Adaptive Trend Filtering for ECG Denoising and Delineation.

Area of Science:

• Biomedical Engineering
• Signal Processing
• Medical Diagnostics

Background:

• Electrocardiogram (ECG) recordings are often degraded by noise and interference.
• This noise compromises the accuracy of ECG analysis and subsequent medical diagnoses.

Purpose of the Study:

• To develop a method for removing noise artifacts from ECG signals.
• To accurately locate the main features (peaks) of ECG pulses.

Main Methods:

• Utilized compressive sensing techniques for noise reduction and feature localization.
• Employed trend filtering with a varying proximal parameter to capture ECG peaks with diverse regularities.
• Implemented an adaptive version of the alternating direction method of multipliers (ADMM) algorithm.

Main Results:

• Demonstrated successful noise artifact removal in ECG signals.
• Achieved highly accurate peak localization in both simulated and real ECG data.
• Results were found to be comparable to existing state-of-the-art approaches.

Conclusions:

• The proposed compressive sensing method effectively removes noise from ECG signals.
• The technique accurately identifies key ECG features, improving diagnostic potential.
• This approach offers a promising solution for enhancing ECG analysis and diagnosis.