Method, system and equipment for evaluating and regulating cognitive state of few-electrode electroencephalogram signals

Abstract

The invention discloses a cognitive state evaluation, regulation and control method, system and equipment for few-electrode electroencephalogram signals. The method comprises the following steps: selecting a blood oxygen level dependent signal feature set related to a preset cognitive state; taking the blood oxygen level dependent signal feature set as a label of the preprocessed scalp electroencephalogram signal, and determining a trained first prediction model based on the label and the preprocessed scalp electroencephalogram signal; inputting the target scalp electroencephalogram signal into a trained first prediction model to obtain a target few-electrode electroencephalogram signal; filtering the target few-electrode electroencephalogram signal, and inputting the filtered target few-electrode electroencephalogram signal into a constructed cognitive state evaluation model to obtain a cognitive state output by the cognitive state evaluation model; and according to the cognitive state output by the cognitive state evaluation model, regulating and controlling the current tested target cognitive state. According to the method, the accuracy of cognitive state evaluation is improved, and the spatial specificity of online cognitive regulation is improved.

Description

Technical Field

[0001] The present application relates to the technical field of cognitive state assessment, and in particular to a method, system and device for cognitive state assessment and regulation based on a few-electrode electroencephalogram (EEG) signal. Background Art

[0002] As the core element of the human-computer interaction system, the cognitive state of humans has a significant impact on the stability and safety of the system. Accurately assessing the cognitive state and then adjusting the operator's cognitive level is of great significance to maintaining high work efficiency, enhancing the operating experience, ensuring system safety, and reducing the accident rate.

[0003] Cognitive ability is related to specific brain regions, neural circuits or brain networks in the human brain. Magnetic resonance imaging, as one of the most commonly used non-invasive brain imaging techniques, has a very high spatial resolution and can accurately locate cognitive-related brain regions, but MRI equipment is less portable and expensive. EEG signals, as another most commonly used non-invasive brain imaging signal, have the characteristics of high temporal resolution, good portability and economy, but due to the low spatial resolution, they cannot accurately locate the neural signals of specific brain regions or neural circuits related to cognitive tasks. The EEG tracing method can locate the functional brain regions related to specific tasks, but generally requires 64 or more electrodes, and has poor economy and portability.

[0004] Therefore, how to restore the blood oxygen level-dependent (BOLD) signal of specific brain areas related to cognition through few-electrode EEG signals, so as to realize the assessment of cognitive state using few-electrode EEG signals and perform online cognitive regulation is an urgent problem to be solved. Summary of the invention

[0005] The present application aims to propose a method, system and device for evaluating and regulating cognitive state using few-electrode EEG signals, which can restore blood oxygen level-dependent signals of specific brain areas related to cognition through few-electrode EEG signals, thereby improving the accuracy of cognitive state evaluation using few-electrode EEG signals and improving the spatial specificity of online cognitive regulation.

[0006] In a first aspect, an embodiment of the present application provides a method for evaluating and regulating cognitive state based on a few-electrode electroencephalogram signal, the method comprising:

[0007] Acquire synchronously collected scalp EEG signals, functional magnetic resonance imaging data, and structural magnetic resonance imaging data;

[0008] Preprocessing the structural MRI data to obtain preprocessed structural MRI data; preprocessing the functional MRI data based on the preprocessed structural MRI data to obtain preprocessed functional MRI data; and preprocessing the scalp EEG signal to obtain a preprocessed scalp EEG signal;

[0009] Selecting a blood oxygen level dependent signal feature set related to a preset cognitive state according to the preprocessed functional magnetic resonance imaging data, wherein the blood oxygen level dependent signal feature set includes at least one of a brain region blood oxygen level dependent signal, a neural circuit blood oxygen level dependent signal, and a brain network blood oxygen level dependent signal;

[0010] Using the blood oxygen level dependent signal feature set as a label of the preprocessed scalp EEG signal, and determining a trained first prediction model based on the label and the preprocessed scalp EEG signal;

[0011] Inputting the target scalp EEG signal into the trained first prediction model to obtain a target few-electrode EEG signal, wherein the target few-electrode EEG signal is an EEG signal of a small number of electrodes associated with the label;

[0012] Filtering the target few-electrode EEG signal, and inputting the filtered target few-electrode EEG signal into the constructed cognitive state assessment model to obtain the cognitive state output by the cognitive state assessment model;

[0013] According to the cognitive state output by the cognitive state assessment model, the target cognitive state of the current subject is regulated.

[0014] Compared with the prior art, the first aspect of the present application has the following beneficial effects:

[0015] The method can provide a good data basis for the later training of the first prediction model by preprocessing the structural magnetic resonance imaging data, preprocessing the functional magnetic resonance imaging data, and preprocessing the scalp electroencephalogram (EEG) signal; according to the preprocessed functional magnetic resonance imaging data, a blood oxygen level dependent signal feature set related to a preset cognitive state is selected, the blood oxygen level dependent signal feature set is used as a label of the preprocessed scalp electroencephalogram (EEG) signal, and based on the label and the preprocessed scalp electroencephalogram (EEG) signal, a trained first prediction model is determined, and the target scalp electroencephalogram (EEG) signal is input into the trained first prediction model to obtain a target few-electrode EEG signal; by comparing the brain region blood oxygen level dependent signal, the neural circuit blood oxygen level dependent signal, and the brain network blood oxygen level dependent signal with the preprocessed scalp electroencephalogram (EEG) signal, the trained first prediction model is determined, and the target few-electrode EEG signal is obtained by comparing the brain region blood oxygen level dependent signal, the neural circuit blood oxygen level dependent signal, and the brain network blood oxygen level dependent signal with the preprocessed scalp electroencephalogram (EEG) signal. The first prediction model is trained by establishing association with electrical signals, so that the trained first prediction model can predict EEG signals related to brain region blood oxygen level dependence signals, neural circuit blood oxygen level dependence signals and brain network blood oxygen level dependence signals according to the collected EEG signals; the target few-electrode EEG signals are filtered, and the filtered target few-electrode EEG signals are input into the constructed cognitive state assessment model to obtain the cognitive state output by the cognitive state assessment model, and the target cognitive state of the current subject is regulated according to the cognitive state output by the cognitive state assessment model, which can improve the recovery of blood oxygen level dependence signals of specific brain regions related to cognition through few-electrode EEG signals, thereby improving the accuracy of cognitive state assessment using few-electrode EEG signals and improving the spatial specificity of online cognitive regulation.

[0016] In some embodiments, selecting a blood oxygen level dependent signal feature set associated with a preset cognitive state based on the preprocessed functional magnetic resonance imaging data includes:

[0017] Dividing the preprocessed functional magnetic resonance imaging data of the whole brain into a plurality of sets of blood oxygen level dependent signal feature sets, wherein the functional magnetic resonance imaging data includes at least one of functional magnetic resonance imaging signals of brain regions, functional magnetic resonance imaging signals of neural circuits, and functional magnetic resonance imaging signals of brain networks;

[0018] A multivariate pattern analysis method is used to select a blood oxygen level dependent signal feature set related to a preset cognitive state from the multiple groups of blood oxygen level dependent signal feature sets.

[0019] In some embodiments, determining a trained first prediction model based on the label and the preprocessed scalp EEG signal includes:

[0020] Filtering the preprocessed scalp EEG signal to obtain a multi-band scalp EEG signal;

[0021] A sliding window method is used on the multi-band scalp EEG signal to obtain a multi-time period multi-band EEG signal;

[0022] The multi-time period and multi-frequency band EEG signals are input into a first prediction model, and the first prediction model is trained according to the labels to obtain a trained first prediction model.

[0023] In some embodiments, the first prediction model includes a channel attention layer, a spatial attention layer, and a neural network prediction layer, and the multi-period multi-band EEG signal is input into the first prediction model, and the first prediction model is trained according to the label to obtain a trained first prediction model, including:

[0024] Inputting the multi-period and multi-band EEG signals into the channel attention layer to obtain an output result of the channel attention layer;

[0025] Input the output result of the channel attention layer to the spatial attention layer to obtain the output result of the spatial attention layer;

[0026] Inputting the output result of the spatial attention layer into the neural network prediction layer to obtain the output result of the neural network prediction layer;

[0027] Constructing a first loss function according to the output result of the neural network prediction layer and the label;

[0028] The first prediction model is trained according to the first loss function until the first loss function converges or reaches a preset number of iterations, thereby obtaining a trained first prediction model.

[0029] In some embodiments, the cognitive state assessment model includes a time feature extraction module, a spatial feature extraction module and a classification module, and the filtered target few-electrode EEG signal is input into the constructed cognitive state assessment model to obtain the cognitive state output by the cognitive state assessment model, including:

[0030] Inputting the filtered target few-electrode EEG signal into the spatial feature extraction module to obtain the output result of the spatial feature extraction module;

[0031] Inputting the output result of the spatial feature extraction module into the temporal feature extraction module to obtain the output result of the temporal feature extraction module;

[0032] The output result of the time feature extraction module is input into the classification module to obtain the cognitive state output by the cognitive state assessment model.

[0033] In some embodiments, regulating the target cognitive state of the current subject according to the cognitive state output by the cognitive state assessment model includes:

[0034] Obtain the few-electrode EEG signal of the current subject;

[0035] Filtering the few-electrode EEG signal of the current subject to obtain a few-electrode EEG signal in the same frequency band as the target few-electrode EEG signal;

[0036] Calculating a first time-frequency feature according to a few-electrode EEG signal in the same frequency band as the target few-electrode EEG signal;

[0037] According to the cognitive state output by the cognitive state assessment model and the first time-frequency feature, the target cognitive state of the current subject is regulated.

[0038] In some embodiments, regulating the target cognitive state of the current subject according to the cognitive state output by the cognitive state assessment model and the first time-frequency feature includes:

[0039] Obtaining the cognitive state output by the cognitive state assessment model for multiple test times;

[0040] Selecting the EEG signal corresponding to each cognitive state predicted correctly from the cognitive states output by the cognitive state assessment model of the plurality of test times;

[0041] Calculating the second time-frequency feature corresponding to each cognitive state according to the EEG signal corresponding to each correctly predicted cognitive state;

[0042] Compare the second time-frequency feature with the first time-frequency feature to obtain a difference;

[0043] According to the difference, the target cognitive state of the current subject is adjusted.

[0044] In a second aspect, the present application also provides a system for evaluating and regulating cognitive status based on a few-electrode electroencephalogram signal, the system comprising:

[0045] A data acquisition unit, used to acquire synchronously collected scalp EEG signals, functional magnetic resonance imaging data and structural magnetic resonance imaging data;

[0046] A data processing unit, configured to preprocess the structural MRI data to obtain preprocessed structural MRI data; preprocess the functional MRI data based on the preprocessed structural MRI data to obtain preprocessed functional MRI data; and preprocess the scalp EEG signal to obtain a preprocessed scalp EEG signal;

[0047] a data selection unit, configured to select a blood oxygen level dependent signal feature set related to a preset cognitive state according to the preprocessed functional magnetic resonance imaging data, wherein the blood oxygen level dependent signal feature set includes at least one of a brain region blood oxygen level dependent signal, a neural circuit blood oxygen level dependent signal, and a brain network blood oxygen level dependent signal;

[0048] a model training unit, configured to use the blood oxygen level dependent signal feature set as a label of the preprocessed scalp EEG signal, and determine a trained first prediction model based on the label and the preprocessed scalp EEG signal;

[0049] A model output unit, used for inputting the target scalp EEG signal into the trained first prediction model to obtain a target few-electrode EEG signal, wherein the target few-electrode EEG signal is an EEG signal of a small number of electrodes associated with the label;

[0050] A data evaluation unit, used for filtering the target few-electrode EEG signal, and inputting the filtered target few-electrode EEG signal into the constructed cognitive state evaluation model to obtain the cognitive state output by the cognitive state evaluation model;

[0051] The data control unit is used to control the target cognitive state of the current subject according to the cognitive state output by the cognitive state assessment model.

[0052] In the third aspect, an embodiment of the present application also provides an electronic device, comprising at least one control processor and a memory for communicating with the at least one control processor; the memory stores instructions that can be executed by the at least one control processor, and the instructions are executed by the at least one control processor so that the at least one control processor can execute the cognitive state assessment and regulation method of a few-electrode EEG signal as described above.

[0053] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to enable a computer to execute the cognitive state assessment and regulation method of a few-electrode EEG signal as described above.

[0054] It can be understood that the beneficial effects of the second to fourth aspects compared with the related art are the same as the beneficial effects of the first aspect compared with the related art. Please refer to the relevant description in the first aspect, and no further details will be given here. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which:

[0056] Figure 1 It is a flowchart of an embodiment of a method for evaluating and regulating cognitive state based on a few-electrode EEG signal provided in the present application;

[0057] Figure 2 It is a flowchart of the offline training phase and the online testing phase in the embodiment of the cognitive state assessment and regulation method based on a few-electrode EEG signal provided by the present application;

[0058] Figure 3 It is a flow chart of the first prediction model in an embodiment of the method for evaluating and regulating cognitive state based on few-electrode EEG signals provided in the present application;

[0059] Figure 4 It is a flow chart of a cognitive state assessment model in an embodiment of a method for cognitive state assessment and regulation based on a few-electrode EEG signal provided in the present application;

[0060] Figure 5 It is a structural schematic diagram of an embodiment of a cognitive state assessment and regulation system based on a few-electrode EEG signal provided in the present application. DETAILED DESCRIPTION

[0061] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application, and cannot be understood as limiting the present application.

[0062] In the description of this application, if there is a description of first, second, etc., it is only for the purpose of distinguishing technical features, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features or implicitly indicating the order of the indicated technical features.

[0063] In the description of the present application, it should be understood that the descriptions involving orientation, such as the orientation or positional relationship indicated as up, down, etc., are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present application.

[0064] In the description of this application, it should be noted that, unless otherwise clearly defined, terms such as setting, installing, connecting, etc. should be understood in a broad sense, and technicians in the relevant technical field can reasonably determine the specific meaning of the above terms in this application based on the specific content of the technical solution.

[0065] Cognitive ability is related to specific brain regions, neural circuits or brain networks in the human brain. Magnetic resonance imaging, as one of the most commonly used non-invasive brain imaging techniques, has a very high spatial resolution and can accurately locate cognitive-related brain regions, but MRI equipment is less portable and expensive. EEG signals, as another most commonly used non-invasive brain imaging signal, have the characteristics of high temporal resolution, good portability and economy, but due to the low spatial resolution, they cannot accurately locate the neural signals of specific brain regions or neural circuits related to cognitive tasks. The EEG tracing method can locate the functional brain regions related to specific tasks, but generally requires 64 or more electrodes, and has poor economy and portability.

[0066] Therefore, how to restore the blood oxygen level-dependent signals of specific brain areas related to cognition through few-electrode EEG signals, so as to realize the assessment of cognitive state using few-electrode EEG signals and perform online cognitive regulation is an urgent problem to be solved.

[0067] In order to solve the problem of how to restore the blood oxygen level dependence signal of specific brain areas related to cognition through few-electrode EEG signals, thereby improving the accuracy of cognitive state assessment using few-electrode EEG signals and improving the spatial specificity of online cognitive regulation, the present application proposes a cognitive state assessment and regulation method, system and device using few-electrode EEG signals.

[0068] Reference Figure 1 The present application embodiment provides a method for evaluating and regulating cognitive state based on a few-electrode electroencephalogram signal, the method comprising the following steps:

[0069] Step S100, acquiring synchronously collected scalp EEG signals, functional magnetic resonance imaging data and structural magnetic resonance imaging data;

[0070] Step S200, preprocessing the structural MRI data to obtain preprocessed structural MRI data; preprocessing the functional MRI data according to the preprocessed structural MRI data to obtain preprocessed functional MRI data; and preprocessing the scalp EEG signal to obtain a preprocessed scalp EEG signal;

[0071] Step S300, selecting a blood oxygen level dependent signal feature set related to a preset cognitive state according to the preprocessed functional magnetic resonance imaging data, wherein the blood oxygen level dependent signal feature set includes at least one of a brain region blood oxygen level dependent signal, a neural circuit blood oxygen level dependent signal, and a brain network blood oxygen level dependent signal;

[0072] Step S400: using the blood oxygen level dependent signal feature set as a label of the preprocessed scalp EEG signal, and determining a trained first prediction model based on the label and the preprocessed scalp EEG signal;

[0073] Step S500: input the target scalp EEG signal into the trained first prediction model to obtain a target few-electrode EEG signal, wherein the target few-electrode EEG signal is an EEG signal of a small number of electrodes associated with the label;

[0074] Step S600, filtering the target few-electrode EEG signal, and inputting the filtered target few-electrode EEG signal into the constructed cognitive state assessment model to obtain the cognitive state output by the cognitive state assessment model;

[0075] Step S700: According to the cognitive state output by the cognitive state assessment model, the target cognitive state of the current subject is regulated.

[0076] In this embodiment, by preprocessing the structural MRI data, preprocessing the functional MRI data, and preprocessing the scalp EEG signal, a good data basis can be provided for the later training of the first prediction model; according to the preprocessed functional MRI data, a blood oxygen level dependent signal feature set related to the preset cognitive state is selected, and the blood oxygen level dependent signal feature set is used as a label of the preprocessed scalp EEG signal, and based on the label and the preprocessed scalp EEG signal, a trained first prediction model is determined, and the target scalp EEG signal is input into the trained first prediction model to obtain a target few-electrode EEG signal, and the brain region blood oxygen level dependent signal, the neural circuit blood oxygen level dependent signal, and the brain network blood oxygen level dependent signal are compared with the preprocessed scalp EEG signal. The first prediction model is trained by establishing association with cortical EEG signals, so that the trained first prediction model can predict EEG signals related to brain region blood oxygen level dependence signals, neural circuit blood oxygen level dependence signals and brain network blood oxygen level dependence signals according to the collected EEG signals; the target few-electrode EEG signals are filtered, and the filtered target few-electrode EEG signals are input into the constructed cognitive state assessment model to obtain the cognitive state output by the cognitive state assessment model, and the target cognitive state of the current subject is regulated according to the cognitive state output by the cognitive state assessment model, which can improve the recovery of blood oxygen level dependence signals of specific brain regions related to cognition through few-electrode EEG signals, thereby improving the accuracy of cognitive state assessment using few-electrode EEG signals and improving the spatial specificity of online cognitive regulation.

[0077] In some embodiments, based on the preprocessed functional magnetic resonance imaging data, a blood oxygen level dependent signal feature set associated with a preset cognitive state is selected, including:

[0078] dividing the preprocessed functional magnetic resonance imaging data of the whole brain into a plurality of sets of blood oxygen level dependent signal feature sets, wherein the functional magnetic resonance imaging data includes at least one of functional magnetic resonance imaging signals of brain regions, functional magnetic resonance imaging signals of neural circuits, and functional magnetic resonance imaging signals of brain networks;

[0079] A multivariate pattern analysis method is used to select a blood oxygen level dependent signal feature set related to a preset cognitive state from multiple groups of blood oxygen level dependent signal feature sets.

[0080] In this embodiment, by selecting a blood oxygen level dependent signal feature set composed of blood oxygen level dependent signals of brain regions, neural circuits or brain networks related to a preset cognitive state, a good data foundation can be laid for the subsequent training of the first prediction model.

[0081] In some embodiments, determining a trained first prediction model based on the label and the preprocessed scalp EEG signal includes:

[0082] Filtering the preprocessed scalp EEG signal to obtain a multi-band scalp EEG signal;

[0083] The sliding window method is used for multi-band scalp EEG signals to obtain multi-period multi-band EEG signals;

[0084] The multi-time period and multi-frequency band EEG signals are input into the first prediction model, and the first prediction model is trained according to the labels to obtain a trained first prediction model.

[0085] In this embodiment, the first prediction model is trained by using brain region blood oxygen level dependence signals, neural circuit blood oxygen level dependence signals and brain network blood oxygen level dependence signals as labels for model training and establishing associations with preprocessed scalp EEG signals, so that the trained first prediction model can predict EEG signals related to brain region blood oxygen level dependence signals, neural circuit blood oxygen level dependence signals and brain network blood oxygen level dependence signals based on the collected EEG signals.

[0086] In some embodiments, the first prediction model includes a channel attention layer, a spatial attention layer, and a neural network prediction layer, and multi-time period and multi-band EEG signals are input into the first prediction model, and the first prediction model is trained according to the label to obtain a trained first prediction model, including:

[0087] Input multi-period and multi-band EEG signals into the channel attention layer to obtain the output result of the channel attention layer;

[0088] Input the output result of the channel attention layer to the spatial attention layer to obtain the output result of the spatial attention layer;

[0089] Input the output result of the spatial attention layer into the neural network prediction layer to obtain the output result of the neural network prediction layer;

[0090] Construct the first loss function based on the output results and labels of the neural network prediction layer;

[0091] The first prediction model is trained according to the first loss function until the first loss function converges or reaches a preset number of iterations, thereby obtaining a trained first prediction model.

[0092] In this embodiment, a first loss function is constructed according to the output results and labels of the neural network prediction layer, and then the first prediction model is trained according to the first loss function until the first loss function converges or reaches a preset number of iterations, thereby obtaining a trained first prediction model. The trained first prediction model can predict few-electrode EEG signals related to brain region blood oxygen level-dependent signals, neural circuit blood oxygen level-dependent signals, and brain network blood oxygen level-dependent signals, thereby improving the prediction accuracy of the first prediction model.

[0093] In some embodiments, the cognitive state assessment model includes a time feature extraction module, a spatial feature extraction module and a classification module, and the filtered target few-electrode EEG signal is input into the constructed cognitive state assessment model to obtain the cognitive state output by the cognitive state assessment model, including:

[0094] Inputting the filtered target few-electrode EEG signal into the spatial feature extraction module to obtain the output result of the spatial feature extraction module;

[0095] The output result of the spatial feature extraction module is input into the temporal feature extraction module to obtain the output result of the temporal feature extraction module;

[0096] The output result of the time feature extraction module is input into the classification module to obtain the cognitive state output by the cognitive state assessment model.

[0097] In this embodiment, after the filtered target few-electrode EEG signal passes through the time feature extraction module, the space feature extraction module and the classification module, a relatively accurate cognitive state can be obtained, thereby realizing the evaluation of the cognitive state using the few-electrode EEG signal.

[0098] In some embodiments, according to the cognitive state output by the cognitive state assessment model, regulating the target cognitive state of the current subject includes:

[0099] Obtain the few-electrode EEG signal of the current subject;

[0100] Filter the few-electrode EEG signal of the current subject to obtain a few-electrode EEG signal in the same frequency band as the target few-electrode EEG signal;

[0101] Calculate the first time-frequency feature according to the few-electrode EEG signal in the same frequency band as the target few-electrode EEG signal;

[0102] According to the cognitive state and the first time-frequency characteristics output by the cognitive state assessment model, the target cognitive state of the current subject is regulated.

[0103] In this embodiment, by calculating the first time-frequency feature based on the few-electrode EEG signal in the same frequency band as the target few-electrode EEG signal, the target cognitive state of the current subject is regulated according to the cognitive state and the first time-frequency feature output by the cognitive state assessment model, so as to obtain accurate data in the later stage and train the prediction model based on the accurate data, the spatial specificity of cognitive state assessment and cognitive state regulation can be further improved.

[0104] In some embodiments, regulating the target cognitive state of the current subject according to the cognitive state and the first time-frequency feature output by the cognitive state assessment model includes:

[0105] Obtaining the cognitive state output by the cognitive state assessment model for multiple trial times;

[0106] From the cognitive states output by the cognitive state evaluation model of multiple trials, select the EEG signal corresponding to each cognitive state predicted correctly;

[0107] According to the EEG signal corresponding to each correctly predicted cognitive state, the second time-frequency feature corresponding to each cognitive state is calculated;

[0108] Compare the second time-frequency feature with the first time-frequency feature to obtain a difference;

[0109] According to the difference, adjust the current target cognitive state of the subject.

[0110] In this embodiment, the second time-frequency feature corresponding to each cognitive state is calculated based on the EEG signal corresponding to each correctly predicted cognitive state, and then the second time-frequency feature is compared with the first time-frequency feature to obtain the difference, and then the target cognitive state of the current subject is adjusted according to the difference, thereby improving the spatial specificity of the regulation of the target cognitive state.

[0111] To facilitate understanding by those skilled in the art, a set of best embodiments is provided below:

[0112] like Figure 2 As shown, the embodiment of the present application is divided into two stages: offline training and online testing. In the offline stage, the prediction model (i.e., the first prediction model) of scalp EEG and BOLD signals (i.e., blood oxygen level dependent signals) of specific brain regions and the cognitive state assessment model of few-electrode EEG are trained. In the online testing stage, the few-electrode EEG signals of the subjects in different cognitive states are collected in real time, the cognitive state is evaluated, and visual / auditory feedback signals are constructed for neurofeedback training. Specifically:

[0113] In the offline training phase, an experimental paradigm is designed (which may be a task paradigm in which subjects are asked to watch various types of emotional videos) to obtain multimodal brain imaging data including synchronously collected scalp EEG signals, functional magnetic resonance imaging data and structural magnetic resonance imaging data; the multimodal imaging data are preprocessed to obtain preprocessed multimodal imaging data; the fMRI signals in the preprocessed multimodal imaging data (i.e., preprocessed functional magnetic resonance imaging data) are used to locate the BOLD signals of brain regions, neural circuits or brain networks related to specific cognition using a multivariate pattern analysis method; a first prediction model for the association between scalp EEG signals and BOLD signals of specific brain regions, neural circuits or brain networks is established using a deep neural network to find the few-electrode EEG signals of electrode channels and frequency bands that are more important for predicting blood oxygen level dependent signals (i.e., BOLD signals); the few-electrode EEG signals of specific electrode channels and frequency bands found are used to establish a few-electrode EEG cognitive state assessment model with the cognitive state under a specific experimental paradigm. During the online testing phase, few-electrode EEG signals under different cognitive paradigms are collected in real time. EEG signals in specific frequency bands are calculated through preprocessing and filtering. The signals are input into a trained cognitive state assessment model to identify the cognitive state, and a neural feedback signal of visual / auditory feedback is constructed based on the identified cognitive state for cognitive enhancement.

[0114] This embodiment specifically includes the following steps:

[0115] 1. Design an experimental paradigm to obtain multimodal brain imaging data including synchronously collected scalp EEG signals, functional MRI data, and structural MRI data.

[0116] The acquisition device used can be a nuclear magnetic resonance scanner, and the fast gradient echo mode is selected. Before scanning, the scanning parameters such as scanning cycle, layer thickness, number of layers, scanning matrix, and scanning field of view need to be set. First, the individual structural magnetic resonance imaging data of the subjects are collected, and then the functional magnetic resonance imaging data and scalp EEG signals of the subjects under four types of emotional video stimulation paradigms are collected at the same time. During the scanning process, the subjects need to wear a 64-conductor magnetic resonance electrode cap, and the subjects are required to keep their heads still. The scalp EEG signal data is collected while the functional magnetic resonance imaging data is collected. The placement of the EEG electrodes is arranged according to the international electrode 10-10 system. During the acquisition process, the subjects repeatedly watch emotional stimulation videos, which contain four types of emotions: happiness, sadness, neutrality, and fear.

[0117] 2. Preprocess multimodal imaging data.

[0118] (1) Structural MRI data were preprocessed by motion correction, registration, segmentation, spatial normalization, and smoothing, which were implemented in MATLAB using the toolkit Statistic Parametric Mapping (SPM12).

[0119] (2) Preprocess the scalp EEG signal. According to the marker of the scalp EEG signal at the beginning of fMRI acquisition, extract the scalp EEG signal during the fMRI scanning period, calculate the noise template, use adaptive artifact subtraction to subtract the fMRI gradient noise template contained in the scalp EEG signal, and use the principal component analysis method to remove the gradient artifact of the EEG signal. The above preprocessing process is implemented in the software Brain Vision; the EEG signal is baseline corrected and bandpass filtered (0.1Hz to 45Hz), and the independent component analysis method is used to remove the influence of noise such as electrooculography, electrocardiography, and head movement. This process is completed using the MATLAB toolkit eeglab.

[0120] (3) Preprocessing the functional magnetic resonance imaging data includes: firstly performing inter-slice time correction and head motion correction processing; then registering the functional magnetic resonance image and the structural magnetic resonance image, mapping all data to the MNI (Montreal Neurological Institute) standard space for standardization processing based on the segmentation information of the structural magnetic resonance image; using the specified frequency band space to perform time domain filtering on the functional magnetic resonance imaging data to remove linear drift, and in this embodiment, the specified frequency band interval is 0.01 Hz to 0.08 Hz. The above preprocessing process is implemented in MATLAB using the toolkit SPM12.

[0121] It should be noted that when the fMRI signal starts scanning, a scan start mark is placed on the scalp EEG signal, which ensures that the fMRI data and the scalp EEG signal are collected simultaneously.

[0122] The preprocessed structural MRI data are registered with the functional MRI data, and the functional MRI data are mapped to the MNI standard space; the preprocessed functional MRI data are used in step 3 to locate the BOLD signal of the brain region, neural circuit or brain network related to specific cognition; the preprocessed scalp EEG signal is used in step 4 to establish a first prediction model of the association between the scalp EEG signal and the BOLD signal of the specific brain region, neural circuit or brain network;

[0123] 3. Use fMRI signals to locate brain areas, neural circuits or brain networks related to specific cognition.

[0124] The preprocessed fMRI signals (i.e., preprocessed functional magnetic resonance imaging data) are analyzed using a multivariate pattern analysis method, and the brain regions related to the emotional video stimulation paradigm are located using a whole-brain-based method (i.e., using the fMRI signals of the whole brain). Specifically, fMRI signals of several brain regions are selected as variables from the fMRI signals of the whole brain in turn, and the fMRI signals of the whole brain are traversed. The multivariate pattern analysis method is used to fully consider the relationship between the feature combination pattern and the cognitive state variable, and the feature subset related to the cognitive state variable (i.e., the blood oxygen level dependent signal feature set) is selected for decoding, and the BOLD signals of the brain regions, neural circuits, or brain networks related to the emotional video stimulation paradigm (i.e., the preset cognitive state) are determined. Among them, the feature subset includes at least one of the selected brain region BOLD signals, neural circuit BOLD signals, and brain network BOLD signals.

[0125] The BOLD signals of the brain regions, neural circuits or brain networks selected in step 3 are used as labels of the preprocessed scalp EEG signals. The preprocessed scalp EEG signals are input into the first prediction model for training, and the scalp EEG signals that can predict the BOLD signals of the brain regions, neural circuits or brain networks are selected. Step 4 is to select the scalp EEG signals of certain specific electrodes from the scalp EEG signals of all electrodes by establishing the first prediction model that associates the scalp EEG signals with the fMRI signals.

[0126] 4. Use a deep neural network to establish the first prediction model of the association between scalp EEG signals and BOLD signals in specific brain regions, neural circuits or brain networks, and find the few-electrode EEG signals (i.e., target few-electrode EEG signals) of electrode channels and frequency bands that are more important for BOLD signal prediction.

[0127] The preprocessed scalp EEG signals are band-pass filtered and divided into five frequency bands of delta (0.5-3Hz), theta (3-8Hz), alpha (8-12Hz), beta (12-30Hz), and gamma (30-60Hz) (i.e., multi-band scalp EEG signals);

[0128] The response time of the BOLD signal may vary. The collected scalp EEG signals were aligned with the synchronously collected fMRI signals using a sliding window method. The window length was set to 2 s and slid forward in steps of 0.5 s to obtain EEG samples of multiple time periods and multiple frequency bands (i.e., multi-time period and multi-frequency band EEG signals).

[0129] A deep neural network prediction model (i.e., the first prediction model) combining the channel attention layer, the spatial attention layer, and the neural network prediction layer is constructed. The input of the deep neural network prediction model is the EEG signals of multiple time periods and multiple frequency bands. The output of the deep neural network prediction model is the predicted few-electrode EEG signals related to the BOLD signals of specific brain regions, neural circuits, or brain networks in the feature subset (i.e., the target few-electrode EEG signals). The loss function of the deep neural network prediction model (i.e., the first loss function) is the root mean square error value between the BOLD signals of the brain regions, neural circuits, or brain networks related to the predicted few-electrode EEG signals and the corresponding real brain regions, real neural circuits, or real brain networks (i.e., the brain region BOLD signals, neural circuit BOLD signals, or brain network BOLD signals in the feature subset), that is:

[0130] loss1=||y-y'||2

[0131] Where y′ represents the BOLD signal of the brain region, neural circuit or brain network related to the predicted few-electrode EEG signal, and y represents the BOLD signal of the real brain region, real neural circuit or real brain network corresponding to the BOLD signal of the predicted brain region, neural circuit or brain network.

[0132] Reference Figure 3 First, the collected 64-channel scalp EEG signals are bandpass filtered in the five frequency bands of delta, theta, alpha, beta, and gamma. Then, considering the time delay of the BOLD signal, the multi-band EEG signals of different time periods (i.e., multi-time and multi-band EEG signals) are obtained by sliding windows. The multi-time and multi-band EEG signals are input into the channel attention layer, and different weights are attached to the EEG signals of different channels; then the output results of the channel attention layer are input into the spatial attention layer; finally, the output results of the spatial attention layer are input into the neural network prediction layer, and different weights are attached to the EEG signals of different frequency bands. Among them, the channel attention layer can be constructed by Squeeze-and-Excitation Networks, the spatial attention layer can be constructed by Spatial Transformer Networks, and the neural network prediction layer can be constructed by 2D-Convolutional Neural Networks. The electrode channels related to the BOLD signals of specific brain regions, neural circuits, or brain networks are determined by the channel attention layer, and the frequency band range related to the BOLD signals of specific brain regions, neural circuits, or brain networks is determined by the spatial attention layer.

[0133] 5. Use the selected few-electrode EEG signals of specific channels and frequency bands to establish a few-electrode EEG cognitive state assessment model under a specific paradigm.

[0134] According to the few-electrode EEG signals and specific frequency bands selected in step 4 with good prediction effects on BOLD signals of specific brain regions, neural circuits or brain networks, the few-electrode EEG signals selected under the emotional stimulation task paradigm are filtered to extract the few-electrode EEG signals of the specific frequency bands of the EEG signals (i.e., the filtered target few-electrode EEG signals);

[0135] Reference Figure 4 , a cognitive state assessment model of few-electrode EEG signals based on a deep neural network is constructed. The deep neural network includes a time feature extraction module, a spatial feature extraction module and a classification module. Among them, the time feature extraction module can be constructed by a long short-term memory network, the spatial feature extraction model can be constructed by a 2D CNN network, and the classification model can be constructed by an MLP multi-layer perceptron network. The input of the cognitive state assessment model is the few-electrode EEG signals of a specific frequency band, and the output of the cognitive state assessment model is the discriminated cognitive state. Specifically, the spatial features of the few-electrode EEG signals of a specific frequency band are extracted by the spatial feature extraction module (i.e., the output result of the spatial feature extraction module); then the spatial features are input into the time feature extraction module to extract features including spatial features and time features (i.e., the output result of the time feature extraction module), and then the features including spatial features and time features are input into the classification module to obtain the discriminated cognitive state (i.e., the cognitive state output by the cognitive state assessment model).

[0136] The loss function of the deep neural network is the cross entropy error between the cognitive state predicted by the cognitive state assessment model and the actual cognitive state. The deep neural network is used to learn the correlation characteristics of multi-band few-electrode EEG signals and cognitive states. The loss function is:

[0137]

[0138] Among them, p i represents the cognitive state output by the cognitive state assessment model, y i Indicates the true cognitive state.

[0139] 6. Use the selected few-electrode EEG signals of specific channels and frequency bands to construct the control signal of neural feedback.

[0140] According to the few-electrode EEG signals selected in step 4 and the few-electrode EEG signal cognitive state assessment model established in step 5, only the few-electrode EEG signals of the subjects performing the emotional stimulation paradigm task are collected during the online experiment, and the few-electrode EEG signals are pre-processed by online filtering according to the specific frequency band determined in step 4;

[0141] Using the filtered EEG signal as a neural feedback signal, calculating the time-frequency characteristics of the EEG signal (i.e., the first time-frequency characteristics), such as power spectrum density, entropy, etc., constructing visual and auditory feedback signals according to the calculated time-frequency characteristics, and presenting the visual and auditory feedback signals in real time on a display screen or headphones;

[0142] According to the few-electrode EEG cognitive state assessment model established in step 5, in the offline analysis, the number of correctly predicted trials in the experimental trials of each type of cognitive state is counted, and certain time-frequency characteristics (i.e., second time-frequency characteristics) of the EEG signals of these accurately predicted trials are calculated, such as power spectral density, entropy, etc. The calculated time-frequency characteristics are used as the control target value of the online feedback experiment of this category of cognitive state. The difference between the time-frequency characteristics of the EEG signals of the subject's current state and the control target value will be displayed in real time. According to the difference, the subject's own state is adjusted in real time to narrow the gap with the control target value, so as to help the subject control the current target cognitive state (i.e., any cognitive state of the four emotions of happiness, sadness, neutrality and fear).

[0143] In this embodiment, by effectively integrating multimodal brain imaging data collected synchronously under the emotional stimulation paradigm, a first prediction model and a few-electrode EEG cognitive state assessment model were established. These models can improve the spatial specificity of cognitive state assessment and neural regulation.

[0144] Reference Figure 5 The embodiment of the present application also provides a cognitive state evaluation and control system for a few-electrode EEG signal, the system comprising a data acquisition unit 100, a data processing unit 200, a data selection unit 300, a model training unit 400, a model output unit 500, a data evaluation unit 600 and a data control unit 700, wherein:

[0145] The data acquisition unit 100 is used to acquire synchronously collected scalp EEG signals, functional magnetic resonance imaging data and structural magnetic resonance imaging data;

[0146] The data processing unit 200 is used to preprocess the structural MRI data to obtain preprocessed structural MRI data; preprocess the functional MRI data according to the preprocessed structural MRI data to obtain preprocessed functional MRI data; and preprocess the scalp EEG signal to obtain the preprocessed scalp EEG signal;

[0147] The data selection unit 300 is used to select a blood oxygen level dependent signal feature set related to a preset cognitive state according to the preprocessed functional magnetic resonance imaging data, wherein the blood oxygen level dependent signal feature set includes at least one of a brain region blood oxygen level dependent signal, a neural circuit blood oxygen level dependent signal, and a brain network blood oxygen level dependent signal;

[0148] A model training unit 400 is used to use the blood oxygen level dependent signal feature set as a label of the preprocessed scalp EEG signal, and determine a trained first prediction model based on the label and the preprocessed scalp EEG signal;

[0149] The model output unit 500 is used to input the target scalp EEG signal into the trained first prediction model to obtain a target few-electrode EEG signal, wherein the target few-electrode EEG signal is an EEG signal of a small number of electrodes related to the label;

[0150] The data evaluation unit 600 is used to filter the target few-electrode EEG signal, and input the filtered target few-electrode EEG signal into the constructed cognitive state evaluation model to obtain the cognitive state output by the cognitive state evaluation model;

[0151] The data control unit 700 is used to control the target cognitive state of the current subject according to the cognitive state output by the cognitive state assessment model.

[0152] It should be noted that since the cognitive state assessment and regulation system of a few-electrode EEG signal in this embodiment and the cognitive state assessment and regulation method of a few-electrode EEG signal mentioned above are based on the same inventive concept, the corresponding contents in the method embodiment are also applicable to the system embodiment and will not be described in detail here.

[0153] An embodiment of the present application also provides an electronic device, comprising: at least one control processor and a memory for communicating with the at least one control processor.

[0154] The memory, as a non-transient computer-readable storage medium, can be used to store non-transient software programs and non-transient computer executable programs. In addition, the memory may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory may optionally include a memory remotely disposed relative to the processor, and these remote memories may be connected to the processor 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.

[0155] The non-transient software program and instructions required to implement the method for evaluating and regulating the cognitive state of a few-electrode EEG signal in the above embodiment are stored in the memory. When executed by the processor, the method for evaluating and regulating the cognitive state of a few-electrode EEG signal in the above embodiment is executed, for example, the above described method is executed. Figure 1 The method comprises steps S100 to S700.

[0156] The system embodiments described above are merely illustrative, and the units described as separate components may or may not be physically separated, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0157] The present application also provides a computer-readable storage medium storing computer-executable instructions, which are executed by one or more control processors to enable the one or more control processors to execute a method for evaluating and regulating the cognitive state of a few-electrode EEG signal in the above method embodiment, for example, to execute the above described Figure 1 The functions of method steps S100 to S700 in the method.

[0158] It will be appreciated by those skilled in the art that all or some of the steps and systems in the disclosed method above may be implemented as software, firmware, hardware and appropriate combinations thereof. Some physical components or all physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor or a microprocessor, or may be implemented as hardware, or may be implemented as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or a non-transitory medium) and a communication medium (or a temporary medium). As known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, disk storage or other magnetic storage devices, or any other medium that may be used to store desired information and may be accessed by a computer. Furthermore, it is well known to those skilled in the art that communication media typically embodies computer readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.

[0159] The above is a specific description of the preferred implementation of the embodiments of the present application, but the embodiments of the present application are not limited to the above-mentioned implementation methods. Technical personnel familiar with the field can also make various equivalent modifications or substitutions without violating the spirit of the embodiments of the present application. These equivalent modifications or substitutions are all included in the scope defined by the claims of the embodiments of the present application.

Claims

1. A method for evaluating and regulating cognitive state based on few-electrode EEG signals, characterized in that: The method comprises: Acquire synchronously collected scalp EEG signals, functional magnetic resonance imaging data, and structural magnetic resonance imaging data; Preprocessing the structural MRI data to obtain preprocessed structural MRI data; preprocessing the functional MRI data based on the preprocessed structural MRI data to obtain preprocessed functional MRI data; and preprocessing the scalp EEG signal to obtain a preprocessed scalp EEG signal; Selecting a blood oxygen level dependent signal feature set related to a preset cognitive state according to the preprocessed functional magnetic resonance imaging data, wherein the blood oxygen level dependent signal feature set includes at least one of a brain region blood oxygen level dependent signal, a neural circuit blood oxygen level dependent signal, and a brain network blood oxygen level dependent signal; Using the blood oxygen level dependent signal feature set as a label of the preprocessed scalp EEG signal, and determining a trained first prediction model based on the label and the preprocessed scalp EEG signal; Inputting the target scalp EEG signal into the trained first prediction model to obtain a target few-electrode EEG signal, wherein the target few-electrode EEG signal is an EEG signal of a small number of electrodes associated with the label; Filtering the target few-electrode EEG signal, and inputting the filtered target few-electrode EEG signal into the constructed cognitive state assessment model to obtain the cognitive state output by the cognitive state assessment model; According to the cognitive state output by the cognitive state assessment model, the target cognitive state of the current subject is regulated.

2. The method for evaluating and regulating cognitive status based on few-electrode EEG signals according to claim 1, characterized in that: The method of selecting a blood oxygen level dependent signal feature set related to a preset cognitive state based on the preprocessed functional magnetic resonance imaging data includes: Dividing the preprocessed functional magnetic resonance imaging data of the whole brain into a plurality of sets of blood oxygen level dependent signal feature sets, wherein the functional magnetic resonance imaging data includes at least one of functional magnetic resonance imaging signals of brain regions, functional magnetic resonance imaging signals of neural circuits, and functional magnetic resonance imaging signals of brain networks; A multivariate pattern analysis method is used to select a blood oxygen level dependent signal feature set related to a preset cognitive state from the multiple groups of blood oxygen level dependent signal feature sets.

3. The method for evaluating and regulating cognitive status based on few-electrode EEG signals according to claim 1, characterized in that: The step of determining a trained first prediction model based on the label and the preprocessed scalp EEG signal comprises: Filtering the preprocessed scalp EEG signal to obtain a multi-band scalp EEG signal; A sliding window method is used on the multi-band scalp EEG signal to obtain a multi-time period multi-band EEG signal; The multi-time period and multi-frequency band EEG signals are input into a first prediction model, and the first prediction model is trained according to the labels to obtain a trained first prediction model.

4. The method for evaluating and regulating cognitive status based on few-electrode EEG signals according to claim 3, characterized in that: The first prediction model includes a channel attention layer, a spatial attention layer and a neural network prediction layer. The multi-period multi-band EEG signal is input into the first prediction model, and the first prediction model is trained according to the label to obtain a trained first prediction model, including: Inputting the multi-period and multi-band EEG signals into the channel attention layer to obtain an output result of the channel attention layer; Input the output result of the channel attention layer to the spatial attention layer to obtain the output result of the spatial attention layer; Inputting the output result of the spatial attention layer into the neural network prediction layer to obtain the output result of the neural network prediction layer; Constructing a first loss function according to the output result of the neural network prediction layer and the label; The first prediction model is trained according to the first loss function until the first loss function converges or reaches a preset number of iterations, thereby obtaining a trained first prediction model.

5. The method for evaluating and regulating cognitive status based on few-electrode EEG signals according to claim 1, characterized in that: The cognitive state evaluation model includes a time feature extraction module, a spatial feature extraction module and a classification module. The filtered target few-electrode EEG signal is input into the constructed cognitive state evaluation model to obtain the cognitive state output by the cognitive state evaluation model, including: Inputting the filtered target few-electrode EEG signal into the spatial feature extraction module to obtain the output result of the spatial feature extraction module; Inputting the output result of the spatial feature extraction module into the temporal feature extraction module to obtain the output result of the temporal feature extraction module; The output result of the time feature extraction module is input into the classification module to obtain the cognitive state output by the cognitive state assessment model.

6. The method for evaluating and regulating cognitive status based on few-electrode EEG signals according to claim 1, characterized in that: The step of regulating the target cognitive state of the current subject according to the cognitive state output by the cognitive state assessment model includes: Obtain the few-electrode EEG signal of the current subject; Filtering the few-electrode EEG signal of the current subject to obtain a few-electrode EEG signal in the same frequency band as the target few-electrode EEG signal; Calculating a first time-frequency feature according to a few-electrode EEG signal in the same frequency band as the target few-electrode EEG signal; According to the cognitive state output by the cognitive state assessment model and the first time-frequency feature, the target cognitive state of the current subject is regulated.

7. The method for evaluating and regulating cognitive status based on few-electrode EEG signals according to claim 6, characterized in that: The step of regulating the target cognitive state of the current subject according to the cognitive state output by the cognitive state assessment model and the first time-frequency feature includes: Obtaining the cognitive state output by the cognitive state assessment model for multiple test times; Selecting the EEG signal corresponding to each cognitive state predicted correctly from the cognitive states output by the cognitive state assessment model of the plurality of test times; Calculating the second time-frequency feature corresponding to each cognitive state according to the EEG signal corresponding to each correctly predicted cognitive state; Compare the second time-frequency feature with the first time-frequency feature to obtain a difference; According to the difference, the target cognitive state of the current subject is adjusted.

8. A cognitive state assessment and control system based on few-electrode EEG signals, characterized in that: The system comprises: A data acquisition unit, used to acquire synchronously collected scalp EEG signals, functional magnetic resonance imaging data and structural magnetic resonance imaging data; A data processing unit, configured to preprocess the structural MRI data to obtain preprocessed structural MRI data; preprocess the functional MRI data based on the preprocessed structural MRI data to obtain preprocessed functional MRI data; and preprocess the scalp EEG signal to obtain a preprocessed scalp EEG signal; a data selection unit, configured to select a blood oxygen level dependent signal feature set related to a preset cognitive state according to the preprocessed functional magnetic resonance imaging data, wherein the blood oxygen level dependent signal feature set includes at least one of a brain region blood oxygen level dependent signal, a neural circuit blood oxygen level dependent signal, and a brain network blood oxygen level dependent signal; a model training unit, configured to use the blood oxygen level dependent signal feature set as a label of the preprocessed scalp EEG signal, and determine a trained first prediction model based on the label and the preprocessed scalp EEG signal; A model output unit, used for inputting the target scalp EEG signal into the trained first prediction model to obtain a target few-electrode EEG signal, wherein the target few-electrode EEG signal is an EEG signal of a small number of electrodes associated with the label; A data evaluation unit, used for filtering the target few-electrode EEG signal, and inputting the filtered target few-electrode EEG signal into the constructed cognitive state evaluation model to obtain the cognitive state output by the cognitive state evaluation model; The data control unit is used to control the target cognitive state of the current subject according to the cognitive state output by the cognitive state assessment model.

9. An electronic device, characterized in that: It includes at least one control processor and a memory for communicating with the at least one control processor; the memory stores instructions that can be executed by the at least one control processor, and the instructions are executed by the at least one control processor so that the at least one control processor can execute the cognitive state assessment and regulation method of few-electrode EEG signals as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to enable a computer to execute the cognitive state assessment and regulation method based on a few-electrode EEG signal as described in any one of claims 1 to 7.

Images