Precision Heart Rate Estimation Using a PPG Sensor Patch Equipped with New Algorithms of Pre-Quality Checking and...

Area of Science:

• Biomedical Engineering
• Signal Processing
• Wearable Technology

Background:

• Accurate heart rate estimation from wearable sensors is crucial for health monitoring.
• Motion artifacts significantly degrade the quality of photoplethysmography (PPG) signals.
• Existing methods struggle with reliable heart rate monitoring during physical activity.

Purpose of the Study:

• To develop a robust method for accurate heart rate estimation using a single PPG signal and accelerations.
• To effectively mitigate motion artifacts introduced by hand movements and walking.
• To improve the reliability of wearable heart rate monitoring devices.

Main Methods:

• A two-sub-algorithm approach: pre-quality checking and motion artifact removal (MAR).
• MAR utilizes Hankel matrix decomposition and Singular Value Decomposition (SVD).
• Data collected using a wearable device with PPG and 3-axis accelerometer sensors.

Main Results:

• Achieved an average error of 0.7345 ± 8.1129 bpm and MAE of 1.86 bpm for walking.
• Demonstrated the second-highest accuracy for single PPG and 3-axis accelerometer methods during walking.
• Attained the best accuracy (3.78 bpm MAE) for hand-moving scenarios among reported studies.

Conclusions:

• The proposed method effectively removes motion artifacts, enabling accurate heart rate estimation.
• This technique enhances the usability of wearable devices for continuous heart rate monitoring.
• The study presents a significant advancement in wearable-based physiological signal processing.