Heart Rate Estimation Considering Reconstructed Signal Features Based on Variational Mode Decomposition via...

Area of Science:

• Biomedical Engineering
• Signal Processing
• Radar Technology

Background:

• Accurate heart rate estimation is crucial for non-invasive health monitoring.
• Existing radar-based methods face challenges in separating heartbeat signals from respiration and motion artifacts.
• Privacy concerns and the need for through-clothing measurement highlight the demand for advanced techniques.

Purpose of the Study:

• To develop a novel heart rate estimation method using MIMO FMCW radar and VMD.
• To improve the accuracy and robustness of radar-based heart rate monitoring.
• To enable reliable heart rate estimation in scenarios with multiple subjects.

Main Methods:

• Utilizing Multiple-Input Multiple-Output (MIMO) Frequency Modulated Continuous Wave (FMCW) radar for signal acquisition.
• Applying Variational Mode Decomposition (VMD) to decompose radar signals into Intrinsic Mode Functions (IMFs).
• Extracting the heartbeat-specific IMF based on its center frequency and reconstructing the heart rate signal.

Main Results:

• The proposed VMD-based MIMO FMCW radar method achieved a Mean Absolute Error (MAE) of 2.54 BPM for a single subject.
• Demonstrated accurate simultaneous heart rate estimation for two subjects with an MAE of 2.28 BPM.
• Outperformed traditional Doppler radar methods, showing significantly lower error rates.

Conclusions:

• The VMD-based MIMO FMCW radar approach offers a robust and accurate solution for non-contact heart rate estimation.
• The method effectively isolates heartbeat signals, overcoming limitations of conventional techniques.
• This technology holds promise for enhanced privacy-preserving remote health monitoring.