Advancing our understanding the basic mechanisms of atrial fibrillation using novel computational approaches

Atrial fibrillation (AF) is the most common heart rhythm disturbance and is associated with substantial morbidity and mortality. The overall prevalence of AF varies from 2% to 5% of the general population worldwide, and it is projected to more than double in the following couple of decades, becoming a global epidemic.

However, the current clinical treatment of AF is suboptimal. There are three types of clinical treatment for patients with AF: 1) pharmacological approaches for rate and rhythm control; 2) electrical cardioversion; and 3) cryoablation/catheter ablation/surgical (maze) ablation. Recent population-based studies suggest that all three treatment options often lose their effectiveness and have side effects, especially for patients with persistent or permanent AF or AF patients with concurrent diseases. The high-profile clinical trials, including the Substrate and Trigger Ablation for Reduction of Atrial Fibrillation Part 2 (STAR-AF II), Focal Impulse and Rotor Modulation (FIRM) trial and subsequent studies using the FIRM approach by different international groups, have generated mixed outcomes from ablation treatment using existing ablation strategies.

The main reasons for the poor performance of current clinical treatment for AF are due to 1) lack of basic understanding of the underlying patient-specific atrial substrate which sustains AF directly; 2) incomplete knowledge of potential risk factors of AF and nonexistence of effective upstream approaches for AF prevention; 3) need for quantitative tools to investigate effective strategies which are impossible under clinical/experimental settings.

Computational approaches are the driving force behind our advancing understanding of AF. For example, computer models of atrial electrical activation provide a powerful analysis framework primarily for three purposes, 1) to illustrate the basic electrical and structural mechanisms behind cardiac arrhythmias; 2) to test the dynamic impact of antiarrhythmic drugs from cell to organ level; and 3) to investigate optimal ablation lesions and ablation treatment strategies. Other typical computational approaches include the forward/inverse computing approach used in body surface mapping, phase singularity analysis adapted in the FIRM studies and more recent machine learning approach in facilitating automatic analysis of ECG data.

The aim of our group is to enhance our understanding and treatment of AF in the development or use of novel computational approaches such as structural analysis, computer simulations, signal processing and machine learning.

The ideal candidates for the PhD positions will have a Masters’ or a Bachelors’ degree with Honours (Second Class Honours, Division One or better) in Engineering or Mathematics, and are good at least one computer language (Matlab/C/C++/Fortran). Experience in scientific/numerical computing would be an advantage. The candidate needs to be passionate about research.