Dynamic Tracking of State Anxiety via Multi-Modal Data and Machine Learning.

Area of Science:

• Psychology
• Neuroscience
• Biomedical Engineering

Background:

• State anxiety research often uses low-temporal-resolution self-report measures.
• Understanding immediate psychological and physiological responses to anxiety is crucial.
• Existing methods limit the dynamic tracking of state anxiety.

Purpose of the Study:

• To develop a high-temporal-resolution tracking model for state anxiety.
• To create a quantitative measure for dynamic anxiety level changes.
• To enable fine-grained analysis of anxiety in real-time.

Main Methods:

• Induced state anxiety using aversive stimuli (pictures, electric shocks).
• Collected multi-modal data: self-reports, electrocardiogram (ECG), galvanic skin response (GSR).
• Trained and validated machine learning models to predict state anxiety from physiological and psychological features.

Main Results:

• Achieved a high correlation between predicted and self-reported state anxiety levels.
• Demonstrated the model's ability to capture dynamic changes in state anxiety.
• Validated the use of multi-modal physiological data for anxiety prediction.

Conclusions:

• A novel, high-resolution quantitative model for state anxiety tracking was developed.
• The model offers sensitive, fine-grained measures of anxiety levels.
• This approach supports future affective brain-computer interaction and anxiety modulation studies.