On the Design and Implementation of a Highly Accurate Pulse Predictor for Exercise Equipment

Area of Science:

• Biomedical Engineering
• Signal Processing
• Wearable Technology

Background:

• Accurate heart rate monitoring during exercise is crucial for performance and health.
• Existing methods often struggle in noisy environments or require user adjustments.
• Hand-held contact signals offer a potential but challenging source for heart rate data.

Purpose of the Study:

• To develop a highly accurate heart rate monitoring system using hand-held contact signals.
• To achieve reliable heart rate estimation in noisy exercise environments.
• To compare the performance of the developed system against a commercial heart rate monitor.

Main Methods:

• Developed a statistical model using a Laplacian random process to capture signal patterns and uncertainties.
• Derived and implemented a highly accurate pulse predictor (HAPPEE) in real-time on a Cypress PSoC 5LP board.
• Conducted real-time experiments comparing HAPPEE with a POLAR commercial heart rate monitor on seven healthy subjects.

Main Results:

• HAPPEE achieved 99.2% credible estimates, significantly higher than POLAR's 93.6%.
• HAPPEE demonstrated a lower Mean Square Error (MSE) of 3.1 compared to POLAR's 7.7.
• The results indicate superior accuracy of HAPPEE in heart rate monitoring.

Conclusions:

• HAPPEE accurately monitors heart rate from hand-held contact signals in noisy exercise conditions.
• HAPPEE offers advantages over existing methods by not requiring pulse detection or parameter tuning.
• The system's real-time implementation on low-cost hardware makes it suitable for exercise equipment.