A Database of Simultaneously Recorded ECG Signals With and Without EMG Noise

Area of Science:

• Biomedical Engineering
• Signal Processing
• Cardiology

Background:

• Noise in electrocardiographic (ECG) signals can compromise clinical interpretation.
• Electromyographic (EMG) noise overlaps with the ECG's QRS complex, complicating removal.
• Current methods for evaluating ECG denoising lack a standardized database.

Purpose of the Study:

• To introduce a novel database for evaluating ECG denoising algorithms.
• To provide a standardized resource for comparing different noise-removal techniques.
• To address the absence of a unique, open-source database for ECG signal processing.

Main Methods:

• Developed a novel acquisition method for recording ECG signals.
• Recorded both genuine EMG-noise-free and EMG-noise-contaminated ECG signals.
• Created the SimEMG database using this direct recording approach.

Main Results:

• Successfully acquired and compiled a database of genuine ECG signals with and without EMG noise.
• The SimEMG database offers a direct comparison of noise-free and contaminated signals.
• The database is made available as an open-source resource.

Conclusions:

• The SimEMG database provides a unique and standardized platform for testing ECG denoising methods.
• Facilitates objective comparison and advancement of algorithms for removing EMG noise from ECG.
• Promotes reproducible research in biomedical signal processing.