WO2025030833A1 - 面向稳态视觉诱发电位的低训练成本脑电特征解码方法 - Google Patents

面向稳态视觉诱发电位的低训练成本脑电特征解码方法 Download PDF

Info

Publication number
WO2025030833A1
WO2025030833A1 PCT/CN2024/081370 CN2024081370W WO2025030833A1 WO 2025030833 A1 WO2025030833 A1 WO 2025030833A1 CN 2024081370 W CN2024081370 W CN 2024081370W WO 2025030833 A1 WO2025030833 A1 WO 2025030833A1
Authority
WO
WIPO (PCT)
Prior art keywords
steady
visual evoked
state visual
evoked potential
sub
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
PCT/CN2024/081370
Other languages
English (en)
French (fr)
Inventor
柯余峰
刘爽
明东
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Tianjin University
Original Assignee
Tianjin University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Tianjin University filed Critical Tianjin University
Publication of WO2025030833A1 publication Critical patent/WO2025030833A1/zh
Anticipated expiration legal-status Critical
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
    • A61B5/25—Bioelectric electrodes therefor
    • A61B5/251—Means for maintaining electrode contact with the body
    • A61B5/256—Wearable electrodes, e.g. having straps or bands
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
    • A61B5/316—Modalities, i.e. specific diagnostic methods
    • A61B5/369—Electroencephalography [EEG]
    • A61B5/377—Electroencephalography [EEG] using evoked responses
    • A61B5/378—Visual stimuli
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7235—Details of waveform analysis
    • A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
    • A61B5/7267—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems involving training the classification device
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/02—Preprocessing
    • G06F2218/04—Denoising
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/08—Feature extraction
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/12—Classification; Matching

Definitions

  • Embodiments of the present application relate to coding and decoding technologies for brain-computer interface signals, for example, to a low-training-cost EEG feature decoding method and a brain-computer interface system for steady-state visual evoked potentials.
  • BCI brain-computer interfaces
  • VEP visually evoked potentials
  • SSVEP Steady-State Visual Evoked Potentials
  • the SSVEP decoding algorithm represented by the Task-Related Component Analysis (TRCA) algorithm needs to rely on a large amount of training data to obtain high recognition accuracy.
  • TRCA Task-Related Component Analysis
  • pre-training helps the algorithm learn from individual EEG data and generate spatial filters and EEG template signals that meet the EEG characteristics of the subjects, so that the algorithm can obtain satisfactory recognition results.
  • the principle of the TRCA algorithm itself is to maximize the repeated signal components between trials, the recognition performance tends to drop significantly when there is less training data, especially when there is only a single trial training data for each frequency, the recognition performance is very poor.
  • the present application provides a low-training-cost EEG feature decoding method and brain-computer interface for steady-state visual evoked potentials, so as to achieve the technical effect of improving the accuracy of identifying unknown visual stimuli.
  • the embodiments of the present application provide a low-training-cost EEG feature decoding method for steady-state visual evoked potentials, comprising:
  • N f visual stimuli with frequencies of f 1 ,...f i ,...f n ,,...f N f are presented on the screen, and the original steady-state visual evoked potential signals of the subjects looking at these visual stimuli are collected as training data.
  • Nc is the number of EEG channels
  • Ns is the number of sampling points
  • Nt is the number of trials
  • Nfb filter banks are used to preprocess the original steady-state visual evoked potential Xn to obtain the preprocessed steady-state visual evoked potential of Nfb sub-bands
  • m 1, 2, ..., N fb ;
  • the correlation coefficients [ ⁇ 1 , ... ⁇ i , ..., ⁇ n , ..., ⁇ Nf ] between the initial steady-state visual evoked potentials [X 1 , ...X i , ...,X n , ...,X Nf ] are as follows: the i-th correlation coefficient ⁇ i is the largest, and the visual stimulation frequency of the unknown steady-state visual evoked potential K is fi .
  • the present application also provides a low-training-cost EEG feature decoding device for steady-state visual evoked potentials, comprising:
  • the preprocessing module is set to present N f visual stimuli with frequencies of f 1 ,...f i ,...f n ,,...f Nf on the screen, and collect the original steady-state visual evoked potential signals of the subjects looking at these visual stimuli as training data.
  • Nc is the number of EEG channels
  • Ns is the number of sampling points
  • Nt is the number of trials
  • Nfb filter banks are used to preprocess the original steady-state visual evoked potential Xn to obtain the preprocessed steady-state visual evoked potential of Nfb sub-bands
  • m 1, 2, ..., N fb ;
  • Sub-period division module set to pre-process the steady-state visual evoked potential Divided into P n sub-periods
  • Integrated spatial filter acquisition module set according to Get the spatial filter of the mth sub-band
  • the spatial filter for the mth sub-band of N f frequencies Integrate to obtain the integrated spatial filter of the mth sub-band
  • Template signal acquisition module set to Taking P n sub-periods as an average, we get Copy multiple connections Synthesize the steady-state visual evoked potential template signal of the required length
  • the correlation coefficient calculation module is set to integrate the spatial filter and steady-state visual evoked potential Template signal Calculate the correlation coefficient of N fb sub-bands For N fb The weighted sum is used to obtain the correlation coefficient ⁇ n between the unknown steady-state visual evoked potential K and the original steady-state visual evoked potential X n ;
  • the visual stimulation frequency acquisition module is configured to obtain the correlation coefficient [ ⁇ 1 ,... ⁇ i ,..., ⁇ n ,..., ⁇ Nf ] between the unknown steady-state visual evoked potential K and all original steady-state visual evoked potentials [X 1 ,...X i ,...,X n ,...,X Nf ] whose frequencies are f 1 ,...f i ,...f n ,,...f Nf respectively. If the i-th correlation coefficient ⁇ i is the largest, the visual stimulation frequency of the unknown steady-state visual evoked potential K is fi .
  • the embodiment of the present application further provides a low-training-cost brain-computer interface system for steady-state visual evoked potentials, comprising:
  • a memory configured to store at least one program
  • the at least one processor When the at least one program is executed by the at least one processor, the at least one processor implements the low-training-cost EEG feature decoding method for steady-state visual evoked potentials as described in any one of the embodiments.
  • an embodiment of the present application further provides a computer-readable storage medium comprising a computer program, which, when executed by a processor, implements a low-training-cost EEG feature decoding method for steady-state visual evoked potentials as described in any one of the embodiments.
  • FIG1 is a schematic diagram of the composition structure of a steady-state visual evoked potential (SSVEP) brain-computer interface system;
  • SSVEP steady-state visual evoked potential
  • FIG2 is a schematic diagram of the signal-to-noise ratio characteristics of a typical steady-state visual evoked potential in the frequency domain;
  • FIG3 is a flow chart of a low-training-cost EEG feature decoding method for steady-state visual evoked potentials provided in Example 1 of the present application;
  • FIG4 is a graph showing the percentage of recognition accuracy of unknown visual stimuli using different algorithms
  • FIG5( a ) is a schematic diagram showing the effect of the test data length on the recognition accuracy of different algorithms when the number of trials is 1;
  • FIG5( b ) is a schematic diagram showing the effect of the test data length on the information transmission efficiency of different algorithms when the number of trials is 1;
  • FIG6 is a comparison chart of the recognition accuracy of the ePRCA algorithm and the eTRCA algorithm with different numbers of trials;
  • FIG7 is a schematic diagram of the structure of a low-training-cost EEG feature decoding device for steady-state visual evoked potentials provided in Example 2 of the present application;
  • FIG8 is a schematic diagram of the structure of a low-training-cost brain-computer interface system for steady-state visual evoked potentials provided in Example 3 of the present application.
  • FIG. 1 is a schematic diagram of the composition structure of the steady-state visual evoked potential (SSVEP) brain-computer interface system, which mainly includes a stimulation module, an acquisition module and a processing module.
  • the stimulation module refers to the visual stimulation of a fixed frequency encoded by the display.
  • a steady-state visual evoked potential can be induced in the occipital area of the subject's brain.
  • the acquisition module obtains electroencephalographic physiological signals representing brain thinking and cognitive activities from the subject's cerebral cortex or brain cavity through electrodes.
  • the processing module extracts the corresponding electroencephalographic characteristic signals from the electroencephalographic physiological signals through a series of signal processing methods and performs pattern recognition, thereby converting the electroencephalographic characteristic signals into machine language instructions.
  • FIG 2 is a schematic diagram of the signal-to-noise ratio characteristics of a typical steady-state visual evoked potential in the frequency domain.
  • the steady-state visual evoked potential (SSVEP) signal is induced by a visual stimulus of a fixed frequency.
  • the visual stimulus is presented periodically at a specific frequency (such as flickering, graphic flipping, graphic scaling, etc.)
  • the subject's visual system will be affected and produce an evoked response with stable frequency domain characteristics.
  • the response contains the frequency component of the visual stimulus and its higher harmonic components.
  • Figure 2 is the result of averaging the steady-state visual evoked potential of multiple trials. Averaging multiple trials can offset random noise and improve the signal-to-noise ratio.
  • the high signal-to-noise ratio effect at the visual stimulus frequency and higher harmonic frequency shown in Figure 2 may not be achieved.
  • Steady-state visual evoked potentials also have spatial characteristics. For example, different frequency components of the SSVEP signal come from different brain areas of the subject. Therefore, when the SSVEP signal is transmitted to different parts of the subject's scalp surface, the signal-to-noise ratio of each harmonic is also different.
  • FIG3 is a flow chart of a low-training-cost EEG feature decoding method for steady-state visual evoked potentials provided in Example 1 of the present application.
  • This embodiment can be applied to provide a spatial filter to decode the EEG features of steady-state visual evoked potentials.
  • the training of the spatial filter can be based on experimental data of fewer trials, or even on single experimental data.
  • the method can be executed by a computer, for example, including the following steps:
  • step S310-step S340 is the process of training to obtain the spatial filter
  • step S350-step S370 uses the trained spatial filter to filter and analyze the unknown steady-state visual evoked potential, so as to obtain the meaning represented by the unknown steady-state visual evoked potential, that is, to obtain the frequency information of the visual stimulus that the subject is looking at.
  • Nc is the number of EEG channels
  • Ns is the number of sampling points
  • Nt is the number of trials
  • Nfb filter banks are used to preprocess the original steady-state visual evoked potential Xn to obtain the preprocessed steady-state visual evoked potential of Nfb sub-bands
  • m 1,2,...,N fb .
  • n 1,2,..., Nf ; Nf is the number of frequency types of visual stimulation, Nc , Ns , Nt are the number of EEG channels, the number of sampling points and the number of trials respectively.
  • the collected multi-channel EEG signal Xn can be band-pass filtered and frequency-notched to filter out obvious noise signals.
  • the frequencies of visual stimulation and high-order harmonics are usually between 8 Hz and 80 Hz, so the multi-channel EEG signal Xn is subjected to a wide range of band-pass filtering to obtain a signal with a frequency between 8 Hz and 80 Hz.
  • the original steady-state visual evoked potential Xn is preprocessed by using N fb filter banks, that is, the original steady-state visual evoked potential Xn is divided into different sub-band signals to obtain N fb sub-band signals of specific frequency bands.
  • the above-mentioned operation of preprocessing the original steady-state visual evoked potential Xn by using the filter bank is not repeated here.
  • the original steady-state visual evoked potential Xn of the subject's multiple EEG channels is collected.
  • the visual stimulus is presented in the form of flashes, and one trial is a flash lasting 2 seconds (the appropriate duration can also be selected according to the situation). Within these 2 seconds, it flashes multiple times, thereby inducing the subject's original steady-state visual evoked potential Xn .
  • the steady-state visual evoked potential (SSVEP) is a periodically repeated sinusoidal signal, and the SSVEP is formed by the temporal superposition of transient visual evoked potentials.
  • PRC periodically repeated component
  • PRCA periodically repeated component analysis
  • the number of sub-cycle sampling points is calculated using fn , so the number of sub-cycle sampling points Ln of the original steady-state visual evoked potential Xn can also be applied to the preprocessed steady-state visual evoked potential Bit Split according to the number of sub-period sampling points L n Get a new three-dimensional tensor It contains data of N t P n cycles, which can be called The periodically repeating component (PRC) matrix of .
  • PRC periodically repeating component
  • the spatial resolution of the collected subject's scalp EEG signal is not high, only reaching the centimeter level.
  • the collected EEG signal is also very weak, generally at the microvolt level, and the noise signal is larger than the useful signal amplitude, so the signal-to-noise ratio of the collected scalp EEG signal is relatively low.
  • the signal-to-noise ratio of scalp EEG signals will affect the subsequent extraction and pattern recognition of EEG feature signals, and then affect the performance of the brain-computer interface.
  • the basic principle of the spatial filtering method is: convert the EEG signals recorded by different EEG channels, enhance the intensity of specific signal components, and weaken the common noise in the EEG channels to obtain a projection method that can better distinguish useful signals from noise signals, that is, a spatial filter.
  • the spatial filter in this application is an N c -dimensional vector, and each component corresponds to the weight of an EEG channel.
  • Get the spatial filter of the mth sub-band include:
  • the spatial filter for the mth sub-band of N f frequencies Integrate to obtain the integrated spatial filter of the mth sub-band That is, N f spatial filters with the same m Combined use to obtain a combined spatial filter
  • SSVEP steady-state visual evoked potential
  • synthesize a steady-state visual evoked potential template signal of the desired length include:
  • the unknown steady-state visual evoked potential K is subjected to the training integrated spatial filter.
  • the same N fb filter groups are used for preprocessing to obtain the unknown steady-state visual evoked potential K (m) after preprocessing.
  • the meaning of the preprocessing is detailed in S310 and will not be repeated here.
  • Calculate the correlation coefficient of N fb sub-bands include:
  • the correlation coefficient is calculated according to the following formula
  • ⁇ represents the calculation of the two-dimensional correlation coefficient.
  • K (m) of the unknown steady-state visual evoked potential K and the mth sub-band signal of the EEG signal induced by the visual stimulus with frequency fn
  • the unknown steady-state visual evoked potential K is obtained by comparing it with the original steady-state visual evoked potential Xn.
  • the correlation coefficients ⁇ n between them include:
  • the correlation coefficient ⁇ n is obtained according to the following formula
  • the correlation coefficient ⁇ n between the unknown steady-state visual evoked potential K and the original steady-state visual evoked potential X n is obtained by weighted summation.
  • a(m) is the correlation coefficient of each specific frequency band
  • the weighting coefficient can be determined by grid search based on the final recognition accuracy as the objective function.
  • the above method for obtaining the correlation coefficient ⁇ n between the unknown steady-state visual evoked potential K and the original steady-state visual evoked potential X n is called an Ensemble Periodically Repeated Component Analysis (ePRCA) algorithm.
  • ePRCA Ensemble Periodically Repeated Component Analysis
  • Figure 4 shows the percentage of recognition accuracy of unknown visual stimuli using different algorithms.
  • the comparison results of the unknown visual stimulus recognition accuracy of the ePRCA algorithm described in the application and the ensemble TRCA (eTRCA) algorithm of the related technology are as follows: when the trial length Ttest is 0.5s and 1s, the recognition accuracy of the ePRCA algorithm of the present application is always better than that of the eTRCA algorithm under different numbers of EEG channels and different numbers of trials. And with the decrease of the number of training trials and the number of EEG channels, the recognition accuracy advantage of the present application compared to the eTRCA algorithm is more prominent.
  • the method of the present application can obtain a recognition accuracy equivalent to the eTRCA algorithm training trial number of 5 when the number of training trials is 1.
  • the number of training trials is the number of unit durations of a single frequency visual stimulus applied to the subject when the spatial filter is trained.
  • the number of EEG channels is the number of electrodes worn at different positions on the subject's head.
  • Figure 5(a) is a schematic diagram showing the effect of the test data length of different algorithms with a trial number of 1 on the recognition accuracy.
  • Figure 5(b) is a schematic diagram showing the effect of the test data length of different algorithms with a trial number of 1 on the information transmission efficiency, where T train is the data length of the training trial.
  • T train is the data length of the training trial.
  • the results of Figure 5 show that the recognition accuracy and information transmission efficiency of the ePRCA algorithm of the present application are always higher than those of the eTRCA algorithm.
  • Figure 4(a) and Figure 5(b) it can be seen that the ePRCA algorithm of the present application can solve a more reliable spatial filter with a small number of trials, thereby enabling SSVEP-BCI to obtain better recognition effects.
  • Figure 6 is a comparison chart of the recognition accuracy of the ePRCA algorithm and the eTRCA algorithm at different trial numbers.
  • the recognition accuracy of the ePRCA algorithm described in this application has advantages under different trial numbers, especially when the trial number is 1, the average recognition accuracy of the ePRCA algorithm exceeds 90%, while the average recognition accuracy of the eTRCA algorithm does not exceed 40%.
  • the technical solution of this embodiment is to present N f visual stimuli with frequencies of f 1 ,...f i ,...f n ,,...f N f on the screen, and collect the original steady-state visual evoked potential signals of the subjects looking at these visual stimuli as training data.
  • the original steady-state visual evoked potential of the subjects' multi-electroencephalogram channels collected when the subjects look at the visual stimulus with a frequency of f n is
  • the original steady-state visual evoked potential Xn is preprocessed using Nfb filter banks to obtain preprocessed steady-state visual evoked potentials of Nfb sub-bands.
  • the i-th correlation coefficient ⁇ i is the largest, and the visual stimulation frequency of the unknown steady-state visual evoked potential K is fi .
  • Figure 7 is a schematic diagram of the structure of a low-training-cost EEG feature decoding device for steady-state visual evoked potentials provided in Example 2 of the present application.
  • the EEG feature decoding device can execute the low-training-cost EEG feature decoding method for steady-state visual evoked potentials provided in Example 1 of the present application, and has the corresponding functional modules and effects of the execution method.
  • a low-training-cost electroencephalogram feature decoding device for steady-state visual evoked potential comprising:
  • the preprocessing module 610 is configured to present N f visual stimuli with frequencies of f 1 , ...f i , ...f n , ...f Nf on the screen, and collect the original steady-state visual evoked potential signals of the subject looking at these visual stimuli respectively.
  • Nc is the number of EEG channels
  • Ns is the number of sampling points
  • Nt is the number of trials
  • Nfb filter banks are used to preprocess the original steady-state visual evoked potential Xn to obtain the preprocessed steady-state visual evoked potential of Nfb sub-bands
  • m 1, 2, ..., N fb ;
  • the sub-period division module 620 is configured to divide the pre-processed steady-state visual evoked potential into Divided into P n sub-periods
  • the integrated spatial filter acquisition module 630 is configured according to Get the spatial filter of the mth sub-band The spatial filter for the mth sub-band of N f frequencies Integrate to obtain the integrated spatial filter of the mth sub-band
  • the template signal acquisition module 640 is configured to Taking P n sub-periods as an average, we get Copy multiple connections Synthesize the steady-state visual evoked potential template signal of the required length
  • the correlation coefficient calculation module 660 is configured to calculate the correlation coefficient by integrating the spatial filter. and steady-state visual evoked potential template signal Calculate the correlation coefficient of N fb sub-bands For N fb The weighted sum is used to obtain the correlation coefficient ⁇ n between the unknown steady-state visual evoked potential K and the original steady-state visual evoked potential X n ;
  • the visual stimulation frequency acquisition module 670 is configured to obtain the correlation coefficients [ ⁇ 1 , ... ⁇ i , ..., ⁇ n , ..., ⁇ Nf ] between the unknown steady-state visual evoked potential K and all original steady-state visual evoked potentials [X 1 , ... Xi , ..., Xn , ..., XNf ] with frequencies f 1 , ... fi , ...fn , ... fNf ] respectively.
  • the i-th correlation coefficient ⁇ i is the largest, and the unknown steady-state visual evoked potential K is the original steady-state visual evoked potential K.
  • the visual stimulation frequency of the sensory evoked potential K is fi .
  • the sub-period division module 620 is configured as follows:
  • the template signal acquisition module 640 is configured as:
  • the integrated spatial filter acquisition module 630 is configured as:
  • the correlation coefficient calculation module 660 is set to:
  • the correlation coefficient is calculated according to the following formula
  • ⁇ represents the calculation of the two-dimensional correlation coefficient.
  • the correlation coefficient calculation module 660 is set to:
  • the correlation coefficient ⁇ n is obtained according to the following formula
  • the technical solution of this embodiment solves the technical problem that when the eTRCA algorithm is used, the recognition performance drops significantly when there is less training data, especially when there is only a single trial training data, the recognition performance is very poor, and achieves the technical effect of improving the accuracy of recognizing unknown visual stimuli.
  • Figure 8 is a structural schematic diagram of a low-training-cost brain-computer interface system for steady-state visual evoked potentials provided in Example 3 of the present application.
  • the system includes a processor 710, a memory 720, an input device 730, and an output device 740; the number of processors 710 in the system can be at least one, and Figure 8 takes one processor 710 as an example; the processor 710, memory 720, input device 730, and output device 740 in the system can be connected via a bus or other means, and Figure 8 takes connection via a bus as an example.
  • the memory 720 can be used to store software programs, computer executable programs and modules, such as the program instructions/modules corresponding to the low-training-cost EEG feature decoding method for steady-state visual evoked potentials in the embodiments of the present application (for example, a preprocessing module 610, a sub-cycle division module 620, an integrated spatial filter acquisition module 630, a template signal acquisition module 640, an unknown steady-state visual evoked potential acquisition module 650, a correlation coefficient calculation module 660 and a visual stimulation frequency acquisition module 670 in a low-training-cost EEG feature decoding device for steady-state visual evoked potentials).
  • a preprocessing module 610 for example, a sub-cycle division module 620, an integrated spatial filter acquisition module 630, a template signal acquisition module 640, an unknown steady-state visual evoked potential acquisition module 650, a correlation coefficient calculation module 660 and a visual stimulation frequency acquisition module 670 in a low-training-cost EEG feature decoding device for steady-state visual
  • the processor 710 executes various functional applications and data processing of the system by running the software programs, instructions and modules stored in the memory 720, that is, realizing the above-mentioned low-training-cost EEG feature decoding method for steady-state visual evoked potentials.
  • the memory 720 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system and at least one application required for a function; the data storage area may store data created according to the use of the terminal, etc.
  • the memory 720 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device.
  • the memory 720 may include a memory remotely arranged relative to the processor 710, and these remote memories may be connected to the system via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.
  • the input device 730 may be used to receive input digital or character information and generate key signal input related to user settings and function control of the system.
  • the output device 740 may include display devices such as a display screen.
  • Embodiment 4 of the present application further provides a computer-readable storage medium containing a computer program, wherein when the computer program is executed by a processor, it is configured to execute a low-training-cost EEG feature decoding method for steady-state visual evoked potentials, the method comprising:
  • N f visual stimuli with frequencies of f 1 ,...f i ,...f n ,,...f N f are presented on the screen, and the original steady-state visual evoked potential signals of the subjects looking at these visual stimuli are collected as training data.
  • Nc is the number of EEG channels
  • Ns is the number of sampling points
  • Nt is the number of trials
  • Nfb filter banks are used to preprocess the original steady-state visual evoked potential Xn to obtain the preprocessed steady-state visual evoked potential of Nfb sub-bands
  • m 1, 2, ..., N fb ;
  • the computer-readable storage medium containing a computer program provided in an embodiment of the present application is not limited to the method operations described above, and can also execute related operations in the low-training-cost EEG feature decoding method for steady-state visual evoked potentials provided in any embodiment of the present application.
  • the low-training-cost EEG feature decoding method for steady-state visual evoked potentials described in this application can also be applied to signal-to-noise ratio enhancement and signal recognition of steady-state auditory evoked potentials, steady-state somatosensory evoked potentials and other types of signals, and the corresponding brain-computer interface system can be built based on this.
  • the technicians in the relevant field can clearly understand that the present application can be implemented with the help of software and necessary general hardware, and of course it can also be implemented by hardware, but in many cases the former is a better implementation method.
  • the technical solution of the present application can be essentially or the part that contributes to the relevant technology can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk or optical disk, etc., including multiple instructions for a computer device (which can be a personal computer, server, or network device, etc.) to execute the method described in the embodiment of the present application.
  • a computer-readable storage medium such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk or optical disk, etc.

Landscapes

  • Engineering & Computer Science (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Physics & Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Computation (AREA)
  • Medical Informatics (AREA)
  • General Health & Medical Sciences (AREA)
  • General Physics & Mathematics (AREA)
  • Biophysics (AREA)
  • Pathology (AREA)
  • Biomedical Technology (AREA)
  • Heart & Thoracic Surgery (AREA)
  • General Engineering & Computer Science (AREA)
  • Molecular Biology (AREA)
  • Surgery (AREA)
  • Animal Behavior & Ethology (AREA)
  • Evolutionary Biology (AREA)
  • Public Health (AREA)
  • Veterinary Medicine (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Psychiatry (AREA)
  • Mathematical Physics (AREA)
  • Psychology (AREA)
  • Signal Processing (AREA)
  • Physiology (AREA)
  • Fuzzy Systems (AREA)
  • Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)

Abstract

本申请公开了面向稳态视觉诱发电位的低训练成本脑电特征解码方法，该方法包括：采集受试者的原始稳态视觉诱发电位信号作为训练数据；采用滤波器组对原始稳态视觉诱发电位Xn进行预处理，得到预处理后稳态视觉诱发电位Xn (m)；将预处理后稳态视觉诱发电位Xn (m)划分为Pn个子周期（I）；根据（I）得到第m子频段的空间滤波器（II）；对空间滤波器（II）进行集成，得到集成空间滤波器（III）；对（IV）进行平均，得到（V）；复制连接多个（V），合成得到所需长度的稳态视觉诱发电位模板信号（VI）；通过集成空间滤波器（III）和稳态视觉诱发电位模板信号（VI）对未知稳态视觉诱发电位K进行识别。

Description

面向稳态视觉诱发电位的低训练成本脑电特征解码方法
本申请要求在2023年8月7日提交中国专利局、申请号为202310981789.2的中国专利申请的优先权，该申请的全部内容通过引用结合在本申请中。
技术领域
本申请实施例涉及脑机接口信号的编码及解码技术，例如涉及一种面向稳态视觉诱发电位的低训练成本脑电特征解码方法及脑机接口系统。
背景技术
近年来，随着科研工作者对脑电波(Electroencephalogram，EEG)的深入研究，基于头皮脑电的非侵入式脑机接口(Brain Computer Interface，BCI)得到了突飞猛进的发展。其中，基于视觉诱发电位(Visually Evoked Potential，VEP)的BCI因其较高的信息传输率和解码可靠性受到广泛关注。稳态视觉诱发电位(Steady-State Visual Evoked Potentials，SSVEP)BCI范式更因其具有诱发特征稳定、信噪比高等特点，已成为BCI的主流范式之一。
相关技术中，以任务相关成分分析(Task-Related Component Analysis，TRCA)算法为代表的SSVEP解码算法需要依赖大量的训练数据才能获得较高识别精度。相对于使用标准正余弦作为模板而言，预先训练有助于算法从个体的脑电数据中学习并生成符合被试脑电特征的空间滤波器和脑电模板信号，从而使算法获得满意的识别结果。由于TRCA算法本身的原理是最大化试次间的重复信号成分，因此当训练数据较少时识别性能往往会大幅度下降，特别是在每个频率仅有单试次训练数据时，识别性能很差。
发明内容
本申请提供一种面向稳态视觉诱发电位的低训练成本脑电特征解码方法及脑机接口，以实现提高未知视觉刺激识别准确率的技术效果。
第一方面，本申请实施例提供了一种面向稳态视觉诱发电位的低训练成本脑电特征解码方法，包括：
在屏幕上呈现Nf个频率分别为f1,…fi,…fn,,…fNf的视觉刺激，采集受试者分别注视这些视觉刺激的原始稳态视觉诱发电位信号作为训练数据，受试者注视频率为fn的视觉刺激时所采集的受试者的多脑电通道的原始稳态视觉诱发电位为其中，Nc为脑电通道数，Ns为采样点数，Nt为试次数；采用Nfb个滤波器组对所述原始稳态视觉诱发电位Xn进行预处理，得到Nfb个子频段的预处理后稳态视觉诱发电位其中，m＝1,2,…,Nfb；
将所述预处理后稳态视觉诱发电位划分为Pn个子周期
根据得到第m子频段的空间滤波器对Nf个频率的第m子频段的空间滤波器进行集成，得到第m子频段的集成空间滤波器
对以Pn个子周期进行平均，得到复制连接多个合成得到所需长度的稳态视觉诱发电位模板信号
采集受试者的未知稳态视觉诱发电位采用所述Nfb个滤波器组对所述未知稳态视觉诱发电位K进行预处理，得到Nfb个子频段的预处理后未知稳态视觉诱发电位K(m)，其中，m＝1,2,…,Nfb；
通过所述集成空间滤波器和所述稳态视觉诱发电位模板信号计算Nfb个子频段的相关系数对Nfb个加权求和得到所述未知稳态视觉诱发电位K与所述原始稳态视觉诱发电位Xn之间的相关系数ρn；
得到所述未知稳态视觉诱发电位K与频率分别为f1,…fi,…fn,,…fNf的全部原 始稳态视觉诱发电位[X1,…Xi,…,Xn,...,XNf]之间的相关系数[ρ1,…ρi,…,ρn,…,ρNf]，第i个相关系数ρi最大，则所述未知稳态视觉诱发电位K的视觉刺激频率为fi。
第二方面，本申请还提供了一种面向稳态视觉诱发电位的低训练成本脑电特征解码装置，包括：
预处理模块，设置为在屏幕上呈现Nf个频率分别为f1,…fi,…fn,,…fNf的视觉刺激，采集受试者分别注视这些视觉刺激的原始稳态视觉诱发电位信号作为训练数据，受试者注视频率为fn的视觉刺激时所采集的受试者的多脑电通道的原始稳态视觉诱发电位为其中，Nc为脑电通道数，Ns为采样点数，Nt为试次数；采用Nfb个滤波器组对所述原始稳态视觉诱发电位Xn进行预处理，得到Nfb个子频段的预处理后稳态视觉诱发电位其中，m＝1,2,…,Nfb；
子周期划分模块，设置为将预处理后稳态视觉诱发电位划分为Pn个子周期
集成空间滤波器获取模块，设置为根据得到第m子频段的空间滤波器对Nf个频率的第m子频段的空间滤波器进行集成，得到第m子频段的集成空间滤波器
模板信号获取模块，设置为对以Pn个子周期进行平均，得到复制连接多个合成得到所需长度的稳态视觉诱发电位模板信号
未知稳态视觉诱发电位采集模块，设置为采集受试者的未知稳态视觉诱发电位采用所述Nfb个滤波器组对所述未知稳态视觉诱发电位K进行预处理，得到Nfb个子频段的预处理后未知稳态视觉诱发电位K(m)，其中，m＝1,2,…,Nfb；
相关系数计算模块，设置为通过集成空间滤波器和稳态视觉诱发电位 模板信号计算Nfb个子频段的相关系数对Nfb个加权求和得到未知稳态视觉诱发电位K与所述原始稳态视觉诱发电位Xn之间的相关系数ρn；
视觉刺激频率获取模块，设置为得到未知稳态视觉诱发电位K与频率分别为f1,…fi,…fn,,…fNf的全部原始稳态视觉诱发电位[X1,…Xi,…,Xn,...,XNf]之间的相关系数[ρ1,…ρi,…,ρn,…,ρNf]，第i个相关系数ρi最大，则所述未知稳态视觉诱发电位K的视觉刺激频率为fi。
第三方面，本申请实施例还提供了一种面向稳态视觉诱发电位的低训练成本脑机接口系统，包括：
至少一个处理器；
存储器，设置为存储至少一个程序，
当所述至少一个程序被所述至少一个处理器执行时，使得所述至少一个处理器实现如实施例中任一所述的面向稳态视觉诱发电位的低训练成本脑电特征解码方法。
第四方面，本申请实施例还提供了一种包含计算机程序的计算机可读存储介质，所述计算机程序在由处理器执行时实现如实施例中任一所述的面向稳态视觉诱发电位的低训练成本脑电特征解码方法。
附图说明
图1为稳态视觉诱发电位(SSVEP)脑机接口系统的组成结构示意图；
图2为典型的稳态视觉诱发电位在频域上的信噪比特征示意图；
图3为本申请实施例一提供的一种面向稳态视觉诱发电位的低训练成本脑电特征解码方法的流程图；
图4为不同算法的未知视觉刺激的识别正确率百分比图；
图5(a)为试次数为1的不同算法的测试数据长度对识别正确率的影响示意图；
图5(b)为试次数为1的不同算法的测试数据长度对信息传输效率的影响示意图；
图6为不同试次数的ePRCA算法和eTRCA算法的识别正确率对比图；
图7为本申请实施例二提供的一种面向稳态视觉诱发电位的低训练成本脑电特征解码装置的结构示意图；
图8为本申请实施例三提供的一种面向稳态视觉诱发电位的低训练成本脑机接口系统的结构示意图。
具体实施方式
实施例一
基于头皮脑电的BCI可以无创地建立大脑与外部设备之间的信息传输通道，从而帮助患有严重运动障碍的病人或者其他有需求的健康人群与外界环境进行友好沟通。图1为稳态视觉诱发电位(SSVEP)脑机接口系统的组成结构示意图，该系统主要包括刺激模块、采集模块以及处理模块。其中，刺激模块是指由显示器编码出固定频率的视觉刺激。当受试者注视视觉刺激时，在受试者的大脑枕区能够诱发出稳态视觉诱发电位。采集模块通过电极从受试者的大脑皮层或大脑颅腔内获取表征大脑思维及认知活动的脑电生理信号。处理模块则通过一系列的信号处理方法从脑电生理信号中提取出相应的脑电特征信号并进行模式识别，从而将脑电特征信号转换为机器语言指令。
图2为典型的稳态视觉诱发电位在频域上的信噪比特征示意图。稳态视觉诱发电位(SSVEP)信号是由固定频率的视觉刺激所诱发的，当视觉刺激以特定频率进行周期性呈现(如闪烁、图形翻转、图形缩放等)，受试者的视觉系统会受到影响而产生稳定频域特征的诱发响应，该响应包含视觉刺激的频率成分及其高次谐波成分。图2为多个试次的稳态视觉诱发电位平均之后的结果，多个试次进行平均能够抵消随机噪声、提高信噪比。在单个试次的情况下，可能并不能达到图2所示的在视觉刺激频率和高次谐波频率处的高信噪比效果。 稳态视觉诱发电位还具有空域特征，例如，SSVEP信号的不同频率成分来自于受试者不同的脑区，因此当SSVEP信号传导至受试者头皮表面的不同处时各次谐波的信噪比也不尽相同。
图3为本申请实施例一提供的一种面向稳态视觉诱发电位的低训练成本脑电特征解码方法的流程图。本实施例可适用于提供一种空间滤波器来对稳态视觉诱发电位的脑电特征进行解码，该空间滤波器的训练可以基于较少试次的实验数据，甚至基于单次实验数据。该方法可以由计算机执行，例如包括如下步骤：
其中，步骤S310-步骤S340为训练得到空间滤波器的过程，步骤S350-步骤S370则利用训练好的空间滤波器对未知稳态视觉诱发电位进行滤波分析，从而得到未知稳态视觉诱发电位所表征的含义，即得到受试者所注视的视觉刺激的频率信息。
S310、在屏幕上呈现Nf个频率分别为f1,…fi,…fn,,…fNf的视觉刺激，采集受试者分别注视这些视觉刺激的原始稳态视觉诱发电位信号作为训练数据，受试者注视频率为fn的视觉刺激时所采集的受试者的多脑电通道的原始稳态视觉诱发电位为其中，Nc为脑电通道数，Ns为采样点数，Nt为试次数；采用Nfb个滤波器组对所述原始稳态视觉诱发电位Xn进行预处理，得到Nfb个子频段的预处理后稳态视觉诱发电位其中，m＝1,2,…,Nfb。
假设受试者关注频率为fn的视觉刺激，采集受试者此时的脑电信号，得到原始稳态视觉诱发电位其中，n＝1,2,…,Nf；Nf为视觉刺激的频率类型数，Nc、Ns、Nt分别为脑电通道数、采样点数以及试次数。
可选的，在对所述原始稳态视觉诱发电位Xn进行预处理之前，可以对采集到的多脑电通道脑电信号Xn进行带通滤波和工频陷波，滤除掉明显的噪声信号，例如，视觉刺激与高次谐波的频率通常在8Hz～80Hz之间，则对多脑电通道脑电信号Xn进行大范围的带通滤波，得到频率为8Hz～80Hz之间的信号。
采用Nfb个滤波器组对所述原始稳态视觉诱发电位Xn进行预处理，即将原始稳态视觉诱发电位Xn进行频率划分，得到不同的子带信号，得到Nfb个特定频段的子带信号其中，m＝1,2,…,Nfb；Nfb为滤波器组的个数，滤波器组中的滤波器可根据实际需要进行设置。例如，m＝2的滤波器组中有8个滤波器，通带分别为16Hz～80Hz、24Hz～80Hz、32Hz～80Hz、40Hz～80Hz、48Hz～80Hz、56Hz～80Hz、64Hz～80Hz和72Hz～80Hz。上述采用滤波器组对所述原始稳态视觉诱发电位Xn进行预处理的操作在此不再赘述。
S320、将预处理后稳态视觉诱发电位划分为Pn个子周期
受试者注视频率为fn的视觉刺激时，采集受试者的多脑电通道的原始稳态视觉诱发电位Xn。例如，该视觉刺激以闪烁的形式呈现，一个试次为持续2秒闪烁(也可以根据情况选择合适的时长)，在这2s内，闪烁了多次从而诱发了受试者的原始稳态视觉诱发电位Xn。与多次闪烁相对应的，稳态视觉诱发电位(SSVEP)是周期性重复的正弦类信号，且SSVEP是由瞬态视觉诱发电位的时间叠加形成的。考虑到以上事实，本申请提出了最大化SSVEP的周期性重复成分(Periodically Repeated Component，PRC)的方法，即周期性重复成分分析(Periodically Repeated Component Analysis，PRCA)算法，该算法改进了TRCA算法的性能。
可选的，将预处理后稳态视觉诱发电位划分为Pn个子周期包括：将原始稳态视觉诱发电位为Xn划分为Pn个子周期，Pn＝floor(Nsfn/Fs)，每个子周期采样点的数量为Ln＝round(Fs/fn)，其中，Fs为采样频率，floor为向下取整函数，round为四舍五入函数；根据子周期采样点的数量Ln分割得到
由于在计算协方差矩阵时，高次谐波的周期长度太短容易受噪声影响，因此对于所有的子周期采样点的数量均采用fn计算，因而原始稳态视觉诱发电位Xn的子周期采样点的数量Ln同样可以应用于预处理后稳态视觉诱发电 位根据子周期采样点的数量Ln分割得到新的三维张量 其中包含NtPn个周期的数据，可将称为的周期性重复成分(PRC)矩阵。
通过应用floor函数和round函数，给出了对预处理后稳态视觉诱发电位进行子周期划分的一种方案，进而得到NtPn个周期的数据。
S330、根据得到第m子频段的第p子周期空间滤波器对Nf个频率的第m子频段的空间滤波器进行集成，得到第m子频段的集成空间滤波器
由于容积导体效应与封闭电场的影响，采集受试者头皮脑电信号的空间分辨率不高，只能达到厘米级，同时采集到的脑电信号也非常微弱，一般在微伏级别，而噪声信号相对于有用信号幅度偏大，故采集得到的头皮脑电信号的信噪比较低。
头皮脑电信号的信噪比会影响后续脑电特征信号的提取和模式识别，继而影响脑机接口的性能。研究证明空间滤波方法在降低SSVEP脑电信号的信噪比上发挥了重要作用，空间滤波方法的基本原理是：将不同脑电通道记录到的脑电信号进行信号转换，增强特定信号成分的强度，减弱脑电通道中的共有噪声，得到可以较好地区分有用信号与噪声信号的投影方式，即空间滤波器。例如，本申请中的空间滤波器是Nc维的向量，每一个分量对应一个脑电通道的权重。
可选的，根据得到第m子频段的空间滤波器包括：
根据如下广义特征值分解进行计算

其中，这里是跨周期协方差矩阵的总和，是周期自协方差的总和，和表示中第p1和第p2个周期。函数Cov表示协方差运算，为了使方程(1)具有有限解，方程(1)的分母被限制为1，那么得到的空间滤波器就是矩阵的最大特征值对应的特征向量。
上述计算过程给出了根据求解第m子频段的空间滤波器的计算过程。
对Nf个频率的第m子频段的空间滤波器进行集成，得到第m子频段的集成空间滤波器也就是说，将相同m的Nf个空间滤波器组合使用，得到组合空间滤波器
S340、对以Pn个子周期进行平均，得到复制连接多个合成得到所需长度的稳态视觉诱发电位模板信号
计算稳态视觉诱发电位(SSVEP)信号的子周期的模板，即周期性重复成分 为了可以同TRCA算法那样应用基于SSVEP模板的模板匹配方法，可以通过复制和连接来重建所需长度的SSVEP模板重建获取的理论依据在于，不考虑可能的瞬时响应时，SSVEP本质上是由PRC组成的。
可选的，合成得到所需长度的稳态视觉诱发电位模板信号包括：
复制连接多个合成得到所需长度的稳态视觉诱发电位模板信号 其中， 为的 一部分，的采样点数为Ns。
是矩阵从第一个采样点到第l个采样点的部分。在后续未知稳态视觉诱发电位K的采样点数为Ns的情况下，的采样点数也合成为Ns，其中，n＝1,2,……,Nf；m＝1,2,…,Nfb。
根据后续未知稳态视觉诱发电位K的采样点数来合成所需长度的稳态视觉诱发电位模板信号可以根据情况选择合适的采样点数，增加了的适用范围。
S350、采集受试者的未知稳态视觉诱发电位采用所述Nfb个滤波器组对所述未知稳态视觉诱发电位K进行预处理，得到Nfb个子频段的预处理后未知稳态视觉诱发电位K(m)，其中，m＝1,2,…,Nfb。
对未知稳态视觉诱发电位K采用与训练集成空间滤波器时相同的Nfb个滤波器组进行预处理，得到预处理后未知稳态视觉诱发电位K(m)，该预处理的含义，详见S310，在此不再赘述。
S360、通过集成空间滤波器和稳态视觉诱发电位模板信号计算Nfb个子频段的相关系数对Nfb个加权求和得到未知稳态视觉诱发电位K与所述原始稳态视觉诱发电位Xn之间的相关系数ρn。
可选的，通过集成空间滤波器和稳态视觉诱发电位模板信号计算Nfb个子频段的相关系数包括：
根据如下公式求取相关系数
其中，ρ表示计算二维相关系数。为未知稳态视觉诱发电位K的第m个子频段信号K(m)与频率为fn的视觉刺激所诱发的脑电信号的第m个子频段信号之间的相关系数。
可选的，得到未知稳态视觉诱发电位K与所述原始稳态视觉诱发电位Xn之 间的相关系数ρn，包括：
根据如下公式求取相关系数ρn
其中，a(m)被定义为a(m)＝m-s+β，α和β取值可通过网格搜索确定最优参数，也可取经验值s＝1.25，β＝0.25。
对Nfb个加权求和得到所述未知稳态视觉诱发电位K与所述原始稳态视觉诱发电位Xn之间的相关系数ρn，其中，a(m)为各个特定频段的相关系数的加权系数，该加权系数可以最终识别正确率为目标函数根据网格搜索确定。
上述得到未知稳态视觉诱发电位K与原始稳态视觉诱发电位Xn之间的相关系数ρn的方法称为集成周期性重复成分分析(Ensemble Periodically Repeated Component Analysis，ePRCA)算法。
S370、得到未知稳态视觉诱发电位K与频率分别为f1,…fi,…fn,,…fNf的全部原始稳态视觉诱发电位[X1,…Xi,…,Xn,...,XNf]之间的相关系数[ρ1,…ρi,…,ρn,…,ρNf]，第i个相关系数ρi最大，则所述未知稳态视觉诱发电位K的视觉刺激频率为fi。
稳态视觉诱发电位识别本质上是为了确定所采集的未知稳态视觉诱发电位是受试者注视哪一频率视觉刺激所诱发的信号。因此，在得到相关系数ρn之后，根据全部Nf个ρn中的最大值确定未知稳态视觉诱发电位K是哪一频率视觉刺激对应的信号，即识别结果为τ＝maxnρn，未知稳态视觉诱发电位K与哪一频率的原始稳态视觉诱发电位的相关系数最大，则该未知稳态视觉诱发电位K即为受试者关注哪一频率的视觉刺激所诱发的信号，即实现了对未知视觉刺激的识别。
图4为不同算法的未知视觉刺激的识别正确率百分比图。如图4所示，本 申请所述的ePRCA算法与相关技术的集成任务相关成分分析(ensemble TRCA，eTRCA)算法的未知视觉刺激识别正确率的比较结果如下：在试次长度Ttest为0.5s和1s时，不同脑电通道数与不同试次数下，本申请的ePRCA算法的识别正确率始终优于eTRCA算法。并且随着训练试次数和脑电通道数的减少，本申请相比较于eTRCA算法的识别正确率优势更加突出。值得关注的是，本申请方法在训练试次数为1的情况下能够获得与eTRCA算法训练试次数为5相当的识别正确率。其中，训练试次数为在对空间滤波器进行训练时，对受试者施加单个频率视觉刺激的单位时长的次数。脑电通道数为佩戴在受试者头部不同位置处的电极数量。
图5(a)为试次数为1的不同算法的测试数据长度对识别正确率的影响示意图。图5(b)为试次数为1的不同算法的测试数据长度对信息传输效率的影响示意图，其中Ttrain为训练试次的数据长度。图5的结果表明本申请的ePRCA算法的识别正确率和信息传输效率始终高于eTRCA算法。结合图4、图5(a)和图5(b)，可以看出本申请的ePRCA算法能够在试次数少的情况下求解出更可靠空间滤波器，从而使SSVEP-BCI获得更好的识别效果。
图6为不同试次数的ePRCA算法和eTRCA算法的识别正确率对比图。如图6所示，本申请所述的ePRCA算法的识别正确率在不同试次数下均有优势，特别是在试次数为1时，ePRCA算法平均识别正确率超过90％，而eTRCA算法的平均识别正确率不超过40％。
本实施例的技术方案，通过在屏幕上呈现Nf个频率分别为f1,…fi,…fn,,…fNf的视觉刺激，采集受试者分别注视这些视觉刺激的原始稳态视觉诱发电位信号作为训练数据，受试者注视频率为fn的视觉刺激时所采集的受试者的多脑电通道的原始稳态视觉诱发电位为采用Nfb个滤波器组对所述原始稳态视觉诱发电位Xn进行预处理，得到Nfb个子频段的预处理后稳态视觉诱发电位将预处理后稳态视觉诱发电位划分为Pn个子周期根据 得到第m子频段的空间滤波器对Nf个频率的第m子频段的空间滤波器进行集成，得到第m子频段的集成空间滤波器 对以Pn个子周期进行平均，得到复制连接多个合成得到所需长度的稳态视觉诱发电位模板信号采集受试者的未知稳态视觉诱发电位采用所述Nfb个滤波器组对所述未知稳态视觉诱发电位K进行预处理，得到Nfb个子频段的预处理后未知稳态视觉诱发电位K(m)；通过集成空间滤波器和稳态视觉诱发电位模板信号计算Nfb个子频段的相关系数对Nfb个加权求和得到未知稳态视觉诱发电位K与所述原始稳态视觉诱发电位Xn之间的相关系数ρn；得到未知稳态视觉诱发电位K与频率分别为f1,…fi,…fn,,…fNf的全部原始稳态视觉诱发电位[X1,…Xi,…,Xn,...,XNf]之间的相关系数[ρ1,…ρi,…,ρn,…,ρNf]，第i个相关系数ρi最大，则所述未知稳态视觉诱发电位K的视觉刺激频率为fi。解决了使用TRCA算法时当训练数据较少时识别性能大幅度下降，特别是在仅有单试次训练数据时，识别性能很差的技术问题，达到了提高未知视觉刺激识别准确率的技术效果。
实施例二
图7为本申请实施例二提供的一种面向稳态视觉诱发电位的低训练成本脑电特征解码装置的结构示意图。该脑电特征解码装置可以执行本申请实施例一所提供的面向稳态视觉诱发电位的低训练成本脑电特征解码方法，具备执行方法相应的功能模块和效果。
一种面向稳态视觉诱发电位的低训练成本脑电特征解码装置，包括：
预处理模块610，设置为在屏幕上呈现Nf个频率分别为f1,…fi,…fn,,…fNf的视觉刺激，采集受试者分别注视这些视觉刺激的原始稳态视觉诱发电位信号作 为训练数据，受试者注视频率为fn的视觉刺激时所采集的受试者的多脑电通道的原始稳态视觉诱发电位为其中，Nc为脑电通道数，Ns为采样点数，Nt为试次数；采用Nfb个滤波器组对所述原始稳态视觉诱发电位Xn进行预处理，得到Nfb个子频段的预处理后稳态视觉诱发电位其中，m＝1,2,…,Nfb；
子周期划分模块620，设置为将预处理后稳态视觉诱发电位划分为Pn个子周期
集成空间滤波器获取模块630，设置为根据得到第m子频段的空间滤波器对Nf个频率的第m子频段的空间滤波器进行集成，得到第m子频段的集成空间滤波器
模板信号获取模块640，设置为对以Pn个子周期进行平均，得到复制连接多个合成得到所需长度的稳态视觉诱发电位模板信号
未知稳态视觉诱发电位采集模块650，设置为采集受试者的未知稳态视觉诱发电位采用所述Nfb个滤波器组对所述未知稳态视觉诱发电位K进行预处理，得到Nfb个子频段的预处理后未知稳态视觉诱发电位K(m)，其中，m＝1,2,…,Nfb；
相关系数计算模块660，设置为通过集成空间滤波器和稳态视觉诱发电位模板信号计算Nfb个子频段的相关系数对Nfb个加权求和得到未知稳态视觉诱发电位K与所述原始稳态视觉诱发电位Xn之间的相关系数ρn；
视觉刺激频率获取模块670，设置为得到未知稳态视觉诱发电位K与频率分别为f1,…fi,…fn,,…fNf的全部原始稳态视觉诱发电位[X1,…Xi,…,Xn,...,XNf]之间的相关系数[ρ1,…ρi,…,ρn,…,ρNf]，第i个相关系数ρi最大，则所述未知稳态视 觉诱发电位K的视觉刺激频率为fi。
可选的，所述子周期划分模块620设置为：
将原始稳态视觉诱发电位为Xn划分为Pn个子周期，Pn＝floor(Nsfn/Fs)，每个子周期采样点的数量为Ln＝round(Fs/fn)，其中，Fs为采样频率，floor为向下取整函数，round为四舍五入函数；根据子周期采样点的数量Ln分割得到
可选的，所述模板信号获取模块640设置为：
复制连接多个合成得到所需长度的稳态视觉诱发电位模板信号 其中， 为的一部分，的采样点数为Ns。
可选的，所述集成空间滤波器获取模块630设置为：
根据如下广义特征值分解进行计算


其中，这里是跨周期协方差矩阵的总和，是周期自协方差矩阵的总和，和表示中第p1和第p2个周期。
可选的，所述相关系数计算模块660设置为：
根据如下公式求取相关系数
其中，ρ表示计算二维相关系数。
可选的，所述相关系数计算模块660设置为：
根据如下公式求取相关系数ρn
其中，a(m)被定义为a(m)＝m-s+β，s和β取值可通过网格搜索确定最优参数，也可取经验值s＝1.25，β＝0.25。
本实施例的技术方案，解决了使用eTRCA算法时当训练数据较少时识别性能大幅度下降，特别是在仅有单试次训练数据时，识别性能很差的技术问题，达到了提高未知视觉刺激识别准确率的技术效果。
实施例三
图8为本申请实施例三提供的一种面向稳态视觉诱发电位的低训练成本脑机接口系统的结构示意图，如图8所示，该系统包括处理器710、存储器720、输入装置730和输出装置740；系统中处理器710的数量可以是至少一个，图8中以一个处理器710为例；系统中的处理器710、存储器720、输入装置730和输出装置740可以通过总线或其他方式连接，图8中以通过总线连接为例。
存储器720作为一种计算机可读存储介质，可用于存储软件程序、计算机可执行程序以及模块，如本申请实施例中的面向稳态视觉诱发电位的低训练成本脑电特征解码方法对应的程序指令/模块(例如，一种面向稳态视觉诱发电位的低训练成本脑电特征解码装置中的预处理模块610、子周期划分模块620、集成空间滤波器获取模块630、模板信号获取模块640、未知稳态视觉诱发电位采集模块650、相关系数计算模块660和视觉刺激频率获取模块670)。处理器710通过运行存储在存储器720中的软件程序、指令以及模块，从而执行系统的各种功能应用以及数据处理，即实现上述的面向稳态视觉诱发电位的低训练成本脑电特征解码方法。
存储器720可主要包括存储程序区和存储数据区，其中，存储程序区可存储操作系统、至少一个功能所需的应用程序；存储数据区可存储根据终端的使用所创建的数据等。此外，存储器720可以包括高速随机存取存储器，还可以包括非易失性存储器，例如至少一个磁盘存储器件、闪存器件、或其他非易失性固态存储器件。在一些实例中，存储器720可包括相对于处理器710远程设置的存储器，这些远程存储器可以通过网络连接至系统。上述网络的实例包括但不限于互联网、企业内部网、局域网、移动通信网及其组合。
输入装置730可用于接收输入的数字或字符信息，以及产生与系统的用户设置以及功能控制有关的键信号输入。输出装置740可包括显示屏等显示设备。
实施例四
本申请实施例四还提供一种包含计算机程序的计算机可读存储介质，所述计算机程序在由处理器执行时设置为执行一种面向稳态视觉诱发电位的低训练成本脑电特征解码方法，该方法包括：
在屏幕上呈现Nf个频率分别为f1,…fi,…fn,,…fNf的视觉刺激，采集受试者分别注视这些视觉刺激的原始稳态视觉诱发电位信号作为训练数据，受试者注视频率为fn的视觉刺激时所采集的受试者的多脑电通道的原始稳态视觉诱发电位为其中，Nc为脑电通道数，Ns为采样点数，Nt为试次数；采用Nfb个滤波器组对所述原始稳态视觉诱发电位Xn进行预处理，得到Nfb个子频段的预处理后稳态视觉诱发电位其中，m＝1,2,…,Nfb；
将预处理后稳态视觉诱发电位划分为Pn个子周期
根据得到第m子频段的空间滤波器对Nf个频率的第m子频段的空间滤波器进行集成，得到第m子频段的集成空间滤波器
对以Pn个子周期进行平均，得到复制连接多个 合成得到所需长度的稳态视觉诱发电位模板信号
采集受试者的未知稳态视觉诱发电位采用所述Nfb个滤波器组对所述未知稳态视觉诱发电位K进行预处理，得到Nfb个子频段的预处理后未知稳态视觉诱发电位K(m)，其中，m＝1,2,…,Nfb；
通过集成空间滤波器和稳态视觉诱发电位模板信号计算Nfb个子频段的相关系数对Nfb个加权求和得到未知稳态视觉诱发电位K与所述原始稳态视觉诱发电位Xn之间的相关系数ρn；
得到未知稳态视觉诱发电位K与频率分别为f1,…fi,…fn,,…fNf的全部原始稳态视觉诱发电位[X1,…Xi,…,Xn,...,XNf]之间的相关系数[ρ1,…ρi,…,ρn,…,ρNf]，第i个相关系数ρi最大，则所述未知稳态视觉诱发电位K的视觉刺激频率为fi。
当然，本申请实施例所提供的一种包含计算机程序的计算机可读存储介质，其计算机程序不限于如上所述的方法操作，还可以执行本申请任意实施例所提供的面向稳态视觉诱发电位的低训练成本脑电特征解码方法中的相关操作。
需要注意的是，本申请所述的面向稳态视觉诱发电位的低训练成本脑电特征解码方法也可应用于稳态听觉诱发电位、稳态体感诱发电位等类型信号的信噪比增强和信号识别，并基于此构建相应的脑机接口系统。
通过以上关于实施方式的描述，所属领域的技术人员可以清楚地了解到，本申请可借助软件及必需的通用硬件来实现，当然也可以通过硬件实现，但很多情况下前者是更佳的实施方式。基于这样的理解，本申请的技术方案本质上或者说对相关技术做出贡献的部分可以以软件产品的形式体现出来，该计算机软件产品可以存储在计算机可读存储介质中，如计算机的软盘、只读存储器(Read-Only Memory，ROM)、随机存取存储器(Random Access Memory，RAM)、闪存(FLASH)、硬盘或光盘等，包括多个指令用以使得一台计算机设备(可以是个人计算机，服务器，或者网络设备等)执行本申请实施例所述的方法。
值得注意的是，上述面向稳态视觉诱发电位的低训练成本脑电特征解码装置的实施例中，所包括的各个单元和模块只是按照功能逻辑进行划分的，但并不局限于上述的划分，只要能够实现相应的功能即可；另外，各功能单元的名称也只是为了便于相互区分，并不用于限制本申请的保护范围。

Claims (9)

  1. 一种面向稳态视觉诱发电位的低训练成本脑电特征解码方法，包括：
    在屏幕上呈现Nf个频率分别为f1,…fi,…fn,,…fNf的视觉刺激，采集受试者分别注视这些视觉刺激的原始稳态视觉诱发电位信号作为训练数据，受试者注视频率为fn的视觉刺激时所采集的受试者的多脑电通道的原始稳态视觉诱发电位为其中，Nc为脑电通道数，Ns为采样点数，Nt为试次数；采用Nfb个滤波器组对所述原始稳态视觉诱发电位Xn进行预处理，得到Nfb个子频段的预处理后稳态视觉诱发电位其中，m＝1,2,…,Nfb；
    将所述预处理后稳态视觉诱发电位划分为Pn个子周期
    根据得到第m子频段的空间滤波器对Nf个频率的第m子频段的空间滤波器进行集成，得到第m子频段的集成空间滤波器
    对以Pn个子周期进行平均，得到复制连接多个合成得到所需长度的稳态视觉诱发电位模板信号
    采集受试者的未知稳态视觉诱发电位采用所述Nfb个滤波器组对所述未知稳态视觉诱发电位K进行预处理，得到Nfb个子频段的预处理后未知稳态视觉诱发电位K(m)，其中，m＝1,2,…,Nfb；
    通过所述集成空间滤波器和所述稳态视觉诱发电位模板信号计算Nfb个子频段的相关系数对Nfb个加权求和得到所述未知稳态视觉诱发电位K与所述原始稳态视觉诱发电位Xn之间的相关系数ρn；
    得到所述未知稳态视觉诱发电位K与频率分别为f1,…fi,…fn,,…fNf的全部原始稳态视觉诱发电位[X1,…Xi,…,Xn,...,XNf]之间的相关系数[ρ1,…ρi,…,ρn,…,ρNf]，第i个相关系数ρi最大，则所述未知稳态视觉诱发电位K的视觉刺激频率为fi。
  2. 根据权利要求1所述的面向稳态视觉诱发电位的低训练成本脑电特征解 码方法，其中，所述将所述预处理后稳态视觉诱发电位划分为Pn个子周期包括：
    将所述原始稳态视觉诱发电位为Xn划分为Pn个子周期，Pn＝floor(Nsfn/Fs)，每个子周期采样点的数量为Ln＝round(Fs/fn)，其中，Fs为采样频率，floor为向下取整函数，round为四舍五入函数；根据子周期采样点的数量Ln分割得到
  3. 根据权利要求1所述的面向稳态视觉诱发电位的低训练成本脑电特征解码方法，其中，所述合成得到所需长度的稳态视觉诱发电位模板信号包括：
    合成得到所需长度的稳态视觉诱发电位模板信号 其中，为的一部分，的采样点数为Ns。
  4. 根据权利要求1所述的面向稳态视觉诱发电位的低训练成本脑电特征解码方法，其中，所述根据得到第m子频段的空间滤波器包括：
    根据如下广义特征值分解进行计算


    其中，这里是跨周期协方差矩阵的总和，是周期自协方差矩阵的总和，和表示中第p1和第p2个周期。
  5. 根据权利要求1所述的面向稳态视觉诱发电位的低训练成本脑电特征解 码方法，其中，所述通过所述集成空间滤波器和所述稳态视觉诱发电位模板信号计算Nfb个子频段的相关系数包括：
    根据如下公式求取相关系数
    其中，ρ表示计算二维相关系数。
  6. 根据权利要求1所述的面向稳态视觉诱发电位的低训练成本脑电特征解码方法，其中，所述得到所述未知稳态视觉诱发电位K与所述原始稳态视觉诱发电位Xn之间的相关系数ρn，包括：
    根据如下公式求取相关系数ρn
    其中，a(m)被定义为a(m)＝m-s+β，s和β取值可通过网格搜索确定最优参数，也可取经验值s＝1.25，β＝0.25。
  7. 一种面向稳态视觉诱发电位的低训练成本脑电特征解码装置，包括：
    预处理模块(610)，设置为在屏幕上呈现Nf个频率分别为f1,…fi,…fn,,…fNf的视觉刺激，采集受试者分别注视这些视觉刺激的原始稳态视觉诱发电位信号作为训练数据，受试者注视频率为fn的视觉刺激时所采集的受试者的多脑电通道的原始稳态视觉诱发电位为其中，Nc为脑电通道数，Ns为采样点数，Nt为试次数；采用Nfb个滤波器组对所述原始稳态视觉诱发电位Xn进行预处理，得到Nfb个子频段的预处理后稳态视觉诱发电位其中，m＝1,2,…,Nfb；
    子周期划分模块(620)，设置为将预处理后稳态视觉诱发电位划分为Pn个子周期
    集成空间滤波器获取模块(630)，设置为根据得到第m子频段的空间 滤波器对Nf个频率的第m子频段的空间滤波器进行集成，得到第m子频段的集成空间滤波器
    模板信号获取模块(640)，设置为对以Pn个子周期进行平均，得到复制连接多个合成得到所需长度的稳态视觉诱发电位模板信号
    未知稳态视觉诱发电位采集模块(650)，设置为采集受试者的未知稳态视觉诱发电位采用所述Nfb个滤波器组对所述未知稳态视觉诱发电位K进行预处理，得到Nfb个子频段的预处理后未知稳态视觉诱发电位K(m)，其中，m＝1,2,…,Nfb；
    相关系数计算模块(660)，设置为通过集成空间滤波器和稳态视觉诱发电位模板信号计算Nfb个子频段的相关系数对Nfb个加权求和得到未知稳态视觉诱发电位K与所述原始稳态视觉诱发电位Xn之间的相关系数ρn；
    视觉刺激频率获取模块(670)，设置为得到未知稳态视觉诱发电位K与频率分别为f1,…fi,…fn,,…fNf的全部原始稳态视觉诱发电位[X1,…Xi,…,Xn,...,XNf]之间的相关系数[ρ1,…ρi,…,ρn,…,ρNf]，第i个相关系数ρi最大，则所述未知稳态视觉诱发电位K的视觉刺激频率为fi。
  8. 一种面向稳态视觉诱发电位的低训练成本脑机接口系统，包括：
    至少一个处理器；
    存储器，设置为存储至少一个程序，
    当所述至少一个程序被所述至少一个处理器执行时，使得所述至少一个处理器实现如权利要求1-6中任一所述的面向稳态视觉诱发电位的低训练成本脑电特征解码方法。
  9. 一种包含计算机程序的计算机可读存储介质，所述计算机程序在由处理器执行时实现如权利要求1-6中任一所述的面向稳态视觉诱发电位的低训练成 本脑电特征解码方法。
PCT/CN2024/081370 2023-08-07 2024-03-13 面向稳态视觉诱发电位的低训练成本脑电特征解码方法 Pending WO2025030833A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202310981789.2A CN117009766A (zh) 2023-08-07 2023-08-07 面向稳态视觉诱发电位的低训练成本脑电特征解码方法
CN202310981789.2 2023-08-07

Publications (1)

Publication Number Publication Date
WO2025030833A1 true WO2025030833A1 (zh) 2025-02-13

Family

ID=88561503

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2024/081370 Pending WO2025030833A1 (zh) 2023-08-07 2024-03-13 面向稳态视觉诱发电位的低训练成本脑电特征解码方法

Country Status (2)

Country Link
CN (1) CN117009766A (zh)
WO (1) WO2025030833A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN120597079A (zh) * 2025-04-25 2025-09-05 天津大学 微弱非对称性视觉刺激的脑电信号分类模型的训练方法及应用
CN121040925A (zh) * 2025-11-03 2025-12-02 湘江实验室 面向脑控康复设备的控制策略生成方法、装置、设备及存储介质
CN121365243A (zh) * 2025-12-22 2026-01-20 上海术理智能科技有限公司 基于迁移学习的双域双尺度跨被试ssvep解码方法及装置

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117009766A (zh) * 2023-08-07 2023-11-07 天津大学 面向稳态视觉诱发电位的低训练成本脑电特征解码方法
CN119149938A (zh) * 2024-08-19 2024-12-17 西安电子科技大学 基于样本数据增强的稳态视觉诱发电位脑电特征解码方法

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107748622A (zh) * 2017-11-08 2018-03-02 中国医学科学院生物医学工程研究所 一种基于人脸感知的稳态视觉诱发电位脑‑机接口方法
CN113100786A (zh) * 2021-05-19 2021-07-13 南京邮电大学 一种稳态视觉诱发电位信号的频率识别方法
CN114371784A (zh) * 2022-01-14 2022-04-19 天津大学 一种面向稳态视觉诱发电位的脑-机接口解码方法
CN116451017A (zh) * 2023-03-14 2023-07-18 天津大学 面向稳态视觉诱发电位的脑电特征解码方法及脑机接口
CN117009766A (zh) * 2023-08-07 2023-11-07 天津大学 面向稳态视觉诱发电位的低训练成本脑电特征解码方法

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111580643B (zh) * 2020-04-10 2022-04-26 天津大学 一种基于稳态不对称视觉诱发电位的脑-机接口方法
CN113288181B (zh) * 2021-06-21 2022-09-27 杭州电子科技大学 基于稳态视觉诱发电位脑电信号识别的个体模板重构方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107748622A (zh) * 2017-11-08 2018-03-02 中国医学科学院生物医学工程研究所 一种基于人脸感知的稳态视觉诱发电位脑‑机接口方法
CN113100786A (zh) * 2021-05-19 2021-07-13 南京邮电大学 一种稳态视觉诱发电位信号的频率识别方法
CN114371784A (zh) * 2022-01-14 2022-04-19 天津大学 一种面向稳态视觉诱发电位的脑-机接口解码方法
CN116451017A (zh) * 2023-03-14 2023-07-18 天津大学 面向稳态视觉诱发电位的脑电特征解码方法及脑机接口
CN117009766A (zh) * 2023-08-07 2023-11-07 天津大学 面向稳态视觉诱发电位的低训练成本脑电特征解码方法

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN120597079A (zh) * 2025-04-25 2025-09-05 天津大学 微弱非对称性视觉刺激的脑电信号分类模型的训练方法及应用
CN121040925A (zh) * 2025-11-03 2025-12-02 湘江实验室 面向脑控康复设备的控制策略生成方法、装置、设备及存储介质
CN121365243A (zh) * 2025-12-22 2026-01-20 上海术理智能科技有限公司 基于迁移学习的双域双尺度跨被试ssvep解码方法及装置

Also Published As

Publication number Publication date
CN117009766A (zh) 2023-11-07

Similar Documents

Publication Publication Date Title
WO2025030833A1 (zh) 面向稳态视觉诱发电位的低训练成本脑电特征解码方法
CN110969108B (zh) 一种基于自主运动想象脑电的肢体动作识别方法
Burke et al. A parametric feature extraction and classification strategy for brain-computer interfacing
CN102722727B (zh) 基于脑功能网络邻接矩阵分解的脑电特征提取方法
US11221672B2 (en) Asymmetric EEG-based coding and decoding method for brain-computer interfaces
Wang et al. Common spatial pattern method for channel selelction in motor imagery based brain-computer interface
Chiang et al. Cross-subject transfer learning improves the practicality of real-world applications of brain-computer interfaces
CN102184019B (zh) 基于隐性注意的视听联合刺激脑-机接口方法
CN111184509A (zh) 一种基于传递熵的情绪诱导脑电信号分类方法
Miao et al. Automated CCA-MWF algorithm for unsupervised identification and removal of EOG artifacts from EEG
CN115640827B (zh) 对电刺激数据处理的智能闭环反馈网络方法及系统
CN115414051A (zh) 一种脑电信号自适应窗口的情绪分类识别方法
CN115620751B (zh) 一种基于说话人语音诱发的脑电信号预测方法
Ferrante et al. Data-efficient hand motor imagery decoding in EEG-BCI by using Morlet wavelets & common spatial pattern algorithms
Feng et al. Feature extraction algorithm based on csp and wavelet packet for motor imagery eeg signals
Kaewwit et al. High accuracy EEG biometrics identification using ICA and AR model
Melinda et al. Innovative portable quantitative EEG data acquisition based cyton biosensing board using psd analysis
CN120162529B (zh) 一种基于噪声注意力机制的脑电信号时频域去噪方法
CN120803248B (zh) 本体脑波音频听觉感知同步反馈方法、装置、系统及电子设备
Ozdenizci et al. On the use of generative deep neural networks to synthesize artificial multichannel EEG signals
US20250120831A1 (en) Intelligent ward round method and system based on ssvep electroencephalogram signal
CN116451017B (zh) 面向稳态视觉诱发电位的脑电特征解码方法及脑机接口
Biswas et al. Lateralization of brain during EEG based covert speech classification
Dhole et al. A Review of EEG Artifact Removal Techniques for Brain-Computer Interface
CN109078262B (zh) 一种基于外周神经电刺激的mi-bci训练方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 24850493

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE