CN116269446A - Method, electronic equipment and medium for classifying format tower electroencephalogram signals based on algebraic topology - Google Patents

Abstract

The invention discloses a method for classifying Gestalt EEG signals based on algebraic topology, electronic equipment, and a medium, including performing a visual stimulation experiment on subjects with contour recognition pictures to obtain original Gestalt EEG signal data; Preprocessing the Gestalt EEG signal data; performing algebraic topology processing on the preprocessed Gestalt EEG signal data; constructing a Gestalt EEG signal classification network to classify the Gestalt EEG signal data after algebraic topology processing. The present invention introduces an algebraic topology tool, extracts the topological space features of the Gestalt EEG signal and digitizes it as a classification object, which reduces the data size by an order of magnitude, greatly reduces the computational complexity, saves computational resources, and can effectively improve the classification Accuracy.

Description

Gestalt EEG signal classification method based on algebraic topology, electronic equipment, medium

technical field

The invention belongs to the field of electroencephalogram signal classification, and in particular relates to a method for classifying gestalt electroencephalogram signals based on algebraic topology, electronic equipment and media.

Background technique

In recent years, the rapid development of deep learning has promoted many research fields, especially in the fields of computer vision, pattern recognition, signal processing and detection. With the continuous improvement and wide application of neural networks, deep learning has also been introduced into the processing of electroencephalogram signals. The EEG signals classified by the existing deep neural network can be mainly divided into several types such as emotion recognition, motor imagery, mental load test, epilepsy detection, and sleep stage scoring. They all belong to the EEG signals generated by perception and intuitive stimulation. Traditional EEG signal classification usually includes an important step of feature extraction, which is highly dependent on domain knowledge and is time-consuming and laborious; neural networks can automatically extract distinguishable features based on a large amount of data training, which greatly improves the Classification accuracy of EEG signals. However, when LSTM, which is good at processing time-series signals, and the current relatively powerful Transformer network classify gestalt and non-gestalt EEG signals at the conscious level, the classification accuracy is low.

Contents of the invention

Aiming at the deficiencies of the prior art, the present invention provides a method for classifying Gestalt EEG signals based on algebraic topology, electronic equipment, and media.

According to the first aspect of the embodiments of the present application, a method for classifying Gestalt EEG signals based on algebraic topology is provided, which specifically includes the following steps:

Step S1, subjecting subjects to a visual stimulation experiment of contour recognition pictures to obtain original Gestalt EEG signals;

Step S2, preprocessing the original Gestalt EEG signal;

Step S3, performing algebraic topology processing on the preprocessed Gestalt EEG signal;

Step S4, constructing a Gestalt EEG signal classification network to classify the Gestalt EEG signals after the algebraic topology processing.

According to a second aspect of the embodiments of the present application, an electronic device is provided, including a memory and a processor, the memory is coupled to the processor; wherein the memory is used to store program data, and the processor is used to The program data is executed to realize the above-mentioned algebraic topology-based Gestalt EEG signal classification method.

According to a third aspect of the embodiments of the present application, there is provided a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, the above-mentioned algebraic topology-based Gestalt EEG signal classification method is implemented.

Different from the prior art, the beneficial effects of the present invention are as follows: the present invention provides a method for classifying Gestalt EEG signals based on algebraic topology, by introducing an algebraic topology tool, extracting the topological space features of Gestalt EEG signals and Digitization as the classification object reduces the data size by orders of magnitude, greatly reduces the computational complexity, saves computing resources, and can effectively improve the classification accuracy.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings that need to be used in the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those skilled in the art, other drawings can also be obtained based on these drawings without any creative effort.

Fig. 1 is a flow chart of a method for classifying Gestalt EEG signals based on algebraic topology provided by an embodiment of the present invention;

Figure 2 is a gestalt and confusion map;

3 is a schematic diagram of collecting a single Gestalt EEG signal;

4 is a schematic diagram of preprocessing the original EEG signal;

Fig. 5 is a schematic diagram of algebraic topology processing;

6 is a schematic diagram of a Gestalt EEG signal classification network;

Fig. 7 is a schematic diagram of an electronic device provided by an embodiment of the present invention.

Detailed ways

The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some, not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.

It should be noted that, in the case of no conflict, the features in the following embodiments and implementation manners can be combined with each other.

The present invention provides a method for classifying Gestalt EEG signals based on algebraic topology. The Gestalt EEG signals are generated by subjects when viewing Gestalt and non-Gestalt (chaotic graph) pictures. An example of this is shown in Figure 2. The special feature of this kind of picture is that it is easy for people to distinguish them visually. The former three small circles with gaps form an obvious triangular outline, while the latter is just On the basis of the former, turn the small circles slightly to form, and the triangle outline is obviously not formed between the small circles. It is easy for people to distinguish the two kinds of pictures with the naked eye, but the neural network cannot. The classification results of the classic CNN, LSTM, Transformer, etc. for the two kinds of pictures are only about 50% accurate. Experiments have proved that neural networks have little effect on improving the classification effect of EEG signals. Therefore, the present invention introduces algebraic topology tools. Topology is based on the volume research of network shape and structure. Combined with algebraic analysis, it will be more Abstract high-dimensional spatial features are mapped or dimensionally reduced into easy-to-understand digital features, and then digital quantities are analyzed.

Fig. 1 shows a flow chart of a method for classifying Gestalt EEG signals based on algebraic topology provided by an embodiment of the present invention, the method specifically includes the following sub-steps:

In step S1, the subject is subjected to a visual stimulation experiment of contour recognition pictures to obtain original Gestalt EEG signal data.

Specifically, in this example, the process of obtaining original Gestalt EEG signal data through a visual stimulation experiment in step S1 is as follows:

Step S101, recruit 20 volunteers with normal eyesight and mental state, and healthy physical condition, and let the subjects receive the experimental stimulation and signal collection and recording alone in a shielded room without interference.

Step S102, each subject observes gestalt pictures 10 times and chaotic picture pictures 30 times, the duration of each observation is 10s, and the interval between each time is 1s, the specific process is shown in Figure 3.

Step S103, checking the validity of the collected Gestalt EEG signal data.

Step S2, preprocessing the original Gestalt EEG signal acquired in step S1, making it a data object suitable for processing as an algebraic topology tool.

As shown in Figure 4, the specific process of processing the original gestalt EEG signal in step S2 is as follows:

Step S201 , electrode selection; there are 64 electrodes in the collected original Gestalt EEG signals, and channels 1-59 are selected as processing objects, and channels 60-64 are positioning electrodes, which need to be deleted.

Step S202, filtering; using band-pass filtering to limit the frequency of the Gestalt EEG signal within the range of 0.1-45hz, which not only removes the interference signals of the higher and lower bands, but also basically retains all effective EEG signals;

Step S203, artifact removal; artifact interference from various other biological signals will inevitably occur during the signal acquisition process, so it is necessary to remove various artifact noises such as blinking, eye movement, and electromechanical noise. Based on the standard of the normal signal of the human brain, a noise amplitude threshold is set, and the trials with amplitude peaks exceeding the noise threshold are screened out;

Step S204, baseline calibration; the mean value of the EEG signal in the state 1 s before the start of each experiment is used as the baseline of the spontaneous EEG, and then the baseline is subtracted from the EEG signal to obtain a corresponding calibration signal.

Step S3, performing algebraic topology processing on the preprocessed Gestalt EEG signal.

Specifically, after the processing of step S2, the signal of each trial period is:

Step S301, performing Hilbert transform and integrating each signal in the signal F _EEG of each trial period, that is, each row;

Among them, N is the data length, and M is the number of electrodes for EEG signal collection.

Among them, t represents the time variable, and τ is the dummy variable in the convolution integral;

Step S302, calculate the instantaneous phase of each electrode:

Among them, f _i is the i-th line in F _EEG ;

Step S303, using the phase locking value (Phase Locking Value, PLV) phase synchronization analysis method to calculate the value of the corresponding element of the correlation matrix;

Among them, j is the imaginary number unit, φ ^P (n), φ ^q (n) represent the instantaneous phase of the nth sampling moment in electrodes p and q. Take the absolute value of the calculation result of formula (5), and normalize the entire matrix to get the final correlation matrix C _M×M :

Step S304, constructing a VR complex based on the correlation matrix C _M×M .

Specifically, the number of rows and columns in the incidence matrix C _M×M represents the initial number of independent components, and the value recorded in each row represents the "distance" between it and other components. It is conceivable that this series of M independent components in a high-dimensional space The components are fixedly arranged in space according to their position and distance, so that a VR (Vietoris-Rips complex) complex can be constructed. The filtering value process from zero connection to complex connection point set is shown in Figure 5. In this process, the connection relationship changes as the filtering value grows.

Step S305, calculate the algebraic object of the n-dimensional hole number in the topological space based on the VR complex—Betty number, which has three dimensions of β ₀ , β ₁ , and β ₂ features, and take β ₀ as the input of the Gestalt EEG signal classification network .

Step S4, constructing a Gestalt EEG signal classification network to classify the Gestalt EEG signals after the algebraic topology processing.

In this example, a Gestalt EEG signal classification network is constructed with a long short-term memory network (LSTM, Long Short-Term Memory) and a Transformer encoder as the basic network.

Further, the step S4 also includes: training the Gestalt EEG signal classification network.

Step S401, dividing the Gestalt EEG signal data to be classified into a training set and a testing set according to a ratio of 8:2.

Step S402, using the training set to train the Gestalt EEG signal classification network, and saving the model parameters.

Step S403, using the test set to test the accuracy of the Gestalt EEG signal classification network until reaching a self-defined accuracy threshold.

As shown in FIG. 7 , an embodiment of the present application provides an electronic device, which includes a memory 101 for storing one or more programs; and a processor 102 . When one or more programs are executed by the processor 102, the method according to any one of the first aspect above is realized.

It also includes a communication interface 103, the memory 101, the processor 102 and the communication interface 103 are electrically connected to each other directly or indirectly, so as to realize data transmission or interaction. For example, these components can be electrically connected to each other through one or more communication buses or signal lines. The memory 101 can be used to store software programs and modules, and the processor 102 executes various functional applications and data processing by executing the software programs and modules stored in the memory 101 . The communication interface 103 can be used for signaling or data communication with other node devices.

Wherein, memory 101 can be but not limited to, random access memory 101 (Random Access Memory, RAM), read-only memory 101 (Read Only Memory, ROM), programmable read-only memory 101 (Programmable Read-Only Memory, PROM), Erasable Programmable Read-Only Memory 101 (Erasable Programmable Read-Only Memory, EPROM), Electrically Erasable Programmable Read-Only Memory 101 (Electric Erasable Programmable Read-Only Memory, EEPROM) and the like.

The processor 102 may be an integrated circuit chip with signal processing capabilities. The processor 102 can be a general-purpose processor 102, including a central processing unit 102 (Central Processing Unit, CPU), a network processor 102 (Network Processor, NP), etc.; it can also be a digital signal processor 102 (Digital Signal Processing, DSP ), Application Specific Integrated Circuit (ASIC), Field Programmable Gate Array (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, and discrete hardware components.

In the embodiments provided in this application, it should be understood that the disclosed method and system may also be implemented in other ways. The method and system embodiments described above are only illustrative, for example, the flowcharts and block diagrams in the accompanying drawings show the system of possible implementations of methods and systems, methods and computer program products according to multiple embodiments of the present application Architecture, function and operation. In this regard, each block in a flowchart or block diagram may represent a module, program segment, or part of code that includes one or more Executable instructions. It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks in succession may, in fact, be executed substantially concurrently, or they may sometimes be executed in the reverse order, depending upon the functionality involved. It should also be noted that each block of the block diagrams and/or flowchart illustrations, and combinations of blocks in the block diagrams and/or flowchart illustrations, can be implemented by a dedicated hardware-based system that performs the specified function or action , or may be implemented by a combination of dedicated hardware and computer instructions.

In addition, each functional module in each embodiment of the present application may be integrated to form an independent part, each module may exist independently, or two or more modules may be integrated to form an independent part.

On the other hand, an embodiment of the present application provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by the processor 102, the method according to any one of the above-mentioned first aspects is implemented. If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or the 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 are used to make a computer device (which may be a personal computer, a server, or a network device, etc.) execute all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory 101 (ROM, Read-Only Memory), random access memory 101 (RAM, RandomAccess Memory), magnetic disk or optical disk, etc., which can store program codes. medium.

Other embodiments of the present application will readily occur to those skilled in the art from consideration of the specification and practice of the disclosure herein. This application is intended to cover any modification, use or adaptation of the application, these modifications, uses or adaptations follow the general principles of the application and include common knowledge or conventional technical means in the technical field not disclosed in the application . The specification and examples are to be considered as illustrative only.

It should be understood that the present application is not limited to the precise constructions which have been described above and shown in the accompanying drawings, and various modifications and changes may be made without departing from the scope thereof.

Claims (8)

1. a kind of Gestalt EEG signal classification method based on algebraic topology, it is characterized in that, described method comprises: Step S1, subjecting subjects to a visual stimulation experiment of contour recognition pictures to obtain original Gestalt EEG signals; Step S2, preprocessing the original Gestalt EEG signal; Step S3, performing algebraic topology processing on the preprocessed Gestalt EEG signal; Step S4, constructing a Gestalt EEG signal classification network to classify the Gestalt EEG signals after the algebraic topology processing. 2. the gestalt EEG signal classification method based on algebraic topology according to claim 1, is characterized in that, carrying out preprocessing to original gestalt EEG signal data comprises: Delete the Gestalt EEG signals collected by the positioning electrodes; Perform band-pass filtering on the Gestalt EEG signal, so that the frequency of the Gestalt EEG signal is within the range of 0.1-45hz; Remove noise from Gestalt EEG signals; Baseline calibration of Gestalt EEG signals. 3. the gestalt EEG signal classification method based on algebraic topology according to claim 2, is characterized in that, carrying out baseline calibration to gestalt EEG signal comprises: The mean value of the Gestalt EEG signal in the 1s state before the start of each visual stimulation experiment was used as the signal baseline, and then the collected Gestalt EEG signal was subtracted from the signal baseline to obtain the corresponding calibration signal. 4. the gestalt electroencephalographic signal classification method based on algebraic topology according to claim 1, is characterized in that, carrying out algebraic topology processing to the preprocessed gestalt electroencephalogram signal comprises: Step S301, perform Hilbert transform and integrate each signal in _{the FEEG} preprocessed in each visual stimulation experiment period, and the calculation formula is as follows:

Wherein, N is the data length, and M is the number of electrodes collected by the EEG signal;

Among them, t represents the time variable, and τ is the dummy variable in the convolution integral; Step S302, calculate the instantaneous phase φ of each electrode:

Among them, f _i is the i-th line in F _EEG ; Step S303, using the phase lock value phase synchronization analysis method to calculate the value of the corresponding element of the correlation matrix;

Among them, j is the imaginary number unit, φ ^P (n) and φ ^q (n) represent the instantaneous phase of the nth sampling moment in electrodes p and q; take the absolute value of the calculation result of formula (5), and normalize the whole matrix One can get the final correlation matrix C _M×M :

Step S304, constructing a VR complex based on the correlation matrix C _M×M ; Step S305, calculate the algebraic object of the n-dimensional hole number in the topological space based on the VR complex, that is, the Betti number, obtain the three-dimensional features of β ₀ , β ₁ , and β ₂ , and take β ₀ as the input of the Gestalt EEG signal classification network . 5. the algebraic topology-based Gestalt EEG signal classification method according to claim 1, wherein the Gestalt EEG signal classification network is constructed based on a long short-term memory network and a Transformer encoder. 6. The method for classifying Gestalt EEG signals based on algebraic topology according to claim 1 or 5, wherein said step S4 further comprises: training the Gestalt EEG signal classification network. 7. An electronic device comprising a memory and a processor, wherein the memory is coupled to the processor; wherein the memory is used to store program data, and the processor is used to execute the program data to Realize the algebraic topology-based Gestalt EEG signal classification method described in any one of claims 1-6. 8. A computer-readable storage medium on which a computer program is stored, characterized in that, when the program is executed by a processor, it realizes the Gestalt EEG based on algebraic topology as claimed in any one of claims 1-6 Signal Classification Method.

Images