Reference for Electrocardiographic Imaging-Based T-Wave Alternans Estimation

Area of Science:

• Cardiology
• Biomedical Engineering
• Signal Processing

Background:

• Sudden cardiac death (SCD) is a significant cause of mortality.
• T-wave alternans (TWA) are reliable predictors of SCD.
• Electrocardiographic imaging offers spatial-temporal insights but lacks specific references and segmentation methods for TWA analysis.

Purpose of the Study:

• To develop a reference standard for evaluating TWA estimation methods.
• To introduce a novel T-wave segmentation procedure for electrocardiographic imaging data.
• To create a bootstrap-based classifier for TWA detection.

Main Methods:

• A novel T-wave segmentation procedure was developed and compared to a standard method.
• A spatial-temporal Gaussian function was used to integrate TWA into epicardial signals, creating a reference.
• A bootstrap-based classifier was implemented for TWA detection.

Main Results:

• The novel T-wave segmentation procedure demonstrated superior performance compared to the commonly used method.
• The generated reference enabled the detection of TWA with amplitudes as low as approximately 0.15 mV amidst noise.
• The developed classifier showed consistent performance, detecting TWA with amplitudes around 0.08 mV.

Conclusions:

• This study provides a crucial spatial-temporal reference for evaluating TWA estimation methods.
• The developed methods enhance the accuracy and reliability of TWA detection.
• Establishing a gold standard for TWA analysis can aid in reducing SCD incidence.