An ICA based method to process simultaneous EEG-fMRI in Epilepsy
Figueiredo P, Marques J, Sales F, Castelo-Branco M
Universidade de Coimbra
A methodology based on independent component analysis (ICA) of the EEG signal recorded simultaneous to an fMRI acquisition is explored in this abstract. Due to the large amplitude of EEG signal during epileptic seizures, ICA has little problem in separating out the component(s) relating to that interictal activity. The relevant components are then used to model the time course of the event related design. This approach, when compared to the more conventional methodology, seems to have increased sensitivity to BOLD changes relating to interictal activity, whilst preserving the most relevant regions of activity found with the conventional methodology.