Impact of Sympathetic Activation in Imaging Photoplethysmography

Alexander Woyczyk, Stefan Rasche, Sebastian ZaunsederWorkshops @ ICCV 2019 · DOI

Introduction

Imaging photoplethysmography (iPPG) has emerged as a promising non-contact method for monitoring physiological parameters like heart rate (HR) and heart rate variability (HRV). By using standard cameras to detect subtle color changes in the skin caused by blood volume variations, iPPG offers a convenient alternative to traditional contact-based PPG sensors. However, as researchers attempt to extract more complex morphological data—such as stress levels or blood pressure—from these signals, a critical question arises: does iPPG behave identically to contact PPG under physiological stress?

This study investigates how sympathetic nervous system activation, triggered by pain and cold, affects iPPG signals from the face compared to traditional finger-based PPG. The findings challenge the assumption that iPPG is simply a remote version of contact PPG.

Experimental Methodology

To induce sympathetic activation, researchers employed the Cold Pressor Test (CPT). Participants (n=22) immersed one hand in 3°C water for up to three minutes, a standard protocol for eliciting a "fight or flight" response.

Data collection involved several synchronized sensors:

  • Imaging Sensors: An RGB camera (utilizing the green channel for optimal signal quality) and a Near-Infrared (NIR) camera captured facial video.
  • Region of Interest (ROI): A specific area on the forehead was manually defined to extract the iPPG signal, as shown in the experimental setup.
  • Reference Monitoring: Simultaneous recordings were taken using a traditional finger PPG sensor, ECG, and a continuous blood pressure monitor (Finometer).
Forehead Region of Interest (ROI) used for iPPG signal extraction.

The researchers used a "beat template" approach, averaging multiple pulse waves to create a clean ensemble wave for measuring the AC component (pulse amplitude) accurately despite the noise inherent in camera-based monitoring.

Physiological Divergence: Face vs. Finger

The study's most significant finding is the stark divergence between facial iPPG and finger PPG signals during sympathetic activation.

In traditional finger PPG, the pulse amplitude shows a persistent and significant decrease. This is a well-known effect of peripheral vasoconstriction, where the body narrows blood vessels in the extremities to redirect blood to vital organs during stress.

In contrast, the facial iPPG signal exhibited a brief initial dip followed by a significant increase in amplitude above the baseline. This suggests that the face does not share the same high level of sympathetic responsiveness as the fingers. Instead, the facial signal is dominated by systemic factors: the increase in cardiac output and pulse pressure caused by the stress response. Because facial vessels do not constrict as aggressively, they reflect the increased pressure of the blood being pumped by the heart.

Key Results and Insights

  • Finger PPG: Demonstrated a significant and persistent decrease in the alternating component (AC) during the stimulus (p < 0.001).
  • Facial iPPG: Showed a significant eventual increase in AC amplitude (p < 0.01) after the initial stimulus.
  • Systemic Correlation: The rise in facial iPPG amplitude correlated with a significant increase in systemic pulse pressure, confirming that the facial signal tracks central cardiovascular changes rather than local vasoconstriction.
  • Signal Depth: The researchers noted that iPPG (especially green light) captures microcirculation in shallower skin layers compared to the deeper macrocirculation monitored by infrared contact PPG.

Conclusion and Future Directions

The results prove that iPPG is not merely a "contactless version" of traditional PPG; the two methods capture different physiological responses due to regional vascular differences. This has major implications for using iPPG to monitor stress or blood pressure, as models based on finger PPG data may not be directly applicable to facial camera data.

Future research will focus on pulse wave decomposition to better understand signal morphology and the development of more robust ROI tracking to mitigate movement artifacts during stress-inducing scenarios, which often involve involuntary physical reactions to pain or discomfort.

Summary updated

BibTeX
@inproceedings{woyczyk2019impact,
  author = {Alexander Woyczyk and Stefan Rasche and Sebastian Zaunseder},
  title = {{Impact of Sympathetic Activation in Imaging Photoplethysmography}},
  booktitle = {Workshops @ ICCV},
  year = {2019},
  doi = {10.1109/iccvw.2019.00210},
}
Figure