Localization of epileptic foci from preictal EEG data using standardized shrinking LORETA-FOCUSS algorithm

Abstract

To localize epileptic foci, EEG source localization methods are often applied to interictal or ictal EEG data. However, ictal epileptiform is often interfered with artifacts caused by the movement of the patient. In this paper, we use an algorithm called Standardized Shrinking LORETA-FOCUSS (SSLOFO) with a three-shell head model to reconstruct the sources from the EEG data of an epileptic patient during four subperiods, with three preictal and one ictal. The results demonstrate that using preictal EEG, SSLOFO can accurately localize the epileptic foci in the left frontal lobe, as has been confirmed by intracranial recordings. The present study also suggests that we may use the trends of the estimated source energy with time to predict epileptic seizures.