Mental state determination method, device and equipment based on electroencephalogram signals and medium

Abstract

The invention provides a mental state determination method, device and equipment based on an electroencephalogram signal and a medium, and relates to the technical field of electroencephalogram signal processing, the method comprises the following steps: acquiring an acquired preprocessed target electroencephalogram signal of a target object; performing multi-dimensional feature extraction on the target electroencephalogram signal to obtain a relaxation degree feature, a fatigue degree feature, a concentration degree feature, a target frequency feature and a signal dynamic feature; and determining the mental state level of the target object based on the relaxation degree feature, the fatigue degree feature, the concentration degree feature, the target frequency feature and the signal dynamic feature. According to the method, a more complete and more accurate mental state assessment result can be generated by integrating multi-dimensional features, and the accuracy and comprehensiveness of mental state assessment are improved.

Description

Technical Field

[0001] The present application relates to the technical field of electroencephalogram (EEG) signal processing, and in particular to a method, device, equipment and medium for determining a mental state based on EEG signals. Background Art

[0002] Electroencephalogram (EEG) is an electrophysiological signal that reflects brain neuron activity and contains rich information about emotions and cognitive states. Because it uses a non-invasive method and has low acquisition costs and real-time characteristics, EEG signal analysis has become an important tool for studying human cognitive activities and mental states. In recent years, with the rapid development of brain-computer interface technology, EEG-based mental state detection technology has made significant progress. Studies have shown that different frequency bands in EEG signals are closely related to human cognitive and emotional states: Waves reflect cognitive activity and attention levels; waves are associated with memory and emotional processing; The EEG wave is related to the state of relaxation and attention suppression. These research findings have laid a theoretical foundation for the development of EEG-based mental state detection systems. Therefore, how to accurately evaluate the user's mental state based on EEG signals is an urgent problem to be solved. Summary of the invention

[0003] The present application provides a method, device, equipment and medium for determining mental state based on EEG signals, which can improve the accuracy of mental state assessment based on EEG signals. The technical solution is as follows: According to one aspect of the present application, a method for determining a mental state based on an electroencephalogram signal is provided, the method comprising: Acquire the collected target EEG signal of the target object after preprocessing; Performing multi-dimensional feature extraction on the target EEG signal to obtain relaxation features, fatigue features, concentration features, target frequency features, and signal dynamic features; The mental state level of the target object is determined based on the relaxation feature, the fatigue feature, the concentration feature, the target frequency feature, and the signal dynamic feature.

[0004] According to another aspect of the present application, a device for determining a mental state based on an electroencephalogram signal is provided, the device comprising: An acquisition module is used to acquire the target EEG signal of the target object after preprocessing; A feature extraction module is used to extract multi-dimensional features of the target EEG signal to obtain relaxation features, fatigue features, concentration features, target frequency features and signal dynamic features; The first determination module is used to determine the mental state level of the target object based on the relaxation feature, the fatigue feature, the concentration feature, the target frequency feature and the signal dynamic feature.

[0005] According to one aspect of the present application, an electronic device is provided, comprising: a processor and a memory storing a program, wherein the program comprises instructions, and when the instructions are executed by the processor, the processor executes the mental state determination method based on EEG signals as described above.

[0006] According to another aspect of the present application, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to cause the computer to execute the mental state determination method based on EEG signals as described above.

[0007] According to another aspect of the present application, a computer program product is provided, the computer program product comprising computer instructions, the computer instructions being stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the above-mentioned method for determining a mental state based on an EEG signal.

[0008] The beneficial effects brought by the technical solution provided by the embodiment of the present application include at least: By extracting multi-dimensional features from the target EEG signal of the target object, mainly including: relaxation features, fatigue features, concentration features, target frequency features and signal dynamic features; and combining the multi-dimensional features to provide users with a complete mental state assessment result, that is, the mental state level; compared with the single indicator used in related technologies to determine the mental state, the comprehensive multi-dimensional features can generate a more complete and accurate mental state assessment result, which improves the accuracy and comprehensiveness of the mental state assessment, and thereby improves the accuracy of subsequent adjustments and processing based on the mental state assessment results. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Further details, features and advantages of the present application are disclosed in the following description of exemplary embodiments in conjunction with the accompanying drawings, in which: Figure 1 A schematic diagram showing an example system in which various methods described herein may be implemented according to an exemplary embodiment of the present application; Figure 2 A flowchart of a method for determining a mental state based on an EEG signal according to an exemplary embodiment of the present application is shown; Figure 3 is a schematic diagram of determining a mental state based on an EEG signal provided by an exemplary embodiment of the present application; Figure 4A flowchart of another method for determining a mental state based on an EEG signal according to an exemplary embodiment of the present application is shown; Figure 5 is a schematic diagram of determining a mental state based on an EEG signal provided by another exemplary embodiment of the present application; Figure 6 A flowchart of another method for determining a mental state based on an EEG signal according to an exemplary embodiment of the present application is shown; Figure 7 is a schematic diagram of the structure of a device for determining a mental state based on an electroencephalogram signal provided in an embodiment of the present application; Figure 8 A structural block diagram of an exemplary electronic device that can be used to implement the embodiments of the present application is shown. DETAILED DESCRIPTION

[0010] The embodiments of the present application will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present application are shown in the accompanying drawings, it should be understood that the present application can be implemented in various forms and should not be construed as being limited to the embodiments described herein. Instead, these embodiments are provided to provide a more thorough and complete understanding of the present application. It should be understood that the drawings and embodiments of the present application are only for exemplary purposes and are not intended to limit the scope of protection of the present application.

[0011] It should be understood that the various steps described in the method implementation of the present application can be performed in different orders and / or performed in parallel. In addition, the method implementation may include additional steps and / or omit the steps shown. The scope of the present application is not limited in this respect.

[0012] The term "including" and its variations used herein are open inclusions, i.e., "including but not limited to". The term "based on" means "based at least in part". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one other embodiment"; the term "some embodiments" means "at least some embodiments". The relevant definitions of other terms will be given in the following description. It should be noted that the concepts of "first", "second", etc. mentioned in this application are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units. It should be noted that the modifications of "one" and "multiple" mentioned in this application are illustrative rather than restrictive. Those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more". The names of the messages or information exchanged between multiple devices in the embodiments of this application are only for illustrative purposes, and are not used to limit the scope of these messages or information.

[0013] The following describes the solution of the present invention with reference to the accompanying drawings, and illustrates the technical solution provided by the embodiments of the present application in detail through specific embodiments and their application scenarios.

[0014] Most current EEG monitoring devices are limited to outputting a single feature. Although this single feature can reflect the user's mental state to a certain extent, the output of a single indicator is often difficult to fully reflect the user's overall mental state, resulting in low accuracy in the detection of the mental state, which in turn affects the accuracy of subsequent actual operations performed based on the mental state detection results. Therefore, in order to solve the problem of low accuracy in mental state detection results caused by extracting a single feature in related technologies, the embodiment of the present application provides a method for comprehensive mental state assessment by integrating multiple indicators such as concentration, fatigue, relaxation, sample entropy, and individual alpha frequency, thereby improving the accuracy and comprehensiveness of mental state assessment.

[0015] Figure 1 A schematic diagram of an example system in which various methods described herein may be implemented according to an exemplary embodiment of the present application is shown. The system includes at least one device end 110 and a cloud server 120.

[0016] The device end 110 is a device that can collect and process EEG signals. For example, the device end 110 can be a portable head-mounted device or a head-mounted massage device. After the device end 110 obtains the collected EEG signals, it pre-processes the EEG signals and extracts multi-dimensional features to obtain preset multi-dimensional features, and sends the multi-dimensional features to the cloud server 120.

[0017] Optionally, the device end 110 may also be a device capable of receiving an EEG signal sent by an EEG signal acquisition device (or an EEG signal acquisition component or apparatus).

[0018] A mental state classification model is deployed in the cloud server 120. After the cloud server 120 receives the multi-dimensional features fed back by the device end 110, the multi-dimensional features can be input into the mental state classification model to obtain the mental state classification result, i.e., the mental state level, and then the mental state level is fed back to the device end 110 so that the device end 110 can perform subsequent operations based on the mental state level.

[0019] Optionally, the mental state classification model may also be deployed in the device end 110, and the mental state classification purpose can be accomplished by the device end 110 itself.

[0020] Please refer to Figure 2 , which shows a flow chart of a method for determining a mental state based on an EEG signal according to an exemplary embodiment of the present application. Figure 1 The device end 110 shown in the figure is used as an example for explanation. Figure 2As shown, the method includes: Step 201, obtaining the target EEG signal of the target object after preprocessing.

[0021] In one possible implementation, the EEG signals of the target object are collected by an EEG signal collection component (or an EEG signal collection device), and the collected EEG signals are transmitted to the device end. After obtaining the collected EEG signals of the target object, the device end first performs a series of preprocessing operations on the collected EEG signals to obtain the target EEG signals that can be used for subsequent mental state analysis.

[0022] Among them, the preprocessing operation includes two filtering processes and artifact removal processes. The specific preprocessing process is: first, a bandpass filter is used to perform the first filtering process on the collected EEG signal to limit the signal frequency band to 0.5-50Hz, retain the target frequency range of EEG, which is a frequency range closely related to the mental state, remove low-frequency and high-frequency non-target noise, reduce the influence of external interference, and obtain the first EEG signal. Secondly, in order to effectively eliminate the power frequency noise, a notch filter is used to process the first EEG signal to remove the 50Hz power frequency interference and obtain the second EEG signal. After that, the independent component analysis (ICA) method is used to process the second EEG signal to further separate and eliminate physiological artifacts caused by eye movements, muscle activities, etc., to obtain more accurate EEG data, and finally obtain the preprocessed target EEG signal.

[0023] Optionally, the embodiment of the present application adopts a single-channel EEG acquisition device (or a single-channel EEG acquisition module) to collect the user's EEG signals in real time. The sampling rate is set to 250Hz, and 250 data points are collected per second to ensure that the time and frequency resolution is sufficient to capture subtle changes in brain activity. In terms of electrode selection, FP2 is used as the main acquisition electrode, FP1 is used as the reference electrode, and FPZ is used as the ground electrode. The single-channel EEG acquisition module is based on the brain area signal acquisition of FP2, combined with a low-power hardware architecture and a high-precision analog-to-digital conversion circuit. The collected EEG signals are transmitted to the data processing module in real time wirelessly via Bluetooth to ensure signal stability and convenience of data acquisition. Moreover, the single-channel EEG acquisition module can effectively reduce the hardware complexity and volume of the device end while ensuring that the EEG signals that can be used for subsequent mental state analysis are collected, reducing the cumbersome steps for users to wear, and improving portability and wearing comfort.

[0024] Step 202, extract multi-dimensional features of the target EEG signal to obtain relaxation features, fatigue features, concentration features, target frequency features and signal dynamic features.

[0025] Different from the related art, the embodiment of the present application extracts multi-dimensional features closely related to the mental state from the target EEG signal for comprehensive analysis of the mental state of the target object. In a corresponding possible implementation, the device performs multi-dimensional feature extraction on the pre-processed target EEG signal to obtain features of five dimensions, namely, relaxation features, fatigue features, concentration features, target frequency features, and signal dynamic features, so as to comprehensively capture the features related to the mental state in the EEG signal.

[0026] Step 203, determining the mental state level of the target object based on the relaxation characteristics, fatigue characteristics, concentration characteristics, target frequency characteristics and signal dynamic characteristics. After obtaining multi-dimensional features, the device can analyze and determine the mental state level of the target object based on relaxation characteristics, fatigue characteristics, concentration characteristics, target frequency characteristics and signal dynamic characteristics, so as to provide users with a complete mental state assessment result. Without the need for users to interpret various indicators item by item, users can understand their overall mental state in a short time, which is more convenient for application in daily life and work.

[0027] Figure 3 FIG. 1 is a schematic diagram of determining a mental state based on an EEG signal provided by an exemplary embodiment of the present application. Figure 3 As shown, the collected EEG signal 301 is preprocessed to obtain the target EEG signal 302; multi-dimensional feature extraction is performed on the target EEG signal 302 to obtain relaxation feature 303, fatigue feature 304, concentration feature 305, target frequency feature 306 and signal dynamic feature 307; the multi-dimensional features are combined to predict the mental state level 308 of the target object.

[0028] In summary, the embodiment of the present application provides a method for determining mental state based on EEG signals: by extracting multi-dimensional features from the target EEG signals of the target object, mainly including: relaxation features, fatigue features, concentration features, target frequency features and signal dynamic features; and combining the multi-dimensional features to provide the user with a complete mental state assessment result, that is, the mental state level; compared with the single indicator used in related technologies to determine the mental state, the comprehensive multi-dimensional features can generate a more complete and accurate mental state assessment result, which improves the accuracy and comprehensiveness of the mental state assessment, and thereby improves the accuracy of subsequent adjustment processing based on the mental state assessment results.

[0029] Considering that EEG signals are mainly Wave, wave and The wave is related to the user's mental state. When extracting multi-dimensional features from the target EEG signal, the calculation of each feature depends on Wave, wave and Wave related characteristics.

[0030] Please refer to Figure 4 , which shows a flowchart of another method for determining a mental state based on an EEG signal according to an exemplary embodiment of the present application. Figure 1 The device end 110 shown in the figure is used as an example for explanation. Figure 4 As shown, the method includes: Step 401, obtaining the target EEG signal of the target object after preprocessing.

[0031] The implementation of step 401 may refer to step 201 , and this embodiment will not be described in detail here.

[0032] Step 402: Perform spectrum estimation processing on the target EEG signal to obtain a target power spectrum density.

[0033] When extracting multi-dimensional features of the target EEG signal, the Welch method is first used to estimate the spectrum of the target EEG signal. The Welch method mainly reduces spectrum leakage and noise influence through segmentation, windowing and averaging, improves the smoothness and stability of spectrum estimation, and calculates the target power spectrum density corresponding to the target EEG signal, that is, the Welch power spectrum estimate.

[0034] Specifically, the process of performing spectrum estimation processing on the target EEG signal to obtain the target power spectrum density may include the following steps: (1) The target EEG signal is segmented with a window length of 1 second, and each segment has a 50% overlap, that is, the sliding window step is 0.5 seconds. Each segment of the signal is weighted using a Hanning window. The windowed signal is expressed as: (1) in, represents the signal after each window is added, It represents the nth segment signal after the target EEG signal is divided. is the window function.

[0035] (2) Apply Fast Fourier Transform (FFT) to each windowed signal to convert the time domain signal into a frequency domain representation. The calculation formula of FFT is: (2) in, represents the signal after each window is added, represents the kth frequency component of the frequency domain signal, and N is the number of sampling points in the window.

[0036] (3) The power spectrum density of each signal segment is calculated using the FFT result, using the formula: (3) Where N is the number of sampling points in the window, f s is the sampling rate, P k [ f ] is the power spectral density of the kth signal.

[0037] (4) The power spectral density of all segment signals is averaged to obtain the final Welch power spectrum estimate. The corresponding calculation formula is as follows: (4) Among them, P k [ f ] is the power spectrum density of the kth signal, M is the total number of signal segments, P Welch [ f ] is the target power spectral density.

[0038] By using the above formulas (1) to (4) to perform spectrum estimation processing on the target EEG signal, the target power spectrum density corresponding to the target EEG signal can be obtained.

[0039] Step 403: Determine the relaxation characteristic, fatigue characteristic, concentration characteristic and target frequency characteristic based on the target power spectrum density.

[0040] Among the multi-dimensional features, the relaxation feature, fatigue feature, concentration feature and target frequency feature are all related to the target power spectrum density corresponding to the target EEG signal. Based on the target power spectrum density, the relaxation feature, fatigue feature, concentration feature and target frequency feature corresponding to the target EEG signal can be calculated respectively.

[0041] Considering that the EEG signals related to mental state are mainly Wave, wave and Wave, when calculating the above characteristics based on the target power spectrum density, it mainly involves Wave, wave and In a corresponding exemplary example, step 403 may also include steps 403A to 403C.

[0042] Step 403A, based on the target power spectrum density, determine The energy of the first frequency band of the wave, The second frequency band energy of the wave and The third frequency band energy of waves.

[0043] because Wave, wave and If the frequency band of the wave is different, the target power spectrum density and frequency band can be combined to determine the The energy of the first frequency band of the wave, The second frequency energy of the wave and The specific process includes: obtaining The first frequency range of the wave, The second frequency range of the wave and The third frequency range of waves; based on The first frequency band range of the wave and the target power spectrum density can be determined The energy of the first frequency band of the wave; based on The second frequency band range of the wave and the target power spectrum density can be determined The second frequency band energy of the wave; based on The third frequency band range and target power spectrum density of the wave can be determined The third frequency band energy of waves.

[0044] Exemplarily, the formula for calculating the frequency band energy may be as shown in formula (5).

[0045] (5) in, f low and f high They are respectively the lower and upper frequency limits of the frequency bands corresponding to different waves. is the target power spectral density calculated using the Welch method. Wave, wave and The frequency band range of the wave is calculated by integrating the total energy within the corresponding frequency band, that is, the frequency band energy.

[0046] For example, The first frequency band corresponding to the wave is 8-12Hz, that is, the upper limit frequency is 12Hz and the lower limit frequency is 8Hz. The second frequency band corresponding to the wave is 4-8Hz, that is, the upper limit frequency is 8Hz and the lower limit frequency is 4Hz; The third frequency band corresponding to the wave is in the range of 12-30 Hz, that is, the upper limit frequency is 30 Hz and the lower limit frequency is 12 Hz. Substituting the upper limit frequency and lower limit frequency corresponding to each type of wave into formula (5) and combining them with the target power spectrum density, the frequency band energy corresponding to each type of wave can be obtained by integral calculation.

[0047] Step 403B, based on the energy of the first frequency band, the energy of the second frequency band and the energy of the third frequency band, determine the relaxation characteristic, the fatigue characteristic and the concentration characteristic.

[0048] Further comprehensive The energy of the first frequency band of the wave, The second frequency energy of the wave and The energy of the third frequency band of the wave is used to calculate the relaxation characteristics, fatigue characteristics and concentration characteristics.

[0049] Among them, the relaxation index (RI) (relaxation index is also the relaxation characteristic) is used to evaluate the individual's current relaxation state. Wave, wave and The energy ratio of the wave reflects the degree of relaxation. The calculation formula corresponding to the relaxation characteristic can be shown as formula (6): (6) Among them, RI represents the relaxation characteristic, yes The frequency band energy of the wave (i.e. the second frequency band energy), yes The frequency band energy of the wave band (i.e. the energy of the first frequency band), yes The energy of the third frequency band (i.e. the energy of the third frequency band). In formula (6), the higher Wave energy is often associated with states of relaxation and quietude, while wave and The energy of the wave is relatively low, indicating a distracted or sleepy state. The energy accounts for a large proportion and When the energy is relatively small, the relaxation index RI value will increase, indicating that the individual is in a more relaxed state. wave and As the energy of the wave increases, the RI value will decrease, which means the degree of relaxation decreases. From formula (6), we can see that after obtaining The energy of the first frequency band of the wave, The second frequency energy of the wave and After the third frequency band energy of the wave, first calculate The second frequency band energy of the wave The sum of the first energy of the third frequency band energy of the wave, and then The ratio of the energy of the first frequency band of the wave to the sum of the first energies is determined as the relaxation characteristic.

[0050] Among them, fatigue index (FI) is an important indicator for measuring individual fatigue status. It evaluates the degree of fatigue by weighted summation of energy in different frequency bands in EEG signals. Wave energy is associated with states of relaxation and sleepiness, while Wave energy is associated with wakefulness and alertness. wave and Wave energy relative to When the wave energy ratio increases, the fatigue index FI will increase accordingly, which indicates that the individual's fatigue state is enhanced. However, related studies have shown that changes in fatigue state may not only be related to the spectrum distribution, but also involve the volatility of the signal in the time dimension. The dynamic change of wave energy can capture the fatigue change trend of the signal over time and provide time sensitivity for the fatigue index. Based on the above analysis, the calculation formula of the fatigue index (i.e. fatigue characteristics) can be shown as formula (7): (7) Among them, w1, w2, and w3 are weight coefficients, which are used to control the influence of different items on fatigue characteristics, satisfying the relationship that w1 is greater than w2 and greater than w3, and the sum of w1, w2, and w3 is not 1. yes The frequency band energy of the wave (i.e. the second frequency band energy), yes The frequency band energy of the wave band (i.e. the energy of the first frequency band), yes The energy of the third frequency band. for The dynamic change average value of the energy of the corresponding frequency band of the wave can be calculated as shown in formula (8): (8) in, N Represents the number of windows, Represents the distance between adjacent sliding windows The relative change in wave energy, and The calculation formula can be shown as formula (9).

[0051] (9) in, is the window corresponding to time t The frequency band energy of the wave, Δt is the time interval, which is consistent with the step size of the sliding window (0.5 seconds).

[0052] From formula (7) to formula (9), we can see that after obtaining The energy of the first frequency band of the wave, The second frequency energy of the wave and After the third frequency band energy of the wave, first calculate The dynamic change mean of the energy of the corresponding frequency band of the wave is then calculated separately The second frequency band energy of the wave The first ratio between the energy of the third frequency band of the wave, and the calculation The energy of the first frequency band of the wave is The second ratio between the third frequency band energies of the waves, and then based on the first weight coefficient, the first ratio, the second weight coefficient, the second ratio, the third weight coefficient and The dynamic change mean of the energy in the corresponding frequency band of the wave is used to calculate the fatigue characteristics.

[0053] Among them, the concentration index (CI) is an important indicator used to measure an individual’s current state of concentration. Wave, wave and The energy ratio of the waves is calculated to reflect the characteristics of the individual's EEG activity. Waves are usually associated with higher-order cognitive activities such as alertness and concentration, while wave and The wave is related to states such as relaxation and distraction. The calculation formula of the corresponding concentration index (i.e. concentration characteristics) can be shown in formula (10).

[0054] (10) in, yes The frequency band energy of the wave (i.e. the second frequency band energy), yes The frequency band energy of the wave band (i.e. the energy of the first frequency band), yes The energy of the third frequency band (i.e. the energy of the third frequency band). A higher CI value indicates a stronger state of concentration.

[0055] From formula (10), we can see that after obtaining The energy of the first frequency band of the wave, The second frequency energy of the wave and After the third frequency band energy of the wave, first calculate The energy of the first frequency band of the wave and The sum of the second energy of the second frequency band energy of the wave, and then The ratio of the second frequency band energy of the wave to the sum of the second energies is determined as the relaxation characteristic.

[0056] Step 403C, based on the target power spectrum density and The first frequency band range of the wave determines the target frequency characteristics.

[0057] Since Individual Alpha Frequency (IAF) is unique to each person The peak frequency, usually in the range of 8-12 Hz, reflects the cognitive activity of an individual in a resting state. Therefore, the introduction of IAF in the process of mental state assessment has important physiological significance. The corresponding The first frequency band of the wave is used to determine the target frequency characteristic (i.e., IAF). Its calculation formula can be shown in formula (11).

[0058] (11) in, is the target power spectral density calculated by the Welch method, and f is the frequency. The calculated IAF is used for subsequent mental state analysis and classification. Higher IAF values ​​are usually associated with higher attention or higher cognitive arousal levels, while lower IAF values ​​may indicate that the user is in a more relaxed or less alert state. This feature has important reference value for real-time monitoring and evaluation of the user's mental state.

[0059] From formula (11), we can see that after obtaining the target power spectrum density and After the first frequency band of the wave is obtained, it can be substituted into formula (11) to calculate the target frequency characteristics, that is, the individual alpha frequency.

[0060] Step 404, obtaining the signal dynamic characteristics by performing segmented similarity calculation processing on the target EEG signal.

[0061] Sample entropy is a nonlinear dynamic indicator used to measure signal complexity. It can measure the randomness and complexity of the signal and help analyze the dynamic changes of EEG signals in different states (such as fatigue and relaxation). If the sample entropy value is high, it means that the signal has higher irregularity and complexity, which may correspond to a higher level of neural activity, such as concentration or wakefulness. On the contrary, a lower sample entropy means that the signal is more regular and more structured, which may reflect fatigue, relaxation or unconsciousness. Therefore, the sample entropy feature is also introduced in the embodiment of the present application for subsequent analysis of mental state. Exemplarily, the calculation process of sample entropy (that is, the dynamic feature of the signal) may include the following steps: (1) Calculate the matching probability :Divide the target EEG signal of each time window into time segments of length m=2. , the fragment of length 2 is: ,For each time segment, calculate its similarity with other segments, and the judgment criterion is that their difference (the absolute difference of all data points) is less than the threshold r.

[0062] (2) Calculate the matching probability : Increase the length of each time segment and construct a length of time segment and calculate the similarity again.

[0063] (3) Apply the sample entropy calculation formula, that is, formula (12), to calculate the sample entropy of the target EEG signal: (12) Where m is the embedding dimension, i.e., the length of the EEG signal segmentation in each time window. r is the similarity threshold, which is usually 20% to 30% of the signal standard deviation and is used to measure the similarity between segments.

[0064] Step 405 , normalizing the relaxation feature, fatigue feature, concentration feature, target frequency feature, and signal dynamic feature to obtain normalized multi-dimensional features.

[0065] Taking into account that the feature scales of different features may be different, in order to eliminate the influence of different feature scales on the subsequent mental state assessment, the relaxation features, fatigue features, concentration features, target frequency features and signal dynamic features are first normalized to enhance the comparability between different features and improve the classification effect of the mental state classification model.

[0066] For example, the normalization processing method can be a Z-Score normalization processing method, where for each feature X in the feature set, its mean is calculated. and standard deviation , and each eigenvalue x is standardized, and the calculation formula is as follows: (13) Among them, Z is the standardized eigenvalue, X is the original eigenvalue, is the mean of feature X, is the standard deviation of feature X. Through Z-Score standardization, all features are converted to the same scale, so that the classification model can weigh different features more effectively and improve classification accuracy.

[0067] Based on formula (13), after obtaining the relaxation feature, fatigue feature, concentration feature, target frequency feature and signal dynamic feature, first calculate the mean value among the five features and the standard deviation among the five features; secondly, based on the mean, standard deviation and relaxation feature, determine the normalized relaxation feature; based on the mean, standard deviation and fatigue feature, determine the normalized fatigue feature; based on the mean, standard deviation and concentration feature, determine the normalized concentration feature; based on the mean, standard deviation and target frequency feature, determine the normalized target frequency feature; based on the mean, standard deviation and signal dynamic feature, determine the normalized signal dynamic feature, thereby obtaining the normalized multi-dimensional features. Step 406, sending the normalized multi-dimensional features to the cloud server. After receiving the normalized multi-dimensional features, the cloud server is used to input the normalized multi-dimensional features into the mental state classification model to obtain the mental state level.

[0068] In order to reduce the power consumption of the device side, the mental state classification model can be deployed in the cloud server, which solves the problem of computing resource limitations on the device side in real-time signal analysis. The device only performs basic signal preprocessing locally and leaves the core computing tasks to the cloud. This not only reduces the power consumption of portable devices and extends battery life, but also improves classification accuracy and response speed. This combination of device and cloud ensures both real-time performance and detection accuracy, enabling the device to stably provide users with efficient and accurate mental state detection under limited resource conditions.

[0069] In a possible implementation, after the normalized multi-dimensional features are obtained on the device side, the normalized multi-dimensional features can be sent to the cloud server. After receiving the normalized multi-dimensional features, the cloud server inputs the normalized multi-dimensional features into the mental state classification model, outputs the probability of each classification result, and then determines the classification result with the highest probability as the mental state level of the target object.

[0070] Exemplarily, if the preset classification results include three levels: "excellent mental state", "good mental state" and "poor mental state", the mental state classification model can predict the probability of the mental state level belonging to these three levels based on the multi-dimensional features of the input: the probability of belonging to the level of "excellent mental state" is 10%, the probability of belonging to the level of "good mental state" is 5%, and the probability of belonging to the level of "poor mental state" is 85%. The level with the highest probability (i.e., poor mental state) is determined as the mental state level of the target object.

[0071] Optionally, in other possible implementations, different feature weights can be set for the five features respectively, and the feature weights and the five features can be input into the mental state classification model to predict the final mental state level. Since the feature weights of different features are set in advance, when predicting the mental state level, different features have different degrees of influence on the final classification result.

[0072] For example, the feature weights of different features can be adjusted according to the application scenario of the mental state level. For example, in the learning scenario, the user's concentration may be more concerned, so the concentration feature can be set to have a higher feature weight; in the vehicle driving scenario, the user's fatigue level may be more concerned, so the fatigue feature can be set to have a higher feature weight; for example, in the massage scenario, the user's relaxation level may be more concerned, so the relaxation feature can be set to have a higher feature weight. By adjusting the feature weights of different features, the mental state level prediction can be more adapted to the needs of the current application scenario while taking into account comprehensiveness. Step 407, receiving the mental state level of the target object fed back by the cloud server.

[0073] After predicting the mental state level of the target object, the cloud server can feed back the mental state level to the device end, and the device end receives the mental state level of the target object fed back by the cloud server.

[0074] Step 408, input the normalized multi-dimensional features into the mental state classification model to obtain the mental state level output by the mental state classification model.

[0075] Optionally, if the computing resources on the device side are sufficient, the mental state classification model can also be deployed on the device side. After the normalized multi-dimensional features are obtained on the device side, the normalized multi-dimensional features can be directly input into the mental state classification model, and the probabilities of each classification result can be output, and the classification result with the highest probability can be determined as the mental state level of the target object.

[0076] Figure 5 FIG. 1 is a schematic diagram of determining a mental state based on an EEG signal provided by another exemplary embodiment of the present application. Figure 5 As shown, the collected EEG signal 501 is preprocessed to obtain a target EEG signal 502; the target EEG signal 502 is subjected to spectrum estimation processing to obtain a target power spectrum density 503; based on the power spectrum density 503, the target The energy of the first frequency band of the wave is 504. The second frequency band energy of the wave is 505 and The third frequency band energy 506 of the wave; based on the first frequency band energy 504, the second frequency band energy 505 and the third frequency band energy 506, respectively determine the relaxation characteristic 507, the fatigue characteristic 508 and the concentration characteristic 509; based on the target power spectrum density 503 and The first frequency band range of the wave is used to determine the target frequency characteristics 510; and the signal dynamic characteristics 511 are obtained by performing segmented similarity calculation on the target EEG signal; and the normalized multi-dimensional characteristics are input into the mental state classification model 512 to obtain the mental state level 513 of the target object.

[0077] In order to enable the mental state classification model to have the mental state classification function, the mental state classification model needs to be trained in advance. The construction process of the mental state classification model may include the following steps: (1) Obtaining sample EEG data of multiple sample subjects; (2) Labeling the mental state level based on the subjective ratings of the sample subjects; (3) Filtering and removing artifacts from sample EEG data; (4) Extract features from the sample EEG data to obtain sample relaxation features, sample fatigue features, sample concentration features, sample frequency features, and sample signal dynamic features, and perform normalization processing.

[0078] (5) Construct a mental state classification model.

[0079] Specifically, the normalized multi-dimensional sample features are input into the mental state classification model to obtain the prediction results of the mental state classification model, and then the mental state classification model is trained based on the loss between the prediction results and the annotation combination (mental state level annotation); after multiple rounds of training, until the loss value of the mental state classification model is less than the preset value, it is determined that the mental state classification model is completed.

[0080] In other possible scenarios, the device side can also continuously accumulate sample data based on the user's feedback operations to update the mental state classification model, so that the mental state classification model can gradually learn the personalized mental state characteristics of the user of this device side. Exemplarily, after predicting the mental state level, the device side can feedback the mental state level to the user. If the user feedbacks that the mental state level assessment is incorrect, the feedback operation can be triggered. The corresponding device side uses the multi-dimensional features and mental state level involved in this prediction process as negative samples, and uses the multi-dimensional features and the correct mental state level (user feedback) as positive samples. After the number of negative samples and positive samples accumulates to a preset number, they can be used to update the model parameters of the mental state classification model.

[0081] Optionally, if the mental state classification model is deployed on the device side, the model parameters can be updated directly; if the mental state classification model is deployed on a cloud server, the device side can periodically send the collected positive and negative samples to the cloud server, and the cloud server is used to update the model parameters of the mental state classification model.

[0082] In this embodiment, a method for extracting relaxation features, fatigue features, concentration features, target frequency features and signal dynamic features is provided, so that features related to mental state can be captured from multiple aspects, thereby improving the accuracy of subsequent mental state classification; moreover, by deploying the mental state classification model on a cloud server, the computing resources of the cloud server can be used to quickly and accurately respond to mental state classification requests.

[0083] After obtaining the mental state classification result, that is, the mental state level, the mental state level can be applied to a variety of scenarios. For example, if the device is a portable head-mounted massage device, the massage parameters can be adjusted based on the mental state level; or, if the device is a portable learning device, personalized learning prompts can be given based on the mental state level; if the device is a vehicle-mounted terminal, driver safety prompts can be given based on the mental state level; if the device is an electroencephalogram detection head-mounted device, personalized work and rest prompts can be given based on the mental state level, etc.

[0084] Take the portable head-mounted massage device as an example, please refer to Figure 6 , which shows a flowchart of another method for determining a mental state based on an EEG signal according to an exemplary embodiment of the present application. Figure 1 The device end 110 shown in the figure is used as an example for explanation. Figure 6 As shown, the method includes: Step 601, obtaining the target EEG signal of the target object after preprocessing.

[0085] Step 602, performing multi-dimensional feature extraction on the target EEG signal to obtain relaxation features, fatigue features, concentration features, target frequency features and signal dynamic features.

[0086] Step 603, determining the mental state level of the target object based on the relaxation characteristics, fatigue characteristics, concentration characteristics, target frequency characteristics and signal dynamic characteristics. The implementation of steps 601 to 603 may refer to the above embodiment, and will not be described in detail in this embodiment.

[0087] Step 604, determining target massage parameters that match the mental state level, with different mental state levels corresponding to different target massage parameters.

[0088] Among them, the target head device is a head-mounted massage device, in which different massage parameters can be pre-configured according to different mental state levels. The massage parameters may include massage intensity, massage rhythm, massage mode, etc.

[0089] When a user uses a head-mounted massage device, the user's target EEG signal can be collected in real time to determine the user's current mental state level, and then the corresponding target massage parameters are matched according to the mental state level to achieve personalized massage services.

[0090] Step 605: Control the target head device to perform a massage operation based on the target massage parameters.

[0091] Then, the target head device is controlled to perform a massage operation based on the target massage parameters. For example, when the user's mental state level is "poor mental state", the target head device can automatically switch to a relaxation mode, with soothing music or a specific massage program, to help the user relieve fatigue and improve the massage experience.

[0092] In other possible scenarios, personalized suggestions can be made for users' work, study, and rest based on their mental state levels. For example, if the user's mental state level is in "excellent mental state", the device will suggest that the user make full use of the current state to complete efficient work or complex tasks to maximize work efficiency. When the user's mental state level is in "good mental state", the system will suggest that the user complete work of general difficulty. When it is detected that the user's mental state level is in a "poor mental state" state, the user will be prompted to rest or perform relaxation exercises in time to prevent excessive fatigue. This grading mechanism can not only effectively monitor the user's mental state, but also provide dynamically adjusted personalized intervention plans based on the state classification results, helping users to gradually improve their own psychological and physiological states, thereby improving their overall health and work efficiency.

[0093] In this embodiment, by matching different massage parameters to different mental state levels, different massage functions can be personalized according to the user's current mental state, thereby amplifying the comfort of the corresponding massage function of the target head device. Please refer to Figure 7 , which is a schematic diagram of the structure of a device for determining a mental state based on an EEG signal provided in an embodiment of the present application. For example, Figure 7 As shown, the device 700 includes: An acquisition module 701 is used to acquire a target EEG signal of a target object after preprocessing; The feature extraction module 702 is used to extract multi-dimensional features of the target EEG signal to obtain relaxation features, fatigue features, concentration features, target frequency features and signal dynamic features; The first determination module 703 is used to determine the mental state level of the target object based on the relaxation feature, the fatigue feature, the concentration feature, the target frequency feature and the signal dynamic feature.

[0094] Optionally, the feature extraction module 702 is further used to: Performing spectrum estimation processing on the target EEG signal to obtain a target power spectrum density; Based on the target power spectrum density, determining the relaxation characteristic, the fatigue characteristic, the concentration characteristic and the target frequency characteristic; The signal dynamic characteristics are obtained by performing segmented similarity calculation processing on the target EEG signal.

[0095] Optionally, the feature extraction module 702 is further used to: Based on the target power spectral density, determine The energy of the first frequency band of the wave, The second frequency band energy of the wave and The third frequency band energy of the wave; Determining the relaxation characteristic, the fatigue characteristic, and the concentration characteristic based on the first frequency band energy, the second frequency band energy, and the third frequency band energy; Based on the target power spectral density and the The first frequency band range of the wave determines the target frequency characteristics.

[0096] Optionally, the feature extraction module 702 is further used to: Get the The first frequency range of the wave, the The second frequency range of the wave and the The third frequency range of waves; Based on the first frequency band range and the target power spectrum density, determine the The energy of the wave in the first frequency band; Based on the second frequency band and the target power spectrum density, determine the the energy of the second frequency band of the wave; Based on the third frequency band and the target power spectrum density, determine the The third frequency band energy of the wave.

[0097] Optionally, the first determining module 703 is further configured to: Normalizing the relaxation feature, the fatigue feature, the concentration feature, the target frequency feature, and the signal dynamic feature to obtain normalized multi-dimensional features; The normalized multi-dimensional features are input into a mental state classification model to obtain the mental state level output by the mental state classification model.

[0098] Optionally, the first determining module 703 is further configured to: Normalizing the relaxation feature, the fatigue feature, the concentration feature, the target frequency feature, and the signal dynamic feature to obtain normalized multi-dimensional features; Sending the normalized multi-dimensional features to a cloud server, wherein the cloud server is configured to input the normalized multi-dimensional features into a mental state classification model after receiving the normalized multi-dimensional features to obtain the mental state level; Receive the mental state level of the target object fed back by the cloud server.

[0099] Optionally, the device further comprises: A second determination module is used to determine a target massage parameter that matches the mental state level, and different mental state levels correspond to different massage parameters; A control module is used to control the target head device to perform a massage operation based on the target massage parameters.

[0100] In summary, the embodiment of the present application provides a method for determining mental state based on EEG signals: by extracting multi-dimensional features from the target EEG signals of the target object, mainly including: relaxation features, fatigue features, concentration features, target frequency features and signal dynamic features; and combining the multi-dimensional features to provide the user with a complete mental state assessment result, that is, the mental state level; compared with the single indicator used in related technologies to determine the mental state, the comprehensive multi-dimensional features can generate a more complete and accurate mental state assessment result, which improves the accuracy and comprehensiveness of the mental state assessment, and thereby improves the accuracy of subsequent adjustment processing based on the mental state assessment results.

[0101] The exemplary embodiment of the present application also provides an electronic device, comprising: at least one processor; and a memory connected to the at least one processor. The memory stores a computer program that can be executed by the at least one processor, and when the computer program is executed by the at least one processor, it is used to enable the electronic device to perform the mental state determination method based on EEG signals according to the embodiment of the present application. It should be noted that the electronic device is the device end shown in the above embodiment.

[0102] An exemplary embodiment of the present application also provides a non-transitory computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor of a computer, is used to cause the computer to perform a method for determining a mental state based on electroencephalogram signals according to an embodiment of the present application.

[0103] An exemplary embodiment of the present application also provides a computer program product, including a computer program, wherein when the computer program is executed by a processor of a computer, it is used to cause the computer to perform the method for determining a mental state based on an electroencephalogram signal according to an embodiment of the present application.

[0104] refer to Figure 8 , the structural block diagram of the electronic device 800 that can be used as the server or client of the present application will now be described, which is an example of the hardware device that can be applied to various aspects of the present application. The electronic device is intended to represent various forms of digital electronic computer equipment, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices and other similar computing devices. The components shown herein, their connections and relationships, and their functions are only examples, and are not intended to limit the implementation of the present application described and / or required herein.

[0105] like Figure 8 As shown, the electronic device 800 includes a computing unit 801, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 802 or a computer program loaded from a storage unit 808 to a random access memory (RAM) 803. In the RAM 803, various programs and data required for the operation of the device 800 can also be stored. The computing unit 801, the ROM 802, and the RAM 803 are connected to each other via a bus 804. An input / output (I / O) interface 805 is also connected to the bus 804.

[0106] A plurality of components in the electronic device 800 are connected to the I / O interface 805, including: an input unit 806, an output unit 807, a storage unit 808, and a communication unit 809. The input unit 806 may be any type of device capable of inputting information to the electronic device 800, and the input unit 806 may receive input digital or character information, and generate key signal inputs related to user settings and / or function control of the electronic device. The output unit 807 may be any type of device capable of presenting information, and may include, but is not limited to, a display, a speaker, a video / audio output terminal, a vibrator, and / or a printer. The storage unit 808 may include, but is not limited to, a disk, an optical disk. The communication unit 809 allows the electronic device 800 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks, and may include, but is not limited to, a modem, a network card, an infrared communication device, a wireless communication transceiver, and / or a chipset, such as a Bluetooth device, a WiFi device, a WiMax device, a cellular communication device, and / or the like.

[0107] Optionally, the electronic device 800 is further provided with a single-channel EEG signal acquisition module (not shown in the figure) for acquiring EEG signals and transmitting the EEG signals to a signal processor of the electronic device 800 for processing the EEG signals.

[0108] The computing unit 801 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 801 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 801 performs the various methods and processes described above. For example, in some embodiments, Figure 2 , Figure 4 , Figure 6 The method shown may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 808. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 800 via the ROM 802 and / or the communication unit 809. In some embodiments, the computing unit 801 may be configured to execute the computer program by any other suitable means (e.g., by means of firmware). Figure 2 , Figure 4 , Figure 6 The method shown.

[0109] The program code for implementing the method of the present application can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, implements the functions / operations specified in the flow chart and / or block diagram. The program code can be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0110] In the context of the present application, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or device, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0111] As used herein, the terms "machine-readable medium" and "computer-readable medium" refer to any computer program product, apparatus, and / or device (e.g., disk, optical disk, memory, programmable logic device (PLD)) for providing machine instructions and / or data to a programmable processor, including a machine-readable medium that receives machine instructions as a machine-readable signal. The term "machine-readable signal" refers to any signal for providing machine instructions and / or data to a programmable processor.

[0112] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0113] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0114] A computer system may include clients and servers. Clients and servers are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship to each other.

Claims

1. A method for determining mental state based on EEG signals, characterized in that: The method comprises: Acquire the collected target EEG signal of the target object after preprocessing; Performing multi-dimensional feature extraction on the target EEG signal to obtain relaxation features, fatigue features, concentration features, target frequency features, and signal dynamic features; The mental state level of the target object is determined based on the relaxation feature, the fatigue feature, the concentration feature, the target frequency feature, and the signal dynamic feature.

2. The method according to claim 1, characterized in that The multi-dimensional feature extraction of the target EEG signal is performed to obtain relaxation features, fatigue features, concentration features, target frequency features and signal dynamic features, including: Performing spectrum estimation processing on the target EEG signal to obtain a target power spectrum density; Based on the target power spectrum density, determining the relaxation characteristic, the fatigue characteristic, the concentration characteristic and the target frequency characteristic; The signal dynamic characteristics are obtained by performing segmented similarity calculation processing on the target EEG signal.

3. The method according to claim 2, characterized in that The determining, based on the target power spectrum density, the relaxation characteristic, the fatigue characteristic, the concentration characteristic and the target frequency characteristic comprises: Based on the target power spectral density, determine The energy of the first frequency band of the wave, The second frequency band energy of the wave and The third frequency band energy of the wave; Determining the relaxation characteristic, the fatigue characteristic, and the concentration characteristic based on the first frequency band energy, the second frequency band energy, and the third frequency band energy; Based on the target power spectral density and the The first frequency band range of the wave determines the target frequency characteristics.

4. The method according to claim 3, characterized in that The determining, based on the target power spectral density, The energy of the first frequency band of the wave, The second frequency band energy of the wave and The third frequency band energy of waves includes: Get the The first frequency range of the wave, the The second frequency range of the wave and the The third frequency range of waves; Based on the first frequency band range and the target power spectrum density, determine the The energy of the wave in the first frequency band; Based on the second frequency band and the target power spectrum density, determine the the energy of the second frequency band of the wave; Based on the third frequency band and the target power spectrum density, determine the The third frequency band energy of the wave.

5. The method according to any one of claims 1 to 4, characterized in that: The determining the mental state level of the target object based on the relaxation feature, the fatigue feature, the concentration feature, the target frequency feature and the signal dynamic feature includes: Normalizing the relaxation feature, the fatigue feature, the concentration feature, the target frequency feature, and the signal dynamic feature to obtain normalized multi-dimensional features; The normalized multi-dimensional features are input into a mental state classification model to obtain the mental state level output by the mental state classification model.

6. The method according to any one of claims 1 to 4, characterized in that: The determining the mental state level of the target object based on the relaxation feature, the fatigue feature, the concentration feature, the target frequency feature and the signal dynamic feature also includes: Normalizing the relaxation feature, the fatigue feature, the concentration feature, the target frequency feature, and the signal dynamic feature to obtain normalized multi-dimensional features; Sending the normalized multi-dimensional features to a cloud server, wherein the cloud server is configured to input the normalized multi-dimensional features into a mental state classification model after receiving the normalized multi-dimensional features to obtain the mental state level; Receive the mental state level of the target object fed back by the cloud server.

7. The method according to any one of claims 1 to 4, characterized in that: The method further comprises: Determining target massage parameters that match the mental state level, where different mental state levels correspond to different target massage parameters; The target head device is controlled to perform a massage operation based on the target massage parameter.

8. A device for determining mental state based on EEG signals, characterized in that: The device comprises: An acquisition module is used to acquire the target EEG signal of the target object after preprocessing; A feature extraction module is used to extract multi-dimensional features of the target EEG signal to obtain relaxation features, fatigue features, concentration features, target frequency features and signal dynamic features; The first determination module is used to determine the mental state level of the target object based on the relaxation feature, the fatigue feature, the concentration feature, the target frequency feature and the signal dynamic feature.

9. An electronic device, comprising: processor; as well as Memory for storing programs, The program includes instructions, which, when executed by the processor, cause the processor to perform the method for determining a mental state based on an electroencephalogram signal according to any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause the computer to execute the method for determining a mental state based on an electroencephalogram signal according to any one of claims 1 to 7.

Images