CN113288050B - Multidimensional enhanced epileptic seizure prediction system based on graph convolution network - Google Patents
Multidimensional enhanced epileptic seizure prediction system based on graph convolution network Download PDFInfo
- Publication number
- CN113288050B CN113288050B CN202110443244.7A CN202110443244A CN113288050B CN 113288050 B CN113288050 B CN 113288050B CN 202110443244 A CN202110443244 A CN 202110443244A CN 113288050 B CN113288050 B CN 113288050B
- Authority
- CN
- China
- Prior art keywords
- graph
- data
- space
- electroencephalogram
- time
- 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
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7271—Specific aspects of physiological measurement analysis
- A61B5/7275—Determining trends in physiological measurement data; Predicting development of a medical condition based on physiological measurements, e.g. determining a risk factor
-
- 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
- A61B5/4094—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
- A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Public Health (AREA)
- Surgery (AREA)
- Veterinary Medicine (AREA)
- General Health & Medical Sciences (AREA)
- Animal Behavior & Ethology (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Molecular Biology (AREA)
- Psychiatry (AREA)
- Physiology (AREA)
- Artificial Intelligence (AREA)
- Neurology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Signal Processing (AREA)
- Neurosurgery (AREA)
- Evolutionary Computation (AREA)
- Fuzzy Systems (AREA)
- Mathematical Physics (AREA)
- Psychology (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The scheme takes a multi-channel spatial relationship as a breakthrough, explores the contribution degree of the multi-channel spatial relationship to epileptic seizure prediction from three dimensions of frequency, space and time, and comprises three components, namely an information reconstruction space, a graph encoder and a space-time predictor. The information reconstruction space and the image encoder mentioned by the model allow feature enhancement and feature extraction of richer epileptic electroencephalogram signals, and particularly explore the correlation among channels of electroencephalogram, so that feature representation is enhanced, and the accuracy of epileptic electroencephalogram prediction attack is improved; meanwhile, a core structure of the space-time predictor in the scheme adopts a gate control circulation unit to explore the rule of epileptic electroencephalogram signals in a time layer, and the goal of improving the operation efficiency of the model is achieved by reducing the scale of network parameters.
Description
Technical Field
The disclosure belongs to the technical field of electroencephalogram signal processing, and particularly relates to a multi-dimensional enhanced epileptic seizure prediction system based on a graph convolution network.
Background
The statements in this section merely provide background information related to the present disclosure and may not necessarily constitute prior art.
Epilepsy is a chronic disease of functional disorder caused by abnormal discharge of cerebral neurons, and the symptom is paroxysmal and transient. Moreover, the prevalence rate is high, and drug treatment and surgical resection are the main treatment forms, and the occurrence of the disease causes great psychological stress and serious life trouble for patients due to unpredictability. In the clinical diagnosis, medical workers usually perform long-time electroencephalogram monitoring on patients, manually analyze relevant electroencephalogram characteristics, and diagnose the disease conditions, but doctors usually have difficulty in predicting the occurrence of diseases.
Seizure prediction consists in detecting changes in the characteristics of the signal prior to a seizure. Numerous studies have found that there is a transition between the inter-and intra-seizure phase, referred to as pre-seizure. Thus, the prediction of epileptic seizures can be considered as being directed to early detection of pre-seizure.
In the task oriented to epileptic seizure prediction, the inventor finds that more existing methods mainly focus on signal data of each channel, neglect correlation between the channels, and excavate spatial relationships between the channels to have insufficient depth and shallow exploration dimensions, which results in the problem of low accuracy of the existing epileptic seizure prediction system.
Disclosure of Invention
In order to solve the above problems, the present disclosure provides a Multi-dimensional Enhanced Seizure Prediction system (MESPF) based on a graph convolution network, where the scheme performs feature enhancement and feature extraction on richer epileptic electroencephalogram signals through an information reconstruction space and a graph encoder, and improves the correlation among channels of the electroencephalogram, thereby enhancing feature representation and improving the accuracy of epileptic electroencephalogram Prediction seizures; meanwhile, the rule of the epilepsia electroencephalogram signals in the time layer is explored through a space-time predictor, and the purpose of improving the operation efficiency of the model is achieved by reducing the scale of network parameters.
According to a first aspect of embodiments of the present disclosure, there is provided a multidimensional enhanced seizure prediction system based on a graph-volume network, including:
the data acquisition unit is configured to acquire electroencephalogram data to be detected and preprocess the acquired electroencephalogram data to be detected;
the information reconstruction space unit is configured to decompose and reconstruct each channel of the preprocessed electroencephalogram data by using the information reconstruction space model, calculate the correlation among the channels based on the reconstructed data and generate graph data;
a graph encoder unit configured to perform feature extraction and encoding on the graph data using a graph encoder;
and the space-time predictor unit is configured to input the coded graph state code data into the space-time predictor to serve as state input data of the space-time predictor in the current time period, further input the graph state code data in sequence according to a time sequence, and then obtain a prediction result through the multilayer perceptron.
Further, decomposing and reconstructing each channel of the preprocessed electroencephalogram data by using the information reconstruction space model specifically comprise: decomposing all epilepsia electroencephalogram signals in high-frequency and low-frequency ranges by utilizing a wavelet packet decomposition technology, calculating energy value characteristics of each frequency band, and reconstructing the representation of each channel data unit; and calculating the correlation among multiple channels by using a Pearson correlation coefficient calculation method to generate graph data.
Further, the performing feature extraction and encoding on the graph data by using a graph encoder specifically includes: the graph encoder adopts a graph convolution network, the input part of the graph encoder comprises a graph characteristic vector obtained in an information reconstruction space and the correlation relation between channels of an electroencephalogram signal, and after the graph space characteristic is extracted through a graph convolution layer, the characteristic is subjected to weighted synthesis through a full connection layer to generate a graph state code.
According to a third aspect of the embodiments of the present disclosure, there is provided an electronic device, including a memory, a processor, and a computer program stored in the memory and running on the memory, wherein the processor, when executing the program, implements the functions of the graph-volume-network-based multi-dimensional enhanced seizure prediction system.
According to a fourth aspect of embodiments of the present disclosure, there is provided a non-transitory computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the functionality of the graph volume network based multi-dimensional enhanced seizure prediction system.
Compared with the prior art, the beneficial effect of this disclosure is:
the scheme takes a multi-channel spatial relationship as a breakthrough, explores the contribution degree of the multi-channel spatial relationship to epileptic seizure prediction from three dimensions of frequency, space and time, and comprises three components, namely an information reconstruction space, a graph encoder and a space-time predictor. The information reconstruction space and the graph encoder mentioned by the model allow feature enhancement and feature extraction of richer epilepsia electroencephalogram signals, and particularly explore the correlation among channels of the electroencephalogram, so that feature representation is enhanced, and the accuracy of epilepsia electroencephalogram prediction attack is improved. And secondly, a core structure of the space-time predictor mentioned by the model adopts a gate control circulation unit to explore the rule of epileptic electroencephalogram signals in a time layer, and the goal of improving the running efficiency of the model is achieved by reducing the scale of network parameters.
Advantages of additional aspects of the disclosure will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the disclosure.
Drawings
The accompanying drawings, which are included to provide a further understanding of the disclosure, illustrate embodiments of the disclosure and together with the description serve to explain the disclosure and are not to limit the disclosure.
Fig. 1 is a composition diagram of a multidimensional enhanced seizure prediction system based on a graph-volume network according to a first embodiment of the disclosure.
FIG. 2 is an exemplary graph of an epileptic brain electrical signal waveform obtained from a CHB-MIT data set according to a first embodiment of the disclosure;
FIG. 3 is a graph illustrating criteria for dividing the time intervals of seizures during data preprocessing as described in the first embodiment of the present disclosure;
fig. 4 is a graph of the relationship among 18 channels extracted from the epileptic electroencephalogram signal according to the first embodiment of the present disclosure;
FIG. 5 is a block diagram of logic within a block diagram encoder according to one embodiment of the present disclosure;
fig. 6 is an internal structure diagram of a space-time predictor according to a first embodiment of the present disclosure.
Detailed Description
The present disclosure is further illustrated by the following examples in conjunction with the accompanying drawings.
It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the disclosure. 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 disclosure 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 example embodiments according to the present disclosure. 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.
The embodiments and features of the embodiments in the present disclosure may be combined with each other without conflict.
The first embodiment is as follows:
the embodiment aims to provide a multidimensional enhanced epileptic seizure prediction system based on a graph convolution network.
With reference to fig. 1, a multidimensional enhanced seizure prediction system based on a graph-volume network includes:
the data acquisition unit is configured to acquire electroencephalogram data to be detected and preprocess the acquired electroencephalogram data to be detected;
the acquiring of the electroencephalogram signal data specifically comprises the following steps:
acquiring epilepsia electroencephalogram signal data: using the CHB-MIT data set as a source of epileptic electroencephalogram data, and acquiring required signal data from the CHB-MIT data set; as shown in fig. 2, an example of an epileptic brain electrical signal waveform obtained from the CHB-MIT data set is shown;
the pretreatment specifically comprises:
preprocessing the acquired epilepsia electroencephalogram signals: first, the electroencephalographic recording of each case in the data set is divided into time periods. That is, the individual periods of the epileptic seizure are defined as the period of seizure, the period of seizure pre-period, and the period of seizure inter-period, which are respectively referenced to the gray, light gray, and dark gray labeled periods in the figure, which shows the criteria for dividing the different periods in detail, the period of seizure pre-period is defined as the signal data between 5 minutes and 1 hour before the seizure, and the period of seizure inter-period is defined as the signal data four hours before the seizure or four hours after the seizure. And secondly, dividing epilepsia electroencephalogram signal data units. Since the signal of the CHB-MIT data set is sampled at 256 sample points per second, we define 1024 points as one data unit. Because the data duration of the epileptic seizure interval is longer than that of the epileptic seizure in the early stage, in order to ensure the scientificity of experimental performance evaluation, the balance of positive and negative samples is carried out before the experiment, namely, the invention increases the sample amount in the early stage of epileptic seizure by setting the overlapping rate of the window in the early stage of epileptic seizure to be 50 percent in the experiment. In this experiment, the pre-episode samples were defined as positive samples and the inter-episode samples as negative samples.
The information reconstruction space unit is configured to decompose and reconstruct each channel of the preprocessed electroencephalogram data by using the information reconstruction space model, calculate the correlation among the channels based on the reconstructed data and generate graph data;
the information reconstruction space module specifically comprises the following operations:
in an information reconstruction space, the model introduces a wavelet packet decomposition technology, 5-layer wavelet packet decomposition is carried out on all epileptic electroencephalogram signals in high-frequency and low-frequency ranges to obtain 32-segment frequency band wavelet packet coefficients, energy value characteristics of each sub-frequency band are calculated, and 32 pieces of energy characteristic values are reconstructed to represent data of each channel. Furthermore, in order to enhance the relationship between the epileptic brain electrical signal channels, correlation between the multiple channels is calculated using a pearson correlation coefficient calculation method, and graph data is generated. It should be noted that the number of channels is 18, and the correlation between the multiple channels obtained by calculation is an 18 × 18 data matrix.
A graph encoder unit configured to perform feature extraction and encoding on the graph data using a graph encoder;
wherein the graph encoder module includes the operations of:
the image encoder performs image data feature extraction and encoding on the reconstruction information, and specifically comprises the following steps: in order to process the graph data more efficiently, the graph encoder builds a relevant encoding model by relying on a graph convolution network. The model structure comprises an input layer, a first graph volume layer, a second graph volume layer, a third graph volume layer, a full connection layer and a graph state code output layer. The input of the model comprises two parts, one part is a graph feature vector obtained in an information reconstruction space, the specific dimension is 18 x 32, the graph feature vector is up-sampled, the dimension is expanded to 18 x 256, and the dimension is used as one of the input of a graph encoder. The other part is the correlation relationship among multiple channels of the epileptic brain electrical signals, and the specific dimension is 18 multiplied by 18. The graph space feature extraction is carried out by three layers of graph volume layers, the dimension of the MESPF first graph volume layer is 18 multiplied by 128, the dimension of the MESPF second graph volume layer is 18 multiplied by 64, and the dimension of the MESPF third graph volume layer is 18 multiplied by 32. And carrying out weighted synthesis on the features through the full connection layer. Finally, a graph state code is generated, which includes 18 feature values to characterize 18 channels. It should be noted that, each segment of the calculated state code of the graph is executed according to a time sequence, and each segment of the state code represents a state of a current time period and is sequentially input to the space-time predictor.
The space-time predictor unit is configured to input the coded graph state code characteristic data into the space-time predictor and execute a prediction task;
wherein the spatial predictor module specifically comprises the operations of:
and exploring the rule of the graph state code data in a time sequence layer by using a space-time predictor, specifically, coding the data units by using a graph coder according to a time sequence, and sequentially using the graph state code completing the process as the input of the space-time predictor. The core structure of the space-time predictor is a gate control circulation unit, the structure combines a forgetting gate and an input gate into an updating gate N, and combines a unit state U and a state H into a composite state C. Reset gate R t The graph state code data for controlling the last time is inputted to R t Aggregated data, R t The smaller the value of (a), the less the graph state code data representing the last time is written, and the operation formula is: r t ＝σ(W r [C t-1 ，S MESPF_t ]) Wherein W is r Representing a weight matrix, σ representsSigmoid function, mapping data to 0 to 1. Updating door N t For controlling the degree to which the picture-state code data at the previous moment are input to the current state, N t The smaller the value of (a), the less graph state code data representing the last time is written, and the calculation formula is as follows: n is a radical of hydrogen t ＝σ(W N [C t-1 ，S MESPF_t ]) (ii) a The calculation formula of the candidate hidden layer is as follows:
wherein, tanh represents tanh function, the mapping range of which is-1 to +1, and hidden layer information C is output t The calculation formula of (2):
and then, outputting a prediction result through a multilayer perceptron, namely, estimating the state output of the next time interval by a graph state code through a space-time predictor, and further outputting the seizure prediction result of the epileptic electroencephalogram signal through the multilayer perceptron. In further embodiments, there is also provided:
an electronic device comprising a memory and a processor, and computer instructions stored on the memory and executed on the processor, the computer instructions, when executed by the processor, implement the functions of the system of the first embodiment. For brevity, no further description is provided herein.
It should be understood that in this embodiment, the processor may be a central processing unit CPU, and the processor may also be other general purpose processors, digital signal processors DSP, application specific integrated circuits ASIC, off-the-shelf programmable gate arrays FPGA or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, and so on. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
The memory may include both read-only memory and random access memory and may provide instructions and data to the processor, and a portion of the memory may also include non-volatile random access memory. For example, the memory may also store device type information.
A computer readable storage medium storing computer instructions which, when executed by a processor, implement the functions of the system of the first embodiment.
The system in the first embodiment may be implemented by code, and may be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules in a processor. The software modules may be located in ram, flash, rom, prom, or eprom, registers, among other storage media as is well known in the art. The storage medium is located in a memory, and a processor reads information in the memory and completes the steps of the method in combination with hardware of the processor. To avoid repetition, it is not described in detail here.
Those of ordinary skill in the art will appreciate that the various illustrative elements, i.e., algorithm steps, described in connection with the embodiments disclosed herein may be implemented as electronic hardware or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.
The multidimensional enhancement epileptic seizure prediction system based on the graph convolution network can be realized and has wide application prospect.
The above description is only a preferred embodiment of the present disclosure and is not intended to limit the present disclosure, and various modifications and changes may be made to the present disclosure by those skilled in the art. Any modification, equivalent replacement, improvement and the like made within the spirit and principle of the present disclosure should be included in the protection scope of the present disclosure.
Although the present disclosure has been described with reference to specific embodiments, it should be understood that the scope of the present disclosure is not limited thereto, and those skilled in the art will appreciate that various modifications and changes can be made without departing from the spirit and scope of the present disclosure.
Claims (7)
1. A multidimensional enhanced epileptic seizure prediction system based on a graph convolution network is characterized by comprising:
the data acquisition unit is configured to acquire electroencephalogram data to be detected and preprocess the acquired electroencephalogram data to be detected;
the information reconstruction space unit is configured to decompose and reconstruct each channel of the preprocessed electroencephalogram data by using an information reconstruction space model, calculate the correlation among the channels based on the reconstructed data and generate graph data;
a graph encoder unit configured to perform feature extraction and encoding on the graph data using a graph encoder; the graph encoder adopts a graph convolution network, the input part of the graph encoder comprises a graph characteristic vector obtained in an information reconstruction space and a correlation relation between channels of an electroencephalogram signal, and after graph space characteristics are extracted through a graph convolution layer, the characteristics are subjected to weighted synthesis through a full connection layer to generate a graph state code;
the space-time predictor unit is configured to input the coded graph state code data into the space-time predictor to serve as state input data of the space-time predictor in the current time period, further, the graph state code data are sequentially input according to a time sequence, and then a prediction result is obtained through the multilayer perceptron;
the space-time predictor is composed of a gating circulation unit, receives the graph state code output by the graph encoder according to the time sequence and estimates the state output of the next time period; in the space-time predictor, a forgetting gate and an input gate are combined into an updating gate N, and a unit state U and a state H are combined into a composite state C; reset gate R t Graph state code data for controlling the previous time is inputted to R t Aggregated data; updating door N t For controlling the degree to which the picture state code data at the previous time is input to the current state.
2. The system of claim 1, wherein the spatial information reconstruction module is used to decompose and reconstruct each channel of the electroencephalogram data, and specifically comprises: decomposing all epilepsia electroencephalogram signals in high-frequency and low-frequency ranges by utilizing a wavelet packet decomposition technology, calculating energy value characteristics of each frequency band, and reconstructing the representation of each channel data unit; and calculating the correlation among multiple channels by using a Pearson correlation coefficient calculation method to generate graph data.
3. The system as claimed in claim 2, wherein 5 layers of wavelet packet decomposition are performed on all epileptic brain electrical signals in high and low frequency ranges to obtain 32 band wavelet coefficients, energy value characteristics of each band are calculated, and each channel data is reconstructed to 32 energy characteristic values.
4. The system of claim 1, wherein the graph encoder comprises an input layer, a first graph convolution layer, a second graph convolution layer, a third graph convolution layer, a full link layer, and a graph status code output layer.
5. The system according to claim 1, wherein the preprocessing of the acquired electroencephalogram data to be detected specifically comprises:
carrying out time division on the electroencephalogram record of each case in the data set; dividing epilepsia electroencephalogram signal data units; positive and negative sample data are balanced.
6. An electronic device comprising a memory, a processor and a computer program stored on the memory for execution by the processor to implement the functionality of the graph convolution network based multi-dimensional enhanced seizure prediction system according to any of claims 1-5.
7. A non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the functionality of the graph-volume network based multi-dimensional enhanced seizure prediction system according to any one of claims 1-5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110443244.7A CN113288050B (en) | 2021-04-23 | 2021-04-23 | Multidimensional enhanced epileptic seizure prediction system based on graph convolution network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110443244.7A CN113288050B (en) | 2021-04-23 | 2021-04-23 | Multidimensional enhanced epileptic seizure prediction system based on graph convolution network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113288050A CN113288050A (en) | 2021-08-24 |
| CN113288050B true CN113288050B (en) | 2022-10-04 |
Family
ID=77320270
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110443244.7A Active CN113288050B (en) | 2021-04-23 | 2021-04-23 | Multidimensional enhanced epileptic seizure prediction system based on graph convolution network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113288050B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113842118B (en) * | 2021-12-01 | 2022-03-25 | 浙江大学 | Epileptic seizure real-time detection monitoring system for epileptic video electroencephalogram examination |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018102815A1 (en) * | 2016-12-02 | 2018-06-07 | Thomas Jefferson University | Signal processing method for distinguishing and characterizing high-frequency oscillations |
| CN110399857A (en) * | 2019-08-01 | 2019-11-01 | 西安邮电大学 | A kind of brain electricity emotion identification method based on figure convolutional neural networks |
| CN110960191A (en) * | 2019-11-29 | 2020-04-07 | 杭州电子科技大学 | Epilepsia electroencephalogram signal classification method based on frequency spectrum energy diagram |
| CN112294337A (en) * | 2020-09-29 | 2021-02-02 | 山东师范大学 | Epilepsy detection system based on linear graph convolution neural network |
| CN112294338A (en) * | 2020-09-29 | 2021-02-02 | 山东师范大学 | Epilepsy detection system, equipment and medium based on graph attention network |
| CN112350899A (en) * | 2021-01-07 | 2021-02-09 | 南京信息工程大学 | Network flow prediction method based on graph convolution network fusion multi-feature input |
| CN112489740A (en) * | 2020-12-17 | 2021-03-12 | 北京惠及智医科技有限公司 | Medical record detection method, training method of related model, related equipment and device |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106388814B (en) * | 2016-10-11 | 2019-06-18 | 天津大学 | Epileptic EEG Signal recognition methods based on optimal core time-frequency distributions Visual Graph |
| EP3522171A1 (en) * | 2018-02-02 | 2019-08-07 | Université de Rennes 1 | Method, device and program for determining a disease score |
| CN109165387A (en) * | 2018-09-20 | 2019-01-08 | 南京信息工程大学 | A kind of Chinese comment sentiment analysis method based on GRU neural network |
| CN109447244A (en) * | 2018-10-11 | 2019-03-08 | 中山大学 | A kind of advertisement recommended method of combination gating cycle unit neural network |
| CN110693493B (en) * | 2019-10-12 | 2022-05-31 | 北京工业大学 | Epilepsia electroencephalogram prediction feature extraction method based on convolution and recurrent neural network combined time multiscale |
| CN111657935B (en) * | 2020-05-11 | 2021-10-01 | 浙江大学 | Epilepsia electroencephalogram recognition system based on hierarchical graph convolutional neural network, terminal and storage medium |
-
2021
- 2021-04-23 CN CN202110443244.7A patent/CN113288050B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018102815A1 (en) * | 2016-12-02 | 2018-06-07 | Thomas Jefferson University | Signal processing method for distinguishing and characterizing high-frequency oscillations |
| CN110399857A (en) * | 2019-08-01 | 2019-11-01 | 西安邮电大学 | A kind of brain electricity emotion identification method based on figure convolutional neural networks |
| CN110960191A (en) * | 2019-11-29 | 2020-04-07 | 杭州电子科技大学 | Epilepsia electroencephalogram signal classification method based on frequency spectrum energy diagram |
| CN112294337A (en) * | 2020-09-29 | 2021-02-02 | 山东师范大学 | Epilepsy detection system based on linear graph convolution neural network |
| CN112294338A (en) * | 2020-09-29 | 2021-02-02 | 山东师范大学 | Epilepsy detection system, equipment and medium based on graph attention network |
| CN112489740A (en) * | 2020-12-17 | 2021-03-12 | 北京惠及智医科技有限公司 | Medical record detection method, training method of related model, related equipment and device |
| CN112350899A (en) * | 2021-01-07 | 2021-02-09 | 南京信息工程大学 | Network flow prediction method based on graph convolution network fusion multi-feature input |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113288050A (en) | 2021-08-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110693493B (en) | Epilepsia electroencephalogram prediction feature extraction method based on convolution and recurrent neural network combined time multiscale | |
| CN110584649B (en) | Method and system for identifying abnormal electrocardiosignals in dynamic single-lead electrocardiogram | |
| Boon et al. | Paroxysmal atrial fibrillation prediction method with shorter HRV sequences | |
| CN113786204B (en) | Epileptic intracranial brain electrical signal early warning method based on deep convolution attention network | |
| CN111297349A (en) | Machine learning-based heart rhythm classification system | |
| Sriraam | Correlation dimension based lossless compression of EEG signals | |
| CN115470827B (en) | Self-supervision learning and twin network-based noise reduction method for resistant electrocardiosignals | |
| WO2020228420A1 (en) | Denoising autoencoder training method, electrocardiography signal denoising method, and apparatuses | |
| KR102141185B1 (en) | A system of detecting epileptic seizure waveform based on coefficient in multi-frequency bands from electroencephalogram signals, using feature extraction method with probabilistic model and machine learning | |
| Li et al. | Patient-specific seizure prediction from electroencephalogram signal via multi-channel feedback capsule network | |
| Raj et al. | Automated recognition of cardiac arrhythmias using sparse decomposition over composite dictionary | |
| CN113288050B (en) | Multidimensional enhanced epileptic seizure prediction system based on graph convolution network | |
| CN115886840B (en) | Epileptic prediction method based on domain-oriented multi-level deep convolution feature fusion network | |
| CN115470832A (en) | Electrocardiosignal data processing method based on block chain | |
| Sriraam et al. | An adaptive error modeling scheme for the lossless compression of EEG signals | |
| CN114648048A (en) | Electrocardiosignal noise reduction method based on variational self-coding and PixelCNN model | |
| Miller et al. | Physiological waveform imputation of missing data using convolutional autoencoders | |
| Chatterjee et al. | An input-delay neural-network-based approach for piecewise ECG signal compression | |
| CN116942172A (en) | Wavelet double-channel single-lead electrocardiograph denoising method based on coding and decoding structure | |
| Salafian et al. | Efficient epileptic seizure detection using CNN-aided factor graphs | |
| Sow et al. | Real-time prognosis of ICU physiological data streams | |
| Yang et al. | Superimposed semantic communication for iot-based real-time ecg monitoring | |
| CN115017960B (en) | Electroencephalogram signal classification method based on space-time combined MLP network and application | |
| Xiao et al. | Online MECG compression based on incremental tensor decomposition for wearable devices | |
| CN115844326A (en) | Self-adaptive sleep staging method and system based on key brain 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 |