Nonstationary harmonic modeling for ECG removal in surface EMG signals

Area of Science:

• Biomedical Engineering
• Signal Processing
• Neuroscience

Background:

• Surface electromyography (EMG) signals are often contaminated by electrocardiogram (ECG) artifacts.
• Variability in heart rate and QRS complex causes amplitude and frequency fluctuations in ECG, complicating artifact removal.

Purpose of the Study:

• To develop a compact and efficient method for mitigating ECG interference in EMG signals.
• To preserve low-frequency EMG signal content during artifact suppression.

Main Methods:

• Utilized time-variant harmonic modeling of cardiac artifacts.
• Employed third-order constant-coefficient polynomials to capture ECG variability.
• Modulated a stationary harmonic basis within the analysis window.
• Applied a least-squares solution to a linear system for efficient parameter estimation.

Main Results:

• Successfully suppressed ECG artifacts from mixed EMG-ECG signals.
• Preserved significant low-frequency EMG signal content (below 20 Hz).
• The proposed method demonstrated superior performance compared to two reference methods in EMG preservation.

Conclusions:

• The time-variant harmonic modeling approach offers an effective and compact solution for ECG artifact removal in EMG.
• This method enhances the quality of EMG recordings by minimizing cardiac interference while retaining valuable physiological information.