Quantifying Irregular Morphology Electrograms in Atrial Fibrillation Using Fractional Fourier Domains

Abstract

A fractional Fourier transform-based strategy is proposed for the analysis of morphological irregularities of 933 simulated electrograms recorded from a 3D simulated human atria under atrial fibrillation. Optimum fractional domains are considered to quantify the non-stationary features in the signals after preprocessing and segmentation. Optimum order maps obtained with the proposed strategy are compared with approximate entropy maps generated in previous studies. The results indicate that distinct types of electrical activity can be associated with different fractional domains and that the FrFT can be sensitive to different underlying propagation patterns where the approximate entropy exhibits uniform behavior.