Video-based sympathetic arousal assessment via peripheral blood flow estimation

Area of Science:

• Physiology
• Biomedical Engineering
• Computer Vision

Background:

• Electrodermal activity (EDA) is a key indicator of sympathetic nervous system activity.
• Conventional EDA measurement necessitates direct skin contact via electrodes, limiting its application.
• Non-invasive methods for assessing sympathetic arousal are highly desirable.

Purpose of the Study:

• To investigate the feasibility of inferring sympathetic arousal using only optical sensors (RGB cameras).
• To develop and validate a novel approach measuring peripheral blood flow optically.
• To compare optical measurements with gold-standard EDA and photoplethysmography (PPG) signals.

Main Methods:

• Development of a method to optically measure peripheral blood flow from facial and palm videos.
• Collection of a synchronized dataset including facial/palm videos, EDA, and PPG signals from 21 participants.
• Correlation analysis between optically inferred signals and gold-standard EDA.

Main Results:

• Optically inferred peripheral blood flow signals closely correlate with ground truth EDA.
• Median correlations ranging from 0.57 to 0.63 were achieved using palm or forehead videos and PPG signals.
• Sympathetic arousal was most effectively inferred from optical signals of the forehead, finger, or palm.

Conclusions:

• Optical measurement of peripheral blood flow offers a viable non-invasive alternative to traditional EDA for assessing sympathetic arousal.
• RGB cameras can be utilized to monitor physiological stress responses remotely.
• Forehead, finger, and palm regions are optimal for optical-based sympathetic arousal monitoring.