CN110327043B - Event-related potential waveform map solving method based on sparse modeling - Google Patents
Event-related potential waveform map solving method based on sparse modeling Download PDFInfo
- Publication number
- CN110327043B CN110327043B CN201910715578.8A CN201910715578A CN110327043B CN 110327043 B CN110327043 B CN 110327043B CN 201910715578 A CN201910715578 A CN 201910715578A CN 110327043 B CN110327043 B CN 110327043B
- Authority
- CN
- China
- Prior art keywords
- waveform
- cognitive
- dictionary
- ewa
- eeg
- 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
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
- A61B5/40—Detecting, measuring or recording for evaluating the nervous system
- A61B5/4076—Diagnosing or monitoring particular conditions of the nervous system
- A61B5/4088—Diagnosing of monitoring cognitive diseases, e.g. Alzheimer, prion diseases or dementia
-
- 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
- A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
-
- 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
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Public Health (AREA)
- Surgery (AREA)
- Veterinary Medicine (AREA)
- General Health & Medical Sciences (AREA)
- Psychiatry (AREA)
- Animal Behavior & Ethology (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Molecular Biology (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Signal Processing (AREA)
- Physiology (AREA)
- Neurology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Psychology (AREA)
- Evolutionary Computation (AREA)
- Fuzzy Systems (AREA)
- Mathematical Physics (AREA)
- Developmental Disabilities (AREA)
- Neurosurgery (AREA)
- Hospice & Palliative Care (AREA)
- Child & Adolescent Psychology (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention provides a sparse modeling-based solution method for event-related potential waveform maps. The method comprises the following steps: data preprocessing, drying removal, framing and energy normalization; sparse dictionary learning, dictionary initialization, sparse modeling, dictionary learning and convergence judgment; clustering a waveform dictionary; and judging the properties of the waveform dictionary. The invention has the advantages that: the ERP waveform is obtained according to the inherent event-related potential activity pattern in the data instead of the label given by the cognitive experiment, the EEG data under the experimental environment can be processed, the EEG data obtained under the real environment without the label can also be processed, and the method has better universality and practicability. In addition, the signal processing method facing single-time and single-frame data processing is used, the flexibility of EEG analysis processing can be effectively improved, and the method has a remarkable application prospect in the field of brain-computer interfaces.
Description
Technical Field
The invention relates to the technical field of Waveform maps, in particular to a method for solving an Event-related potential Waveform map (EWA) for brain activity state analysis and brain activity pattern recognition.
Background
In the existing field of brain cognitive state analysis, a cognitive state analysis means which is data-driven and does not depend on a label manually given in the cognitive experiment process is very lacking. EEG is a potential reflection of brain electrophysiological activity directly corresponding to cognitive activity on the scalp, and contains important information related to a large amount of cognitive activity, especially ERP (event related) components therein, which can be considered as a direct representation of event-induced neuron cluster activity, and analyzing the activity can not only help to understand the cognitive activity mechanism of the brain, but also generate great value in the field of BCI (brain-computer interface). However, the existing ERP analysis method usually depends on event labels in the cognitive test process and is obtained by using a superposition average method; although the method can obtain an approximate ERP waveform related to a specific stimulation event, the interference of some unexpected events outside the experimental design on the cognitive process is completely ignored, and the influence of the change of the subjective cognitive state (such as fatigue, vague nerve and the like) on the cognitive process is not considered. In addition, the characteristic that multi-frame EEG data superposition is needed to obtain results greatly limits the application of the method in the fields of cognitive state analysis and BCI. Other data-driven methods, such as Principal Component Analysis (PCA), Independent Component Analysis (ICA), and Empirical Mode Decomposition (EMD), are also used for analyzing EEG signals, and attempt to separate mixed components in EEG signals based on statistical characteristics, but due to the highly time-varying characteristics of EEG signals, the statistical characteristics are often unstable, so that the results of the components separated by these methods have no clear cognitive meaning, cannot accurately characterize a single cognitive process, and cannot be well applied to real-time cognitive state analysis and pattern recognition.
Disclosure of Invention
Aiming at the defects in the prior art, the invention provides a potential waveform map solving method based on sparse modeling, and the defects in the prior art are overcome.
In order to realize the purpose, the technical scheme adopted by the invention is as follows:
a potential waveform map solving method based on sparse modeling comprises the following steps:
step 1: the event-related potential waveform map is characterized by the following four points (1): acquiring an EEG component time domain waveform dictionary containing an EWA and a spontaneous component dictionary by using any dictionary learning method based on sparse decomposition based on EEG waveform data containing ERP components; (2) carrying out unsupervised clustering on the waveform dictionary, wherein a clustering method can be selected at will; (3) determining the cognitive significance of the clustered EEG component time domain waveforms of each type; the cognitive significance can be directly obtained by analyzing the time domain waveform, and a cognitive experiment data label with higher association degree with the category component time domain waveform can also be referred; generally speaking, the cognition related component and the non-cognition related (spontaneous electroencephalogram) component can be identified through the frequency distribution, the maximum intensity and the peak position of a class of time-domain waveforms; (4) and averaging the waveforms of all types of cognitive related electroencephalogram components to obtain an ERP component waveform map with cognitive significance for subsequent analysis or identification.
Step 2: data preprocessing, comprising the following substeps:
denoising in step 201: EEG signals from different sources generally contain white noise with different intensities, the noise needs to be removed before analysis and processing, and any feasible noise removal method such as Kalman filtering, wavelet denoising, sparse denoising and the like can be used;
step 202 framing: decomposing EEG sample data into single-channel data frames starting at the cognitive stimulation time and taking the whole cognitive activity process as the frame length;
step 203 energy normalization: in order to ensure that the waveform of the related map entry is not distorted due to the energy difference between frames, energy normalization processing is required to be carried out on each frame of data, namely a coefficient vector is divided by the square root of the sum of squares of all sampling points, and the energy difference can be recovered through coefficient compensation in the reconstruction process;
and step 3: sparse dictionary learning, comprising the following substeps:
step 301 dictionary initialization: in general, a plurality of samples with the lowest correlation in the training data can be used as the initial waveform dictionary atoms;
step 302 sparse modeling: performing sparse modeling on all training samples to obtain a sparse model with excellent samples under the current dictionary;
step 303 dictionary learning: correcting the existing waveform dictionary by using the sparse model obtained in the previous step;
step 304 convergence judgment: if the current error and the sparsity degree meet the preset convergence condition, entering the next step, otherwise, returning to the step 202;
step 4, clustering the waveform dictionary, which comprises the following substeps:
step 401SOM scale setting: the number of SOM self-organizing nodes needs to be significantly more than the total number of ERP component waveform classes and EEG spontaneous component classes;
step 402SOM self-organizing clustering: obtaining a clustering result of waveform dictionary atoms by using an SOM algorithm;
step 5, judging the properties of the waveform dictionary, comprising the following substeps:
step 501, evaluating an ERP waveform based on stationarity measurement: firstly, averaging waveform atoms of each type of EEG components, and then evaluating the order degree of the averaged waveform by using a method based on difference stationarity measurement, thereby discovering potential ERP waveforms caused by unknown cognitive activities;
the stationarity Smoothness of the residual error is determined according to the following indexes:
wherein e is a reduced residual error obtained after sparse modeling is carried out on an EEG waveform by using the existing waveform dictionary atoms, e ' is a result obtained after median filtering is carried out on the residual error e, max (e ') is the maximum value of the residual error median filtering result, and min (e ') is the minimum value of the median filtering result
Step 502, analyzing the cognitive meaning of the EWA map entry related to the cognitive event: if the training data contains information related to the cognitive event category, the cognitive significance of the EWA entry can be analyzed through the correlation between the cognitive event category and the EWA map entry category;
step 503 is based on EWA extension of new data: sparse modeling is carried out on new training data by using an EWA obtained through training and a spontaneous component waveform dictionary, if an error exceeds a certain water criterion, 30HZ low-pass filtering is carried out on a residual error, then the stationarity index in the step 401 is used for measuring the stationarity of the residual error, if the stationarity index is lower than 7, a new ERP component waveform is considered to exist, and the residual error is taken as an alternative EWA map entry and is included in the EWA; when a plurality of new sample data are associated with the candidate EWA entry, it is considered that the cognitive level waveform actually exists, and the cognitive level waveform is included as a formal entry in the EWA.
Further, in step 502, specifically, the pearson correlation coefficient of the label vector and the entry type vector may be used for determining, and if the label vector and the entry type vector are significantly correlated, it may be determined that the EWA entry indicates the corresponding cognitive event type.
Compared with the prior art, the invention has the advantages that:
the ERP waveform is obtained according to the inherent event-related potential activity pattern in the data instead of the label given by the cognitive experiment, the EEG data under the experimental environment can be processed, the EEG data obtained under the real environment without the label can also be processed, and the method has better universality and practicability. In addition, the signal processing method facing single-time and single-frame data processing is used, the flexibility of EEG analysis processing can be effectively improved, and the method has a remarkable application prospect in the field of brain-computer interfaces.
Drawings
Fig. 1 is a schematic diagram of an EWA map according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be further described in detail below with reference to the accompanying drawings by way of examples.
Existing brain-cognitive research work has demonstrated that regardless of the experimental paradigm of design, the ERP components obtained are necessarily of very high similarity, e.g. error-related potentials, if the same cognitive activities are triggered. A single extraction ERP method study shows that an ERP component waveform can be found from an EEG signal. Then, if an ERP waveform with cognitive significance can be extracted to construct an ERP waveform map, ERP extraction and brain activity pattern analysis can be performed based on cognitive activities rather than experimental paradigms. An ERP Waveform Atlas (EWA) is a collection of a series of ERP waveforms with cognitive significance, wherein each Waveform represents a specific cognitive process, and the EWA is researched to find a reliable scalp electroencephalogram dictionary corresponding to the human brain cognitive process, which has important significance for brain activity research and development of brain-computer interface technology.
The potential waveform map has the following characteristics:
1. each entry of the atlas is a time domain waveform of an ERP component, or other cognitively relevant component, in an EEG signal, each entry generally corresponding to a particular cognitive process, with a clear cognitive meaning.
2. Generally, EWAs are trained using data-driven methods using EEG signals associated with cognitive events, and the training process may use, but does not rely on, labels that the EEG data is associated with a class of cognitive events. Some specific entries in the EWA may be acquired using other methods as well, but should be considered as part of the EWA as claimed in claim 1 as long as they are time domain waveforms derived from EEG signal analysis and have a clear cognitive meaning.
3. In the application process, the EWA can be directly used as a part of a sparse modeling dictionary in the sparse modeling and analysis process of the EEG signal; and the coefficient of the sparse model obtained by using the EWA dictionary has clear cognitive significance, the cognitive significance is consistent with the cognitive significance of the corresponding EWA map entry, and the strength of the corresponding cognitive process is represented by the absolute value of the coefficient.
4. The profile may be used for identification of specific cognitive processes. Because of the explicit cognitive properties of the EWA entry, various signal processing methods can be used to perform component analysis on the EEG signal and then compare the cognitively relevant portions thereof to the EWA entry to identify any particular cognitive process.
5. The atlas can be used in the field of BCI to realize the control of any brain-computer interface equipment; the specific method is that on the basis of the specific cognitive process identification in 4), an identification result is converted into a control signal of brain-computer interface equipment; in this application scenario, the EWA needs to be trained for application requirements.
As described above, the specific definition of EWA is: a set of time domain waveforms of event related components derived from component analysis of an EEG signal, wherein each entry has a well-defined cognitive meaning. The application environment of the EWA includes, but is not limited to, the ranges 3) to 5) above, and also includes the application of any EEG component time domain waveform map satisfying the above definition in the fields of EEG analysis and recognition. FIG. 1 is an example of an EWA map;
sparse modeling-based EWA solving method
In addition to the definition and application range of the EWA, the present patent also includes a feasible EWA construction method, which comprises the following specific steps:
s1: acquiring an EEG component time domain waveform dictionary containing an EWA and a spontaneous component dictionary by using any dictionary learning method based on sparse decomposition based on EEG waveform data containing ERP components;
s2: and carrying out unsupervised clustering on the waveform dictionary, wherein the clustering method can be selected at will.
S3: and determining the cognitive significance of the clustered EEG component time domain waveforms of each type. The cognitive significance can be directly obtained by analyzing the time domain waveform, and a cognitive experiment data label with higher association degree with the category component time domain waveform can also be referred. Generally speaking, cognitively related components and non-cognitively related (spontaneous brain electrical) components can be identified by the frequency distribution, maximum intensity and peak position of a class of time domain waveforms.
S4: and averaging the waveforms of all types of cognitive related electroencephalogram components to obtain an ERP component waveform map with cognitive significance for subsequent analysis or identification.
The solving method of the potential waveform map comprises the following steps:
step 1: data pre-processing
The ERP-containing electroencephalogram signals recorded in different environments need to be preprocessed so as to ensure the consistency of the ERP-containing electroencephalogram signals in the aspects of amplitude, energy and noise intensity. The step comprises the following substeps:
step 101, denoising: EEG signals from different sources generally contain white noise of different intensities, and the noise needs to be removed before analysis and processing, and any feasible noise removal method such as Kalman filtering, wavelet denoising, sparse denoising and the like can be used.
Step 102 framing: decomposing EEG sample data into single-channel data frames starting at the cognitive stimulation time and taking the whole process of cognitive activity as the frame length.
Step 103, energy normalization: in order to ensure that the waveform of the related map entry is not distorted by the energy difference between frames, each frame of data must be subjected to energy normalization processing, namely dividing a coefficient vector by the square root of the sum of squares of all sampling points, and the energy difference can be recovered through coefficient compensation in the reconstruction process.
Step 2: sparse dictionary learning
The preprocessed data can enter a sparse dictionary learning stage. Any dictionary learning algorithm such as K-SVD can be used, and in general, this step can be divided into the following sub-steps:
step 201 dictionary initialization: in general, the samples with the lowest correlation in the training data may be used as the initial waveform dictionary atoms.
Step 202 sparse modeling: and performing sparse modeling on all training samples to obtain a sparse model with excellent samples under the current dictionary.
Step 203 dictionary learning: and correcting the existing waveform dictionary by using the sparse model obtained in the last step.
Step 204, convergence judgment: and if the current error and the sparsity degree meet the preset convergence condition, entering the next step, otherwise, returning to the step 202.
Step 3 waveform dictionary clustering
Clustering is performed by taking all waveform dictionary atoms as clustering objects, and an SOM self-organizing clustering algorithm can be used to avoid accurate estimation of the ERP waveform category number in advance. This step can be summarized as the following substeps:
step 301SOM scale setting: the number of SOM self-organizing nodes needs to be significantly more than the total number of ERP component waveform classes and EEG spontaneous component classes.
Step 302SOM self-organizing clustering: and (5) obtaining a clustering result of the waveform dictionary atoms by using an SOM algorithm.
Step 4 waveform dictionary property determination
The EWA map entry is distinguished from the spontaneous component waveform dictionary based on the statistical characteristics. The method comprises the following substeps:
step 401, stability measurement-based ERP waveform evaluation: the method comprises the steps of firstly averaging waveform atoms of each type of EEG component, and then evaluating the order degree of the averaged waveform by using a method based on difference stationarity measurement, so as to find potential ERP waveforms caused by unknown cognitive activities.
The stationarity Smoothness of the residual error is determined according to the following indexes:
wherein e is a reduced residual error obtained after sparse modeling is carried out on an EEG waveform by using the existing waveform dictionary atoms, e ' is a result obtained after median filtering is carried out on the residual error e, max (e ') is the maximum value of the residual error median filtering result, and min (e ') is the minimum value of the median filtering result
Step 402, analyzing the cognitive meaning of the EWA map entry related to the cognitive event: if the training data contains information related to the cognitive event category, the cognitive meaning of the EWA entry can be analyzed through the correlation between the cognitive event category and the EWA map entry category. Specifically, the pearson correlation coefficient of the label vector and the entry type vector may be used for determination, and if the label vector and the entry type vector are significantly correlated, it may be considered that the EWA entry indicates the corresponding cognitive event type.
Step 403 is based on the EWA extension of the new data: and (3) carrying out sparse modeling on new training data by using the EWA acquired by training and a spontaneous component waveform dictionary, firstly carrying out 30HZ low-pass filtering on the residual error if the error exceeds a certain water criterion, then measuring the stationarity by using the stationarity index in the step 401, considering that a new ERP component waveform exists if the stationarity index is lower than 7, and taking the residual error as an alternative EWA map entry to be incorporated into the EWA. When a plurality of new sample data are associated with the candidate EWA entry, it is considered that the cognitive level waveform actually exists, and the cognitive level waveform is included as a formal entry in the EWA.
It will be appreciated by those of ordinary skill in the art that the examples described herein are intended to assist the reader in understanding the manner in which the invention is practiced, and it is to be understood that the scope of the invention is not limited to such specifically recited statements and examples. Those skilled in the art can make various other specific changes and combinations based on the teachings of the present invention without departing from the spirit of the invention, and these changes and combinations are within the scope of the invention.
Claims (2)
1. A sparse modeling-based event-related potential oscillogram solving method is characterized by comprising the following steps:
step 1: the event-related potential waveform map is characterized by the following four points (1): acquiring an EEG component time domain waveform dictionary containing an EWA and a spontaneous component dictionary by using any dictionary learning method based on sparse decomposition based on EEG waveform data containing ERP components; (2) carrying out unsupervised clustering on the waveform dictionary, wherein a clustering method can be selected at will; (3) determining the cognitive significance of the clustered EEG component time domain waveforms of each type; the cognitive significance can be directly obtained by analyzing the time domain waveform, and a cognitive experiment data label with higher association degree with the category component time domain waveform can also be referred; generally speaking, cognitively relevant components and non-cognitively relevant components can be identified through the frequency distribution, the maximum intensity and the peak position of a class of time domain waveforms; (4) averaging waveforms of all types of cognition related electroencephalogram components to obtain an ERP component waveform map with cognitive significance for subsequent analysis or identification;
step 2: data preprocessing, comprising the following substeps:
denoising in step 201: EEG signals of different origins typically contain white noise of different intensities, which needs to be removed before analysis and processing, and the noise removal method comprises: kalman filtering, wavelet denoising and sparse denoising;
step 202 framing: decomposing EEG sample data into single-channel data frames starting at the cognitive stimulation time and taking the whole cognitive activity process as the frame length;
step 203 energy normalization: in order to ensure that the waveform of the related map entry is not distorted due to the energy difference between frames, energy normalization processing is required to be carried out on each frame of data, namely a coefficient vector is divided by the square root of the sum of squares of all sampling points, and the energy difference can be recovered through coefficient compensation in the reconstruction process;
and step 3: sparse dictionary learning, comprising the following substeps:
step 301 dictionary initialization: taking a plurality of samples with the lowest correlation degree in the training data as initial waveform dictionary atoms;
step 302 sparse modeling: performing sparse modeling on all training samples to obtain a sparse model with excellent samples under the current dictionary;
step 303 dictionary learning: correcting the existing waveform dictionary by using the sparse model obtained in the previous step;
step 304 convergence judgment: if the current error and the sparsity degree meet the preset convergence condition, entering the next step, otherwise, returning to the step 202;
step 4, clustering the waveform dictionary, which comprises the following substeps:
step 401SOM scale setting: the number of SOM self-organizing nodes needs to be significantly more than the total number of ERP component waveform classes and EEG spontaneous component classes;
step 402SOM self-organizing clustering: obtaining a clustering result of waveform dictionary atoms by using an SOM algorithm;
step 5, judging the properties of the waveform dictionary, comprising the following substeps:
step 501, evaluating an ERP waveform based on stationarity measurement: firstly, averaging waveform atoms of each type of EEG components, and then evaluating the order degree of the averaged waveform by using a method based on difference stationarity measurement, thereby discovering potential ERP waveforms caused by unknown cognitive activities; the stationarity Smoothness of the residual error is determined according to the following indexes:
wherein e is a reduced residual error obtained after sparse modeling is carried out on an EEG waveform by using an existing waveform dictionary atom, e ' is a result obtained after median filtering is carried out on the residual error e, max (e ') is the maximum value of the residual error median filtering result, and min (e ') is the minimum value of the median filtering result;
step 502, analyzing the cognitive meaning of the EWA map entry related to the cognitive event: if the training data contains information related to the cognitive event category, the cognitive significance of the EWA entry can be analyzed through the correlation between the cognitive event category and the EWA map entry category;
step 503 is based on EWA extension of new data: sparse modeling is carried out on new training data by using an EWA obtained through training and a spontaneous component waveform dictionary, if an error exceeds a certain water criterion, 30HZ low-pass filtering is carried out on a residual error, then the stationarity index in the step 501 is used for measuring the stationarity of the residual error, if the stationarity index is lower than 7, a new ERP component waveform is considered to exist, and the residual error is taken as an alternative EWA map entry and is included in the EWA; when a plurality of new sample data are associated with the candidate EWA entry, it is considered that the cognitive level waveform actually exists, and the cognitive level waveform is included as a formal entry in the EWA.
2. The sparse modeling based event-related potential oscillogram solving method according to claim 1, wherein: specifically, in step 502, the pearson correlation coefficient of the label vector and the entry type vector is used for judgment, and if the label vector and the entry type vector are significantly correlated, it can be considered that the EWA entry indicates the corresponding cognitive event type.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910715578.8A CN110327043B (en) | 2019-08-05 | 2019-08-05 | Event-related potential waveform map solving method based on sparse modeling |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910715578.8A CN110327043B (en) | 2019-08-05 | 2019-08-05 | Event-related potential waveform map solving method based on sparse modeling |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110327043A CN110327043A (en) | 2019-10-15 |
| CN110327043B true CN110327043B (en) | 2022-04-01 |
Family
ID=68148727
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910715578.8A Active CN110327043B (en) | 2019-08-05 | 2019-08-05 | Event-related potential waveform map solving method based on sparse modeling |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110327043B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114343675B (en) * | 2021-12-27 | 2023-05-30 | 深圳航天科技创新研究院 | Electroencephalogram component extraction method |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101937207A (en) * | 2010-08-27 | 2011-01-05 | 上海交通大学 | Intelligent visual monitoring and diagnosing method of mechanical equipment state |
| CN103117546A (en) * | 2013-02-28 | 2013-05-22 | 武汉大学 | Ultrashort-term slide prediction method for wind power |
| CN104515905A (en) * | 2013-09-29 | 2015-04-15 | 哈尔滨工业大学 | CQT multi-resolution based subject EEG signal adaptive spectrum analysis method |
| CN104545900A (en) * | 2014-12-29 | 2015-04-29 | 中国医学科学院生物医学工程研究所 | Event related potential analyzing method based on paired sample T test |
| CN105748067A (en) * | 2016-02-05 | 2016-07-13 | 电子科技大学 | Evoked potential extracting method based on random gradient adaptive filtering |
| CN107547269A (en) * | 2017-08-14 | 2018-01-05 | 浙江大学 | The construction method of intelligent substation communication flow threshold model based on FARIMA |
| CN109875580A (en) * | 2019-03-01 | 2019-06-14 | 安徽工业大学 | A kind of Cognitive efficiency can computation model |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060122529A1 (en) * | 2004-12-06 | 2006-06-08 | Yang Tsau | Low noise amplifier for electro-physiological signal sensing |
| WO2016162820A1 (en) * | 2015-04-07 | 2016-10-13 | Tata Consultancy Services Limited | System and method for estimating cognitive traits |
| US10928472B2 (en) * | 2016-01-14 | 2021-02-23 | Technion Research & Development Foundation Limited | System and method for brain state classification |
-
2019
- 2019-08-05 CN CN201910715578.8A patent/CN110327043B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101937207A (en) * | 2010-08-27 | 2011-01-05 | 上海交通大学 | Intelligent visual monitoring and diagnosing method of mechanical equipment state |
| CN103117546A (en) * | 2013-02-28 | 2013-05-22 | 武汉大学 | Ultrashort-term slide prediction method for wind power |
| CN104515905A (en) * | 2013-09-29 | 2015-04-15 | 哈尔滨工业大学 | CQT multi-resolution based subject EEG signal adaptive spectrum analysis method |
| CN104545900A (en) * | 2014-12-29 | 2015-04-29 | 中国医学科学院生物医学工程研究所 | Event related potential analyzing method based on paired sample T test |
| CN105748067A (en) * | 2016-02-05 | 2016-07-13 | 电子科技大学 | Evoked potential extracting method based on random gradient adaptive filtering |
| CN107547269A (en) * | 2017-08-14 | 2018-01-05 | 浙江大学 | The construction method of intelligent substation communication flow threshold model based on FARIMA |
| CN109875580A (en) * | 2019-03-01 | 2019-06-14 | 安徽工业大学 | A kind of Cognitive efficiency can computation model |
Non-Patent Citations (3)
| Title |
|---|
| Development and evalution of the sparse decomposition method with mixed over-complete dictionart for evoked potential estimation;Xu Peng;《COMPUTERS IN BIOLOGY AND MEDICINE》;20071231;第37卷(第12期);第1731-1740页 * |
| 基于时序模型的变形预报研究;谢瑞等;《测绘科学》;20120720;第37卷(第04期);第96-98页 * |
| 面向事件相关电位成分的稀疏字典构建方法;李海峰，丰上等;《信号处理》;20180825;第34卷(第8期);第891-896页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110327043A (en) | 2019-10-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Sanei et al. | A new adaptive line enhancer based on singular spectrum analysis | |
| Camilleri et al. | Automatic detection of spindles and K-complexes in sleep EEG using switching multiple models | |
| CN109886068B (en) | Motion data-based action behavior identification method | |
| US20210267530A1 (en) | Multiclass classification method for the estimation of eeg signal quality | |
| Sameni et al. | An iterative subspace denoising algorithm for removing electroencephalogram ocular artifacts | |
| Boumbarov et al. | ECG personal identification in subspaces using radial basis neural networks | |
| CN109009098B (en) | Electroencephalogram signal feature identification method under motor imagery state | |
| CN115414051A (en) | Emotion classification and recognition method of electroencephalogram signal self-adaptive window | |
| CN110327043B (en) | Event-related potential waveform map solving method based on sparse modeling | |
| CN107045624B (en) | Electroencephalogram signal preprocessing and classifying method based on maximum weighted cluster | |
| Akhavan et al. | Subject-based discriminative sparse representation model for detection of concealed information | |
| Pentari et al. | Graph denoising of impulsive EEG signals and the effect of their graph representation | |
| CN114330422A (en) | Cross-test migration learning method for estimating electroencephalogram emotional characteristics in real time | |
| CN116807496B (en) | Method, device, equipment and medium for positioning epileptic interval brain wave abnormal signals | |
| CN112861629A (en) | Multi-window distinguishing typical pattern matching method and brain-computer interface application | |
| CN110516711B (en) | Training set quality evaluation method of MI-BCI system and optimization method of single training sample | |
| Rudas et al. | On activity identification pipelines for a low-accuracy EEG device | |
| CN114098764A (en) | Data processing method and device, electronic equipment and storage medium | |
| Spinnato et al. | Finding EEG space-time-scale localized features using matrix-based penalized discriminant analysis | |
| Shulyak et al. | Analysis of the biomedical signals' structure in the problem of recognition | |
| Kiamini et al. | A wavelet based algorithm for ocular artifact detection in the EEG signals | |
| Sattar et al. | Time-scale-based segmentation for degraded PCG signals using NMF | |
| CN111783857A (en) | Motor imagery brain-computer interface based on nonlinear network information graph | |
| CN110647915A (en) | Dynamic mode judgment method for consistency analysis of high-dimensional data | |
| CN111631711B (en) | Method for analyzing and processing brain electrical data of schizophrenia and auditory hallucination symptoms |
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 |