A Singular-Value-Based Map to Highlight Abnormal Regions Associated With Atrial Fibrillation Using High-Resolution...

Area of Science:

• Cardiovascular Electrophysiology
• Biomedical Signal Processing
• Medical Diagnostics

Background:

• Atrial fibrillation (AF) severity is typically assessed using metrics like conduction block (CB) and continuous conduction delay and block (cCDCB) from epicardial electrograms.
• Existing methods focus on conduction velocity and wavefront propagation, overlooking crucial information from atrial action potential morphology.
• A need exists for novel analytical approaches to capture complementary electrophysiological properties for improved AF assessment.

Purpose of the Study:

• To derive and evaluate new features based on atrial potential waveform morphology for detecting variations associated with atrial fibrillation.
• To explore the utility of singular value decomposition of epicardial measurements for characterizing AF.
• To investigate the potential of this non-parametric method for identifying electropathological regions.

Main Methods:

• Utilized singular value decomposition (SVD) on epicardial measurement matrices to analyze spatial variations in atrial potential morphology during a single beat.
• Developed a non-parametric method requiring minimal preprocessing.
• Conducted simultaneous measurements of electrograms (EGMs) and multi-lead electrocardiograms (ECGs) to compare invasive and non-invasive data.

Main Results:

• Normalized singular values were significantly higher during AF compared to sinus rhythm (SR).
• The difference in normalized singular values between AF and SR was more pronounced in non-invasive ECG data than in EGM data under favorable electrode placement.
• Singular value maps effectively highlighted areas susceptible to conduction fractionation and block.

Conclusions:

• Singular value-based features derived from atrial potential morphology offer a valuable tool for detecting and evaluating atrial fibrillation.
• The proposed method provides a promising approach for identifying electropathological regions without the need for local activation time estimation.
• This technique enhances the understanding of AF pathophysiology by incorporating waveform morphology.