EarSet: A Multi-Modal Dataset for Studying the Impact of Head and Facial Movements on In-Ear PPG Signals

Area of Science:

• Biomedical Engineering
• Wearable Technology
• Signal Processing

Background:

• Photoplethysmography (PPG) measures blood volume changes but is susceptible to motion artifacts.
• Motion artifacts are a significant challenge for ear-based PPG sensors due to head and facial movements.
• Existing in-ear PPG datasets are insufficient for robust artifact mitigation research.

Purpose of the Study:

• To develop and validate a novel ear-worn sensing device for multi-channel PPG and motion data.
• To create a comprehensive dataset of in-ear PPG and motion signatures during natural movements.
• To advance research on accurate and robust in-ear vital signs monitoring using PPG.

Main Methods:

• Designed an ear tip integrating a 3-channel PPG sensor and a 6-axis motion sensor.
• Collected multi-modal data from 30 participants performing 16 distinct natural motions.
• Enabled simultaneous sensing of PPG signals at multiple wavelengths and corresponding motion data from both ears.

Main Results:

• Successfully collected a unique, multi-modal in-ear dataset.
• The dataset captures PPG variations and motion signatures across various natural movements.
• The device facilitates synchronized, multi-wavelength PPG and motion sensing.

Conclusions:

• The developed ear tip and dataset are crucial for addressing motion artifacts in in-ear PPG.
• This resource will accelerate the development of reliable cardiovascular health monitoring with earables.
• Unlocks the potential of next-generation PPG-equipped earable devices for accurate vital signs sensing.