Atrial fibrillation subtypes classification using the General Fourier-family Transform

Area of Science:

• Cardiology
• Biomedical Engineering
• Signal Processing

Background:

• Atrial fibrillation (AF) presents in paroxysmal, persistent, and permanent forms, distinguished by temporal patterns.
• Surface electrocardiograms (ECGs) do not inherently reveal these AF subtype classifications.
• Accurate differentiation is crucial for appropriate patient management and treatment strategies.

Purpose of the Study:

• To develop and validate a novel classification method for discriminating between AF subtypes.
• To utilize short ECG segments for subtype classification, overcoming limitations of standard surface ECGs.
• To assess the feasibility of integrating this method into a semi-automated diagnostic system.

Main Methods:

• Processing of electrocardiogram (ECG) recordings using advanced time-frequency analysis techniques.
• Development of a classification algorithm designed to differentiate between paroxysmal, persistent, and permanent atrial fibrillation.
• Evaluation of the method's performance on real patient cases.

Main Results:

• The proposed classification method achieved a global accuracy of 80% in distinguishing AF subtypes.
• Time-frequency analysis proved effective in extracting differentiating features from ECG segments.
• Initial evaluations on real cases demonstrated promising diagnostic potential.

Conclusions:

• A novel ECG-based method can effectively classify atrial fibrillation subtypes with high accuracy.
• Time-frequency techniques offer a viable approach for analyzing ECGs to differentiate AF patterns.
• The developed method shows potential for implementation in semi-automated diagnostic tools for AF management.