CN111419221A - Electroencephalogram signal analysis method based on graph convolution network - Google Patents
Electroencephalogram signal analysis method based on graph convolution network Download PDFInfo
- Publication number
- CN111419221A CN111419221A CN202010093974.4A CN202010093974A CN111419221A CN 111419221 A CN111419221 A CN 111419221A CN 202010093974 A CN202010093974 A CN 202010093974A CN 111419221 A CN111419221 A CN 111419221A
- Authority
- CN
- China
- Prior art keywords
- brain
- electroencephalogram
- lead
- network
- electroencephalogram signal
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000004458 analytical method Methods 0.000 title claims abstract description 25
- 210000004556 brain Anatomy 0.000 claims abstract description 42
- 230000003925 brain function Effects 0.000 claims abstract description 30
- 238000000034 method Methods 0.000 claims abstract description 17
- 238000002372 labelling Methods 0.000 claims abstract description 3
- 239000011159 matrix material Substances 0.000 claims description 24
- 230000006870 function Effects 0.000 claims description 15
- 238000010586 diagram Methods 0.000 claims description 11
- 238000004364 calculation method Methods 0.000 claims description 6
- 238000001228 spectrum Methods 0.000 claims description 4
- 230000004913 activation Effects 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 3
- 201000010099 disease Diseases 0.000 abstract description 10
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 abstract description 10
- 210000000653 nervous system Anatomy 0.000 description 3
- 208000012902 Nervous system disease Diseases 0.000 description 2
- 208000025966 Neurological disease Diseases 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 230000002269 spontaneous effect Effects 0.000 description 2
- 230000004931 aggregating effect Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000007177 brain activity Effects 0.000 description 1
- 210000004958 brain cell Anatomy 0.000 description 1
- 230000002490 cerebral effect Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 210000003792 cranial nerve Anatomy 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000005684 electric field Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000005036 nerve Anatomy 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000001020 rhythmical effect Effects 0.000 description 1
- 210000004761 scalp Anatomy 0.000 description 1
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/369—Electroencephalography [EEG]
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/40—Detecting, measuring or recording for evaluating the nervous system
- A61B5/4029—Detecting, measuring or recording for evaluating the nervous system for evaluating the peripheral nervous systems
- A61B5/4041—Evaluating nerves condition
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/40—Detecting, measuring or recording for evaluating the nervous system
- A61B5/4058—Detecting, measuring or recording for evaluating the nervous system for evaluating the central nervous system
- A61B5/4064—Evaluating the brain
-
- 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
- A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
- A61B5/7267—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems involving training the classification device
Abstract
The invention relates to an electroencephalogram signal analysis method, which comprises the following steps: s1: establishing a brain functional network according to the electroencephalogram signal; s2: labeling different brain function networks according to the brain function networks to complete the training of the convolutional network model; s3: and according to the trained graph convolution network model, carrying out type recognition on the new brain function network to finish electroencephalogram classification. Compared with the prior art, the electroencephalogram signal analysis method can accurately describe the detail characteristics and the state of the brain, and can provide a brand-new method for intelligent judgment, early warning and treatment of the diseases by applying the method to the brain functional network of the patient with the diseases related to the brain nervous system.
Description
Technical Field
The invention belongs to the field of electroencephalogram signal analysis, and particularly relates to an electroencephalogram signal analysis method based on a graph convolution network.
Background
Electroencephalograms (EEG) are patterns obtained by recording spontaneous biopotentials of the brain from the scalp by means of a precise electronic instrument in an enlarged manner, and are spontaneous and rhythmic electrical activities of brain cell groups recorded by means of electrodes. The electroencephalogram signal has higher time precision and can dynamically observe the state change of the brain, so the electroencephalogram signal is an important tool for diagnosing different neurological disorders and diseases. The interaction and the mutual association between the neurons of the brain make the brain become a complex kinetic system. The brain is considered as a functional network, and the time correlation among nerve clusters of different areas of the brain is researched, so that the exploration of the relation among the different areas of the brain is a method of a brain dynamic system. At present, a method for establishing a brain function network generally establishes a functional network according to a pearson correlation coefficient of EEG signals acquired between different electrodes, or establishes a functional network according to a phase-locked value of EEG signals between different electrodes, so that the discharge characteristics of the EEG signals at different moments are ignored, and the established brain function network cannot better reflect the real state of the brain. In addition, the traditional electroencephalogram signal analysis method is also generally used for analyzing the EEG signal difference between a patient with brain nervous system related diseases and a normal person from the characteristics (amplitude, frequency spectrum and phase) of the signal per se, and the EEG signal cannot be analyzed through the whole communication relation of the brain, so that some high-order hidden valuable information of a brain network is ignored.
Disclosure of Invention
The invention aims to provide an electroencephalogram signal analysis method based on a graph convolution network.
The invention is realized by the following technical scheme:
an electroencephalogram signal analysis method based on a graph convolution network, wherein electroencephalogram signals are collected by a plurality of lead electrodes, and the method comprises the following steps:
s1: establishing a brain functional network according to an electroencephalogram signal, comprising the following specific steps:
s11: the power of the whole brain domain GFP is calculated,
wherein K is the total number of conducting electrodes of the leads and represents K nodes, i is the ith lead, and Vi(t) EEG signal for the ith lead, VmeanAnd (t) is the average value of K lead signals at the t moment.
S12: the peak time of GFP is obtained and is recorded as the time series t ═ t1,t2,t3...tn]；
S13: in thatTime series [ t ]1,t2,t3...tn]At each time ti(i∈[1,n]) Calculating the phase locking value between each two paths of lead electroencephalogram signals within a time range of the central time;
s14: reflecting the relation between the nodes according to the phase locking value index between each two paths of lead electroencephalogram signals so as to establish a brain function network.
Compared with the prior art, the electroencephalogram signal analysis method can accurately describe the detail characteristics and the state of the brain
Further, the calculation formula of the graph convolution network in step S2 is as follows:
wherein
To add a self-connected degree matrix to each node in the brain function network map,
adding the unit matrix I, H to the adjacent matrix of the brain function network diagram(l+1)Outputs results for l +1 layers of GCN, H(l)Is a GCN output of l layers, H(0)For initialising the feature matrix of the node, W(l)The parameter moment to be trained of the layer l, wherein l is an integer greater than or equal to 0 and represents the layer of the GCN, and sigma (-) is an activation function, and a Re L U function or a Sigmoid function can be selected.
Further, the initialized feature matrix H of the step S2(0)The electroencephalogram signal characteristics of each path of lead node comprise an initialization characteristic matrix which is composed of one or more characteristics of amplitude characteristic values, frequency characteristic values, power spectrum characteristic values and brain area position characteristics of the node electroencephalogram signals.
Further, the graph convolution network training model of step S2 is classified by softmax function, and then W is completed by using cross entropy loss function back propagation(l)And (5) parameter training.
Further, the method can be used for preparing a novel materialThe phase-locked value P L V is calculated by the following formula:
wherein
Is the phase relation value of two-way lead EEG signal, N is time sequence [ t1,t2,t3...tn]At each time ti(i∈[1,n]) The number of sampling points in a time range of the central time.
Further, the phase relation value of the two-way lead electroencephalogram signals
May be calculated by time-frequency analysis of the signal.
In order that the invention may be more clearly understood, specific embodiments thereof will be described hereinafter with reference to the accompanying drawings.
Drawings
Fig. 1 is a flow chart of a method of electroencephalographic signal analysis of the present invention.
FIG. 2 is a diagram showing the time of the peak power in the whole brain domain
Fig. 3 is a brain function network creation representation.
Fig. 4 is a diagram of a brain function network.
Detailed Description
The invention provides an electroencephalogram signal analysis method based on a graph convolution network, which can effectively analyze a brain functional network. Referring to fig. 1, the electroencephalogram signal analysis method of the present invention includes the following steps:
s1: a brain function network is established from the electroencephalogram signals.
Since electroencephalogram information is typically acquired from multiple lead electrodes, an associative relationship between different electrodes can be established, thereby establishing a brain functional network. Defining the measured area of each lead electrode as a node in the graph, abstracting the multipath EEG signal into a graph G consisting of a set of points and a set of edges: g ═ V, E, where V is the set of nodes corresponding to the lead nodes for EEG brain number acquisition and E is the connected edge.
The Global Field Power (GFP) represents the electric field strength on each brain at an instant, and is therefore generally used for measuring the response of the Global brain to an event or characterizing the rapid change of brain activity, and the peak position of a GFP curve represents the instant of the strongest field strength and the highest signal-to-noise ratio terrain.
S11: the power of the whole brain domain GFP is calculated,
wherein K is the total number of the conducting electrodes which represent K nodes, i is the ith lead, and Vi(t) EEG signal for the ith lead, VmeanAnd (t) is the average value of K lead signals at the t moment.
S12: the peak time of GFP is obtained and is recorded as the time series t ═ t1,t2,t3...tn]。
As shown in FIG. 2, [ t ]1,t2,t3...tn]The moment of alignment is the peak moment of the GFP curve.
S13: in the time series [ t ]1,t2,t3...tn]At each time ti(i∈[1,n]) And calculating the phase locking value between each two paths of lead electroencephalogram signals within a time range of the central time.
The calculation formula of the phase-locked value P L V is as follows:
wherein
Phase gating for two-way lead brain electrical signalsCoefficient, N is a time series [ t ]1,t2,t3...tn]At each time ti(i∈[1,n]) The number of sampling points in a time range of the central time. And the phase relation value of two-way lead EEG signal
The calculation can be performed by time-frequency analysis of the signal, including short-time fourier transform, wavelet transform, hilbert-yellow transform, and Cohen-like time-frequency distribution.
S14: reflecting the relation between the nodes according to the phase locking value index between each two paths of lead electroencephalogram signals so as to establish a brain function network.
The time series t is calculated in step S131,t2,t3...tn]Each time ti(i∈[1,n]) The average value of the multiple phase-locked values or the maximum value thereof is taken as the phase-locked value index between each two paths of lead electroencephalogram signals to describe the relation between two nodes, so that a weighted brain function network diagram can be established, and the weight value between the nodes is the phase-locked value index; and an unauthorized brain function network graph can be established by comparing the phase-locked value index with a threshold value, judging that the nodes have a connection relation if the phase-locked value index is greater than or equal to the threshold value, or judging that the nodes have no connection relation if the phase-locked value index is less than the threshold value, thereby establishing the unauthorized brain function network graph.
For the brain functional network diagram G, the unweighted adjacency matrix of the diagram is A, and the element α thereofijThe calculation is as follows:
if the phase lock value P L V between EEG signals of any two nodes i and jijIf the value is greater than or equal to the threshold value, the nodes i and j in the brain area are considered to have a connection relation, and the element α corresponding to the adjacent matrix W is adjacentij1, otherwise αij0, indicates the absence or presence between nodes i and j of the brain regionThere is a connection relationship.
The whole process of establishing the brain function network diagram is shown in fig. 3, the phase-locked value of each two paths of electroencephalogram signals is calculated to obtain an adjacency matrix with weight, and then the phase-locked value is compared with a threshold value to establish a non-weight adjacency matrix of the brain network connection diagram. As shown in fig. 4, a functional brain network is established for 128-lead electroencephalogram signals.
S2: and labeling different brain function networks according to the brain function networks to finish the training of the convolutional network model.
The core idea of the Graph Convolution Network (GCN) is to generate a new "node representation" by "aggregating" node information "with" side information "to generate a new" node representation "and the node can transmit and receive messages, so that the GCN can find the internal detail association of data from the association of global data, and thus can describe graph data more in detail and deeply, compared with the conventional graph analysis method.
The calculation formula of each layer graph convolution network is as follows:
wherein
To add a self-connected degree matrix to each node in the brain function network map,
adding an identity matrix I, H to an adjacency matrix of a brain function network diagram(l+1)Outputs results for l +1 layers of GCN, H(l)Is a GCN output of l layers, H(0)For initialising a feature matrix of nodes, W(l)The parameter moment to be trained of the layer l, wherein l is an integer greater than or equal to 0 and represents the GCN layer, sigma (·) is an activation function, and a Re L U function or a Sigmoid function and the like can be selected.
Initializing feature matrix H(0)For each path of lead nodeThe electroencephalogram signal characteristics comprise an initialization characteristic matrix which is composed of one or more characteristics of amplitude characteristic values, frequency characteristic values, power spectrum characteristic values and brain area position characteristics of the nodes of the electroencephalogram signals.
Obtaining Z through N layers of image convolution neural network operation, summing characteristic values of all nodes represented by Z, then carrying out softmax function classification, matching with the label of electroencephalogram signals, such as electroencephalogram signals of normal people or electroencephalogram signals of patients with nervous system related diseases, and then utilizing cross entropy loss function to carry out back propagation to complete W(l)And (3) parameter training, so that a GCN model for distinguishing electroencephalogram signals of normal people and patients with diseases related to the nervous system is established, and intelligent diagnosis of the diseases in the aspect of cranial nerves is realized.
S3: and according to the trained graph convolution network model, carrying out type recognition on a new brain function network to finish electroencephalogram classification.
According to the trained GCN model, the newly input electroencephalogram signals can be intelligently recognized and classified only by establishing a brain function network diagram according to the method in the step S1 to obtain an adjacency matrix and inputting the adjacency matrix into the trained GCN model, and whether the modified electroencephalogram signals belong to certain brain nervous system related diseases or not can be judged.
The following describes the procedures and specific applications of the method of electroencephalographic signal analysis of the present invention.
Firstly, according to an electroencephalogram signal to be analyzed, judging the connection relation of nodes through phase-locked values of all-brain-domain power peaks of every two paths of lead nodes at moments near the peak value, and establishing a brain function network; then training an optimal GCN model according to different types of electroencephalogram signals; and finally, classifying the new electroencephalogram signal by using the trained GCN model to judge whether the new electroencephalogram signal belongs to a certain disease related to the cerebral nervous system.
Compared with the prior art, the electroencephalogram signal analysis method based on the graph convolution network establishes the brain function network through the phase-locked value at the moment near the full brain domain power peak value, and then completes electroencephalogram classification through the graph convolution network. The method can accurately describe the detail characteristics and the state of the brain, and can provide a brand-new method for the intelligent judgment, early warning and treatment of the diseases by applying the method to the brain functional network of the patient with the diseases related to the brain nervous system.
The present invention is not limited to the above-described embodiments, and various modifications and variations of the present invention are intended to be included within the scope of the claims and the equivalent technology of the present invention if they do not depart from the spirit and scope of the present invention.
Claims (6)
1. An electroencephalogram signal analysis method, wherein electroencephalogram signals are collected by multi-channel lead electrodes, and the method is characterized in that: the method comprises the following steps:
s1: establishing a brain functional network according to an electroencephalogram signal, comprising the following specific steps:
s11: the power of the whole brain domain GFP is calculated,
wherein K is the total number of conducting electrodes of the leads and represents K nodes, i is the ith lead, and Vi(t) EEG signal for the ith lead, VmeanAnd (t) is the average value of K lead signals at the t moment.
S12: the peak time of GFP is obtained and is recorded as the time series t ═ t1,t2,t3...tn]；
S13: in the time series [ t ]1,t2,t3...tn]At each time ti(i∈[1,n]) Calculating the phase locking value between each two paths of lead electroencephalogram signals within a time range of the central time;
s14: reflecting the relation between nodes according to the phase locking value index between each two paths of lead electroencephalogram signals so as to establish a brain function network.
S2: labeling different brain function networks according to the brain function networks to complete the training of the convolutional network model;
s3: and according to the trained graph convolution network model, carrying out type recognition on the new brain function network to finish electroencephalogram classification.
2. The electroencephalogram signal analysis method according to claim 1, characterized in that: the formula of the graph convolution network in step S2 is:
wherein
To add a self-connected degree matrix to each node in the brain function network map,
adding an identity matrix I, H to an adjacency matrix of a brain function network diagram(l+1)Outputs results for l +1 layers of GCN, H(l)Is a GCN output of l layers, H(0)For initialising a feature matrix of nodes, W(l)The parameter moment to be trained of the layer l, wherein l is an integer greater than or equal to 0 and represents the layer of the GCN, and sigma (-) is an activation function, and a Re L U function or a Sigmoid function can be selected.
3. The electroencephalogram signal analysis method according to claim 2, characterized in that: initializing feature matrix H of the step S2(0)The electroencephalogram signal characteristics of each path of lead node comprise an initialization characteristic matrix which is composed of one or more characteristics of amplitude characteristic values, frequency characteristic values, power spectrum characteristic values and brain area position characteristics of the node electroencephalogram signals.
4. The electroencephalogram signal analysis method according to claim 3, characterized in that: the graph convolution network training model of the step S2 is classified by adopting a softmax function, and W is completed by utilizing cross entropy loss function reverse propagation(l)And (5) parameter training.
5. The method of claim 4The electroencephalogram signal analysis method is characterized in that the calculation formula of the phase-locked value P L V is as follows:
wherein
Is the phase relation value of two-way lead EEG signal, N is time sequence [ t1,t2,t3...tn]At each time ti(i∈[1,n]) The number of sampling points in a time range of the central time.
6. The electroencephalogram signal analysis method according to claim 5, characterized in that: phase relation value of the two-path lead electroencephalogram signal
May be calculated by time-frequency analysis of the signal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010093974.4A CN111419221A (en) | 2020-02-14 | 2020-02-14 | Electroencephalogram signal analysis method based on graph convolution network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010093974.4A CN111419221A (en) | 2020-02-14 | 2020-02-14 | Electroencephalogram signal analysis method based on graph convolution network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111419221A true CN111419221A (en) | 2020-07-17 |
Family
ID=71547060
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010093974.4A Pending CN111419221A (en) | 2020-02-14 | 2020-02-14 | Electroencephalogram signal analysis method based on graph convolution network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111419221A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113476056A (en) * | 2021-06-25 | 2021-10-08 | 西北工业大学 | Motor imagery electroencephalogram signal classification method based on frequency domain graph convolution neural network |
| CN113509148A (en) * | 2021-04-28 | 2021-10-19 | 东北大学 | Schizophrenia detection system based on mixed high-order brain network |
| CN113729729A (en) * | 2021-08-13 | 2021-12-03 | 北京航空航天大学 | Schizophrenia early detection system based on graph neural network and brain network |
| CN113951883A (en) * | 2021-11-12 | 2022-01-21 | 上海交通大学 | Gender difference detection method based on electroencephalogram signal emotion recognition |
| CN113974650A (en) * | 2021-06-29 | 2022-01-28 | 华南师范大学 | Electroencephalogram network function analysis method and device, electronic equipment and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107358022A (en) * | 2017-06-02 | 2017-11-17 | 常州大学 | A kind of Modularity analysis method of cerebral function network |
| CN108577835A (en) * | 2018-05-17 | 2018-09-28 | 太原理工大学 | A kind of brain function network establishing method based on micro- state |
| CN108648095A (en) * | 2018-05-10 | 2018-10-12 | 浙江工业大学 | A kind of nodal information hidden method accumulating gradient network based on picture scroll |
| CN110399857A (en) * | 2019-08-01 | 2019-11-01 | 西安邮电大学 | A kind of brain electricity emotion identification method based on figure convolutional neural networks |
| CN110473592A (en) * | 2019-07-31 | 2019-11-19 | 广东工业大学 | The multi-angle of view mankind for having supervision based on figure convolutional network cooperate with lethal gene prediction technique |
| CN110522448A (en) * | 2019-07-12 | 2019-12-03 | 东南大学 | A kind of brain network class method based on figure convolutional neural networks |
-
2020
- 2020-02-14 CN CN202010093974.4A patent/CN111419221A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107358022A (en) * | 2017-06-02 | 2017-11-17 | 常州大学 | A kind of Modularity analysis method of cerebral function network |
| CN108648095A (en) * | 2018-05-10 | 2018-10-12 | 浙江工业大学 | A kind of nodal information hidden method accumulating gradient network based on picture scroll |
| CN108577835A (en) * | 2018-05-17 | 2018-09-28 | 太原理工大学 | A kind of brain function network establishing method based on micro- state |
| CN110522448A (en) * | 2019-07-12 | 2019-12-03 | 东南大学 | A kind of brain network class method based on figure convolutional neural networks |
| CN110473592A (en) * | 2019-07-31 | 2019-11-19 | 广东工业大学 | The multi-angle of view mankind for having supervision based on figure convolutional network cooperate with lethal gene prediction technique |
| CN110399857A (en) * | 2019-08-01 | 2019-11-01 | 西安邮电大学 | A kind of brain electricity emotion identification method based on figure convolutional neural networks |
Non-Patent Citations (1)
| Title |
|---|
| 秦云: "脑电的脑网络方法及应用研究" * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113509148A (en) * | 2021-04-28 | 2021-10-19 | 东北大学 | Schizophrenia detection system based on mixed high-order brain network |
| CN113509148B (en) * | 2021-04-28 | 2022-04-22 | 东北大学 | Schizophrenia detection system based on mixed high-order brain network |
| CN113476056A (en) * | 2021-06-25 | 2021-10-08 | 西北工业大学 | Motor imagery electroencephalogram signal classification method based on frequency domain graph convolution neural network |
| CN113476056B (en) * | 2021-06-25 | 2024-03-15 | 西北工业大学 | Motor imagery electroencephalogram signal classification method based on frequency domain graph convolution neural network |
| CN113974650A (en) * | 2021-06-29 | 2022-01-28 | 华南师范大学 | Electroencephalogram network function analysis method and device, electronic equipment and storage medium |
| CN113729729A (en) * | 2021-08-13 | 2021-12-03 | 北京航空航天大学 | Schizophrenia early detection system based on graph neural network and brain network |
| CN113729729B (en) * | 2021-08-13 | 2022-06-14 | 北京航空航天大学 | Schizophrenia early detection system based on graph neural network and brain network |
| CN113951883A (en) * | 2021-11-12 | 2022-01-21 | 上海交通大学 | Gender difference detection method based on electroencephalogram signal emotion recognition |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111227829A (en) | Electroencephalogram signal analysis method based on complex network characteristic indexes | |
| CN111419221A (en) | Electroencephalogram signal analysis method based on graph convolution network | |
| Paraskevopoulou et al. | Feature extraction using first and second derivative extrema (FSDE) for real-time and hardware-efficient spike sorting | |
| Maccione et al. | A novel algorithm for precise identification of spikes in extracellularly recorded neuronal signals | |
| Gibson et al. | Comparison of spike-sorting algorithms for future hardware implementation | |
| CN104586387A (en) | Method for extracting and fusing time, frequency and space domain multi-parameter electroencephalogram characters | |
| CN111062250A (en) | Multi-subject motor imagery electroencephalogram signal identification method based on depth feature learning | |
| CN111150393B (en) | Electroencephalogram epilepsy spike discharge joint detection method based on LSTM multichannel | |
| CN109497996B (en) | Method for constructing and analyzing complex network of micro-state EEG time domain features | |
| CN104720796A (en) | Automatic detecting system and method for epileptic attack time period | |
| CN111227827B (en) | Electroencephalogram signal analysis method based on community division algorithm | |
| CN110598793A (en) | Brain function network feature classification method | |
| Chen et al. | Multiattention adaptation network for motor imagery recognition | |
| CN113509148B (en) | Schizophrenia detection system based on mixed high-order brain network | |
| CN109271887A (en) | A kind of composite space filtering and template matching method for the identification of brain power mode | |
| CN114492513A (en) | Electroencephalogram emotion recognition method for adaptation to immunity domain based on attention mechanism in cross-user scene | |
| Xiong et al. | An unsupervised compressed sensing algorithm for multi-channel neural recording and spike sorting | |
| CN111227826B (en) | Electroencephalogram signal analysis method based on network motif | |
| CN111931578B (en) | Electroencephalogram identification method based on minimum spanning tree and regional double-layer network | |
| CN114081505A (en) | Electroencephalogram signal identification method based on Pearson correlation coefficient and convolutional neural network | |
| CN109325410A (en) | A kind of EEG signals EEG feature extracting method based on convolutional neural networks | |
| CN110738093B (en) | Classification method based on improved small world echo state network electromyography | |
| CN111227828A (en) | Method for establishing brain function network | |
| Behrouzi et al. | Understanding power of graph convolutional neural network on discriminating human EEG signal | |
| CN115886735A (en) | Severe encephalitis patient epileptic seizure automatic checkout device based on brain network optimization |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200717 |
|
| RJ01 | Rejection of invention patent application after publication |