Cardiac Resynchronization Therapy (CRT) is an effective treatment for patient with heart failure and electrical dyssynchrony. However, about 30% of patients do not respond, usually due to suboptimal activation of the left ventricle. The clinical choice for the activation on the left ventricle is based on the latest activation time (LAT) in the epicardial veins. The mapping procedure to obtain LAT is invasive and often associated to X-ray exposure. We propose a pipeline to support this clinical procedure, with the aim at reducing the required mapping procedure. In particular, we consider the monodomain model for the mathematical description of the activation times in the context of a normal excitation, i.e. in absence of arrhythmias and fibrillations. We studied about ten patients (seven of them with scars) who suffered from LBBB and used the corresponding maps of activation times obtained by catheters to estimate the conduction properties of the patient. In particular, a first set (half) of data were used to estimate the conductivities (also in the scars) and a second half to compute the errors of the numerical simulations. We found an excellent agreement between measures and numerical results. Thus, our model is able to predict with great accuracy the activation in the epircadial veins. In addition, since we are able to compute the activation times in all the reconstructed veins, we can explore also those points not mapped by the catheter and in principle localize the effective point with LAT. This makes our model a promising tool to compute the LAT in a more precise way and with a reduced mapping procedure.