Within the euHeart project, WP6 aims at developing personalized biophysical models of the atria and ventricles in order to improve the planning of radiofrequency ablation (RFA) on patients suffering from atrial fibrillation (AF) and ventricular tachycardia (VT). This deliverable makes the link between the data acquisition performed at King's College London and the electrophysiological models developed by IBT and INRIA for the simulation of AF and VT. Indeed, model personalization aims at estimating model parameters, boundary conditions, computational meshes that are coherent with the available dataset. Ideally, the simulated output of a personalized model should match the measurements extracted from patient specific images and biosignals. In WP6, the available datasets consist mostly of anatomical MR images and Body Surface Potential Mapping acquired on the torso for the atria and noncontact endocardial electrophysiological mappings for the ventricles. In this document, we describe the adjustment of atrial models to better fit certain high risk patient groups and to account for electrical remodeling due to AF. Furthermore, with some examples of personalization on individual patients, we introduce some work in progress methods to estimate the electrophysiological properties of atrial models in simple and robust ways from non-invasive data. For the ventricles, a complete personalization of electrophysiological models has been performed from two ischemic cases with a fairly good agreement between simulation and measurements. The estimated patient-specific parameters include tissue conductivity and restitution curves of action potential duration and conductivity. Those parameters and especially their spatial heterogeneity is key to the initiation and sustainability of VT.