Using Electrooculography and Electrodermal Activity During a Cold Pressor Test to Identify Physiological Biomarkers...

Area of Science:

• Physiological monitoring
• Wearable technology
• Mental health assessment

Background:

• State anxiety (s-anxiety) impacts mental and physical health, with links to cardiovascular issues.
• Traditional monitoring methods lack real-time contextual sensitivity for anxiety.
• Electrooculography (EOG) and electrodermal activity (EDA) are promising biosignals for wearable-based anxiety detection.

Purpose of the Study:

• Identify novel biomarkers for s-anxiety using EOG and EDA in real-world settings.
• Evaluate noninvasive wearable technology for real-time physiological stress monitoring.
• Distinguish anxiety signals from artifacts in noisy environments.

Main Methods:

• Utilized two datasets: BLINKEO for EOG blink identification and EMOCOLD for EOG/EDA during a cold pressor test (CPT).
• Analyzed blink rate variability, skin conductance, and arousal metrics.
• Employed Shapley additive explanations (SHAP) for model interpretation and biomarker refinement.

Main Results:

• BLINKEO achieved 98.17% accuracy in distinguishing blinks from noise.
• EMOCOLD data showed elevated anxiety during CPT, normalizing post-test.
• SHAP analysis identified specific EDA (Hjorth activity, spectral entropy) and EOG (opening phase energy, signal height) features as key predictors of s-anxiety.

Conclusions:

• Combined EOG and EDA analysis significantly improves real-time anxiety marker detection.
• Wearable technology holds potential for personalized health monitoring and mental health interventions.
• Developed context-sensitive models for anxiety assessment using wearable data.