Non-linear EEG analysis of idiopathic hypersomnia

Abstract

The electroencephalogram (EEG) signals are used to analyse and quantify the "depth" of sleep and its dynamic behaviour during night. In this work, we investigate a direct data-driven nonlinear and non-stationary quantitative analysis of sleep EEG issued from patients suffering from idopathic hypersomnia. We show that the minimum weighted average instantaneous frequency appears to be a specific intrinsic characteristic of brain function mechanism in these patients. It could be an interesting new parameter for the quantification of sleep. © Springer-Verlag Berlin Heidelberg 2010.