Abstract

Signal processing methods may constitute a substantial complement to visual analysis of SEEG signals in providing quantified information on signals (e.g. morphological characteristics) and in computing meaningful quantities that are not accessible to visual inspection (e.g. spectral properties or synchrony). In addition, and complementary to signal processing, computational neuroscience aims at developing models of epileptogenic networks and ultimately explaining some mechanisms involved in the generation of epileptiform activity. This chapter reviews a number of signal processing methods (time–frequency analysis, epileptogenicity index, and nonlinear correlation analysis) and computational models (at micro- and mesoscopic levels). The methods and models described illustrate the insight that can be gained about the information conveyed by SEEG signals recorded from epileptogenic networks observed during interictal (spikes and high-frequency oscillations) and ictal (fast-onset discharges) periods. Provided examples show that appropriate processing/modelling methods applied to electrophysiological signals can considerably improve the interpretation of SEEG recordings.