An efficient ECG baseline removal filter based on frequency response masking technique for wearable applications

Area of Science:

• Biomedical Engineering
• Signal Processing
• Digital Filters

Background:

• Baseline wander in electrocardiogram (ECG) signals degrades diagnostic accuracy.
• Conventional finite impulse response (FIR) filters for baseline wander removal can be computationally intensive.

Purpose of the Study:

• To develop a computationally efficient FIR filter for effective ECG baseline wander removal.
• To optimize filter specifications for high signal-to-noise ratio (SNR) and minimal coefficient count.

Main Methods:

• Utilized the frequency-response masking (FRM) technique to design the FIR filter.
• Optimized filter specifications to maximize SNR and minimize computational complexity.
• Evaluated filter performance using ECG signals from the MIT-BIH Arrhythmia Database and simulated baseline noise.

Main Results:

• The proposed FRM filter achieved a 90% reduction in multiplications compared to conventional FIR filters.
• Demonstrated effective removal of baseline wander from ECG signals.
• Preserved the integrity of the essential ECG signal components during noise reduction.

Conclusions:

• The FRM-based FIR filter offers a computationally efficient solution for ECG baseline wander removal.
• The filter provides a favorable trade-off between performance (SNR) and complexity.
• This method holds promise for real-time ECG signal processing applications.