Removing ECG Artifact from the Surface EMG Signal Using Adaptive Subtraction Technique

Area of Science:

• Biomedical Engineering
• Signal Processing
• Neuroscience

Background:

• Electrocardiogram (ECG) artifact significantly contaminates electromyogram (EMG) signals from upper trunk muscles.
• This contamination renders the recorded EMG signals unusable for analysis.

Purpose of the Study:

• To develop and evaluate a method for removing ECG contamination from EMG signals.
• To improve the utility of EMG recordings from the upper trunk.

Main Methods:

• Clean EMG and ECG signals were recorded from leg muscles and V4 lead, respectively.
• A contaminated EMG signal was simulated by combining clean EMG and ECG artifact.
• An adaptive subtraction method, involving QRS detection, ECG template averaging, and low-pass filtering, was applied for artifact removal.

Main Results:

• The adaptive subtraction method achieved a signal-to-noise ratio of 10.493, relative error of 0.04, and cross-correlation of 97%.
• Performance was assessed using qualitative and quantitative metrics, including power spectrum density and coherence.

Conclusions:

• The adaptive subtraction method demonstrates effectiveness in removing ECG artifact from contaminated EMG signals.
• The proposed method yields acceptable results for cleaning upper trunk EMG recordings.