Detecting Physiological Responses Using Multimodal Earbud Sensors.

Area of Science:

• Biomedical Engineering
• Physiological Monitoring
• Wearable Technology

Background:

• Continuous stress negatively impacts well-being, affecting physiological responses like heart rate and breathing.
• Traditional stress detection methods use uncomfortable sensors (e.g., chestbands) to collect signals like electrocardiogram (ECG).
• There is a need for non-invasive, comfortable methods for continuous physiological monitoring.

Purpose of the Study:

• To investigate the feasibility of using earbuds for passive physiological monitoring.
• To assess the accuracy of earbud-based photoplethysmography (PPG) for measuring heart rate and heart rate variability.
• To evaluate the effectiveness of earbud-derived signals in classifying stress-induced physiological responses.

Main Methods:

• A lab study involving 18 participants exposed to stress-inducing and relaxation tasks.
• Simultaneous collection of earbud photoplethysmography (PPG), core body temperature, and inertial measurements.
• Comparison with laboratory-grade reference measurements including ECG, impedance cardiography (ICG), and blood pressure.

Main Results:

• Earbud PPG sensors reliably captured heart rate and heart rate variability.
• Earbud signals classified stress responses with high accuracy (91.30% recall, 80.52% precision, 85.12% F1-score) using a random forest classifier.
• Multi-modal sensing using earbuds further improved classification accuracy.

Conclusions:

• Earbuds can be effectively used for passively monitoring physiological responses to stress.
• This technology offers a comfortable and convenient alternative to traditional, obtrusive monitoring systems.
• The findings support the development of earbuds as a tool for unobtrusive, continuous health monitoring.