Epileptic seizure detection method and device based on feature fusion and network equipment

Abstract

The embodiment of the invention provides an epileptic seizure detection method and device based on feature fusion and network equipment, and the method comprises the steps: obtaining channel data of a channel of each electroencephalogram signal, and carrying out the feature extraction of the channel data according to a convolutional neural network, and obtaining a first feature of the channel; performing correlation analysis between the channels according to the first feature to determine a second feature between the channels; fusing the first feature and the second feature to obtain a fused feature; and inputting the fusion features into a classification model containing a multi-head attention mechanism, and determining an epilepsy analysis result. According to the scheme, the intra-channel features and the inter-channel features are fused, the correlation between the channel data and the channel data is comprehensively considered, and the epilepsy seizure can be identified more accurately.

Description

Technical Field

[0001] The present application relates to the field of computer information technology, and in particular to a method, device and network equipment for detecting epileptic seizures based on feature fusion. Background Art

[0002] Epilepsy is a serious neurological disease. Detecting epileptic seizures from long-term EEG monitoring is a time-consuming and laborious task. In addition, different types of epilepsy usually originate from different brain regions, and epileptic seizures show complexity and diversity. These factors make accurate epilepsy detection a critical but challenging task in clinical practice. Electroencephalogram (EEG) signals, as an important tool for recording brain activity, are widely used in the detection of epileptic seizures.

[0003] Existing solutions only consider the intra-channel features of EEG, but not the inter-channel features of EEG, resulting in low accuracy of recognition results. Summary of the invention

[0004] The embodiments of the present application provide a method, device and network device for detecting epileptic seizures based on feature fusion, which can analyze intra-channel features and inter-channel features of EEG, and analyze them after fusion to more accurately identify epileptic seizures. The technical solution is as follows:

[0005] In a first aspect, the present application provides an epileptic seizure detection method based on feature fusion, the method comprising: obtaining channel data of each channel of each EEG signal, and performing feature extraction on the channel data based on a convolutional neural network to obtain a first feature of the channel; performing correlation analysis between channels based on the first feature to determine a second feature between channels; fusing the first feature with the second feature to obtain a fused feature; and inputting the fused feature into a classification model including a multi-head attention mechanism to determine the epilepsy analysis result.

[0006] Furthermore, the method of performing correlation analysis between channels based on the first feature to determine a second feature between channels includes: analyzing the Pearson correlation coefficient between channels based on the first feature to perform correlation analysis between channels and obtain a feature matrix between channels; incorporating adaptively adjustable weight information into the feature matrix, and learning the weight information based on the first feature to determine the second feature between channels, the weight information corresponding to the correlation relationship between channels.

[0007] Furthermore, the method of analyzing the Pearson correlation coefficient between channels based on the first feature to perform correlation analysis between channels and obtain a feature matrix between channels includes: obtaining first data of the first feature of the first channel and second data of the second feature of the second channel; obtaining a first average value of the first channel and a second average value of the second channel; analyzing the Pearson correlation coefficient between channels based on the first data, the second data, the first average value and the second average value to form a correlation matrix between channels; processing the values ​​in the correlation matrix according to a preset threshold to obtain a feature matrix between channels; after the feature matrix is ​​processed by the preset threshold, the values ​​in the correlation matrix that are greater than or equal to the preset threshold are 1, and the values ​​that are less than the preset threshold are 0.

[0008] Furthermore, the method incorporates adaptively adjustable weight information into the feature matrix, and learns the weight information based on the first feature to determine the second feature between channels, including: generating a diagonal matrix of learnable parameters, the parameters in the diagonal matrix are related to the weight information of the correlation relationship between channels; integrating the diagonal matrix into the feature matrix to obtain a self-loop perception matrix, and learning based on the first feature; calculating the degree matrix of the self-loop perception matrix, and normalizing the self-loop perception matrix and the degree matrix to obtain a symmetric normalized Laplace matrix as the second feature between channels.

[0009] Furthermore, the fusing of the first feature and the second feature to obtain the fused feature includes: performing hierarchical linear transformation processing on the first feature and the second feature to obtain an intermediate representation; and activating the intermediate representation using a hyperbolic tangent function to obtain the fused feature.

[0010] Furthermore, the performing hierarchical linear transformation processing on the first feature and the second feature to obtain an intermediate representation includes: performing hierarchical linear transformation processing on the first feature and the second feature through a first fully connected layer to obtain an intermediate matrix; and processing the intermediate matrix through a second fully connected layer to obtain an intermediate representation.

[0011] Furthermore, the classification model including the multi-head attention mechanism is used to: obtain the time information of the fused feature as the time feature; determine the relevant features of the fused feature based on the time feature of the fused feature; determine the attention weights corresponding to the relevant features before and after the fused feature, and integrate the relevant features into the fused feature according to the attention weights to determine the pooled feature; use the activation function to process the pooled feature to determine the epilepsy analysis result.

[0012] Furthermore, the method of obtaining channel data of the channels of each EEG signal includes: obtaining channel-related information of the channels of each EEG signal, and performing bandpass filtering within a preset frequency band; the channel-related information includes electroencephalogram data and time information, and the channel-related information includes first information corresponding to epileptic seizures and second information corresponding to non-epileptic seizures; dividing the filtered channel-related information according to the time information and the preset time interval to obtain information segments, and performing normalization processing to obtain channel data.

[0013] In the second aspect, the present application provides an epileptic seizure detection device based on feature fusion, the device comprising: a first feature acquisition module, used to acquire channel data of channels of each EEG signal, and perform feature extraction on the channel data based on a convolutional neural network to obtain a first feature of the channel; a second feature acquisition module, used to perform correlation analysis between channels based on the first feature to determine a second feature between channels; a fusion feature acquisition module, used to fuse the first feature with the second feature to obtain a fusion feature; a classification model analysis module, used to input the fusion feature into a classification model including a multi-head attention mechanism to determine the epilepsy analysis result.

[0014] In a third aspect, the present application provides a network device comprising: a memory, a transceiver, and a processor; wherein the memory is used to store a computer program; the transceiver is used to send and receive data under the control of the processor; and the processor is used to read the computer program in the memory and execute the method described in the first aspect.

[0015] In a fourth aspect, the present application provides a storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the method described in the first aspect.

[0016] The beneficial effects of the technical solution provided by this application are:

[0017] The scheme of the present application can be applied to the scenario of disease identification based on EEG data. For example, it can be used to identify whether the patient has epilepsy based on EEG data so as to timely warn and other treatments. This scheme can obtain the channel data of the channel of the EEG signal to extract channel features, analyze the correlation between channels, extract inter-channel features, and then fuse the channel features and inter-channel features to adopt a classification model to analyze the fused features to determine whether an epileptic seizure occurs. This scheme can more accurately identify the onset of epilepsy by fusing the intra-channel features and inter-channel features, comprehensively considering the channel data itself and the correlation between the channel data. Specifically, this scheme can obtain the channel data of the channel of each EEG signal, and extract features from the channel data based on the convolutional neural network to obtain the first feature of the channel; perform correlation analysis between channels based on the first feature to determine the second feature between channels; fuse the first feature with the second feature to obtain a fused feature; input the fused feature into a classification model containing a multi-head attention mechanism to fuse the relevant features before and after the feature and analyze it to determine the epilepsy analysis result. This solution can pre-train convolutional neural networks and classification models including multi-head attention mechanisms for feature extraction and feature analysis. This solution considers the spatial feature correlation between channels and the temporal correlation of the features before and after the channels, which can perform epilepsy analysis more accurately. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in describing the embodiments of the present application are briefly introduced below.

[0019] Figure 1 is a flowchart of an epileptic seizure detection method based on feature fusion according to an embodiment of the present application;

[0020] Figure 2 is a flowchart of an epileptic seizure detection method based on feature fusion according to another embodiment of the present application;

[0021] Figure 3 is a structural schematic diagram of an epileptic seizure detection device based on feature fusion according to an embodiment of the present application;

[0022] Figure 4 It is a structural block diagram of a network device according to an embodiment of the present application;

[0023] Figure 5 This is a structural block diagram of a user device according to an embodiment of the present application. DETAILED DESCRIPTION

[0024] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout identify the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application, and cannot be interpreted as limiting the present application.

[0025] It will be understood by those skilled in the art that, unless specifically stated, the singular forms "one", "said", and "the" used herein may also include plural forms, and "multiple" refers to two or more, and other quantifiers are similar. It should be further understood that the term "comprising" used in the specification of the present application refers to the presence of the features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or their groups. It should be understood that when we say that an element is "connected" or "coupled" to another element, it can be directly connected or coupled to other elements, or there may be intermediate elements. In addition, the "connection" or "coupling" used here may include wireless connection or wireless coupling. The term "and / or" used here describes the association relationship of associated objects, indicating that there may be three relationships, for example, A and / or B, which can represent: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the objects associated before and after are in an "or" relationship.

[0026] This application proposes a method for epileptic seizure detection. This application first uses the Pearson correlation coefficient combined with learnable self-loop weights to dynamically adjust channel weights during training to extract inter-channel features, and then combines these features with intra-channel features to obtain a more comprehensive view of brain activity. Then, a self-attention mechanism-based graph convolutional network (LATGCN) is proposed to identify inter-channel and intra-channel features of EEG channels to effectively capture spatial and temporal information for epileptic seizure detection, so as to assist doctors in achieving seizure detection during the clinical diagnosis of epilepsy.

[0027] Specifically, in order to improve the accuracy of epileptic seizure detection, the present application proposes an epileptic seizure detection method, the flow chart of which is as follows: Figure 1 The whole process is divided into four stages: data preprocessing, inter-channel feature extraction, inter-channel and intra-channel feature fusion, and epilepsy detection.

[0028] Data preprocessing: Electroencephalogram (EEG) data is bandpass filtered between 4-30Hz, which covers important frequency components closely related to brain activity, such as theta waves (4-7Hz), alpha waves (8-12Hz), and beta waves (13-30Hz). Then, the application splits the EEG recordings of epileptic and non-epileptic seizures into 2-second intervals to reduce the computational burden of the model. Finally, the application normalizes these EEG segments to eliminate amplitude differences between different recordings.

[0029] Inter-channel and intra-channel feature fusion: In EEG analysis, inter-channel features highlight the functional connections and interactions between different brain regions, revealing how these brain regions are coordinated in different cognitive processes. Analysis of inter-channel features helps detect epileptic seizures from long-term EEG signal monitoring. The Pearson correlation coefficient is a commonly used functional connectivity measure that quantifies the strength and direction of the linear relationship between two signals, revealing the functional coupling between different brain regions. Therefore, this application extracts inter-channel features by calculating the Pearson correlation coefficient of different EEG signal channels.

[0030] First, the preprocessed EEG signal is input into the convolutional neural network to extract features from each channel. Subsequently, all features of each channel are collected and the Pearson correlation coefficient between each channel is calculated. The convolutional neural network here consists of three convolutional layers, three pooling layers, and a fully connected layer to extract channel features. In order for the network to accurately extract channel features, the network is pre-trained until the training loss converges and the verification accuracy exceeds 90%. The Pearson correlation coefficient is calculated as follows:

[0031]

[0032] In formula 1, x i and i are the data values ​​corresponding to the two channels, and is the average value of the two channel data. The value range of the Pearson correlation coefficient is [-1, 1]. A value close to 1 indicates a strong positive correlation between the two variables, a value close to -1 indicates a strong negative correlation, and a value of 0 indicates no linear correlation between the variables. In order to extract as many features of brain regions with strong functional connections as possible and filter out interfering brain regions with weak connections, this application sets a threshold for the correlation matrix. Values ​​above the threshold are set to 1, indicating that there is a direct connection between channels, and values ​​below the threshold are set to 0, indicating that there is no connection between channels. Finally, the feature matrix between channels can be obtained.

[0033] Secondly, based on the extracted features, the inter-channel feature matrix is ​​calculated and a learnable self-loop Laplacian matrix, called the self-loop-aware Laplacian matrix, is introduced to enable the network to adaptively adjust these weights during training. For a given N×N inter-channel feature matrix A, let it be compared with the learnable parameter θ s =[θ s1 ,θ s2 ..., θ sN ] is added to obtain the Laplace matrix of self-loop perception And calculate the degree matrix The calculation formula is as follows:

[0034]

[0035] The symmetric normalized Laplace matrix L can be further calculated by the results calculated by the above formula 2 and formula 3. The calculation formula is as follows:

[0036]

[0037] Where I represents the identity matrix.

[0038] The input features are processed by the symmetric normalized Laplacian matrix, and the information between channels is integrated through matrix multiplication to form an intermediate representation. Then it is linearly transformed through two fully connected layers and activated by the hyperbolic tangent function. This hierarchical linear transformation can further mine and integrate the node features in the graph, improving the convergence speed of the model. The node feature update formula of each layer of convolutional neural network (GCN) is as follows:

[0039]

[0040] Where H i represents the feature matrix of the i-th layer node of GCN, and is the weight matrix of the two fully connected layers of the i-th layer, and H represents the intermediate features after the first linear transformation.

[0041] Finally, through Laplacian matrix multiplication and hierarchical linear transformation, the fused features within and between channels are obtained so that a holistic view of brain activity covering both local dynamics and global network properties can be obtained.

[0042] Epileptic seizure detection based on self-loop-aware Transformer-graph convolutional network: r-In the LATGCN, the GCN layer is first used to process the relationship between nodes, and the fused features of intra-channel and inter-channel are used as the input of the network. Secondly, the Transformer encoder after the GCN layer can obtain temporal information. The Transformer encoder contains an attention mechanism that enables each input token to interact with other tokens, thereby effectively capturing global temporal dependencies. The multi-head attention mechanism allows the model to simultaneously focus on different temporal features of the input through queries, keys, and values. The position feedforward network processes each temporal position independently, further improving the expressiveness of temporal features. Residual connections and layer normalization enhance the stability of the model, while position encoding provides the model with position information of the time series. These components work together to enable the Transformer encoder to effectively model complex temporal dependencies.

[0043] The features of all nodes are aggregated through a pooling layer with an attention mechanism, and the pooled feature vector p is activated by the Sigmoid function after passing through the fully connected layer to obtain the final prediction result. The calculation formula of the attention pooling layer is as follows:

[0044] p = ∑ softmax(α(x i ))·x i Formula 7

[0045] In the formula, α(x i ) is the attention mechanism for each node feature x i The attention weight assigned. The attention mechanism can capture features that are beneficial to EEG signal recognition and filter out irrelevant features, which can improve the recognition accuracy of EEG signals and achieve accurate detection of epileptic seizures.

[0046] Specifically, the present application provides a method for detecting epileptic seizures based on feature fusion, such as Figure 2 As shown, the method includes:

[0047] Step 202: Acquire channel data of each channel of the EEG signal, and perform feature extraction on the channel data based on a convolutional neural network to obtain a first feature of the channel. The first feature may also be referred to as a channel feature or an intra-channel feature.

[0048] Step 204: Perform correlation analysis between channels based on the first feature to determine a second feature between channels. The second feature may also be called an inter-channel feature.

[0049] Step 206: Fuse the first feature with the second feature to obtain a fused feature.

[0050] Step 208: Input the fused features into a classification model including a multi-head attention mechanism to determine the epilepsy analysis result.

[0051] It should be noted that part of the implementation process of this solution has been described in the above embodiments, and the implementation process of the above embodiments may be referred to, which will not be repeated here.

[0052] The scheme of the present application can be applied to the scenario of disease identification based on EEG data. For example, it can be used to identify whether the patient has epilepsy based on EEG data so as to timely warn and other treatments. This scheme can obtain the channel data of the channel of the EEG signal to extract channel features, analyze the correlation between channels, extract inter-channel features, and then fuse the channel features and inter-channel features to adopt a classification model to analyze the fused features to determine whether an epileptic seizure occurs. This scheme can more accurately identify the onset of epilepsy by fusing the intra-channel features and inter-channel features, comprehensively considering the channel data itself and the correlation between the channel data. Specifically, this scheme can obtain the channel data of the channel of each EEG signal, and extract features from the channel data based on the convolutional neural network to obtain the first feature of the channel; perform correlation analysis between channels based on the first feature to determine the second feature between channels; fuse the first feature with the second feature to obtain a fused feature; input the fused feature into a classification model containing a multi-head attention mechanism to fuse the relevant features before and after the feature and analyze it to determine the epilepsy analysis result. This solution can pre-train convolutional neural networks and classification models including multi-head attention mechanisms for feature extraction and feature analysis.

[0053] This solution can use a convolutional neural network to extract channel features corresponding to channel data, analyze the Pearson correlation coefficient between channels, and incorporate learnable weight information to analyze the correlation between channels as channel-to-channel features. Specifically, as an optional embodiment, the correlation analysis between channels based on the first feature to determine the second feature between channels includes: analyzing the Pearson correlation coefficient between channels based on the first feature to perform correlation analysis between channels to obtain a feature matrix between channels; incorporating adaptively adjustable weight information into the feature matrix, and learning the weight information based on the first feature to determine the second feature between channels, and the weight information corresponds to the correlation relationship between channels.

[0054] This solution can determine the correlation matrix by analyzing the channel data and the channel average value, and convert the values ​​in the correlation matrix that are greater than or equal to the preset threshold value to 1, and the values ​​that are less than the preset threshold value to 0 according to the preset threshold value, to form a feature matrix. Specifically, as an optional embodiment, the Pearson correlation coefficient between channels is analyzed based on the first feature to perform correlation analysis between channels to obtain the feature matrix between channels, including: obtaining the first data of the first feature of the first channel and the second data of the second feature of the second channel; obtaining the first average value of the first channel and the second average value of the second channel; analyzing the Pearson correlation coefficient between channels based on the first data, the second data, the first average value, and the second average value to form the correlation matrix between channels; processing the values ​​in the correlation matrix according to the preset threshold value to obtain the feature matrix between channels; after the feature matrix is ​​processed by the preset threshold value, the values ​​in the correlation matrix that are greater than or equal to the preset threshold value are 1, and the values ​​that are less than the preset threshold value are 0.

[0055] After obtaining the feature matrix, a diagonal matrix of learnable parameters is added to the feature matrix, and learning is performed based on the training data to determine the parameters corresponding to the diagonal matrix to determine the second feature. Specifically, as an optional embodiment, the adaptively adjustable weight information is incorporated into the feature matrix, and the weight information is learned based on the first feature to determine the second feature between channels, including: generating a diagonal matrix of learnable parameters, the parameters in the diagonal matrix are related to the weight information of the correlation relationship between channels; integrating the diagonal matrix into the feature matrix to obtain a self-loop perception matrix, and learning based on the first feature; calculating the degree matrix of the self-loop perception matrix, and normalizing the self-loop perception matrix and the degree matrix to obtain a symmetric normalized Laplace matrix as the second feature between channels.

[0056] This solution can perform linear transformation through two fully connected layers, and adopt hyperbolic tangent function activation to form fused features. Specifically, as an optional embodiment, the fusion of the first feature and the second feature to obtain the fused feature includes: performing hierarchical linear transformation processing on the first feature and the second feature to obtain an intermediate representation; and adopting hyperbolic tangent function to activate the intermediate representation to obtain the fused feature. Specifically, as an optional embodiment, the hierarchical linear transformation processing on the first feature and the second feature to obtain the intermediate representation includes: performing hierarchical linear transformation processing on the first feature and the second feature through the first fully connected layer to obtain an intermediate matrix; and processing the intermediate matrix through the second fully connected layer to obtain the intermediate representation.

[0057] After learning the spatial features between channels, this solution can also learn the temporal correlation of fused features based on a pre-trained classification model containing a multi-head attention mechanism, and integrate the features before and after the fused features to form pooled features for analysis. Specifically, as an optional embodiment, the classification model containing a multi-head attention mechanism is used to: obtain the time information of the fused features as a time feature; determine the relevant features of the fused features based on the time features of the fused features; determine the attention weights corresponding to the relevant features before and after the fused features, and integrate the relevant features into the fused features according to the attention weights to determine the pooled features; and use an activation function to process the pooled features to determine the results of the epilepsy analysis.

[0058] Before analyzing the data, the scheme can be preprocessed. Specifically, as an optional embodiment, the channel data of the channel of each EEG signal is obtained, including: obtaining the channel-related information of the channel of each EEG signal, and performing bandpass filtering within a preset frequency band; the channel-related information includes electroencephalogram data and time information, and the channel-related information includes first information corresponding to epileptic seizures and second information corresponding to non-epileptic seizures; the filtered channel-related information is segmented according to the time information and the preset time interval to obtain information segments, and normalized to obtain channel data. The preset frequency band is 4-30Hz, and the preset time interval can be 2s.

[0059] On the basis of the above embodiments, the present application also provides an epileptic seizure detection device based on feature fusion, such as Figure 3 As shown, the device comprises:

[0060] The first feature acquisition module 302 is used to acquire channel data of each channel of the EEG signal, and perform feature extraction on the channel data according to a convolutional neural network to obtain a first feature of the channel.

[0061] The second feature acquisition module 304 is used to perform correlation analysis between channels based on the first feature to determine the second feature between the channels.

[0062] The fusion feature acquisition module 306 is used to fuse the first feature with the second feature to obtain a fusion feature.

[0063] The classification model analysis module 308 is used to input the fused features into the classification model including the multi-head attention mechanism to determine the epilepsy analysis results.

[0064] The implementation of the embodiment of the present application is similar to the implementation of the above-mentioned method embodiment. The specific implementation can refer to the specific implementation of the above-mentioned method embodiment, which will not be repeated here.

[0065] The scheme of the present application can be applied to the scenario of disease identification based on EEG data. For example, it can be used to identify whether the patient has epilepsy based on EEG data so as to timely warn and other treatments. This scheme can obtain the channel data of the channel of the EEG signal to extract channel features, analyze the correlation between channels, extract inter-channel features, and then fuse the channel features and inter-channel features to adopt a classification model to analyze the fused features to determine whether an epileptic seizure occurs. This scheme can more accurately identify the onset of epilepsy by fusing the intra-channel features and inter-channel features, comprehensively considering the channel data itself and the correlation between the channel data. Specifically, this scheme can obtain the channel data of the channel of each EEG signal, and extract features from the channel data based on the convolutional neural network to obtain the first feature of the channel; perform correlation analysis between channels based on the first feature to determine the second feature between channels; fuse the first feature with the second feature to obtain a fused feature; input the fused feature into a classification model containing a multi-head attention mechanism to fuse the relevant features before and after the feature and analyze it to determine the epilepsy analysis result. This solution can pre-train convolutional neural networks and classification models including multi-head attention mechanisms for feature extraction and feature analysis.

[0066] It should be noted that the division of units and / or modules in the embodiments of the present application is schematic and is only a logical function division. There may be other division methods in actual implementation. In addition, each functional unit and / or module in each embodiment of the present application may be integrated into a processing unit and / or module, or each unit and / or module may exist physically separately, or two or more units and / or modules may be integrated into one unit and / or module. The above-mentioned integrated units and / or modules may be implemented in the form of hardware or in the form of software functional units and / or modules.

[0067] If the integrated units and / or modules are implemented in the form of software functional units and / or modules and sold or used as independent products, they can be stored in a processor-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the relevant technology or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) or a processor (processor) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), disk or optical disk and other media that can store program codes.

[0068] In addition, the data transmission device and the data transmission method provided in the above embodiments are based on the same application concept. Since the principles of solving problems by the method and the device are similar, the implementation of the device and the method can refer to each other, and the repeated parts will not be repeated.

[0069] Figure 4 A structural block diagram of a network device is shown according to an exemplary embodiment.

[0070] like Figure 4 As shown, the network device 1100 includes at least: a processor 1110 , a memory 1120 , and a transceiver 1130 .

[0071] The transceiver 1130 is used to receive and send data under the control of the processor 1110 .

[0072] exist Figure 4 In the embodiment, the bus architecture may include any number of interconnected buses and bridges, specifically linking together various circuits of one or more processors represented by processor 1110 and memory represented by memory 1120. The bus architecture may also link together various other circuits such as peripherals, voltage regulators, and power management circuits, which are well known in the art and are therefore not further described herein. The bus interface provides an interface. The transceiver 1130 may be a plurality of components, namely, a transmitter and a receiver, providing units and / or modules for communicating with various other devices on a transmission medium, which may include a wireless channel, a wired channel, an optical cable, and other transmission media.

[0073] The processor 1110 is responsible for managing the bus architecture and general processing, and the memory 1120 can store data used by the processor 1110 when performing operations.

[0074] Optionally, the processor 1110 may be a central processing unit (CPU), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or a complex programmable logic device (CPLD), and the processor 1110 may also adopt a multi-core architecture. The processor 1110 and the memory 1120 may also be physically arranged separately.

[0075] The processor 1110 calls the computer program stored in the memory 1120 to execute any one of the methods for allocating a cell radio network temporary identifier provided in the above embodiments of the present application according to the obtained executable instructions.

[0076] Figure 5 A structural block diagram of a user equipment is shown according to an exemplary embodiment.

[0077] like Figure 5 As shown, the user equipment 1300 includes at least: a processor 1310 , a memory 1320 , and a transceiver 1330 .

[0078] The transceiver 1330 is used to receive and send data under the control of the processor 1310 .

[0079] exist Figure 5 In the embodiment, the bus architecture may include any number of interconnected buses and bridges, specifically one or more processors represented by processor 1310 and various circuits of memory represented by memory 1320 are linked together. The bus architecture may also link together various other circuits such as peripherals, voltage regulators, and power management circuits, which are well known in the art and are therefore not further described herein. The bus interface provides an interface. The transceiver 1330 may be a plurality of components, namely, a transmitter and a receiver, providing units and / or modules for communicating with various other devices on a transmission medium, including transmission media such as wireless channels, wired channels, and optical cables. For different user devices, the user interface 1340 may also be an interface capable of externally and internally connecting required devices, and the connected devices include but are not limited to a keypad, a display, a speaker, a microphone, a joystick, and the like.

[0080] The processor 1310 is responsible for managing the bus architecture and general processing, and the memory 1320 can store data used by the processor 1310 when performing operations.

[0081] Optionally, the processor 1310 may be a CPU (central processing unit), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array) or a CPLD (Complex Programmable Logic Device), and the processor 1310 may also adopt a multi-core architecture. The processor 1310 and the memory 1320 may also be physically arranged separately.

[0082] The processor 1310 calls the computer program stored in the memory 1320 to execute any one of the methods for allocating a cell radio network temporary identifier provided in the above embodiments of the present application according to the obtained executable instructions.

[0083] It should be noted here that the above-mentioned device provided in the embodiment of the present application can implement all the method steps implemented in the above-mentioned method embodiment, and can achieve the same technical effect. The parts and beneficial effects of this embodiment that are the same as those in the method embodiment will not be described in detail here.

[0084] In addition, a storage medium is provided in an embodiment of the present application, on which a computer program is stored, and when the computer program is executed by the processor, the data transmission method in the above embodiments is implemented. The storage medium can be any available medium or data storage device that the processor can access, including but not limited to magnetic storage (such as floppy disk, hard disk, tape, magneto-optical disk (MO), etc.), optical storage (such as CD, DVD, BD, HVD, etc.), and semiconductor storage (such as ROM, EPROM, EEPROM, non-volatile memory (NAND FLASH), solid state drive (SSD)), etc.

[0085] In an embodiment of the present application, a program product is provided, for example, the program product is an FPGA chip or a DSP chip, and the program product includes executable instructions, and the executable instructions are stored in a storage medium. The processor reads the executable instructions from the storage medium, so that when the executable instructions are executed by the processor, the data transmission method in each of the above embodiments is implemented.

[0086] The scheme of the present application can be applied to the scenario of disease identification based on EEG data. For example, it can be used to identify whether the patient has epilepsy based on EEG data so as to timely warn and other treatments. This scheme can obtain the channel data of the channel of the EEG signal to extract channel features, analyze the correlation between channels, extract inter-channel features, and then fuse the channel features and inter-channel features to adopt a classification model to analyze the fused features to determine whether an epileptic seizure occurs. This scheme can more accurately identify the onset of epilepsy by fusing the intra-channel features and inter-channel features, comprehensively considering the channel data itself and the correlation between the channel data. Specifically, this scheme can obtain the channel data of the channel of each EEG signal, and extract features from the channel data based on the convolutional neural network to obtain the first feature of the channel; perform correlation analysis between channels based on the first feature to determine the second feature between channels; fuse the first feature with the second feature to obtain a fused feature; input the fused feature into a classification model containing a multi-head attention mechanism to fuse the relevant features before and after the feature and analyze it to determine the epilepsy analysis result. This solution can pre-train convolutional neural networks and classification models including multi-head attention mechanisms for feature extraction and feature analysis.

[0087] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage and optical storage, etc.) that contain computer-usable program code.

[0088] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer executable instructions. These computer executable instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0089] These processor executable instructions may also be stored in a processor readable memory that can direct a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the processor readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0090] These processor-executable instructions may also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable device to implement the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0091] It should be understood that, although the steps in the flowchart of the accompanying drawings are displayed in sequence as indicated by the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least a part of the steps in the flowchart of the accompanying drawings may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be executed in turn or alternately with other steps or at least a part of the sub-steps or stages of other steps.

[0092] The above description is only a partial implementation method of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A method for detecting epileptic seizures based on feature fusion, characterized in that: The method comprises: Acquire channel data of each channel of the EEG signal, and perform feature extraction on the channel data according to a convolutional neural network to obtain a first feature of the channel; Performing correlation analysis between channels based on the first feature to determine a second feature between the channels; The first feature is fused with the second feature to obtain a fused feature; The fused features were input into a classification model containing a multi-head attention mechanism to determine the epilepsy analysis results.

2. The method according to claim 1, characterized in that The performing correlation analysis between channels according to the first feature to determine the second feature between the channels includes: According to the first feature, the Pearson correlation coefficient between channels is analyzed to perform correlation analysis between channels and obtain a feature matrix between channels; Adaptively adjustable weight information is incorporated into the feature matrix, and the weight information is learned based on the first feature to determine the second feature between channels. The weight information corresponds to the correlation relationship between channels.

3. The method according to claim 2, characterized in that According to the first feature, the Pearson correlation coefficient between channels is analyzed to perform correlation analysis between channels, and a feature matrix between channels is obtained, including: Obtain first data of a first feature of a first channel and second data of a second feature of a second channel; Obtain a first average value of the first channel and a second average value of the second channel; Analyze the Pearson correlation coefficient between the channels according to the first data, the second data, the first average value, and the second average value to form a correlation relationship matrix between the channels; According to the preset threshold, the values ​​in the association relationship matrix are processed to obtain the feature matrix between channels; after the feature matrix is ​​processed by the preset threshold, the values ​​in the association relationship matrix that are greater than or equal to the preset threshold are 1, and the values ​​that are less than the preset threshold are 0.

4. The method according to claim 3, characterized in that The step of incorporating adaptively adjustable weight information into the feature matrix and learning the weight information based on the first feature to determine the second feature between channels includes: Generate a diagonal matrix of learnable parameters, where the parameters in the diagonal matrix are related to the weight information of the association relationship between channels; Integrate the diagonal matrix into the feature matrix to obtain the self-loop perception matrix, and learn based on the first feature; The degree matrix of the self-loop perception matrix is ​​calculated, and normalization is performed based on the self-loop perception matrix and the degree matrix to obtain a symmetric normalized Laplace matrix as the second feature between channels.

5. The method according to claim 4, characterized in that The step of fusing the first feature with the second feature to obtain a fused feature includes: Performing hierarchical linear transformation on the first feature and the second feature to obtain an intermediate representation; The intermediate representation is activated using a hyperbolic tangent function to obtain fused features.

6. The method according to claim 5, characterized in that The performing hierarchical linear transformation on the first feature and the second feature to obtain an intermediate representation includes: Through the first fully connected layer, the first feature and the second feature are subjected to hierarchical linear transformation to obtain an intermediate matrix; Through the second fully connected layer, the intermediate matrix is ​​processed to obtain the intermediate representation.

7. The method according to claim 6, characterized in that The classification model including the multi-head attention mechanism is used to: Obtain the time information of the fused features as the time feature; Determine relevant features of the fused features according to the temporal features of the fused features; Determine the attention weights corresponding to the relevant features before and after the fusion feature, and integrate the relevant features into the fusion feature according to the attention weights to determine the pooled feature; The activation function is used to process the pooled features to determine the epilepsy analysis results.

8. The method according to claim 1, characterized in that The step of obtaining channel data of each channel of the electroencephalogram signal comprises: Obtaining channel-related information of each EEG signal channel, and performing bandpass filtering within a preset frequency band; the channel-related information includes EEG data and time information, and the channel-related information includes first information corresponding to an epileptic seizure and second information corresponding to a non-epileptic seizure; The filtered channel-related information is segmented according to time information and preset time intervals to obtain information segments, which are then normalized to obtain channel data.

9. An epileptic seizure detection device based on feature fusion, characterized in that: The device comprises: A first feature acquisition module is used to acquire channel data of each channel of the EEG signal, and perform feature extraction on the channel data according to a convolutional neural network to obtain a first feature of the channel; A second feature acquisition module, used for performing correlation analysis between channels based on the first feature to determine a second feature between channels; A fusion feature acquisition module, used for fusing the first feature with the second feature to obtain a fusion feature; The classification model analysis module is used to input the fused features into the classification model including the multi-head attention mechanism to determine the epilepsy analysis results.

10. A network device, characterized in that: include: A memory, a transceiver, and a processor; wherein the memory is used to store a computer program; the transceiver is used to send and receive data under the control of the processor; The processor is used to read the computer program in the memory and execute the method according to claims 1-8.

Images