On the Role of Skull Parcellation in the Computational Modeling of Human Head Conductivity

In this paper we propose a method to include skull inhomogeneities in the EEG forward/inverse calculation by parcellating the skull based on a combination of skull anatomy and thickness. We show that it is enough to determine a few skull parameters (4 6) in the inverse search to capture the skull inhomogeneities if we also include the thickness. The validity of the linear relation between skull conductivities and skull thickness is investigated, our results show that this relation applies. The effect of skull inhomogeneities on the conductivity reconstruction of the human head tissues is presented. Our simulation results show here that the models approximating the skull as a homogenous tissue could produce an order of magnitude error in the retrieved conductivities in comparison with the models based on parcellated skull.