High Density Frequency Patterns and Phase Mapping in Human Persistent Atrial Fibrillation

Atrial fibrillation (AF) is the most common cardiac arrhythmia in clinical practice, and it increases the chance of stroke fivefold. The mechanisms underlying the initiation and preparation of AF are still not clearly understood. Dominant frequency (DF) analysis has been widely used as a feature for the analysis of atrial electrograms. Atrial regions that contain high DF (HDF) signals are believed to correspond to the underlying AF drivers. HDF sites have been proven largely unstable in time and space. However, cyclic behaviours of HDF reappearance were noticed in the LA during persAF, suggesting that AF is not totally random. Similarly, re-entrant activity, also known as ‘rotors’ or spiral waves, have been shown to exist in atrial arrhythmias, and are believed to drive AF. Recently, drifting and unstable rotor behaviours were observed in persAF. The current work aims to develop techniques and tools to better track important features, especially when they are spatiotemporal unstable, such as HDF and phase singularities (PSs), to study the underlying AF mechanisms. In this work, a novel interactive graphic user interface was implemented, compatible with a commercial electro-anatomical mapping system (EnSite, St Jude Medical), providing additional features that are not currently available in commercial systems, to guide catheter ablation of persAF. In addition, a new algorithm has been developed to identify reappearing HDF patterns, and these recurring patterns showed high organisation, which could be important atrial sites for ablation. This is the first algorithm that could track recurrent patterns in DF analysis. Lastly, the performance of state-of-art PS detection methods was firstly investigated in non-contact mapping, suggesting that PS detection are method-dependent. Optimised parameters of the methods were proposed to increase detection accuracy for PSs detection in AF.