Cognitive state evaluation method in closed environment

Abstract

The invention relates to a cognitive state evaluation method and device in a closed environment, electronic equipment and a storage medium. The method comprises the following steps: based on a preset emotion cognition model and a preset attention cognition model respectively, acquiring emotion cognition features and attention cognition features; carrying out dimension reduction processing on the emotion cognition features and the attention cognition features; and by taking the emotion cognition features and the attention cognition features subjected to dimension reduction processing as input, based on a support vector machine (SVM) classifier learning cognition feature model, evaluating a cognition state, and generating an evaluation result. The brain cognitive state is evaluated from subjective and objective angles by adopting the self-evaluation scale and the detected electroencephalogram features and behavioral features, recognition and classification are performed by using a machine learning method, the brain evaluation state is evaluated, and a method basis is provided for detection of the brain cognitive state under a long-term closed operation condition.

Description

Background Art

[0002] In many operations, operators or workers need to work and live in a specific closed environment. Studies have shown that a closed isolation environment is different from the normal environment in which humans live. There is social isolation and space closure. Moreover, people who stay in a closed isolation environment for a long time will have abnormal reactions such as brain cognitive function.

[0003] EEG has recently been used in two closed environment (space) studies to explore how closed environments affect the physiological and emotional state of the brain. One study found that three months of isolation led to a general decrease in brain activity, as well as a decrease in physical state and motivation. Another study investigated the impact of closed environments on stress responses. The study evaluated the brain activity of participants from six different countries who lived in a simulated spacecraft for 520 days in complete isolation. The study found that global alpha and beta activity in the cerebral cortex decreased significantly from the beginning to the end of the mission. A recent study on the mental health of soldiers stationed in a fully closed environment at the plateau found that the fully closed environment had a certain impact on the mental health of the soldiers stationed in the training, mainly manifested in emotional irritability, anxiety, depression and other aspects.

[0004] A good level of attention of operators in the task is an important condition for completing the task. Posner et al. (2002, 2014) first subdivided the attention network into three subnetworks: alerting, orienting, and executive control. The three attention networks reflect the ability of individuals to obtain and maintain alertness to a certain type of information or target, selective attention to external useful information, and processing conflicting information. These three components of attention can be measured by the Attention Network Test (ANT) developed by Fan et al. ANT has been used by a large number of researchers and has been proven to be an effective tool for measuring the efficiency of different attention networks (Macleod et al., 2010). The attention network experimental paradigm has been well theoretically demonstrated and experimentally proven. Due to its high operability and short testing time, it is widely used in psychology, neuroscience, and medicine. ANT relies on performance measurements of stimuli under different conditions, such as response time and accuracy, and the calculation of difference scores for target alertness, orientation, and executive control.

[0005] Therefore, one or more methods are needed to solve the above problems.

[0006] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present disclosure, and therefore may include information that does not constitute the prior art known to ordinary technicians in the field. Summary of the invention

[0007] The purpose of the present disclosure is to provide a method, device, electronic device and computer-readable storage medium for evaluating cognitive status in a closed environment, thereby overcoming one or more problems caused by limitations and defects of related technologies at least to a certain extent.

[0008] According to one aspect of the present disclosure, a method for evaluating cognitive status in a closed environment is provided, comprising:

[0009] Collecting emotion cognition features and attention cognition features based on a preset emotion cognition model and a preset attention cognition model respectively;

[0010] Performing dimensionality reduction processing on the emotion cognitive features and attention cognitive features;

[0011] The emotion cognitive features and attention cognitive features that have undergone dimensionality reduction processing are used as input, and the cognitive state is evaluated based on a support vector machine (SVM) classifier learning cognitive feature model to generate an evaluation result.

[0012] In an exemplary embodiment of the present disclosure, the method of collecting emotion recognition features based on a preset emotion recognition model further includes:

[0013] Based on the Self-Rating Anxiety Scale (SAS), the Self-Rating Depression Scale (SDS), the Self-Rating Psychological Stress Questionnaire (PSET), and the Self-Rating Symptom Checklist (SCL-90), we obtained the scores of SAS, SDS, PSET, and the scores of somatization, obsession, interpersonal sensitivity, depression, anxiety, hostility, terror, paranoia, and psychosis in SCL-90.

[0014] Based on the emotional stroop EEG experiment, behavioral characteristics and EEG feature ERP were extracted to obtain a preset number of emotional behavioral characteristics and the amplitude of ERP components.

[0015] In an exemplary embodiment of the present disclosure, the method of collecting attention cognition features based on a preset attention cognition model further includes:

[0016] An attention network ANT EEG experiment was designed, and the behavioral characteristics of attention orientation network efficiency, attention alertness network efficiency, and attention executive control network efficiency were collected based on the attention network ANT experiment.

[0017] In an exemplary embodiment of the present disclosure, the method further includes:

[0018] The emotion cognitive features and attention cognitive features are subjected to dimensionality reduction processing based on factor analysis.

[0019] In an exemplary embodiment of the present disclosure, the method further includes:

[0020] Recruit a preset number of experimenters to participate in the emotion experiment and attention experiment;

[0021] In the emotion induction experiment, movie clips from the preset emotional video database were used to induce specific positive, neutral and negative emotions in the subjects to complete the collection of training samples;

[0022] In the attention experiment, based on the attention network ANT EEG experiment, after one hour of simulated driving task, the subjects' fatigue self-assessment results and ANT EEG data were collected to complete the training sample collection.

[0023] In an exemplary embodiment of the present disclosure, the method further includes:

[0024] Using a multi-kernel learning support vector machine to learn the reduced-dimensional emotion cognitive features and attention cognitive features, respectively, to establish an emotion state and attention level classification model, and to achieve an evaluation of the emotion state and attention level;

[0025] Based on the kernel function vector and its weight vector obtained by the EasyMKL algorithm, the combined kernel vector is calculated;

[0026] Combine the SVM classifier and the test set to complete the classification experiment.

[0027] In one aspect of the present disclosure, a cognitive state assessment device in a closed environment is provided, comprising:

[0028] A feature collection module, used to collect emotion recognition features and attention recognition features based on a preset emotion recognition model and a preset attention recognition model respectively;

[0029] A dimensionality reduction processing module, used for performing dimensionality reduction processing on the emotion cognitive features and attention cognitive features;

[0030] The cognitive state evaluation module is used to evaluate the cognitive state based on the support vector machine (SVM) classifier learning cognitive feature model using the emotion cognitive features and attention cognitive features that have undergone dimensionality reduction processing as input, and generate an evaluation result.

[0031] In one aspect of the present disclosure, there is provided an electronic device, comprising:

[0032] Processor; and

[0033] A memory having computer-readable instructions stored thereon, wherein the computer-readable instructions, when executed by the processor, implement the method according to any one of the above items.

[0034] In one aspect of the present disclosure, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the method according to any one of the above items is implemented.

[0035] In an exemplary embodiment of the present disclosure, a method for evaluating cognitive state in a closed environment is provided, wherein the method comprises: collecting emotional cognitive features and attention cognitive features based on a preset emotional cognitive model and a preset attention cognitive model respectively; performing dimensionality reduction processing on the emotional cognitive features and attention cognitive features; using the emotional cognitive features and attention cognitive features that have undergone dimensionality reduction processing as input, evaluating the cognitive state based on a support vector machine (SVM) classifier learning cognitive feature model, and generating an evaluation result. The present disclosure uses a self-assessment scale and measured EEG features and behavioral features to evaluate the brain's cognitive state from a subjective and objective perspective, respectively, and uses a machine learning method for identification and classification to achieve brain assessment status evaluation, providing a methodological basis for detecting the brain's cognitive state under long-term closed working conditions.

[0036] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The above and other features and advantages of the present disclosure will become more apparent by describing in detail example embodiments thereof with reference to the attached drawings.

[0038] Figure 1 A flowchart of a method for evaluating cognitive status in a closed environment according to an exemplary embodiment of the present disclosure is shown;

[0039] Figure 2 A schematic diagram of a method for evaluating cognitive status in a closed environment according to an exemplary embodiment of the present disclosure is shown;

[0040] Figure 3 A schematic diagram of an ANT experimental paradigm of a cognitive state assessment method in a closed environment according to an exemplary embodiment of the present disclosure is shown;

[0041] Figure 4 A schematic block diagram of a cognitive state assessment device in a closed environment according to an exemplary embodiment of the present disclosure is shown;

[0042] Figure 5 A block diagram schematically shows an electronic device according to an exemplary embodiment of the present disclosure;

[0043] Figure 6 A schematic diagram of a computer-readable storage medium according to an exemplary embodiment of the present disclosure is schematically shown. DETAILED DESCRIPTION

[0044] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in many forms and should not be construed as limited to the embodiments set forth herein; rather, these embodiments are provided so that this disclosure will be comprehensive and complete and will fully convey the concepts of the example embodiments to those skilled in the art. The same reference numerals in the figures represent the same or similar parts, and thus their repeated description will be omitted.

[0045] In addition, the described features, structures or characteristics may be combined in one or more embodiments in any suitable manner. In the following description, many specific details are provided to provide a full understanding of the embodiments of the present disclosure. However, those skilled in the art will appreciate that the technical solutions of the present disclosure may be practiced without one or more of the specific details, or other methods, components, materials, devices, steps, etc. may be adopted. In other cases, known structures, methods, devices, implementations, materials or operations are not shown or described in detail to avoid blurring the various aspects of the present disclosure.

[0046] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities may be implemented in software form, or these functional entities or parts of functional entities may be implemented in one or more software hardened modules, or these functional entities may be implemented in different networks and / or processor devices and / or microcontroller devices.

[0047] In this example embodiment, a method for evaluating cognitive status in a closed environment is first provided; Figure 1 As shown in , the cognitive state assessment method in a closed environment may include the following steps:

[0048] Step S110, collecting emotion cognition features and attention cognition features based on a preset emotion cognition model and a preset attention cognition model respectively;

[0049] Step S120, performing dimensionality reduction processing on the emotion cognitive features and attention cognitive features;

[0050] Step S130, using the emotion cognitive features and attention cognitive features that have undergone dimensionality reduction processing as input, evaluating the cognitive state based on a support vector machine (SVM) classifier learning cognitive feature model, and generating an evaluation result.

[0051] In an exemplary embodiment of the present disclosure, a method for evaluating cognitive state in a closed environment is provided, wherein the method comprises: collecting emotional cognitive features and attention cognitive features based on a preset emotional cognitive model and a preset attention cognitive model respectively; performing dimensionality reduction processing on the emotional cognitive features and attention cognitive features; using the emotional cognitive features and attention cognitive features that have undergone dimensionality reduction processing as input, evaluating the cognitive state based on a support vector machine (SVM) classifier learning cognitive feature model, and generating an evaluation result. The present disclosure uses a self-assessment scale and measured EEG features and behavioral features to evaluate the brain's cognitive state from a subjective and objective perspective, respectively, and uses a machine learning method for identification and classification to achieve brain assessment status evaluation, providing a methodological basis for detecting the brain's cognitive state under long-term closed working conditions.

[0052] Next, a cognitive state assessment method in a closed environment in this exemplary embodiment will be further described.

[0053] Embodiment 1:

[0054] In step S110, emotion recognition features and attention recognition features may be collected based on a preset emotion recognition model and a preset attention recognition model, respectively.

[0055] In the embodiment of this example, the method of collecting emotion recognition features based on a preset emotion recognition model further includes:

[0056] Based on the Self-Rating Anxiety Scale (SAS), the Self-Rating Depression Scale (SDS), the Self-Rating Psychological Stress Questionnaire (PSET), and the Self-Rating Symptom Checklist (SCL-90), we obtained the scores of SAS, SDS, PSET, and the scores of somatization, obsession, interpersonal sensitivity, depression, anxiety, hostility, terror, paranoia, and psychosis in SCL-90.

[0057] Based on the emotional stroop EEG experiment, behavioral characteristics and EEG feature ERP were extracted to obtain a preset number of emotional behavioral characteristics and the amplitude of ERP components.

[0058] In the embodiment of this example, the method of collecting attention cognition features based on a preset attention cognition model further includes:

[0059] An attention network ANT EEG experiment was designed, and the behavioral characteristics of attention orientation network efficiency, attention alertness network efficiency, and attention executive control network efficiency were collected based on the attention network ANT experiment.

[0060] In step S120, the emotion recognition features and attention recognition features may be subjected to dimensionality reduction processing.

[0061] In the embodiment of this example, the method further includes:

[0062] The emotion cognitive features and attention cognitive features are subjected to dimensionality reduction processing based on factor analysis.

[0063] In step S130, the emotion cognitive features and attention cognitive features that have undergone dimensionality reduction processing may be used as inputs, and the cognitive state may be evaluated based on a support vector machine (SVM) classifier learning cognitive feature model to generate an evaluation result.

[0064] In the embodiment of this example, the method further includes:

[0065] Recruit a preset number of experimenters to participate in the emotion experiment and attention experiment;

[0066] In the emotion induction experiment, movie clips from the preset emotional video database were used to induce specific positive, neutral and negative emotions in the subjects to complete the collection of training samples;

[0067] In the attention experiment, based on the attention network ANT EEG experiment, after one hour of simulated driving task, the subjects' fatigue self-assessment results and ANT EEG data were collected to complete the training sample collection.

[0068] In the embodiment of this example, the method further includes:

[0069] Using a multi-kernel learning support vector machine to learn the reduced-dimensional emotion cognitive features and attention cognitive features, respectively, to establish an emotion state and attention level classification model, and to achieve an evaluation of the emotion state and attention level;

[0070] Based on the kernel function vector and its weight vector obtained by the EasyMKL algorithm, the combined kernel vector is calculated;

[0071] Combine the SVM classifier and the test set to complete the classification experiment.

[0072] Embodiment 2:

[0073] In this exemplary embodiment, if Figure 2 As shown in the figure, the evaluation of brain cognition in a closed environment mainly considers the evaluation of emotional cognitive function and attention cognitive function. In the emotional evaluation, two measurement methods, scale and EEG, were used. Four emotional measurement scales were selected to extract the scale score characteristics. The EEG experiment of emotional stroop was used to obtain the EEG ERP amplitude characteristics and reaction time behavioral characteristics. The evaluation of attention cognitive function used the EEG experiment of attention network test to extract the behavioral characteristics related to the efficiency of attention alertness network, orientation network, and control network, as well as the EEG ERP amplitude characteristics.

[0074] In this exemplary embodiment, the scale measurement includes:

[0075] The scales used in this patent include the self-rating anxiety scale (SAS), the self-rating depression scale (SDS), the psychological stress self-evaluation test (PSET) and the symptom check list 90 (SCL-90) assessment tools.

[0076] The cut-off value of the SAS standard score is 50 points. 50-59 points indicate mild anxiety, 60-69 points indicate moderate anxiety, and ≥70 points indicate severe anxiety. A SAS standard score of ≥50 is considered positive.

[0077] The cut-off value of the SDS standard score is 53 points. 53-62 points indicate mild depression, 63-72 points indicate moderate depression, and >72 points indicate severe depression. An SDS standard score ≥53 points is considered positive.

[0078] PSET scoring standard, the sum of the scores of each item is the original score, and then converted into a standard score T. T = 50 + 10 (XM) / SD, where X is the original score, M is the mean, and SD is the standard deviation. T < 70 points indicates a normal level of psychological stress, and T ≥ 70 points indicates a high level of psychological stress.

[0079] SCL-90 includes 90 items, including somatization, obsession, interpersonal sensitivity, depression, anxiety, hostility, phobia, paranoia and psychosis. The 5-level score is 1 point for no symptoms, 2 points for mild symptoms, 3 points for moderate symptoms, 4 points for severe symptoms and 5 points for severe symptoms. The higher the score, the more severe the symptoms of the research subject, and the worse the mental health condition.

[0080] The parameters obtained in the self-rating scale include SAS score, SDS score, PSET score and the scores of somatization, obsession, interpersonal sensitivity, depression, anxiety, hostility, phobia, paranoia and psychoticism factors in SCL-90, totaling 12 subjective characteristics.

[0081] In the embodiment of this example, the emotional stroop EEG experiment and feature extraction include:

[0082] Emotional stroop EEG experiment design:

[0083] The experimental materials of the emotional Stroop task used pictures with faces and words superimposed. The experiment adopted an experimental design with two emotion types (pleasure and fear) and two conflict types (consistent and inconsistent). The face pictures used in the experiment were from the Chinese Facial Emotion Expression Picture System (CFAPS), from which 5 male faces of pleasure and fear and 5 female faces of pleasure and fear were selected, totaling 20 pictures. The face images were processed, and the words "pleasure" and "fear" were annotated on the face pictures to obtain 40 expression images. The experimental task included two conditions: consistent and inconsistent. The valence of the face and the annotated words was consistent, which was the stimulus of the consistent condition, and the valence of the face and the annotated words was inconsistent, which was the stimulus of the inconsistent condition. The task required to judge the emotional valence of the face. Under the inconsistent condition, the meaning of the word would interfere with the judgment of the face valence, and under the consistent condition, the meaning of the word would promote the judgment of the face valence. The emotional face images used in the practice were different from the pictures used in the formal experiment, but the processing principles were consistent.

[0084] In the experiment, the subjects were asked to sit about 80 cm in front of the computer. First, the subjects were shown the introductory words to let them understand the experimental conditions. Then, they clicked the space bar, and a "+" fixation point was displayed on the screen for 500ms, prompting the subjects to concentrate. Then, the stimulus picture was displayed, and the time was displayed. When they saw a stimulus picture, they needed to make an accurate judgment of the facial emotion as quickly as possible. If they judged it as joy, they pressed the "F" key, and if they judged it as fear, they pressed the "J" key. After pressing the key, the picture would disappear. The subjects needed to respond to the facial emotion within 1000ms. After 1000ms, the picture stimulus automatically disappeared and was considered to be a wrong judgment. After a blank interval of 800-1200ms, the next test would be carried out. During this process, the subjects needed to focus on suppressing the interference signals of the two words "joy" and "fear". There were 15 practice sessions before the formal test, and 120 trials in the formal test. There were 120 trials in the formal experiment, which were divided into 4 blocks, and the number of trials in the consistent and inconsistent conditions was the same. After each test, the researchers will give the subjects a certain amount of rest to reduce fatigue and ensure the credibility of the experiment.

[0085] Behavioral feature extraction:

[0086] The extracted behavioral features include reaction time and accuracy. The experiment is divided into four conditions, namely "happy face and labeled text are consistent", "happy face and labeled text are inconsistent", "fearful face and labeled text are consistent" and "fearful face and labeled text are inconsistent". The reaction time and accuracy of the subjects when the stimulus is presented in these four conditions are recorded respectively, for a total of 8 emotion-related behavioral features.

[0087] EEG feature extraction:

[0088] The EEG signal preprocessing process was performed in the EEGLAB plug-in (http: / / sccn.ucsd.edu / eeglab / ) in MATLAB 2017a (MathWorks, Natick, MA, USA) and self-written scripts for processing and analysis.

[0089] A series of offline preprocessing work is performed on the collected EEG raw signals. First, all the lead data are re-referenced and the reference electrode is reset to the whole brain average. Then the EEG data is band-pass filtered at 0.1Hz-30Hz using the Basic FIR filter. The ICA method is used to identify and remove eye movement artifacts in the EEG signal. Finally, the experimental data processing personnel browse all the EEG data with the naked eye, and manually remove the fluctuations of the EEG amplitude exceeding 100μV as artifacts for the experimental trials.

[0090] Each ERP segment was extracted from the EEG data recorded for each stimulus from 200ms before the stimulus onset to 1000ms after the stimulus presentation, of which the first 200ms data was used for baseline correction. The ERP segments under the four conditional stimuli were superimposed and averaged to obtain the ERP waveforms under the four conditions, and the amplitudes of the four ERP components were extracted: N1 (80-120ms), P2 (140-220ms), N2 (220-290ms) and P3 (300-390ms), a total of 16 emotion-related EEG features.

[0091] In the embodiment of this example, the attention network (ANT) EEG experiment includes:

[0092] Experimental design:

[0093] like Figure 3As shown in Figure 1, ANT consists of 1 practice block (24 trials) and 4 experimental blocks (4×96 trials). Subjects must complete the practice block before the formal experiment of ANT, and the entire test block has a total of 408 trials. The data of the practice block are excluded from the statistical analysis. Subjects need to respond to the direction of the central target arrow. For example, if the arrow points to the left, the left index finger presses the "F" key, and if the arrow points to the right, the right index finger presses the "J" key. The experimental task is presented through E-prime, and the test cues will be displayed randomly. Each test block is about 7 minutes. At the beginning of each trial, there is a cross fixation point in the center of the screen, which is randomly displayed for 400 to 1600 milliseconds, and then a 100-millisecond "*" cue prompt is displayed. The cue prompts are divided into four types, namely, no cue prompt, double cue prompt, central cue prompt and spatial cue prompt, followed by an 800-millisecond fixation point. Within 1700 milliseconds after the cue disappears, the target stimulus is displayed above or below the fixation point. There are two conditions: "target-interference consistency" (the five arrows are in the same direction) and "target-interference inconsistency" (the middle arrow is in the opposite direction to the other four arrows). The subjects must press the keys as quickly and accurately as possible. When the keys are responded to, the arrows will disappear. If the subjects do not respond to the corresponding stimulus within 1700 milliseconds, the arrows will automatically disappear. The next step is to display the fixation point, and the change in its presentation time depends on the presentation duration of the initial fixation point and the reaction speed of the subjects. The total time for each trial is 4000ms. Only in the spatial cue condition, the cue and the target are completely matched, while in the other three cue conditions, the correspondence between the cue and the target is uncertain. This is because the four cue conditions have different connections with the target stimulus.

[0094] Behavioral characteristics:

[0095] First, the reaction time (RT) of each cue condition was obtained, and incorrect key presses, data with reaction times exceeding 1600 milliseconds, and data with reaction times below 100 milliseconds were excluded from the behavioral data. The attention alerting network efficiency, attention orienting network efficiency, and attention executive control network efficiency were calculated according to the following method, for a total of three attention-related behavioral characteristics.

[0096] Attention alert network efficiency = RT 无线索条件 —RT 双线索条件. Under the no-cue condition, the subjects have the greatest uncertainty about the target occurrence time. According to the test criteria, any cue can be used as an alert signal because attention is more dispersed under both cues, and the alert network score is calculated by subtracting the average reaction time under the no-cue condition from the average reaction time under the dual-cue condition. However, under the dual-cue condition, the subjects received more alert reminders than under the no-cue condition, and their attention was focused on the location where the target stimulus appeared, thereby improving the efficiency of the response. The higher the alert network efficiency value, the more efficient the warning network, that is, the subjects can maintain a higher state of alertness based on the cues, and the efficiency of attention processing is higher.

[0097] Note that the efficiency of the directed network = RT 中心线索条件 —RT 空间线索条件 . Under the central cue condition, participants can stay alert based on the spatial location information presented by the stimulus, thereby improving reaction performance, but the central cue does not provide effective information about the target. In contrast, spatial cues can also be used to generate alertness, which is 100% about the location information of the target. Therefore, the orientation score is obtained by subtracting the average RT of the central cue condition from the average RT of the spatial cue condition. The greater the difference in reaction time between the two cues, the more the subjects benefit from reliable spatial cues and the higher the orientation efficiency.

[0098] Attention executive control network efficiency = RT 不一致条件 —RT 一致条件 The inconsistent instruction condition requires higher conflict resolution ability than the consistent instruction condition, so the reaction time of the former is longer than that of the latter. The smaller the difference between the two conditions, the higher the conflict resolution ability and the higher the efficiency of the executive control network.

[0099] In summary, a total of 3 attention-related behavioral features were extracted.

[0100] EEG characteristics:

[0101] The EEG signal preprocessing process was performed in the EEGLAB plug-in (http: / / sccn.ucsd.edu / eeglab / ) in MATLAB 2017a (MathWorks, Natick, MA, USA) and self-written scripts for processing and analysis.

[0102] A series of offline preprocessing work is performed on the collected EEG raw signals. First, all the lead data are re-referenced and the reference electrode is reset to the whole brain average. The EEG data is then band-pass filtered at 0.1Hz-30Hz using a Basic FIR filter. The ICA (independent component analysis) method is used to identify and remove eye movement artifacts in the EEG signal. Finally, the experimental data processing personnel browse all the EEG data with the naked eye, and manually remove the fluctuations in the EEG amplitude exceeding 100μV as artifacts for the experimental trials.

[0103] The “cue” stimulus selected 400ms before stimulus as the baseline, and extracted the segment from 400ms before stimulus to 1000ms after stimulus presentation; the “target” stimulus selected 1300ms before stimulus to 900ms before stimulus as the baseline (consistent with the baseline segment of the cue stimulus), and extracted the segment from 1300ms before stimulus to 1700ms after stimulus presentation. Subsequently, the four “cue” stimuli and two “target” stimuli of each subject were superimposed and averaged to obtain waveforms at the group level under different conditions. The ERP components in the following time periods were extracted: N1 component after “cue” stimulation (150-230ms), CNV after “cue” stimulation (O wave, 600-800ms; E wave, when “target” stimulation appears, i.e. 800-1000ms after “cue” stimulation), N200 component after “cue” stimulation (250-400ms); N1 component after “target” stimulation (150-230ms), N200 component after “target” stimulation (300-450ms), P300 component after “target” stimulation (250-450ms). In total, the amplitudes of 22 ERP components were obtained.

[0104] In the embodiment of this example, feature dimensionality reduction includes:

[0105] Through the above methods, a series of features related to emotion and attention were obtained, including the features of the subjective evaluation of the subjects and the objective features obtained from the electrophysiological signals, as shown in the following table. Before classification, the above features were first reduced in dimension by factor analysis to avoid overfitting in each training, keep the factors with feature values ​​above 1, and achieve feature dimensionality reduction. On the one hand, all features of emotion and attention were reduced in dimension, and at the same time, the emotional features and attention features were reduced in dimension respectively, and the features after dimensionality reduction were put into the subsequent classification work.

[0106]

[0107] In the embodiment of this example, the training sample collection includes:

[0108] Thirty healthy subjects without mental illness were recruited to participate in emotion and attention experiments.

[0109] In the emotion induction experiment, 15 carefully selected Chinese movie clips from the emotional video database established by the BCMI laboratory of Shanghai Jiao Tong University were used to induce specific emotions (positive, neutral and negative) in the subjects. All 30 subjects conducted 15 trials. Before the start of each trial, there was a 5-second prompt, followed by watching a movie clip lasting about 4 minutes. After watching the clip, they immediately reported their emotional response to the movie clip they had just watched, and immediately performed emotional self-assessment, emotional scale measurement and Stroop EEG experiment. Then rest for 15 seconds before watching the next movie clip. The order of emotional presentation in the experiment was 1, 0, -1, -1, 0, 1, -1, 0, 1, 0, -1, 0, -1, where 1 represents positive emotion, 0 represents neutral emotion, and -1 represents negative emotion.

[0110] In the attention experiment, the subjects first performed the ANT EEG data acquisition experiment mentioned above at the beginning. After the subjects completed the one-hour simulated driving task, the subjects' fatigue self-assessment results were collected, and the subjects' ANT EEG data were collected again.

[0111] In the embodiment of this example, building a classification model includes:

[0112] In the present disclosure, a support vector machine (SVM) classifier is used to learn cognitive features to classify different cognitive states in a closed environment. SVM is a classic machine learning algorithm that is widely used in classification and regression problems. It achieves classification or regression tasks by finding a hyperplane that can maximize the distance between samples. The SVM algorithm performs well in many practical applications, such as image recognition, text classification, bioinformatics, etc. However, with the continuous increase in the amount of data and the increase in the complexity of the problem, the traditional SVM algorithm faces problems such as high computational complexity and poor model generalization ability. In order to overcome the problems of the traditional SVM algorithm, a multi-kernel learning method is used in this patent. Multi-kernel learning (MKL) can better capture the potential structure of the data by learning a combination of multiple kernel functions, thereby improving the learning performance of SVM. The multi-kernel learning method can better capture the nonlinear relationship of the data and improve the performance of the classifier by introducing multiple kernel functions. Different kernel functions are suitable for different types of data, for example, the linear kernel is suitable for linearly separable data, the polynomial kernel is suitable for polynomially separable data, and the Gaussian kernel is suitable for nonlinearly separable data. By linearly combining multiple kernel functions, a more powerful kernel function can be obtained to improve the classification performance of SVM. The combination of synthetic kernels is shown in the following formula, where η represents the kernel function weight coefficient and p represents the number of kernel functions.

[0113]

[0114] Optimization of Multi-core Learning Support Vector Machines:

[0115]

[0116]

[0117] ξ i ≥0, i=1, 2, ..., n

[0118] Among them, C is a regularization parameter used to balance the model complexity and error term, ξ i is a slack variable used to deal with noise and outliers in the data, w represents the direction of the classification hyperplane in the high-dimensional feature space, and is the mapping function. The classification decision function is:

[0119]

[0120] The multi-kernel learning support vector machine is used to learn the emotion-related features and attention-related features after dimensionality reduction, and the emotional state and attention level classification model is established to achieve the evaluation of emotional state and attention level. Specifically, the multi-kernel learning support vector machine is constructed by pre-defining 10 polynomial kernel functions (poly) and 10 radial basis kernel functions (rbf) with different parameters to construct the kernel vector, and the weight vector corresponding to the kernel vector is learned on the training set by the EasyMKL[1] method. Both of these kernel functions are commonly used and relatively effective kernel function types when using SVM for EEG classification tasks. Since the parameters of the kernel function actually have continuous and infinite values ​​to choose from, only some commonly used empirical discrete values ​​are selected within the commonly used parameter range for research. The parameter of the polynomial kernel function is the degree of the highest order term. In this chapter, we set it to (degree = 1, 2, 3, 4, 5, 6, 7, 8, 9, 10) for a total of 10 types, and set the parameter gamma of the radial basis kernel to (gamma = 0.00001, 0.0001, 0.001, 0.01, 0.1, 1, 10, 100, 1000, 10000) for a total of 10 types. The kernel function vector and its weight vector obtained by the EasyMKL algorithm can be used to calculate a combined kernel vector, and then combined with the SVM classifier and the test set to complete the classification experiment.

[0121] In the embodiment of this example, the present disclosure provides a method for evaluating the cognitive state of a person's brain in a closed environment, which uses a combination of subjective and objective characteristics to evaluate the level of the brain's cognitive state, providing a methodological basis for detecting the brain's cognitive state under long-term closed working conditions.

[0122] It should be noted that, although the steps of the method in the present disclosure are described in a specific order in the drawings, this does not require or imply that the steps must be performed in this specific order, or that all the steps shown must be performed to achieve the desired results. Additionally or alternatively, some steps may be omitted, multiple steps may be combined into one step, and / or one step may be decomposed into multiple steps, etc.

[0123] In addition, in this exemplary embodiment, a cognitive state assessment device in a closed environment is also provided. Figure 4 As shown, the cognitive state evaluation device 400 in a closed environment may include: a feature acquisition module 410, a dimensionality reduction processing module 420 and a cognitive state evaluation module 430. Among them:

[0124] A feature collection module 410 is used to collect emotion recognition features and attention recognition features based on a preset emotion recognition model and a preset attention recognition model respectively;

[0125] A dimension reduction processing module 420 is used to perform dimension reduction processing on the emotion recognition features and attention recognition features;

[0126] The cognitive state evaluation module 430 is used to evaluate the cognitive state based on the support vector machine (SVM) classifier learning cognitive feature model using the emotion cognitive features and attention cognitive features that have undergone dimensionality reduction processing as inputs, and generate an evaluation result.

[0127] The specific details of each module of the cognitive state assessment device in a closed environment have been described in detail in the corresponding cognitive state assessment method in a closed environment, so they will not be repeated here.

[0128] It should be noted that, although several modules or units of the cognitive state assessment device 400 in a closed environment are mentioned in the above detailed description, this division is not mandatory. In fact, according to the embodiments of the present disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided into multiple modules or units for embodiment.

[0129] In addition, in an exemplary embodiment of the present disclosure, an electronic device capable of implementing the above method is also provided.

[0130] It will be appreciated by those skilled in the art that various aspects of the present invention may be implemented as systems, methods or program products. Therefore, various aspects of the present invention may be specifically implemented in the following forms, namely: complete hardware embodiments, complete software embodiments (including firmware, microcode, etc.), or embodiments combining hardware and software aspects, which may be collectively referred to herein as "circuits", "modules" or "systems".

[0131] Refer to the following Figure 5 An electronic device 500 according to such an embodiment of the present invention is described. Figure 5 The electronic device 500 shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0132] like Figure 5 As shown, the electronic device 500 is in the form of a general computing device. The components of the electronic device 500 may include but are not limited to: at least one processing unit 510, at least one storage unit 520, a bus 530 connecting different system components (including the storage unit 520 and the processing unit 510), and a display unit 540.

[0133] The storage unit stores program codes, which can be executed by the processing unit 510, so that the processing unit 510 performs the steps according to various exemplary embodiments of the present invention described in the above “Exemplary Method” section of this specification. For example, the processing unit 510 can perform the following steps: Figure 1 Steps S110 to S130 shown in FIG.

[0134] The storage unit 520 may include a readable medium in the form of a volatile storage unit, such as a random access storage unit (RAM) 5201 and / or a cache storage unit 5202 , and may further include a read-only storage unit (ROM) 5203 .

[0135] The storage unit 520 may also include a program / utility 5204 having a set (at least one) of program modules 5205, such program modules 5205 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment.

[0136] Bus 530 may represent one or more of several types of bus structures, including a memory unit bus or memory unit controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus architectures.

[0137] The electronic device 500 may also communicate with one or more external devices 570 (e.g., keyboards, pointing devices, Bluetooth devices, etc.), may also communicate with one or more devices that enable a user to interact with the electronic device 500, and / or may communicate with any device that enables the electronic device 500 to communicate with one or more other computing devices (e.g., routers, modems, etc.). Such communication may be performed via an input / output (I / O) interface 550. Furthermore, the electronic device 500 may also communicate with one or more networks (e.g., local area networks (LANs), wide area networks (WANs), and / or public networks, such as the Internet) via a network adapter 560. As shown, the network adapter 560 communicates with other modules of the electronic device 500 via a bus 530. It should be understood that, although not shown in the figure, other hardware and / or software modules may be used in conjunction with the electronic device 500, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, etc.

[0138] Through the description of the above embodiments, it is easy for those skilled in the art to understand that the example embodiments described here can be implemented by software, or by software combined with necessary hardware. Therefore, the technical solution according to the embodiment of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, and includes several instructions to enable a computing device (which can be a personal computer, a server, a terminal device, or a network device, etc.) to execute the method according to the embodiment of the present disclosure.

[0139] In an exemplary embodiment of the present disclosure, a computer-readable storage medium is also provided, on which a program product capable of implementing the above method of the present specification is stored. In some possible embodiments, various aspects of the present invention can also be implemented in the form of a program product, which includes a program code, and when the program product is run on a terminal device, the program code is used to enable the terminal device to perform the steps according to various exemplary embodiments of the present invention described in the above "Exemplary Method" section of the present specification.

[0140] refer to Figure 6 As shown, a program product 600 for implementing the above method according to an embodiment of the present invention is described, which can adopt a portable compact disk read-only memory (CD-ROM) and include program code, and can be run on a terminal device, such as a personal computer. However, the program product of the present invention is not limited thereto. In this document, a readable storage medium can be any tangible medium containing or storing a program, which can be used by or in combination with an instruction execution system, an apparatus or a device.

[0141] The program product may use any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable 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 above.

[0142] Computer readable signal media may include data signals propagated in baseband or as part of a carrier wave, in which readable program code is carried. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Readable signal media may also be any readable medium other than a readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0143] The program code embodied on the readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wired, optical cable, RF, etc., or any suitable combination of the foregoing.

[0144] Program code for performing the operations of the present invention may be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java, C++, etc., and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as a separate software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving a remote computing device, the remote computing device may be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computing device (e.g., through the Internet using an Internet service provider).

[0145] In addition, the above-mentioned figures are only schematic illustrations of the processes included in the method according to an exemplary embodiment of the present invention, and are not intended to be limiting. It is easy to understand that the processes shown in the above-mentioned figures do not indicate or limit the time sequence of these processes. In addition, it is also easy to understand that these processes can be performed synchronously or asynchronously, for example, in multiple modules.

[0146] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary technical means in the art that are not disclosed in the present disclosure. The specification and examples are to be considered as exemplary only, and the true scope and spirit of the present disclosure are indicated by the claims.

[0147] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. A method for evaluating cognitive status in a closed environment, characterized in that: The method comprises: Collecting emotion cognition features and attention cognition features based on a preset emotion cognition model and a preset attention cognition model respectively; Performing dimensionality reduction processing on the emotion cognitive features and attention cognitive features; The emotion cognitive features and attention cognitive features that have undergone dimensionality reduction processing are used as input, and the cognitive state is evaluated based on a support vector machine (SVM) classifier learning cognitive feature model to generate an evaluation result.

2. The method according to claim 1, characterized in that The method of collecting emotion recognition features based on a preset emotion recognition model also includes: Based on the Self-Rating Anxiety Scale (SAS), the Self-Rating Depression Scale (SDS), the Self-Rating Psychological Stress Questionnaire (PSET), and the Self-Rating Symptom Checklist (SCL-90), we obtained the scores of SAS, SDS, PSET, and the scores of somatization, obsession, interpersonal sensitivity, depression, anxiety, hostility, terror, paranoia, and psychosis in SCL-90. Based on the emotional stroop EEG experiment, behavioral characteristics and EEG feature ERP were extracted to obtain a preset number of emotional behavioral characteristics and the amplitude of ERP components.

3. The method according to claim 1, characterized in that The method of collecting attention cognition features based on a preset attention cognition model also includes: An attention network ANT EEG experiment was designed, and the behavioral characteristics of attention orientation network efficiency, attention alertness network efficiency, and attention executive control network efficiency were collected based on the attention network ANT experiment.

4. The method according to claim 1, characterized in that The method further comprises: The emotion cognitive features and attention cognitive features are subjected to dimensionality reduction processing based on factor analysis.

5. The method according to claim 1, characterized in that The method further comprises: Recruit a preset number of experimenters to participate in the emotion experiment and attention experiment; In the emotion induction experiment, movie clips from the preset emotional video database were used to induce specific positive, neutral and negative emotions in the subjects to complete the collection of training samples; In the attention experiment, based on the attention network ANT EEG experiment, after one hour of simulated driving task, the subjects' fatigue self-assessment results and ANT EEG data were collected to complete the training sample collection.

6. The method according to claim 1, characterized in that The method further comprises: Using a multi-kernel learning support vector machine to learn the reduced-dimensional emotion cognitive features and attention cognitive features, respectively, to establish an emotion state and attention level classification model, and to achieve an evaluation of the emotion state and attention level; Based on the kernel function vector and its weight vector obtained by the EasyMKL algorithm, the combined kernel vector is calculated; Combine the SVM classifier and the test set to complete the classification experiment.

7. A cognitive status assessment device in a closed environment, characterized in that: The device comprises: A feature collection module, used to collect emotion recognition features and attention recognition features based on a preset emotion recognition model and a preset attention recognition model respectively; A dimensionality reduction processing module, used for performing dimensionality reduction processing on the emotion cognitive features and attention cognitive features; The cognitive state evaluation module is used to evaluate the cognitive state based on the support vector machine (SVM) classifier learning cognitive feature model using the emotion cognitive features and attention cognitive features that have undergone dimensionality reduction processing as input, and generate an evaluation result.

8. An electronic device, characterized in that: include Processor; and A memory having computer-readable instructions stored thereon, wherein the computer-readable instructions, when executed by the processor, implement the method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.

Images