A Novel ECG Denoising Scheme Using the Ensemble Kalman Filter.

Area of Science:

• Biomedical Engineering
• Signal Processing
• Wearable Technology

Background:

• Electrocardiogram (ECG) monitoring is essential for detecting cardiovascular anomalies.
• Wearable devices enable home-based physiological signal acquisition but often yield noisy data.
• Efficient ECG filtering is critical for reliable signal interpretation.

Purpose of the Study:

• To develop a novel Ensemble Kalman Filter (EnKF) method for denoising ECG signals.
• To compare the EnKF method against various established filtering algorithms.
• To evaluate the performance of ECG denoising techniques using motion artifacts.

Main Methods:

• Development of a novel Ensemble Kalman Filter (EnKF) for ECG signal denoising.
• Comparative analysis with Savitzky-Golay (SG), EEMD, NLMS, RLS, TVD, Wavelet, and EKF filters.
• Utilized data from the MIT-BIH Noise Stress Test database with added motion artifacts.

Main Results:

• The proposed EnKF method achieved an average signal-to-noise ratio (SNR) of 10.96.
• Demonstrated a Percentage Root Difference of 150.45.
• Achieved a correlation coefficient of 0.959, indicating high signal fidelity.

Conclusions:

• The novel Ensemble Kalman Filter (EnKF) offers a robust solution for denoising ECG signals in the presence of noise and motion artifacts.
• EnKF outperforms other tested filtering methods in enhancing ECG signal quality for wearable health monitoring.
• This advanced ECG filtering technique is vital for accurate cardiovascular anomaly detection in home settings.