CN-121971089-A - Intelligent psychological abnormal behavior screening method for college student group

CN121971089ACN 121971089 ACN121971089 ACN 121971089ACN-121971089-A

Abstract

The invention discloses a psychological abnormal behavior intelligent screening method for college students, which relates to the technical field of psychological screening and artificial intelligence, and specifically comprises the following steps of inducing a subject response through a standardized cognitive task; the method comprises the steps of synchronously collecting and processing eye movement and electroencephalogram signals, extracting multi-mode physiological characteristics and carrying out fusion calculation to generate a time sequence cognition coupling characteristic value, training a time sequence abnormality judging model based on normal group data, finally inputting the characteristics of a person to be tested into the model, calculating judging coefficients and outputting grading screening results, thereby realizing noninvasive, objective and efficient psychological state assessment and early warning.

Inventors

JIANG WEIWEI
LIU GUILIN
FANG CAN
SHEN YING

Assignees

湖南工程职业技术学院

Dates

Publication Date: 20260505
Application Date: 20260312

Claims (9)

1. The intelligent psychological abnormal behavior screening method for college students is characterized by comprising the following specific steps of: S100, inducing cognitive tasks, namely sequentially presenting emotion stimulus browsing, attention test and memory extraction three types of cognitive task sequences to a college student subject by adopting a campus intelligent tablet computer terminal, and forming cognitive responses by means of canonical task stimulus; S200, synchronously acquiring signals, namely synchronously acquiring eye movement track signals and electroencephalogram physiological signals of a subject according to a unified timestamp standard in the whole process of executing three types of cognitive tasks, and carrying out artifact removal and smoothing treatment on the acquired signals to enable the time sequences of the two types of signals to be aligned; s300, performing multi-mode coupling processing, namely respectively extracting features of the eye movement track signals and the electroencephalogram physiological signals which are aligned in time sequence to obtain eye movement micro-feature integrated values and electroencephalogram rhythm feature integrated values, determining eye movement and electroencephalogram feature weight coefficients and time sequence attenuation factors, and calculating cognitive coupling feature values at each sampling moment by adopting a multi-mode cognitive coupling algorithm; S400, training a discrimination model, namely building a psychological abnormal dynamic discrimination model based on a time sequence abnormal discrimination algorithm, taking a cognitive coupling characteristic value as a model input, and determining the total number of time sequence sampling points, a cognitive coupling characteristic mean value and an extremum of a normal college student group and a physiological adaptation correction coefficient; S500, outputting screening results, namely repeating the steps of cognitive task induction, signal synchronous acquisition and multi-mode coupling treatment on a college student subject to be tested to obtain a full-time cognitive coupling characteristic value, inputting the characteristic value into a trained psychological abnormality dynamic discrimination model, calculating by a time sequence abnormality discrimination algorithm to obtain a psychological abnormality discrimination coefficient, and outputting a corresponding psychological abnormality screening classification result according to a preset threshold value.
2. The intelligent screening method for the psychological abnormal behaviors of college students according to claim 1 is characterized in that in the step S100, in the induction of cognitive tasks, three types of cognitive tasks mainly comprise emotion stimulus browsing tasks which are positive, neutral and negative emotion materials, including emotion face pictures and daily scene pictures, the quantity of the emotion pictures is equal, the sizes of the pictures are uniform, the presentation time of each picture is fixed according to random sequence, attention test tasks adopt a digital symbol matching range, 0-9 numbers and corresponding unique geometric symbols are presented, the corresponding description of the digital symbols with fixed time is presented before the tasks, the numbers are presented randomly in the test process, a subject is required to click the corresponding geometric symbols and synchronously record response conditions, a memory extraction task adopts a digital n-back range, n values are fixed to be 2, a random number string with fixed length is presented first, the presentation rate of the number string is fixed, the other string of numbers is presented after the presentation is completed, the subject is required to recognize the same numbers as the first two digits and synchronously record recognition conditions, and click confirmation is completed after the recognition.
3. The intelligent screening method for the psychological abnormal behaviors of the college student group is characterized in that in S200, in the signal synchronous acquisition, eye movement track signals are signals capable of reflecting eye dynamic changes of a subject in the cognitive task execution process, and comprise a gazing point position, a gazing amplitude, a gazing direction, gazing duration, pupil diameters and pupil dilation rate, brain electric physiological signals are signals capable of reflecting brain-related brain area electric activity states of the subject, electric activity data of forehead leaves, temporal leaves, top leaves, pillow leaves, psychological states and cognitive response-related brain areas are mainly captured, eye movement track signals are acquired through a campus intelligent tablet computer terminal, eye movement track signals are acquired in real time and are stored in an electric signal form, in the acquisition process, a dry electrode is attached to the scalp of the subject, in real time, brain area electric activities are captured and are converted into processable electric signals, an eye movement acquisition module and a brain electrical head ring are synchronized with a uniform time stamp of a tablet computer terminal, and in the synchronous acquisition process, pseudo-trace elimination and smooth processing are synchronously carried out on the two types of signals.
4. The intelligent screening method for psychology abnormal behaviors of college students according to claim 1, wherein in the S200 signal synchronous acquisition, artifact removal and smoothing process specifically comprises: Artifact rejection is carried out on two types of signals in a corresponding mode, wherein artifacts of an electroencephalogram signal mainly comprise myoelectric artifacts and ocular artifacts, a preset amplitude threshold value is +/-75 mu V, signal fragments with amplitude exceeding the threshold value in the acquired electroencephalogram signal are judged to be artifacts and are directly rejected, and the signal missing fragments are complemented after rejection; The smoothing processing sets the size of a sliding window to 5 sampling points for the electroencephalogram signals, performs point-by-point sliding average calculation on the electroencephalogram signals with artifacts removed, sets the size of the sliding window to 3 sampling points for the eye movement track signals, performs point-by-point sliding average calculation on the eye movement signals with artifacts removed, and reserves signal data of each time sequence sampling moment in the sliding average calculation process.
5. The intelligent screening method for the psychological abnormal behaviors of the college student population according to claim 1 is characterized in that in the S300 multi-mode coupling processing, eye movement micro-feature comprehensive values are obtained by extracting the characteristics of gaze point density, eye jump amplitude, pupil dilation rate, gaze duration, eye jump latency and pupil diameter variation coefficient through normalization processing and weighting calculation, and brain electrical rhythm feature comprehensive values are obtained by carrying out wavelet decomposition on brain electrical physiological signals after time sequence alignment and extracting alpha wave, theta wave, delta wave and gamma wave phase joint rhythm indexes including alpha wave inhibition rate, theta wave absolute power, theta wave and beta wave power ratio, delta wave relative power and gamma wave event related synchronization characteristics.
6. The intelligent college student group-oriented psychological abnormal behavior screening method according to claim 1, wherein in the S300 multi-modal coupling processing, a multi-modal cognitive coupling algorithm adopts the following mathematical expression: Wherein, the The cognitive coupling characteristic value is the cognitive coupling characteristic value at the t time sequence sampling moment, wherein t is the index of the time sequence sampling moment; The eye movement micro-feature integrated value at the t time sequence sampling moment under the i-th cognitive task; The characteristic comprehensive value of the brain electrical rhythm at the t time sequence sampling moment under the i-th cognitive task; the eye movement characteristic weight coefficient is; is the characteristic weight coefficient of brain electricity; The time sequence attenuation factor at the time of the t time sequence sampling under the i-th class cognitive task.
7. The intelligent screening method for the psychological abnormal behavior of the college student group is characterized in that in the training of the S400 discrimination model, a time sequence formed by cognitive coupling characteristic values at each time sequence sampling moment is used as input for the psychological abnormal dynamic discrimination model, the characteristic sampling quantity in the whole period of a corresponding cognitive task is corresponding to the psychological abnormal dynamic discrimination model, the model adopts a deep neural network structure based on a time sequence attention mechanism and comprises an input layer, a plurality of time sequence coding layers and an abnormal discrimination output layer, the time sequence coding layers are used for capturing dynamic modes of characteristic changes along with time in the cognitive process, the abnormal discrimination output layer outputs abnormal scores at each sampling moment, the training stage adopts the cognitive coupling characteristic time sequence of the normal college student group as a training set, the reconstruction loss is used as an optimization target, firstly, the input characteristic is preprocessed based on the cognitive coupling characteristic average value, the extreme value and the physiological adaptation correction coefficient of the normal group, the group difference and the dimension influence are eliminated, the model parameter is updated through iteration until the reconstruction error converges to a preset threshold value of 0.001, the model calibration and training are completed, and after the training is completed, the model can identify the characteristic deviating from the normal time sequence of the abnormal samples.
8. The intelligent college student group-oriented psychological abnormal behavior screening method according to claim 1, wherein in the step S500 of screening result output, a mathematical expression of a time sequence abnormal discrimination algorithm is as follows: Wherein, the Is a psychological abnormality discrimination coefficient; For the whole period of cognitive tasks is a time-series sampling point total number; The cognitive coupling characteristic value of the college student subject to be tested at the t time sequence sampling moment is obtained, and t is the index of the time sequence sampling moment; The cognitive coupling characteristic average value of the normal college student group at the t time sequence sampling moment is obtained; maximum value of cognitive coupling characteristics for normal college student group; The minimum value of cognitive coupling characteristics is the group of normal college students; The correction coefficient is physiologically adapted to college student groups.
9. The intelligent screening method for psychology abnormal behavior of college students according to claim 1, wherein in the step S500 of screening result output, the decision rule of the psychology abnormal screening classification result is as follows: when the psychology abnormal discrimination coefficient M is less than or equal to 0.3, judging that the psychology state is normal, and outputting a screening result without obvious psychology abnormal tendency; When M is less than or equal to 0.3 and less than or equal to 0.6, judging that the mild psychological abnormal tendency exists, outputting the mild psychological abnormal tendency, and suggesting to carry out daily psychological adjustment and the focused screening result; when M is less than or equal to 0.8 and is less than or equal to 0.6, judging that the moderate psychological abnormal tendency exists, outputting the screening result of the professional psychological consultation and assessment is suggested; When M is more than 0.8, judging that the severe psychological abnormality tendency exists, outputting the severe psychological abnormality tendency, suggesting that the professional psychological intervention and the medical diagnosis and treatment screening result are immediately carried out, The grading result is output in a standardized report form and comprises a subject cognitive coupling characteristic sequence, an abnormality discrimination coefficient and a corresponding grading suggestion, so that a complete psychological abnormality screening process is formed.

Description

Intelligent psychological abnormal behavior screening method for college student group Technical Field The invention relates to the technical field of psychological screening and artificial intelligence, in particular to an intelligent psychological abnormal behavior screening method for college students. Background The psychological state of the current college student group is influenced by multiple factors such as academic stress, social relations, employment anxiety and the like, the occurrence probability of psychological abnormal behaviors is in an ascending trend, the importance and urgency of campus psychological health management work are increasingly highlighted, accurate and efficient psychological abnormal screening becomes a core link of a campus psychological health service system, the external appearance of the psychological abnormal behaviors is closely related to internal cognition and physiological states, the psychological states are researched and judged by capturing physiological signal changes in the cognition process, the important research direction in the field of psychological health screening is formed, the portability of physiological signal acquisition equipment is continuously improved along with the development of intelligent perception technology and artificial intelligent algorithms, the multi-mode signal analysis and time sequence model discrimination technology is gradually matured, technical support is provided for intelligent screening of psychological abnormal behaviors, the popularization of intelligent terminals and portable perception equipment in a campus scene also enables the normative psychological screening of the college student group to have implementation conditions, and the intelligent screening method based on the cognition task induction and the physiological signal analysis becomes a research and application direction which is in agreement with actual demands. The traditional university student psychological abnormality screening method mainly comprises the step of carrying out result judgment based on a scale questionnaire, is easily influenced by factors such as emotion, cognition deviation and the like, is difficult to reflect a real psychological physiological state, and the screening result lacks objective physiological data support, so that the problems of large judgment deviation and insufficient accuracy exist, the screening method based on the physiological signals only adopts a single signal type for analysis, psychological physiological associated characteristics in the cognition process cannot be comprehensively captured, the integrity and the accuracy of characteristic characterization are limited, meanwhile, the signal acquisition and the processing lack of standardized flow, the acquisition results in different scenes are difficult to be uniformly compared, the discrimination model is often constructed based on static characteristics, the dynamic change rule of the cognition process central physiological state cannot be captured, the screening timeliness and pertinence are insufficient, the campus scale and normalized screening requirements are difficult to adapt, and the grading judgment and the accurate intervention of the psychological abnormality tendency cannot be realized. Disclosure of Invention The invention aims to make up the defects of the prior art and provides an intelligent screening method for the psychological abnormal behaviors of college students, which is characterized in that the method induces the response of a subject through three cognitive tasks of emotion, attention and memory, synchronously collects eye movement and brain electrical signals, performs time sequence alignment and denoising treatment, combines the eye movement and brain electrical characteristics by adopting a multi-mode cognitive coupling algorithm, builds a time sequence dynamic discrimination model, calculates the deviation degree of the characteristics of a person to be tested and a normal group, outputs a grading screening result, and realizes objective and efficient recognition and early warning of psychological abnormal tendency. The invention provides a psychological abnormal behavior intelligent screening method for college students, which aims to solve the technical problems and comprises the following specific steps: S100, inducing cognitive tasks, namely sequentially presenting emotion stimulus browsing, attention test and memory extraction three types of cognitive task sequences to a college student subject by adopting a campus intelligent tablet computer terminal, and forming cognitive responses by means of canonical task stimulus; S200, synchronously acquiring signals, namely synchronously acquiring eye movement track signals and electroencephalogram physiological signals of a subject according to a unified timestamp standard in the whole process of executing three kinds of cognitive tasks, and carrying out artifact elimination and smoo