CN104515905B - The EEG signals adaptive spectrum analysis method of subject based on CQT multiresolution - Google Patents
The EEG signals adaptive spectrum analysis method of subject based on CQT multiresolution Download PDFInfo
- Publication number
- CN104515905B CN104515905B CN201310450516.1A CN201310450516A CN104515905B CN 104515905 B CN104515905 B CN 104515905B CN 201310450516 A CN201310450516 A CN 201310450516A CN 104515905 B CN104515905 B CN 104515905B
- Authority
- CN
- China
- Prior art keywords
- frequency
- cqt
- eeg signal
- eeg
- multiresolution
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 32
- 238000010183 spectrum analysis Methods 0.000 title claims abstract description 19
- 230000003044 adaptive effect Effects 0.000 title claims abstract description 11
- 238000001228 spectrum Methods 0.000 claims abstract description 18
- 238000004364 calculation method Methods 0.000 claims abstract description 15
- 210000001595 mastoid Anatomy 0.000 claims description 11
- 230000003595 spectral effect Effects 0.000 claims description 7
- 230000002146 bilateral effect Effects 0.000 claims description 3
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 210000003205 muscle Anatomy 0.000 claims description 3
- 238000004458 analytical method Methods 0.000 abstract description 17
- 210000004556 brain Anatomy 0.000 abstract description 16
- 239000000284 extract Substances 0.000 abstract description 7
- 238000000605 extraction Methods 0.000 abstract 1
- 230000009466 transformation Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 230000007812 deficiency Effects 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 230000019771 cognition Effects 0.000 description 2
- 238000003672 processing method Methods 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 230000001149 cognitive effect Effects 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 230000006698 induction Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 210000000337 motor cortex Anatomy 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
Landscapes
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention discloses a kind of EEG signals adaptive spectrum analysis methods of subject based on CQT multiresolution, successively the difference by being pre-processed to raw EEG signal, automatically finding subject by the harmonic components and fine structure characteristic of EEG signal, the analysis based on CQT multiresolution and each frequency band samples bandwidth of calculating.The present invention can according to original brain wave signal it is adaptive find subject difference characteristic, can more accurately extract the frequency spectrum specific characteristics of EEG signals；Frequency spectrum analysis method based on multiresolution, it is contemplated that the other influence of EEG signals frequency band length difference improves practicability；Compared with classical CQT frequency spectrum analysis method, calculation times are reduced, efficiency significantly improves；It is then much more flexible using variable-resolution in the present invention, improve the accuracy and speed of feature extraction.
Description
Technical field
The present invention relates to field brain-computer interface field and brain wave signal analysis methods, especially a kind of to be based on more points of CQT
The EEG signals adaptive spectrum analysis method of the subject of resolution.
Background technique
With going deep into for brain-computer interface technical research, more and more electroencephalogramsignal signal analysis methods are emerged.At this stage
Brain-computer interface although have been realized in the control function of a part, but this control still more original primary control, also
There are many problems to have to be solved.The target of brain machine interface system is to realize direct, natural control model by brain electricity.For up to
To this target, it is important to collect the feature EEG signals mode that can reflect that human thinking is intended to, and pass through data processing mould
Block realizes the decoding of thinking to control.However, due to brain wave signal low signal-to-noise ratio and the random feature of non-stationary, even for same
One subject, the brain wave signal of induction are also likely to be present certain difference.For the difference problem between subject, row is found
Efficient adaptive brain wave signal difference identification technology have great importance.
The processing method of EEG signals is mainly based on time-domain analysis and frequency-domain analysis at present.Due to the timing of EEG signals
Property, existing EEG signals mostly use Time Domain Analysis, such as zero passage point analysis, histogram analysis, variance analysis, peak value inspection
The methods of survey.Wave character, intuitive and explicit physical meaning are mainly directly extracted in time-domain analysis.But point analysis knot in time domain
Fruit tends to rely on the distribution of sampled point, and EEG signal sampling is more dispersed, cannot often distinguish effect well.Cause
This just has many scholars that time domain EEG signal is transformed into frequency domain, extracts its frequency domain character to be analyzed, be identified.Such as: Barry
Eye closing (Eyes Closed, EC) is had found using EEG spectrum analysis technique respectively with Chen and opens eyes (Eyes Open, EO)
It is different.In engineering, electroencephalogram spectrum signature is extracted using frequency Power estimation, distinguishes different feelings, movement or cognitive activities,
It realizes brain-computer interface (brain-computer interface, BCI).Famous Wadsworth BCI is exactly to pass through calculating sense
It is mobile to control cursor to feel mu the and beta frequency band feature of the EEG of motor cortex.It is using more in EEG analysis at present
Autoregressive (AR) model spectra estimation technology.This method wants the linear of signal processed, stationarity and signal-to-noise ratio
Ask higher, thus be not suitable for it is long when EEG data be analyzed and processed.To improve spectrum estimation performance, Bartlett and Welch
The power spectrum Nonparametric Estimation based on Fourier transformation is proposed respectively.Its by overall length be N it is long when data be divided into M
Section, every segment length are L, are averaging after calculating separately each section of power spectral density.The method of Welch is changed on this basis
Into data segment allows to be overlapped, and uses adding window to each segment data, plays smoothing effect to former power spectrum.
But existing spectrum analysis is that a series of equally spaced frequencies are calculated by uniform sampling at equal intervals
Method.However, the band separation of EEG signals be it is unequal, the frequency band division methods being often used such as us are defined as follows: δ
(0.5~4Hz), θ (4~8Hz), α (8~13Hz), β (13~20Hz) and γ (30~50Hz).If between our uses etc.
Frequency band energy is calculated every method, γ frequency band is since span is larger, and frequency range is considerably beyond δ and θ frequency band.It is this to draw at equal intervals
Frequency dividing spectral analysis method has ignored frequency band length difference, certainly will will cause certain error.In addition, using EEG signals as each of representative
Kind cognition signal all has considerably complicated harmonic components and fine structure, various ambient noises and different types of cognition letter
Number its harmonic components and fine structure are different but have certain regularity, and different people same class signal also has it
People's feature.Current frequency spectrum analysis method does not all account for the difference problem between subject.
Summary of the invention
It is above-mentioned based on poor between the deficiency and subject that divide frequency spectrum analysis method at equal intervals the purpose of the present invention is being directed to
Different big problem proposes a kind of EEG signals adaptive spectrum analysis method of subject based on CQT multiresolution, overcomes single
The deficiency of resolution analytical procedures can efficiently extract the frequency spectrum specific characteristics of brain wave signal.
In order to achieve the above objectives, the present invention is implemented according to following technical scheme:
A kind of EEG signals adaptive spectrum analysis method of the subject based on CQT multiresolution, includes the following steps:
1) raw EEG signal is pre-processed:
A. bilateral mastoid process is carried out as the reference of reference electrode to EEG signal and is turned as reference electrode using unilateral mastoid process
It changes, it is assumed that left mastoid process is M1 electrode, and right mastoid process is M2 electrode, and conversion formula is as follows:
B. every one-dimensional EEG signal is segmented, segment length 2s, without overlapping between each section, by any one in every section
The numerical value of point is greater than 80 μ v, and whole section is given up, and excludes the interference of blink artefact and Muscle artifacts；
2) difference of subject is automatically found by the harmonic components of EEG signal and fine structure characteristic:
A. assume initially that subject EEG signal isIt ties up (m is number of poles, and n is number of samples), from signalIn with
Machine takes out one-dimensional, and taking out a length of L one sectionCalculate average power spectra PlWith frequency sequence fl, calculation formula is as follows:
Wherein frequency sequence
B. using spline method in flOn the basis of improve frequency resolution, and EEG signal is divided into 5 sections, δ (0.5
~4Hz), θ (4~8Hz), α (8~13Hz), β (13~20Hz) and γ (30~50Hz)；
The position of frequency is as subject difference characteristic when c. finding out each band frequency Energy maximum value；
3) it analyzes and calculates each frequency band samples bandwidth:
A. estimate that centre frequency, calculation formula are as follows based on CQT:
N in formulakIt is to calculate kth frequency fkCQ transformation when
Corresponding window length, wNk(n) be length be NkWindow function, Q be CQ transformation in invariant, k be sequence C Q spectrum frequency
Rate subscript；
B. interval (bandwidth) B of adjacent spectral line is calculatedk, calculation formula Bk=fk+1-fk；
C. each frequency band samples bandwidth N is calculatedk, calculation formula Nk=Fs/Bk。
Compared with prior art, the invention has the benefit that
1. can according to original brain wave signal it is adaptive find subject difference characteristic, can more accurately extract brain telecommunications
Number frequency spectrum specific characteristics；
2. the frequency spectrum analysis method based on multiresolution, it is contemplated that the other influence of EEG signals frequency band length difference improves
Practicability.Traditional digital signal processing method generallys use single resolution spectrum analysis method such as Fourier transform, due to brain
Electric signal frequency band unequal interval divides feature, by some features of processing meeting lossing signal, so that error rate and false recognition rate
It is very high；
3. reducing calculation times, efficiency significantly improves compared with classical CQT frequency spectrum analysis method.The classical side CQT
The number of low frequency range Frequency point is a much larger than high-frequency region Frequency point by exponential distribution rule, therefore in calculating process for method
Number, cannot show the feature of brain wave signal well.It is then much more flexible using variable-resolution in the present invention, it improves feature and mentions
The accuracy and speed taken.
Detailed description of the invention
Fig. 1 is flow chart of the invention；
Fig. 2 is present invention EEG signal schematic diagram after pretreatment；
Fig. 3 is that frequency spectrum analysis method extracts brain wave signal spectrum signature schematic diagram at equal intervals using single resolution ratio；
Fig. 4 is that tradition CQT analysis method extracts brain wave signal spectrum signature schematic diagram；
Fig. 5 is the brain wave signal spectrum signature schematic diagram that analysis method of the invention is extracted.
Specific embodiment
With reference to the accompanying drawing and specific embodiment the invention will be further described, in the illustrative examples of the invention
And explanation is used to explain the present invention, but not as a limitation of the invention.
The EEG signals adaptive spectrum analysis method of subject as shown in Figure 1 and Figure 2 based on CQT multiresolution, including
Following step:
1) raw EEG signal is pre-processed:
A. bilateral mastoid process is carried out as the reference of reference electrode to EEG signal and is turned as reference electrode using unilateral mastoid process
It changes, it is assumed that left mastoid process is M1 electrode, and right mastoid process is M2 electrode, and conversion formula is as follows:
B. every one-dimensional EEG signal is segmented, segment length 2s, without overlapping between each section, by any one in every section
The numerical value of point is greater than 80 μ v, and whole section is given up, and excludes the interference of blink artefact and Muscle artifacts；
2) difference of subject is automatically found by the harmonic components of EEG signal and fine structure characteristic:
A. assume initially that subject EEG signal isIt ties up (m is number of poles, and n is number of samples), from signalIn with
Machine takes out one-dimensional, and taking out a length of L one sectionCalculate average power spectra PlWith frequency sequence fl, calculation formula is as follows:
Wherein frequency sequence
B. using spline method in flOn the basis of improve frequency resolution, and EEG signal is divided into 5 sections, δ (0.5
~4Hz), θ (4~8Hz), α (8~13Hz), β (13~20Hz) and γ (30~50Hz)；
The position of frequency is as subject difference characteristic when c. finding out each band frequency Energy maximum value；
3) it analyzes and calculates each frequency band samples bandwidth:
A. estimate that centre frequency, calculation formula are as follows based on CQT:
N in formulakIt is to calculate kth frequency fkCQ transformation when
Corresponding window length, wNk(n) be length be NkWindow function, Q be CQ transformation in invariant, k be sequence C Q spectrum frequency
Rate subscript；
B. interval (bandwidth) B of adjacent spectral line is calculatedk, calculation formula Bk=fk+1-fk；
C. each frequency band samples bandwidth N is calculatedk, calculation formula Nk=Fs/Bk。
It can be seen that the analysis side with analysis method effect resolution ratio single ratio FFT of the invention in conjunction with Fig. 3, Fig. 4, Fig. 5
Method obtains spectral line amplitude close to constant amplitude, and sidelobe magnitudes are smaller.Technical solutions according to the invention can be believed in original E.E.G
The difference characteristic that the EEG signals being tested in signal are found in number has still maintained more points of CQT while reducing calculation times
The characteristics of resolution, obtained spectral line amplitude is close to constant amplitude.Technical solutions according to the invention overcome single resolution ratio well
Extract spectrum signature deficiency, reduce calculation times simultaneously can also well brain wave signal feature.
The limitation that technical solution of the present invention is not limited to the above specific embodiments, it is all to do according to the technique and scheme of the present invention
Technology deformation out, falls within the scope of protection of the present invention.
Claims (1)
1. a kind of EEG signals adaptive spectrum analysis method of subject based on CQT multiresolution, it is characterised in that: including under
State step:
1) EEG signal is pre-processed:
A. bilateral mastoid process is carried out as the reference of reference electrode to EEG signal and is converted as reference electrode using unilateral mastoid process, it is false
If left mastoid process is M1 electrode, conversion formula is as follows:
Wherein Xn (m)It is EEG signal, m is number of poles, and n is number of samples；
B. it takes and is segmented per one-dimensional EEG signal, segment length 2s, without overlapping between each section, if occurred in certain section any one
A numerical value is greater than the point of 80 μ v, then gives up whole section, excludes the interference of blink artefact and Muscle artifacts；
2) subject difference characteristic is automatically found by the harmonic components of EEG signal and fine structure characteristic:
A. from EEG signal X '(m) nIn take one section of certain one-dimensional, a length of L at random, calculate average power spectra PlWith frequency sequence fl, meter
It is as follows to calculate formula:
Wherein, centre frequency corresponding to the l articles spectral line isL=0,1..., N-1；
B. using spline method in frequency sequence flOn the basis of improve frequency resolution, and above-mentioned EEG signal is divided into 5
Section: δ (0.5~4Hz), θ (4~8Hz), α (8~13Hz), β (13~20Hz) and γ (30~50Hz)；
The position of frequency is as subject difference characteristic when c. finding out each band frequency Energy maximum value；
3) analyze and calculate each frequency band samples bandwidth；
A. estimate that centre frequency, calculation formula are as follows based on CQT:
N in formulakIt is to calculate kth frequency fkCQT when corresponding window length,Be length be NkWindow function, Q is
Invariant in CQT, k are the frequency index of sequence C Q spectrum；
B. the interval B of adjacent spectral line is calculatedk, calculation formula Bk=fk+1-fk；
C. each frequency band samples bandwidth N is calculatedk, calculation formula Nk=Fs/Bk。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310450516.1A CN104515905B (en) | 2013-09-29 | 2013-09-29 | The EEG signals adaptive spectrum analysis method of subject based on CQT multiresolution |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310450516.1A CN104515905B (en) | 2013-09-29 | 2013-09-29 | The EEG signals adaptive spectrum analysis method of subject based on CQT multiresolution |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104515905A CN104515905A (en) | 2015-04-15 |
| CN104515905B true CN104515905B (en) | 2019-09-10 |
Family
ID=52791514
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310450516.1A Active CN104515905B (en) | 2013-09-29 | 2013-09-29 | The EEG signals adaptive spectrum analysis method of subject based on CQT multiresolution |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104515905B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105067129A (en) * | 2015-08-26 | 2015-11-18 | 云南师范大学 | Power-spectrum-based detection method for analyzing light variability period of quasi-stellar object |
| CN106618561A (en) * | 2016-12-27 | 2017-05-10 | 深圳职业技术学院 | High-robustness analysis method for electroencephalogram (ECG) signal of disabled people in motion state |
| CN106933247B (en) * | 2017-03-30 | 2020-05-29 | 歌尔科技有限公司 | Unmanned aerial vehicle control method, device and system |
| CN107463792B (en) * | 2017-09-21 | 2023-11-21 | 北京大智商医疗器械有限公司 | Nerve feedback device, system and method |
| CN110327043B (en) * | 2019-08-05 | 2022-04-01 | 哈尔滨工业大学 | Event-related potential waveform map solving method based on sparse modeling |
| CN114252698A (en) * | 2020-09-24 | 2022-03-29 | 广州慧睿思通科技股份有限公司 | Frequency spectrum processing method, system, electronic equipment and storage medium |
| CN113017651B (en) * | 2021-03-16 | 2022-06-21 | 哈尔滨工业大学 | Brain function network analysis method for emotion EEG |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4941477A (en) * | 1987-09-09 | 1990-07-17 | University Patents, Inc. | Method and apparatus for detection of deception |
| US6574573B1 (en) * | 1998-10-27 | 2003-06-03 | Gram Corporation | Spectrum analysis and display method time series |
| CN101352337A (en) * | 2007-12-25 | 2009-01-28 | 天津大学 | Method for capturing signals and extracting characteristics of stand imagination action brain wave |
| CN100506147C (en) * | 2001-06-13 | 2009-07-01 | 康普麦迪克斯有限公司 | Method and apparatus for monitoring consciousness |
| CN102697493A (en) * | 2012-05-03 | 2012-10-03 | 北京工业大学 | Method for rapidly and automatically identifying and removing ocular artifacts in electroencephalogram signal |
| CN102973264A (en) * | 2012-12-07 | 2013-03-20 | 哈尔滨工业大学深圳研究生院 | Electrocardiosignal preprocessing method based on morphological multiresolution decomposition |
-
2013
- 2013-09-29 CN CN201310450516.1A patent/CN104515905B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4941477A (en) * | 1987-09-09 | 1990-07-17 | University Patents, Inc. | Method and apparatus for detection of deception |
| US6574573B1 (en) * | 1998-10-27 | 2003-06-03 | Gram Corporation | Spectrum analysis and display method time series |
| CN100506147C (en) * | 2001-06-13 | 2009-07-01 | 康普麦迪克斯有限公司 | Method and apparatus for monitoring consciousness |
| CN101352337A (en) * | 2007-12-25 | 2009-01-28 | 天津大学 | Method for capturing signals and extracting characteristics of stand imagination action brain wave |
| CN102697493A (en) * | 2012-05-03 | 2012-10-03 | 北京工业大学 | Method for rapidly and automatically identifying and removing ocular artifacts in electroencephalogram signal |
| CN102973264A (en) * | 2012-12-07 | 2013-03-20 | 哈尔滨工业大学深圳研究生院 | Electrocardiosignal preprocessing method based on morphological multiresolution decomposition |
Non-Patent Citations (1)
| Title |
|---|
| 6 ～ 12岁儿童额叶自发脑电频谱相干的特点;张琼等;《应用心理学》;20071231;第13卷(第4期);311-316 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104515905A (en) | 2015-04-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104515905B (en) | The EEG signals adaptive spectrum analysis method of subject based on CQT multiresolution | |
| CN107361766B (en) | Emotion electroencephalogram signal identification method based on EMD domain multi-dimensional information | |
| CN104706349B (en) | Electrocardiosignal construction method based on pulse wave signals | |
| CN107157477A (en) | EEG signals Feature Recognition System and method | |
| CN105956624B (en) | Mental imagery brain electricity classification method based on empty time-frequency optimization feature rarefaction representation | |
| CN108764043B (en) | Dynamic function connection-based electroencephalogram classification method for entropy values | |
| Cheng et al. | The optimal wavelet basis function selection in feature extraction of motor imagery electroencephalogram based on wavelet packet transformation | |
| Chen et al. | Feature extraction of EEG signals from epilepsy patients based on Gabor transform and EMD decomposition | |
| CN105286860A (en) | Motor imagery brain electrical signal recognition method based on dual-tree complex wavelet energy difference | |
| CN107411739A (en) | EEG signals Emotion identification feature extracting method based on dual-tree complex wavelet | |
| CN107612865A (en) | A kind of signal de-noising method applied to low-voltage powerline carrier communication | |
| CN106485208A (en) | The automatic removal method of eye electrical interference in single channel EEG signals | |
| CN117421539B (en) | Brain-computer interface system data optimization processing method | |
| CN112401902A (en) | Electrocardio identity recognition method and system based on neural network time-frequency analysis combination | |
| CN103932687A (en) | Method and device for preprocessing pulse condition signal | |
| Wu et al. | Fast, accurate localization of epileptic seizure onset zones based on detection of high-frequency oscillations using improved wavelet transform and matching pursuit methods | |
| CN113128384B (en) | Brain-computer interface software key technical method of cerebral apoplexy rehabilitation system based on deep learning | |
| Wang et al. | Research on denoising algorithm for ECG signals | |
| CN103915102B (en) | Method for noise abatement of LFM underwater sound multi-path signals | |
| CN104778342A (en) | Heart sound characteristic extraction method based on wavelet singularity entropy | |
| CN105997067A (en) | Adaptive electromyographic signal detecting and processing method based on fractional Fourier transform | |
| Meng et al. | An empirical envelope estimation algorithm | |
| CN112466322B (en) | Noise signal feature extraction method for electromechanical equipment | |
| CN114358041A (en) | Characteristic waveform extraction method and analysis method based on hybrid algorithm | |
| Yang et al. | Motor imagery EEG recognition based on FBCSP and PCA |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20210318 Address after: Room 206-10, building 16, 1616 Chuangxin Road, Songbei District, Harbin City, Heilongjiang Province Patentee after: Harbin jizuo technology partnership (L.P.) Patentee after: Harbin Institute of Technology Asset Management Co.,Ltd. Address before: 150000 No. 92, West Da Zhi street, Nangang District, Harbin, Heilongjiang. Patentee before: HARBIN INSTITUTE OF TECHNOLOGY |
|
| DD01 | Delivery of document by public notice | ||
| DD01 | Delivery of document by public notice |
Addressee: Li Haifeng Document name: Notification of eligibility |
|
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20210611 Address after: Room 206-12, building 16, 1616 Chuangxin Road, Songbei District, Harbin City, Heilongjiang Province Patentee after: Harbin Institute of Technology Institute of artificial intelligence Co.,Ltd. Address before: Room 206-10, building 16, 1616 Chuangxin Road, Songbei District, Harbin City, Heilongjiang Province Patentee before: Harbin jizuo technology partnership (L.P.) Patentee before: Harbin Institute of Technology Asset Management Co.,Ltd. |