CN115778330A - Automatic epileptic seizure detection system and device based on video electroencephalogram - Google Patents
Automatic epileptic seizure detection system and device based on video electroencephalogram Download PDFInfo
- Publication number
- CN115778330A CN115778330A CN202310071069.2A CN202310071069A CN115778330A CN 115778330 A CN115778330 A CN 115778330A CN 202310071069 A CN202310071069 A CN 202310071069A CN 115778330 A CN115778330 A CN 115778330A
- Authority
- CN
- China
- Prior art keywords
- electroencephalogram
- video
- epileptic
- module
- per
- 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
- 206010010904 Convulsion Diseases 0.000 title claims abstract description 109
- 206010015037 epilepsy Diseases 0.000 title claims abstract description 99
- 238000001514 detection method Methods 0.000 title claims abstract description 95
- 208000028329 epileptic seizure Diseases 0.000 title claims abstract description 72
- 238000000605 extraction Methods 0.000 claims abstract description 64
- 230000001037 epileptic effect Effects 0.000 claims abstract description 28
- 238000007781 pre-processing Methods 0.000 claims abstract description 25
- 238000000034 method Methods 0.000 claims abstract description 23
- 238000013528 artificial neural network Methods 0.000 claims description 39
- 210000004556 brain Anatomy 0.000 claims description 26
- 230000005611 electricity Effects 0.000 claims description 16
- 238000001228 spectrum Methods 0.000 claims description 16
- 230000015654 memory Effects 0.000 claims description 14
- 230000004913 activation Effects 0.000 claims description 13
- 238000005070 sampling Methods 0.000 claims description 8
- 230000006870 function Effects 0.000 claims description 7
- 238000013075 data extraction Methods 0.000 claims description 4
- 230000007274 generation of a signal involved in cell-cell signaling Effects 0.000 claims description 2
- 238000010606 normalization Methods 0.000 description 10
- 238000010586 diagram Methods 0.000 description 3
- 208000003443 Unconsciousness Diseases 0.000 description 2
- 230000036461 convulsion Effects 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 208000017667 Chronic Disease Diseases 0.000 description 1
- 208000012902 Nervous system disease Diseases 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000002490 cerebral effect Effects 0.000 description 1
- 238000003759 clinical diagnosis Methods 0.000 description 1
- 230000019771 cognition Effects 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 230000001787 epileptiform Effects 0.000 description 1
- 238000010304 firing Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000007787 long-term memory Effects 0.000 description 1
- 230000003340 mental effect Effects 0.000 description 1
- 230000004630 mental health Effects 0.000 description 1
- 210000000653 nervous system Anatomy 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 206010042772 syncope Diseases 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
Images
Landscapes
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention discloses an automatic epileptic seizure detection system and device based on video electroencephalogram, comprising: the system comprises an electroencephalogram acquisition module, an electroencephalogram data preprocessing module, an electroencephalogram data feature extraction module, an epileptic electroencephalogram detection module, a video acquisition module, a video feature extraction module and an epileptic seizure detection module. The epileptic seizure detection method provided by the invention simultaneously analyzes the electroencephalogram information and the video information, and can more accurately detect whether the patient has epileptic seizure; the method analyzes the electroencephalogram signals, analyzes the video data when the electroencephalogram signals detect epileptic electroencephalogram, and can greatly reduce the calculated amount on the premise of not reducing the detection precision.
Description
Technical Field
The invention relates to the technical field of medical information, in particular to an automatic epileptic seizure detection system and device based on video electroencephalogram.
Background
Epilepsy (Epilepsy) is a common disease of the nervous system, a chronic disease that results in transient cerebral dysfunction due to sudden abnormal firing of cerebral neurons. Epidemiological data show that the total prevalence rate of epilepsy in China is 7.0 per mill. Because of the difference of the origin brain area and the propagation path of the epileptic seizure, the epileptic seizure has multiple clinical manifestations, such as sudden consciousness loss, syncope, four limbs convulsion and the like, and causes disturbance in cognition and mental aspects, thereby seriously damaging the physical and mental health of patients and even endangering life.
At present, a traditional automatic epilepsy detection system only analyzes electroencephalogram signals and does not combine clinical manifestations of patients, electroencephalogram epilepsy-like discharge is generated when the patients attack epilepsy, on one hand, 0.3% -3% of people in normal people do not have epilepsy, but electroencephalogram epilepsy-like discharge exists, on the other hand, clinical manifestations of clinical diagnosis of epilepsy, such as consciousness loss and four-limb convulsion, need patient accompanying labels and doctors to make manual judgment through videos, and doctors need to combine clinical manifestations of the patients and whether electroencephalogram epilepsy discharge occurs to comprehensively judge whether the patients have epilepsy.
Therefore, a system and a device for automatically detecting epileptic seizure based on video electroencephalogram are provided to solve the technical problems.
Disclosure of Invention
The invention provides an automatic epileptic seizure detection system and device based on video electroencephalogram.
The technical scheme adopted by the invention is as follows:
an automatic epileptic seizure detection system based on video electroencephalogram, comprising:
the electroencephalogram acquisition module: the electroencephalogram data preprocessing module is used for acquiring electroencephalogram signals and inputting the electroencephalogram signals to the electroencephalogram data preprocessing module;
the electroencephalogram data preprocessing module: the electroencephalogram data extraction module is used for preprocessing the electroencephalogram signals to obtain the frequency domain characteristics of the electroencephalogram signals per second, and inputting the frequency domain characteristics of the electroencephalogram signals per second to the electroencephalogram data characteristic extraction module;
the electroencephalogram data feature extraction module: the electroencephalogram detection module is used for extracting the frequency domain characteristics of the electroencephalogram signals per second through the electroencephalogram characteristics to obtain corresponding electroencephalogram characteristics per second and probability that the electroencephalogram signals per second are epileptic-like, inputting the epileptic-like probability to the epileptic-like electroencephalogram detection module, and inputting the electroencephalogram characteristics per second to the epileptic seizure detection module;
the epileptic electroencephalogram detection module comprises: the electroencephalogram signal processing module is used for judging the probability of epileptic appearance of the electroencephalogram signal per second in a continuous time range, judging whether a patient generates epileptic appearance electroencephalogram or not and inputting the result of the epileptic appearance electroencephalogram of the patient into the video feature extraction module;
the video acquisition module: the system comprises a video feature extraction module, a hospital bed area video acquisition module and a hospital bed area video acquisition module, wherein the video feature extraction module is used for acquiring a hospital bed area video of a patient and inputting the hospital bed area video into the video feature extraction module;
the video feature extraction module: sampling the sickbed area video into a picture sequence according to a seizure electroencephalogram result of a patient by one frame per second, acquiring the video characteristics of each frame of picture from each frame of picture sequence by utilizing an OpenPose neural network, and inputting the video characteristics of each frame of picture into a seizure detection module;
a seizure detection module: the electroencephalogram features are input to the seizure detection neural network for obtaining the probability of seizure of the patient, and the probability of seizure of the patient is used for judging whether the patient has seizure or not.
Further, the preprocessing in the electroencephalogram data preprocessing module specifically comprises:
the system is used for fixedly dividing the electroencephalogram signals according to one second, performing Fourier transform on the electroencephalogram signals per second, and calculating the power spectrum of the multichannel electroencephalogram signals in the time range per second, wherein the frequency resolution is 1Hz;
the method is used for carrying out Fourier transform on the electroencephalogram signals per second and data of each channel of three continuous seconds one second before and after the electroencephalogram signals per second, calculating a power spectrum of the multichannel electroencephalogram signals in a continuous three-second range, and then normalizing the frequency resolution to 1Hz by adopting an average method;
the method is used for splicing the multichannel brain electricity signal power spectrum in the time range of each second and the multichannel brain electricity signal power spectrum in the continuous three-second range to serve as the frequency domain characteristic corresponding to the brain electricity signal of each second.
Further, the electroencephalogram data feature extraction module specifically comprises:
the characteristic extraction part is used for extracting the frequency domain characteristics of the electroencephalogram signals per second through the electroencephalogram characteristic extraction neural network to obtain corresponding electroencephalogram characteristics per second;
the channel for the electroencephalogram feature is used as a time axis and is input to a prediction part in the electroencephalogram feature extraction neural network, and the probability that the electroencephalogram signal is epileptic per second is obtained;
the electroencephalogram detection module is used for inputting epileptic probability to the epileptic electroencephalogram detection module and inputting electroencephalogram characteristics every second to the epileptic seizure detection module.
Further, the epileptiform electroencephalogram detection module specifically comprises: the electroencephalogram characteristic extraction module is used for judging the probability that the electroencephalogram signals are epileptic-like per second in a continuous time range, when the epileptic-like probability of more than 50% of the electroencephalogram signals in the continuous time range is greater than 80%, the patients are epileptic-like electroencephalogram, otherwise, the patients are not epileptic-like electroencephalogram, and the result of the epileptic-like electroencephalogram generation of the patients is input to the video characteristic extraction module.
Further, the video feature extraction module specifically includes:
the system is used for sampling the sickbed area video in the continuous time including the corresponding front and back preset time when the epileptic electroencephalogram occurs to a patient into a picture sequence according to one frame per second;
video characteristics of each frame of image of the front 10 layers of VGG-19, which are used for outputting each frame of image sequence by utilizing an OpenPose neural network;
and the video characteristics of each frame of picture are input into the epileptic seizure detection module.
Further, the seizure detection module specifically includes:
the electroencephalogram characteristic processing part is used for inputting the electroencephalogram characteristics in a continuous any time range to the epileptic seizure detection neural network to acquire time sequence electroencephalogram characteristics;
the video feature processing part is used for inputting the video features in continuous arbitrary time ranges into the epileptic seizure detection neural network to obtain time sequence video features;
the time sequence electroencephalogram characteristic and the time sequence video characteristic are spliced and input into a prediction part of a seizure detection neural network to obtain the probability of seizure of the patient;
the method is used for judging that when the probability of seizure of the patient exceeds 85%, the patient is considered to have seizure, and otherwise, the patient is considered to have no seizure.
Furthermore, the electroencephalogram acquisition module acquires electroencephalogram signals in real time according to a 10-20 standard lead system.
Furthermore, the activation layer of the brain electrical characteristic extraction neural network in the brain electrical data characteristic extraction module adopts a Relu activation function.
The invention also provides a device for automatically detecting the epileptic seizure based on the video electroencephalogram, which comprises a memory and one or more processors, wherein executable codes are stored in the memory, and when the one or more processors execute the executable codes, the device is used for realizing any one of the above-mentioned systems for automatically detecting the epileptic seizure based on the video electroencephalogram.
The invention also provides a computer readable storage medium, which stores a program, and when the program is executed by a processor, the system realizes the automatic epileptic seizure detection system based on the video electroencephalogram.
The invention has the beneficial effects that:
1. the epileptic seizure detection method provided by the invention simultaneously analyzes the electroencephalogram information and the video information, and can more accurately detect whether the patient has epileptic seizure.
2. The method analyzes the electroencephalogram signals, analyzes the video data when the electroencephalogram signals detect epileptic electroencephalogram, and can greatly reduce the calculated amount on the premise of not reducing the detection precision.
Drawings
FIG. 1 is a schematic structural diagram of an automatic epileptic seizure detection system based on video electroencephalogram;
fig. 2 is a schematic structural diagram of an automatic epileptic seizure detection device based on video electroencephalogram.
Detailed Description
The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the invention, its application, or uses. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1, an automatic epileptic seizure detection system based on video electroencephalogram includes:
the electroencephalogram acquisition module: the electroencephalogram data preprocessing module is used for acquiring electroencephalogram signals and inputting the electroencephalogram signals to the electroencephalogram data preprocessing module;
the electroencephalogram acquisition module acquires electroencephalogram signals in real time according to a 10-20 standard lead system.
The electroencephalogram data preprocessing module: the electroencephalogram data extraction module is used for preprocessing the electroencephalogram signals to obtain the frequency domain characteristics of the electroencephalogram signals per second, and inputting the frequency domain characteristics of the electroencephalogram signals per second to the electroencephalogram data characteristic extraction module;
the preprocessing in the electroencephalogram data preprocessing module specifically comprises the following steps:
the system is used for fixedly dividing the electroencephalogram signals according to one second, performing Fourier transform on the electroencephalogram signals per second, and calculating the power spectrum of the multichannel electroencephalogram signals in the time range per second, wherein the frequency resolution is 1Hz;
the method is used for carrying out Fourier transform on the electroencephalogram signals per second and data of each channel of three continuous seconds one second before and after the electroencephalogram signals per second, calculating a power spectrum of the multichannel electroencephalogram signals in a continuous three-second range, and then normalizing the frequency resolution to 1Hz by adopting an average method;
the method is used for splicing the multichannel brain electricity signal power spectrum in the time range of each second and the multichannel brain electricity signal power spectrum in the continuous three-second range to serve as the frequency domain characteristic corresponding to the brain electricity signal of each second.
The electroencephalogram data feature extraction module: the electroencephalogram detection module is used for extracting the frequency domain characteristics of the electroencephalogram signals per second through an electroencephalogram characteristic extraction neural network, acquiring corresponding electroencephalogram characteristics per second and probability that the electroencephalogram signals per second are epileptic-like, inputting the epileptic-like probability to the epileptic-like electroencephalogram detection module, and inputting the electroencephalogram characteristics per second to the epileptic seizure detection module;
the activation layer of the brain electrical characteristic extraction neural network in the brain electrical data characteristic extraction module adopts a Relu activation function;
the electroencephalogram data feature extraction module specifically comprises:
the characteristic extraction part is used for extracting the frequency domain characteristics of the electroencephalogram signals per second through electroencephalogram characteristics to obtain corresponding electroencephalogram characteristics per second;
the channel for the electroencephalogram feature is used as a time axis and is input to a prediction part in the electroencephalogram feature extraction neural network, and the probability that the electroencephalogram signal is epileptic per second is obtained;
the electroencephalogram detection module is used for inputting epileptic probability to the epileptic electroencephalogram detection module and inputting electroencephalogram characteristics every second to the epileptic seizure detection module.
The epilepsy-like electroencephalogram detection module comprises: the electroencephalogram signal processing module is used for judging the probability of epileptic appearance of the electroencephalogram signal per second in a continuous time range, judging whether a patient generates epileptic appearance electroencephalogram or not and inputting the result of the epileptic appearance electroencephalogram of the patient into the video feature extraction module;
the epilepsy electroencephalogram detection module specifically comprises: the electroencephalogram characteristic extraction module is used for judging the probability that the electroencephalogram signals are epileptic-like per second in a continuous time range, when the epileptic-like probability of more than 50% of the electroencephalogram signals in the continuous time range is greater than 80%, the patients are epileptic-like electroencephalogram, otherwise, the patients are not epileptic-like electroencephalogram, and the result of the epileptic-like electroencephalogram generation of the patients is input to the video characteristic extraction module.
The video acquisition module: the system is used for acquiring a video of a sickbed area where a patient is located and inputting the video of the sickbed area to a video feature extraction module.
The video feature extraction module: sampling the sickbed area video into a picture sequence according to a seizure electroencephalogram result of a patient by one frame per second, acquiring the video characteristics of each frame of picture from each frame of picture sequence by utilizing an OpenPose neural network, and inputting the video characteristics of each frame of picture into a seizure detection module;
the video feature extraction module specifically comprises:
the system is used for sampling the sickbed area video in the continuous time including the corresponding front and back preset time when the epileptic brain wave of the patient occurs into a picture sequence according to one frame per second;
video characteristics of each frame of image of the front 10 layers of VGG-19, which are used for outputting each frame of image sequence by utilizing an OpenPose neural network;
the video characteristics of each frame of picture are input into the epileptic seizure detection module.
A seizure detection module: the electroencephalogram characteristics are input to the epileptic seizure detection neural network every second to obtain the probability of epileptic seizure of the patient, and the probability of epileptic seizure of the patient is used for judging whether the patient has epileptic seizure;
the epileptic seizure detection module specifically comprises:
the electroencephalogram characteristic processing part is used for inputting the electroencephalogram characteristics in a continuous any time range to the epileptic seizure detection neural network to acquire time sequence electroencephalogram characteristics;
the video feature processing part is used for inputting the video features in continuous arbitrary time ranges into the epileptic seizure detection neural network to obtain time sequence video features;
the time sequence electroencephalogram characteristic and the time sequence video characteristic are spliced and input into a prediction part of an epileptic seizure detection neural network to obtain the probability of epileptic seizure of a patient;
the method is used for judging that when the probability of seizure of the patient exceeds 85%, the patient is considered to have seizure, and otherwise, the patient is considered to have no seizure.
Example (b): an automatic epileptic seizure detection system based on video electroencephalogram, comprising:
the electroencephalogram acquisition module: the electroencephalogram data preprocessing module is used for acquiring electroencephalogram signals and inputting the electroencephalogram signals to the electroencephalogram data preprocessing module;
the electroencephalogram acquisition module acquires electroencephalogram signals in real time according to a 10-20 standard lead system.
The electroencephalogram data preprocessing module: the electroencephalogram data extraction module is used for preprocessing the electroencephalogram signals to obtain the frequency domain characteristics of the electroencephalogram signals per second, and inputting the frequency domain characteristics of the electroencephalogram signals per second to the electroencephalogram data characteristic extraction module;
the preprocessing in the electroencephalogram data preprocessing module specifically comprises the following steps:
the system is used for fixedly dividing the electroencephalogram signals according to one second, performing Fourier transform on the electroencephalogram signals per second, and calculating the power spectrum of the multichannel electroencephalogram signals in the time range per second, wherein the frequency resolution is 1Hz;
the method is used for carrying out Fourier transform on the electroencephalogram signals per second and data of each channel of three continuous seconds one second before and after the electroencephalogram signals per second, calculating a power spectrum of the multichannel electroencephalogram signals in a continuous three-second range, and then normalizing the frequency resolution to 1Hz by adopting an average method;
the method is used for splicing the multichannel brain electricity signal power spectrum in the time range of each second and the multichannel brain electricity signal power spectrum in the continuous three-second range to serve as the frequency domain Feature _ Freq corresponding to the brain electricity signal of each second.
The electroencephalogram data feature extraction module: the electroencephalogram detection module is used for extracting the frequency domain Feature _ Freq of the electroencephalogram signal per second through an electroencephalogram Feature extraction neural network, acquiring the electroencephalogram Feature H _ EEG corresponding to the electroencephalogram Feature per second and the probability that the electroencephalogram signal per second is epileptic, inputting the probability of epileptic to the epileptic electroencephalogram detection module, and inputting the electroencephalogram Feature H _ EEG per second to the epileptic seizure detection module;
the structure of the feature extraction section includes: a convolution layer c1, a batch normalization layer bn1, an active layer a1; a convolution layer c2, a batch normalization layer bn2, an active layer a2; a convolution layer c3, a batch normalization layer bn3, an active layer a3; a convolution layer c4, a batch normalization layer bn4, an active layer a4; a convolution layer c5, a batch normalization layer bn5, an active layer a5; a convolution layer c6, a batch normalization layer bn6, an active layer a6; a convolution layer c7, a batch normalization layer bn7, an active layer a7; a convolution layer c8, a batch normalization layer bn8, an active layer a8; a convolution layer c9, a batch normalization layer bn9, an active layer a9; a convolution layer c10, a batch normalization layer bn10, an active layer a10;
the prediction part structure comprises: long and short term memory network layer LSTM; a full link layer L1; a full connection layer L2; a full connection layer L3; a developing layer Flaten1; a full link layer L4; a full interconnect layer L5; a full connection layer L6; and outputting the layer Out.
The activation layer of the brain electrical characteristic extraction neural network in the brain electrical data characteristic extraction module adopts a Relu activation function;
the electroencephalogram data feature extraction module specifically comprises:
the Feature extraction part is used for extracting the frequency domain Feature _ Freq of the electroencephalogram signal per second through an electroencephalogram Feature extraction neural network to obtain the corresponding electroencephalogram Feature H _ EEG per second;
the channel for the electroencephalogram characteristic H _ EEG is used as a time axis and is input to a prediction part in the electroencephalogram signal characteristic extraction neural network, and the probability that the electroencephalogram signal is epileptic per second is obtained;
inputting the long-short term memory network layer LSTM by taking the channel of the electroencephalogram characteristic H _ EEG of the second as a time axis, and then sequentially inputting the long-short term memory network layer LSTM and the full connection layer L1; a full link layer L2; a full connection layer L3; the unfolding layer, flaten1; a full link layer L4; a full interconnect layer L5; a full connection layer L6; the output layer Out is used for acquiring the probability that the second electroencephalogram signal is epileptic;
the electroencephalogram detection module is used for inputting the epilepsy probability to the epilepsy electroencephalogram detection module and inputting the electroencephalogram characteristics H _ EEG to the epilepsy attack detection module every second.
The epilepsy-like electroencephalogram detection module comprises: the electroencephalogram signal generation module is used for judging the probability of the electroencephalogram signal per second in the continuous time range, the probability of the epilepsy is used for judging whether a patient generates epilepsy-like electroencephalograms, and the result of the occurrence of the epilepsy-like electroencephalograms of the patient is input to the video feature extraction module;
the epilepsy electroencephalogram detection module specifically comprises: the electroencephalogram characteristic extraction module is used for judging the probability that the electroencephalogram signals are epileptic-like per second in a continuous time range, when the epileptic-like probability of more than 50% of the electroencephalogram signals in the continuous time range is greater than 80%, the patients are epileptic-like electroencephalogram, otherwise, the patients are not epileptic-like electroencephalogram, and the result of the epileptic-like electroencephalogram generation of the patients is input to the video characteristic extraction module.
The video acquisition module: the system is used for acquiring the video of the sickbed area where the patient is located and inputting the video of the sickbed area into the video feature extraction module.
The video feature extraction module: sampling the sickbed area video into a picture sequence according to a seizure electroencephalogram result of a patient, acquiring video characteristics H _ IMG of each frame of picture from each frame of picture sequence by using an OpenPose neural network, and inputting the video characteristics H _ IMG of each frame of picture to a seizure detection module;
the video feature extraction module specifically comprises:
the system is used for sampling the sickbed area video in the continuous time including the corresponding front and back preset time when the epileptic electroencephalogram occurs to a patient into a picture sequence according to one frame per second;
video characteristics H _ IMG of each frame of picture of the front 10 layers of VGG-19, which are used for outputting each frame of picture sequence by utilizing an OpenPose neural network;
the video characteristic H _ IMG of each frame of picture is input to the epileptic seizure detection module.
A seizure detection module: the probability of the patient seizure is obtained by inputting the video features of each frame of picture and the electroencephalogram features per second into the seizure detection neural network, and the probability of the patient seizure is used for judging whether the patient seizures.
The epileptic seizure detection module is composed of an epileptic seizure detection neural network, and the epileptic seizure detection neural network is composed of an electroencephalogram characteristic processing part, a video characteristic processing part and a prediction part; the structure of the electroencephalogram feature processing part comprises: a long-short term memory network layer LSTM1, a full connection layer L1 and an activation layer a1; the structure of the video feature processing section includes: a long-short term memory network layer LSTM2, a full connection layer L2 and an activation layer a2; the structure of the prediction part includes: a convolutional layer c1, a convolutional layer c2, an active layer a3; a full connection layer L3, a full connection layer L4, an activation layer a4, a full connection layer L5, an output layer Out. Wherein the activation layer adopts Relu activation function;
the epileptic seizure detection module specifically comprises:
the electroencephalogram characteristic processing part is used for inputting the electroencephalogram characteristic H _ EEG in any continuous time range to the epileptic seizure detection neural network to obtain a time sequence electroencephalogram characteristic F _ EEG;
a video feature processing part for inputting the video features H _ IMG in continuous arbitrary time range to the epileptic seizure detection neural network to obtain a time sequence video feature F _ IMG;
the time sequence electroencephalogram characteristic F _ EEG and the time sequence video characteristic F _ IMG are spliced and input into a prediction part of an epileptic seizure detection neural network to obtain the probability of epileptic seizure of a patient;
the method is used for judging that when the probability of the patient suffering from the epilepsia exceeds 85%, the patient is considered to suffer from the epilepsia, otherwise, the patient is considered to suffer from the epilepsia without seizures.
Corresponding to the embodiment of the automatic epileptic seizure detection system based on the video electroencephalogram, the invention also provides an embodiment of an automatic epileptic seizure detection device based on the video electroencephalogram.
Referring to fig. 2, an automatic epileptic seizure detection device based on video brain electricity according to an embodiment of the present invention includes a memory and one or more processors, where the memory stores executable codes, and when the one or more processors execute the executable codes, the one or more processors are configured to implement an automatic epileptic seizure detection system based on video brain electricity in the foregoing embodiment.
The embodiment of the automatic epileptic seizure detection device based on the video electroencephalogram can be applied to any equipment with data processing capability, such as computers and other equipment or devices. The device embodiments may be implemented by software, or by hardware, or by a combination of hardware and software. The software implementation is taken as an example, and as a logical device, the device is formed by reading corresponding computer program instructions in the nonvolatile memory into the memory for running through the processor of any device with data processing capability. In terms of hardware, as shown in fig. 2, a hardware structure diagram of any device with data processing capability where the device for automatically detecting epileptic seizure based on video electroencephalogram is located according to the present invention is shown in fig. 2, except for the processor, the memory, the network interface, and the nonvolatile memory shown in fig. 2, in the embodiment, any device with data processing capability where the device is located may generally include other hardware according to the actual function of the any device with data processing capability, which is not described again.
The implementation process of the functions and actions of each unit in the above device is specifically described in the implementation process of the corresponding step in the above method, and is not described herein again.
For the device embodiments, since they substantially correspond to the method embodiments, reference may be made to the partial description of the method embodiments for relevant points. The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the scheme of the invention. One of ordinary skill in the art can understand and implement without inventive effort.
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. An automatic epileptic seizure detection system based on video electroencephalogram is characterized by comprising:
an electroencephalogram acquisition module: the electroencephalogram data preprocessing module is used for acquiring electroencephalogram signals and inputting the electroencephalogram signals to the electroencephalogram data preprocessing module;
the electroencephalogram data preprocessing module: the electroencephalogram data extraction module is used for preprocessing the electroencephalogram signals to obtain the frequency domain characteristics of the electroencephalogram signals per second, and inputting the frequency domain characteristics of the electroencephalogram signals per second to the electroencephalogram data characteristic extraction module;
the electroencephalogram data feature extraction module: the electroencephalogram detection module is used for extracting the frequency domain characteristics of the electroencephalogram signals per second through an electroencephalogram characteristic extraction neural network, acquiring corresponding electroencephalogram characteristics per second and probability that the electroencephalogram signals per second are epileptic-like, inputting the epileptic-like probability to the epileptic-like electroencephalogram detection module, and inputting the electroencephalogram characteristics per second to the epileptic seizure detection module;
the epileptic electroencephalogram detection module comprises: the electroencephalogram signal generation module is used for judging the probability of the electroencephalogram signal per second in the continuous time range, the probability of the epilepsy is used for judging whether a patient generates epilepsy-like electroencephalograms, and the result of the occurrence of the epilepsy-like electroencephalograms of the patient is input to the video feature extraction module;
the video acquisition module: the system comprises a video feature extraction module, a data acquisition module and a data processing module, wherein the video feature extraction module is used for acquiring a video of a sickbed area where a patient is located and inputting the video of the sickbed area to the video feature extraction module;
the video feature extraction module: sampling the sickbed area video into a picture sequence according to a seizure electroencephalogram result of a patient by one frame per second, acquiring the video characteristics of each frame of picture from each frame of picture sequence by utilizing an OpenPose neural network, and inputting the video characteristics of each frame of picture into a seizure detection module;
an epileptic seizure detection module: the electroencephalogram features are input to the seizure detection neural network for obtaining the probability of seizure of the patient, and the probability of seizure of the patient is used for judging whether the patient has seizure or not.
2. The system for automatically detecting epileptic seizure based on video electroencephalogram (EEG) of claim 1, wherein the preprocessing in the EEG data preprocessing module specifically comprises:
the system is used for fixedly dividing the electroencephalogram signals according to one second, performing Fourier transform on the electroencephalogram signals per second, and calculating the power spectrum of the multichannel electroencephalogram signals in the time range per second, wherein the frequency resolution is 1Hz;
the method is used for carrying out Fourier transform on the electroencephalogram signals per second and data of each channel of three continuous seconds one second before and after the electroencephalogram signals per second, calculating a power spectrum of the multichannel electroencephalogram signals in a continuous three-second range, and then normalizing the frequency resolution to 1Hz by adopting an average method;
the method is used for splicing the multichannel brain electricity signal power spectrum in the time range of each second and the multichannel brain electricity signal power spectrum in the continuous three-second range to serve as the frequency domain characteristics corresponding to the brain electricity signals per second.
3. The system for automatically detecting epileptic seizure based on video electroencephalogram as claimed in claim 1, wherein said electroencephalogram data feature extraction module specifically comprises:
the characteristic extraction part is used for extracting the frequency domain characteristics of the electroencephalogram signals per second through electroencephalogram characteristics to obtain corresponding electroencephalogram characteristics per second;
the channel for the electroencephalogram feature is used as a time axis and is input to a prediction part in the electroencephalogram feature extraction neural network, and the probability that the electroencephalogram signal is epileptic per second is obtained;
the electroencephalogram detection module is used for inputting epilepsy-like probability into the epilepsy-like electroencephalogram detection module and inputting electroencephalogram characteristics every second into the epilepsy attack detection module.
4. The system for automatically detecting epileptic seizure based on video electroencephalogram (VCEEG) as claimed in claim 1, wherein said epileptic electroencephalogram detection module specifically comprises: the electroencephalogram characteristic extraction module is used for judging the probability that the electroencephalogram signals are epileptic-like per second in a continuous time range, when the epileptic-like probability of more than 50% of the electroencephalogram signals in the continuous time range is greater than 80%, the patients are epileptic-like electroencephalogram, otherwise, the patients are not epileptic-like electroencephalogram, and the result of the epileptic-like electroencephalogram generation of the patients is input to the video characteristic extraction module.
5. The system for automatically detecting epileptic seizure based on video electroencephalogram as claimed in claim 1, wherein said video feature extraction module specifically comprises:
the system is used for sampling the sickbed area video in the continuous time including the corresponding front and back preset time when the epileptic electroencephalogram occurs to a patient into a picture sequence according to one frame per second;
video characteristics of each frame of image of the front 10 layers of VGG-19, which are used for outputting each frame of image sequence by utilizing an OpenPose neural network;
the video characteristics of each frame of picture are input into the epileptic seizure detection module.
6. The system for automatically detecting the epileptic seizure based on the video electroencephalogram, as claimed in claim 1, wherein said epileptic seizure detection module specifically comprises:
the electroencephalogram characteristic processing part is used for inputting the electroencephalogram characteristics in a continuous any time range to the epileptic seizure detection neural network to acquire time sequence electroencephalogram characteristics;
the video feature processing part is used for inputting the video features in continuous arbitrary time ranges into the epileptic seizure detection neural network to obtain time sequence video features;
the time sequence electroencephalogram characteristic and the time sequence video characteristic are spliced and input into a prediction part of an epileptic seizure detection neural network to obtain the probability of epileptic seizure of a patient;
the method is used for judging that when the probability of the patient suffering from the epilepsia exceeds 85%, the patient is considered to suffer from the epilepsia, otherwise, the patient is considered to suffer from the epilepsia without seizures.
7. The video-electroencephalogram-based automatic epileptic seizure detection system of claim 1, wherein the electroencephalogram acquisition module acquires electroencephalogram signals in real time according to a 10-20 standard lead system.
8. The system for automatically detecting epileptic seizure based on video electroencephalogram (EEG) of claim 1, wherein an activation layer of a brain electrical feature extraction neural network in the EEG data feature extraction module adopts a Relu activation function.
9. An automatic epileptic seizure detection device based on video brain electricity is characterized by comprising a memory and one or more processors, wherein the memory stores executable codes, and the one or more processors are used for realizing the automatic epileptic seizure detection system based on video brain electricity, wherein the executable codes are executed by the one or more processors.
10. A computer-readable storage medium, having a program stored thereon, which when executed by a processor, implements a video brain electrical based seizure automatic detection system as claimed in any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310071069.2A CN115778330A (en) | 2023-02-07 | 2023-02-07 | Automatic epileptic seizure detection system and device based on video electroencephalogram |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310071069.2A CN115778330A (en) | 2023-02-07 | 2023-02-07 | Automatic epileptic seizure detection system and device based on video electroencephalogram |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115778330A true CN115778330A (en) | 2023-03-14 |
Family
ID=85430153
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310071069.2A Pending CN115778330A (en) | 2023-02-07 | 2023-02-07 | Automatic epileptic seizure detection system and device based on video electroencephalogram |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115778330A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117033988A (en) * | 2023-09-27 | 2023-11-10 | 之江实验室 | Epileptiform spike processing method and device based on nerve electric signal |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20170354341A1 (en) * | 2016-06-10 | 2017-12-14 | Mocxa, LLC | Procedure and a portable apparatus for diagnosis of seizures |
| CN108647645A (en) * | 2018-05-11 | 2018-10-12 | 广州飞宇智能科技有限公司 | A kind of multi-modal Diagnosis of Epilepsy system and method based on video analysis |
| WO2019158824A1 (en) * | 2018-02-16 | 2019-08-22 | Neuro Event Labs Oy | Method for detecting and classifying a motor seizure |
| CN110236536A (en) * | 2019-06-04 | 2019-09-17 | 电子科技大学 | A kind of brain electricity high-frequency oscillation signal detection system based on convolutional neural networks |
| CN110991289A (en) * | 2019-11-25 | 2020-04-10 | 达闼科技成都有限公司 | Abnormal event monitoring method and device, electronic equipment and storage medium |
| CN111598003A (en) * | 2020-05-18 | 2020-08-28 | 温州大学 | Time-frequency image classification method for electroencephalogram signals of epileptics |
| CN113095428A (en) * | 2021-04-23 | 2021-07-09 | 西安交通大学 | Video emotion classification method and system fusing electroencephalogram and stimulus information |
| CN113723206A (en) * | 2021-08-04 | 2021-11-30 | 三峡大学 | Brain wave identification method based on quantum neural network algorithm |
| CN113842118A (en) * | 2021-12-01 | 2021-12-28 | 浙江大学 | Epileptic seizure real-time detection monitoring system for epileptic video electroencephalogram examination |
| CN114093501A (en) * | 2021-10-19 | 2022-02-25 | 杭州电子科技大学 | Intelligent auxiliary analysis method for children's motor epilepsy based on synchronous video and electroencephalogram |
| CN114145755A (en) * | 2021-12-21 | 2022-03-08 | 上海理工大学 | Household epileptic seizure interactive intelligent monitoring system and method |
| US20220160291A1 (en) * | 2020-11-23 | 2022-05-26 | Mocxa Health Private Limited | System for recording of seizures |
-
2023
- 2023-02-07 CN CN202310071069.2A patent/CN115778330A/en active Pending
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20170354341A1 (en) * | 2016-06-10 | 2017-12-14 | Mocxa, LLC | Procedure and a portable apparatus for diagnosis of seizures |
| WO2019158824A1 (en) * | 2018-02-16 | 2019-08-22 | Neuro Event Labs Oy | Method for detecting and classifying a motor seizure |
| CN108647645A (en) * | 2018-05-11 | 2018-10-12 | 广州飞宇智能科技有限公司 | A kind of multi-modal Diagnosis of Epilepsy system and method based on video analysis |
| CN110236536A (en) * | 2019-06-04 | 2019-09-17 | 电子科技大学 | A kind of brain electricity high-frequency oscillation signal detection system based on convolutional neural networks |
| CN110991289A (en) * | 2019-11-25 | 2020-04-10 | 达闼科技成都有限公司 | Abnormal event monitoring method and device, electronic equipment and storage medium |
| CN111598003A (en) * | 2020-05-18 | 2020-08-28 | 温州大学 | Time-frequency image classification method for electroencephalogram signals of epileptics |
| US20220160291A1 (en) * | 2020-11-23 | 2022-05-26 | Mocxa Health Private Limited | System for recording of seizures |
| CN113095428A (en) * | 2021-04-23 | 2021-07-09 | 西安交通大学 | Video emotion classification method and system fusing electroencephalogram and stimulus information |
| CN113723206A (en) * | 2021-08-04 | 2021-11-30 | 三峡大学 | Brain wave identification method based on quantum neural network algorithm |
| CN114093501A (en) * | 2021-10-19 | 2022-02-25 | 杭州电子科技大学 | Intelligent auxiliary analysis method for children's motor epilepsy based on synchronous video and electroencephalogram |
| CN113842118A (en) * | 2021-12-01 | 2021-12-28 | 浙江大学 | Epileptic seizure real-time detection monitoring system for epileptic video electroencephalogram examination |
| CN114145755A (en) * | 2021-12-21 | 2022-03-08 | 上海理工大学 | Household epileptic seizure interactive intelligent monitoring system and method |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117033988A (en) * | 2023-09-27 | 2023-11-10 | 之江实验室 | Epileptiform spike processing method and device based on nerve electric signal |
| CN117033988B (en) * | 2023-09-27 | 2024-03-12 | 之江实验室 | Epileptiform spike processing method and device based on nerve electric signal |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US8700141B2 (en) | Method and apparatus for automatic evoked potentials assessment | |
| Minasyan et al. | Patient-specific early seizure detection from scalp electroencephalogram | |
| CN115778330A (en) | Automatic epileptic seizure detection system and device based on video electroencephalogram | |
| CN114615924B (en) | System and method for seizure detection based on electroencephalogram (EEG) nonlinear variation | |
| DE102008002898A1 (en) | Detection of anomalies in the measurement of anesthesia | |
| Stefánsson et al. | Auditory event-related potentials, auditory digit span, and clinical symptoms in chronic schizophrenic men on neuroleptic medication | |
| Karpov et al. | Noise amplification precedes extreme epileptic events on human EEG | |
| JP2023099043A (en) | Systems and methods for seizure detection based on changes in electroencephalogram (eeg) non-linearities | |
| CN115804572A (en) | Automatic monitoring system and device for epileptic seizure | |
| Sun et al. | Brain monitoring of sedation in the intensive care unit using a recurrent neural network | |
| Husain et al. | Differentiation of epileptic and psychogenic nonepileptic seizures using single-channel surface electromyography | |
| CN116035593B (en) | Electrocerebral noise reduction method based on generation countermeasure type parallel neural network | |
| CN112802606A (en) | Data screening model establishing method, data screening device, data screening equipment and data screening medium | |
| CN115844336A (en) | Automatic real-time monitoring system and device for epileptic seizure | |
| Najafi et al. | Brain waves characteristics in individuals with obsessive-compulsive disorder: A preliminary study | |
| Das et al. | Discrimination of scalp EEG signals in wavelet transform domain and channel selection for the patient-invariant seizure detection | |
| Bernabei et al. | A full-stack application for detecting seizures and reducing data during continuous electroencephalogram monitoring | |
| Abdi-Sargezeh et al. | Mapping scalp to intracranial EEG using generative adversarial networks for automatically detecting interictal epileptiform discharges | |
| Suárez et al. | Wavelet transform and cross-correlation as tools for seizure prediction | |
| Sharanreddy et al. | An improved approximate entropy based epilepsy seizure detection using multi-wavelet and artificial neural networks | |
| Tolmacheva et al. | Estimation of inter-channel phase synchronization of EEG signals in the ridges of their wavelet spectrograms in patients with traumatic brain injury | |
| Ohya et al. | Eyelid myoclonia with absences occurring during the clinical course of cryptogenic myoclonic epilepsy of early childhood | |
| Bhuiyan et al. | A subband correlation-based method for the automatic detection of epilepsy and seizure in the dual tree complex wavelet transform domain | |
| KR101659941B1 (en) | Telemedicine system | |
| Flasbeck et al. | Heartbeat evoked potentials and autonomic arousal during dissociative seizures: insights from electrophysiology and neuroimaging |
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: 20230314 |
|
| RJ01 | Rejection of invention patent application after publication |