A multichannel decision-level fusion method for T wave alternans detection.

Area of Science:

• Cardiology
• Biomedical Engineering
• Signal Processing

Background:

• Sudden cardiac death (SCD) is a leading cause of mortality in cardiac patients.
• Ventricular arrhythmia, a predictor of SCD, can be identified by T wave alternans (TWA).
• Detecting TWA from body-surface electrocardiograms (ECG) is crucial for SCD prevention.

Purpose of the Study:

• To develop a novel multichannel decision-level fusion method for TWA detection.
• To enhance the accuracy and robustness of TWA detection in ECG signals.
• To address challenges posed by multi-source TWA information, nonlinear propagation, and noise.

Main Methods:

• Proposed a multichannel decision-level fusion method based on the Dezert-Smarandache Theory.
• Utilized a redistribution mechanism to handle competitive information from different ECG channels.
• Evaluated performance using real ECG records from the Physikalisch-Technische Bundesanstalt database.

Main Results:

• The proposed method achieved higher detection accuracy and robustness compared to single-channel methods.
• Demonstrated significant performance improvement (12%-20%) over traditional 1D decision methods.
• Showcased potential for low-cost instruments and portable applications due to reduced synchronous sampling demands.

Conclusions:

• The Dezert-Smarandache Theory-based fusion method effectively improves TWA detection from multichannel ECG.
• This approach offers a more reliable and robust strategy for predicting SCD.
• The method's efficiency and reduced sampling requirements make it suitable for practical clinical use.