CN102783946B - Automatic brain source locating method and device - Google Patents
Automatic brain source locating method and device Download PDFInfo
- Publication number
- CN102783946B CN102783946B CN201210296457.2A CN201210296457A CN102783946B CN 102783946 B CN102783946 B CN 102783946B CN 201210296457 A CN201210296457 A CN 201210296457A CN 102783946 B CN102783946 B CN 102783946B
- Authority
- CN
- China
- Prior art keywords
- mrow
- msub
- frequency band
- mfrac
- math
- 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.)
- Expired - Fee Related
Links
- 210000004556 brain Anatomy 0.000 title claims abstract description 70
- 238000000034 method Methods 0.000 title claims abstract description 18
- 238000005516 engineering process Methods 0.000 claims abstract description 12
- 239000011159 matrix material Substances 0.000 claims description 39
- 230000001360 synchronised effect Effects 0.000 claims description 28
- 238000001228 spectrum Methods 0.000 claims description 8
- 230000004807 localization Effects 0.000 claims description 6
- 238000005070 sampling Methods 0.000 claims description 5
- 230000011218 segmentation Effects 0.000 claims description 4
- 229910021607 Silver chloride Inorganic materials 0.000 claims description 2
- HKZLPVFGJNLROG-UHFFFAOYSA-M silver monochloride Chemical compound [Cl-].[Ag+] HKZLPVFGJNLROG-UHFFFAOYSA-M 0.000 claims description 2
- 238000000605 extraction Methods 0.000 claims 1
- 230000006399 behavior Effects 0.000 abstract description 4
- 230000008569 process Effects 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 230000007246 mechanism Effects 0.000 description 3
- 210000004761 scalp Anatomy 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 230000001149 cognitive effect Effects 0.000 description 2
- 238000003384 imaging method Methods 0.000 description 2
- 230000010365 information processing Effects 0.000 description 2
- 210000002569 neuron Anatomy 0.000 description 2
- 230000010355 oscillation Effects 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000036982 action potential Effects 0.000 description 1
- 230000007177 brain activity Effects 0.000 description 1
- 230000003925 brain function Effects 0.000 description 1
- 210000005013 brain tissue Anatomy 0.000 description 1
- 210000004027 cell Anatomy 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000010304 firing Methods 0.000 description 1
- 238000012880 independent component analysis Methods 0.000 description 1
- 210000005036 nerve Anatomy 0.000 description 1
- 230000000926 neurological effect Effects 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000000513 principal component analysis Methods 0.000 description 1
- 230000000638 stimulation Effects 0.000 description 1
Images
Landscapes
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention discloses an automatic brain source locating method and a device. The automatic brain source locating method is characterized in that: aiming at multi-channel electroencephalogram signals, firstly, the signals are segmented by utilizing the moving window technology; then, for each segment of data, a instantaneous amplitude value and phase information at a specified frequency range are extracted through harmonic wavelet transform; next, on the basis of synchronization technology, synchronization between two channels at the specified frequency range is calculated; and at last, synchronizing information at the segments is integrated, and synchronization degree of electroencephalogram signals of the channels is obtained, so that the location of a brain source is determined. The invention provides the automatic brain source locating method and the device which can help researchers to determine the brain source rapidly, thus providing the objective basis for understanding the relationship between the brain source and behaviors.
Description
Technical Field
The invention relates to the field of bioengineering, in particular to an automatic brain source positioning method and device, which provide a technical means for revealing the activity of a brain area.
Background
Electroencephalogram signals have become an important means for studying brain science. The brain electrical signals are the comprehensive expression of the action potential of the neuron group in the brain tissue. Normal human whole brain is only active in millionths of volts of electrical potential, mainly because the firing of normal nerves is asynchronous. When the brain needs to process information and reflect external stimulation, most cells of neuron groups synchronously discharge, and if the behavior is extremely extreme, so-called super-synchronization is caused, so that the amplitude of brain waves is rapidly increased, and abnormal waves such as spike waves, sharp waves, spine/spike slow complex waves and the like appear on the brain signals. Researchers can determine the location of the brain sources by analyzing these brain signals. However, long-time electroencephalogram signal recording not only increases the burden of researchers, but also inevitably adds subjective selectivity of some people; meanwhile, a large amount of electroencephalogram signals are interpreted for a long time, and the over-fatigue of researchers is easily caused to cause the rise of the misjudgment rate. Therefore, it is an urgent problem to improve the efficiency of scientific research by using digital signal processing technology to realize the preparation and judgment of brain sources.
The existing brain source localization method generally extracts relevant features from electroencephalogram data through multi-channel electroencephalogram signals recorded for a long time to determine possible brain source regions, wherein the features include waveform features in a time domain, power values in a frequency domain (such as documents: Michel CM, Murray M, Lantz G, Gonzalez S, Spineli L, Grave de Peralta R. EEG source imaging. clin neurological 115,2004: 2195. 2222), or electroencephalogram signal feature components extracted by using component analysis technology (independent component analysis or principal component analysis, etc.), and the magnitude of the contribution of each channel electroencephalogram signal to each channel is taken as a feature (such as documents: Y.Stern, M.Y.Neufeld, S.Kipervassser, A.Zilbertennin 1, I.Fried, M.Teherc, E.Adi-Jaurce relating source imaging of electroencephalogram 2. clinical analysis, 2009-PCA.21. EEG, etc.).
The technology is not designed from the basic principle of brain information processing, most algorithms are from the perspective of signal processing, and the basic rule of brain signal processing is not described essentially; on the other hand, the phase information of the signals is not considered, and a communication mechanism of the phases of the brain signals between different brain areas is not mined. The current research results suggest that the main mechanism of brain processing information is synchronous oscillation, which mainly means that the brain region can integrate and process more complex cognitive information through an inherent synchronous mechanism.
Disclosure of Invention
The invention aims to provide an accurate and automatic brain source positioning method and device, which can directly integrate and express the oscillation synchronization process of brain information processing, provide new information for brain cognitive science research and contribute to the deep research of the relation between brain functions and behaviors.
In order to achieve the above object, the technical scheme of the invention is as follows:
an automatic brain source positioning device comprises
The electroencephalogram signal acquisition equipment is used for acquiring original electroencephalogram signals;
the A/D converter is used for converting the analog electroencephalogram signals into digital signals;
the electroencephalogram signal amplifier is used for amplifying the electroencephalogram signals;
the electroencephalogram signal processor is used for segmenting the electroencephalogram signals; then extracting instantaneous amplitude and phase information of the multichannel electroencephalogram signals on a specified frequency band; then determining the synchronization strength between every two channels on the designated frequency band; determining a brain source positioning matrix on each designated frequency band; finally, determining the average synchronous strength between each channel and all other channels; the average synchronization strength is used for measuring the synchronization strength between the brain area near the electrode and other areas, and the larger the value, the stronger the synchronization, the more likely the brain area is.
The electroencephalogram signal is segmented by adopting a moving window technology.
Instantaneous amplitude and phase information on a specified frequency band is extracted using harmonic wavelet transform.
The specific steps of extracting instantaneous amplitude and phase information on a specified frequency band by utilizing harmonic wavelet transform are as follows: setting the frequency band to be examined as fL,fH]，
Harmonic wavelets have a compact frequency domain representation, defined as
Wherein m and n are scale parameters; the harmonic wavelet transform is realized by fast Fourier transform, and the algorithm is completed by three steps: (a) carrying out Fourier transform on a signal x (t) to be processed to obtain a frequency spectrum X (w) of the signal x (t); (b) multiplying the conjugate of the box-shaped spectrum H (w) of the harmonic wavelet on the (m, n) scale by X (w) to obtain W (w); frequency band of interest [ fL,fH]The relationship with the scale parameters m, n is as follows:
wherein f issThe sampling frequency is N, and the number of points of Fourier transform is N; (c) and performing inverse Fourier transform on W (w) to obtain a harmonic wavelet transform coefficient w (t), wherein the coefficient is a complex number, an imaginary part of the coefficient is Hilbert transform of a real part, and an instantaneous phase of the harmonic wavelet transform coefficient w (t) can be calculated as follows:
φ(t)=tan-1(imag(w(t))/real(w(t)))(3)
instantaneous amplitude of
Thus obtaining EEG signals of all channelsi(t), i is 1,2, … …, L (L is the number of channels) is in [ fL,fH]Instantaneous amplitude A over a frequency bandi(t) and phase information
The specific steps for determining the synchronization strength between every two channels on the specified frequency band are as follows: setting two channels for calculating synchronous intensity as EEG respectivelyi(t) and EEGj(t), i, j =1,2, … …, L; i ≠ j, corresponding instantaneous amplitude Ai(t)、Aj(t) and phase information
Then the phase difference is defined as
Defining a phase synchronization strength factor
Wherein M is the signal length, and the synchronization strength factor is redefined as follows in consideration of the influence of the amplitude:
wherein,<·>representing the average over time, normalized to yield nIPSIij：
Will nIPSIijArranging the signals according to a matrix form, wherein the position of i-j, namely elements on a diagonal line of the matrix all take 0, and obtaining a synchronous matrix between every two channels on the frequency band
The specific steps for determining the brain source positioning matrix on each designated frequency band are as follows: with the moving window, the synchronization matrix E between every two channels on the appointed frequency band in each time period can be obtainedk [i,j]K represents the selected number of time segments K =1,2, … …, K; taking the average value of the synchronous matrixes in all time periods as a brain source positioning matrix E on the designated frequency bandave[i,j]， <math>
<mrow>
<msub>
<mi>E</mi>
<mi>ave</mi>
</msub>
<mo>[</mo>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>]</mo>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<mi>K</mi>
</mfrac>
<munderover>
<mi>Σ</mi>
<mrow>
<mi>k</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>K</mi>
</munderover>
<msub>
<mi>E</mi>
<mi>k</mi>
</msub>
<mo>[</mo>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>]</mo>
<mo>;</mo>
</mrow>
</math>
Determining the average synchronization strength S between each channel and all other channelsi，
An automatic brain source localization method, comprising the steps of:
(1) segmenting an electroencephalogram signal;
(2) extracting instantaneous amplitude and phase information of the multichannel electroencephalogram signals on a specified frequency band;
(3) determining the synchronization strength between every two channels on the designated frequency band;
(4) determining a brain source positioning matrix on each designated frequency band;
(5) determining the average synchronous strength between each channel and all other channels; the average synchronization strength is used for measuring the synchronization strength between the brain area near the electrode and other areas, and the larger the value, the stronger the synchronization, the more likely the brain area is.
And (2) segmenting the electroencephalogram signals in the step (1), and performing segmentation processing by adopting a moving window technology.
And (3) extracting instantaneous amplitude and phase information on the specified frequency band by using harmonic wavelet transform in the step (2).
The specific steps of extracting instantaneous amplitude and phase information on a specified frequency band by utilizing harmonic wavelet transform are as follows: setting the frequency band to be examined as fL,fH]，
Harmonic wavelets have a compact frequency domain representation, defined as
Wherein m and n are scale parameters; the harmonic wavelet transform is realized by fast Fourier transform, and the algorithm is completed by three steps: (a) carrying out Fourier transform on a signal x (t) to be processed to obtain a frequency spectrum X (w) of the signal x (t); (b) multiplying the conjugate of the box-shaped spectrum H (w) of the harmonic wavelet on the (m, n) scale by X (w) to obtain W (w); frequency band of interest [ fL,fH]The relationship with the scale parameters m, n is as follows:
wherein f issThe sampling frequency is N, and the number of points of Fourier transform is N; (c) and performing inverse Fourier transform on W (w) to obtain a harmonic wavelet transform coefficient w (t), wherein the coefficient is a complex number, an imaginary part of the coefficient is Hilbert transform of a real part, and an instantaneous phase of the harmonic wavelet transform coefficient w (t) can be calculated as follows:
φ(t)=tan-1(imag(w(t))/real(w(t)))
instantaneous amplitude of
Thus obtaining EEG signals of all channelsi(t), i is 1,2, … …, L (L is the number of channels) is in [ fL,fH]Instantaneous amplitude A over a frequency bandi(t) and phase information
The specific steps of determining the synchronization strength between every two channels on the designated frequency band in the step (3) are as follows: setting two channels for calculating synchronous intensity as EEG respectivelyi(t) and EEGj(t), i, j =1,2, … …, L; i ≠ j, corresponding instantaneous amplitude Ai(t)、Aj(t) and phase information
Then the phase difference is defined as
Defining a phase synchronization strength factor
Wherein M is the signal length, and the synchronization strength factor is redefined as follows in consideration of the influence of the amplitude:
wherein,<·>representing the average over time, normalized to yield nIPSIij：
Will nIPSIijAccording toArranging in a matrix form, wherein the position of i-j, namely elements on the diagonal line of the matrix all take 0, and obtaining a synchronous matrix between every two channels on the frequency band
The specific steps of determining the brain source positioning matrix on each designated frequency band in the step (4) are as follows: with the moving window, the synchronization matrix E between every two channels on the appointed frequency band in each time period can be obtainedk[i,j]K represents the selected number of time segments K =1,2, … …, K; taking the average value of the synchronous matrixes in all time periods as a brain source positioning matrix E on the designated frequency bandave[i,j]， <math>
<mrow>
<msub>
<mi>E</mi>
<mi>ave</mi>
</msub>
<mo>[</mo>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>]</mo>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<mi>K</mi>
</mfrac>
<munderover>
<mi>Σ</mi>
<mrow>
<mi>k</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>K</mi>
</munderover>
<msub>
<mi>E</mi>
<mi>k</mi>
</msub>
<mo>[</mo>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>]</mo>
<mo>;</mo>
</mrow>
</math>
Determining the average synchronous strength S between each channel and all other channels in the step (5)i， <math>
<mrow>
<msub>
<mi>S</mi>
<mi>i</mi>
</msub>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<mrow>
<mi>L</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
</mfrac>
<munderover>
<mi>Σ</mi>
<mrow>
<mi>j</mi>
<mo>=</mo>
<mn>1</mn>
<mo>,</mo>
<mi>j</mi>
<mo>≠</mo>
<mi>i</mi>
</mrow>
<mi>L</mi>
</munderover>
<msub>
<mrow>
<mo>(</mo>
<msub>
<mi>E</mi>
<mi>ave</mi>
</msub>
<mo>)</mo>
</mrow>
<mi>ij</mi>
</msub>
<mo>.</mo>
</mrow>
</math>
Finally, the average synchronous strength S of each channel is displayed on a two-dimensional graph by using an image processing technology.
Compared with the prior art, the invention has the advantages that:
(1) the invention utilizes harmonic wavelet transformation to extract the instantaneous information of the multi-channel electroencephalogram signal on any required frequency band, lays a foundation for further synchronous calculation, and simultaneously makes the analysis of the brain source on a plurality of frequency bands possible;
(2) the invention provides an improved phase synchronization method as a new synchronization intensity factor, which is used for calculating a synchronization intensity matrix between every two channels in different frequency bands so as to obtain a total synchronization intensity index between a brain region near each electrode and other regions, and the total synchronization intensity index is used for identifying the region of a brain source.
Drawings
Fig. 1 is a schematic view of the working process of the present invention.
Fig. 2 is a multi-channel electroencephalogram signal collected from the scalp, wherein a brain activity synchronization period is between two vertical virtual lines.
Fig. 3 is the result of localization of brain sources. The first row is respectively a synchronous information matrix between every two channels on different frequency bands (theta (4-8Hz), alpha (8-12 Hz), beta (12-30 Hz) and gamma (30-80 Hz)) from left to right; the second row is the average synchronization strength between each electroencephalogram channel and all other channels on different frequency bands; the last row is the result of displaying the average synchronization strength between each channel and all other channels on a two-dimensional graph over different frequency bands.
Detailed Description
The device of the invention firstly collects the brain electrical signal. The electroencephalogram signal acquisition is completed by an electroencephalogram signal acquisition device, a special Ag/AgCl electrode or an existing special electroencephalogram electrode cap is generally adopted to be matched with an electroencephalogram amplifier to acquire the electroencephalogram signal, then the acquired electroencephalogram analog signal is converted into a digital signal through an A/D converter, and the digital signal is input into a processor. This embodiment is a long scalp EEG signal recording as shown in fig. 2. When collecting EEG signals, the electrodes are placed according to 10-20 international standard leads, and two temples T1 and T2 are added. Sampling frequency fs=250Hz, using average reference electrode recording. And filtering the signal by a 0.5-80Hz band pass filter to remove clutter interference. FIG. 1 is a flow chart of the present invention, and the steps of the processor for processing the electroencephalogram signal are as follows:
step 1, carrying out segmentation processing on the selected electroencephalogram signal data by adopting a moving window technology. Each segment was selected to be 4 seconds long with 50% overlap.
And 2, extracting instantaneous amplitude and phase information on theta (4-8Hz), alpha (8-12 Hz), beta (12-30 Hz) and gamma (30-80Hz) frequency bands of each segment of signals by utilizing harmonic wavelet transformation, wherein the theta (4-8Hz) frequency band is taken as an example in the embodiment to calculate the synchronous strength between every two channels.
And 3, calculating the synchronization strength between every two channels based on the instantaneous amplitude and phase information of the electroencephalogram signals of each channel on the theta (4-8Hz) frequency band. Assuming that the two channels for which the synchronisation strength is to be calculated are EEG's respectivelyi(t) and EEGj(t), i, j =1,2, … …, L; i ≠ j, i.e. a total of L channels, the phase information on the frequency band extracted by these two channels to be calculated being
And
amplitude information of Ai(t) and Aj(t) of (d). Calculating the phase difference using equation (5)
Then, the formula (6) is applied to calculate the phase synchronization factor PSI between the two channelsij. But the magnitude factor is not taken into account, an improved synchronization strength factor IPSI is calculated using equation (7)ijDue to the elements IPSIijIs not between 0 and 1, and each element is normalized by using the largest element to obtain <math>
<mrow>
<msub>
<mi>nIPSI</mi>
<mi>ij</mi>
</msub>
<mo>=</mo>
<mfrac>
<msub>
<mi>IPSI</mi>
<mi>ij</mi>
</msub>
<mrow>
<munder>
<mi>max</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>,</mo>
<mi>i</mi>
<mo>≠</mo>
<mi>j</mi>
</mrow>
</munder>
<mrow>
<mo>(</mo>
<msub>
<mi>IPSI</mi>
<mi>ij</mi>
</msub>
<mo>)</mo>
</mrow>
</mrow>
</mfrac>
<mo>.</mo>
</mrow>
</math>
Will nIPSIijArranging the signals according to a matrix form, wherein the position of i-j, namely elements on a diagonal line of the matrix all take 0, and obtaining a synchronous matrix between every two channels on the frequency band
Step 4, with the sliding window moving, repeating the steps 2 and 3, and calculating to obtain a synchronization matrix E between a plurality of pairwise channels in each time period and theta (4-8Hz) frequency bandk[i,j]K represents the selected time segment number K is 1,2, … …, K, the total length of the electroencephalogram signal selected in this embodiment is 1min, i.e. 60s, each segment is 4 seconds, and 50% overlap is selected, i.e. the 1 st time segment is 0-4s, the 2 nd time segment is 2-6s, the 3 rd time segment is 4-8s, and so on, and the last time segment is56-60s, the total number is 29, so that K is 29, and the synchronous matrix between every two channels from the 1 st time period to the last 29 th time period in the theta (4-8Hz) frequency band is E1[i,j],E2[i,j]……E29[i,j]. Taking the difference between the segments into consideration, taking the average value of the synchronization matrix in all time periods as the brain source positioning matrix E in the specified frequency bandave[i,j]，
The brain source location matrixes of other frequency bands can be obtained by the same method, as shown in the first row of fig. 3, the brain source location matrixes E between every two channels on different frequency bands (theta, alpha, beta and gamma) are respectively obtained from left to rightave[i,j]。
In order to more intuitively display the calculation result, the synchronicity size of each channel is displayed on the corresponding position of the brain topographic map by using an image processing technology. As shown in the last row of fig. 3, the result is calculated for the synchronization strength of each channel in different frequency bands.
Fig. 3 is the result of localization of brain sources. The first row is respectively a brain source positioning matrix between every two channels on different frequency bands (theta (4-8Hz), alpha (8-12 Hz), beta (12-30 Hz) and gamma (30-80 Hz)) from left to right; when the synchronization intensity among the channels is larger, the color of the image is darker; comparing 4 different frequency bands, the brain source localization matrix has certain difference. The second row is the average synchronization strength between each electroencephalogram channel and all other channels on different frequency bands, and the influence degree of different channels under different frequency bands on the whole synchronization can be found. The last row is the result of displaying the average intensity of synchrony between each channel and all other channels on a two-dimensional map over different frequency bands, compared to the fact that the brain sources (darker in color) are centrally distributed in the front of the right brain.
Claims (2)
1. An automatic brain source positioning device comprises
The electroencephalogram signal acquisition equipment is used for acquiring original electroencephalogram signals;
the A/D converter is used for converting the analog electroencephalogram signal into a digital signal;
the electroencephalogram signal amplifier is used for amplifying the electroencephalogram signals;
the electroencephalogram signal segmentation method is characterized by further comprising an electroencephalogram signal processor, wherein the electroencephalogram signal processor firstly segments the electroencephalogram signal; then extracting instantaneous amplitude and phase information of the multichannel electroencephalogram signals on a specified frequency band; then determining the synchronization strength between every two channels on the designated frequency band; determining a brain source positioning matrix on each designated frequency band; finally, determining the average synchronous strength between each channel and all other channels; the average synchronization strength is used for measuring the degree of synchronism between a brain region near an electrode of the electroencephalogram signal acquisition equipment and other regions, and the larger the numerical value is, the stronger the synchronism is, the more likely the brain region is represented;
the electroencephalogram signal is segmented by adopting a moving window technology; extracting instantaneous amplitude and phase information on a specified frequency band by utilizing harmonic wavelet transform;
the specific steps of extracting instantaneous amplitude and phase information on a specified frequency band by utilizing harmonic wavelet transform are as follows: setting the frequency band to be examined as fL,fH]，
Harmonic wavelets have a compact frequency domain representation, defined as
Wherein m and n are scale parameters; the harmonic wavelet transform is realized by fast Fourier transform, and the algorithm is completed by three steps: (a) carrying out Fourier transform on a signal x (t) to be processed to obtain a frequency spectrum X (w) of the signal x (t); (b) multiplying the conjugate of the box-shaped spectrum H (w) of the harmonic wavelet on the (m, n) scale by X (w) to obtain W (w); frequency band of interest [ fL,fH]The relationship with the scale parameters m, n is as follows:
wherein f issThe sampling frequency is N, and the number of points of Fourier transform is N; (c) and (w) performing inverse Fourier transform to obtain harmonic wavelet transform coefficient w (t), wherein the coefficient is a complex number, the imaginary part is Hilbert transform of a real part, and the instantaneous phase is calculated as follows:
φ(t)=tan-1(imag(w(t))/real(w(t)))
instantaneous amplitude of
Thus obtaining EEG signals of all channelsi(t), i =1,2, ·, L, L is the number of channels, [ f ]L,fH]Instantaneous amplitude A over a frequency bandi(t) and phase information
The specific steps for determining the synchronization strength between every two channels on the specified frequency band are as follows: setting two channels for calculating synchronous intensity as EEG respectivelyi(t) and EEGj(t), i, j =1,2, ·, L i ≠ j, corresponding to the instantaneous amplitude ai(t)、Aj(t) and phase information
Then the phase difference is defined as
Defining a phase synchronization strength factor
Wherein M is the signal length, and the synchronization strength factor is redefined as follows in consideration of the influence of the amplitude:
wherein,<Ai(t)>and<Aj(t)>respectively representing the average value of the corresponding instantaneous amplitude values of the i channel and the j channel in time, and normalizing the average values to obtain nIPSIij：
Will nIPSIijArranging according to a matrix form, wherein the positions of i = j, namely elements on the diagonal line of the matrix are all 0, and obtaining a synchronous matrix between every two channels on the frequency band
The specific steps for determining the brain source positioning matrix on each designated frequency band are as follows: with the moving window, a plurality of synchronous matrixes E between every two channels on the appointed frequency band in each time period can be obtainedk[i,j]K represents the number of selected time segments K =1,2, ·, K; taking the average value of the synchronous matrixes in all time periods as a brain source positioning matrix E on the designated frequency bandave[i,j]，
Determining the average synchronization strength S between each channel and all other channelsi，
2. An automatic brain source localization method is characterized by comprising the following steps:
(1) collecting an electroencephalogram signal by using an Ag/AgCl electrode, and segmenting the electroencephalogram signal;
(2) extracting instantaneous amplitude and phase information of the multichannel electroencephalogram signals on a specified frequency band;
(3) determining the synchronization strength between every two channels on the designated frequency band;
(4) determining a brain source positioning matrix on each designated frequency band;
(5) determining the average synchronous strength between each channel and all other channels; the average synchronization strength is used for measuring the synchronization strength between the brain area near the electrode and other areas, and the larger the value is, the stronger the synchronization is, the more likely the brain area is represented;
in the step (1), the electroencephalogram signals are segmented, and a moving window technology is adopted for segmentation processing; extracting instantaneous amplitude and phase information on a specified frequency band by utilizing harmonic wavelet transform in the step (2);
extraction of instantaneous amplitude and phase information on specified frequency band by harmonic wavelet transformThe method comprises the following steps: setting the frequency band to be examined as fL,fH]，
Harmonic wavelets have a compact frequency domain representation, defined as
Wherein m and n are scale parameters; the harmonic wavelet transform is realized by fast Fourier transform, and the algorithm is completed by three steps: (a) carrying out Fourier transform on a signal x (t) to be processed to obtain a frequency spectrum X (w) of the signal x (t); (b) multiplying the conjugate of the box-shaped spectrum H (w) of the harmonic wavelet on the (m, n) scale by X (w) to obtain W (w); frequency band of interest [ fL,fH]The relationship with the scale parameters m, n is as follows:
wherein f issThe sampling frequency is N, and the number of points of Fourier transform is N; (c) and (w) performing inverse Fourier transform to obtain harmonic wavelet transform coefficient w (t), wherein the coefficient is a complex number, the imaginary part is Hilbert transform of a real part, and the instantaneous phase is calculated as follows:
φ(t)=tan-1(imag(w(t))real(w(t)))
instantaneous amplitude of
Thus obtaining EEG signals of all channelsi(t)，i=1,2,···,L，L is the number of channels, in [ f ]L,fH]Instantaneous amplitude A over a frequency bandi(t) and phase information
The specific steps of determining the synchronization strength between every two channels on the designated frequency band in the step (3) are as follows: setting two channels for calculating synchronous intensity as EEG respectivelyi(t) and EEGj(t), i, j =1,2, ·, L i ≠ j, corresponding to the instantaneous amplitude ai(t)、Aj(t) and phase informationThen the phase difference is defined as
Defining a phase synchronization strength factor
Wherein M is the signal length, and the synchronization strength factor is redefined as follows in consideration of the influence of the amplitude:
wherein,<Ai(t)>and<Aj(t)>respectively representing the average value of the corresponding instantaneous amplitude values of the i channel and the j channel in time, and normalizing the average values to obtain nIPSIij：
Will nIPSIijArranging according to a matrix form, wherein the positions of i = j, namely elements on the diagonal line of the matrix are all 0, and obtaining a synchronous matrix between every two channels on the frequency band
The specific steps of determining the brain source positioning matrix on each designated frequency band in the step (4) are as follows: with the moving window, the synchronous matrix E between a plurality of channels on the appointed frequency band in each time period can be obtainedk[i,j]K represents the number of selected time segments K =1,2, ·, K; taking the average value of the synchronous matrixes in all time periods as a brain source positioning matrix E on the designated frequency bandave[i,j]，
Determining the average synchronous strength S between each channel and all other channels in the step (5)i， <math>
<mrow>
<msub>
<mi>S</mi>
<mi>i</mi>
</msub>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<mrow>
<mi>L</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
</mfrac>
<munderover>
<mi>Σ</mi>
<mrow>
<mi>j</mi>
<mo>=</mo>
<mn>1</mn>
<mo>,</mo>
<mi>j</mi>
<mo>≠</mo>
<mi>i</mi>
</mrow>
<mi>L</mi>
</munderover>
<msub>
<mrow>
<mo>(</mo>
<msub>
<mi>E</mi>
<mi>ave</mi>
</msub>
<mo>)</mo>
</mrow>
<mi>ij</mi>
</msub>
<mo>.</mo>
</mrow>
</math>
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210296457.2A CN102783946B (en) | 2012-08-20 | 2012-08-20 | Automatic brain source locating method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210296457.2A CN102783946B (en) | 2012-08-20 | 2012-08-20 | Automatic brain source locating method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102783946A CN102783946A (en) | 2012-11-21 |
| CN102783946B true CN102783946B (en) | 2014-06-25 |
Family
ID=47149689
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210296457.2A Expired - Fee Related CN102783946B (en) | 2012-08-20 | 2012-08-20 | Automatic brain source locating method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102783946B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP6142354B2 (en) * | 2013-02-27 | 2017-06-07 | 国立研究開発法人理化学研究所 | EEG signal processing apparatus, EEG signal processing method, program, and recording medium |
| CN103505203B (en) * | 2013-09-30 | 2015-06-03 | 西安交通大学 | Method for detecting human metal states based on brain electrical source locating method |
| CN104510468A (en) * | 2014-12-30 | 2015-04-15 | 中国科学院深圳先进技术研究院 | Character extraction method and device of electroencephalogram |
| CN113143293B (en) * | 2021-04-12 | 2023-04-07 | 天津大学 | Continuous speech envelope nerve entrainment extraction method based on electroencephalogram source imaging |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP3983989B2 (en) * | 2001-03-19 | 2007-09-26 | 独立行政法人科学技術振興機構 | Brain motor function analysis and diagnosis device |
| CN101449974A (en) * | 2007-12-05 | 2009-06-10 | 李小俚 | Method for automatic real-time estimating anesthesia depth |
| RU2419383C1 (en) * | 2009-12-28 | 2011-05-27 | Государственное образовательное учреждение высшего профессионального образования "Воронежская государственная медицинская академия им. Н.Н. Бурденко Федерального агентства по здравоохранению и социальному развитию" | Method of diagnosing affective disorders in patients with ihd by data of encephalographic examination |
| KR20120018963A (en) * | 2010-08-24 | 2012-03-06 | 연세대학교 산학협력단 | Method for producing two-dimensional spatiospectral erd/ers patterns from electroencephalogram, method for classifying mental tasks based on the two-dimensional spatiospectral patterns and brain-computer interface system using classified electroencephalogram by the classifying method as input signal |
| KR20120082689A (en) * | 2011-01-14 | 2012-07-24 | 한국과학기술원 | Method and apparatus for quantifying risk of developing schizophrenia using brainwave synchronization level, and computer-readable media recording codes for the method |
-
2012
- 2012-08-20 CN CN201210296457.2A patent/CN102783946B/en not_active Expired - Fee Related
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP3983989B2 (en) * | 2001-03-19 | 2007-09-26 | 独立行政法人科学技術振興機構 | Brain motor function analysis and diagnosis device |
| CN101449974A (en) * | 2007-12-05 | 2009-06-10 | 李小俚 | Method for automatic real-time estimating anesthesia depth |
| RU2419383C1 (en) * | 2009-12-28 | 2011-05-27 | Государственное образовательное учреждение высшего профессионального образования "Воронежская государственная медицинская академия им. Н.Н. Бурденко Федерального агентства по здравоохранению и социальному развитию" | Method of diagnosing affective disorders in patients with ihd by data of encephalographic examination |
| KR20120018963A (en) * | 2010-08-24 | 2012-03-06 | 연세대학교 산학협력단 | Method for producing two-dimensional spatiospectral erd/ers patterns from electroencephalogram, method for classifying mental tasks based on the two-dimensional spatiospectral patterns and brain-computer interface system using classified electroencephalogram by the classifying method as input signal |
| KR20120082689A (en) * | 2011-01-14 | 2012-07-24 | 한국과학기술원 | Method and apparatus for quantifying risk of developing schizophrenia using brainwave synchronization level, and computer-readable media recording codes for the method |
Non-Patent Citations (6)
| Title |
|---|
| 吴祈耀，吴祈宗.脑电信号的现代谱分析技术.《北京理工大学学报》.1995,(第02期), * |
| 基于Harmonic小波时变相干的人脸识别脑电特征提取;许慰玲等;《电子测量与仪器学报》;20060630(第03期);第14~18页 * |
| 王蕾等.基于小波变换的癫痫脑电相位同步化分析方法.《医疗卫生装备》.2007,(第10期), * |
| 许慰玲等.基于Harmonic小波时变相干的人脸识别脑电特征提取.《电子测量与仪器学报》.2006,(第03期), |
| 赵丽娜等.基于信号处理的脑电相位同步性分析方法研究.《生物医学工程学杂志》.2008,(第02期), * |
| 郁洪强等.过度使用互联网对脑电事件相关电位gamma节律同步效应的影响研究.《仪器仪表学报》.2008,(第11期), * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102783946A (en) | 2012-11-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111340142B (en) | Epilepsia magnetoencephalogram spike automatic detection method and tracing positioning system | |
| CN102697493B (en) | Method for rapidly and automatically identifying and removing ocular artifacts in electroencephalogram signal | |
| CN101596101B (en) | Method for determining fatigue state according to electroencephalogram | |
| CN105496363A (en) | Method for classifying sleep stages on basis of sleep EGG (electroencephalogram) signal detection | |
| CN104305993B (en) | A kind of artifacts removing method based on Granger Causality | |
| CN104914994A (en) | Aircraft control system and fight control method based on steady-state visual evoked potential | |
| CN104173046B (en) | A kind of extracting method of color indicia Amplitude integrated electroencephalogram | |
| CN102783946B (en) | Automatic brain source locating method and device | |
| CN103405225B (en) | A kind of pain that obtains feels the method for evaluation metrics, device and equipment | |
| CN106805945A (en) | The removing method of Muscle artifacts in a kind of EEG signals of a small number of passages | |
| CN105342605A (en) | Method for removing myoelectricity artifacts from brain electrical signals | |
| CN111973179B (en) | Brain wave signal processing method, brain wave signal processing device, electronic device, and storage medium | |
| WO2012151453A2 (en) | Seizure detection and epileptogenic lesion localization | |
| CN104510468A (en) | Character extraction method and device of electroencephalogram | |
| CN106901727A (en) | A kind of depression Risk Screening device based on EEG signals | |
| CN100538713C (en) | A kind of brain electricity fluctuation signal analysis equipment | |
| Anwar et al. | Use of portable EEG sensors to detect meditation | |
| Meenakshi et al. | Frequency analysis of healthy & epileptic seizure in EEG using fast Fourier transform | |
| CN101433460B (en) | Spatial filtering method of lower limb imaginary action potential | |
| Chen et al. | A hybrid method for muscle artifact removal from EEG signals | |
| CN110338787A (en) | A kind of analysis method of pair of static EEG signals | |
| CN113576498B (en) | Visual and auditory aesthetic evaluation method and system based on electroencephalogram signals | |
| CN113143296B (en) | Intelligent assessment method and system for communication disorder | |
| CN106821318A (en) | A kind of multiple dimensioned quantitative analysis method of EEG signals | |
| Dilber et al. | EEG based detection of epilepsy by a mixed design approach |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20140625 Termination date: 20150820 |
|
| EXPY | Termination of patent right or utility model |