to claim 2 or 3, special
Sign is: the method for obtaining sample set is: the electro-ocular signal of four-way being sent into signal interception module first, obtains eye electricity
Sample set;Then eye electricity sample set training convolutional self-encoding encoder is used, its study is made to arrive the peak value and local mean value of eye artefact
Feature obtains optimized parameter collection, is stored in eye artefact identification module.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910635391.7A CN110353672B (en) | 2019-07-15 | 2019-07-15 | System and method for removing eye artifacts in electroencephalogram signals |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910635391.7A CN110353672B (en) | 2019-07-15 | 2019-07-15 | System and method for removing eye artifacts in electroencephalogram signals |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110353672A true CN110353672A (en) | 2019-10-22 |
CN110353672B CN110353672B (en) | 2022-02-22 |
Family
ID=68219222
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910635391.7A Active CN110353672B (en) | 2019-07-15 | 2019-07-15 | System and method for removing eye artifacts in electroencephalogram signals |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110353672B (en) |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110974217A (en) * | 2020-01-03 | 2020-04-10 | 苏州大学 | Dual-stage electrocardiosignal noise reduction method based on convolution self-encoder |
CN111012341A (en) * | 2020-01-08 | 2020-04-17 | 东南大学 | Artifact removal and electroencephalogram signal quality evaluation method based on wearable electroencephalogram equipment |
CN111338482A (en) * | 2020-03-04 | 2020-06-26 | 太原理工大学 | Brain-controlled character spelling recognition method and system based on supervised self-encoding |
CN111543984A (en) * | 2020-04-13 | 2020-08-18 | 重庆邮电大学 | Method for removing ocular artifacts of electroencephalogram signals based on SSDA (steady state data acquisition) |
CN111728607A (en) * | 2020-06-23 | 2020-10-02 | 北京脑陆科技有限公司 | Eye noise removing method based on brain wave signal characteristics |
CN111756352A (en) * | 2020-05-18 | 2020-10-09 | 北京大学 | Pulse array time domain filtering method, device, equipment and storage medium |
CN112220482A (en) * | 2020-09-25 | 2021-01-15 | 北京大学 | Method for detecting and eliminating magnetoencephalogram eye movement artifact based on neural network and electronic device |
CN112472107A (en) * | 2020-12-09 | 2021-03-12 | 中国科学技术大学 | Electroencephalogram artifact removing method and device |
CN113208631A (en) * | 2021-04-06 | 2021-08-06 | 北京脑陆科技有限公司 | Winking detection method and system based on EEG brain waves |
CN113253850A (en) * | 2021-07-05 | 2021-08-13 | 中国科学院西安光学精密机械研究所 | Multitask cooperative operation method based on eye movement tracking and electroencephalogram signals |
CN117257327A (en) * | 2023-10-24 | 2023-12-22 | 沈阳工业大学 | Electroencephalogram signal quality assessment method and system |
Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103258120A (en) * | 2013-04-19 | 2013-08-21 | 杭州电子科技大学 | Apoplexy recovery degree index calculation method based on brain electrical signals |
CN104427932A (en) * | 2012-01-18 | 2015-03-18 | 布赖恩斯科普公司 | Method and device for multimodal neurological evaluation |
CN104809434A (en) * | 2015-04-22 | 2015-07-29 | 哈尔滨工业大学 | Sleep staging method based on single-channel electroencephalogram signal ocular artifact removal |
CN106333682A (en) * | 2016-10-10 | 2017-01-18 | 天津大学 | Acute ischemic thalamic stroke early diagnosis method based on electroencephalogram nonlinear dynamic characteristics |
CN106491129A (en) * | 2016-10-10 | 2017-03-15 | 安徽大学 | A kind of Human bodys' response system and method based on EOG |
US9659355B1 (en) * | 2015-12-03 | 2017-05-23 | Motorola Mobility Llc | Applying corrections to regions of interest in image data |
CN108366730A (en) * | 2015-10-05 | 2018-08-03 | 塔塔咨询服务公司 | Method and system of the pretreatment for cognitive load measurement is carried out to EEG signal |
CN108836321A (en) * | 2018-05-03 | 2018-11-20 | 江苏师范大学 | A kind of EEG signals preprocess method based on adaptive noise cancel- ation system |
CN109157214A (en) * | 2018-09-11 | 2019-01-08 | 河南工业大学 | A method of the online removal eye electricity artefact suitable for single channel EEG signals |
CN109497997A (en) * | 2018-12-10 | 2019-03-22 | 杭州妞诺科技有限公司 | Based on majority according to the seizure detection and early warning system of acquisition |
CN109820503A (en) * | 2019-04-10 | 2019-05-31 | 合肥工业大学 | The synchronous minimizing technology of a variety of artefacts in single channel EEG signals |
-
2019
- 2019-07-15 CN CN201910635391.7A patent/CN110353672B/en active Active
Patent Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104427932A (en) * | 2012-01-18 | 2015-03-18 | 布赖恩斯科普公司 | Method and device for multimodal neurological evaluation |
CN103258120A (en) * | 2013-04-19 | 2013-08-21 | 杭州电子科技大学 | Apoplexy recovery degree index calculation method based on brain electrical signals |
CN104809434A (en) * | 2015-04-22 | 2015-07-29 | 哈尔滨工业大学 | Sleep staging method based on single-channel electroencephalogram signal ocular artifact removal |
CN108366730A (en) * | 2015-10-05 | 2018-08-03 | 塔塔咨询服务公司 | Method and system of the pretreatment for cognitive load measurement is carried out to EEG signal |
US9659355B1 (en) * | 2015-12-03 | 2017-05-23 | Motorola Mobility Llc | Applying corrections to regions of interest in image data |
CN106333682A (en) * | 2016-10-10 | 2017-01-18 | 天津大学 | Acute ischemic thalamic stroke early diagnosis method based on electroencephalogram nonlinear dynamic characteristics |
CN106491129A (en) * | 2016-10-10 | 2017-03-15 | 安徽大学 | A kind of Human bodys' response system and method based on EOG |
CN108836321A (en) * | 2018-05-03 | 2018-11-20 | 江苏师范大学 | A kind of EEG signals preprocess method based on adaptive noise cancel- ation system |
CN109157214A (en) * | 2018-09-11 | 2019-01-08 | 河南工业大学 | A method of the online removal eye electricity artefact suitable for single channel EEG signals |
CN109497997A (en) * | 2018-12-10 | 2019-03-22 | 杭州妞诺科技有限公司 | Based on majority according to the seizure detection and early warning system of acquisition |
CN109820503A (en) * | 2019-04-10 | 2019-05-31 | 合肥工业大学 | The synchronous minimizing technology of a variety of artefacts in single channel EEG signals |
Non-Patent Citations (1)
Title |
---|
张晨洁: "基于CNN脑电信号伪迹检测与去除的EEMD方法", 《长春理工大学学报(自然科学版)》 * |
Cited By (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110974217A (en) * | 2020-01-03 | 2020-04-10 | 苏州大学 | Dual-stage electrocardiosignal noise reduction method based on convolution self-encoder |
CN110974217B (en) * | 2020-01-03 | 2022-08-09 | 苏州大学 | Dual-stage electrocardiosignal noise reduction method based on convolution self-encoder |
CN111012341B (en) * | 2020-01-08 | 2022-04-22 | 东南大学 | Artifact removal and electroencephalogram signal quality evaluation method based on wearable electroencephalogram equipment |
CN111012341A (en) * | 2020-01-08 | 2020-04-17 | 东南大学 | Artifact removal and electroencephalogram signal quality evaluation method based on wearable electroencephalogram equipment |
CN111338482A (en) * | 2020-03-04 | 2020-06-26 | 太原理工大学 | Brain-controlled character spelling recognition method and system based on supervised self-encoding |
CN111338482B (en) * | 2020-03-04 | 2023-04-25 | 太原理工大学 | Brain-controlled character spelling recognition method and system based on supervision self-coding |
CN111543984A (en) * | 2020-04-13 | 2020-08-18 | 重庆邮电大学 | Method for removing ocular artifacts of electroencephalogram signals based on SSDA (steady state data acquisition) |
CN111543984B (en) * | 2020-04-13 | 2022-07-01 | 重庆邮电大学 | Method for removing ocular artifacts of electroencephalogram signals based on SSDA |
CN111756352A (en) * | 2020-05-18 | 2020-10-09 | 北京大学 | Pulse array time domain filtering method, device, equipment and storage medium |
CN111728607A (en) * | 2020-06-23 | 2020-10-02 | 北京脑陆科技有限公司 | Eye noise removing method based on brain wave signal characteristics |
CN112220482A (en) * | 2020-09-25 | 2021-01-15 | 北京大学 | Method for detecting and eliminating magnetoencephalogram eye movement artifact based on neural network and electronic device |
CN112472107B (en) * | 2020-12-09 | 2022-04-19 | 中国科学技术大学 | Electroencephalogram artifact removing method and device |
CN112472107A (en) * | 2020-12-09 | 2021-03-12 | 中国科学技术大学 | Electroencephalogram artifact removing method and device |
CN113208631A (en) * | 2021-04-06 | 2021-08-06 | 北京脑陆科技有限公司 | Winking detection method and system based on EEG brain waves |
CN113253850A (en) * | 2021-07-05 | 2021-08-13 | 中国科学院西安光学精密机械研究所 | Multitask cooperative operation method based on eye movement tracking and electroencephalogram signals |
CN117257327A (en) * | 2023-10-24 | 2023-12-22 | 沈阳工业大学 | Electroencephalogram signal quality assessment method and system |
Also Published As
Publication number | Publication date |
---|---|
CN110353672B (en) | 2022-02-22 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110353672A (en) | Eye artefact removal system and minimizing technology in a kind of EEG signals | |
CN110765920B (en) | Motor imagery classification method based on convolutional neural network | |
CN110353702A (en) | A kind of emotion identification method and system based on shallow-layer convolutional neural networks | |
Mammone et al. | Automatic artifact rejection from multichannel scalp EEG by wavelet ICA | |
CN102835955B (en) | Method of automatically removing ocular artifacts from electroencephalogram signal without setting threshold value | |
CN104809434B (en) | A kind of sleep stage method removed based on single channel EEG signals eye electricity artefact | |
CN112381008B (en) | Electroencephalogram emotion recognition method based on parallel sequence channel mapping network | |
CN104688220B (en) | Method for removing ocular artifacts in EEG signals | |
CN109009092A (en) | A method of removal EEG signals noise artefact | |
CN109325586B (en) | System for denoising electroencephalogram signal | |
CN109871831B (en) | Emotion recognition method and system | |
CN111000555A (en) | Training data generation method, automatic recognition model modeling method and automatic recognition method for epilepsia electroencephalogram signals | |
CN114947883A (en) | Time-frequency domain information fusion deep learning electroencephalogram noise reduction method | |
CN113842115B (en) | Improved EEG signal feature extraction method | |
CN115659207A (en) | Electroencephalogram emotion recognition method and system | |
Gao et al. | EEG reconstruction with a dual-scale CNN-LSTM model for deep artifact removal | |
CN113128353B (en) | Emotion perception method and system oriented to natural man-machine interaction | |
CN116172576A (en) | Electroencephalogram signal artifact removing method based on multi-module neural network | |
CN114757236B (en) | Electroencephalogram signal denoising optimization method and system based on TQWT and SVMD | |
CN115462803A (en) | BG-Attention-based electroencephalogram signal denoising method, device and storage medium | |
CN110516711B (en) | Training set quality evaluation method of MI-BCI system and optimization method of single training sample | |
CN116035594B (en) | Electroencephalogram artifact removing method based on segmentation-noise reduction network | |
CN112450946A (en) | Electroencephalogram artifact restoration method based on loop generation countermeasure network | |
CN114533083B (en) | Motor imagery state identification method based on multi-fusion convolutional neural network | |
Labate et al. | Remarks about wavelet analysis in the EEG artifacts detection |
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 |