CN111956221A - Temporal lobe epilepsy classification method based on wavelet scattering factor and LSTM neural network model - Google Patents
Temporal lobe epilepsy classification method based on wavelet scattering factor and LSTM neural network model Download PDFInfo
- Publication number
- CN111956221A CN111956221A CN202010932631.2A CN202010932631A CN111956221A CN 111956221 A CN111956221 A CN 111956221A CN 202010932631 A CN202010932631 A CN 202010932631A CN 111956221 A CN111956221 A CN 111956221A
- Authority
- CN
- China
- Prior art keywords
- temporal lobe
- scattering
- wavelet
- lobe epilepsy
- neural network
- 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.)
- Granted
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
- A61B5/05—Detecting, measuring or recording for diagnosis by means of electric currents or magnetic fields; Measuring using microwaves or radio waves
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/0033—Features or image-related aspects of imaging apparatus classified in A61B5/00, e.g. for MRI, optical tomography or impedance tomography apparatus; arrangements of imaging apparatus in a room
- A61B5/004—Features or image-related aspects of imaging apparatus classified in A61B5/00, e.g. for MRI, optical tomography or impedance tomography apparatus; arrangements of imaging apparatus in a room adapted for image acquisition of a particular organ or body part
- A61B5/0042—Features or image-related aspects of imaging apparatus classified in A61B5/00, e.g. for MRI, optical tomography or impedance tomography apparatus; arrangements of imaging apparatus in a room adapted for image acquisition of a particular organ or body part for the brain
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/40—Detecting, measuring or recording for evaluating the nervous system
- A61B5/4076—Diagnosing or monitoring particular conditions of the nervous system
- A61B5/4094—Diagnosing or monitoring seizure diseases, e.g. epilepsy
-
- 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
- A61B5/7253—Details of waveform analysis characterised by using transforms
- A61B5/7257—Details of waveform analysis characterised by using transforms using Fourier transforms
-
- 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
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B2576/00—Medical imaging apparatus involving image processing or analysis
- A61B2576/02—Medical imaging apparatus involving image processing or analysis specially adapted for a particular organ or body part
- A61B2576/026—Medical imaging apparatus involving image processing or analysis specially adapted for a particular organ or body part for the brain
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Biomedical Technology (AREA)
- Veterinary Medicine (AREA)
- Public Health (AREA)
- General Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Animal Behavior & Ethology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Molecular Biology (AREA)
- Surgery (AREA)
- Artificial Intelligence (AREA)
- Neurology (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Physiology (AREA)
- Radiology & Medical Imaging (AREA)
- Signal Processing (AREA)
- Psychiatry (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Mathematical Physics (AREA)
- Neurosurgery (AREA)
- Evolutionary Computation (AREA)
- Fuzzy Systems (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention discloses a temporal lobe epilepsy classification research method based on the combination of feature extraction of wavelet scattering factors of MEG signals and an LSTM network model, which comprises the following steps: (1) extracting MEG signals of 262 channels of the left temporal lobe epilepsy, the right temporal lobe epilepsy and the normal person; (2) constructing a wavelet scattering network, and calculating scattering factors of all channels; (3) taking the scattering factors as the input of an LSTM neural network and carrying out training and learning; (4) using average classification accuracy, standard deviation and F1And evaluating the result of the network training. The method has excellent effect, and can effectively distinguish epilepsy of left temporal lobe, epilepsy of right temporal lobe and normal person.
Description
Technical Field
The invention belongs to the field of biomedicine, and particularly relates to a temporal lobe epilepsy classification research method based on MEG signal wavelet scattering network construction and feature extraction combined with an LSTM neural network model.
Background
Temporal Lobe Epilepsy (Temporal Lobe Epilepsy) is a brain disease caused by sudden abnormal discharge of neuron groups in the brain, and clinical symptoms mainly include the symptoms of neuron degeneration with reduction, degeneration and necrosis of anterior Temporal and hippocampal neurons, and ganglionic phenomenon with gliocyte hyperplasia. In recent years, researches show that the left and right side nieal epilepsy can cause different influences on language behaviors of the brain and even on the sexual cognitive function, so that the left and right side temporal epilepsy can be distinguished, and the auxiliary diagnosis and treatment and postoperative recovery can be facilitated.
At present, the detection means of epilepsy mainly comprises fMRI and EEG, and because the time resolution of fMRI data is low, EEG signals are mostly non-invasive signals, and the acquisition of voltage signals is unstable. The MEG signal is acquired in a non-invasive manner, is a non-stationary and non-linear complex signal, has high time and space resolution and is beneficial to characterization of the epileptic signal. However, the MEG signals of the left temporal lobe epilepsy and the right temporal lobe epilepsy have confusability, the time-frequency characteristics of the MEG signals can be effectively extracted by adopting wavelet scattering, the multi-dimensional characteristics of the signals can be extracted, and three different groups of people can be effectively distinguished by combining the LSTM with the wavelet scattering network.
Disclosure of Invention
The purpose of the invention is as follows: in order to find an effective method for distinguishing left temporal lobe epilepsy, right temporal lobe epilepsy and healthy adults, the invention provides a research method for classifying and identifying temporal lobe epilepsy based on the combination of wavelet scattering network feature extraction and an LSTM neural network model.
The technical scheme of the invention is as follows:
a temporal lobe epilepsy classification method based on wavelet scattering factors and an LSTM neural network model is characterized by comprising the following steps:
introducing a preprocessed MEG magnetoencephalogram signal;
step (2) constructing a sample library through data enhancement;
scattering each channel MEG signal through the constructed wavelet scattering network, and calculating wavelet factors of 0 to 2 orders of the MEG signal to obtain 15982-dimensional scattering factors;
inputting 15982-dimensional scattering factors into an LSTM neural network for training and learning;
step (5) using the average classification accuracy, standard deviation and F1And evaluating the network learning result.
The step (1) is specifically as follows:
analyzing MEG signals of a normal person, left temporal lobe epilepsy and right temporal lobe epilepsy; the database source is brain hospital affiliated to Nanjing medical university; data are from 39 left temporal lobe epileptics, 32 right temporal lobe epileptics and 16 normal persons, and MEG signals of 275 channels of each subject are collected and processed by Brainstarm software to obtain MEG data of 262 channels and the sampling frequency is 300 Hz.
The step (2) is specifically as follows:
and (2-1) sorting the imported data, defining 10s of data as a window, and dividing the data into 10-second segments, wherein the number of samples of the left temporal lobe epilepsy is 7380, the number of samples of the normal person is 456, the number of samples of the right temporal lobe epilepsy is 5364, and the total number of samples is 13200.
The step (3) is specifically as follows:
step (3-1) combines Fourier transform and traditional wavelet transform to obtain wavelet scattering transform of limited energy finite long-time signal x (t)
Wherein φ (t) represents a mother wave function;
step (3-2) through a formula of wavelet scattering transformation, we can obtain a propagation formula of a nonlinear correlation factor as follows:
we can thus obtain as input features a set of scattering factors, which are combined into
The step (4) is specifically as follows:
and (4-1) performing dimension stretching operation on the wavelet scattering factors of the MEG signals extracted from the step (3-1) to the step (3-2) to transform the MEG signals from 262 × 61 to 1 × 15982.
And (4-2) sending the scattering characteristics subjected to dimension conversion in the step (4-1) as input characteristics into an LSTM network for training.
The step (5) is specifically as follows:
the method comprises the following steps of (5-1) adopting an average accuracy, a class accuracy and a class recall rate after five times of cross experiments as evaluation indexes;
and (5-2) analyzing and comparing the average accuracy, the class accuracy and the class recall rate under different parameter conditions.
Has the advantages that: the temporal lobe epilepsy identification and analysis method is researched, the wavelet scattering coefficient of the MEG signal is extracted to be used as a 15982-dimensional scattering factor, and the wavelet scattering coefficient is sent to an LSTM neural network for learning. The invention has the beneficial effects that: the system can accurately classify the left temporal lobe epilepsy, the right temporal lobe epilepsy and normal people, and can help the recovery of the temporal lobe epilepsy.
Drawings
FIG. 1 is a flow chart of the disclosed method;
FIG. 2 is a graph of the left temporal lobe epileptic partial channel MEG signal;
FIG. 3 is a graph of the MEG signal of the right temporal lobe epileptic partial channel;
FIG. 4 is a diagram of a normal human partial channel MEG signal;
FIG. 5 is a diagram of a wavelet scattering network structure;
FIG. 6 shows the network training process when the number of hidden nodes is 50 and the learning rate is 0.0002;
FIG. 7 shows the network training process when the number of hidden nodes is 100 and the learning rate is 0.0004;
FIG. 8 is a graph showing the result when the number of hidden nodes is 50 and the learning rate is 0.0002;
FIG. 9 is a graph showing the result when the number of hidden nodes is 100 and the learning rate is 0.0004;
Detailed Description
The invention is further elucidated with reference to the drawings and the detailed description.
As shown in fig. 1, the flowchart of the temporal lobe epilepsy identification and classification method based on MEG signal and LSTM neural network model disclosed in the present invention specifically includes the following steps:
introducing a preprocessed MEG magnetoencephalogram signal;
step (1-1) the present invention analyzes MEG signals of normal persons, left temporal lobe epilepsy and right temporal lobe epilepsy. The database source is brain hospital affiliated to Nanjing medical university. Data are from 39 left temporal lobe epileptics, 32 right temporal lobe epileptics and 16 normal persons, 275-channel MEG signals of each subject are collected and subjected to data preprocessing by Brainstarm software, and the processed data are 262-channel MEG data with the sampling frequency of 300 Hz.
Step (2), enhancing data and constructing a database;
the method comprises the following specific steps:
after the imported data are sorted, ten seconds of data are defined as a window, the data are divided into ten seconds of segments, the number of samples of left temporal lobe epilepsy is 7380, the number of samples of normal people is 456, the number of samples of right temporal lobe epilepsy is 5364, and the total number of samples is 13200.
Step (3) constructing a wavelet scattering network, scattering MEG signals of all channels, and calculating wavelet factors of 0-2 orders of the MEG signals; the method comprises the following specific steps:
step (3-1) combines Fourier transform and traditional wavelet transform to obtain wavelet scattering transform of limited energy finite long-time signal x (t)
Where φ (t) represents the mother wave function.
And (3-2) obtaining a propagation formula of the nonlinear related factors through a formula of wavelet scattering transformation as follows:
obtaining a set of scattering factors as input features:
inputting 15982D scattering factors into an LSTM neural network for training and learning, and the method comprises the following steps:
and (4-1) performing dimension stretching operation on the wavelet scattering factors of the MEG signals extracted from the step (3-1) to the step (3-2) to transform the MEG signals from 262 × 61 to 1 × 15982.
And (4-2) sending the scattering characteristics subjected to dimension conversion in the step (4-1) as input characteristics into an LSTM network for training.
Step (5) using the average classification accuracy, standard deviation and F1Evaluating the network learning result by the value, wherein the specific steps are as follows;
the method comprises the following steps of (5-1) adopting an average accuracy, a class accuracy and a class recall rate after five times of cross experiments as evaluation indexes;
and (5-2) analyzing and comparing the average accuracy, the class accuracy and the class recall rate under different parameter conditions.
The experimental steps are as follows:
by adopting the step (1) in the specific embodiment, the MEG data of 87 subjects in the database is imported, wherein the MEG data comprises 39 left temporal lobe epileptic patients, 32 right temporal lobe epileptic patients and 16 normal subjects without brain disease history, and the ownership of the database is affiliated encephalic hospital of Nanjing medical university; cutting the collected segment of each patient into 10-second segments by adopting the step (2) in the specific embodiment, wherein the sample size of the left temporal lobe epilepsy is 7380, the sample size of the normal person is 456, the sample size of the right temporal lobe epilepsy is 5364, and the total number of samples is 13200;
constructing a wavelet scattering network with the highest order of 1 by adopting the step (3) in the specific implementation mode, respectively using Morlet waves and Gabor waves as band-pass and low-pass filters by using the filter, wherein the sampling interval of scattering factors is 8, scattering each channel of each segment to obtain the characteristic of 1 × 61, and combining 262 channels to obtain a scattering factor set of 262 × 61;
in step (4) of the specific embodiment, after the 262-dimensional feature is stretched to 15982-dimensional feature, the 15982-dimensional scattering factor is input to the LSTM neural network for learning. The size parameter of the network input layer is the dimension of a wavelet scattering factor sequence, which is 15982, the LSTM layer can learn the long-term dependence relationship in time sequence data with different step lengths, the output mode is last which represents that the network is classified aiming at sequence-labels, the size of the full connection layer is 3 which is the number of the classified categories, the probabilities of the categories are output through the softmax layer, and the final classification layer outputs the final classification result. The number of hidden nodes of the network is directly related to the number of samples and the requirements of classification, too many hidden nodes prolong the learning time of the network, too few hidden nodes cause low fault tolerance of the network, and the classification identification capability on a test sample set is reduced. In the invention, the number of samples of the left temporal lobe epilepsy is 7380, the number of samples of the normal person is 456, the number of samples of the right temporal lobe epilepsy is 5364, and the total number of samples is 13200. The method comprises the steps of disturbing a total sample, dividing the total sample into five partitions, and carrying out five times of cross validation, wherein one partition is adopted as a test sample in each time of cross validation, and the rest nine partitions are adopted as training samples;
and (4) evaluating the network learning result by using the average classification accuracy and the standard deviation. In the present invention, let ltle (left low temporal epilepsy) be left temporal epilepsia, rtle (right low temporal epilepsy), h (health) be normal person, TP (True Positive) be Positive type sample predicted as Positive type by the model, TN (True Negative) be Negative type sample predicted as Negative type by the model, FP (False Positive) be Negative type sample predicted as Positive type by the model, FN (False Negative) be Positive type sample predicted as Negative type by the model, i ═ 1, 2, 3, …, 5, accuracy is defined as probability of correctly classifying all samples:
the average accuracy of five cross-validation experiments is then:
the accuracy is the ratio of the actual positive class in the samples predicted to be the positive class:
precision is the average after five-layer cross validation.
The recall ratio is the proportion determined as positive class in the samples actually being positive class:
recall is the average value after five-layer cross validation.
The F1 value comprehensively considers the precision rate and the recall rate, is a harmonic mean of the precision rate and the recall rate and is often used as a final evaluation method of a machine learning classification method, and the higher the F1 value of each class is, the better the classification result is. The F1 values under each category are expressed as:
F1the mean value after five-layer cross validation.
The classification accuracy of the total samples at different hidden node numbers and learning rates are listed in table 1:
table 1 number of hidden nodes and average accuracy, average recall rate and classification accuracy under learning rate.
As can be seen from table 1, the resolution of temporal lobe epilepsy recognition is improved as the hidden layer is improved, is improved as the learning rate is reduced, and gradually converges, the highest accuracy rate of 97.19% can be achieved when the learning rate is 0.0002 and the hidden layer is 100 layers (the training process of the network is shown in fig. 2 and 3), and the standard deviation of five times of cross validation is about 7%, so that the accuracy rate and the stability are high, and the target is in line with the actual expected target.
As can be seen from the experimental results of Table 1, in the comparative experiment, when the learning rate was 0.0002 and the number of hidden nodes was 100, the classification of the epilepsy of the left temporal lobe, the epilepsy of the right temporal lobe and the normal person was F1The values reach respective maximum values.
As can be seen from the experimental results of Table 1, the hidden layerWhen the number of nodes is 100 and the learning rate is 0.0002, the accuracy of temporal lobe epilepsy identification and classification is highest, and F of each classification1The value is also highest. Therefore, the classification accuracy of the total samples and the F of each type of sample are comprehensively compared1The wavelet scattering network based on the MEG signal and the LSTM classification model are combined to be effective in the classification of temporal lobe, and even under the condition that the sample distribution is not uniform, the good classification effect can be obtained.
Claims (6)
1. A temporal lobe epilepsy classification method based on wavelet scattering factors and an LSTM neural network model is characterized by comprising the following steps:
introducing a preprocessed MEG magnetoencephalogram signal;
step (2) constructing a sample library through data enhancement;
scattering each channel MEG signal through the constructed wavelet scattering network, and calculating wavelet factors of 0 to 2 orders of the MEG signal to obtain 15982-dimensional scattering factors;
inputting 15982-dimensional scattering factors into an LSTM neural network for training and learning;
step (5) using the average classification accuracy, standard deviation and F1And evaluating the network learning result.
2. The temporal lobe epilepsy classification method of the wavelet scattering factor and LSTM neural network model according to claim 1, wherein the step (1) is specifically:
analyzing MEG signals of a normal person, left temporal lobe epilepsy and right temporal lobe epilepsy; the database source is brain hospital affiliated to Nanjing medical university; data are from 39 left temporal lobe epileptics, 32 right temporal lobe epileptics and 16 normal persons, and MEG signals of 275 channels of each subject are collected and processed by Brainstarm software to obtain MEG data of 262 channels and the sampling frequency is 300 Hz.
3. The temporal lobe epilepsy classification method of the wavelet scattering factor and LSTM neural network model according to claim 1, wherein the step (2) is specifically as follows:
and (2-1) sorting the imported data, defining 10s of data as a window, and dividing the data into 10-second segments, wherein the number of samples of the left temporal lobe epilepsy is 7380, the number of samples of the normal person is 456, the number of samples of the right temporal lobe epilepsy is 5364, and the total number of samples is 13200.
4. The temporal lobe epilepsy classification method of the wavelet scattering factor and LSTM neural network model according to claim 1, wherein the step (3) is specifically:
step (3-1) combines Fourier transform and traditional wavelet transform to obtain wavelet scattering transform of limited energy finite long-time signal x (t)
Wherein φ (t) represents a mother wave function;
step (3-2) through a formula of wavelet scattering transformation, we can obtain a propagation formula of a nonlinear correlation factor as follows:
we can thus obtain as input features a set of scattering factors, which are combined into
5. The temporal lobe epilepsy classification method of the wavelet scattering factor and LSTM neural network model according to claim 1, wherein the step (4) is specifically as follows:
and (4-1) performing dimension stretching operation on the wavelet scattering factors of the MEG signals extracted from the step (3-1) to the step (3-2) to transform the MEG signals from 262 × 61 to 1 × 15982.
And (4-2) sending the scattering characteristics subjected to dimension conversion in the step (4-1) as input characteristics into an LSTM network for training.
6. The temporal lobe epilepsy recognition and classification method of the wavelet scattering factor and LSTM neural network model according to claim 1, wherein the step (5) is specifically:
the method comprises the following steps of (5-1) adopting an average accuracy, a class accuracy and a class recall rate after five times of cross experiments as evaluation indexes;
and (5-2) analyzing and comparing the average accuracy, the class accuracy and the class recall rate under different parameter conditions.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010932631.2A CN111956221B (en) | 2020-09-07 | 2020-09-07 | Temporal lobe epilepsy classification method based on wavelet scattering factor and LSTM neural network model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010932631.2A CN111956221B (en) | 2020-09-07 | 2020-09-07 | Temporal lobe epilepsy classification method based on wavelet scattering factor and LSTM neural network model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111956221A true CN111956221A (en) | 2020-11-20 |
| CN111956221B CN111956221B (en) | 2022-06-07 |
Family
ID=73392523
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010932631.2A Active CN111956221B (en) | 2020-09-07 | 2020-09-07 | Temporal lobe epilepsy classification method based on wavelet scattering factor and LSTM neural network model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111956221B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113077795A (en) * | 2021-04-06 | 2021-07-06 | 重庆邮电大学 | Voiceprint recognition method under channel attention spreading and aggregation |
| CN113796873A (en) * | 2021-10-26 | 2021-12-17 | 山东建筑大学 | Wearable dynamic electrocardiosignal classification method and system |
| CN114081462A (en) * | 2021-11-19 | 2022-02-25 | 齐齐哈尔大学 | Heart health monitoring system based on multi-dimensional physiological information |
| CN114358161A (en) * | 2021-12-27 | 2022-04-15 | 中国人民解放军国防科技大学 | Falling detection method and device based on wavelet scattering network and electronic equipment |
| CN116172531A (en) * | 2023-04-25 | 2023-05-30 | 南京信息工程大学 | Blood pressure change trend estimation method based on wavelet dispersion transformation |
| CN116898454A (en) * | 2023-07-11 | 2023-10-20 | 湖北大学 | Epileptic classification method and system based on electroencephalogram feature fusion deep learning model |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106821376A (en) * | 2017-03-28 | 2017-06-13 | 南京医科大学 | A kind of epileptic attack early warning system and method based on deep learning algorithm |
| CN106874694A (en) * | 2017-03-20 | 2017-06-20 | 江南大学 | A kind of intelligent diagnosis system towards epileptic EEG Signal identification |
| CN108549875A (en) * | 2018-04-19 | 2018-09-18 | 北京工业大学 | A kind of brain electricity epileptic attack detection method based on the perception of depth channel attention |
| CN109691985A (en) * | 2018-12-19 | 2019-04-30 | 北京工业大学 | A kind of temporal epilepsy aided diagnosis method based on DTI technology and SVM |
| CN109700463A (en) * | 2019-03-12 | 2019-05-03 | 四川大学 | A kind of magneticencephalogram epileptic spike recognition methods and system |
| CN109934089A (en) * | 2018-10-31 | 2019-06-25 | 北京航空航天大学 | Multistage epileptic EEG Signal automatic identifying method based on supervision gradient lifter |
| CN109924974A (en) * | 2019-01-18 | 2019-06-25 | 天津职业技术师范大学(中国职业培训指导教师进修中心) | A kind of coding/decoding method and cloud decoding system based on EcoG epilepsy nerve signal |
| CN111222442A (en) * | 2019-12-31 | 2020-06-02 | 中国电子科技集团公司第三十六研究所 | Electromagnetic signal classification method and device |
| GB2578929A (en) * | 2018-11-13 | 2020-06-03 | King Abdulaziz City Of Science And Tech | Automatic brain disorder tendency screening using statistical features of MEG data |
| CN111340142A (en) * | 2020-05-14 | 2020-06-26 | 南京慧脑云计算有限公司 | Epilepsia magnetoencephalogram spike automatic detection method and tracing positioning system |
-
2020
- 2020-09-07 CN CN202010932631.2A patent/CN111956221B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106874694A (en) * | 2017-03-20 | 2017-06-20 | 江南大学 | A kind of intelligent diagnosis system towards epileptic EEG Signal identification |
| CN106821376A (en) * | 2017-03-28 | 2017-06-13 | 南京医科大学 | A kind of epileptic attack early warning system and method based on deep learning algorithm |
| CN108549875A (en) * | 2018-04-19 | 2018-09-18 | 北京工业大学 | A kind of brain electricity epileptic attack detection method based on the perception of depth channel attention |
| CN109934089A (en) * | 2018-10-31 | 2019-06-25 | 北京航空航天大学 | Multistage epileptic EEG Signal automatic identifying method based on supervision gradient lifter |
| GB2578929A (en) * | 2018-11-13 | 2020-06-03 | King Abdulaziz City Of Science And Tech | Automatic brain disorder tendency screening using statistical features of MEG data |
| CN109691985A (en) * | 2018-12-19 | 2019-04-30 | 北京工业大学 | A kind of temporal epilepsy aided diagnosis method based on DTI technology and SVM |
| CN109924974A (en) * | 2019-01-18 | 2019-06-25 | 天津职业技术师范大学(中国职业培训指导教师进修中心) | A kind of coding/decoding method and cloud decoding system based on EcoG epilepsy nerve signal |
| CN109700463A (en) * | 2019-03-12 | 2019-05-03 | 四川大学 | A kind of magneticencephalogram epileptic spike recognition methods and system |
| CN111222442A (en) * | 2019-12-31 | 2020-06-02 | 中国电子科技集团公司第三十六研究所 | Electromagnetic signal classification method and device |
| CN111340142A (en) * | 2020-05-14 | 2020-06-26 | 南京慧脑云计算有限公司 | Epilepsia magnetoencephalogram spike automatic detection method and tracing positioning system |
Non-Patent Citations (3)
| Title |
|---|
| JOAN BRUNA等: "Invariant Scattering Convolution Networks", 《IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE》 * |
| PHILIP A. WARRICK等: "Hybrid scattering-LSTM networks for automated detection of sleep", 《PHYSIOL. MEAS.》 * |
| 骆春桃: "基于脑电（EEG）的帕金森疾病早期诊断的研究", 《中国优秀硕士学位论文全文数据库 (医药卫生科技辑)》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113077795A (en) * | 2021-04-06 | 2021-07-06 | 重庆邮电大学 | Voiceprint recognition method under channel attention spreading and aggregation |
| CN113077795B (en) * | 2021-04-06 | 2022-07-15 | 重庆邮电大学 | Voiceprint recognition method under channel attention spreading and aggregation |
| CN113796873A (en) * | 2021-10-26 | 2021-12-17 | 山东建筑大学 | Wearable dynamic electrocardiosignal classification method and system |
| CN113796873B (en) * | 2021-10-26 | 2024-05-03 | 山东建筑大学 | Wearable dynamic electrocardiosignal classification method and system |
| CN114081462A (en) * | 2021-11-19 | 2022-02-25 | 齐齐哈尔大学 | Heart health monitoring system based on multi-dimensional physiological information |
| CN114358161A (en) * | 2021-12-27 | 2022-04-15 | 中国人民解放军国防科技大学 | Falling detection method and device based on wavelet scattering network and electronic equipment |
| CN114358161B (en) * | 2021-12-27 | 2024-02-02 | 中国人民解放军国防科技大学 | Fall detection method and device based on wavelet scattering network and electronic equipment |
| CN116172531A (en) * | 2023-04-25 | 2023-05-30 | 南京信息工程大学 | Blood pressure change trend estimation method based on wavelet dispersion transformation |
| CN116172531B (en) * | 2023-04-25 | 2023-07-21 | 南京信息工程大学 | Blood pressure change trend estimation method based on wavelet dispersion transformation |
| CN116898454A (en) * | 2023-07-11 | 2023-10-20 | 湖北大学 | Epileptic classification method and system based on electroencephalogram feature fusion deep learning model |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111956221B (en) | 2022-06-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111956221B (en) | Temporal lobe epilepsy classification method based on wavelet scattering factor and LSTM neural network model | |
| Wen et al. | Deep convolution neural network and autoencoders-based unsupervised feature learning of EEG signals | |
| Hamad et al. | Feature extraction of epilepsy EEG using discrete wavelet transform | |
| Acharya et al. | Automated diagnosis of epileptic EEG using entropies | |
| CN108564039A (en) | A kind of epileptic seizure prediction method generating confrontation network based on semi-supervised deep layer | |
| Bairy et al. | Automated diagnosis of depression electroencephalograph signals using linear prediction coding and higher order spectra features | |
| CN109431497B (en) | Electroencephalogram signal processing method and epilepsy detection system | |
| Prasanth et al. | Deep learning for interictal epileptiform spike detection from scalp EEG frequency sub bands | |
| Mustafa et al. | The analysis of eeg spectrogram image for brainwave balancing application using ann | |
| Wessam et al. | K-complexes detection in EEG signals using fractal and frequency features coupled with an ensemble classification model | |
| Sunaryono et al. | Gradient boosting machines fusion for automatic epilepsy detection from EEG signals based on wavelet features | |
| Jadhav et al. | Automated sleep stage scoring using time-frequency spectra convolution neural network | |
| CN114366124A (en) | Epilepsia electroencephalogram identification method based on semi-supervised deep convolution channel attention single classification network | |
| CN112641451A (en) | Multi-scale residual error network sleep staging method and system based on single-channel electroencephalogram signal | |
| Sheykhivand et al. | Sleep stage scoring of single-channel EEG signal based on RUSBoost classifier | |
| Thanaraj et al. | Multichannel interictal spike activity detection using time–frequency entropy measure | |
| Kbah et al. | Epileptic EEG activity detection for children using entropy-based biomarkers | |
| Satapathy et al. | A Deep Learning Approach to Automated Sleep Stages Classification Using Multi-Modal Signals | |
| Sidulova et al. | Towards Explainable Image Analysis for Alzheimer's Disease and Mild Cognitive Impairment Diagnosis | |
| Tripathi et al. | Automatic seizure detection and classification using super-resolution superlet transform and deep neural network-A preprocessing-less method | |
| Mohi-ud-Din et al. | Detection of Autism Spectrum Disorder from EEG signals using pre-trained deep convolution neural networks | |
| Li et al. | A one-dimensional convolutional neural network model for automated localization of epileptic foci | |
| Kaur et al. | Comparison of Classification Models Using Entropy Based Features from Sub-bands of EEG. | |
| Romaissa et al. | Epileptic seizure detection from imbalanced EEG signal | |
| Chavan et al. | EEG signals classification and diagnosis using wavelet transform and artificial neural network |
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 |