Aim: Despite atrial fibrillation being the most commonly occurring sustained cardiac arrhythmia its treatment remains the subject of great debate. DC cardioversion is a treatment option often used to return normal cardiac function. We aimed to assess relapse rates following DC cardioversion and to determine whether transoesophageal echocardiography, P wave signal averaged electrocardiography or heart rate variability measurements had a role in identifying patients likely to relapse to atrial fibrillation. Methods: Patients who were referred for DC cardioversion of chronic non-valvular atrial fibrillation were enrolled into the study. Transoesophageal echocardiography was performed in order to measure left atrial size, left atrial appendage area, and flow velocity within the left atrial appendage, left upper pulmonary vein and across the mitral valve. Patients in whom cardiac thrombus had been excluded proceeded to DC cardioversion. Those patients who achieved sinus rhythm had ap wave signal averaged electrocardiogram recorded one hour following the procedure. At forty eight hours those patients remaining in sinus rhythm had a second p wave signal averaged ECG recorded and a Holter recording in order to determine heart rate variability. Patients were reviewed at three and six months for relapse to atrial fibrillation. Results: DC cardioversion was initially successful in 66 of the 81 patients (8 1 %). At 48 hours 23 patients (35%) had relapsed to AF. No PSAECG measurement differed significantly between these groups. Mean mitral valve flow velocity differed significantly between those who relapsed to AF within 48 hours and those who remained in sinus rhythm (SR group = 83.98cm/s, AF group 71.05cm/s, p=0.048). At three months 48 patients had relapsed to AF (73%) this increased to 51 patients (77%) at six months. No significant difference was observed in any of the TOE, PSAECG or HRV measurements in these groups. Conclusion: No PSAECG or HRV variable helped to predict long term success. TOE measurement of mitral valve flow velocity may allow prediction of early relapse. DC cardioversion without antiarrhythmic prophylaxis leads to a high relapse rate.

Death due to lethal cardiac arrhythmias is the leading cause of mortality in Western society. Many of the fundamental mechanisms underlying the onset of arrthythmias, their maintenance and termination, still remain poorly understood. The specialized conduction (or His-Purkinje) system is fundamental to ventricular electrophysiological function and is a key player in various cardiac diseases. In recent years, computational simulation has become an important tool in im- proving our understanding ofthese mechanisms. Current state-of-the-art computational ventric- ular electrophysiology models often do not feature a detailed representation of the specialized conduction system. Ventricular models that do incorporate the specialized conduction system often use a simplified anatomical description and are commonly based on the monodomain equations, rather than the more general bidomain equations. Thus, using computational simula- tion to investigate both normal physiological function of the specialized conduction system and pathologies in which it is involved presents difficulties. This thesis develops the techniques and tools required to model the specialized conduction sys- tem at the ventricular scale. We derive one-dimensional bidomain equations that model elec- trical propagation in the system by reducing the equations associated with a three-dimensional fibre. To complement the derived equations, we develop a numerical solution scheme for the model that is efficient enough to allow ventricular simulations. The one-dimensional bido- main model allows defibrillation studies to be performed with the specialized conduction sys- tem. Secondly, we investigate the imaging and mesh generation tools required to integrate an anatomically detailed mesh of the specialized conduction system into a current state-of-the-art ventricular mesh. Using these tools, a highly detailed rabbit-specific specialized conduction system anatomical model is developed. Simulations are performed that dem~strate the re- sponse of the specialized conduction system to defibrillation strength shocks and we compare activation sequences generated using the model to experimental recordings. Finally, we investi- gate variability in the anatomy of the system. The tools and ventricular model presented in this thesis fulfil an important role in allowing the study of the e1ectrophysiological function of the specialized conduction system at the ventricular scale.