CN-122025024-A - Social anxiety disorder intervention training method based on eye visual direction attention mechanism

CN122025024ACN 122025024 ACN122025024 ACN 122025024ACN-122025024-A

Abstract

The invention relates to the technical field of anxiety disorder intervention training, and discloses a social anxiety disorder intervention training method based on an eye visual direction attention mechanism, which can generate a multi-scene social dialogue environment based on a social dialogue environment label, so that a patient can perform dialogue training with virtual characters independently under a plurality of social environments under the condition of doctor's advice, and perform repeated social dialogue training by oneself, thereby reducing the medical difficulty of a doctor, extracting the sensitive social environment corresponding to the patient, and performing repeated reinforcement through targeted training, so that the method can identify weak links of different patients, perform reinforced training, extract time thresholds belonging to individuals based on real social reactions of different users, generate personal reports belonging to personal people after the targeted training is finished, facilitate the real-time mastering of own information of the patient, and ensure that the mastered information is more comprehensive compared with the traditional doctor-patient dialogue and is not easy to miss.

Inventors

WANG HAILING
BAI YIRU
ZHANG XINYUE
MA SIYU
WANG JUNXING
LIU SHUYA

Assignees

山东师范大学

Dates

Publication Date: 20260512
Application Date: 20260113

Claims (10)

1. A method for intervention training of social anxiety disorder based on eye-ward attention mechanisms, comprising the steps of: Firstly, deploying virtual AI social equipment, and enabling a user to input identity information through the wearable virtual AI social equipment to conduct social environment training; Step two, filtering social environment scenes, generating a plurality of social scenes by the virtual AI social equipment, completing social training by a new user according to requirements, deploying eye tracking equipment in the virtual AI social equipment, and synchronously collecting data; Step three, identifying sensitive social environments, and identifying each sensitive social environment based on the focus points of the eye directions of each virtual social environment; Performing targeted training by the virtual AI social equipment aiming at each identified sensitive social environment; Step five, summarizing training, and generating a personal report for the result of targeted training; And step six, generating a personal training plan, logging in again or continuing training, directly entering targeted training by the virtual AI social equipment, training a sensitive social environment, and realizing closed-loop targeted reinforcement training.
2. The method for intervention training of social anxiety disorder based on eye-ward-looking attention mechanism of claim 1, wherein the virtual AI social equipment comprises a wearable VR head display and a wearable earphone, a host client of the virtual AI social equipment generates a main interface for user identity information input and logs in the virtual AI social equipment, the virtual AI social equipment guides a user to enter a rest state of a section of neutral content, heart rate signals and skin electric signals of the user are collected through a heart rate sensor and a skin electric sensor, and personal physiological baseline values belonging to the user are calculated, and the method specifically comprises the following steps: after the user enters a central content resting state, the virtual AI social equipment guides the user to enter a resting preparation link with a duration of T seconds; The heart rate sensor collects the instantaneous heart rate value of the user at a set sampling frequency, wherein the value unit is times/min; the skin electric sensor collects the instantaneous skin electric conduction value of the user at a set adoption frequency, wherein the unit of the value is micro Siemens; acquiring a heart rate data set consisting of instantaneous heart rate values and a skin electricity data set consisting of skin conductance values respectively after T seconds of acquisition; and secondly, summing a heart rate data set consisting of the acquired instantaneous heart rate values, dividing the heart rate data set by the acquisition times to obtain a heart rate resting baseline value, and summing a skin electricity data set consisting of the acquired skin conductance values, dividing the skin electricity data set by the acquisition times to obtain a skin electricity resting baseline value.
3. The method for training intervention of social anxiety disorder based on eye viewing attention mechanism as set forth in claim 2, wherein the virtual AI social device generates a multi-scene social session environment based on a set social session environment label, a user focuses on completing current social session training on the premise of excluding external interference factors, and completes session training according to prompts provided by a social environment scene, an eye tracking device is integrated in the virtual AI social device, and counts the number of eye viewing removals in each social session environment during each eye viewing removals.
4. The method for training intervention of social anxiety disorder based on eye-ward-looking attention mechanism as set forth in claim 3, wherein an initial time threshold is set in the virtual AI social device for filtering normal line-of-sight movement behavior, normal direction-of-sight movement time and number of times are filtered, the valid direction-of-sight movement time and the valid direction-of-sight movement times are marked only when the direction-of-sight movement time of the eyes of the user exceeds the initial time threshold, and the heart rate value and the skin conductance value of the user are detected within the valid direction-of-sight movement time, and the anxiety weight of each corresponding social dialogue environment is calculated by using the valid direction-of-sight movement time and the valid direction-of-sight movement times, and the method comprises the specific steps of: recording the total dialogue time of each social dialogue environment, the times of the user moving away from the sight direction due to anxiety in each social dialogue environment and the corresponding time; Summing the removal view time to obtain a total removal time; total avoidance time ratio for each social interaction environment; when the heart rate value exceeds the heart rate resting baseline value by a first preset threshold value in the corresponding effective vision direction moving time, marking the heart rate value as a primary heart rate response time; when the skin conductance value exceeds the skin electric resting baseline value by a second preset threshold value in the corresponding effective vision direction removing time, marking one-time skin electric response time; Calculating physiological verification factors in the corresponding social dialogue environment; dividing the total removal times with the total dialogue time to obtain a total removal frequency; Combining the total moving frequency and the total avoidance time ratio to obtain the anxiety weight of each corresponding social dialogue environment; and fourthly, performing descending ranking according to the anxiety weight of each social dialogue environment, identifying each social dialogue environment, and judging whether the dialogue smoothness reaches the standard and whether the limb actions are standard or not.
5. The method for training intervention of social anxiety disorder based on eye-ward-looking attention mechanism as set forth in claim 1, wherein the specific training generates the corresponding number of social dialogue environments according to the anxiety weights of the social dialogue environments according to the total number of the social dialogue environments, and the specific steps are as follows: Step one, total social dialogue environment Weights with anxiety one by one Multiplying to obtain a corresponding result, rounding to form a corresponding social dialogue environment number; step two, adding the number of the social dialogue environments to obtain a sum, and when the sum is equal to the sum If the two phases are equal, preparing a next step; When the sum is smaller than Generating the social dialogue environment with the highest anxiety weight by the total residual quantity; when the sum is greater than Anxiety weights according to individual social conversation environments Descending ranking, one by one singular deduction, when anxiety weight occurs When the social interaction scenes are juxtaposed, deducting is carried out according to the sequence of the social interaction scenes until the sum is equal to the sequence of the social interaction scenes Equal; And thirdly, generating the social dialogue environments with the corresponding quantity according to the processed social dialogue environment corresponding quantity to form a targeted training social dialogue environment.
6. The method for intervention training of social anxiety disorder based on eye-direction attention mechanisms of claim 1, wherein the eye-direction removal time and the number of times of the user's eye-direction removal are collected in real time by the eye-direction tracking device integrated with the virtual AI social device in the targeted training, and anxiety weight calculation is performed again on each social dialogue environment in the targeted training after the targeted training is finished.
7. The method for training intervention of social anxiety disorder based on eye-ward and attention mechanisms of claim 4, wherein anxiety weights are screened out when social dialogue scene filtering is performed on new users in the filtered social environment scene Minimum social dialogue environment, with its corresponding total removal time With corresponding number of visual direction shifts The division, average, defined as individual time threshold, is taken and used in place of the initial time threshold and in subsequent targeted training.
8. The method for intervention training of social anxiety disorder based on eye gaze fixation mechanism as recited in claim 7, wherein the targeted training uses an obtained personal time threshold to re-determine gaze fixation time during user session training.
9. The method for intervention training of social anxiety disorder based on eye-ward and attention mechanism as set forth in claim 8, wherein the training is performed by obtaining the latest social session context anxiety weight after the targeted training is completed, again based on the total number of the next social session contexts And calculating the number of each social dialogue environment for the next targeted training to generate a next personal training program, and after the current training is finished, directly performing the targeted training after the user continues training or logs in again next time.
10. The method for training intervention of social anxiety disorder based on eye gaze fixation mechanism of claim 8, wherein the personal report summarizing the training comprises a user's gaze fixation time, a number of gaze fixation movements, anxiety weights, a personal time threshold, a fluency of conversation, limb movements, and historical data, and wherein the personal report is generated in a tree diagram.

Description

Social anxiety disorder intervention training method based on eye visual direction attention mechanism Technical Field The invention relates to the technical field of anxiety disorder intervention training, in particular to a social anxiety disorder intervention training method based on an eye visual direction attention mechanism. Background Social anxiety disorder (also known as social phobia) is a common mental disorder, and patients can generate significant and unreasonable stress and fear in social interaction and are accompanied by avoidance behavior, and the disorder is frequently generated in people of 17 to 30 years old, and can seriously impair social functions and life quality of patients for a long time. Traditional treatment schemes rely on a large number of high frequency dialogue exercises with the patient, the effective intervention of which is a lengthy process, extremely dependent on repetitive exercises to correct the fear response of the patient, but in reality, limited psychologists are forced to fall into such "high repetition, long period" one-to-one dialogue exercises, which is a huge waste of medical human resources, and also makes a large number of patients unavailable for timely, adequate treatment. In order to break the key of the stiffness, the application aims to develop a social anxiety disorder intervention training method based on an eye vision direction attention mechanism, and under supervision and guidance of doctors, patients can automatically complete a large amount of basic and repeated training, so that the doctor can be greatly liberated from generating force, a rehabilitation support which can be carried out at any time and any place as required can be provided for the patients, and the method is a powerful supplement to the existing medical system. Disclosure of Invention (0) Technical problem to be solved The main object of the present invention is to provide a social anxiety disorder intervention training method based on eye-ward attention mechanisms, so as to solve the problems presented in the background. (II) technical scheme In order to achieve the aim, the invention provides the following technical scheme that the social anxiety disorder intervention training method based on the eye visual direction attention mechanism comprises the following steps: Firstly, deploying virtual AI social equipment, and enabling a user to input identity information through the wearable virtual AI social equipment to conduct social environment training; Step two, filtering social environment scenes, generating a plurality of social scenes by the virtual AI social equipment, completing social training by a new user according to requirements, deploying eye tracking equipment in the virtual AI social equipment, and synchronously collecting data; Step three, identifying sensitive social environments, and identifying each sensitive social environment based on the focus points of the eye directions of each virtual social environment; Performing targeted training by the virtual AI social equipment aiming at each identified sensitive social environment; Step five, summarizing training, and generating a personal report for the result of targeted training; And step six, generating a personal training plan, logging in again or continuing training, directly entering targeted training by the virtual AI social equipment, training a sensitive social environment, and realizing closed-loop targeted reinforcement training. Preferably, the virtual AI social device includes a wearable VR head display and a headset, and the host client of the virtual AI social device generates a main interface for user identity information input and logs into the virtual AI social device. Preferably, the virtual AI social device generates a multi-scenario social dialogue environment based on a social dialogue environment label, a user focuses on the current social dialogue environment in various social environments by excluding external interference factors, and completes dialogue training according to requirements and prompts provided by the virtual AI social device, and an eye movement tracking device is integrated in the virtual AI social device to time the removal of each eye view direction in the social dialogue environment process and record the number of times of the removal of each eye view direction in each social environment. Preferably, an initial time threshold is set in the virtual AI social device, and is used for filtering out normal sight movement behaviors, filtering normal sight removal time and times, marking valid sight removal time and sight removal times only when the sight removal time of eyes of a user exceeds the initial time threshold, and calculating anxiety weights of each corresponding social dialogue environment by using the valid sight removal time and sight removal times. Preferably, the targeted training is based on the total number of the built-in social dialogue environmentsAnxiety