Decomposing atrial activity signal by combining ICA and WABS

Area of Science:

• Biomedical Engineering
• Signal Processing
• Cardiology

Background:

• Atrial Fibrillation (AF) is a common arrhythmia.
• Analyzing Atrial Activity (AA) in Electrocardiogram (ECG) is crucial for diagnosing and managing AF.
• Existing methods for AA decomposition have limitations, especially for complex AF cases.

Purpose of the Study:

• To propose a novel technique for decomposing Atrial Activity (AA) signals in Electrocardiogram (ECG) during Atrial Fibrillation (AF).
• To address the limitations of current Blind Source Separation (BSS) algorithms in extracting AA signals from multi-lead ECG in AF patients.

Main Methods:

• A hybrid statistical approach combining Independent Component Analysis (ICA) and a new Weighted Average Beat Subtraction (WABS) method.
• Application of the proposed technique to decompose multi-lead AA signals from surface ECG data in AF patients.

Main Results:

• The proposed technique successfully decomposes multi-lead AA signals from surface ECG during AF.
• Demonstrated feasibility and effectiveness of the novel AA decomposition method through clinical data verification.

Conclusions:

• The novel technique offers an effective solution for AA signal decomposition in AF.
• This method overcomes limitations of existing BSS algorithms for complex AF scenarios.
• The validated approach shows promise for improved AF analysis using ECG.