CN106821375A - A kind of EEG feature extraction method based on CSP algorithms and AR model imagination action poteutials - Google Patents
A kind of EEG feature extraction method based on CSP algorithms and AR model imagination action poteutials Download PDFInfo
- Publication number
- CN106821375A CN106821375A CN201710072658.7A CN201710072658A CN106821375A CN 106821375 A CN106821375 A CN 106821375A CN 201710072658 A CN201710072658 A CN 201710072658A CN 106821375 A CN106821375 A CN 106821375A
- Authority
- CN
- China
- Prior art keywords
- eeg
- eeg signals
- feature extraction
- poteutials
- csp
- 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
- A61B5/24—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
- A61B5/72—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
-
- 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
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2132—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on discrimination criteria, e.g. discriminant analysis
- G06F18/21322—Rendering the within-class scatter matrix non-singular
-
- 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
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2132—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on discrimination criteria, e.g. discriminant analysis
- G06F18/21322—Rendering the within-class scatter matrix non-singular
- G06F18/21324—Rendering the within-class scatter matrix non-singular involving projections, e.g. Fisherface techniques
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Molecular Biology (AREA)
- General Health & Medical Sciences (AREA)
- Veterinary Medicine (AREA)
- Public Health (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Animal Behavior & Ethology (AREA)
- Surgery (AREA)
- Data Mining & Analysis (AREA)
- Psychiatry (AREA)
- Theoretical Computer Science (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Mathematical Physics (AREA)
- Signal Processing (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physiology (AREA)
- Psychology (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention discloses a kind of EEG feature extraction method of the imagination action poteutial based on CSP algorithms and AR models, including the collection of the EEG signals for performing successively, tentatively filtering and utilization cospace pattern algorithm(CSP algorithms)With autoregression model(AR models)Being combined carries out characteristic extraction step.The present invention improves the accuracy of EEG feature extraction, has opened up application field, possesses wide prospect.
Description
Technical field
The present invention relates to BCI fields, and in particular to a kind of to imagine right-hand man's action potential based on CSP algorithms and AR models
EEG feature extraction method.
Background technology
EEG signals are highly important bioelectrical signals in human body, are the essence embodiments of brain complex work.Brain telecommunications
The many physiology of human body, psychographic information are contained in number, which includes the thought of people.In the research of EEG signals, gradually
Form another gate technique --- brain-computer interface technology(BCI).Brain-computer interface technology is referred to by brain and computer etc.
Passage is set up between electronic equipment, people can directly pass through the technology of the motion of the command operating or control machine of brain.Wherein
Mostly important is exactly the feature extraction of EEG signals, and it directly determines the accuracy that EEG signals are understood.The present invention is devised
It is a kind of to be based on two kinds of EEG feature extraction methods of algorithm, the accuracy of EEG feature extraction can be improved, open up it
The prospect of application aspect.
In terms of feature extraction is carried out, at present with the presence of many algorithms, such as wavelet transformation, autoregression algorithm, cospace mould
Formula scheduling algorithm, wavelet transformation is a kind of temporal analysis, there is good resolution ratio in terms of time domain;Autoregression algorithm is by width
The EEG signals that degree is changed over time are changed into power spectrum chart, so as to extract the frequency domain character of EEG signals;Cospace pattern algorithm
The spatial feature of EEG signals is extracted using the estimate covariance matrix of two class EEG signals.
How three of the above method respectively a little, synthesis and improves autoregression algorithm and cospace pattern algorithm, prominent brain
The frequency domain and spatial feature of electric signal improve the feature extraction accuracy of EEG signals to reach, and just turn into skill urgently to be resolved hurrily
Art problem.
The content of the invention
Goal of the invention:In order to overcome the deficiencies in the prior art, the present invention is in existing EEG feature extraction
In method, with reference to and improve two kinds of common data processing methods, i.e. cospace pattern algorithm and autoregression model algorithm, with
The effect for improving identification EEG signals accuracy rate is reached, there is provided herein one kind based on CSP algorithms and AR models imagination right-hand man
The EEG feature extraction method of action potential.
Technical scheme:In order to solve the above technical problems, the present invention combines cospace pattern algorithm(CSP)With autoregression mould
Type algorithm(AR), the EEG signals frequency domain and spatial feature for characterizing imagination right-hand man's action potential are extracted, specifically include following
Step:
Step one:EEG signals to imagining right-hand man's action potential are acquired.
Step 2:EEG signals to gathering tentatively are filtered.
Step 3:Use cospace pattern(CSP algorithms)And autoregression model(AR models)To filtered brain line signal
Processed.
Further, in step one, electroencephalogramsignal signal acquisition module is brain wave acquisition headgear, using Emotiv EPOC+ 14
Conductive polar cap is with the frequency collection EEG signal of 128Hz.Electroencephalogramsignal signal acquisition module is used to gather subject in Mental imagery or so
The EEG signals in brain frontal cortex region, middle section and top region under hands movement, and by data transmission module transmit to
The preliminary filtering part of EEG signals;
Further, in step 2, the preliminary filtering using wavelet analysis tool box in MATLAB to EEG signals, for filtering off
The clutter that the work ripple and surrounding environment of EEG signals bring, and the waveform unrelated with imagination right-hand man's action potential.
Further, in step 3, EEG signals after filtering are carried out with CSP algorithm filtering, then initial signal is entered
Row AR models carry out computing, and two for obtaining matrix is integrated, as the characteristic vector of gained.
Beneficial effect:The present invention improves the accuracy of EEG feature extraction.Compared to traditional CSP feature extractions
Method, the present invention is not only researched and analysed in terms of the Spatial characteristic of signal, is more carried out in terms of the poor and characteristic of signal
Treatment.Set about in terms of two different, significantly improve the accuracy rate of signal characteristic abstraction, opened up application field, have
Standby wide prospect.
Except invention described above solve technical problem, constitute technical scheme technical characteristic and by these skills
Outside the advantage that the technical characteristic of art scheme is brought, included in other technologies problem, technical scheme that the present invention can be solved
The advantage that other technical characteristics and these technical characteristics bring, will be described in more detail with reference to accompanying drawing.
Brief description of the drawings
Fig. 1 is brain wave acquisition card internal structure block diagram;
Fig. 2 is implementing procedure figure of the present invention.
Specific embodiment
Below in conjunction with accompanying drawing, it is preferable to carry out being described in detail to of the invention.
In the method, it is manual to imagination left and right with the frequency of 128Hz first by the conductive polar caps of Emotiv EPOC+ 14
The EEG signals for making current potential are acquired, and the EEG signals for collecting are carried out using wavelet analysis tool box in MATLAB then
Filtering, filters off the clutter that the 50Hz works ripple and environment of EEG signals bring, and filters off as far as possible and imagination right-hand man's action potential
Unrelated waveform, then EEG signals after filtering are carried out with CSP algorithm filtering, then AR models are carried out to initial signal transport
Calculate, two matrixes that will be obtained are integrated, resulting characteristic vector.
Comprise the following steps that:
Step one:EEG signals to imagining right-hand man's action potential are acquired.
As shown in figure 1, electroencephalogramsignal signal acquisition module is specially brain wave acquisition headgear, it is conductive using Emotiv EPOC+ 14
Polar cap carries wireless module and is connected with computer with the frequency collection EEG signal of 128Hz, by brain wave acquisition headgear;Brain telecommunications
The brain in brain frontal cortex region, middle section and top region of number acquisition module collection subject under different motion Imaginary Movement
Electric signal.Brain wave acquisition card is provided with brain wave acquisition headgear, brain wave acquisition card includes acquisition electrode, reference electrode, brain wave acquisition
Chip and carry wireless launcher, acquisition electrode, reference electrode and carry wireless launcher and be connected with brain wave acquisition chip.Brain
Electric acquisition chip uses the EMOTIV EPOC+ headgears of EMOTIV companies of the U.S., 14 acquisition electrodes and 2 reference electrodes, collection
The original E.E.G value of electrode collection user simultaneously transmits it to brain wave acquisition chip, the original E.E.G value of brain wave acquisition chip treatment
EEG EEG signals after being amplified, EEG EEG signals are Scalp Potential at each electrode, and will be made up of EEG signals
Packet is passed on computer in real time by carrying wireless transport module.
Step 2:EEG signals to gathering tentatively are filtered.
The EEG signals for collecting are filtered using wavelet analysis tool box in MATLAB, filter off EEG signals
The clutter that 50Hz works ripple and environment bring, filters off the waveform unrelated with imagination right-hand man's action potential as far as possible.
Step 3:Use cospace pattern(CSP algorithms)And autoregression model(AR models)To filtered brain line signal
Processed.
As shown in Fig. 2 to the signal after filtering(Call filtering signal in the following text)CSP algorithms are first carried out, is comprised the following steps that:
(1)The data set that the data set life that imagination left hand is moved is moved with the imagination right hand is sorted out respectively, and obtaining two dimensions is
Matrix, be respectively designated as M, N.(Sample is sampled point number, and channel is port number(Channel is in the present invention
14), trail is sampling number)
(2)Calculate two groups of average covariance matrices of data , and ask them and C=。
(3)Using PCA, whitening matrix P is obtained.,
(4)Construction right-hand man's spatial filter.。
WhereinIt is characterized valueThe corresponding characteristic vector of maximum,It is characterized valueMaximum it is corresponding
Characteristic vector.
(5)The selection of characteristic vector f,,.Final
To oneMatrix.
AR model algorithm treatment is carried out to filtering signal, is comprised the following steps that:
(1)The data set that the data set that imagination left hand is moved is moved with the imagination right hand is sorted out respectively, obtains two dimensions
Matrix, is respectively designated as P, Q.(Sample is sampled point number, and channel is port number(Channel is 14 in the present invention),
Trail is sampling number)
(2)It is determined that using the passage of AR models, what the present invention was applied is EMOTIV EPOC+ headgears, right in 14 acquisition electrodes
The FC5 for answering and FC6 passages.
(3)Confirm the optimal factor n of AR models, optimal n is obtained by test of many times.
(4)With optimal exponent number n, to FC5 passages, each trail uses AR model treatments, finally gives a data,
With optimal exponent number n, to FC5 passages, each trail uses AR model treatments again, a data is finally given, by this two groups of numbers
According to being integrated, the matrix of is obtained, i.e. for the characteristic vector that AR models are obtained.
It is comprehensive、The matrix f, f that one can be obtained are the characteristic vector obtained by the present invention.
Embodiments of the present invention are described in detail above in association with accompanying drawing, but the present invention is not limited to described reality
Apply mode.For one of ordinary skill in the art, in the range of principle of the invention and technological thought, to these implementations
Mode carries out various changes, modification, replacement and deformation and still falls within protection scope of the present invention to enclose.
Claims (4)
1. a kind of EEG feature extraction method based on CSP algorithms and AR model imagination action poteutials, it is characterised in that bag
Include following steps:
Step one, the EEG signals of collection imagination right-hand man's action potential;
Step 2, the EEG signals to gathering tentatively are filtered;
Step 3, is processed preliminary filtered brain line signal using cospace pattern and autoregression model.
2. the EEG feature extraction side based on CSP algorithms and AR model imagination action poteutials according to claim 1
Method, it is characterised in that:
In the step one, using brain volume of the electroencephalogramsignal signal acquisition module collection subject under Mental imagery or so hands movement
The EEG signals in leaf region, middle section and top region, and transmitted to EEG signals by data transmission module and tentatively filter
Part;
In the step 2, tentatively filter for filtering off the clutter that the work ripple and surrounding environment of EEG signals bring, Yi Jiyu
The unrelated waveform of imagination right-hand man's action potential;
In the step 3, CSP algorithm filtering is carried out to the EEG signals by preliminary filtering, then AR moulds are carried out to initial signal
Type carries out computing, and two for obtaining matrix is integrated into the characteristic vector of gained.
3. the EEG feature extraction side based on CSP algorithms and AR model imagination action poteutials according to claim 2
Method, it is characterised in that:The electroencephalogramsignal signal acquisition module is brain wave acquisition headgear, using the conductive polar caps of Emotiv EPOC+ 14
With the frequency collection EEG signal of 128Hz.
4. the EEG feature extraction side based on CSP algorithms and AR model imagination action poteutials according to claim 3
Method, it is characterised in that:When carrying out AR models to initial signal and carrying out computing, in only taking 14 crosslinking electrodes of brain wave acquisition headgear
Data in FC5 and FC6.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710072658.7A CN106821375A (en) | 2017-02-10 | 2017-02-10 | A kind of EEG feature extraction method based on CSP algorithms and AR model imagination action poteutials |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710072658.7A CN106821375A (en) | 2017-02-10 | 2017-02-10 | A kind of EEG feature extraction method based on CSP algorithms and AR model imagination action poteutials |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106821375A true CN106821375A (en) | 2017-06-13 |
Family
ID=59122350
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710072658.7A Pending CN106821375A (en) | 2017-02-10 | 2017-02-10 | A kind of EEG feature extraction method based on CSP algorithms and AR model imagination action poteutials |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106821375A (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108904980A (en) * | 2018-08-01 | 2018-11-30 | 国家康复辅具研究中心 | Upper limb initiative rehabilitation method and device based on brain electricity and functional electrostimulation |
CN109299647A (en) * | 2018-07-24 | 2019-02-01 | 东南大学 | A kind of extraction of multitask Mental imagery brain electrical feature and mode identification method towards vehicle control |
CN109657646A (en) * | 2019-01-07 | 2019-04-19 | 哈尔滨工业大学(深圳) | The character representation and extracting method, device and storage medium of physiological time sequence |
CN111110230A (en) * | 2020-01-09 | 2020-05-08 | 燕山大学 | Motor imagery electroencephalogram feature enhancement method and system |
CN112826451A (en) * | 2021-03-05 | 2021-05-25 | 中山大学 | Anesthesia depth and sleep depth assessment method and device |
CN113197585A (en) * | 2021-04-01 | 2021-08-03 | 燕山大学 | Neuromuscular information interaction model construction and parameter identification optimization method |
CN118319330A (en) * | 2024-06-14 | 2024-07-12 | 东北电力大学 | Brain wave analysis method for diagnosing depression |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103690160A (en) * | 2013-11-18 | 2014-04-02 | 浙江大学 | Electroencephalogram feature extraction method based on non-Gaussian time sequence model |
CN105563495A (en) * | 2016-02-01 | 2016-05-11 | 浙江大学 | Mechanical arm system controlled on basis of refined motor imagination electroencephalogram signals and method |
-
2017
- 2017-02-10 CN CN201710072658.7A patent/CN106821375A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103690160A (en) * | 2013-11-18 | 2014-04-02 | 浙江大学 | Electroencephalogram feature extraction method based on non-Gaussian time sequence model |
CN105563495A (en) * | 2016-02-01 | 2016-05-11 | 浙江大学 | Mechanical arm system controlled on basis of refined motor imagination electroencephalogram signals and method |
Non-Patent Citations (1)
Title |
---|
郭凯: "基于运动想象的脑机接口的研究及其应用", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109299647A (en) * | 2018-07-24 | 2019-02-01 | 东南大学 | A kind of extraction of multitask Mental imagery brain electrical feature and mode identification method towards vehicle control |
CN109299647B (en) * | 2018-07-24 | 2022-02-11 | 东南大学 | Vehicle control-oriented multitask motor imagery electroencephalogram feature extraction and mode recognition method |
CN108904980A (en) * | 2018-08-01 | 2018-11-30 | 国家康复辅具研究中心 | Upper limb initiative rehabilitation method and device based on brain electricity and functional electrostimulation |
CN109657646A (en) * | 2019-01-07 | 2019-04-19 | 哈尔滨工业大学(深圳) | The character representation and extracting method, device and storage medium of physiological time sequence |
CN109657646B (en) * | 2019-01-07 | 2023-04-07 | 哈尔滨工业大学(深圳) | Method and device for representing and extracting features of physiological time series and storage medium |
CN111110230A (en) * | 2020-01-09 | 2020-05-08 | 燕山大学 | Motor imagery electroencephalogram feature enhancement method and system |
CN112826451A (en) * | 2021-03-05 | 2021-05-25 | 中山大学 | Anesthesia depth and sleep depth assessment method and device |
CN113197585A (en) * | 2021-04-01 | 2021-08-03 | 燕山大学 | Neuromuscular information interaction model construction and parameter identification optimization method |
CN113197585B (en) * | 2021-04-01 | 2022-02-18 | 燕山大学 | Neuromuscular information interaction model construction and parameter identification optimization method |
CN118319330A (en) * | 2024-06-14 | 2024-07-12 | 东北电力大学 | Brain wave analysis method for diagnosing depression |
CN118319330B (en) * | 2024-06-14 | 2024-08-06 | 东北电力大学 | Brain wave analysis method for diagnosing depression |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106821375A (en) | A kind of EEG feature extraction method based on CSP algorithms and AR model imagination action poteutials | |
CN103793058B (en) | A kind of active brain-computer interactive system Mental imagery classification of task method and device | |
CN104548347A (en) | Pure idea nerve muscle electrical stimulation control and nerve function evaluation system | |
WO2020186651A1 (en) | Smart sports earphones based on eeg thoughts and implementation method and system thereof | |
CN106963372B (en) | A kind of brain electricity-electromyography signal fusing device and fusion method | |
CN104997581B (en) | Artificial hand control method and apparatus for driving EEG signals on the basis of facial expressions | |
CN106383579A (en) | EMG and FSR-based refined gesture recognition system and method | |
CN103955269A (en) | Intelligent glass brain-computer interface method based on virtual real environment | |
WO2020042511A1 (en) | Motion potential brain-machine interface encoding and decoding method based on spatial filtering and template matching | |
CN105022488A (en) | Wireless BCI (Brain Computer Interface) input system based on SSVEP (Steady-State Visual Evoked Potentials) brain electric potential | |
CN113274032A (en) | Cerebral apoplexy rehabilitation training system and method based on SSVEP + MI brain-computer interface | |
Zhao et al. | Deep CNN model based on serial-parallel structure optimization for four-class motor imagery EEG classification | |
CN105286860A (en) | Motor imagery brain electrical signal recognition method based on dual-tree complex wavelet energy difference | |
CN112732090B (en) | Muscle cooperation-based user-independent real-time gesture recognition method | |
CN105962935A (en) | Brain electrical nerve feedback training system and method for improving motor learning function | |
CN111544855A (en) | Pure idea control intelligent rehabilitation method based on distillation learning and deep learning and application | |
CN102306303A (en) | Electroencephalography signal characteristic extraction method based on small training samples | |
CN105708587A (en) | Lower-limb exoskeleton training method and system triggered by brain-computer interface under motion imagination pattern | |
CN106983509A (en) | A kind of portable brain electric acquisition system with real-time mark function | |
CN110074779A (en) | A kind of EEG signal identification method and device | |
CN106909226A (en) | A kind of polymorphic brain machine interface system | |
CN109858537A (en) | EEG feature extraction method of the improved EEMD in conjunction with CSP | |
CN103294192A (en) | LED lamp switch control device and control method thereof based on motor imagery | |
CN109299647A (en) | A kind of extraction of multitask Mental imagery brain electrical feature and mode identification method towards vehicle control | |
CN108364062A (en) | Deep learning model building method based on MEMD and the application in Mental imagery |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170613 |
|
RJ01 | Rejection of invention patent application after publication |