CN113229781A - Cross-patient epilepsy detection system based on multiple views - Google Patents
Cross-patient epilepsy detection system based on multiple views Download PDFInfo
- Publication number
- CN113229781A CN113229781A CN202110411678.9A CN202110411678A CN113229781A CN 113229781 A CN113229781 A CN 113229781A CN 202110411678 A CN202110411678 A CN 202110411678A CN 113229781 A CN113229781 A CN 113229781A
- Authority
- CN
- China
- Prior art keywords
- patient
- epilepsy
- view
- data
- module
- 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
- 206010015037 epilepsy Diseases 0.000 title claims abstract description 77
- 238000001514 detection method Methods 0.000 title claims abstract description 53
- 238000000926 separation method Methods 0.000 claims abstract description 25
- 230000001037 epileptic effect Effects 0.000 claims abstract description 18
- 238000005215 recombination Methods 0.000 claims abstract description 8
- 230000006798 recombination Effects 0.000 claims abstract description 8
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 7
- 206010010904 Convulsion Diseases 0.000 claims description 17
- 208000028329 epileptic seizure Diseases 0.000 claims description 8
- 238000000034 method Methods 0.000 claims description 7
- 238000011176 pooling Methods 0.000 claims description 7
- 238000007781 pre-processing Methods 0.000 claims description 4
- 230000000694 effects Effects 0.000 abstract description 10
- 208000037170 Delayed Emergence from Anesthesia Diseases 0.000 description 12
- 230000004913 activation Effects 0.000 description 8
- 238000010586 diagram Methods 0.000 description 6
- 230000006870 function Effects 0.000 description 6
- 238000010276 construction Methods 0.000 description 2
- 238000000537 electroencephalography Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 238000000638 solvent extraction Methods 0.000 description 2
- 230000005856 abnormality Effects 0.000 description 1
- 230000003042 antagnostic effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 210000003710 cerebral cortex Anatomy 0.000 description 1
- 230000002490 cerebral effect Effects 0.000 description 1
- 230000036461 convulsion Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 230000001314 paroxysmal effect Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 210000004761 scalp Anatomy 0.000 description 1
- 230000009885 systemic effect Effects 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
- A61B5/40—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
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/047—Probabilistic or stochastic networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H30/00—ICT specially adapted for the handling or processing of medical images
- G16H30/20—ICT specially adapted for the handling or processing of medical images for handling medical images, e.g. DICOM, HL7 or PACS
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients
Landscapes
- Health & Medical Sciences (AREA)
- Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Public Health (AREA)
- Artificial Intelligence (AREA)
- Biophysics (AREA)
- Medical Informatics (AREA)
- Molecular Biology (AREA)
- Theoretical Computer Science (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Pathology (AREA)
- Physiology (AREA)
- Veterinary Medicine (AREA)
- Animal Behavior & Ethology (AREA)
- Surgery (AREA)
- Evolutionary Computation (AREA)
- Heart & Thoracic Surgery (AREA)
- Computational Linguistics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Neurology (AREA)
- Software Systems (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Signal Processing (AREA)
- Computing Systems (AREA)
- Psychiatry (AREA)
- Neurosurgery (AREA)
- Epidemiology (AREA)
- Primary Health Care (AREA)
- Fuzzy Systems (AREA)
- Probability & Statistics with Applications (AREA)
- Databases & Information Systems (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention relates to a multi-view-based cross-patient epilepsy detection system, which divides original electroencephalogram data into samples according to the same size to form a training set and a testing set; performing discrete Fourier transform processing on the obtained data set to obtain frequency domain data, performing wavelet packet decomposition processing on the obtained data set to obtain time-frequency domain data, wherein the time domain data, the frequency domain data and the time-frequency domain data jointly form multi-view data input by a model; decomposing the multi-view data into epilepsy characteristics and patient characteristics through a characteristic separation and recombination module, wherein the epilepsy characteristics and the patient characteristics are respectively used as the input of an epilepsy detection module and a patient classification module to carry out epilepsy detection and patient classification; the trained epileptic and patient characteristics can be recombined into the original multi-view characteristics. The system further improves the effect of the cross-patient epilepsy detection by using the multi-view feature decomposition and recombination, the epilepsy detection and the confrontation training of patient classification.
Description
Technical Field
The invention relates to the field of epilepsy detection, in particular to a multi-view-based cross-patient epilepsy detection system.
Background
The statements in this section merely provide background information related to the present disclosure and may not necessarily constitute prior art.
Epilepsy refers to a clinical phenomenon caused by paroxysmal abnormality and excessive hypersynchronized discharge of cerebral neurons. The epileptic seizure is usually a systemic seizure, which can cause people to lose consciousness, is accompanied by phenomena such as convulsion, and seriously affects the normal life of people, even endangers life.
Electroencephalogram is the most convenient and economical means for diagnosing epileptic seizure at present, and the epileptic seizure is diagnosed by observing electroencephalogram waveforms by doctors. At present, deep learning methods are gradually used to replace or assist doctors in completing epilepsy detection.
In a system for implementing epilepsy detection by using electroencephalogram, there is a serious problem that the current epilepsy detection system has good effect on most patients due to physiological differences among patients, but it is difficult to accurately judge the electroencephalogram characteristics of epilepsy on a part of specific patients.
Disclosure of Invention
In order to solve the technical problems in the background art, the invention provides a multi-view-based cross-patient epilepsy detection system, which separates and recombines epilepsy features and patient features in electroencephalogram data, performs epilepsy detection and patient classification, and solves the problem of unsatisfactory epilepsy detection effect caused by physiological differences of patients.
In order to achieve the purpose, the invention adopts the following technical scheme:
a first aspect of the invention provides a multi-view based cross-patient epilepsy detection system, comprising:
a data pre-processing module configured to: acquiring electroencephalogram data to be detected, and dividing a sample;
a multi-view feature module configured to: based on the divided data, respectively carrying out discrete Fourier transform and wavelet packet decomposition processing to obtain frequency domain data and time-frequency domain data; the frequency domain data, the time-frequency domain data and the time domain data are taken as electroencephalogram data with multi-view characteristics;
a feature separation and reassembly module configured to: separating to obtain epileptic characteristics and patient characteristics based on the electroencephalogram data with the multi-view characteristics; the epilepsy features and the patient features are respectively used as training inputs of an epilepsy detection module and a patient classification module, and the trained epilepsy features and the trained patient features are recombined into original multi-view features;
an epilepsy detection module configured to: classifying the epilepsy characteristics into epileptic seizure and epileptic non-seizure after training based on the epilepsy characteristics of the characteristic separation and recombination module;
a patient classification module configured to: and classifying the patients after training based on the patient characteristics of the characteristic separation and recombination module.
And the detection system finally outputs the patient classification result and the epilepsy detection result.
The time domain signal is the original signal, which is the time domain feature part of the multi-view feature.
The frequency domain signal is obtained by discrete Fourier transform of an original signal, and is used as a frequency domain characteristic part of the multi-view characteristic.
The time-frequency domain signal is obtained by decomposing a wavelet packet of an original signal and is used as a time-frequency domain characteristic part of the multi-view characteristic.
The characteristic separation and characteristic reconstruction module comprises a characteristic separation module and a characteristic reconstruction module.
The feature separation module comprises at least one convolutional layer and at least one max-pooling layer, and separates epileptic features from patient features.
And the characteristic reconstruction module performs deconvolution on the epileptic characteristics through the deconvolution layer.
The feature reconstruction module deconvolves the patient features through a deconvolution layer.
And the characteristic reconstruction module linearly adds the deconvoluted epilepsy characteristics and the patient characteristics to obtain reconstructed characteristics.
Compared with the prior art, the above one or more technical schemes have the following beneficial effects:
1. the epilepsia features and the patient features in the electroencephalogram data are separated and recombined to carry out epilepsia detection and patient classification, so that the problem of unsatisfactory epilepsia detection effect caused by physiological differences of patients is solved.
2. Through analyzing the electroencephalogram signals of a period of time, original electroencephalogram data are transformed and divided into multi-view data, then the multi-view data are separated into epilepsy characteristics and patient characteristics to serve as training input, the trained epilepsy characteristics and patient characteristics can be recombined into original multi-view characteristics, and the detection effect of the system is improved.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, are included to provide a further understanding of the invention, and are incorporated in and constitute a part of this specification, illustrate exemplary embodiments of the invention and together with the description serve to explain the invention and not to limit the invention.
Fig. 1 is an electroencephalogram of epileptic seizures provided by one or more embodiments of the present invention;
fig. 2 is a schematic diagram of an epilepsy detection process provided by one or more embodiments of the present invention;
FIG. 3 is a schematic diagram of electroencephalogram data partitioning provided by one or more embodiments of the present invention;
FIG. 4 is a multi-view feature building diagram provided by one or more embodiments of the invention;
FIG. 5 is a schematic illustration of feature decomposition and reconstruction provided by one or more embodiments of the invention;
FIG. 6 is a schematic diagram of a patient classification module architecture provided in accordance with one or more embodiments of the invention;
fig. 7 is a schematic diagram of an epilepsy detection module architecture provided in one or more embodiments of the present invention;
FIG. 8 is a schematic diagram of output results provided by one or more embodiments of the invention.
Detailed Description
The invention is further described with reference to the following figures and examples.
It is to be understood that the following detailed description is exemplary and is intended to provide further explanation of the invention as claimed. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of exemplary embodiments according to the invention. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, and it should be understood that when the terms "comprises" and/or "comprising" are used in this specification, they specify the presence of stated features, steps, operations, devices, components, and/or combinations thereof, unless the context clearly indicates otherwise.
As described in the background art, in the conventional epilepsy detecting system, it is difficult to accurately judge the electroencephalogram characteristics of epilepsy due to the physiological differences of a part of specific patients in the electroencephalogram, so that the detection effect of the whole system is not ideal.
The following embodiments therefore present a multi-view based, cross-patient epilepsy detection system, an improvement over the system for detecting epilepsy using electroencephalography itself,
the first embodiment is as follows:
as shown in fig. 2, a multi-view based cross-patient epilepsy detection system, comprising:
the data preprocessing module, namely the data partitioning in fig. 2, is configured to: acquiring electroencephalogram data to be detected acquired by an N-channel electrode; carrying out sample division processing on the acquired electroencephalogram data to be detected; n is a positive integer;
a multi-view feature module, i.e., multi-view feature extraction in fig. 2, configured to: performing discrete Fourier transform processing on the data after data preprocessing to acquire frequency domain data; and carrying out wavelet packet decomposition processing based on the divided data to obtain time-frequency domain data. The time domain data, the frequency domain data and the time-frequency domain data are used as three views of the data together.
A feature separation and reassembly module, i.e., the feature separation and reassembly module of fig. 2, configured to: electroencephalogram data based on multi-view characteristics are subjected to epileptic characteristic separation convolution layers, and epileptic characteristics are separated from the electroencephalogram data; separating the physiological characteristics of the patient from the electroencephalogram data through the patient characteristic separation convolutional layer; the epilepsy features and the patient features are respectively used as input of an epilepsy detection module and a patient classification module, and the trained epilepsy features and patient features can be recombined into original multi-view features.
An epilepsy detection module configured to: based on the epileptic features of the feature separation and recombination module, further training of the model is carried out through an M-layer ResNet layer, and finally, the model is classified into epileptic seizure and epileptic non-seizure; m is a positive integer;
a patient classification module configured to: based on the patient characteristics of the characteristic separation and recombination module, further training the model through an M-layer ResNet layer, and finally classifying the patient; m is a positive integer;
acquiring electroencephalogram data to be detected acquired by an N-channel electrode; in the embodiment, the method comprises the following steps: acquiring an electroencephalogram signal containing 23 dual-channel electrodes, and acquiring the electroencephalogram signal to obtain electroencephalogram data to be detected.
Fig. 1 shows an electroencephalogram of seizures acquired, from which changes in brain waves at the time of seizures can be visually perceived. EEG data, which is acquired by placing electrodes in the cerebral cortex, requires that the resistance at each electrode site on the scalp be 0.
As one or more embodiments, as shown in fig. 3, dividing acquired electroencephalogram data to be detected; the method specifically comprises the following steps:
dividing the electroencephalogram data according to time intervals;
in order to enrich data, the data are divided in a sliding window mode.
Further, dividing the electroencephalogram data at time intervals in an overlapping mode; the method specifically comprises the following steps:
and (3) carrying out sample division on the acquired electroencephalogram data, wherein the length of a sample to be tested is 1 second, and when the sample is divided, the division needs to be carried out by overlapping 50%. As shown in fig. 3, of the left, middle and right regions, the left + middle region is a sample example, and the middle + right region is a sample example, with an overlapping portion of 50% (and the middle overlapping region), so that several samples to be tested are obtained.
Furthermore, z-score standardization processing is carried out on electroencephalogram data, the maximum value and the minimum value of all electroencephalogram signals are solved, then the mean value is subtracted, and the standard deviation is removed, so that the model is more favorable for convergence.
As one or more embodiments, multi-view feature construction is performed based on the partitioned data; as shown in fig. 4, the specific steps include:
the data are subjected to feature construction from three views of a time domain, a frequency domain and a time-frequency domain, an original signal is a time domain signal, and the original signal can be used as a time domain feature part of combined features; converting the original signal into a frequency domain signal through discrete Fourier transform, and taking the frequency domain signal as a frequency domain characteristic part of the combined characteristic; and converting the original signal into a time-frequency domain signal through wavelet packet decomposition, and taking the time-frequency domain signal as a time-frequency domain characteristic part of the combined characteristic. Thus, a multi-view feature is constructed.
As one or more embodiments, the multi-view features are input into the feature separation and feature reconstruction module as an input of the model, as shown in fig. 5, the feature separation module decomposes the multi-view features into epilepsy features and patient features, and inputs the epilepsy features and the patient features into the epilepsy detection module and the patient classification module, as shown in fig. 6 and 7, respectively. And the characteristic reconstruction module perfectly reconstructs the separated epileptic characteristic and the patient characteristic into the original characteristic. The feature separation and feature reconstruction module herein includes:
the characteristic separation module is used for separating the epileptic characteristics and comprises a convolution layer and a maximum pooling layer; second, the patient features are separated, including one convolutional layer, one maximally pooled layer.
A feature reconstruction module, which firstly restores the feature dimension to the dimension before feature separation, namely, the epileptic feature is deconvoluted through a deconvolution layer, the patient feature is deconvoluted through a deconvolution layer, and the patient feature is deconvoluted through a deconvolution layer; second, the deconvolved epileptic feature is linearly summed with the patient feature and then divided by 2 to yield the reconstructed feature.
As one or more embodiments, the epilepsy detection module based on the separated epilepsy features specifically includes, as shown in fig. 7, the following specific steps:
the epileptic features first pass through a convolutional layer, a regularization layer, a Relu activation function, a maximum pooling layer, and then three superimposed residual blocks.
Each residual block contains convolutional layers with a convolution kernel of 1 x 1 and a filter number of 4 x i, regularization layers, Relu activation functions, convolutional layers with a convolution kernel of 3 x 3 and a filter number of 4 x i, regularization layers, Relu activation functions, convolutional layers with a convolution kernel of 1 x 1 and a filter number of 4 x (i +1), regularization layers, where residual connection contains convolutional layers with a convolution kernel of 1 x 1 and a filter number of 4 x (i +1), regularization layers. Where i is 1 in the first residual block, i is 2 in the first residual block, and i is 3 in the first residual block.
The output of the last layer of residual block passes through an average pooling layer and is input to a full-link layer, the unit of which is 2, which is determined by the status of epileptic seizure and the interval of epileptic seizure. The last Softmax performs activation classification prediction on the sample.
As one or more embodiments, the patient classification module based on the separated patient features, as shown in fig. 6, includes the following steps:
the patient features first pass through a convolutional layer, a regularization layer, a Relu activation function, a maximum pooling layer, and then three superimposed residual blocks.
Each residual block contains convolutional layers with a convolution kernel of 1 x 1 and a filter number of 4 x i, regularization layers, Relu activation functions, convolutional layers with a convolution kernel of 3 x 3 and a filter number of 4 x i, regularization layers, Relu activation functions, convolutional layers with a convolution kernel of 1 x 1 and a filter number of 4 x (i +1), regularization layers, where residual connection contains convolutional layers with a convolution kernel of 1 x 1 and a filter number of 4 x (i +1), regularization layers. Where i is 1 in the first residual block, i is 2 in the first residual block, and i is 3 in the first residual block.
And the output of the last layer of residual block passes through an average pooling layer and is input into a full-connection layer, and the final Softmax carries out activation classification prediction on the samples.
The degree of characteristic separation directly determines the effect of epilepsy detection, and the separated epilepsy characteristics and patient characteristics are ensured to be restored to the original characteristics without deformation. The game of the two parties forms antagonistic training, and the effect of the epilepsy detection classifier is facilitated.
The system in the embodiment has a good effect on the detection of the epilepsy of the cross-patient, as shown in fig. 8, the electroencephalogram signals of 8 patients are respectively detected, the best detection accuracy rate reaches more than 90%, and the lowest detection accuracy rate is also more than 70%, so that the superiority of the system is fully proved.
The electroencephalogram data can be analyzed to detect whether epileptic seizures occur within a certain period of time, so that the basis for treatment can be provided.
According to the method, the epilepsia characteristics and the patient characteristics in the electroencephalogram data are separated and recombined to carry out epilepsia detection and patient classification, and the problem that the epilepsia detection effect is not ideal due to physiological differences of patients is solved.
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (10)
1. A multi-view based cross-patient epilepsy detection system, characterized by: the method comprises the following steps:
a data pre-processing module configured to: acquiring electroencephalogram data to be detected, and dividing a sample;
a multi-view feature module configured to: based on the divided data, respectively carrying out discrete Fourier transform and wavelet packet decomposition processing to obtain frequency domain data and time-frequency domain data; the frequency domain data, the time-frequency domain data and the time domain data are taken as electroencephalogram data with multi-view characteristics;
a feature separation and reassembly module configured to: separating to obtain epileptic characteristics and patient characteristics based on the electroencephalogram data with the multi-view characteristics; the epilepsy features and the patient features are respectively used as training inputs of an epilepsy detection module and a patient classification module, and the trained epilepsy features and the trained patient features are recombined into original multi-view features;
an epilepsy detection module configured to: classifying the epilepsy characteristics into epileptic seizure and epileptic non-seizure after training based on the epilepsy characteristics of the characteristic separation and recombination module;
a patient classification module configured to: and classifying the patients after training based on the patient characteristics of the characteristic separation and recombination module.
2. A multi-view based cross-patient epilepsy detection system according to claim 1, wherein: the time domain signal is an original signal which is a time domain feature part of the multi-view feature.
3. A multi-view based cross-patient epilepsy detection system according to claim 1, wherein: the frequency domain signal is obtained by discrete Fourier transform of an original signal and is used as a frequency domain characteristic part of the multi-view characteristic.
4. A multi-view based cross-patient epilepsy detection system according to claim 1, wherein: the time-frequency domain signal is obtained by decomposing an original signal through a wavelet packet and is used as a time-frequency domain characteristic part of the multi-view characteristic.
5. A multi-view based cross-patient epilepsy detection system according to claim 1, wherein: the characteristic separation and characteristic reconstruction module comprises a characteristic separation module and a characteristic reconstruction module.
6. A multi-view based cross-patient epilepsy detection system according to claim 5, wherein: the feature separation module comprises at least one convolutional layer and at least one max-pooling layer, and separates epileptic features from patient features.
7. A multi-view based cross-patient epilepsy detection system according to claim 5, wherein: and the characteristic reconstruction module carries out deconvolution on the epileptic characteristics through a deconvolution layer.
8. A multi-view based cross-patient epilepsy detection system according to claim 5, wherein: the feature reconstruction module deconvolves patient features through a deconvolution layer.
9. A multi-view based cross-patient epilepsy detection system according to claim 7 or 8, characterized in that: and the characteristic reconstruction module linearly adds the deconvoluted epilepsy characteristics and the patient characteristics to obtain reconstructed characteristics.
10. A multi-view based cross-patient epilepsy detection system according to claim 1, wherein: the detection system outputs patient classification results and epilepsy detection results.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110411678.9A CN113229781A (en) | 2021-04-16 | 2021-04-16 | Cross-patient epilepsy detection system based on multiple views |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110411678.9A CN113229781A (en) | 2021-04-16 | 2021-04-16 | Cross-patient epilepsy detection system based on multiple views |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113229781A true CN113229781A (en) | 2021-08-10 |
Family
ID=77128323
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110411678.9A Pending CN113229781A (en) | 2021-04-16 | 2021-04-16 | Cross-patient epilepsy detection system based on multiple views |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113229781A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116570297A (en) * | 2023-07-13 | 2023-08-11 | 四川君健万峰医疗器械有限责任公司 | Epileptic brain wave signal identification method |
-
2021
- 2021-04-16 CN CN202110411678.9A patent/CN113229781A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116570297A (en) * | 2023-07-13 | 2023-08-11 | 四川君健万峰医疗器械有限责任公司 | Epileptic brain wave signal identification method |
| CN116570297B (en) * | 2023-07-13 | 2023-10-13 | 四川君健万峰医疗器械有限责任公司 | Epileptic brain wave signal identification method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Cheong et al. | Feature extraction of EEG signal using wavelet transform for autism classification | |
| CN113786204B (en) | Epileptic intracranial brain electrical signal early warning method based on deep convolution attention network | |
| Fathima et al. | Wavelet based features for epileptic seizure detection | |
| AlSharabi et al. | EEG signal processing for Alzheimer’s disorders using discrete wavelet transform and machine learning approaches | |
| CN106236080A (en) | Based on the removing method of myoelectricity noise in multichannel EEG signals | |
| CN115363531A (en) | Epilepsy detection system based on bimodal electroencephalogram signal information bottleneck | |
| Khan et al. | Wavelet-based multi-class discrimination of EEG for seizure detection | |
| CN113229781A (en) | Cross-patient epilepsy detection system based on multiple views | |
| Kbah et al. | Epileptic EEG activity detection for children using entropy-based biomarkers | |
| Christensen et al. | Separation of Parkinson's patients in early and mature stages from control subjects using one EOG channel | |
| Smitha et al. | Classifying subjective emotional stress response evoked by multitasking using EEG | |
| Mohi-ud-Din et al. | Detection of Autism Spectrum Disorder from EEG signals using pre-trained deep convolution neural networks | |
| Melinda et al. | The effect of power spectral density on the electroencephalography of autistic children based on the welch periodogram method | |
| Abásolo1 et al. | Approximate entropy of EEG background activity in alzheimer's disease patients | |
| Rajaei et al. | Pediatric epilepsy: Clustering by functional connectivity using phase synchronization | |
| Singh et al. | Frequency band separation for epilepsy detection using EEG | |
| EP3745947B1 (en) | Method for detecting a conversion from mild cognitive impairment to alzheimer disease | |
| Radhakrishnan et al. | Investigating EEG Signals of Autistic Individuals Using Detrended Fluctuation Analysis. | |
| Fathima et al. | Wavelet based detection of epileptic seizures using scalp EEG | |
| Wei et al. | Automatic recognition of chewing noises in epileptic EEG based on period segmentation | |
| Pedrollo et al. | Spiking neural networks diagnosis of ADHD subtypes through EEG signals evaluation | |
| Melinda et al. | Classification of EEG Signal using Independent Component Analysis and Discrete Wavelet Transform based on Linear Discriminant Analysis | |
| Kalbhor et al. | The review of detection and classification of epilectic seizures using wavelet transform | |
| Mohi ud Din et al. | Automated classification of autism spectrum disorder using EEG signals and convolutional neural networks | |
| Janwattanapong et al. | Connectivity patterns of interictal epileptiform discharges using coherence analysis |
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: 20210810 |
|
| RJ01 | Rejection of invention patent application after publication |