In epileptic patients candidate to surgery, the interpretation of EEG signals recorded either within (depth EEG) or at the surface (scalp EEG) of the head is a crucial issue to determine epileptogenic brain regions and to define subsequent surgical strategy. This task remains difficult as there is no simple relationship between the spatiotemporal features of neuronal generators (convoluted cortical dipole layers) and the electric field potentials recorded by the electrodes. Indeed, this relationship depends on the complex interaction of several factors regarding involved cortical sources: location, area, geometry, and synchronization of neuronal activity. A computational model is proposed to address this issue. It relies on a neurophysiologically relevant model of EEG signals, which combines an accurate description of both the intracerebral sources of activity and the transfer function between dipole layers and recorded field potentials. The model is used, on the one hand, to quantitatively study the influence of source-related parameters on the properties of simulated signals, and on the other hand, to jointly analyze depth EEG and scalp EEG signals. In this article, the authors review some of the results obtained from the model with respect to the literature on the interpretation of EEG signals in the context of epilepsy.
From the *Bernstein Center for Computational Neuroscience, Albert-Ludwigs Universität, Freiburg, Germany; †INSERM, U642; ‡LTSI, Université de Rennes 1, Rennes; §INSERM, U751; ∥Aix Marseille Université, Faculté de Médecine; and ¶Assitance Publique–Hôpitaux de Marseille, Hôpital de la Timone, Service de Neurophysiologie, Clinique, Marseille, France.
Supported by the French Ministry of Research (ACI Neurosciences Intégratives et Computationnelles 2002–2006) and by Inserm (2007–2008).
Address correspondence and reprint requests to Fabrice Wendling, INSERM U642, Laboratoire Traitement du Signal et de l'Image, Université de Rennes 1, Campus de Beaulieu, Bât. 22, 35042 Rennes Cedex, France; e-mail: email@example.com.