Electrodermal Temperature-Adjusted Electrodermal Activity (EDA) for Stress Detection in Virtual Reality

Area of Science:

• Physiological computing
• Affective computing
• Human-computer interaction

Background:

• Electrodermal activity (EDA) is a key indicator of emotional arousal and stress.
• Thermoregulatory mechanisms can independently alter EDA, confounding stress detection in real-world settings, especially virtual reality (VR).
• Existing methods struggle to differentiate between stress-induced EDA and thermally influenced EDA.

Purpose of the Study:

• To develop and validate a temperature-corrected framework for precise stress identification using EDA in VR.
• To distinguish authentic stress responses from heat-associated physiological changes.
• To improve the accuracy and ecological validity of stress detection in VR environments.

Main Methods:

• Utilized the Wearable Emotion Sensing and Detection (WESAD) dataset.
• Synchronized peripheral temperature variations with EDA (electrodermal conductance) patterns.
• Implemented a temperature-corrected framework combining a proportionality model and a data-driven adaptive scaling approach.
• Analyzed feature importance to assess the contribution of temperature-derived parameters.

Main Results:

• The temperature-corrected framework significantly improved the differentiation between stress-related and thermally influenced EDA.
• Statistical analysis confirmed strong distinctions between affective states based on both conductance and peripheral temperature.
• The adaptive scaling model demonstrated superior performance in generating condition-specific patterns compared to the proportionality model.
• Temperature-derived features were identified as crucial for classification consistency.

Conclusions:

• Temperature compensation is an essential preprocessing step for reliable stress identification in VR settings.
• The proposed framework enhances the physiological precision and ecological authenticity of EDA-based stress detection.
• Accurate interpretation of EDA across diverse thermal conditions is achievable with temperature correction.