Source Localization of Normal Variants Seen on EEG

Abstract

Purpose: The EEG is an essential neurological diagnostic tool. EEG abnormalities can guide diagnosis and management of epilepsy. There are also distinctive EEG waveforms that are seen in healthy individuals. It is critical not to misinterpret these as abnormal. To emphasize the importance of these waveforms, we analyzed different normal variants via the source localization technology. Methods: This is a retrospective analysis of EEGs performed at the Duke University Hospital between June 2014 and Dec 2019. We selected samples of vertex waves, Mu, lambda, POSTS, wickets, and sleep spindles for analysis. EEG were imported to Curry 8 (Compumedics) to calculate the dipole and current density. The averaged head model from the Montreal Neurological Institute database was used for reconstruction. Results: Thirty-four patient EEG samples were selected including five vertex, six Mu, four wicket, seven lambda, five POSTS, and seven spindles. Results from source localization showed that vertex waves are localized in the frontocentral area, whereas spindles in the deep midline central region. Mu were identified in the ipsilateral somatosensory cortex. Lambda and POSTS, on the other hand, had maximum results over the bilateral occipital region and wickets in the ipsilateral temporal lobe. Conclusions: Our results confirm and expand previous hypotheses. This allows us to speculate on the origin of these normal EEG variants. Although this study is limited by small sample size, lack of high-density EEG, and patient-specific MRI, our analysis provides an easily replicable three-dimensional visualization of these waveforms.