Electroencephalogram signal quality improving method based on empirical mode decomposition and correlation coefficients

Abstract

The invention relates to the technical field related to electroencephalogram signals, in particular to an electroencephalogram signal quality improving method based on empirical mode decomposition and correlation coefficients. The method comprises the following steps: firstly, carrying out 1-45Hz band-pass filtering preprocessing on an original electroencephalogram signal, and then segmenting the signal into a first segment and a second segment according to a time step difference k; k is far smaller than the fragment length, so that the effective electroencephalogram is highly correlated while the artifacts are low in correlation. Respectively executing empirical mode decomposition on the two fragments in parallel to obtain intrinsic mode functions (IMF) layer by layer, and immediately calculating a Pearson's correlation coefficient rho between the IMF and the IMF of the corresponding layer when each IMF is generated; and if rho is less than a preset threshold 0.7, regarding the IMF as an artifact and deleting the artifact. And finally, reconstructing an electroencephalogram signal by the reserved IMF.

Description

Technical Field

[0001] This application relates to the field of EEG signal technology, specifically to a method for improving the quality of EEG signals based on empirical mode decomposition and correlation coefficient. Background Technology

[0002] Electroencephalogram (EEG) signals are weak bioelectrical signals generated on the scalp surface by clusters of neurons in the cerebral cortex firing. The amplitude is typically only 10–100 μV, and the frequency band covers 0.5 Hz–100 Hz. Due to the non-stationary nature of EEG, its low signal-to-noise ratio (SNR is often below 10 dB), and its susceptibility to multi-source interference, EEG signals are highly susceptible to noise contamination during actual acquisition. Summary of the Invention

[0003] In view of this, the embodiments of this application aim to provide a method for improving the quality of EEG signals based on empirical mode decomposition and correlation coefficient, so as to reduce noise and improve the quality of EEG signals.

[0004] This application provides a method for improving the quality of EEG signals based on empirical mode decomposition and correlation coefficient, including: Acquiring EEG signals; The EEG signal is segmented to obtain a first segment of data from m time steps to n time steps; and a second segment of data from m+k time steps to n+k time steps. Where: m, n, and k are all positive integers; m is less than n; k is a positive integer less than the difference between m and n; Empirical mode decomposition is performed on the first and second segmented data respectively; For each set of intrinsic mode functions, determine the Pearson correlation coefficient between the corresponding intrinsic mode functions in that set; wherein, each set of corresponding intrinsic mode functions includes: an intrinsic mode function from the first segment of decomposed data and an intrinsic mode function from the second segment of decomposed data, and the two have the same frequency level; If the Pearson correlation coefficient is less than a preset value, then delete the intrinsic mode functions corresponding to the Pearson correlation coefficient in the first and second decomposed data. The EEG signal was reconstructed based on the deleted first and second segment data.

[0005] In some embodiments, before performing data segmentation on the electroencephalogram (EEG) signals, the method further includes: The EEG signal is preprocessed to perform preliminary noise reduction.

[0006] In some embodiments, the preprocessing of the EEG signal includes: The EEG signal is initially denoised using a bandpass filter ranging from 1 to 45 Hz.

[0007] In some embodiments, it also includes: A parallel processing approach is adopted, simultaneously executing steps to perform empirical mode decomposition on the first and second segmented data, and step to determine the Pearson correlation coefficient between the corresponding intrinsic mode functions for each group.

[0008] In some embodiments, parallel processing methods include: While performing empirical mode decomposition on the first and second segmented data respectively, for each group of intrinsic mode functions obtained by decomposition, the step is to determine the Pearson correlation coefficient between the corresponding intrinsic mode functions of that group.

[0009] In some embodiments, the preset value is 0.7.

[0010] In some embodiments, the termination condition for empirical mode decomposition is: the residual becomes a monotonic function, or the energy of the residual is lower than a preset threshold.

[0011] This application provides a device for improving the quality of EEG signals based on empirical mode decomposition and correlation coefficient, comprising: The acquisition module is used to acquire electroencephalogram (EEG) signals; The segmentation module is used to segment the EEG signal into data to obtain first segmented data from m time steps to n time steps; and second segmented data from m+k time steps to n+k time steps; Where: m, n, and k are all positive integers; m is less than n; k is less than the difference between m and n by 0.1. The decomposition module is used to perform empirical mode decomposition on the first segmented data and the second segmented data, respectively. The relevant module is used to determine the Pearson correlation coefficient between the corresponding intrinsic mode functions for each group; wherein, each corresponding intrinsic mode function includes: an intrinsic mode function from the first segmented data after decomposition and an intrinsic mode function from the second segmented data after decomposition, and the two have the same frequency level; The deletion module is used to delete the intrinsic mode functions corresponding to the Pearson correlation coefficient in the first segmented data and the second segmented data after decomposition if the Pearson correlation coefficient is less than a preset value. The reconstruction module is used to reconstruct EEG signals based on the deleted first and second segment data.

[0012] This application provides an electronic device, including: A processor, and a memory for storing a processor-executable program; The processor is configured to implement the above-described method for improving the quality of EEG signals based on empirical mode decomposition and correlation coefficients by running a program in the memory.

[0013] This application provides a computer-readable storage medium storing a computer program that, when executed by a processor, causes the processor to perform the above-described method for improving the quality of electroencephalogram (EEG) signals based on empirical mode decomposition and correlation coefficients.

[0014] This application provides a method for improving the quality of electroencephalogram (EEG) signals based on empirical mode decomposition (EMD) and correlation coefficients. The method involves acquiring an EEG signal; segmenting the EEG signal into a first segment (m to n time steps) and a second segment (m+k to n+k time steps); where m, n, and k are positive integers; m is less than n; and k is a positive integer less than the difference between m and n; performing empirical mode decomposition (EMD) on the first and second segments; determining the Pearson correlation coefficient between the corresponding intrinsic mode functions (IMFs) for each group; each IMF group includes one IMF from the first segment and one IMF from the second segment, with consistent frequency levels; if the Pearson correlation coefficient is less than a preset value, deleting the IMF corresponding to the Pearson correlation coefficient from both the first and second segments; and reconstructing the EEG signal based on the deleted first and second segments. The solution provided in this application combines EMD data-driven adaptive decomposition and Pearson correlation coefficient to automatically eliminate low-correlation noise or artifact-dominated IMFs, while retaining the essential characteristics of highly correlated signals. This eliminates the need for preset noise bands or filtering parameters, demonstrating strong adaptability and robustness. It utilizes the time-frequency localization characteristics of EMD to decompose the signal and, combined with the Pearson correlation coefficient, accurately removes artifacts such as electrooculography (EOG), electromyography (EMG), and power line interference that overlap with the frequency domain of EEG signals, while preserving effective components such as alpha and beta waves, ensuring signal integrity and information content. Attached Figure Description

[0015] The above and other objects, features, and advantages of this application will become more apparent from the more detailed description of the embodiments of this application in conjunction with the accompanying drawings. The drawings are provided to further illustrate the embodiments of this application and form part of the specification. They are used together with the embodiments of this application to explain this application and do not constitute a limitation thereof. In the drawings, the same reference numerals generally represent the same components or steps.

[0016] Figure 1This is a flowchart illustrating a method for improving the quality of EEG signals based on empirical mode decomposition and correlation coefficients, provided in one embodiment of this application.

[0017] Figure 2 This is a schematic diagram of a segmentation provided in one embodiment of this application.

[0018] Figure 3 This is a schematic diagram of the structure of an EEG signal quality improvement device based on empirical mode decomposition and correlation coefficient provided in one embodiment of this application.

[0019] Figure 4 This is a schematic diagram of an electronic device structure provided in one embodiment of this application. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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 are within the scope of protection of the present invention.

[0021] This application proposes a three-step approach of "microstep difference-EMD-Pearson" to significantly improve the quality of EEG signals under single-channel, low-latency conditions. Specifically, the original EEG signal is truncated along the time axis into two segments separated by only k sampling points (k ≪ segment length), forming "homogeneous micro-shift" data pairs. This design ensures that the effective EEG components are highly similar, while transient artifacts such as electrooculography (EOG) and electromyography (EMG) have low correlation between the two segments. Empirical mode decomposition (EMD) is performed on each of the two data segments to obtain intrinsic mode functions (IMFs) arranged in frequency hierarchy, IMF_1…IMF_K. The decomposition terminates at "residual monotonicity" or "residual energy < 1–5% of original energy". The Pearson correlation coefficient ρ is calculated for IMF pairs at the same level; a threshold of 0.7 is set. IMFs with ρ ≥ 0.7 are identified as effective EEG IMFs and retained, while those with ρ < 0.7 are identified as artifact IMFs and deleted. Finally, the retained IMFs are superimposed and reconstructed to obtain an EEG signal with high signal-to-noise ratio and low artifacts.

[0022] This method requires no additional reference channel, can be completed in a single channel with a latency of ≤50 ms, and is applicable to various scenarios such as clinical, BCI, and wearable devices.

[0023] After introducing the basic principles of this application, various non-limiting embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0024] Figure 1 This is a flowchart illustrating a method for improving the quality of EEG signals based on empirical mode decomposition and correlation coefficients, provided in one embodiment of this application. Figure 1 As shown in the figure, the method includes the following steps.

[0025] Step S110: Obtain electroencephalogram (EEG) signals; The original EEG signals are weak bioelectrical signals recorded on the scalp surface of the brain's neuron electrical activities, with an amplitude usually of 10 - 100 μV and a frequency range covering 0.5 Hz (slow wave) to 100 Hz (high-frequency γ wave). The original EEG signals have non-stationarity (time-varying statistical characteristics), low signal-to-noise ratio (SNR is often below 10 dB), and are vulnerable to contamination by physiological artifacts (electrooculogram, electromyogram), environmental noise (power frequency interference), baseline drift, etc. The acquisition methods of the original EEG signals include: recording from the scalp surface through electrode acquisition devices, specifically involving multi-channel (such as 64-channel) signals, and the sampling rate can be 250 Hz (which can be adjusted according to the device in practice).

[0026] Step S120: Segment the EEG signals to obtain the first segmented data from the m-th time step to the n-th time step; and the second segmented data from the (m + k)-th time step to the (n + k)-th time step; Where: m, n, and k are all positive integers; m is less than n; k is a positive integer less than the difference between m and n; further, k is less than 0.1 times the difference between m and n.

[0027] In this way, the continuous EEG signals are segmented into two sub-signals with time overlap, and the stability of the signal components is verified through time-shift comparison (the effective signals remain consistent after time shift, while the noise has low consistency due to randomness). The specific operation is as follows: The original signal is divided into sequences according to the time step, and each time step contains multi-channel voltage values, forming an input matrix X ∈ R

[0030] , ,

[0029] , × C (N is the number of time steps, and C is the number of channels).

[0028] The segmentation results are as follows: The first segmented data x1(t): covering the time steps from m to n (m < n); The second segmented data x2(t): covering the time steps from m + k to n + k (k is a positive integer, and k < n - m, ensuring that there is an overlapping area between the two segments of data).

[0029] Example: If m = 1, n = 100, and k = 1, then x1(t) is the 1st to 99th time steps, and x2(t) is the 2nd to 100th time steps, and the two signals overlap by 98 time steps, facilitating subsequent comparative analysis.

[0030] Step S130: Perform empirical mode decomposition on the first segmented data and the second segmented data respectively; Step S140: For each group of corresponding intrinsic mode functions, determine the Pearson correlation coefficient between the corresponding intrinsic mode functions of that group; wherein, each group of corresponding intrinsic mode functions includes: an intrinsic mode function from the first segmented data after decomposition and an intrinsic mode function from the second segmented data after decomposition, and the two have the same frequency level; Since x2(t) is a time-shifted version of x1(t), theoretically, the IMF of the corresponding frequency component should satisfy... .in, The intrinsic mode functions in the second segment of data; To and Correspondingly, the intrinsic mode functions in the first segment of data.

[0031] Frequency hierarchy refers to the sorting of intrinsic mode functions in the first segment of decomposed data according to their frequency from high to low, with the highest frequency in the first layer, the next highest in the second layer, and so on. The same principle applies to the first segment of decomposed data.

[0032] Furthermore, in the actual decomposition process, the eigenmode functions decomposed first generally have higher frequencies than those decomposed later. Therefore, under normal circumstances, the eigenmode functions decomposed first constitute the first layer, the next first the second layer, and so on.

[0033] For each pair of matched IMF components and Calculate its Pearson correlation coefficient. : Step S150: If the Pearson correlation coefficient is less than a preset value, delete the intrinsic mode functions corresponding to the Pearson correlation coefficient in the first segmented data and the second segmented data after decomposition. Specifically, the preset value can be 0.7.

[0034] If ρi,j ≥ 0.7: the IMF component is considered an effective component of the signal and is retained.

[0035] If ρi,j < 0.7: the IMF is considered to be possibly affected by noise pollution or inconsistent decomposition, and is therefore rejected.

[0036] Step S160: Reconstruct the EEG signal based on the deleted first and second segment data.

[0037] The reconstructed signal eliminated low-correlation noise components and retained high-correlation effective EEG components, significantly improving the signal-to-noise ratio and providing high-quality data for subsequent analysis (such as brain-computer interface classification and clinical diagnosis).

[0038] Specifically, before segmenting the EEG signal, the method further includes: preprocessing the EEG signal to perform preliminary noise reduction.

[0039] That is, a preprocessing step is further included between steps S110 and S120 to perform preliminary noise reduction on the raw EEG signal, thereby reducing the impact of baseline drift, electromyography, electrooculography, and power line interference on subsequent EMD decomposition and IMF screening. Preprocessing S115 specifically includes the following sub-step: preliminary noise reduction of the EEG signal using a bandpass filter within a preset frequency range. Specifically, a bandpass filter of 1~45 Hz is used to process the raw EEG signal, the function of which is: Filter out low-frequency drift (such as baseline noise, typically <0.5 Hz) and high-frequency interference (such as electromyography artifacts, power supply noise, etc., typically >45 Hz). Retaining key frequency band information relevant to the task, such as effective components of EEG signals like mu rhythm (8~13 Hz) and beta rhythm (13~30 Hz), lays the foundation for subsequent processing.

[0040] This preprocessing step is an important preliminary step in the entire signal quality improvement process. By initially screening frequency bands, it reduces the total amount of noise that needs to be processed in the subsequent empirical mode decomposition (EMD), thereby improving the efficiency and accuracy of the overall method.

[0041] In some embodiments, the parallel processing steps are used to perform empirical mode decomposition on the first segmented data and the second segmented data respectively, and the steps are used to determine the Pearson correlation coefficient between the corresponding intrinsic mode functions of each group.

[0042] Specifically, EMD decomposition and Pearson calculation were originally cascaded, with sequential execution resulting in added computation time. In this embodiment, two EMD channels are started in parallel within the same clock cycle, and ρ calculation is triggered immediately after each IMF is generated, theoretically achieving a speedup of ≈2. Specifically, the parallel processing methods include: Specifically, while performing empirical mode decomposition on the first and second segmented data respectively, for each group of intrinsic mode functions obtained by decomposition, the step is to determine the Pearson correlation coefficient between the corresponding intrinsic mode functions of that group.

[0043] To address the "computational efficiency and real-time bottleneck" in multi-channel EEG signal processing (traditional methods involve sequential computation channel by channel, resulting in high latency and making it difficult to meet the low latency requirements of <100ms in scenarios such as brain-computer interfaces), a parallel architecture is used to improve processing speed and ensure the feasibility of the method in practical applications.

[0044] The EMD decomposition process for the first and second data segments can be performed simultaneously. Since the two data segments are independent time-shifted signal fragments, their decomposition processes are independent and can be distributed to different computing units (such as multi-threaded, multi-core processors) for synchronous execution, reducing the overall decomposition time.

[0045] During the decomposition process, PCC calculation can be initiated immediately for the intrinsic mode functions (IMFs) that have been decomposed. For example, when the high-frequency IMFs of the first segment of data (such as IMF11) have been decomposed and the high-frequency IMFs corresponding to the second segment of data (such as IMF21) are generated synchronously, PCC can be calculated immediately for this set of matching IMFs without waiting for all IMFs to be decomposed before unified calculation. This pipelined parallelism further shortens the processing cycle.

[0046] In this embodiment, the calculation of EMD and the calculation of Pearson correlation coefficient ρ are designed as a pipeline of "decomposition and matching on the same side" to achieve true zero-wait parallelism.

[0047] By using the parallel processing described above, the total latency of multi-channel signal processing can be significantly reduced, while retaining the adaptive characteristics of EMD decomposition and the accuracy of PCC screening. This allows the method to meet real-time requirements without sacrificing signal quality improvement, making it particularly suitable for time-sensitive applications such as brain-computer interfaces.

[0048] The solution provided in this application will be described below with reference to the above preferred embodiments: First, the raw MI-EEG signal (i.e., electroencephalogram signal) is acquired; then, the raw MI-EEG signal is preliminarily processed by a bandpass filter of 1~45 Hz to filter out low-frequency drift (such as baseline noise) and high-frequency interference (such as electromyographic artifacts and power supply frequency noise), while retaining task-related frequency band information (such as mu rhythm and beta rhythm).

[0049] Reference Figure 2 Data segmentation: The original MI-EEG signal (sampling rate 250 Hz, 64 channels) is divided into sequences according to time steps. Each time step contains voltage values ​​from multiple channels, forming an input matrix X∈RN×C, where N is the number of time steps and C is the number of channels. The current sample segment is divided into two data segments, the first segment... The second data segment is from the first time step to the (n-1)th time step. This ranges from the second time step to the nth time step. This provides two data segments for later comparative analysis.

[0050] Then, empirical mode decomposition is performed: To avoid the impact of amplitude differences on the decomposition, the two signal segments are standardized (Z-score normalization). Then, EMD decomposition is performed on each of the two signal segments separately. right )and The steps for independently performing EMD decomposition are as follows: Let the original signal be r(t) = x(t), and the IMF set... .

[0051] The IMF extraction (iterative process) is as follows: Step 1 (Screening Process): Identify all local extrema (maximum and minimum) of r(t).

[0052] The upper and lower envelopes were fitted using cubic spline interpolation respectively. and .

[0053] Calculate the mean envelope .

[0054] Extract detail components .

[0055] Step 2 (IMF determination): Check whether h(t) satisfies the IMF condition (the number of extreme points and the number of zero crossings do not differ by more than 1, and the mean envelope is close to zero).

[0056] If the condition is not met, repeat step 1 with h(t) as the new r(t); if the condition is met, save h(t) as an IMF and update the residuals. .

[0057] Termination condition: Decomposition stops when the residual r(t) is a monotonic function or the residual energy is below the threshold.

[0058] The final decomposition result is as follows:

[0059] Simultaneous parallel IMF component matching and correlation analysis: To ensure correct matching, the following strategies can be adopted: Matching by frequency level: Starting from high frequency ( From low frequency () Layer-by-layer comparison: The Pearson correlation coefficient is calculated using an algorithm based on the Pearson correlation coefficient. .

[0060] Signal purification and reconstruction: If ρi,j ≥ 0.7: the IMF component is considered an effective component of the signal and is retained.

[0061] If ρi,j < 0.7: the IMF is considered to be possibly affected by noise pollution or inconsistent decomposition, and is therefore rejected.

[0062] The reconstructed and purified signal: Retain the IMFs that satisfy the conditions for x1(t) and x2(t) respectively, and reconstruct the signal:

[0063]

[0064] Among them: I I represents the intrinsic mode functions that were not deleted from the first segment of data after decomposition; These are the intrinsic mode functions that were not deleted from the second segment of data after decomposition; This is the first segment of data to be reconstructed; For the reconstructed second segment of data; further, based on and EEG signals can be spliced ​​together; during the splicing process, non-overlapping time segments can be directly used. and For the corresponding part, for overlapping time periods, it can be directly adopted. and The mean of the corresponding part.

[0065] The apparatus embodiments of this application can be used to execute the method embodiments of this application. For details not disclosed in the apparatus embodiments of this application, please refer to the method embodiments of this application.

[0066] Figure 3 The diagram shown is a block diagram of an EEG signal quality improvement device based on empirical mode decomposition and correlation coefficient, according to an embodiment of this application. Figure 3 As shown, the device includes: Acquisition module 31 is used to acquire electroencephalogram (EEG) signals; The segmentation module 32 is used to segment the EEG signal to obtain first segmented data from m time steps to n time steps; and second segmented data from m+k time steps to n+k time steps. Where: m, n, and k are all positive integers; m is less than n; k is less than the difference between m and n by 0.1. Decomposition module 33 is used to perform empirical mode decomposition on the first segmented data and the second segmented data respectively; The relevant module 34 is used to determine the Pearson correlation coefficient between the corresponding intrinsic mode functions for each group; wherein, each corresponding intrinsic mode function includes: an intrinsic mode function from the first segmented data after decomposition and an intrinsic mode function from the second segmented data after decomposition, and the two have the same frequency level; The deletion module 35 is used to delete the intrinsic mode functions corresponding to the Pearson correlation coefficient in the first segmented data and the second segmented data after decomposition if the Pearson correlation coefficient is less than a preset value. Reconstruction module 36 is used to reconstruct EEG signals based on the deleted first and second segment data.

[0067] Below, for reference Figure 4 This describes an electronic device according to embodiments of the present application. Figure 4 A block diagram of an electronic device according to an embodiment of this application is illustrated.

[0068] like Figure 4 As shown, the electronic device 400 includes one or more processors 410 and memory 420.

[0069] The processor 410 may be a central processing unit (CPU) or other form of processing unit with data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device 400 to perform desired functions.

[0070] The memory 420 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory. The non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 410 may execute the program instructions to implement the EEG signal quality improvement method based on empirical mode decomposition and correlation coefficients of the various embodiments of this application described above, and / or other desired functions. Various contents, such as category correspondences, may also be stored in the computer-readable storage medium.

[0071] In one example, the electronic device 400 may also include an input device 430 and an output device 440, which are interconnected via a bus system and / or other forms of connection mechanism (not shown).

[0072] In addition, the input device 430 may also include, for example, a keyboard, mouse, interface, etc. The output device 440 can output various information to the outside, including analysis results, etc. The output device 440 may include, for example, a display, speaker, printer, and communication network and its connected remote output devices, etc.

[0073] Of course, for the sake of simplicity, Figure 4Only some of the components of the electronic device relevant to this application are shown in this illustration; components such as buses, input / output interfaces, etc., are omitted. In addition, the electronic device may include any other suitable components depending on the specific application.

[0074] In addition to the methods and devices described above, embodiments of this application may also be computer program products, which include computer program instructions that, when executed by a processor, cause the processor to perform the steps in the EEG signal quality improvement method based on empirical mode decomposition and correlation coefficient according to various embodiments of this application as described in the "Exemplary Methods" section of this specification.

[0075] The computer program product can be written in any combination of one or more programming languages ​​to perform the operations of the embodiments of this application. The programming languages ​​include object-oriented programming languages ​​such as Java and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computing device, partially on the user's computing device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0076] Furthermore, embodiments of this application may also be computer-readable storage media storing computer program instructions that, when executed by a processor, cause the processor to perform the steps in the method for improving the quality of EEG signals based on empirical mode decomposition and correlation coefficients according to various embodiments of this application as described in the "Exemplary Methods" section above.

[0077] The computer-readable storage medium may be any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof.

[0078] The above description has been given for purposes of illustration and description. Furthermore, this description is not intended to limit the embodiments of this application to the forms disclosed herein. Although numerous exemplary aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations thereof.

Claims

1. An electroencephalogram signal quality improvement method based on empirical mode decomposition and correlation coefficient, characterized in that, The method comprises the following steps: obtaining an electroencephalogram signal; performing data segmentation on the electroencephalogram signal to obtain first segmented data from m time steps to n time steps and second segmented data from m+k time steps to n+k time steps; wherein m, n, and k are positive integers; m is less than n; and k is a positive integer less than the difference between m and n; respectively performing empirical mode decomposition on the first segmented data and the second segmented data; for each group of corresponding intrinsic mode functions, determining a Pearson correlation coefficient between the corresponding intrinsic mode functions in the group; wherein each group of corresponding intrinsic mode functions comprises one intrinsic mode function from the first segmented data after decomposition and one intrinsic mode function from the second segmented data after decomposition, and the two intrinsic mode functions have consistent frequency levels; if the Pearson correlation coefficient is less than a preset value, deleting the intrinsic mode functions corresponding to the Pearson correlation coefficient from the first segmented data after decomposition and the second segmented data after decomposition; reconstructing the electroencephalogram signal based on the first segmented data after deletion and the second segmented data after deletion. Before the step of performing data segmentation on the electroencephalogram signal, the method further comprises the following step:

2. The method for EEG signal quality improvement based on empirical mode decomposition and correlation coefficient according to claim 1, characterized in that, preprocessing the electroencephalogram signal to preliminarily denoise the electroencephalogram signal. The step of preprocessing the electroencephalogram signal comprises the following step:

3. The method of claim 2, wherein the method is characterized by, preliminarily denoising the electroencephalogram signal by a band-pass filter with a frequency range of 1-45 Hz. The method further comprises the following steps:

4. The method of claim 1, wherein the method is characterized by, respectively performing empirical mode decomposition on the first segmented data and the second segmented data by using a parallel processing mode, and for each group of corresponding intrinsic mode functions, determining a Pearson correlation coefficient between the corresponding intrinsic mode functions in the group by using the parallel processing mode. The parallel processing mode comprises the following steps:

5. The method for EEG signal quality improvement based on empirical mode decomposition and correlation coefficient according to claim 4, characterized in that, respectively performing empirical mode decomposition on the first segmented data and the second segmented data, and for each group of corresponding intrinsic mode functions, determining a Pearson correlation coefficient between the corresponding intrinsic mode functions in the group at the same time. The preset value is 0.

7.

6. The method of claim 1, wherein the method is characterized by, The termination condition of empirical mode decomposition is that a residual error becomes a monotonic function or the energy of the residual error is lower than a preset threshold.

7. The method of claim 1, wherein the method is characterized by, The method comprises the following steps:

8. An electroencephalogram signal quality improvement device based on empirical mode decomposition and correlation coefficient, characterized in that, an obtaining module configured to obtain an electroencephalogram signal; a segmentation module configured to perform data segmentation on the electroencephalogram signal to obtain first segmented data from m time steps to n time steps and second segmented data from m+k time steps to n+k time steps; wherein m, n, and k are positive integers; m is less than n; and k is a positive integer less than the difference between m and n; a decomposition module configured to respectively perform empirical mode decomposition on the first segmented data and the second segmented data; a correlation module configured to, for each group of corresponding intrinsic mode functions, determine a Pearson correlation coefficient between the corresponding intrinsic mode functions in the group; wherein each group of corresponding intrinsic mode functions comprises one intrinsic mode function from the first segmented data after decomposition and one intrinsic mode function from the second segmented data after decomposition, and the two intrinsic mode functions have consistent frequency levels; a deletion module configured to, if the Pearson correlation coefficient is less than a preset value, delete the intrinsic mode functions corresponding to the Pearson correlation coefficient from the first segmented data after decomposition and the second segmented data after decomposition; and ​ ​ The reconstruction module is configured to reconstruct the electroencephalogram signal based on the first segmented data and the second segmented data after the deletion.

9. An electronic device, comprising: The method comprises: a processor, and a memory for storing a program executable by the processor; The processor is configured to implement the electroencephalogram signal quality improvement method based on empirical mode decomposition and correlation coefficient according to any one of claims 1 to 7 by running the program in the memory.

10. A computer-readable storage medium, characterized in that, The computer program is stored on the computer readable storage medium and, when executed by the processor, causes the processor to perform the electroencephalogram signal quality improvement method based on empirical mode decomposition and correlation coefficient according to any one of claims 1 to 7.

Images