Unobtrusive heart rate estimation during physical exercise using photoplethysmographic and acceleration data

Area of Science:

• Biomedical Engineering
• Signal Processing
• Wearable Technology

Background:

• Photoplethysmography (PPG) offers a non-invasive, cost-effective method for heart rate monitoring during physical activity.
• Motion artifacts in wrist-based PPG signals significantly challenge accurate heart rate estimation during exercise.
• Existing methods often struggle with the noise introduced by dynamic movements.

Purpose of the Study:

• To develop and evaluate a robust methodology for accurate heart rate estimation from wrist-type PPG signals during physical exercises.
• To overcome the limitations imposed by motion artifacts through the integration of acceleration data.
• To enhance the reliability of heart rate monitoring for applications in sports and rehabilitation.

Main Methods:

• A four-stage algorithm was proposed: wavelet-based denoising, acceleration-based denoising, frequency-domain heart rate estimation, and post-processing.
• The methodology incorporates accelerometer data to mitigate motion-induced noise in PPG signals.
• Algorithm design and development involved experiments with various exercises, including running and rehabilitation movements.

Main Results:

• The proposed method achieved a mean absolute error of 1.96 bpm and a standard deviation of 2.86 bpm for heart rate estimation.
• A high correlation coefficient of 0.98 was observed between the estimated and actual heart rates.
• The algorithm demonstrated robustness against motion artifacts across different exercise types.

Conclusions:

• The developed methodology effectively addresses the challenge of motion artifacts in PPG-based heart rate monitoring.
• The integration of acceleration information significantly improves the accuracy and reliability of heart rate estimation during physical activities.
• This approach holds promise for accurate heart rate monitoring in sports and rehabilitation settings.