CN116756643A - Fisher score-based electroencephalogram channel selection method - Google Patents
Fisher score-based electroencephalogram channel selection method Download PDFInfo
- Publication number
- CN116756643A CN116756643A CN202310729167.0A CN202310729167A CN116756643A CN 116756643 A CN116756643 A CN 116756643A CN 202310729167 A CN202310729167 A CN 202310729167A CN 116756643 A CN116756643 A CN 116756643A
- Authority
- CN
- China
- Prior art keywords
- channel
- fisher
- channels
- electroencephalogram
- subset
- 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
- 238000010187 selection method Methods 0.000 title claims abstract description 13
- 238000000034 method Methods 0.000 claims abstract description 32
- 238000005457 optimization Methods 0.000 claims abstract description 7
- 238000012360 testing method Methods 0.000 claims description 16
- 210000004556 brain Anatomy 0.000 claims description 10
- 238000001159 Fisher's combined probability test Methods 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 4
- 238000000605 extraction Methods 0.000 claims description 3
- 238000001914 filtration Methods 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 3
- 238000012163 sequencing technique Methods 0.000 claims description 3
- 230000009286 beneficial effect Effects 0.000 abstract 1
- 238000005259 measurement Methods 0.000 abstract 1
- 108091006146 Channels Proteins 0.000 description 124
- 210000004761 scalp Anatomy 0.000 description 7
- 230000000694 effects Effects 0.000 description 5
- 238000013461 design Methods 0.000 description 2
- 238000005868 electrolysis reaction Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 238000007917 intracranial administration Methods 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000007177 brain activity Effects 0.000 description 1
- 210000005013 brain tissue Anatomy 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000001054 cortical effect Effects 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 230000002232 neuromuscular Effects 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- 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
- A61B5/24—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]
- A61B5/37—Intracranial electroencephalography [IC-EEG], e.g. electrocorticography [ECoG]
-
- 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]
- A61B5/372—Analysis of electroencephalograms
- A61B5/374—Detecting the frequency distribution of signals, e.g. detecting delta, theta, alpha, beta or gamma waves
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/01—Input arrangements or combined input and output arrangements for interaction between user and computer
- G06F3/011—Arrangements for interaction with the human body, e.g. for user immersion in virtual reality
- G06F3/015—Input arrangements based on nervous system activity detection, e.g. brain waves [EEG] detection, electromyograms [EMG] detection, electrodermal response detection
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- General Engineering & Computer Science (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- General Physics & Mathematics (AREA)
- Veterinary Medicine (AREA)
- Public Health (AREA)
- Animal Behavior & Ethology (AREA)
- Surgery (AREA)
- Molecular Biology (AREA)
- Pathology (AREA)
- Biophysics (AREA)
- Psychology (AREA)
- Psychiatry (AREA)
- Neurosurgery (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Human Computer Interaction (AREA)
- Neurology (AREA)
- Dermatology (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention discloses an electroencephalogram channel selection method based on Fisher score. The invention adopts a local optimization method based on Fisher score to realize the selection of the channel; according to the invention, firstly, based on a Fisher score which is a statistical measurement index, preliminary division is carried out on the pre-selected channels, then, a local optimization method is adopted to finish the selection of the optimal channel subset, the channel selection rules and the accuracy of different wave bands are analyzed, and meanwhile, the position distribution of the selected channel subset on a sensor domain is provided. The method provided by the invention has the advantages that the classification accuracy of the brain-computer signal motor imagery tasks is obviously improved compared with that of selecting all channels, and the method is beneficial to the portability and the accurate application of brain-computer interfaces.
Description
Technical Field
The invention belongs to the technical fields of brain-computer interface technology, computer technology, biomedical engineering, artificial intelligence and brain science, and particularly relates to an electroencephalogram channel selection method based on Fisher score.
Background
The brain-computer interface (Brain computer interface, BCI) refers to a software and hardware communication system that enables information interaction between the human or animal brain and external devices without relying on conventional neuromuscular physiological systems. The brain-computer interface aims to control external equipment or a computer through brain activities and help disabled people to restore normal life as much as possible. Motor Imagery (MI) refers to Imagery that uses brain ideas to act without performing an action, but with neuronal activity in the primary sensory-Motor area. The brain-computer interface system based on the motor imagery is used for controlling external equipment (a manipulator, a trolley) and the like by collecting brain electricity of a person when the motor imagery task is executed and completing decoding.
Scalp signals acquired by the scalp electrodes have a lower spatial resolution compared to intracranial electroencephalograms (local field potentials and cortical electroencephalograms) because the number of scalp electrodes is fixed and relatively small. And the spatial expression of intracranial neuron signals on the scalp is greatly weakened and disturbed when the brain tissue transmits to different positions of the scalp under the influence of the receptor volume conduction effect. The volume conduction effect can also lead to the similarity of the brain electrical signals collected by the scalp electrodes, and further weaken the accuracy and effectiveness of scalp brain electrical decoding.
The multi-channel EEG signal can improve EEG signal resolution, while a large number of EEG channels can provide more rich EEG activity information, information redundancy and noise interference are increased, and high-dimensional data is generated. Since the amplitude of the EEG signal is in the microvolt range, it is easily disturbed by noise (artifacts), and it is therefore necessary to remove these artifacts from the EEG signal in order to preserve truly valuable information. In order to reduce redundant information in the electroencephalogram signals and obtain more real signals, reasonable electroencephalogram channel selection is necessary. The channel selection not only can reduce the complexity of the BCI system, but also can improve the decoding accuracy, and has very important significance for the portable application of the brain-computer interface.
In addition, selecting channels based on which index is also a worth discussing. Currently, some channel selection methods use numerical feature ranking to select channels. The problem with these methods is that the ordering of the numerical features does not mean a significant improvement in the classification of the electroencephalogram signals or that the numerical features preferably do not mean an optimal classification, but the accuracy is improved when a certain number of channels are selected.
Disclosure of Invention
In order to solve the problems, the invention provides a Fisher score-based local optimization electroencephalogram channel selection method. According to the method, fisher score is used as a numerical characteristic, and the influence of local optimal channel subset combination on classification accuracy is further researched on the basis of the Fisher score, so that higher classification accuracy is achieved compared with the selection of a simple numerical characteristic channel. According to the invention, the influence of the local optimal channel subset combination on the classification precision is further researched on the basis of the numerical feature sequencing, and the classification effect of the motor imagery electroencephalogram signals is improved.
The invention is realized by the following technical scheme.
An electroencephalogram channel selection method based on Fisher score applies a local optimization method to EEG channels selected by Fisher score to select channels; the method comprises the following specific steps:
(1) Filtering the original electroencephalogram signals, calculating Fisher scores of all channels in corresponding frequency bands and sequencing;
(2) And selecting the first K channels with the highest Fisher scores as the optimal channels, namely the TOP-K-Fisher channel combination.
(3) K with highest electroencephalogram information classification accuracy in TOP-K-Fisher channel combination under the Top-K-Fisher method is selected as a K value of a subsequent local optimization method;
(4) For the channel subset of TOP-K-Fisher selected in the step (3), firstly selecting one channel, generating a test channel subset of one channel, taking the test channel subset as the basis of subsequent feature extraction and classification, and marking the test channel subset as the best channel best with highest classification accuracy in the TOP-K-Fisher channel n ；
(5) Deleting the best channel best from the TOP-K-Fisher channel subset n In the current best channel best n Adding one channel in the TOP-K-Fisher channel subset rest channels, and generating a next test channel subset best n+1 Marking the optimal channel with highest classification accuracy according to the electroencephalogram information classification accuracy result of the channel, and marking the optimal channel as a local optimal channel
(6) Repeating the step (5), and gradually updating the test channel subset best for generating the newly added channels n+1 And progressively marking locally optimal channelsUp to the number of test channels n c Equal to K;
(7) And finally, selecting the channel subset with the highest classification accuracy from all the test channel subsets.
In the invention, in the step (1), the original electroencephalogram signal is filtered, and four frequency bands are divided, wherein the four frequency bands are respectively: alpha, beta, gamma, theta.
In the invention, in the step (1), the sampling rate of the original electroencephalogram signal is 250Hz.
In the present invention, in step (2), for the BCI IV IIa dataset, TOP-K (k=5, 10, 15, 20) channel combinations divided into 4 channel numbers; for the self-acquired dataset, TOP-K channel combinations (k=10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 110, 120) divided into 12 channel numbers.
In the invention, in the step (3), a classifier used for classifying the electroencephalogram signals is an SVM, and the classification task is a left-right hand motor imagery task.
Compared with the prior art, the invention has the following advantages:
1) Compared with a TOP-K-Fisher method which simply selects the channels with the highest Fisher scores as the optimal channels, the TOP-K-Fisher method which is used for selecting a series of local optimal channel subsets obtained after channel selection in the TOP-K-Fisher channel set and then selecting the subset with the highest accuracy to obtain a final channel set can further select the channel subset which is effective in decoding the motor imagery signal on the basis of the TOP-K-Fisher channels. For the four bands of alpha, beta, gamma and theta, the accuracy of the TOP-K-LOCS method is always higher than that of all channels as long as the proper K value is selected.
1. When the number of channels is small, the accuracy of the TOP-K-Fisher method cannot be higher than that of all channels. When the TOP-K-LOCS is at K=5, the average accuracy of each band exceeds the accuracy of the full channel. From another aspect it was shown that TOP-K-LOCS selected channel combinations were more efficient than TOP-K-Fisher channel subset combinations.
2. The channel selection method can reduce the number of channels to a certain extent, ensure the improvement of the precision, reduce the number of the selected channels to a certain extent, and facilitate the portability and the accurate application of the brain-computer interface system.
Drawings
FIG. 1 is a flow chart of a method of selecting locally optimized channels according to the present invention.
FIG. 2 shows the Fisher score calculation for a data set over four frequency bins.
Fig. 3 is a graph disclosing the distribution of the locations of nine tested channels selected by the TOP-K-LOCS method of dataset BCI IV IIa in the theta band.
Fig. 4 is a graph disclosing the distribution of the positions of nine tested channels selected by the TOP-K-LOCS method of dataset BCI IV IIa in the alpha band.
Fig. 5 shows the distribution of the positions of the nine tested channels selected by the TOP-K-LOCS method of the public dataset BCI IV IIa in the beta band.
Fig. 6 shows the distribution of the positions of the nine tested channels selected by the TOP-K-LOCS method of the public dataset BCI IV IIa in the gamma band.
Detailed Description
The technical scheme of the invention is clearly described below. It will be apparent that the described embodiments are only some, but not all, embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The invention provides a motor imagery brain electrolysis code method based on channel selection, wherein a flow chart is shown in fig. 1, and the method comprises the following steps:
step one: filtering the collected electroencephalogram signals, wherein a data set is disclosed by using a BCI IV IIa, the sampling rate is 250Hz, the electroencephalogram signals are divided into four frequency bands, namely theta, alpha, beta and gamma, and Fisher scores are calculated and sequenced for all channels in the four frequency bands. The calculated Fisher score is shown in FIG. 2.
Step two: and dividing the electroencephalogram channels into different channel combinations according to the Fisher score according to the result of Fisher score calculation, namely TOP-K-Fisher channels. We divided the BCI IV IIa dataset into TOP-K (k=5, 10, 15, 20) channel combinations of 4 channel numbers.
Step three: for the TOP-K-Fisher subset of channels, only one channel (best n N=1), a test subset of channels is generated as a basis for subsequent feature extraction and classification, and the best channel with highest classification accuracy among TOP-K-Fisher channels is marked. We use SVM for classification.
Step four: according to the marked best channel (best n ) In a next step a subset of the newly added channels (best n+1 ). Specifically, all of the best-tagged channels (best) are deleted from the TOP-K-Fisher channel subset n ) In the best channel and sequentially adding the remaining channels to generate a next subset of test channels (best n+1 )。
Step five: repeating the fourth step, gradually marking the local optimal channels, and updating the subset of the test channels until the number n of the test channels c Equal to K.
Step six: the best subset and corresponding accuracy are selected in TOP-K-LOCS.
In the invention, the Fisher score is calculated as follows:
where Fisher (n) is the Fisher score for the nth channel. S is S B (n) is the inter-class variance of the nth channel, S W (n) is the intra-class variance of the nth channel. S is S B The calculation formula of (n) is as follows:
wherein C is the task type number of the motor imagery electroencephalogram signal, n trails Is the total number of experiments of motor imagery, [ n ] rrails ] i Representing the experiment times of the ith motor imagery task, m i (n) Is the characteristic average value of i-type motor imagery tasks on the nth channel, m (n) Mean characteristic average value, W, of all types of motor imagery EEG on nth channel i Indicating that the current experimental task is a motor imagery task of the i type.
Tables 1 and 2 show the accuracy results of using the disclosed dataset under the TOP-K-Fisher and TOP-K-LOCS methods, respectively, the classifier uses a support vector machine, and the classification task is a left-right hand motor imagery task. From table 1 we can see that as the K value increases (k=5, 10, 15, 20), the classification accuracy of TOP-K-Fisher increases accordingly. As can be seen from table 2, in the θ band, the highest accuracy is 82.50% which is 8.28% higher than the average accuracy of the full channel 74.22%. In the alpha band, the highest accuracy is 81.60%, which is 6.21% higher than 75.39% of the full channel. In the beta band, the highest accuracy is 85.13%, which is 6.03% higher than 79.10% of the full channel. In particular, in the gamma band, the average accuracy of all channels is 72.81%, while in the TOP-K-LOCS method of K=20, the accuracy reaches 80.66%, and the improvement is 7.85%.
Table 1 discloses the accuracy of the data set TOP-K-Fisher at different K values
Table 2 discloses the accuracy of the data set TOP-K-LOCS at different K values
Unlike the TOP-K-Fisher channel selection method, TOP-K-LOCS achieves the highest accuracy at four bands k=20 and the number of channels selected is less than 20, that is, it selects fewer channels than the TOP-K-Fisher method.
The TOP-K-LOCS method may further select a subset of channels that are effective for decoding motor imagery signals based on the TOP-K-Fisher channels, as compared to the TOP-K-Fisher method. From the results of the four bands, it can be seen that the accuracy of the TOP-K-LOCS method is always higher than that of all channels, as long as the appropriate K value is selected. At the same time, this also shows that TOP-K-LOCS selected channel combinations are more efficient than TOP-K-Fisher channel subset combinations.
Fig. 3 shows the TOP-K-LOCS channel selection results for the public dataset in the theta band, we found that the average classification accuracy was 82.50% for the 9 subjects, and it can be seen from subject 5 that the TOP-K-LOCS method discarded some TOP-lying channels, including CZ, C4, C1 and C3. Also, TOP-K-LOCS gave up some top-leaf channels, including CZ, C2, C4 and CP2, for trials 6 and 7, on the basis of improved accuracy. Thus, for the θ -band, the TOP lobe region may contain less effective motor imagery information, so that when a channel is selected based on the TOP-K-LOCS method, part of the channel for that region is not selected.
Table 3 shows the number of channels selected by the TOP-K-LOCS method of the public dataset 9 tested, from which it can be seen that the average number of channels over the four bands is 10.56,8.11, 10.44 and 11.78, respectively, which have been significantly reduced compared to the original 22 electrode channels; in specific application, the electrodes can be set according to the average channel, so that the number of the electrodes in the electrode cap can be reduced, and the classification effect is better.
Table 3 discloses the number of channels selected by the data set 9 tested TOP-K-LOCS method
In summary, the invention designs a motor imagery brain electrolysis code method based on channel selection by using the design method and the data, greatly reduces the number of channels required by motor imagery brain-computer interface classification, analyzes the reasons of the discarded channels on the motor imagery, and provides possible physiological interpretability. In summary, the invention is hopeful to reduce the number of channels for collecting the electroencephalogram signal electrode in the classification of motor imagery, and provides a new direction for portability of brain-computer interface systems.
The foregoing describes a particular implementation of the invention in order to facilitate an understanding of the invention by others skilled in the art. It should be noted that the present invention is not limited to the above embodiments, and that all the inventions utilizing the inventive concept are protected as long as the variations are defined in the appended claims.
Claims (6)
1. The Fisher score-based electroencephalogram channel selection method is characterized in that a local optimization method is applied to EEG channels selected by Fisher scores to select the channels; the method comprises the following specific steps:
(1) Filtering the original electroencephalogram signals, calculating Fisher scores of all channels in corresponding frequency bands and sequencing;
(2) Dividing an electroencephalogram channel into different channel combinations according to Fisher scores according to the result of Fisher score calculation, and selecting a channel with high K before the Fisher score, namely a TOP-K-Fisher channel combination;
(3) K with highest electroencephalogram information classification accuracy in TOP-K-Fisher channel combination under the TOP-K-Fisher method is selected as a K value of a subsequent local optimization method;
(4) For the TOP-K-Fisher channel subset selected in the step (3), firstly selecting one channel, generating a test channel subset of one channel, namely selecting the electroencephalogram data of one channel as the basis of subsequent feature extraction and classification, and marking the electroencephalogram data as the best channel best with highest classification accuracy in the TOP-K-Fisher channel n ；
(5) Deleting the best channel best from the TOP-K-Fisher channel subset n In the current best channel best n Adding one channel in the TOP-K-Fisher channel subset rest channels, and generating a next test channel subset best n+1 Marking the optimal channel with highest classification accuracy according to the electroencephalogram information classification accuracy result of the channel, and marking the optimal channel as a local optimal channel
(6) Repeating the step (5), and gradually updating the test channel subset best for generating the newly added channels n+1 And progressively marking locally optimal channelsUp to the number of test channels n c Equal to K;
(7) And finally, selecting the channel combination with the highest accuracy from all the test channel subsets.
2. The electroencephalogram channel selection method according to claim 1, wherein in the step (1), after the original electroencephalogram signal is filtered, four frequency bands are divided, and the four frequency bands are respectively: θ (4-8 Hz), α (8-12 Hz), β (12-30 Hz), γ (30-45 Hz).
3. The method according to claim 1, wherein in the step (1), the sampling rate of the original brain electrical signal is 250Hz.
4. The method according to claim 1, wherein in step (2), the BCI IV IIa data set is divided into 4 channel combinations of TOP-K, k=5, 10, 15, 20 channels.
5. The electroencephalogram channel selection method according to claim 1, wherein in the step (2), for the self-acquired data set, a TOP-K channel combination divided into 12 channel numbers, k=10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 110, 120.
6. The electroencephalogram channel selection method according to claim 1, wherein the classifier used in classifying the electroencephalogram signals is an SVM, and the classification task is a left-right-hand motor imagery task.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310729167.0A CN116756643A (en) | 2023-06-20 | 2023-06-20 | Fisher score-based electroencephalogram channel selection method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310729167.0A CN116756643A (en) | 2023-06-20 | 2023-06-20 | Fisher score-based electroencephalogram channel selection method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116756643A true CN116756643A (en) | 2023-09-15 |
Family
ID=87958529
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310729167.0A Pending CN116756643A (en) | 2023-06-20 | 2023-06-20 | Fisher score-based electroencephalogram channel selection method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116756643A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117311516A (en) * | 2023-11-28 | 2023-12-29 | 北京师范大学 | Motor imagery electroencephalogram channel selection method and system |
-
2023
- 2023-06-20 CN CN202310729167.0A patent/CN116756643A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117311516A (en) * | 2023-11-28 | 2023-12-29 | 北京师范大学 | Motor imagery electroencephalogram channel selection method and system |
| CN117311516B (en) * | 2023-11-28 | 2024-02-20 | 北京师范大学 | Motor imagery electroencephalogram channel selection method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Chen et al. | Removal of muscle artifacts from the EEG: A review and recommendations | |
| CN110969108B (en) | Limb action recognition method based on autonomic motor imagery electroencephalogram | |
| Graimann et al. | Toward a direct brain interface based on human subdural recordings and wavelet-packet analysis | |
| CN114533086B (en) | Motor imagery brain electrolysis code method based on airspace characteristic time-frequency transformation | |
| Norani et al. | A review of signal processing in brain computer interface system | |
| Obayya et al. | Automatic classification of sleep stages using EEG records based on Fuzzy c-means (FCM) algorithm | |
| Ahmed | Wheelchair movement control VIA human eye blinks | |
| Shi et al. | Feature extraction of brain–computer interface electroencephalogram based on motor imagery | |
| Alamdari et al. | A review of methods and applications of brain computer interface systems | |
| Awan et al. | Effective classification of EEG signals using K-nearest neighbor algorithm | |
| CN116756643A (en) | Fisher score-based electroencephalogram channel selection method | |
| Ahmadi et al. | Brain-computer interface signal processing algorithms: A computational cost vs. accuracy analysis for wearable computers | |
| Tian et al. | Study on classification of left-right hands motor imagery EEG signals based on CNN | |
| Guo et al. | Diagnosing Parkinson’s disease using multimodal physiological signals | |
| Ahmadi et al. | Light-weight single trial EEG signal processing algorithms: computational profiling for low power design | |
| Sun et al. | Assessing features for electroencephalographic signal categorization | |
| Liu et al. | EEG feature extraction and pattern recognition during right and left hands motor imagery in brain-computer interface | |
| Trad et al. | Nonlinear and nonstationary framework for feature extraction and classification of motor imagery | |
| Bawane et al. | Classification of mental task based on EEG processing using self organising feature map | |
| Gupta et al. | A feature ranking strategy to facilitate multivariate signal classification | |
| Sun et al. | MEEG-Transformer: transformer Network based on Multi-domain EEG for emotion recognition | |
| Wang et al. | Classification of four-class motor imagery EEG data using spatial filtering | |
| Samuel et al. | Determining the optimal window parameters for accurate and reliable decoding of multiple classes of upper limb motor imagery tasks | |
| Reaj et al. | Emotion recognition using EEG-based brain computer interface | |
| Zhao et al. | Classifying ECoG signals using probabilistic neural network |
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 |