Female-male Differences Should be Considered in Physical Pain Quantification based on Electrodermal Activity:...

Area of Science:

• Neuroscience
• Biomedical Engineering
• Psychophysiology

Background:

• Objective pain quantification remains a significant challenge.
• Electrodermal activity (EDA) shows promise for pain assessment due to its sensitivity.
• Cognitive stress can interfere with accurate pain estimation using EDA.

Purpose of the Study:

• To investigate the influence of cognitive stress on EDA-based pain quantification.
• To explore sex-specific differences in EDA responses to pain and stress.
• To evaluate the efficacy of machine learning models for sex-specific pain estimation using EDA.

Main Methods:

• Collected EDA signals from 10 subjects (5 male, 5 female) under varying pain (low/high) and cognitive stress (Stroop test) conditions.
• Extracted phasic, tonic, and frequency-domain EDA features.
• Utilized leave-one-subject-out cross-validation for machine learning model evaluation.

Main Results:

• Significant differences in EDA features were observed between painless and painful states, and between stress and no-stress conditions.
• Frequency-domain EDA features increased with stress for both sexes, with higher standard deviation in females.
• Machine learning models achieved higher accuracies when trained on same-sex data (63.5% combined, 72.4% male, 53.2% female) compared to different-sex data (47.6% combined, 57.4% male, 42.7% female).

Conclusions:

• EDA signals are sensitive to both pain and cognitive stress.
• Sex is a critical factor influencing EDA-based pain quantification, particularly under cognitive stress.
• Incorporating patient sex into EDA analysis can enhance the accuracy of objective pain measurement.