Pranav Bhagirath

135 Integrated whole-heart computational workflow for inverse potential mapping and personalized simulations BACKGROUND Inverse potential mapping (IPM) and simulations of cardiac activation are promising computational techniques for non-invasive assessment of rhythm disorders [1,2] . Recent studies have examined the role of simulationmodels for personalizing catheter ablation strategies [3,4] . Furthermore, catheter ablation guidance by IPM shows a higher accuracy when compared to conventional mapping procedures [5,6] . In general, IPM requires a (computational) mesh representing the thoracic and cardiac volumes to reconstruct cardiac activation sequences from body surface potentials (BSP). Similarly, realistic simulations demand for a patient-specific mesh, incorporating functional information about tissue characteristics such as electrical conductivity, mechanical deformation and fiber orientation [3,4,7] . In contrast to meshes used solely for visualization (shells), the computational meshes for these purposes require topologically correct segmentations. Although the currently available models are useful, they are very time consuming [4] , or too comprehensive (multiple parameters) [7] , and therefore not ready for use on a routine basis in the clinical arena. In addition, none of the currently available methods provide an integrated whole-heart (topologically correct) mesh, incorporating both the atria and ventricles. This limits a comprehensive and integrated study of whole-heart electrical interaction. This article proposes a comprehensive finite element model based whole-heart computational workflow suitable for IPM and efficient personalized simulations. The clinical feasibility of reconstructing IPM was explored using BSP’s of healthy volunteers. Subsequently, the simulation features were explored by generating activation maps (isochrones) in different models of human hearts, both normal and with structural heart disease. METHODS The computational workflow for whole-heart electrical assessment consists of four steps ( figure 1 ). These steps involve (1) acquisition of BSP, (2) acquisition of subject specific geometry using cardiac magnetic resonance imaging (MRI), (3) topologically correct segmentation and generation of the computational mesh and (4) utilizing the mesh for reconstructing cardiac surface potentials or conducting simulations.