Real-Time Robust Heart Rate Estimation From Wrist-Type PPG Signals Using Multiple Reference Adaptive Noise...

Area of Science:

• Biomedical Engineering
• Wearable Technology
• Signal Processing

Background:

• Photoplethysmographic (PPG) signals from wrist-based sensors are crucial for wearable devices.
• Motion artifacts (MA) significantly degrade PPG signal quality, complicating accurate heart rate (HR) estimation.
• Robust HR estimation algorithms are vital for the commercial success of wearable health monitors.

Purpose of the Study:

• To develop a robust algorithm for accurate HR estimation from wrist-type PPG signals.
• To address the challenge of severe motion artifacts (MA) in PPG-based HR monitoring.
• To enhance the user experience and reliability of wearable devices.

Main Methods:

• Proposed a novel Multiple Reference Adaptive Noise Cancellation (MURAD) technique.
• Utilized four reference noise signals (RNS): three-axis accelerometer data and the difference between two PPG signals.
• Employed peak verification techniques and a time-window approach for HR estimation.

Main Results:

• The MURAD technique demonstrated a lower average absolute error compared to existing state-of-the-art methods.
• Achieved more accurate HR estimations even under severe motion artifact conditions.
• Successfully reduced the impact of complex motion artifacts on PPG signals.

Conclusions:

• MURAD offers a promising solution for reliable HR monitoring using PPG in wearable devices.
• The method effectively mitigates motion artifacts, improving the accuracy of HR estimation.
• Enhances the feasibility and user experience of wrist-based PPG monitoring devices.