Pranav Bhagirath

224 Appendices correlated with viable myocardium on the scar-map. These findings suggested that a CMR based workflow is feasible and clinically usable to perform a detailed study of the left atrial (LA) geometrical and structural changes using a single examination. Clinical application of this workflow could significantly improve patient selection and procedural guidance of (redo-) AF ablation. In Chapter 5 we studied the effect of LA geometrical changes (sphericity), following the index ablation of AF, on the outcome of redo-ablation procedures. All patients with a reduction of sphericity following the index ablation were free from AF following the redo-ablation, whereas all patients with an increased sphericity, had recurrence of AF. These results suggest that selection for redo-ablation can be improved by including LA sphericity as a stratification factor. In Chapter 6 we examined the requirements and clinical feasibility of a dedicated interventional CMR (iCMR) suite for EP procedures. First, the limitations of current EP procedures and ablation strategies are analyzed and the advantages of an iCMR suite were discussed. Second, the clinical feasibility was examined by presenting the current challenges of working in an MRI environment. Safety, imaging and device related aspects were also reviewed. Finally, the requirement for implementing an iCMR suite and the current state of their development was addressed. In Chapter 7 we proposed a clinically applicable, integratedwhole-heart computational workflow for non-invasive, in-vivo assessment of electrical activation using IPM and cardiac activation simulations. In contrast to currently available methods, the presented work-flowperformed an integrated evaluation of inverse potential mapping and cardiac activation simulations, allowing for a comprehensive clinical study of whole-heart electrical interaction. When integrated with tissue characteristics as provided by cardiac MRI, this workflow provided a realistic, patient-specific model for electrical modeling. In Chapter 8 the feasibility and relevance of computing body volume potentials (BVP) for IPM was investigated using a rectangular tank filled with an electrolytic conductor and a patient specific 3D model. Efficient generation of high quality volume meshes and computation of BVP with a resolution of 5 mm was found feasible. This allowed incorporation of different anisotropies, local tissue characteristics and sigma gradients over different regions which could improve understanding of the genesis of body surface potentials and sources of local inaccuracies.