A spontaneous electroencephalogram data processing method, system, device and storage medium

Abstract

The application relates to the technical field of brain-computer interfaces, and specifically discloses a spontaneous electroencephalogram data processing method, system, device and storage medium. The method comprises the following steps: acquiring spontaneous electroencephalogram data of a target user in a target task state; pre-processing the spontaneous electroencephalogram data of the target user to obtain target data; the pre-processing at least comprises band-pass filtering and noise channel removal; when the number of residual channels corresponding to the target data is greater than 1, based on a preset frequency band corresponding to the target task state, a typical correlation analysis algorithm is used to perform enhancement processing on the target data to obtain a target enhancement signal; feature extraction and state classification are performed on the target enhancement signal to obtain a target classification result. According to the scheme, the preset frequency band obtained based on historical user frequency band evaluation is used to perform CCA enhancement on the pre-processed multi-channel spontaneous electroencephalogram, the signal quality and state classification accuracy can be improved under a low signal-to-noise ratio condition, and the scheme has the advantages of strong cross-user generalization capability, high calculation efficiency and good interpretability.

Description

Technical Field

[0001] This invention relates to the field of brain-computer interface technology, and in particular to a method, system, device, and storage medium for processing spontaneous EEG data. Background Technology

[0002] Current passive brain-computer interfaces often use a few leads (e.g., 1-2 leads) of dry electrodes to collect spontaneous EEG signals, resulting in low signal-to-noise ratio (SNR) and difficulties in cross-user transfer. To address the low SNR issue, existing signal enhancement methods such as wavelet transform, empirical mode decomposition, and deep learning methods either limit themselves to single-lead processing and fail to fully utilize the spatial correlation between leads, leading to poor performance at low SNR; or they require large amounts of labeled training data with limited algorithmic improvement, making it difficult to effectively improve signal quality in practical applications with few leads and low SNR, thus affecting the accuracy of subsequent state classification. Summary of the Invention

[0003] This invention provides a method, system, device, and storage medium for processing spontaneous EEG data, in order to solve the problem that existing spontaneous EEG data processing methods suffer from poor state classification accuracy due to low signal-to-noise ratio and failure to fully utilize the spatial relationship between leads.

[0004] According to one aspect of the present invention, a method for processing spontaneous electroencephalogram (EEG) data is provided, the method comprising: Acquire spontaneous EEG data of the target user in the target task state; The spontaneous EEG data of the target user is preprocessed to obtain the target data; the preprocessing includes at least bandpass filtering and noise lead removal. When the number of remaining leads corresponding to the target data is greater than 1, the target data is enhanced using a typical correlation analysis algorithm based on the preset frequency band corresponding to the target task state to obtain the target enhanced signal; wherein, the preset frequency band is obtained by frequency band evaluation based on the spontaneous EEG data of multiple historical users in the target task state; Feature extraction and state classification are performed on the target enhancement signal to obtain the target classification result.

[0005] According to another aspect of the present invention, a spontaneous EEG data processing system is provided, the system comprising: The data acquisition module is used to acquire spontaneous EEG data of the target user in the target task state; The preprocessing module is used to preprocess the spontaneous EEG data of the target user to obtain the target data; the preprocessing includes at least bandpass filtering and noise lead removal; The signal enhancement module is used to enhance the target data based on the preset frequency band corresponding to the target task state and use the typical correlation analysis algorithm to obtain the target enhanced signal when the number of remaining leads corresponding to the target data is greater than 1. The preset frequency band is obtained by frequency band evaluation based on the spontaneous EEG data of multiple historical users in the target task state. The classification module is used to extract features and classify the state of the target enhancement signal to obtain the target classification result.

[0006] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the spontaneous EEG data processing method according to any embodiment of the present invention.

[0007] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the spontaneous EEG data processing method according to any embodiment of the present invention.

[0008] According to another aspect of the present invention, a computer program product is provided, the computer program product comprising a computer program that, when executed by a processor, implements the spontaneous EEG data processing method according to any embodiment of the present invention.

[0009] The technical solution of this invention involves acquiring spontaneous EEG data of a target user under a target task state; preprocessing the spontaneous EEG data of the target user to obtain target data; the preprocessing includes at least bandpass filtering and noise lead removal; when the number of remaining leads corresponding to the target data is greater than 1, the target data is enhanced using a canonical correlation analysis (CCA) algorithm based on a preset frequency band corresponding to the target task state to obtain a target enhanced signal; wherein, the preset frequency band is obtained by frequency band evaluation based on spontaneous EEG data of multiple historical users under the target task state; feature extraction and state classification are performed on the target enhanced signal to obtain the target classification result. This solution, by preprocessing the spontaneous EEG data and using a preset frequency band obtained from frequency band evaluation based on historical users for CCA signal enhancement, can effectively extract task-related spatial filtering components from multi-lead signals, significantly improve signal quality under low signal-to-noise ratio conditions, and thus improve the accuracy of state classification. Simultaneously, this solution does not require the collection of additional training data for the target user, supports online real-time computation, and has the advantages of strong cross-user generalization ability, high computational efficiency, and good interpretability.

[0010] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 This is a flowchart of a spontaneous EEG data processing method provided in an embodiment of the present invention; Figure 2 This is a flowchart of another spontaneous EEG data processing method provided in an embodiment of the present invention; Figure 3 This is a schematic diagram illustrating the comparison of classification results provided in an embodiment of the present invention; Figure 4 This is another classification result comparison diagram provided by an embodiment of the present invention; Figure 5 This is another classification result comparison diagram provided by an embodiment of the present invention; Figure 6 This is another classification result comparison diagram provided by an embodiment of the present invention; Figure 7This is a schematic diagram of the structure of a spontaneous EEG data processing system provided in an embodiment of the present invention; Figure 8 This is a schematic diagram of the structure of an electronic device that implements the spontaneous EEG data processing method of the present invention. Detailed Implementation

[0013] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0014] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention 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 so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or 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.

[0015] Figure 1 This is a flowchart illustrating a spontaneous EEG data processing method provided in an embodiment of the present invention. This embodiment is applicable to situations where spontaneous EEG signals with low signal-to-noise ratios are enhanced to achieve user state classification. This method can be executed by a spontaneous EEG data processing system, which can be implemented in hardware and / or software and can be configured in an electronic device. Figure 1 As shown in the figure, the spontaneous EEG data processing method provided in this embodiment specifically includes the following steps: S110. Obtain spontaneous EEG data of the target user in the target task state.

[0016] In this context, the target user refers to an individual currently using the brain-computer interface system for state monitoring or interaction. The target task state can refer to a specific task that the user needs to perform or a specific physiological / psychological state that they need to maintain during the experiment, such as an open-eye state, a closed-eye state, a state of focused attention, a relaxed state, or a sleep state. This state corresponds to a specific EEG pattern or frequency band characteristic.

[0017] Spontaneous EEG data refers to EEG signals generated spontaneously by neuronal groups in the cerebral cortex without the need for specific external stimuli. These signals include rhythmic components such as alpha and beta waves, and differ from event-related potentials (ERPs) and are characterized by an inherently low signal-to-noise ratio. It should be noted that, unless otherwise specified, the EEG data used in this embodiment refers to spontaneous EEG data.

[0018] In this embodiment of the invention, after the target user wears a multi-lead EEG acquisition device, spontaneous EEG data of the user during a period of time T under the target task state can be collected. The dimension can be C×N, where C is the number of leads. The number of sampling points. Where is the sampling frequency and T is the sampling time.

[0019] S120. Preprocess the spontaneous EEG data of the target user to obtain the target data; the preprocessing includes at least bandpass filtering and noise lead removal.

[0020] Bandpass filtering refers to a filtering method that allows signals within a specific frequency range to pass through while suppressing frequency components outside that range. It is used to remove low-frequency baseline drift and high-frequency electromagnetic interference from EEG data. Noise lead removal refers to the process of automatically identifying and eliminating lead channels that are of poor quality or completely ineffective due to poor electrode contact, amplifier saturation, or other reasons, by analyzing the statistical or temporal characteristics of each lead signal. Target data refers to the clean EEG data obtained after preprocessing, which has removed significant noise and invalid leads; it serves as the input data for subsequent signal enhancement and analysis.

[0021] In this embodiment of the invention, the collected raw spontaneous EEG data can be processed. Preprocessing operations, including but not limited to bandpass filtering and noise lead removal, are performed to obtain clean target data X. Specifically, the bandpass filtering process may include: selecting an appropriate digital filter type based on application requirements and computational resource constraints, such as, but not limited to, Butterworth filters, Chebyshev filters, elliptic filters, and Finite Impulse Response (FIR) filters; then, determining the passband cutoff frequency based on the EEG rhythms of interest to the target task, and simultaneously setting filter parameters such as the stopband cutoff frequency, maximum passband attenuation, and minimum stopband attenuation. For example, for EEG signals containing alpha waves (8-12Hz), the passband cutoff frequency can be set to 0.5Hz and 30Hz, and the stopband cutoff frequency can be set to 0.1Hz and 40Hz. Finally, the raw spontaneous EEG data is processed according to the set filter parameters. Filtering is then performed. To avoid introducing phase distortion, a zero-phase filtering method can be used, which involves first performing forward filtering on the data, and then performing reverse filtering on the result.

[0022] The process of noise lead removal may include: calculating the characteristic parameters used to measure the noise level for each lead signal after bandpass filtering, such as including but not limited to the variance, standard deviation, peak-to-peak value, entropy, kurtosis, etc.; then, comparing the noise characteristics of each lead signal with the corresponding preset threshold Th; if the noise characteristics of a lead exceed the threshold Th, the lead is determined to be a noise lead, and the corresponding lead signal is removed from the original EEG data, thereby obtaining clean target data X.

[0023] S130. When the number of remaining leads corresponding to the target data is greater than 1, the target data is enhanced using a typical correlation analysis algorithm based on the preset frequency band corresponding to the target task state to obtain the target enhanced signal.

[0024] The remaining lead count refers to the number of valid EEG leads that remain after noise lead removal. The preset frequency band is one or more frequency ranges obtained by frequency band evaluation based on spontaneous EEG data from multiple historical users in the target task state, and is used to transfer the data to the target user.

[0025] Canonical Correlation Analysis (CCA) is a multivariate statistical analysis method used to find the linear projection direction between two sets of variables, maximizing the correlation between the two projected variables. In this scheme, the CCA algorithm is used to calculate the spatial filtering coefficients between the target data X and the reference signal Y to achieve signal enhancement.

[0026] In this embodiment of the invention, for the preprocessed target data X, the number of valid leads that are not identified as noise, i.e., the number of remaining leads, can be counted. ,like Then, based on the preset frequency band corresponding to the target task state, CCA enhancement processing is performed on the target data; if If the data processing flow for the current sampling time T is skipped, the EEG data for the next sampling time T will be collected and processed. The CCA enhancement process may include: acquiring a pre-configured frequency band for the target task state, which can be obtained by evaluating the frequency band of spontaneous EEG data from multiple historical users in the target task state; constructing a corresponding reference signal Y based on the acquired pre-configured frequency band, and using the CCA algorithm to analyze the linear correlation between the target data X and the reference signal Y to obtain spatial filtering coefficients that maximize the correlation between the two; then using these spatial filtering coefficients to linearly weight and combine the target data to obtain the target enhancement signal, thereby significantly increasing the signal energy within the pre-configured frequency band while effectively suppressing irrelevant noise components and greatly improving the signal-to-noise ratio.

[0027] S140. Perform feature extraction and state classification on the target enhancement signal to obtain the target classification result.

[0028] Feature extraction refers to the process of extracting information from the enhanced signal that reflects the essential characteristics of the signal and is effective for subsequent classification tasks. Its purpose is to reduce data dimensionality, remove redundant information, and improve classification accuracy and efficiency. State classification refers to the process of dividing the target user's EEG data into different task state categories based on the extracted feature vectors and using a pre-trained classification model.

[0029] The target classification result can refer to the final output state classification result, which reflects the psychological or physiological state of the target user at the current moment. The target classification result can be a category label, probability value, or other forms of representation.

[0030] In this embodiment of the invention, the corresponding feature vector can be extracted from the target enhancement signal by means of methods not limited to frequency domain transformation and deep learning, and then the target classification result of the current state of the target user can be output through a preset classification model. In one embodiment, the classification model may include, but is not limited to: linear classifiers (such as logistic regression, linear discriminant analysis), nonlinear classifiers (such as support vector machines, random forests, gradient boosting trees), neural networks, etc.

[0031] The spontaneous EEG data processing method provided in this invention preprocesses the spontaneous EEG data of the target user and, when the number of remaining leads is greater than 1, uses a preset frequency band obtained based on historical user group frequency band evaluation to perform CCA signal enhancement. This effectively mines the spatial correlation between multiple leads, significantly improves the signal-to-noise ratio of the preset frequency band signal under low signal-to-noise ratio conditions, and thus improves the accuracy of state classification. At the same time, this method does not require the collection of additional training data for the target user, supports online real-time calculation, and has the advantages of strong cross-user generalization ability, high computational efficiency, and good interpretability.

[0032] Furthermore, based on the above embodiments of the invention, when the number of remaining leads corresponding to the target data is greater than 1, the target data is enhanced using a typical correlation analysis algorithm based on a preset frequency band corresponding to the target task state to obtain a target enhancement signal, including: Obtain the remaining number of leads for the target data; When it is determined that the number of remaining leads is greater than 1, a corresponding reference signal is constructed according to the preset frequency band. Based on the target data and reference signal, the corresponding spatial filtering coefficients are determined using a typical correlation analysis algorithm. The target data is linearly weighted and combined using spatial filtering coefficients to obtain the target enhancement signal.

[0033] The reference signal can refer to an artificially synthesized signal constructed according to a preset frequency band, used to guide the CCA algorithm in extracting the EEG signal components most relevant to the preset frequency band. The spatial filtering coefficients can refer to the projection vectors calculated by the CCA algorithm, whose dimension is the same as the number of leads, and each component represents the weighted weight of the corresponding lead signal.

[0034] In this embodiment of the invention, the process of performing CCA enhancement on the target data specifically includes: S1. Count the number of valid leads from the preprocessed target data, i.e., the number of remaining leads. .

[0035] S2, if If the signal enhancement step is skipped, the EEG data for the next sampling time T will be collected and processed. Then, the following signal enhancement processing flow based on CCA will be executed: ①Based on the acquired preset frequency band Construct a corresponding reference signal. For example, it can be based on a preset frequency band. and sampling rate With a fixed frequency step size If frequency points are selected uniformly within a preset frequency band, then the number of frequency points contained within the preset frequency band is:

[0036] Next, for each frequency point The corresponding sinusoidal basis functions can be constructed. Sum and cosine basis functions Then, by concatenating the sine and cosine basis functions of all frequency points column by column, the corresponding reference signal Y can be obtained, which has a dimension of N×(2M).

[0037] ② Input the target data X and the reference signal Y into the CCA algorithm, and obtain the corresponding spatial filtering coefficients by solving the generalized eigenvalue problem. .

[0038] ③ Based on the calculated spatial filtering coefficients It can be used to perform linear weighted combination of target data X to obtain target enhancement signal. .

[0039] Furthermore, based on the above embodiments of the invention, the corresponding spatial filtering coefficients are determined using a typical correlation analysis algorithm based on the target data and the reference signal, including: Determine the autocovariance matrices of the target data and the reference signal, and the cross-covariance matrix between the target data and the reference signal, respectively; Solving the generalized eigenvalue problem based on the autocovariance matrix and crosscovariance matrix yields multiple canonical correlation coefficients and the eigenvectors corresponding to each canonical correlation coefficient. Select at least one eigenvector corresponding to a canonical correlation coefficient as a spatial filtering coefficient, in descending order of canonical correlation coefficients.

[0040] In this embodiment of the invention, the specific process of solving the spatial filter coefficients using CCA includes: S1. Calculate the autocovariance matrix of the target data X. The autocovariance matrix of the reference signal Y And the cross-covariance matrix between the target data X and the reference signal Y .

[0041] S2. Based on the above covariance matrix, construct the following generalized eigenvalue problem:

[0042] By solving this problem, a set of eigenvalues ​​is obtained. and the eigenvector corresponding to each eigenvalue. Among them, the arithmetic square root of the eigenvalues This is the i-th canonical correlation coefficient, which represents the correlation strength between the projected sequences.

[0043] S3. Sort the canonical correlation coefficients from largest to smallest, and select the eigenvectors corresponding to one or more of the top canonical correlation coefficients as spatial filtering coefficients. If a single eigenvector is selected (e.g., ... If the first k elements are selected, a one-dimensional spatial filter matrix is ​​obtained; if the first k elements are selected (k≥2), a multi-dimensional spatial filter matrix is ​​obtained. .

[0044] Furthermore, based on the above embodiments of the invention, the target data is linearly weighted and combined using spatial filtering coefficients to obtain the target enhancement signal, including: When one spatial filter coefficient is selected, the target data is linearly weighted and combined using the spatial filter coefficient to obtain the target enhancement signal and an intermediate enhancement signal, which is then used as the target enhancement signal. Alternatively, when multiple spatial filter coefficients are selected, the target data is linearly weighted and combined using each spatial filter coefficient to obtain multiple intermediate enhanced signals, and the multi-channel signal composed of multiple intermediate enhanced signals is used as the target enhanced signal.

[0045] In this embodiment of the invention, the process of determining the target enhancement signal based on spatial filtering coefficients and target data includes: S1. When only one spatial filter coefficient is selected, it can be linearly weighted and combined with the target data X to obtain a one-dimensional intermediate enhancement signal, which can then be directly used as the final target enhancement signal. Output.

[0046] S2. When selecting multiple spatial filter coefficients (e.g., k), each spatial filter coefficient can be linearly weighted and combined with the target data matrix X to obtain multiple one-dimensional intermediate enhancement signals. These k intermediate enhancement signals are then stacked row-wise to obtain a k×N dimensional multi-channel signal, and this multi-channel signal matrix is ​​used as the final target enhancement signal. Output.

[0047] This embodiment achieves single-channel or multi-channel enhanced signal output by flexibly controlling the number of spatial filtering coefficients selected. Specifically, when selecting single-channel output, the computational efficiency is high, and the most relevant EEG components to the target task frequency band can be extracted to the greatest extent, significantly improving the signal-to-noise ratio. When selecting multi-channel output, not only can the strongest relevant components be retained, but also the second strongest orthogonal components can be provided, thereby retaining richer spatial pattern information, which is beneficial for subsequent feature extraction and classifiers to obtain higher discrimination accuracy.

[0048] Furthermore, based on the above embodiments of the invention, the target data is linearly weighted and combined using spatial filtering coefficients to obtain the target enhancement signal, including: When multiple spatial filter coefficients are selected, the target data is linearly weighted and combined using each spatial filter coefficient. Multiple intermediate enhancement signals are fused to obtain the target enhancement signal.

[0049] In this embodiment of the invention, the process of determining the target enhancement signal based on spatial filtering coefficients and target data further includes: S1. When multiple spatial filter coefficients are selected (e.g., k), each spatial filter coefficient can be linearly weighted and combined with the target data matrix X to obtain multiple one-dimensional intermediate enhanced signals.

[0050] S2. Merge the obtained k intermediate enhanced signals according to the preset fusion rules to generate the final target enhanced signal. The methods for fusion processing may include, but are not limited to: summation fusion, weighted average, etc.

[0051] This solution fuses the intermediate enhancement signals corresponding to multiple spatial filtering coefficients to output a single-channel enhancement signal with constant dimension. This process eliminates the need for downstream feature extraction and classification modules to adapt to variable channel numbers, thus significantly simplifying system interface design and engineering integration.

[0052] Furthermore, based on the above embodiments of the invention, the process for obtaining the preset frequency band includes: Acquire spontaneous EEG data from multiple historical users in the target task state; The spontaneous EEG data of each historical user were preprocessed to obtain the corresponding historical preprocessed data. For historical preprocessed data with a remaining lead count greater than 1, the corresponding individual frequency bands are determined respectively; The preset frequency band is determined based on the frequency band of each body.

[0053] In this embodiment of the invention, the process of determining the preset frequency band under the target task state through frequency band evaluation specifically includes: S1. Obtain multi-lead spontaneous EEG signals from several (e.g., at least 20) historical users under the same target task state.

[0054] S2. Preprocess the EEG data of each historical user to obtain their corresponding historical preprocessed data; the preprocessing includes at least bandpass filtering and noise lead removal.

[0055] S3. Obtain the remaining lead count corresponding to the historical preprocessed data, and then determine the individual frequency band only for historical users with a remaining lead count greater than 1. The process of determining the individual frequency band may include: (1) Based on the preset wideband search range, sliding window width and sliding step size, generate multiple candidate frequency bands; (2) For each candidate frequency band, the historical preprocessed data is enhanced using the typical correlation analysis algorithm to obtain the corresponding enhanced signal; (3) Determine the enhancement effect evaluation index corresponding to each enhanced signal, and determine the individual frequency band corresponding to the historical user in multiple candidate frequency bands based on the enhancement effect evaluation index.

[0056] Specifically, it can be performed according to a pre-configured wideband search range. Sliding window width W and sliding step size Generate all possible candidate frequency bands ,in The starting frequency is from Increase one by one Until .

[0057] Next, a corresponding reference signal is constructed for each candidate frequency band, and the historical preprocessed data is enhanced using the CCA algorithm to obtain the corresponding enhanced signal. The CCA signal enhancement process can be referred to the above embodiment, and will not be repeated here.

[0058] Finally, based on the obtained enhanced signal, the corresponding enhancement effect evaluation index is determined, and the optimal individual frequency band for the corresponding historical user is determined according to the evaluation index. The individual frequency band can be determined using a single evaluation index, or it can be determined based on the weighted result of multiple evaluation indices; this embodiment does not impose any limitation on this.

[0059] In one embodiment, the evaluation metrics for the enhancement effect may include, but are not limited to: ① Statistical separability, used to measure the separability of two states after enhancement (applicable to binary classification tasks). The larger the value, the more significant the difference between the two states, and the better the enhancement effect. It can be expressed as:

[0060] In the formula, The average power of the enhanced signal in the preset frequency band under state 1; The average power of the enhanced signal within the preset frequency band under state 2; The variance of the power values ​​for each test (or each time period) under state 1; This represents the variance of the power values ​​for each test (or each time period) under state 2.

[0061] ② The maximum canonical correlation coefficient can be obtained directly during the CCA process. The closer the value is to 1, the higher the consistency between the enhanced signal and the reference signal, and the better the enhancement effect.

[0062] ③ Power spectral contrast, used to measure the degree of energy prominence of the enhanced signal relative to adjacent frequency bands within a preset frequency band. The larger the value, the more prominent the rhythm peak and the better the enhancement effect. It can be expressed as:

[0063] In the formula, Preset frequency band Average power within; Low-frequency adjacent band Average power within; High-frequency adjacent band Average power within.

[0064] S4. Based on the optimal individual frequency bands of each historical user, statistical methods can be used to determine the final preset frequency band. For example, based on the lower limit of the individual frequency bands of P historical users... and upper limit The median method can be used to determine the corresponding preset frequency band. ,Right now:

[0065]

[0066] This embodiment analyzes EEG data from multiple historical users on a target task offline. Employing a frequency band evaluation method combining sliding window search and CCA enhancement, it obtains a group-optimal preset frequency band. This band effectively represents the optimal enhancement band for most individuals performing the task, thus avoiding time-consuming frequency band search and label acquisition for each new user. Compared to using fixed physiological frequency bands, the group-optimized preset frequency band significantly improves the signal-to-noise ratio (SNR) enhancement effect of CCA spatial filtering, enabling new users to achieve stable state classification accuracy even under low SNR conditions (such as dry electrodes or hair occlusion). Furthermore, this method does not require large amounts of training data for deep learning, offering advantages such as high computational efficiency and strong interpretability.

[0067] Figure 2 A flowchart illustrating another spontaneous EEG data processing method provided in an embodiment of the present invention. Figure 2 As shown in the specific embodiment of this binary classification task of eye-opening and eye-closed, the raw EEG data of the user's occipital region (hair region) is first collected using 8-lead dry electrodes; then, the data is bandpass filtered and leads with excessive noise are removed. If the number of remaining effective leads is greater than 1, a reference signal is constructed using a preset frequency band determined based on historical users, and the multi-lead signal is spatially filtered by CCA to obtain the enhanced correlation component in the time domain; finally, the enhanced signal is frequency domain transformed to extract the alpha band power feature, and input into the classifier to output the eye-opening or eye-closed state.

[0068] Figure 3 and Figure 4 These are schematic diagrams comparing classification results without CCA signal enhancement and with CCA signal enhancement under normal signal-to-noise ratio (normal electrode contact) conditions. Figure 3 and Figure 4 It can be seen that, under normal electrode contact, regardless of whether CCA signal enhancement is used, both methods can clearly observe the enhancement of the alpha wave when the eyes are closed.

[0069] Figure 5 and Figure 6These are schematic diagrams comparing classification results without CCA signal enhancement and with CCA signal enhancement under low signal-to-noise ratio conditions (such as poor contact due to hair obstruction). Figure 5 and Figure 6 It can be seen that when the signal-to-noise ratio is low, the spectral difference between open and closed eyes is almost indistinguishable without CCA signal enhancement, but the difference is significant after CCA signal enhancement, thus verifying the effectiveness of this method in improving classification accuracy under low signal-to-noise ratio conditions.

[0070] Figure 7 This is a schematic diagram of the structure of a spontaneous EEG data processing system provided in an embodiment of the present invention. Figure 7 As shown, the system includes: The data acquisition module 21 is used to acquire spontaneous EEG data of the target user in the target task state.

[0071] The preprocessing module 22 is used to preprocess the spontaneous EEG data of the target user to obtain the target data; the preprocessing includes at least bandpass filtering and noise lead removal.

[0072] The signal enhancement module 23 is used to enhance the target data based on the preset frequency band corresponding to the target task state and use the typical correlation analysis algorithm to obtain the target enhanced signal when the number of remaining leads corresponding to the target data is greater than 1. The preset frequency band is obtained by frequency band evaluation based on the spontaneous EEG data of multiple historical users in the target task state.

[0073] The classification module 24 is used to extract features and classify the state of the target enhancement signal to obtain the target classification result.

[0074] Furthermore, based on the above embodiments of the invention, the signal enhancement module 23 includes: The remaining lead count acquisition unit is used to acquire the remaining lead count of the target data.

[0075] The reference signal construction unit is used to construct a corresponding reference signal according to a preset frequency band when the number of remaining leads is determined to be greater than 1.

[0076] The analysis unit is used to determine the corresponding spatial filtering coefficients based on the target data and reference signal using typical correlation analysis algorithms.

[0077] The target enhancement signal determination unit is used to linearly weight and combine the target data using spatial filtering coefficients to obtain the target enhancement signal.

[0078] Furthermore, based on the above embodiments of the invention, the analysis unit is specifically used for: Determine the autocovariance matrices of the target data and the reference signal, and the cross-covariance matrix between the target data and the reference signal, respectively; Solving the generalized eigenvalue problem based on the autocovariance matrix and crosscovariance matrix yields multiple canonical correlation coefficients and the eigenvectors corresponding to each canonical correlation coefficient. Select at least one eigenvector corresponding to a canonical correlation coefficient as a spatial filtering coefficient, in descending order of canonical correlation coefficients.

[0079] Furthermore, based on the above embodiments of the invention, the target enhancement signal determination unit is specifically used for: When only one spatial filter coefficient is selected, the target data is linearly weighted and combined using the spatial filter coefficient to obtain an intermediate enhanced signal, which is then used as the target enhanced signal. Alternatively, when multiple spatial filter coefficients are selected, the target data is linearly weighted and combined using each spatial filter coefficient to obtain multiple intermediate enhanced signals, and the multi-channel signal composed of multiple intermediate enhanced signals is used as the target enhanced signal.

[0080] Furthermore, based on the above embodiments of the invention, the target enhancement signal determination unit is specifically used for: When multiple spatial filter coefficients are selected, the target data is linearly weighted and combined using each spatial filter coefficient to obtain multiple intermediate enhanced signals. Multiple intermediate enhancement signals are fused to obtain the target enhancement signal.

[0081] Furthermore, based on the above embodiments of the invention, the process for obtaining the preset frequency band includes: Acquire spontaneous EEG data from multiple historical users in the target task state; The spontaneous EEG data of each historical user were preprocessed to obtain the corresponding historical preprocessed data. For historical preprocessed data with a remaining lead count greater than 1, the corresponding individual frequency bands are determined respectively; The preset frequency band is determined based on the frequency band of each body.

[0082] Furthermore, based on the above embodiments of the invention, determining the individual frequency band includes: Based on the preset wideband search range, sliding window width, and sliding step size, multiple candidate frequency bands are generated; For each candidate frequency band, the historical preprocessed data is enhanced using typical correlation analysis algorithms to obtain the corresponding enhanced signal; Each enhancement signal is evaluated by a specific enhancement performance index, and the individual frequency bands corresponding to historical users are determined from multiple candidate frequency bands based on these evaluation metrics.

[0083] The spontaneous EEG data processing system provided in this embodiment of the invention can execute the spontaneous EEG data processing method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method.

[0084] Figure 8 A schematic diagram of an electronic device 30 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0085] like Figure 8 As shown, the electronic device 30 includes at least one processor 31 and a memory, such as a read-only memory (ROM) 32 or a random access memory (RAM) 33, communicatively connected to the at least one processor 31. The memory stores computer programs executable by the at least one processor. The processor 31 can perform various appropriate actions and processes based on the computer program stored in the ROM 32 or loaded from storage unit 38 into the RAM 33. The RAM 33 can also store various programs and data required for the operation of the electronic device 30. The processor 31, ROM 32, and RAM 33 are interconnected via a bus 34. An input / output (I / O) interface 35 is also connected to the bus 34.

[0086] Multiple components in electronic device 30 are connected to I / O interface 35, including: input unit 36, such as keyboard, mouse, etc.; output unit 37, such as various types of monitors, speakers, etc.; storage unit 38, such as disk, optical disk, etc.; and communication unit 39, such as network card, modem, wireless transceiver, etc. Communication unit 39 allows electronic device 30 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0087] Processor 31 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 31 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 31 performs the various methods and processes described above, such as spontaneous EEG data processing methods.

[0088] In some embodiments, the spontaneous EEG data processing method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 38. In some embodiments, part or all of the computer program may be loaded and / or mounted on electronic device 30 via ROM 32 and / or communication unit 39. When the computer program is loaded into RAM 33 and executed by processor 31, one or more steps of the spontaneous EEG data processing method described above may be performed. Alternatively, in other embodiments, processor 31 may be configured to perform the spontaneous EEG data processing method by any other suitable means (e.g., by means of firmware).

[0089] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0090] In some embodiments, the spontaneous EEG data processing method may be implemented as a computer program, which is implicitly included in a computer program product. When executed by a processor, the computer program implements the spontaneous EEG data processing method of the present invention. The computer program product can be understood as a software product that primarily implements its solution through a computer program. The computer program used to implement the method of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The computer program may be executed entirely on a machine, partially on a machine, partially on a remote machine as a standalone software package, or entirely on a remote machine or server.

[0091] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0092] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0093] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0094] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0095] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0096] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for processing spontaneous EEG data, characterized in that, The method includes: Acquire spontaneous EEG data of the target user in the target task state; The spontaneous EEG data of the target user is preprocessed to obtain the target data; the preprocessing includes at least bandpass filtering and noise lead removal. When the number of remaining leads corresponding to the target data is greater than 1, the target data is enhanced using a typical correlation analysis algorithm based on a preset frequency band corresponding to the target task state to obtain a target enhanced signal; wherein, the preset frequency band is obtained by frequency band evaluation based on spontaneous EEG data of multiple historical users under the target task state; Feature extraction and state classification are performed on the target enhancement signal to obtain the target classification result; Wherein, when the number of remaining leads corresponding to the target data is greater than 1, the target data is enhanced using a typical correlation analysis algorithm based on a preset frequency band corresponding to the target task state to obtain a target enhancement signal, including: Obtain the remaining number of leads for the target data; When it is determined that the number of remaining leads is greater than 1, a corresponding reference signal is constructed according to the preset frequency band; Based on the target data and the reference signal, the corresponding spatial filtering coefficients are determined using the typical correlation analysis algorithm. The target data is linearly weighted and combined using the spatial filtering coefficients to obtain the target enhancement signal; The process for obtaining the preset frequency band includes: Acquire spontaneous EEG data from multiple historical users under the target task state; The spontaneous EEG data of each of the historical users were preprocessed to obtain the corresponding historical preprocessed data. For the historical preprocessed data with a remaining lead count greater than 1, the corresponding individual frequency bands are determined respectively; The preset frequency band is determined based on the frequency bands of each individual frequency band.

2. The method according to claim 1, characterized in that, The step of determining the corresponding spatial filtering coefficients based on the target data and the reference signal using the typical correlation analysis algorithm includes: Determine the autocovariance matrix of the target data and the reference signal, and the crosscovariance matrix between the target data and the reference signal, respectively; Based on the self-covariance matrix and the cross-covariance matrix, the generalized eigenvalue problem is solved to obtain multiple canonical correlation coefficients and the eigenvectors corresponding to each canonical correlation coefficient. In accordance with the order of the canonical correlation coefficients from largest to smallest, at least one feature vector corresponding to the canonical correlation coefficient is selected as the spatial filtering coefficient.

3. The method according to claim 1, characterized in that, The step of linearly weighting and combining the target data using the spatial filtering coefficients to obtain the target enhanced signal includes: When only one spatial filtering coefficient is selected, the target data is linearly weighted and combined using the spatial filtering coefficient to obtain an intermediate enhanced signal, and the intermediate enhanced signal is used as the target enhanced signal. Alternatively, when multiple spatial filtering coefficients are selected, the target data is linearly weighted and combined using each spatial filtering coefficient to obtain multiple intermediate enhanced signals, and the multi-channel signal composed of the multiple intermediate enhanced signals is used as the target enhanced signal.

4. The method according to claim 1, characterized in that, The step of linearly weighting and combining the target data using the spatial filtering coefficients to obtain the target enhanced signal includes: When multiple spatial filtering coefficients are selected, the target data is linearly weighted and combined using each spatial filtering coefficient to obtain multiple intermediate enhanced signals. The target enhanced signal is obtained by fusing multiple intermediate enhanced signals.

5. The method according to claim 1, characterized in that, Determining the individual frequency band includes: Based on the preset wideband search range, sliding window width, and sliding step size, multiple candidate frequency bands are generated; For each candidate frequency band, the historical preprocessed data is enhanced using the typical correlation analysis algorithm to obtain the corresponding enhanced signal; Each enhancement signal is evaluated by an enhancement effect index, and the individual frequency band corresponding to the historical user is determined from among the candidate frequency bands based on the enhancement effect evaluation index.

6. A spontaneous EEG data processing system, characterized in that, The system includes: The data acquisition module is used to acquire spontaneous EEG data of the target user in the target task state; The preprocessing module is used to preprocess the spontaneous EEG data of the target user to obtain target data; the preprocessing includes at least bandpass filtering and noise lead removal; The signal enhancement module is used to enhance the target data using a typical correlation analysis algorithm based on a preset frequency band corresponding to the target task state when the number of remaining leads corresponding to the target data is greater than 1, thereby obtaining a target enhanced signal; wherein, the preset frequency band is obtained by frequency band evaluation based on spontaneous EEG data of multiple historical users under the target task state; The classification module is used to extract features and classify the state of the target enhancement signal to obtain the target classification result; The signal enhancement module includes: The remaining lead count acquisition unit is used to acquire the remaining lead count of the target data; A reference signal construction unit is used to construct a corresponding reference signal according to the preset frequency band when it is determined that the number of remaining leads is greater than 1. The analysis unit is used to determine the corresponding spatial filtering coefficients based on the target data and the reference signal using the typical correlation analysis algorithm; The target enhancement signal determination unit is used to perform linear weighted combination of the target data using the spatial filtering coefficients to obtain the target enhancement signal; The process for obtaining the preset frequency band includes: Acquire spontaneous EEG data from multiple historical users under the target task state; The spontaneous EEG data of each of the historical users were preprocessed to obtain the corresponding historical preprocessed data. For the historical preprocessed data with a remaining lead count greater than 1, the corresponding individual frequency bands are determined respectively; The preset frequency band is determined based on the frequency bands of each individual frequency band.

7. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the spontaneous EEG data processing method according to any one of claims 1-5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the spontaneous EEG data processing method according to any one of claims 1-5.

Images