CN114145754B - EEG cross frequency coupling-based stroke brain function assessment device - Google Patents
EEG cross frequency coupling-based stroke brain function assessment device Download PDFInfo
- Publication number
- CN114145754B CN114145754B CN202111522093.0A CN202111522093A CN114145754B CN 114145754 B CN114145754 B CN 114145754B CN 202111522093 A CN202111522093 A CN 202111522093A CN 114145754 B CN114145754 B CN 114145754B
- Authority
- CN
- China
- Prior art keywords
- brain
- frequency
- functional connection
- segment
- connection matrix
- 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
- 238000010168 coupling process Methods 0.000 title claims abstract description 32
- 238000005859 coupling reaction Methods 0.000 title claims abstract description 32
- 230000008878 coupling Effects 0.000 title claims abstract description 30
- 230000003925 brain function Effects 0.000 title claims abstract description 28
- 210000004556 brain Anatomy 0.000 claims abstract description 119
- 230000006870 function Effects 0.000 claims abstract description 63
- 238000007781 pre-processing Methods 0.000 claims abstract description 15
- 239000011159 matrix material Substances 0.000 claims description 100
- 238000004364 calculation method Methods 0.000 claims description 48
- 208000006011 Stroke Diseases 0.000 claims description 32
- 230000000638 stimulation Effects 0.000 claims description 27
- 238000004458 analytical method Methods 0.000 claims description 25
- 238000000034 method Methods 0.000 claims description 19
- 238000012935 Averaging Methods 0.000 claims description 13
- 230000002490 cerebral effect Effects 0.000 claims description 5
- 238000011156 evaluation Methods 0.000 claims description 5
- 238000000819 phase cycle Methods 0.000 claims description 5
- 230000009466 transformation Effects 0.000 claims description 5
- 241001450674 Saissetia coffeae Species 0.000 claims description 4
- 238000001914 filtration Methods 0.000 claims description 3
- 238000012800 visualization Methods 0.000 claims description 3
- 206010008190 Cerebrovascular accident Diseases 0.000 claims description 2
- 230000005540 biological transmission Effects 0.000 claims description 2
- 230000004069 differentiation Effects 0.000 claims description 2
- 230000010354 integration Effects 0.000 claims description 2
- 238000003012 network analysis Methods 0.000 abstract description 3
- 230000036541 health Effects 0.000 abstract description 2
- 238000012360 testing method Methods 0.000 description 14
- 238000010586 diagram Methods 0.000 description 7
- 238000005516 engineering process Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 210000001550 testis Anatomy 0.000 description 2
- NAWXUBYGYWOOIX-SFHVURJKSA-N (2s)-2-[[4-[2-(2,4-diaminoquinazolin-6-yl)ethyl]benzoyl]amino]-4-methylidenepentanedioic acid Chemical compound C1=CC2=NC(N)=NC(N)=C2C=C1CCC1=CC=C(C(=O)N[C@@H](CC(=C)C(O)=O)C(O)=O)C=C1 NAWXUBYGYWOOIX-SFHVURJKSA-N 0.000 description 1
- 208000026106 cerebrovascular disease Diseases 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000004070 electrodeposition Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 235000019580 granularity Nutrition 0.000 description 1
- 230000001771 impaired effect Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 210000001595 mastoid Anatomy 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 210000004761 scalp Anatomy 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
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
- A61B5/369—Electroencephalography [EEG]
- A61B5/372—Analysis of electroencephalograms
- A61B5/374—Detecting the frequency distribution of signals, e.g. detecting delta, theta, alpha, beta or gamma waves
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/369—Electroencephalography [EEG]
- A61B5/377—Electroencephalography [EEG] using evoked responses
- A61B5/378—Visual stimuli
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7203—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/725—Details of waveform analysis using specific filters therefor, e.g. Kalman or adaptive filters
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Medical Informatics (AREA)
- Surgery (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Veterinary Medicine (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Psychiatry (AREA)
- Molecular Biology (AREA)
- Physics & Mathematics (AREA)
- Animal Behavior & Ethology (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Signal Processing (AREA)
- Psychology (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physiology (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention discloses a stroke brain function assessment device based on EEG cross frequency coupling. Collecting and preprocessing multichannel electroencephalogram data of a stroke state or a health state under a motor imagery task, and extracting stimulated effective data segments; calculating the phase coupling relation between the frequency bands in each data segment; and extracting multi-scale brain network indexes including a whole brain average function connection value, an average function connection value on a hemisphere scale, a characteristic path length and a clustering coefficient index, and evaluating the brain function state according to the Euclidean distance. The invention breaks through the limitation of single-band brain network analysis, and effectively evaluates the brain function of stroke through the cross frequency brain network.
Description
Technical Field
The invention belongs to the stroke assessment field in the technical field of electroencephalogram analysis, and particularly relates to a stroke brain function assessment device based on EEG cross frequency coupling.
Background
Stroke is a global cerebrovascular disease and in most countries is the leading cause of disability in adults. In addition, stroke also severely impedes the daily life of the patient and his family. Electroencephalogram movement is a technology clinically used for rehabilitation of movement functions of stroke, and is beneficial to rehabilitation of movement functions of stroke patients. However, in the field of stroke assessment, doctors evaluate stroke patients mainly by means of medical scales, but this method relies on the clinical experience of doctors, and the judgment result is often time-consuming and not comprehensive enough.
Therefore, related art experts propose techniques for estimating brain functions of stroke patients based on brain electric motor imagery data, such as by a beta band brain network. At present, the technology only considers the coupling relation of the brain electrical signals between different channels in a single frequency band, and omits the brain complex interaction information transferred between cross frequency bands, namely cross-frequency coupling (cross-frequency coupling, CFC).
In conclusion, the invention realizes a cerebral function evaluation device by utilizing the design of the electroencephalogram cross frequency coupling technology so as to objectively assist cerebral function evaluation of a cerebral patient.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provides a stroke brain function assessment device based on EEG cross frequency coupling. The brain network based on cross frequency coupling is constructed by taking brain electrical data of a stroke patient and a healthy tested in left hand and right hand motor imagery as research objects, and effective brain network indexes are extracted from a plurality of time and space granularities so as to assist in evaluating the brain function state of the stroke patient.
To achieve the above object, the stroke brain function assessment device based on EEG cross frequency coupling of the present invention specifically comprises the following modules:
and the brain electrical data acquisition module is used for acquiring multichannel brain electrical data of a healthy state or a cerebral apoplexy state when a tested person moves imagination in the left hand and the right hand.
The electroencephalogram data preprocessing frequency division module is used for preprocessing electroencephalogram data acquired by the electroencephalogram data acquisition module so as to remove artifact components in the electroencephalogram data, improve signal to noise ratio and obtain complete electroencephalogram signal segments and multi-segment electroencephalogram signal subfragments under left and right hand motor imagery stimulus of five frequency bands of delta (0.1-4 hz), theta (4-8 hz), alpha (8-12 hz), low beta (12 hz-20 hz) and high beta (20 hz-28 hz); the preprocessing specifically comprises operations of removing the electrooculogram, re-referencing, correcting a base line, filtering, dividing the electroencephalogram data and the like.
The cross frequency function connection matrix calculation module is used for calculating the phase coupling relation between each electroencephalogram signal sub-segment and the complete electroencephalogram signal segment in the frequency band and between the frequency bands after being processed by the electroencephalogram data preprocessing frequency division module, so as to construct a function connection matrix in the frequency band and a function connection matrix between the frequency bands of each electroencephalogram signal sub-segment and the complete electroencephalogram signal segment; the system comprises a first computing module and a second computing module;
the first calculation module is used for calculating the instantaneous phase sequence of each electroencephalogram signal sub-segment and the complete electroencephalogram signal segment; the method specifically comprises the following steps:
the instantaneous phase sequence phi (t) of the electroencephalogram signal x (t) on each channel is calculated through Hilbert transformation on the electroencephalogram signal processed by the electroencephalogram data preprocessing frequency division module, and the formula is shown as follows:
wherein the method comprises the steps ofAnd the PV is a Keke main value which is the result of the Hilbert transformation of the electroencephalogram signal x (t) at the moment t on a certain channel.
The second calculation module is used for calculating a function connection matrix in the frequency band and a function connection matrix between the frequency bands; the method specifically comprises the following steps:
for the same frequency band (delta, alpha, low beta and high beta frequency bands are selected), selecting the center frequency f of any two channels m And f n F is a signal of (f) m ＝f n Calculate the phase-to-phase coupling relation PSI (f) m ，f n ) I.e. the functional connection value between two channels, the range of values being 0,1]The calculation formula is as follows:
wherein T represents the number of time samples of the signal; j represents an imaginary number; Δφ (f) m ，f n T) represents two central frequency bands f m And f n The calculation method of the instantaneous phase difference of the signals is as follows:
Δφ(f m ，f n ，t)＝φ(f m ，t)-φ(f n t) type (4)
For different frequency bands (delta-alpha two frequency bands, delta-low beta two frequency bands and delta-high beta two frequency bands are selected) Selecting the center frequency f of any two channels with different frequency bands m And f n F is a signal of (f) m ≠f n Calculating the phase-coupling relation PSI (f) between two signals according to the formula (3) m ，f n ) The method comprises the steps of carrying out a first treatment on the surface of the Calculating two central frequency bands as f according to a formula (5) m And f n Wherein n and m are the smallest positive integers satisfying the requirement of equation (6);
Δφ(f m ，f n ，t)＝nφ(f m ，t)-mφ(f n t) type (5)
n×f m ＝m×f n (6)
Constructing an N multiplied by N frequency band internal function connection matrix for each single-frequency band internal phase coupling relation of each electroencephalogram signal sub-segment and the whole electroencephalogram signal segment; constructing an NxN inter-band functional connection matrix for the phase coupling relationship between each different frequency band of each electroencephalogram sub-segment and each different frequency band of the whole electroencephalogram segment; n represents the number of channels.
The multi-scale brain network index calculation and analysis module is used for calculating the functional connection value index on the whole brain and hemisphere scale under the left and right hand motor imagery stimulus and the two-to-two cross frequency brain network index for each brain signal sub-segment and the frequency band functional connection matrix of the complete brain signal segment constructed by the cross frequency functional connection matrix calculation module, and then carrying out index analysis on the functional connection value index; the system comprises a function connection value index calculation module, a pairwise cross frequency brain network index calculation module and an index analysis module;
the function connection value index calculation module is used for calculating function connection value indexes in alpha, low beta single frequency bands and between delta-alpha, delta-low beta and delta-high beta under left and right hand motor imagery stimulus, and comprises a full brain function connection value and a function connection value on a hemisphere scale; the method specifically comprises the following steps:
the cross frequency function connection matrix calculation module obtains a frequency band function connection matrix and an inter-frequency band function connection matrix under two stimulation types because the motor imagination of the left hand and the right hand is performed;
1) PSI (f) between all channels on each intra-frequency functional connection matrix and inter-frequency functional connection matrix for each electroencephalogram signal sub-segment and complete electroencephalogram signal segment under left and right hand motor imagery stimulus m ，f n ) The values are respectively subjected to averaging treatment to obtain full brain function connection values in a single frequency band and between frequency bands under the stimulation of two motor imagery;
2) The functional connection values on the hemispherical scale comprise a left hemispherical functional connection value, a right hemispherical functional connection value and left and right hemispherical functional connection values;
because the left-hand motor imagery stimulus type is corresponding to the right brain motor related brain region and the right-hand motor imagery stimulus type is corresponding to the left brain motor related brain region, PSI (f) between right brain channels on each intra-frequency functional connection matrix and inter-frequency functional connection matrix of each brain electric signal sub-segment and complete brain electric signal segment under the left-hand motor imagery stimulus type m ，f n ) The values are respectively subjected to averaging treatment to obtain right hemisphere function connection values in each frequency band and between frequency bands under the stimulation of two motor imagery;
PSI (f) between left brain channels on each intra-frequency functional connection matrix and inter-frequency functional connection matrix of each electroencephalogram sub-segment and complete electroencephalogram segment under the type of right-hand motor imagery stimulation m ，f n ) The values are respectively subjected to averaging treatment to obtain left hemisphere function connection values in each frequency band and between frequency bands under the stimulation of two motor imagery;
PSI (f) between left brain channel and right brain channel on each intra-frequency functional connection matrix and inter-frequency functional connection matrix of each electroencephalogram signal sub-segment and complete electroencephalogram signal segment under two motor imagery stimulus types m ，f n ) The values are respectively subjected to averaging treatment to obtain left and right hemisphere function connection values in each frequency band and between frequency bands under the stimulation of two motor imagery;
the two-by-two cross frequency brain network index calculation module is used for combining the delta, alpha, low beta and high beta frequency band functional connection matrix and the delta-alpha, delta-low beta and delta-high beta frequency band functional connection matrix, and calculating the characteristic path length and the clustering coefficient index of the two-by-two cross frequency brain network when the left hand and the right hand move imagine; the method specifically comprises the following steps:
1) Splicing the functional connection matrixes in any two frequency bands of each electroencephalogram signal sub-segment and the complete electroencephalogram signal segment under the left and right hand motor imagery stimulus types with the corresponding functional connection matrixes between the frequency bands to construct a functional connection matrix with the size of 2N multiplied by 2N; the diagonal of the functional connection matrix is two intra-frequency-band functional connection matrices, and the non-diagonal is a corresponding inter-frequency-band functional connection matrix; reserving each element meeting the threshold range in the functional connection matrix and serving as the edge of the 2N node pairwise cross frequency brain network; wherein the node refers to a channel;
2) Calculating characteristic path length indexes for the pairwise crossover frequency brain network:
the characteristic path length is used for describing the efficiency and the capability of global function integration and information transmission of the brain network, is defined as the average value of the shortest path between any two nodes in the brain network, and has the following formula:
wherein N represents the number of nodes in the brain network, l ij Representing the shortest path length between node i and node j.
3) Calculating a clustering coefficient index for the pairwise crossover frequency brain network:
the clustering coefficient is used for describing the function differentiation capability of the local part of the brain network, and the clustering coefficient C of the whole brain network defines the clustering coefficient C of all nodes i The formula is shown below:
wherein N represents the number of nodes, C i Clustering coefficients, k, representing node i i The degree of the node i is indicated,representing the weight of the edge between the node a and the node b in the brain network after the scaling, w a，b Representing the weight of the edge between the node a and the node b in the brain network without scaling, and max (w) represents the maximum value of the weight of the edge of any two nodes in the brain network;
in order to better represent the functional state of the brain, the functional connection values, the characteristic path lengths and the cluster coefficients of the whole brain and hemisphere scale calculated by the method take the time scale into consideration and the space scale into consideration.
The index analysis module is used for evaluating the functional state of the brain according to a plurality of indexes of the functional connection value, the characteristic path length and the clustering coefficient of the whole brain and the hemisphere scale when the left hand and right hand of each electroencephalogram sub-segment and the complete electroencephalogram signal segment are in motor imagery, and specifically comprises:
normalizing all indexes of the functional connection values of the whole brain and hemisphere scales, the characteristic path length and the clustering coefficient of each electroencephalogram sub-segment and the whole electroencephalogram signal segment under the stimulation of left and right hand motor imagery to sequentially form a one-dimensional vector X (X) 1 ，x 2 ，...，x p ) P represents the number of indexes; obtaining Euclidean distance value according to a formula (11), wherein the Euclidean distance value represents the degree of the brain function difference between the current tested person and the healthy person;the value range is [0, + ] infinity]The formula is as follows:
wherein (y) 1 ，y 2 ，...，y p ) Representation and (x) 1 ，x 2 ，...，x p ) Vector of each index vector of the corresponding healthy person is averaged;
and the evaluation result visualization module is used for outputting all indexes and index analysis results obtained by the tested multi-scale brain network index calculation and analysis module.
The beneficial effects of the invention are as follows:
1. the device is based on the idea of cross frequency band coupling, integrates the phase coupling relation between the frequency band and the frequency band to construct a brain network, and compared with the conventional brain network analysis method in a specific single frequency band, the brain network is larger and contains more information, so that the brain function state of a tested brain can be effectively depicted.
2. During brain network analysis, effective metrics are extracted from multiple temporal, spatial scales for stroke assessment.
3. The device can objectively assist brain function assessment of a stroke patient, and has a certain practical significance.
Drawings
FIG. 1 is a schematic diagram of the functional modules of the device of the present invention;
FIG. 2 is a schematic diagram of a specific structure of a cross frequency functional connection matrix calculation module and a multi-scale brain network index calculation and analysis module of the device of the present invention;
FIG. 3 is a diagram showing the position of an electroencephalogram electrode according to an embodiment of the present invention;
FIG. 4 is a schematic diagram of a functional connection matrix of delta and alpha frequency bands constructed under the stimulation of left motor imagery of a stroke subject and a brain network with two-by-two cross frequencies thereof: the functional connection matrix (A) is a 56 multiplied by 56 functional connection matrix formed by splicing a functional connection matrix in delta and alpha frequency bands and a functional connection matrix between corresponding frequency bands, and the functional connection matrix (B) is a pairwise cross frequency brain network schematic diagram constructed after elements in 70% of the functional connection matrix are reserved.
Detailed Description
A stroke brain function assessment device based on EEG cross frequency coupling according to the present invention will be described in detail with reference to the accompanying drawings.
EEG cross frequency coupling based stroke brain function assessment device module schematic, as shown in FIG. 1; the specific structure of the cross frequency function connection matrix calculation module and the multi-scale brain network index calculation and analysis module is shown in fig. 2. EEG cross frequency coupling based stroke brain function assessment device mainly comprises:
and the electroencephalogram data acquisition module is used for performing left-hand and right-hand motor imagination on a plurality of stroke testes and health testes under a proper environment and performing electroencephalogram data acquisition by using 30-channel electroencephalogram acquisition equipment. The experimental stimulation materials are left hand and right hand moving pictures, and the E-prime is used for presenting the stimulation pictures. The tested person needs to judge whether the picture is left hand or right hand according to the randomly presented picture, and the judgment is made by pressing a key. The experiment was set up with multiple stimulations, each at 5s intervals. The electroencephalogram electrode channels are distributed according to an international 10-20 system, as shown in fig. 3, wherein 30 black channels (FP 1, FP2, F7, F3, fz, F4, F8, FC5, FC1, FC2, FC6, T7, C3, cz, C4, T8, TP9, CP5, CP1, CP2, CP6, TP10, P7, P3, pz, P4, P8, O1, oz, O2) are electroencephalogram channels collected by the experimental setup. During acquisition, scalp impedance is controlled within 5kΩ, and sampling rate is 1000Hz.
And the electroencephalogram data preprocessing frequency division module is used for preprocessing the original electroencephalogram data acquired by the electroencephalogram data acquisition module to improve the signal-to-noise ratio of the electroencephalogram data. After the original electroencephalogram data of each tested place is subjected to band-pass filtering to 0.1-30 Hz and is subjected to ica algorithm to remove the artifacts such as the electrooculogram and the like Guan Shengli, double-side mastoid (TP 9, TP 10) is re-referenced, the electroencephalogram data of which the picture is at the front of-200 ms-0ms is used as a base line to carry out base line correction to obtain the electroencephalogram data of 28 channels except TP9 and TP10, and then the electroencephalogram data is filtered to delta (0.1-4 Hz), theta (4 Hz-8 Hz), alpha (8 Hz-12 Hz), low beta (12 Hz-20 Hz) and high beta (20 Hz-28 Hz) by using a band-pass filter. And (3) carrying out data segmentation on each tested brain data after the processing, extracting a brain data segment of 0-1000ms after each stimulation picture is presented as a complete brain data segment of motor imagery, and dividing the brain data segment into 3 subfragments, namely brain data subfragments of 0-300ms, 300-800ms and 800-1000ms, for subsequent analysis.
The cross frequency function connection matrix calculation module comprises a first calculation module and a second calculation module.
A first calculation module: and carrying out Hilbert transformation on the electroencephalogram data on each channel of each tested electroencephalogram sub-segment and the complete electroencephalogram segment processed by the electroencephalogram data preprocessing frequency division module according to a formula (1), and then calculating according to a formula (2) to obtain an instantaneous phase sequence phi (t) of the electroencephalogram signal x (t) on each channel.
The second calculation module is used for calculating a functional connection matrix in the frequency band and a functional connection matrix between the frequency bands; the method specifically comprises the following steps:
for the same frequency band, delta, alpha, low beta and high beta frequency bands are specifically selected, and the phase coupling relation between signals is calculated according to a formula (3) for any two channel signals of each electroencephalogram sub-segment and the complete electroencephalogram signal segment to be tested. Center frequency band f of two channel signals within the same frequency band m ＝f n Calculating the phase difference delta phi (f) between the two channel signals according to the formula (4) m ，f n ，t)。
For different frequency bands, specifically selecting a delta-alpha frequency band, a delta-low beta frequency band and a delta-high beta frequency band, and calculating the phase coupling relation between any two channel signals of each electroencephalogram signal sub-segment and the complete electroencephalogram signal segment to be tested according to a formula (3). Center frequency band f of two channel signals between different frequency bands m ≠f n . According to the filtered frequency range, the relation between 5 center frequencies is calculated to be 1:3:5:8:12. When delta-alpha calculation is adopted, the center frequency ratio of any two channel signals is 1:5, and n and m are taken as 5 and 1. When delta-low beta calculation is adopted, the center frequency ratio of any two channel signals is 1:8, and n and m are taken as 8 and 1. When delta-high beta is selected for calculation, the center frequency ratio of any two channel signals is 1:12, and n and m are taken as 12 and 1. Calculating the phase difference delta phi (f) between the two channel signals according to the formula (4) m ，f n ，t)。
For each tested brain obtained aboveConstructing a 28 x 28 intra-frequency band functional connection matrix according to the phase-to-phase coupling relation in each single frequency band of the electric signal sub-segment and the complete electroencephalogram signal segment; and constructing a 28 multiplied by 28 inter-frequency band functional connection matrix for the obtained phase coupling relation between each different frequency band of each tested electroencephalogram signal sub-segment and the whole electroencephalogram signal segment. Each element in the intra-band functional connection matrix and inter-band functional connection matrix is PSI (f) of a two-channel signal m ，f n )。
The multi-scale brain network index calculation and analysis module comprises a functional connection value index calculation module, a two-to-two cross frequency brain network index calculation module and an index analysis module.
The function connection value index calculation module comprises function connection indexes of alpha, low beta single frequency bands and delta-alpha, delta-low beta and delta-high beta under left and right hand motor imagery stimulus, wherein the function connection indexes comprise a full brain function connection value and a function connection value on a hemisphere scale; the method specifically comprises the following steps:
according to the frequency band internal function connection matrix and the frequency band internal function connection matrix of each brain signal sub-segment and the complete brain signal segment which are tested in multiple left hand and right hand motor imagery and are obtained by calculation of the cross frequency function connection matrix calculation module, firstly, according to the types of the left hand and the right hand stimulation, the internal function connection matrix and the frequency band internal function connection matrix are averaged to obtain the internal function connection matrix and the frequency band internal function connection matrix of each brain signal sub-segment and the complete brain signal segment under the stimulation of each tested left hand and right hand motor imagery.
PSI (f) between all channels on each intra-frequency functional connection matrix and inter-frequency functional connection matrix for each electroencephalogram signal sub-segment and complete electroencephalogram signal segment under left and right hand motor imagery stimulus m ，f n ) And (3) respectively carrying out averaging treatment on the values to obtain the full brain function connection values in a single frequency band and between the two frequency bands under the stimulation of the two motor imagery. In the analysis of the data of the full brain function connection values of the multi-bit stroke test and the healthy test, the stroke test is in the alpha, low beta frequency band and the delta-alpha, delta-low beta and delta-high beta frequency bandsThe brain function connection value is obviously weaker than that of a healthy tested (p is less than 0.05), namely, the brain function connection value can effectively measure the functional state of the brain to a certain extent.
According to the electroencephalogram electrode position distribution diagram shown in fig. 3, the left hemisphere and the right hemisphere respectively have 12 electrode channels, and the left brain channel is: fp1, F3, F7, FC1, FC5, C3, T7, CP1, CP5, P3, P7, O1, right brain channel is: fp2, F4, F8, FC2, FC6, C4, T8, CP2, CP6, P4, P8, O2. According to the positions of the left brain channel and the right brain channel, calculating functional connection values of each electroencephalogram sub-segment and each frequency band of the complete electroencephalogram signal segment under the stimulation of the left and right hand motor imagery of each tested site and on hemispherical scales between the frequency bands; the method specifically comprises the following steps:
PSI (f) between the right brain channels on each intra-frequency functional connection matrix and inter-frequency functional connection matrix of each electroencephalogram signal sub-segment and complete electroencephalogram signal segment under each tested left-hand motor imagery stimulus type m ，f n ) And (3) respectively carrying out averaging treatment on the values to obtain right hemisphere function connection values in each frequency band and between frequency bands under the stimulation of two motor imagery.
PSI (f) between left brain channels on each intra-frequency functional connection matrix and inter-frequency functional connection matrix of each electroencephalogram sub-segment and complete electroencephalogram segment under each tested right-hand motor imagery stimulus type m ，f n ) And (3) respectively carrying out averaging treatment on the values to obtain left hemisphere function connection values in each frequency band and between frequency bands under the stimulation of two motor imagery.
PSI (f) between left brain channel and right brain channel on each intra-frequency functional connection matrix and inter-frequency functional connection matrix of each electroencephalogram signal sub-segment and complete electroencephalogram signal segment under each tested two motor imagery stimulus types m ，f n ) And (3) respectively carrying out averaging treatment on the values to obtain left and right hemisphere function connection values in each frequency band and between frequency bands under the stimulation of two motor imagery.
In analysis of hemispherical functional connection value data of a multi-bit stroke test and a healthy test, hemispherical functional connection value differences between the alpha frequency band, the low beta frequency band and the delta-alpha frequency band, the delta-low beta frequency band and the delta-high beta frequency band of the stroke test and the healthy test are obvious.
The two-by-two cross frequency brain network index calculation module is used for combining the delta, alpha, low beta and high beta frequency band functional connection matrix and the delta-alpha, delta-low beta and delta-high beta frequency band functional connection matrix, and calculating the characteristic path length and the clustering coefficient index of the two-by-two cross frequency brain network when the left hand and the right hand move imagine; the method specifically comprises the following steps:
and splicing the delta and alpha intra-frequency band functional connection matrixes of each electroencephalogram signal sub-segment and the complete electroencephalogram signal segment under the stimulation type of the left and right hand motor imagery under each tested position and the inter-frequency band functional connection matrix of delta-alpha. And splicing the functional connection matrix between the frequency bands of the delta and low beta of each electroencephalogram signal sub-segment and the complete electroencephalogram signal segment under the stimulation type of the left and right hand motor imagery under each tested bit. And splicing the functional connection matrix between the frequency bands of the delta and the high beta of each electroencephalogram signal sub-segment and the complete electroencephalogram signal segment under the stimulation type of the left and right hand motor imagery under each tested bit. Each function connection matrix after splicing is 56 multiplied by 56, two function connection matrices in the frequency band are arranged on the diagonal, and the non-diagonal is the corresponding function connection matrix between the frequency bands. And reserving elements of the functional connection matrix meeting a threshold range as edges of the pairwise cross frequency brain network. The function connection value in the single frequency band is larger than the function connection value between different frequency bands, so that the function connection values in the frequency bands and between the frequency bands are reserved at the same time, and therefore the threshold values are respectively arranged on the function connection matrix in the single frequency band and the function connection matrix between the frequency bands. And arranging all element values in each 28×28 frequency band functional connection matrix and each inter-frequency band functional connection matrix from large to small, determining a threshold value as a specific value when the first 70% of elements are reserved, and reserving the elements larger than the corresponding threshold value in each frequency band functional connection matrix and each inter-frequency band functional connection matrix and taking the elements as edges of the pairwise cross frequency brain network. FIG. 4 shows a schematic diagram of delta and alpha functional connection matrix and a pairwise cross frequency brain network for left hand motor imagery of a stroke, wherein the delta and alpha functional connection matrix is a symmetric matrix, the delta and alpha intra-band functional connection matrix with the size of 28×28 on the diagonal, and the delta-alpha inter-band functional connection matrix with the size of 28×28 on the off-diagonal.
And calculating characteristic path lengths of the pairwise crossover frequency brain network according to a formula (7), and calculating clustering coefficients according to a formula (8), a formula (9) and a formula (10). In the index analysis of brain networks with the frequencies of delta and alpha, delta and low beta, delta and high beta of multi-bit stroke test and healthy test, the characteristic path length of the stroke test is obviously larger than that of the healthy test, and the clustering coefficient is obviously smaller than that of the healthy test (p < 0.05).
The index analysis module normalizes all indexes of all brain and hemisphere scale function connection values, characteristic path lengths and clustering coefficients of each brain electric signal sub-segment and complete brain electric signal segment under the stimulation of left and right hand motor imagery to sequentially form a one-dimensional vector X (X) 1 ，x 2 ，...，x p ) P represents the number of indexes. According to formula (11), calculating the vector obtained by averaging each tested index vector and the corresponding index vector of the plurality of healthy peopleEuclidean distance value of (c)Multiple stroke trial->Larger, indicating impaired brain function.
And the evaluation result visualization module is used for outputting all indexes and index analysis results obtained by the multi-scale brain network index calculation and analysis module of each tested.
The above embodiments are not intended to limit the present invention, and the present invention is not limited to the above embodiments, and falls within the scope of the present invention as long as the present invention meets the requirements.
Claims (7)
1. An EEG cross frequency coupling-based stroke brain function assessment device is characterized by comprising the following modules in detail:
the brain electrical data acquisition module is used for acquiring multichannel brain electrical data of a healthy state or a cerebral apoplexy state when a tested person moves imagination in left hand and right hand;
the electroencephalogram data preprocessing frequency division module is used for preprocessing electroencephalogram data acquired by the electroencephalogram data acquisition module to obtain a complete electroencephalogram signal segment and a multi-segment electroencephalogram signal sub-segment under left and right hand motor imagery stimulus of five frequency bands of delta, theta, alpha, lowbeta and high beta;
the cross frequency function connection matrix calculation module is used for calculating the phase coupling relation between each electroencephalogram signal sub-segment and the complete electroencephalogram signal segment in the frequency band and between the frequency bands after being processed by the electroencephalogram data preprocessing frequency division module, so as to construct a function connection matrix in the frequency band and a function connection matrix between the frequency bands of each electroencephalogram signal sub-segment and the complete electroencephalogram signal segment; the system comprises a first computing module and a second computing module;
the multi-scale brain network index calculation and analysis module is used for calculating the functional connection value index on the whole brain and hemisphere scale under the left and right hand motor imagery stimulus and the two-to-two cross frequency brain network index for each brain signal sub-segment and the frequency band functional connection matrix of the complete brain signal segment constructed by the cross frequency functional connection matrix calculation module, and then carrying out index analysis on the functional connection value index; the system comprises a function connection value index calculation module, a pairwise cross frequency brain network index calculation module and an index analysis module;
the evaluation result visualization module is used for outputting all indexes and index analysis results obtained by the multi-scale brain network index calculation and analysis module;
the function connection value index calculation module is used for calculating function connection value indexes in alpha, lowbeta single frequency bands and between delta-alpha, delta-low beta and delta-high beta under the stimulation of left and right hand motor imagery, and comprises a full brain function connection value and a function connection value on a hemisphere scale; the method specifically comprises the following steps:
1) For each electroencephalogram signal sub-segment and complete electroencephalogram signal segment under left and right hand motor imagery stimulusPSI (f) between all channels on intra-band functional connection matrix and inter-band functional connection matrix m ,f n ) The values are respectively subjected to averaging treatment to obtain a full brain function connection value;
2) Calculating functional connection values on hemispherical scales for functional connection matrixes and inter-frequency-band functional connection matrixes in each electroencephalogram signal sub-segment and each complete electroencephalogram signal segment under the stimulation of left and right hand motor imagery; the functional connection values on the hemispherical scale comprise a left hemispherical functional connection value, a right hemispherical functional connection value and left and right hemispherical functional connection values; the method specifically comprises the following steps:
under the stimulation type of left hand motor imagery, PSI (f) between the functional connection matrix in each frequency band and the right brain channel on the functional connection matrix between the frequency bands m ,f n ) The values are respectively subjected to averaging treatment to obtain right hemisphere function connection values;
under the stimulation type of the motor imagery of the right hand, PSI (f) between the functional connection matrix in each frequency band and the left brain channel on the functional connection matrix between the frequency bands m ,f n ) The values are respectively subjected to averaging treatment to obtain left hemisphere function connection values;
under two motor imagery stimulus types, PSI (f) between a left brain channel and a right brain channel on a functional connection matrix in each frequency band and a functional connection matrix between frequency bands m ,f n ) The values are respectively subjected to averaging treatment to obtain left and right hemisphere function connection values;
the two-by-two cross frequency brain network index calculation module is used for combining the delta, alpha, low beta and high beta frequency band functional connection matrix and the delta-alpha, delta-lowbeta, delta-high beta frequency band functional connection matrix to calculate the characteristic path length and the clustering coefficient index of the two-by-two cross frequency brain network under the left and right hand motor imagery stimulus; the method specifically comprises the following steps:
1) Splicing the functional connection matrixes in any two frequency bands of each electroencephalogram signal sub-segment and the complete electroencephalogram signal segment under the left and right hand motor imagery stimulus types with the corresponding functional connection matrixes between the frequency bands to construct a functional connection matrix with the size of 2N multiplied by 2N; the diagonal of the functional connection matrix is two intra-frequency-band functional connection matrices, and the non-diagonal is a corresponding inter-frequency-band functional connection matrix; reserving each element meeting the threshold range in the functional connection matrix and serving as the edge of the 2N node pairwise cross frequency brain network; wherein the node refers to a channel;
2) Calculating characteristic path length indexes for the pairwise crossover frequency brain network:
the characteristic path length is used for describing the efficiency and the capability of global function integration and information transmission of the brain network, is defined as the average value of the shortest path between any two nodes in the brain network, and has the following formula:
wherein N represents the number of nodes in the brain network, l ij Representing the shortest path length between node i and node j;
3) Calculating a clustering coefficient index for the pairwise crossover frequency brain network:
the clustering coefficient is used for describing the function differentiation capability of the local part of the brain network, and the clustering coefficient C of the whole brain network defines the clustering coefficient C of all nodes i The formula is shown below:
wherein N represents the number of nodes, C i Clustering coefficients, k, representing node i i The degree of the node i is indicated,representing the weight of the edge between the node a and the node b in the brain network after the scaling, w a,b Representing the weight of the edge between the node a and the node b in the brain network without scaling, and max (w) represents the maximum value of the weight of the edge of any two nodes in the brain network;
the index analysis module is used for evaluating the functional states of the brain according to the functional connection values, the characteristic path length and the cluster coefficients of the whole brain and the hemisphere scale under the left and right hand motor imagery stimulus of each electroencephalogram signal sub-segment and the whole brain signal segment.
2. The apparatus of claim 1 wherein the delta band is from 0.1 to 4hz, the theta band is from 4 to 8hz, the alpha band is from 8hz to 12hz, the lowbeta band is from 12hz to 20hz, and the highbeta band is from 20hz to 28hz.
3. The device according to claim 1, wherein the preprocessing in the electroencephalogram data preprocessing frequency division module specifically comprises the steps of removing the electrooculogram, re-referencing, correcting a base line, filtering and dividing the electroencephalogram data.
4. The apparatus of claim 1, wherein the cross-frequency functional connection matrix calculation module comprises a first calculation module and a second calculation module;
the first calculation module is used for calculating the instantaneous phase sequence of each electroencephalogram signal sub-segment and the complete electroencephalogram signal segment; specifically, calculating an instantaneous phase sequence phi (t) of an electroencephalogram signal x (t) on each channel through a Hilbert transformation formula on the electroencephalogram signal processed by the electroencephalogram data preprocessing frequency division module;
the second calculation module is used for calculating a functional connection matrix in the frequency band and a functional connection matrix between the frequency bands.
5. The apparatus of claim 4 wherein the Hilbert transform formula is as follows:
wherein the method comprises the steps ofAnd the PV is a Keke main value which is the result of the Hilbert transformation of the electroencephalogram signal x (t) at the moment t on a certain channel.
6. The apparatus of claim 4, wherein the second computing module is specifically configured to:
selecting delta, alpha, low beta and high beta frequency bands, and aiming at the same frequency band, setting the center frequency of any two channels as f m And f n F is a signal of (f) m ＝f n Calculate the phase-to-phase coupling relation PSI (f) m ,f n ) I.e. the functional connection value between two channels, the range of values being 0,1]The calculation formula is as follows:
wherein T represents the number of time samples of the signal; j represents an imaginary number; Δφ (f) m ,f n T) represents two center frequencies f m And f n The calculation method of the instantaneous phase difference of the signals is as follows:
Δφ(f m ,f n ,t)＝φ(f m ,t)-φ(f n t) type (4)
Selecting delta-alpha two frequency bands, delta-lowbeta two frequency bands and delta-highbeta two frequency bands, and aiming at different frequency bands, setting the center frequency of any two channels as f m And f n F is a signal of (f) m ≠f n Calculating the phase-coupling relation PSI (f) between two signals according to the formula (3) m ,f n ) The method comprises the steps of carrying out a first treatment on the surface of the Root of Chinese characterCalculating two center frequencies as f according to formula (5) m And f n Wherein n and m are the smallest positive integers satisfying the requirement of equation (6);
Δφ(f m ,f n ,t)＝nφ(f m ,t)-mφ(f n t) type (5)
n×f m ＝m×f n (6)
Constructing an N multiplied by N frequency band internal function connection matrix for the phase coupling relation between each acquired electroencephalogram signal sub-segment and each single frequency band internal channel signal of the complete electroencephalogram signal segment; constructing an N multiplied by N inter-band functional connection matrix for the phase coupling relationship between the obtained electroencephalogram signal sub-segments and the channel signals between each different frequency band of the complete electroencephalogram signal segment; n represents the number of channels.
7. The apparatus of claim 1, wherein the index analysis module is specifically configured to:
normalizing all indexes of the full brain and hemisphere scale function connection values, the characteristic path length and the clustering coefficient of each electroencephalogram signal sub-segment and the full electroencephalogram signal segment under the left and right hand motor imagery stimulus to sequentially form a one-dimensional vector X (X) 1 ,x 2 ,…,x p ) P represents the number of indexes; obtaining Euclidean distance value according to a formula (11), wherein the Euclidean distance value represents the degree of the brain function difference between the current tested person and the healthy person;the value range is [0, + ] infinity]The formula is as follows:
wherein (y) 1 ,y 2 ,…,y p ) Representation and (x) 1 ,x 2 ,…,x p ) Vector of the vector of each index of the corresponding healthy person is averaged.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111522093.0A CN114145754B (en) | 2021-12-13 | 2021-12-13 | EEG cross frequency coupling-based stroke brain function assessment device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111522093.0A CN114145754B (en) | 2021-12-13 | 2021-12-13 | EEG cross frequency coupling-based stroke brain function assessment device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114145754A CN114145754A (en) | 2022-03-08 |
| CN114145754B true CN114145754B (en) | 2024-04-02 |
Family
ID=80450629
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111522093.0A Active CN114145754B (en) | 2021-12-13 | 2021-12-13 | EEG cross frequency coupling-based stroke brain function assessment device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114145754B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115317000A (en) * | 2022-08-31 | 2022-11-11 | 燕山大学 | Multilayer brain network model construction method based on same frequency and cross-frequency information interaction |
| CN116058851B (en) * | 2023-02-20 | 2023-07-04 | 之江实验室 | Electroencephalogram data processing method, electroencephalogram data processing device, electroencephalogram data analysis system, electronic device and electroencephalogram data processing medium |
-
2021
- 2021-12-13 CN CN202111522093.0A patent/CN114145754B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN114145754A (en) | 2022-03-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114145754B (en) | EEG cross frequency coupling-based stroke brain function assessment device | |
| Chen et al. | A novel EEMD-CCA approach to removing muscle artifacts for pervasive EEG | |
| CN105595961A (en) | Alzheimer's disease detecting system and method based on electroencephalogram signals | |
| Wang et al. | Functional brain networks in Alzheimer’s disease: EEG analysis based on limited penetrable visibility graph and phase space method | |
| WO2006060727A9 (en) | System and method for eeg imaging of cerebral activity using electrode sets | |
| CN103258120A (en) | Apoplexy recovery degree index calculation method based on brain electrical signals | |
| CN103110418A (en) | Electroencephalogram signal characteristic extracting method | |
| CN105342605A (en) | Method for removing myoelectricity artifacts from brain electrical signals | |
| CN110584660A (en) | Electrode selection method based on brain source imaging and correlation analysis | |
| CN111544256A (en) | Brain-controlled intelligent full limb rehabilitation method based on graph convolution and transfer learning | |
| CN109480832A (en) | The removing method of Muscle artifacts in a kind of single pass EEG signals | |
| CN110916656A (en) | Multichannel pelvic floor muscle strength evaluation system and method | |
| Liu et al. | FPGA-based real-time compressed sensing of multichannel EEG signals for wireless body area networks | |
| Kang et al. | Distortions in EEG interregional phase synchrony by spherical spline interpolation: causes and remedies | |
| CN106236080A (en) | Based on the removing method of myoelectricity noise in multichannel EEG signals | |
| CN105249962B (en) | The retrospective epileptic attack point detecting method of scalp EEG signals and system | |
| Diab et al. | Nonlinear estimation of coupling and directionality between signals: Application to uterine EMG propagation | |
| CN113558637B (en) | Music perception brain network construction method based on phase transfer entropy | |
| Fang et al. | A low power biomedical signal processing system-on-chip design for portable brain-heart monitoring systems | |
| CN107092925B (en) | Cerebral function magnetic resonance imaging blind source separation method based on SIM algorithm in groups | |
| CN110786851A (en) | Method for improving wavelet packet decomposition speed based on Mallat algorithm | |
| CN116269445A (en) | Accurate target identification method for SSVEP short time window signal | |
| CN116595437A (en) | Training method, device and storage medium for zero calibration transfer learning classification model | |
| CN108814593A (en) | A kind of EEG feature extraction method based on complex network | |
| Huang et al. | Brain functional network based on mutual information analysis of EEGs and its application to schizophrenia |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |