Hearables: Heart Rate Variability from Ear Electrocardiogram and Ear Photoplethysmogram (Ear-ECG and Ear-PPG)

Area of Science:

• Biomedical Engineering
• Physiological Monitoring
• Wearable Technology

Background:

• Physiological state classification is crucial for health monitoring.
• Heart rate variability (HRV) analysis from electrocardiograms (ECG) is a common method.
• Ear-based sensors offer a convenient alternative for physiological data acquisition.

Purpose of the Study:

• To classify physiological states including normal breathing, controlled slow breathing, and mental exercises.
• To investigate the feasibility of extracting heart rate variability (HRV) features from ear-based electrocardiograms (ear-ECG).
• To evaluate the benefits of integrating ear-ECG with ear-based photoplethysmograms (ear-PPG) for enhanced classification accuracy.

Main Methods:

• Development of an automated R-peak extraction algorithm for low signal-to-noise ratio ear-ECG signals.
• Extraction and analysis of HRV features from ear-ECG recordings.
• Quantitative comparison of different ear-ECG sensor locations.
• Integration of breathing-related features (SpO2, mean breathing interval) from ear-PPG.

Main Results:

• Successful extraction of meaningful HRV features from ear-ECG.
• Demonstrated feasibility of classifying physiological states using ear-ECG derived HRV.
• Identified optimal ear-ECG sites for data acquisition.
• Showcased improved classification performance when combining ear-ECG and ear-PPG features.

Conclusions:

• Ear-ECG is a viable source for extracting HRV features for physiological state classification.
• Combining ear-ECG and ear-PPG data enhances the accuracy of classifying breathing and mental states.
• The proposed R-peak detection scheme effectively handles low signal-to-noise ratio ear-ECG data.