Electroencephalogram signal feature extraction method and device based on frequency domain analysis

Abstract

The application provides a kind of electroencephalogram feature extraction method and device based on frequency domain analysis.The method first adjusts signal according to the characteristics of electroencephalogram and noise environment through dynamic filtering, then accurately captures the time-varying characteristics of the signal through energy distribution extraction on the frequency-time plane, and applies frequency domain coherence analysis technology to predict the functional connection between brain regions, while combining nonlinear dynamics theory to extract frequency domain nonlinear features of the signal.Finally, the extracted frequency domain features are screened by feature importance scoring method, and the target feature set is obtained to identify and classify electroencephalogram based on it.The application introduces dynamic filtering, frequency domain coherence analysis and nonlinear dynamics theory, solves the problems of insufficient non-stationary processing of electroencephalogram, ignoring time-varying characteristics and low accuracy and computational efficiency of feature extraction in the prior art.

Description

Technical Field

[0001] This invention relates to the field of electroencephalogram (EEG) signal processing technology, and more specifically, to an EEG signal feature extraction method based on frequency domain analysis, an EEG signal feature extraction device based on frequency domain analysis, a computer-readable storage medium, and an EEG wave analysis system. Background Technology

[0002] In the field of electroencephalogram (EEG) signal processing, frequency domain feature extraction methods have always been a research hotspot. Early EEG signal feature extraction mainly relied on time domain analysis, such as autoregressive models and the Higuchi algorithm, which could reflect the characteristics of EEG signals to a certain extent. However, with the continuous development of signal processing technology, frequency domain analysis has gradually become the mainstream research method due to its unique advantages. The application of techniques such as Fourier transform, power spectral density, and band power allows researchers to gain a deeper understanding of the intrinsic laws of EEG signals from a frequency perspective. The development of these technologies has provided strong support for the feature extraction, classification, and interpretation of EEG signals.

[0003] However, existing frequency domain analysis techniques still have certain limitations in EEG signal feature extraction. Traditional Fourier transforms are suitable for processing stationary signals, but the non-stationarity of EEG signals can lead to the loss of important signal information during processing. Existing frequency domain analysis methods often ignore the time-varying characteristics of signals at different frequencies, resulting in extracted features that may not fully reflect the dynamic changes of EEG signals. Furthermore, current frequency domain feature extraction methods are computationally expensive, and their feature discrimination capabilities still need improvement, making it difficult to meet the dual requirements of efficiency and accuracy in practical applications. Summary of the Invention

[0004] The main objective of this application is to provide a method for extracting EEG signal features based on frequency domain analysis, a device for extracting EEG signal features based on frequency domain analysis, a computer-readable storage medium, and an EEG wave analysis system, so as to at least solve the problems of insufficient processing of the non-stationarity of EEG signals, neglect of time-varying characteristics, and low accuracy and computational efficiency of feature extraction in the prior art.

[0005] To achieve the above objectives, according to one aspect of this application, a method for analyzing electroencephalogram (EEG) signals based on frequency domain analysis is provided, comprising: acquiring EEG signals to obtain a first target signal; performing dynamic filtering based on the signal characteristics of the first target signal and the noise environment to obtain a second target signal; extracting the energy distribution of the second target signal in the frequency-time plane to determine the dynamic change characteristics of the second target signal, thereby obtaining first distribution data; performing frequency domain coherence analysis on different second target signals based on the first distribution data using coherence spectral analysis technology to predict functional connectivity between different brain regions, thereby obtaining a first analysis result; performing nonlinear feature extraction in the frequency domain on the second target signal based on the first distribution data and nonlinear dynamics theory, thereby obtaining a second analysis result; converting the first analysis result and the second analysis result into a frequency domain feature set, filtering the frequency domain feature set using a feature importance scoring method to obtain a target feature set; and identifying and classifying the EEG signals based on the target feature set.

[0006] Optionally, acquiring EEG signals to obtain a first target signal, and performing dynamic filtering based on the signal characteristics of the first target signal and the noise environment to obtain a second target signal, includes: acquiring historical EEG signal data of different target objects through electrode pads to obtain a third target signal; acquiring physiological parameter data of the target object to obtain first target data; acquiring ambient sound through a microphone to obtain ambient noise data; acquiring vibration data through an accelerometer to obtain second target data; performing cluster analysis based on the third target signal, the first target data, the ambient noise data, and the second target data to obtain different categories of the EEG signals, resulting in multiple EEG signal groups; establishing feature models for each EEG signal group, and initializing filtering parameters corresponding to different categories based on the feature models; performing cluster analysis on the first target signal to obtain the target category corresponding to the first target signal; determining the corresponding target filtering parameters based on the target category; and filtering the first target signal based on the target filtering parameters to obtain the second target signal.

[0007] Optionally, the energy distribution of the second target signal in the frequency-time plane is extracted to determine the dynamic change characteristics of the second target signal, obtaining first distribution data, including: determining the standard deviation of the second target signal based on the second target signal, and determining the scale parameter of the multi-scale analysis window based on the standard deviation. Where λ is the scale parameter, σ(t) is the standard deviation, and ∈ is a value used to avoid division by zero; the second target signal is windowed based on the scale parameter to obtain the third target signal; an adaptive recursive wavelet transform is performed on the third target signal to extract the instantaneous frequency and amplitude of the third target signal, thus obtaining the first distribution data: E(f,t,λ)=|ARWT(x w (k),λ,t)| 2 Where E(f,t,λ) is the first distribution data, used to characterize the three-dimensional distribution of the third target signal in energy-frequency-time, where ARWT(x w (k),λ,t) satisfy: Where, x w (k) is the third target signal. ψ is a parameter used to limit the range of values ​​for the wavelet basis functions. λ (t) is the adaptive phase function, and j is the imaginary unit.

[0008] Optionally, using coherence spectral analysis, frequency domain coherence analysis is performed on different second target signals based on the first distribution data to predict functional connectivity between different brain regions, obtaining a first analysis result, including: performing coherence calculation based on the second target signals of different channels and the first distribution data corresponding to the second target signals to obtain target coherence. Among them, C ij (f,t) represents the target coherence, S ij (f,t) represents the cross-spectral density between the first distribution data corresponding to the second target signals in channels i and j, where S ij (f,t) satisfies: in, For X i(f,t,λ) The conjugate of X i(f,t,λ) The data is extracted from the first distribution data by a time window, where T is the length of the time window; a weighted coherence index is calculated based on the target coherence to characterize the degree of functional connectivity between different brain regions: Among them, Γ ij (f) is the weighted coherence index, and η(f) is the frequency energy weighting factor used to enhance the coherence of a specific frequency band; the weighted coherence index and the target coherence are determined as the first analysis result.

[0009] Optionally, based on the first distribution data and combined with nonlinear dynamics theory, nonlinear features in the frequency domain of the second target signal are extracted to obtain a second analysis result, including: determining the number of boxes based on the first distribution data, and determining the fractal dimension based on the number of boxes. Where D is the fractal dimension, and N(∈) is the number of boxes determined based on the signal energy distribution in the first distribution data; the complexity index of the second target signal is determined based on the fractal dimension and the time series entropy of the second target signal: Ξ ′ =D·H(X); where, Ξ ′ H(X) is the complexity index, and H(X) is the time series entropy; the complexity index is determined as the second analysis result.

[0010] Optionally, converting the first analysis result and the second analysis result into a frequency domain feature set includes: converting the first analysis result and the second analysis result into a feature vector set to obtain the frequency domain feature set.

[0011] Optionally, the frequency domain feature set is filtered using a feature importance scoring method to obtain a target feature set, including: using a portion of the data in the frequency domain feature set as the initial state of the target feature set; and calculating a target score R(f) based on a scoring function on the feature vectors in the frequency domain feature set using the feature importance scoring method. i )=α·Corr(f i ,y)-β·Redun(f i ,S); where R(f i ) is the target score, Corr(f) i ,y) is the eigenvector f i Correlation with the target variable y, Redun(f i S) is the feature vector f i The redundancy of the current state of the target feature set S is calculated, where α is a weighting factor for balancing correlation and β is a weighting factor for balancing redundancy; the target feature set is sorted based on the target score, and the target feature set is updated with a previously preset number of feature vectors; the target score is recalculated based on the updated target feature set, and the target feature combination is updated according to the recalculated target score, until the difference of the target score between two adjacent iterations of the same feature vector is less than a second threshold or the maximum number of iterations is reached, and the target feature set is output.

[0012] According to another aspect of this application, a device for extracting features from electroencephalogram (EEG) signals based on frequency domain analysis is provided. The device includes: a first acquisition unit for acquiring EEG signals to obtain a first target signal, and performing dynamic filtering based on the signal characteristics of the first target signal and a noise environment to obtain a second target signal; a first calculation unit for extracting the energy distribution of the second target signal in the frequency-time plane to determine the dynamic change characteristics of the second target signal and obtain first distribution data; a second calculation unit for performing frequency domain coherence analysis on different second target signals based on the first distribution data using coherence spectral analysis technology to predict functional connectivity between different brain regions and obtain a first analysis result; a third calculation unit for performing nonlinear feature extraction in the frequency domain on the second target signal based on the first distribution data and nonlinear dynamics theory to obtain a second analysis result; a fourth calculation unit for converting the first analysis result and the second analysis result into a frequency domain feature set, filtering the frequency domain feature set using a feature importance scoring method to obtain a target feature set; and a classification unit for identifying and classifying the EEG signals based on the target feature set.

[0013] According to another aspect of this application, a computer-readable storage medium is provided, the computer-readable storage medium including a stored program, wherein, when the program is executed, it controls the device on which the computer-readable storage medium is located to perform any of the methods described.

[0014] According to another aspect of this application, a brainwave analysis system is provided, comprising: one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs including methods for performing any one of the methods described.

[0015] Applying the technical solution of this application, the method first adjusts the signal according to the characteristics of the EEG signal and the noise environment through dynamic filtering. Then, it accurately captures the time-varying characteristics of the signal by extracting energy distribution on the frequency-time plane, and uses frequency domain coherence analysis to predict functional connectivity between brain regions. Simultaneously, it combines nonlinear dynamics theory to extract nonlinear features from the signal in the frequency domain. Finally, it uses a feature importance scoring method to filter the extracted frequency domain features, obtaining a target feature set, and uses this set for EEG signal recognition and classification. This application, by introducing dynamic filtering, frequency domain coherence analysis, and nonlinear dynamics theory, overcomes the shortcomings of traditional methods that may lose key information when processing non-stationary signals. Frequency domain coherence analysis and nonlinear feature extraction deeply explore the time-varying characteristics and nonlinear laws of the signal, while the feature importance scoring method improves the efficiency and discrimination ability of feature selection, thereby comprehensively improving the accuracy and efficiency of EEG signal analysis. It solves the problems of insufficient handling of non-stationarity of EEG signals, neglect of time-varying characteristics, and low accuracy and computational efficiency of feature extraction in existing technologies. Attached Figure Description

[0016] Figure 1 A hardware structure block diagram of a mobile terminal for extracting EEG signal features based on frequency domain analysis according to an embodiment of this application is shown.

[0017] Figure 2 A flowchart illustrating a frequency domain analysis-based method for extracting features from electroencephalogram (EEG) signals according to an embodiment of this application is shown.

[0018] Figure 3 A structural block diagram of an EEG signal feature extraction device based on frequency domain analysis according to an embodiment of this application is shown.

[0019] The above figures include the following reference numerals:

[0020] 102. Processor; 104. Memory; 106. Transmission device; 108. Input / output device. Detailed Implementation

[0021] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0022] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0023] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate for the embodiments of this application described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0024] As described in the background section, existing technologies have shortcomings in EEG signal analysis, mainly in their inability to effectively handle non-stationary signals, neglect of the time-varying characteristics of signals, and low discriminative ability and computational efficiency of feature extraction, making it difficult to meet the requirements of high precision and high efficiency. To address the problems of insufficient handling of non-stationary EEG signals, neglect of time-varying characteristics, and low accuracy and computational efficiency of feature extraction in existing technologies, embodiments of this application provide an EEG signal feature extraction method based on frequency domain analysis, an EEG signal feature extraction device based on frequency domain analysis, a computer-readable storage medium, and an EEG wave analysis system.

[0025] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0026] The methods and embodiments provided in this application can be executed on a mobile terminal, computer terminal, or similar computing device. Taking running on a mobile terminal as an example, Figure 1 This is a hardware structure block diagram of a mobile terminal for a method of extracting EEG signal features based on frequency domain analysis, according to an embodiment of the present invention. Figure 1 As shown, a mobile terminal may include one or more ( Figure 1Only one is shown in the diagram. A processor 102 (which may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.) and a memory 104 for storing data are also shown. The mobile terminal may further include a transmission device 106 for communication functions and an input / output device 108. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the mobile terminal described above. For example, the mobile terminal may also include components that are more... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0027] The memory 104 can be used to store computer programs, such as application software programs and modules, like the computer program corresponding to the device information display method in this embodiment of the invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, thereby implementing the above-described method. The memory 104 may include high-speed random access memory and non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to the mobile terminal via a network. Examples of the aforementioned networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof. The transmission device 106 is used to receive or send data via a network. Specific examples of the aforementioned networks may include wireless networks provided by the mobile terminal's communication provider. In one example, the transmission device 106 includes a network interface controller (NIC), which can be connected to other network devices via a base station to communicate with the Internet. In one example, the transmission device 106 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0028] This embodiment provides a frequency domain analysis-based EEG signal feature extraction method that runs on a mobile terminal, computer terminal, or similar computing device. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Also, although the logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0029] Figure 2 This is a flowchart of a frequency domain analysis-based EEG signal feature extraction method according to an embodiment of this application. Figure 2 As shown, the method includes the following steps:

[0030] Step S201: Acquire EEG signals to obtain a first target signal; perform dynamic filtering based on the signal characteristics of the first target signal and the noise environment to obtain a second target signal.

[0031] Specifically, EEG signal data is collected, and filtering parameters are learned and adjusted in real time to address the complexity of signal processing and analysis caused by differences in EEG signal characteristics and noise environments among individuals. Dynamic filtering is applied to the EEG signal characteristics and noise environments of different individuals to obtain the second target signal. The introduction of dynamic filtering effectively improves the quality of EEG signals and reduces interference caused by noise. The filtering strategy can adjust parameters according to the characteristics of the signal, so that the final acquired second target signal can more accurately reflect EEG activity.

[0032] Step S202: Extract the energy distribution of the second target signal on the frequency-time plane to determine the dynamic change characteristics of the second target signal and obtain the first distribution data;

[0033] Specifically, it accurately captures the dynamic changes of EEG signals in the frequency-time plane to enable effective analysis, accurately assess the functional connectivity between different brain regions, and provide key information for subsequent functional connectivity analysis and nonlinear feature extraction.

[0034] Step S203: Using coherence spectral analysis technology, frequency domain coherence analysis is performed on different second target signals based on the first distribution data to predict the functional connectivity between different brain regions and obtain the first analysis result.

[0035] Specifically, frequency domain coherence analysis is performed on the frequency-time plane data of the second target signal using coherence spectral analysis. This analysis can reveal the functional connectivity between different brain regions, and the coherence spectrum reveals the coordinated activity of different brain regions in specific frequency bands.

[0036] Step S204: Based on the first distribution data and combined with nonlinear dynamics theory, nonlinear features of the second target signal in the frequency domain are extracted to obtain the second analysis result;

[0037] Specifically, based on the frequency domain distribution data in step S202, and combined with nonlinear dynamics theory, nonlinear feature extraction is performed on the second target signal. Nonlinear analysis methods can reveal the complex dynamic characteristics hidden in the signal and capture valuable information that traditional linear methods cannot discover.

[0038] Step S205: Convert the first analysis result and the second analysis result into a frequency domain feature set, and filter the frequency domain feature set by the feature importance scoring method to obtain the target feature set;

[0039] Specifically, the analysis results obtained in steps S203 and S204 are transformed into a frequency domain feature set. Then, a feature importance scoring method is applied to filter these features, extracting a target feature set with high classification efficiency. This step aims to reduce redundant information and improve the accuracy of subsequent classification tasks through scientific feature selection methods. The filtered target feature set contains the information most influential on the classification task, significantly improving classification accuracy and efficiency.

[0040] Step S206: Identify and classify EEG signals based on the target feature set.

[0041] Specifically, after obtaining the target feature set, these features are used to identify and classify EEG signals, enabling efficient identification of different types of EEG signals. This step provides classification results for subsequent EEG signal analysis, supporting practical applications such as clinical diagnosis and neurological function monitoring.

[0042] This embodiment provides a method for extracting features from electroencephalogram (EEG) signals based on frequency domain analysis. It involves acquiring signal data, learning and adjusting filtering parameters in real time to optimize filtering for the characteristics of different individuals' EEG signals and noise environments; extracting the energy distribution of the signal in the frequency-time plane to reveal the dynamic characteristics of the signal; using coherence spectral analysis to identify the frequency domain coherence between different EEG signal channels and assess the functional connectivity between brain regions; combining nonlinear dynamics theory to extract the nonlinear features of the frequency domain signal and analyze the complexity and chaotic nature of the EEG signal; optimizing the extracted frequency domain features to identify the most contributing feature subset; and outputting the optimized frequency domain feature set for EEG signal classification, recognition, and other advanced analyses. This application optimizes the frequency domain analysis algorithm by introducing dynamic filtering, frequency domain coherence analysis, and nonlinear dynamics theory. Dynamic filtering overcomes the shortcomings of traditional methods that may lose key information when processing non-stationary signals. Frequency domain coherence analysis and nonlinear feature extraction deeply explore the time-varying characteristics and nonlinear laws of the signal, while the feature importance scoring rule improves the efficiency and discrimination ability of feature selection, thereby comprehensively improving the accuracy and efficiency of EEG signal analysis and providing more reliable technical support for EEG signal research and application. Compared with existing technologies, it may have significant advantages in processing non-stationary signals, extracting time-varying features, and reducing computational costs. It solves the problems of insufficient handling of non-stationarity of EEG signals, neglect of time-varying characteristics, and low accuracy and computational efficiency of feature extraction in existing technologies.

[0043] As one possible implementation, an electroencephalogram (EEG) signal is acquired to obtain a first target signal. Based on the signal characteristics of the first target signal and the noise environment, dynamic filtering is performed to obtain a second target signal, including:

[0044] Step S301: Collect historical EEG signal data of different target objects through electrode pads to obtain the third target signal;

[0045] Step S302: Collect physiological parameter data of the target object to obtain first target data; collect ambient sound through a microphone to obtain ambient noise data; collect vibration data through an accelerometer to obtain second target data;

[0046] Specifically, electrode pads can precisely acquire the electrical activity of the brain, providing information on neural activity in various brain regions as a basis for subsequent analysis. Voltage time series and physiological parameter data (such as heart rate and body temperature) are recorded using scalp electrodes on different target subjects, while environmental sound (such as background noise) and vibration data are acquired via microphones and accelerometers. This data, as supplementary information, provides additional background information about the target subject and its surrounding environment, making the analysis of EEG signals more accurate and reliable, especially in noisy environments.

[0047] Step S303: Based on the third target signal, the first target data, the environmental noise data, and the second target data, cluster analysis is performed to obtain different categories of EEG signals and obtain multiple EEG signal groups;

[0048] Specifically, a feature model is established for each target object through preliminary EEG signal analysis. Cluster analysis is then used to comprehensively analyze the third target signal (historical EEG signal data), the first target data (physiological parameter data), environmental noise data, and the second target data (vibration data) to obtain different categories of EEG signals. Especially when the target object is in different physiological states or environmental conditions, cluster analysis helps to clearly distinguish the different categories of EEG signals, thus providing an accurate signal classification basis for subsequent filtering processing.

[0049] Step S304: Establish feature models for each EEG signal group, and initialize filter parameters corresponding to different categories based on the feature models;

[0050] Specifically, feature models help describe and summarize the typical characteristics of each signal category, and the initialization of filter parameters is adjusted based on these feature models. Establishing feature models for each EEG signal group (each category) accurately captures the unique attributes of each type of EEG signal, thereby optimizing filter parameter settings. Through proper filter parameter initialization, targeted processing for each signal category can be ensured, minimizing noise interference and improving filtering effectiveness.

[0051] Let Θ i The filter parameter set for the i-th type of target object is initialized as: Θ i ={θ i1,θ i2 ,…,θ in}; where θ ij It is the j-th filter parameter for the i-th type of target object.

[0052] The system monitors the characteristic changes of EEG signals in real time and dynamically adjusts the filtering parameters based on the real-time tracked characteristics to adapt to the current state of the target object.

[0053] Step S305: Perform cluster analysis on the first target signal to obtain the target category corresponding to the first target signal, determine the corresponding target filtering parameters based on the target category, and filter the first target signal based on the target filtering parameters to obtain the second target signal.

[0054] Specifically, cluster analysis is performed on the first target signal to determine its target category. Based on this category, corresponding target filtering parameters are selected to filter the first target signal, resulting in the second target signal. The filtered signal removes unnecessary noise and retains the EEG signals valuable for analysis. By associating EEG signals with specific categories and applying appropriate filtering parameters, more precise noise suppression and signal enhancement can be achieved.

[0055] As one possible implementation, the energy distribution of the second target signal in the frequency-time plane is extracted to determine the dynamic variation characteristics of the second target signal, resulting in first distribution data, including:

[0056] Step S401: Based on the second target signal, determine the standard deviation of the second target signal, and determine the scale parameters of the multi-scale analysis window based on the standard deviation. Where λ is the scale parameter, σ(t) is the standard deviation, and ∈ is a value used to avoid division by zero;

[0057] Step S402: Window the second target signal based on the scale parameter to obtain the third target signal;

[0058] Specifically, extracting the energy distribution of the signal in the frequency-time plane includes automatically selecting an appropriate multi-scale analysis window based on the waveform characteristics of the EEG signal to capture frequency components at different scales. The standard deviation calculation accurately reflects the signal's fluctuations at different time points. Using the scale parameter λ to control the window size helps to adaptively adjust the width of the analysis window in both frequency and time, thereby improving the signal's time-frequency resolution. Based on the scale parameter λ calculated in step S401, the second target signal is further windowed to obtain the third target signal. The windowing process divides the signal into multiple time windows for analysis, with the size of each window adaptively determined by the scale parameter λ. Windowing refines the analysis of the signal in both time and frequency, improving the resolution of local signal characteristics.

[0059] Therefore, the scale-adaptive windowing method can ensure more accurate local analysis of signals in different frequency and time ranges, thereby optimizing subsequent frequency domain analysis and feature extraction.

[0060] Step S403: Perform adaptive recursive wavelet transform based on the third target signal to extract the instantaneous frequency and amplitude of the third target signal, and obtain the first distribution data: E(f,t,λ)=|ARWT(x w (k),λ,t)| 2 ;

[0061] Where E(f,t,λ) is the first distribution data, used to characterize the three-dimensional distribution of the third target signal in energy-frequency-time, where ARWT(x w (k),λ,t) satisfy:

[0062]

[0063] Where, x w (k) represents the third target signal. ψ is a parameter used to limit the range of values ​​for the wavelet basis functions. λ (t) is the adaptive phase function, and j is the imaginary unit.

[0064] Specifically, an adaptive complex modulation wavelet transform is applied to each windowed signal segment to extract the instantaneous frequency and amplitude of the signal, obtaining the three-dimensional energy-frequency-time distribution E(f,t,λ). The adaptive recursive wavelet transform provides high time-frequency resolution, accurately capturing the changes in the instantaneous frequency and amplitude of the signal. The ARWT method employs adaptive phase functions and wavelet basis functions to perform time-frequency analysis on the signal recursively, thereby obtaining the energy distribution of the signal in frequency, time, and scale, thus providing a more detailed description of the signal's dynamic characteristics. This method is particularly suitable for analyzing non-stationary signals, such as electroencephalogram (EEG) signals, efficiently extracting dynamic change information and providing important features for subsequent analysis and classification.

[0065] As one possible implementation, coherence spectral analysis is used to perform frequency domain coherence analysis on different second target signals based on first distribution data to predict functional connectivity between different brain regions, obtaining a first analysis result, including:

[0066] Step S501: Based on the second target signal from different channels and the first distribution data corresponding to the second target signal, perform coherence calculation to obtain the target coherence:

[0067] Among them, C ij (f,t) represents the target coherence, Sij (f,t) represents the cross-spectral density between the first distribution data corresponding to the second target signals of channels i and j.

[0068] Among them, S ij (f,t) satisfies: in, For X i(f,t,λ) The conjugate of X i(f,t,λ) This refers to the data extracted from the first distribution data using a time window, where T is the length of the time window.

[0069] Specifically, coherence calculations are performed based on the second target signals from different channels and their corresponding first distribution data. Coherence measures the linear relationship between signals. Through coherence calculations, the degree of functional connectivity between different brain regions can be effectively assessed, revealing the collaborative working patterns of different regions in EEG signals. Frequency domain coherence analysis can provide more accurate functional connectivity information than time domain analysis, helping to study the dynamic changes in brain networks.

[0070] Step S502: Calculate the weighted coherence index based on target coherence to characterize the degree of functional connectivity between different brain regions.

[0071] Among them, Γ ij (f) is the weighted coherence index, and η(f) is the frequency energy weighting factor, which is used to enhance the coherence of a specific frequency band;

[0072] Specifically, a weighted coherence index is calculated based on target coherence to further quantify the strength of functional connectivity between different brain regions. η(f) helps to emphasize coherence in specific frequency bands, particularly in certain bands such as alpha, beta, or theta waves, which are particularly important for the functional connections of brain activity. The weighted coherence index, through frequency weighting, can focus on specific frequency bands in the brain, further improving the ability to detect important areas of brain electrical activity. This method not only quantifies the functional connectivity between brain regions but also reveals the contribution of specific frequency bands to functional connectivity, contributing to a deeper understanding of the impact of different frequency bands on brain function patterns.

[0073] Step S503: The weighted coherence index and the target coherence are determined as the first analysis result.

[0074] Specifically, by combining target coherence and weighted coherence indices, richer and more precise analyses of brain functional connectivity can be provided. This comprehensive analysis not only focuses on overall coherence but also emphasizes the role of frequency, thus offering profound insights into the interactions and collaborative work between different brain regions.

[0075] As one possible implementation, based on the first distribution data and combined with nonlinear dynamics theory, nonlinear features in the frequency domain of the second target signal are extracted to obtain a second analysis result, including:

[0076] Step S601: Determine the number of boxes based on the first distribution data, and determine the fractal dimension based on the number of boxes: Where D is the fractal dimension, and N(∈) is the number of boxes determined based on the signal energy distribution in the first distribution data;

[0077] Specifically, extracting fractal dimensions can reveal the self-similarity of EEG signals at different scales, helping to analyze the complex structure of brain signals. In neuroscience, fractal dimensions are closely related to the dynamic activity patterns of the brain and can effectively reflect the complexity of brain activity. The box counting method is used to calculate the fractal dimension D. By changing the scale of the boxes, the self-similarity of signals at different scales is analyzed, and N(∈) is calculated by selectively covering points in the high-energy frequency band. By calculating the fractal dimension, the complexity and nonlinear characteristics of the signal can be quantified.

[0078] Step S602: Determine the complexity index of the second target signal based on the fractal dimension and the time series entropy of the second target signal: Ξ ′ =D·H(X);

[0079] Among them, Ξ ′ H(X) is the time series entropy, representing the degree of uncertainty or disorder in the signal. Higher entropy indicates a more irregular and unpredictable signal, reflecting its more complex dynamic characteristics.

[0080] Specifically, the complexity index combines the fractal dimension and time-series entropy of a signal to measure its overall complexity. By incorporating the combination of time-series entropy and fractal dimension, the complexity index can more comprehensively assess the nonlinear characteristics and dynamic changes of EEG signals. A higher complexity index indicates a more complex dynamic structure of the EEG signal, reflecting more complex interaction patterns between brain regions.

[0081] Step S603: The complexity index is determined as the second analysis result.

[0082] Specifically, the complexity index, as a secondary analytical result, provides high-dimensional features for further analysis of brain signals, helping to deepen the understanding of the nonlinear dynamic characteristics of the signals. This analytical result not only improves the accuracy of signal classification but also reveals the functional complexity of different brain regions.

[0083] As one possible implementation, the first analysis result and the second analysis result are converted into a frequency domain feature set, including:

[0084] The first and second analysis results are converted into a set of feature vectors to obtain a frequency domain feature set.

[0085] Specifically, the first analysis result (weighted coherence index) obtained in step S501 (coherence spectrum analysis) and the second analysis result (complexity index) obtained in step S602 (nonlinear dynamics feature extraction) are first converted into a set of feature vectors. The weighted coherence index reflects the degree of functional connectivity between different brain regions. Its value can be regarded as one dimension of a multidimensional feature vector, representing the interaction of brain regions at different frequency bands. The complexity index reflects the dynamic complexity of brain signals, and its value can also be used to construct a feature vector. Through the form of feature vectors, these two analysis results are transformed into a unified frequency domain feature set for subsequent signal recognition and classification.

[0086] As one possible implementation, the extracted frequency domain features are optimized and selected to identify the most contributing feature subset. The frequency domain feature set is filtered using a feature importance scoring method to obtain the target feature set, which includes:

[0087] Step S701: Use a portion of the data in the frequency domain feature set as the initial state of the target feature set;

[0088] Specifically, the extracted frequency domain features are represented as a set of feature vectors, F = {f1, f2, ..., f n}; where f i It is the i-th feature vector. This set contains multiple feature vectors extracted from frequency domain analysis, which represent multi-dimensional information of the EEG signal.

[0089] Step S702: Based on the feature importance scoring method, calculate the target score R(f) for the feature vectors in the frequency domain feature set using a scoring function: i )=α·Corr(f i ,y)-β·Redun(f i ,S);

[0090] This involves initializing a target feature subset S, using a feature importance scoring method, and defining a scoring function R(f i ) is the target score, Corr(f) i ,y) is the eigenvector f i The correlation with the target variable y reflects the impact of this feature on the final identification result. Redun(f i S) is the eigenvector f iThe redundancy of the current state of the target feature set S is intended to remove redundant and highly similar features, thereby reducing the complexity of the feature set. α is a weighting factor for balancing correlation, and β is a weighting factor for balancing redundancy.

[0091] Step S703: Sort the target features based on the target scores and update the target feature set with a previously preset number of feature vectors;

[0092] Specifically, the feature vectors are sorted based on the calculated target scores, and the feature vector with the highest score is selected to update the target feature set.

[0093] Step S704: Recalculate the target score based on the updated target feature set, and update the target feature combination according to the recalculated target score until the difference in target score between two adjacent iterations of the same feature vector is less than the second threshold or the maximum number of iterations is reached, and output the target feature set.

[0094] Specifically, the target score is recalculated based on the updated target feature set, and the target feature set is adjusted in the next iteration. Through multiple iterations, the algorithm continuously adjusts the feature subset, ensuring gradual improvement of the feature subset until the optimal or near-optimal feature subset is found. Furthermore, through an iterative optimization process, the feature subset with the highest fitness is selected. In each iteration, the result of the previous iteration guides the search direction of the next iteration, ensuring the correlation between formulas. This process continues until the change in the target score is less than a set second threshold, or the maximum number of iterations is reached, ultimately outputting the optimal target feature set. The final optimized frequency domain features consist of the selected most contributing feature subset and are used for the classification, recognition, and other advanced analyses of EEG signals.

[0095] To enable those skilled in the art to better understand the technical solution of this application, the implementation process of the frequency domain analysis-based EEG signal feature extraction method of this application will be described in detail below with reference to specific embodiments.

[0096] This embodiment relates to a specific method for extracting features from electroencephalogram (EEG) signals based on frequency domain analysis. To verify the beneficial effects of this application, scientific demonstration is conducted through experiments.

[0097] Twenty healthy adults aged 25 to 35 years were selected as subjects. Before the experiment, all subjects signed informed consent forms and rested in a quiet laboratory environment to acclimatize. Subjects wore a custom-designed high-density EEG acquisition device containing 64 channels, recording EEG activity at a sampling rate of 2,048 Hz. Simultaneously, physiological data such as heart rate and respiration, as well as environmental noise and vibration information, were collected using synchronous equipment. Experimental data are shown in Table 1.

[0098] Table 1

[0099]

[0100] Under the same signal-to-noise ratio (20dB), the feature extraction time of this application (1.2s) is shorter than that of prior art A (1.5s) and B (1.8s), demonstrating higher processing efficiency. Meanwhile, the classification accuracy of the method in this application (95%) is higher than that of prior art A (90%) and B (85%), indicating higher accuracy in EEG signal classification.

[0101] In terms of feature optimization rate, the method of this application achieves 85%, while the prior art A and B only achieve 75% and 70%, respectively. This indicates that the method of this application is more effective in optimizing feature subsets and can better screen out features that contribute to classification.

[0102] When the signal-to-noise ratio (SNR) drops to 15 dB, the method of this application still maintains a high classification accuracy (92%) and feature optimization rate (80%), while the performance of existing technologies declines. This further demonstrates the robustness and superiority of the method of this application in noisy environments.

[0103] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0104] This application also provides a frequency domain analysis-based EEG signal feature extraction device. It should be noted that this frequency domain analysis-based EEG signal feature extraction device can be used to execute the frequency domain analysis-based EEG signal feature extraction method provided in this application. This device is used to implement the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0105] The following describes the electroencephalogram (EEG) signal feature extraction device based on frequency domain analysis provided in the embodiments of this application.

[0106] Figure 3 This is a structural block diagram of a frequency domain analysis-based EEG signal feature extraction device according to an embodiment of this application. Figure 3 As shown, the device includes: a first acquisition unit 10, a first calculation unit 20, a second calculation unit 30, a third calculation unit 40, a fourth calculation unit 50, and a classification unit 60.

[0107] The first acquisition unit 10 is used to acquire EEG signals, obtain a first target signal, and perform dynamic filtering based on the signal characteristics of the first target signal and the noise environment to obtain a second target signal.

[0108] Specifically, EEG signal data is collected, and filtering parameters are learned and adjusted in real time to address the complexity of signal processing and analysis caused by differences in EEG signal characteristics and noise environments among individuals. Dynamic filtering is applied to the EEG signal characteristics and noise environments of different individuals to obtain the second target signal. The introduction of dynamic filtering effectively improves the quality of EEG signals and reduces interference caused by noise. The filtering strategy can adjust parameters according to the characteristics of the signal, so that the final acquired second target signal can more accurately reflect EEG activity.

[0109] The first computing unit 20 is used to extract the energy distribution of the second target signal on the frequency-time plane to determine the dynamic change characteristics of the second target signal and obtain the first distribution data.

[0110] Specifically, it accurately captures the dynamic changes of EEG signals in the frequency-time plane to enable effective analysis, accurately assess the functional connectivity between different brain regions, and provide key information for subsequent functional connectivity analysis and nonlinear feature extraction.

[0111] The second computing unit 30 is used to perform frequency domain coherence analysis on different second target signals based on the first distribution data using coherence spectrum analysis technology, so as to predict the functional connectivity between different regions of the brain and obtain the first analysis result.

[0112] Specifically, frequency domain coherence analysis is performed on the frequency-time plane data of the second target signal using coherence spectral analysis. This analysis can reveal the functional connectivity between different brain regions, and the coherence spectrum reveals the coordinated activity of different brain regions in specific frequency bands.

[0113] The third calculation unit 40 is used to extract nonlinear features in the frequency domain of the second target signal based on the first distribution data and nonlinear dynamics theory, and obtain the second analysis result.

[0114] Specifically, based on frequency domain distribution data and combined with nonlinear dynamics theory, nonlinear feature extraction is performed on the second target signal. Nonlinear analysis methods can reveal complex dynamic characteristics hidden in signals and capture valuable information that traditional linear methods cannot discover.

[0115] The fourth calculation unit 50 is used to convert the first analysis result and the second analysis result into a frequency domain feature set, and to filter the frequency domain feature set by the feature importance scoring method to obtain the target feature set.

[0116] Specifically, the analysis results are transformed into a frequency domain feature set, and then a feature importance scoring method is applied to filter these features, extracting a target feature set with high classification efficiency. This step aims to reduce redundant information and improve the accuracy of subsequent classification tasks through scientific feature selection methods. The filtered target feature set contains the information most influential on the classification task, significantly improving classification accuracy and efficiency.

[0117] The classification unit 60 is used to identify and classify the EEG signal based on the target feature set.

[0118] This embodiment provides a device for extracting EEG signal features based on frequency domain analysis. The device includes: a first acquisition unit, a first calculation unit, a second calculation unit, a third calculation unit, a fourth calculation unit, and a classification unit. The first calculation unit extracts the energy distribution of a second target signal in the frequency-time plane to determine the dynamic characteristics of the second target signal, obtaining first distribution data. The second calculation unit performs frequency domain coherence analysis on different second target signals based on the first distribution data using coherence spectral analysis technology to predict functional connectivity between different brain regions, obtaining a first analysis result. The third calculation unit performs nonlinear feature extraction in the frequency domain on the second target signal based on the first distribution data and nonlinear dynamics theory, obtaining a second analysis result. The fourth calculation unit converts the first and second analysis results into a frequency domain feature set, filters the frequency domain feature set using a feature importance scoring method, and obtains a target feature set. The classification unit identifies and classifies the EEG signal based on the target feature set. This application addresses the problems in existing technologies, such as insufficient handling of the non-stationarity of EEG signals, neglect of time-varying characteristics, and low accuracy and computational efficiency in feature extraction, by optimizing the frequency domain analysis algorithm and introducing dynamic filtering, frequency domain coherence analysis, and nonlinear dynamics theory.

[0119] As one possible implementation, the first acquisition unit includes: a first acquisition module, a second acquisition module, a first analysis module, an establishment module, and a second analysis module.

[0120] The first acquisition module is used to acquire historical data of EEG signals from different target objects through electrode pads to obtain the third target signal;

[0121] The second acquisition module is used to acquire physiological parameter data of the target object to obtain the first target data, acquire ambient sound through a microphone to obtain ambient noise data, and acquire vibration data through an accelerometer to obtain the second target data.

[0122] Specifically, electrode pads can precisely acquire the electrical activity of the brain, providing information on neural activity in various brain regions as a basis for subsequent analysis. Voltage time series and physiological parameter data (such as heart rate and body temperature) are recorded using scalp electrodes on different target subjects, while environmental sound (such as background noise) and vibration data are acquired via microphones and accelerometers. This data, as supplementary information, provides additional background information about the target subject and its surrounding environment, making the analysis of EEG signals more accurate and reliable, especially in noisy environments.

[0123] The first analysis module is used to perform cluster analysis based on the third target signal, the first target data, the environmental noise data, and the second target data to obtain different categories of EEG signals and obtain multiple EEG signal groups.

[0124] Specifically, a feature model is established for each target object through preliminary EEG signal analysis. Cluster analysis is then used to comprehensively analyze the third target signal (historical EEG signal data), the first target data (physiological parameter data), environmental noise data, and the second target data (vibration data) to obtain different categories of EEG signals. Especially when the target object is in different physiological states or environmental conditions, cluster analysis helps to clearly distinguish the different categories of EEG signals, thus providing an accurate signal classification basis for subsequent filtering processing.

[0125] A module is established to create feature models for each EEG signal group and initialize filter parameters for different categories based on the feature models.

[0126] Specifically, feature models help describe and summarize the typical characteristics of each signal category, and the initialization of filter parameters is adjusted based on these feature models. Establishing feature models for each EEG signal group (each category) accurately captures the unique attributes of each type of EEG signal, thereby optimizing filter parameter settings. Through proper filter parameter initialization, targeted processing for each signal category can be ensured, minimizing noise interference and improving filtering effectiveness.

[0127] Let Θ i The filter parameter set for the i-th type of target object is initialized as: Θ i ={θ i1 ,θ i2 ,…,θ in}; where θ ij It is the j-th filter parameter for the i-th type of target object.

[0128] The system monitors the characteristic changes of EEG signals in real time and dynamically adjusts the filtering parameters based on the real-time tracked characteristics to adapt to the current state of the target object.

[0129] The second analysis module is used to perform cluster analysis on the first target signal to obtain the target category corresponding to the first target signal, determine the corresponding target filtering parameters based on the target category, and filter the first target signal based on the target filtering parameters to obtain the second target signal.

[0130] Specifically, cluster analysis is performed on the first target signal to determine its target category. Based on this category, corresponding target filtering parameters are selected to filter the first target signal, resulting in the second target signal. The filtered signal removes unnecessary noise and retains the EEG signals valuable for analysis. By associating EEG signals with specific categories and applying appropriate filtering parameters, more precise noise suppression and signal enhancement can be achieved.

[0131] As one possible implementation, the first calculation unit includes: a standard deviation module, a windowing module, and a transformation module.

[0132] The standard deviation module is used to determine the standard deviation of the second target signal based on the second target signal, and to determine the scale parameters of the multi-scale analysis window based on the standard deviation. Where λ is the scale parameter, σ(t) is the standard deviation, and ∈ is a value used to avoid division by zero;

[0133] The windowing module is used to window the second target signal based on the scale parameter to obtain the third target signal;

[0134] Specifically, extracting the energy distribution of the signal in the frequency-time plane involves automatically selecting an appropriate multi-scale analysis window based on the waveform characteristics of the EEG signal to capture frequency components at different scales. Thus, the scale-adaptive windowing method ensures more accurate local analysis of the signal across different frequency and time ranges, thereby optimizing subsequent frequency domain analysis and feature extraction.

[0135] The transform module performs adaptive recursive wavelet transform based on the third target signal to extract the instantaneous frequency and amplitude of the third target signal, obtaining the first distribution data: E(f,t,λ)=|ARWT(x w (k),λ,t)| 2 ;

[0136] Where E(f,t,λ) is the first distribution data, used to characterize the three-dimensional distribution of the third target signal in energy-frequency-time, where ARWT(x w (k),λ,t) satisfy:

[0137]

[0138] Where, x w (k) represents the third target signal. ψ is a parameter used to limit the range of values ​​for the wavelet basis functions. λ (t) is the adaptive phase function, and j is the imaginary unit.

[0139] Specifically, an adaptive complex modulation wavelet transform is applied to each windowed signal segment to extract the instantaneous frequency and amplitude of the signal, and to obtain the three-dimensional energy-frequency-time distribution E(f,t,λ). The adaptive recursive wavelet transform can provide high time-frequency resolution and accurately capture the changes in the instantaneous frequency and amplitude of the signal.

[0140] As one possible implementation, the second computing unit includes: a first computing module, a second computing module, and a first result module.

[0141] The first calculation module is used to perform coherence calculation based on the second target signal from different channels and the first distribution data corresponding to the second target signal, to obtain the target coherence:

[0142] Among them, C ij (f,t) represents the target coherence, S ij (f,t) represents the cross-spectral density between the first distribution data corresponding to the second target signals of channels i and j.

[0143] Among them, S ij (f,t) satisfies: in, For X i(f,t,λ) The conjugate of X i(f,t,λ) This refers to the data extracted from the first distribution data using a time window, where T is the length of the time window.

[0144] Specifically, coherence calculations are performed based on the second target signals from different channels and their corresponding first distribution data. Coherence measures the linear relationship between signals. Through coherence calculations, the degree of functional connectivity between different brain regions can be effectively assessed, revealing the collaborative working patterns of different regions in EEG signals. Frequency domain coherence analysis can provide more accurate functional connectivity information than time domain analysis, helping to study the dynamic changes in brain networks.

[0145] The second computational module is used to calculate a weighted coherence index based on target coherence, to characterize the degree of functional connectivity between different brain regions:

[0146] Among them, Γ ij (f) is the weighted coherence index, and η(f) is the frequency energy weighting factor, which is used to enhance the coherence of a specific frequency band;

[0147] Specifically, a weighted coherence index is calculated based on target coherence to further quantify the strength of functional connectivity between different brain regions. η(f) helps to emphasize coherence in specific frequency bands, particularly in certain bands such as alpha, beta, or theta waves, which are particularly important for the functional connections of brain activity. The weighted coherence index, through frequency weighting, can focus on specific frequency bands in the brain, further improving the ability to detect important areas of brain electrical activity. This method not only quantifies the functional connectivity between brain regions but also reveals the contribution of specific frequency bands to functional connectivity, contributing to a deeper understanding of the impact of different frequency bands on brain function patterns.

[0148] The first result module is used to determine the weighted coherence index and the target coherence as the first analysis result.

[0149] Specifically, by combining target coherence and weighted coherence indices, richer and more precise analyses of brain functional connectivity can be provided. This comprehensive analysis not only focuses on overall coherence but also emphasizes the role of frequency, thus offering profound insights into the interactions and collaborative work between different brain regions.

[0150] As one possible implementation, the third computational unit includes: a fractal dimension module, a complexity module, and a second result module.

[0151] The fractal dimension module is used to determine the number of boxes based on the first distribution data, and to determine the fractal dimension based on the number of boxes: Where D is the fractal dimension, and N(∈) is the number of boxes determined based on the signal energy distribution in the first distribution data;

[0152] Specifically, the box counting method is used to calculate the fractal dimension D. By changing the scale of the boxes, the self-similarity of the signal at different scales is analyzed, and N(∈) is calculated by selectively covering points in the high-energy frequency band. By calculating the fractal dimension, the complexity and nonlinear characteristics of the signal can be quantified.

[0153] The complexity module is used to determine the complexity index of the second target signal based on the fractal dimension and the time-series entropy of the second target signal: Ξ ′ =D·H(X);

[0154] Among them, Ξ ′ H(X) is the time series entropy, representing the degree of uncertainty or disorder in the signal. Higher entropy indicates a more irregular and unpredictable signal, reflecting its more complex dynamic characteristics.

[0155] Specifically, by incorporating a combination of time-series entropy and fractal dimension, the complexity index can more comprehensively assess the nonlinear characteristics and dynamic changes of EEG signals. A higher complexity index indicates a more complex dynamic structure of the EEG signal, reflecting more complex interaction patterns between brain regions.

[0156] The second results module is used to determine the complexity index as the second analysis result.

[0157] Specifically, the complexity index, as a secondary analytical result, provides high-dimensional features for further analysis of brain signals, helping to deepen the understanding of the nonlinear dynamic characteristics of the signals. This analytical result not only improves the accuracy of signal classification but also reveals the functional complexity of different brain regions.

[0158] As one possible implementation, the fourth computing unit includes a conversion module.

[0159] The conversion module is used to convert the first analysis result and the second analysis result into a set of feature vectors to obtain a frequency domain feature set.

[0160] Specifically, the weighted coherence index reflects the degree of functional connectivity between different brain regions. Its value can be viewed as one dimension of a multidimensional feature vector, representing the interaction between brain regions at different frequency bands. The complexity index reflects the dynamic complexity of brain signals, and its value can also be used to construct a feature vector. By transforming these two analytical results into a unified set of frequency domain features in the form of feature vectors, they can be used for subsequent signal recognition and classification.

[0161] As one possible implementation, the fourth calculation unit also includes: an initialization module, a scoring module, a sorting module, and an iteration module.

[0162] The initial module is used to take a portion of the data in the frequency domain feature set as the initial state of the target feature set;

[0163] Specifically, the extracted frequency domain features are represented as a set of feature vectors, F = {f1, f2, ..., f n}; where f i It is the i-th feature vector. This set contains multiple feature vectors extracted from frequency domain analysis, which represent multi-dimensional information of the EEG signal.

[0164] The scoring module is used to calculate the target score R(f) based on the feature importance scoring method for feature vectors in the frequency domain feature set using a scoring function: i )=α·Corr(f i ,y)-β·Redun(f i ,S);

[0165] This involves initializing a target feature subset S, using a feature importance scoring method, and defining a scoring function R(f i ) is the target score, Corr(f) i ,y) is the eigenvector f i The correlation with the target variable y reflects the impact of this feature on the final identification result. Redun(f i S) is the eigenvector f i The redundancy of the current state of the target feature set S is intended to remove redundant and highly similar features, thereby reducing the complexity of the feature set. α is a weighting factor for balancing correlation, and β is a weighting factor for balancing redundancy.

[0166] The sorting module is used to sort based on the target score and update the target feature set with a preset number of feature vectors.

[0167] Specifically, the feature vectors are sorted based on the calculated target scores, and the feature vector with the highest score is selected to update the target feature set.

[0168] The iteration module is used to recalculate the target score based on the updated target feature set, and update the target feature combination according to the recalculated target score, until the difference of the target score between two adjacent iterations of the same feature vector is less than the second threshold or the maximum number of iterations is reached, and then outputs the target feature set.

[0169] Specifically, the target score is recalculated based on the updated target feature set, and the target feature set is adjusted in the next iteration. Through multiple iterations, the algorithm continuously adjusts the feature subset, ensuring the gradual improvement of the feature subset until the optimal or near-optimal feature subset is found.

[0170] The aforementioned frequency domain analysis-based EEG signal feature extraction device includes a processor and a memory. The first acquisition unit, first calculation unit, second calculation unit, third calculation unit, fourth calculation unit, and classification unit are all stored as program units in the memory. The processor executes these program units stored in the memory to achieve the corresponding functions. All of the above modules are located in the same processor; alternatively, the modules may be located in different processors in any combination.

[0171] The processor contains a kernel, which retrieves the corresponding program units from memory. One or more kernels can be configured, and communication efficiency can be improved by adjusting kernel parameters.

[0172] The memory may include non-permanent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0173] This invention provides a computer-readable storage medium including a stored program, wherein, when the program is executed, it controls the device containing the computer-readable storage medium to perform the frequency domain analysis-based EEG signal feature extraction method.

[0174] This application also provides an electroencephalogram (EEG) analysis system, comprising: one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include methods for performing the above-described frequency domain analysis-based EEG signal feature extraction method.

[0175] It is obvious to those skilled in the art that the modules or steps of the present invention described above can be implemented using general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. They can be implemented using computer-executable program code, and thus can be stored in a storage device for execution by a computing device. In some cases, the steps shown or described can be performed in a different order than those described herein, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. Thus, the present invention is not limited to any particular combination of hardware and software.

[0176] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0177] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0178] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0179] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0180] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0181] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0182] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0183] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0184] As can be seen from the above description, the embodiments of this application achieve the following technical effects:

[0185] 1) The frequency domain analysis-based EEG signal feature extraction method of this application collects signal data, optimizes filtering based on the characteristics of EEG signals of different individuals and noise environment through real-time learning and adjustment of filtering parameters; extracts the energy distribution of the signal in the frequency-time plane to reveal the dynamic change characteristics of the signal; uses coherence spectrum analysis technology to identify the frequency domain coherence between different EEG signal channels and evaluate the functional connectivity between brain regions; combines nonlinear dynamics theory to extract the nonlinear characteristics of the frequency domain signal and analyze the complexity and chaotic nature of the EEG signal; optimizes the selection of extracted frequency domain features to identify the most contributing feature subset; and outputs the optimized frequency domain feature set for the classification, recognition and other advanced analysis of EEG signals. This application optimizes the frequency domain analysis algorithm by introducing dynamic filtering, frequency domain coherence analysis, and nonlinear dynamics theory. Dynamic filtering overcomes the shortcomings of traditional methods that may lose key information when processing non-stationary signals. Frequency domain coherence analysis and nonlinear feature extraction deeply explore the time-varying characteristics and nonlinear laws of the signal, while the feature importance scoring rule improves the efficiency and discrimination ability of feature selection, thereby comprehensively improving the accuracy and efficiency of EEG signal analysis and providing more reliable technical support for EEG signal research and application. Compared with existing technologies, it may have significant advantages in processing non-stationary signals, extracting time-varying features, and reducing computational costs. It solves the problems of insufficient handling of non-stationarity of EEG signals, neglect of time-varying characteristics, and low accuracy and computational efficiency of feature extraction in existing technologies.

[0186] 2) The EEG signal feature extraction device based on frequency domain analysis of this application. The device includes: a first acquisition unit, a first calculation unit, a second calculation unit, a third calculation unit, a fourth calculation unit, and a classification unit. The first calculation unit is used to extract the energy distribution of a second target signal in the frequency-time plane to determine the dynamic change characteristics of the second target signal, obtaining first distribution data; the second calculation unit is used to perform frequency domain coherence analysis on different second target signals based on the first distribution data using coherence spectral analysis technology to predict functional connectivity between different brain regions, obtaining a first analysis result; the third calculation unit is used to perform nonlinear feature extraction in the frequency domain of the second target signal based on the first distribution data and nonlinear dynamics theory, obtaining a second analysis result; the fourth calculation unit is used to convert the first and second analysis results into a frequency domain feature set, and filter the frequency domain feature set using a feature importance scoring method to obtain a target feature set; the classification unit is used to identify and classify the EEG signal based on the target feature set. This application addresses the problems in existing technologies, such as insufficient handling of the non-stationarity of EEG signals, neglect of time-varying characteristics, and low accuracy and computational efficiency in feature extraction, by optimizing the frequency domain analysis algorithm and introducing dynamic filtering, frequency domain coherence analysis, and nonlinear dynamics theory.

[0187] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for electroencephalogram signal analysis based on frequency domain analysis, characterized in that, The method comprises the following steps: acquiring an electroencephalogram signal to obtain a first target signal, performing dynamic filtering based on signal characteristics of the first target signal and a noise environment to obtain a second target signal; extracting energy distribution of the second target signal on a frequency-time plane to determine dynamic change characteristics of the second target signal to obtain first distribution data; performing frequency domain coherence analysis on different second target signals based on the first distribution data by using a coherence spectrum analysis technique to predict functional connections between different regions of the brain to obtain a first analysis result; extracting nonlinear characteristics of the second target signal in the frequency domain based on the first distribution data and a nonlinear dynamics theory to obtain a second analysis result; converting the first analysis result and the second analysis result into a frequency domain feature set, screening the frequency domain feature set by using a feature importance scoring method to obtain a target feature set; recognizing and classifying the electroencephalogram signal based on the target feature set; acquiring an electroencephalogram signal to obtain a first target signal, performing dynamic filtering based on signal characteristics of the first target signal and a noise environment to obtain a second target signal, comprising: collecting historical data of the electroencephalogram signal of different target objects by using an electrode sheet to obtain third target signals; collecting physiological parameter data of a target object to obtain first target data, collecting environmental sound by using a microphone to obtain environmental noise data, and collecting vibration data by using an accelerometer to obtain second target data; performing cluster analysis based on the third target signals, the first target data, the environmental noise data, and the second target data to obtain different categories of the electroencephalogram signal to obtain a plurality of electroencephalogram signal groups; establishing a feature model for each of the electroencephalogram signal groups, and initializing filter parameters corresponding to different categories based on the feature model; performing cluster analysis on the first target signal to obtain a target category corresponding to the first target signal, determining target filter parameters corresponding to the target category, and filtering the first target signal based on the target filter parameters to obtain the second target signal; extracting energy distribution of the second target signal on a frequency-time plane to determine dynamic change characteristics of the second target signal to obtain first distribution data, comprising: based on the second target signal, determining a standard deviation of the second target signal, and determining a scale parameter of a multi-scale analysis window based on the standard deviation: ； wherein, is the scale parameter, is the standard deviation, is a numerical value for avoiding division by zero; based on the scale parameter, windowing the second target signal to obtain third target signals; based on the third target signals, performing adaptive recursive wavelet transform to extract instantaneous frequency and amplitude of the third target signals to obtain the first distribution data: ； wherein is the first distribution data for characterizing the three-dimensional distribution of the third target signal in energy-frequency-time, wherein satisfies: ； wherein, is the third target signal, is a parameter for limiting the value range of the wavelet basis function, is a wavelet basis function, is an adaptive phase function, is an imaginary unit.

2. The method of claim 1, wherein, performing frequency domain coherence analysis on different second target signals based on the first distribution data by using a coherence spectrum analysis technique to predict functional connections between different regions of the brain to obtain a first analysis result, comprising: based on the second target signals of different channels and the first distribution data corresponding to the second target signals, performing coherence calculation to obtain target coherence: ； wherein, is the second target signal for the target coherence, is the cross-spectral density between the first distribution data corresponding to the second target signal pair for channels i and j, wherein, satisfies: ； wherein, is a conjugate of, is the data in the first distribution data intercepted with a time window, and T is the length of the time window. Calculate a weighted coherence index based on the target coherence to represent the degree of functional connection between different regions of the brain: ； wherein, is the weighted coherence index, is a frequency energy weighting factor for enhancing the coherence of a specific frequency band; Determine the weighted coherence index and the target coherence as the first analysis result.

3. The method of claim 2, wherein, Based on the first distribution data, nonlinear dynamics theory is used to extract nonlinear features in the frequency domain of the second target signal to obtain a second analysis result, including: Determine the number of boxes based on the first distribution data, and determine the fractal dimension based on the number of boxes: ； wherein D is the fractal dimension, is the number of boxes determined from the signal energy distribution in the first distribution data; Determine the complexity index of the second target signal based on the fractal dimension and the time series entropy of the second target signal: ； wherein, is the complexity index, is the time series entropy; Determine the complexity index as the second analysis result.

4. The method of claim 3, wherein, Convert the first analysis result and the second analysis result into a frequency domain feature set, including: Convert the first analysis result and the second analysis result into a feature vector set to obtain the frequency domain feature set.

5. The method of claim 4, wherein, Filter the frequency domain feature set by feature importance scoring method to obtain a target feature set, including: Part of the data in the frequency domain feature set is used as the initial state of the target feature set; Based on the feature importance scoring method, the feature vectors in the frequency domain feature set are calculated based on the scoring function to obtain a target score: ； wherein, is the target score, is the feature vector is the correlation with the target variable y, is the feature vector is the redundancy with the current state of the target feature set S, is the balancing correlation weight factor, is the balancing redundancy weight factor; Sort based on the target score, and update the target feature set with the top pre-set number of feature vectors; Recalculate the target score based on the updated target feature set, and update the target feature set according to the recalculated target score until the difference between the target scores of the same feature vector in adjacent two iterations is less than a second threshold or the maximum iteration number is reached, and output the target feature set.

6. An electroencephalogram signal feature extraction device based on frequency domain analysis, characterized by, The device comprises: A first acquisition unit is configured to acquire an electroencephalogram signal to obtain a first target signal, perform dynamic filtering based on signal characteristics and noise environment of the first target signal, and obtain a second target signal; A first calculation unit is configured to extract energy distribution of the second target signal in a frequency-time plane to determine dynamic change characteristics of the second target signal, and obtain first distribution data; A second calculation unit is configured to perform frequency domain coherence analysis on different second target signals based on the first distribution data by coherence spectrum analysis technology to predict functional connection between different regions of the brain, and obtain a first analysis result; A third calculation unit is configured to perform frequency domain nonlinear feature extraction on the second target signal based on the first distribution data combined with nonlinear dynamics theory, and obtain a second analysis result; A fourth calculation unit is configured to convert the first analysis result and the second analysis result into a frequency domain feature set, filter the frequency domain feature set by feature importance scoring method, and obtain a target feature set; A classification unit is configured to identify and classify the electroencephalogram signal based on the target feature set; The first acquisition unit comprises: A first acquisition module is configured to acquire electroencephalogram historical data of different target objects by an electrode sheet to obtain a third target signal; The second acquisition module is configured to acquire physiological parameter data of a target object to obtain first target data, acquire environmental sound through a microphone to obtain environmental noise data, and acquire vibration data through an accelerometer to obtain second target data. The first analysis module is configured to perform clustering analysis based on the third target signal, the first target data, the environmental noise data, and the second target data to obtain different categories of the electroencephalogram signals and obtain a plurality of electroencephalogram groups. The establishment module is configured to establish a feature model for each of the electroencephalogram groups and initialize filter parameters corresponding to different categories based on the feature model. The second analysis module is configured to perform clustering analysis on the first target signal to obtain a target category corresponding to the first target signal, determine target filter parameters corresponding to the target category based on the target category, filter the first target signal based on the target filter parameters, and obtain the second target signal. The first calculation unit includes: The standard deviation module is configured to determine a standard deviation of the second target signal based on the second target signal and determine a scale parameter of a multi-scale analysis window based on the standard deviation. ； wherein, is the scale parameter, is the standard deviation, is a numerical value for avoiding division by zero; The windowing module is configured to window the second target signal based on the scale parameter to obtain a third target signal. The transformation module is configured to perform adaptive recursive wavelet transformation based on the third target signal to extract an instantaneous frequency and an amplitude of the third target signal and obtain the first distribution data. ； wherein is the first distribution data for characterizing the three-dimensional distribution of the third target signal in energy-frequency-time, wherein satisfies: ； wherein, is the third target signal, is a parameter for limiting the value range of the wavelet basis function, is a wavelet basis function, is an adaptive phase function, is an imaginary unit.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored program, wherein the program controls a device in which the computer-readable storage medium is located to perform the method of any one of claims 1 to 5 when the program is executed.

8. An electroencephalogram analysis system characterized by comprising: The device includes one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include a program for performing the method of any one of claims 1 to 5. ​

Images