Quantitative spatio-temporal characterization of epileptic spikes using high density EEG: Differences between NREM sleep and REM sleep

Xuan Kang,Melanie Boly,Benjamin Jones,Aaron F Struck,Klevest Gjini,Graham Findlay,Rama Maganti

doi:10.1038/s41598-020-58612-4

Xuan Kang, Melanie Boly + Show 5 more

Open Access

https://doi.org/10.1038/s41598-020-58612-4

Copy DOI

Journal: Scientific Reports	Publication Date: Feb 3, 2020
Citations: 19	License type: open-access

Affiliation: University of Wisconsin–Madison

Abstract

In this study, we applied high-density EEG recordings (HD-EEG) to quantitatively characterize the fine-grained spatiotemporal distribution of inter-ictal epileptiform discharges (IEDs) across different sleep stages. We quantified differences in spatial extent and duration of IEDs at the scalp and cortical levels using HD-EEG source-localization, during non-rapid eye movement (NREM) sleep and rapid eye movement (REM) sleep, in six medication-refractory focal epilepsy patients during epilepsy monitoring unit admission. Statistical analyses were performed at single subject level and group level across different sleep stages for duration and distribution of IEDs. Tests were corrected for multiple comparisons across all channels and time points. Compared to NREM sleep, IEDs during REM sleep were of significantly shorter duration and spatially more restricted. Compared to NREM sleep, IEDs location in REM sleep also showed a higher concordance with electrographic ictal onset zone from scalp EEG recording. This study supports the localizing value of REM IEDs over NREM IEDs and suggests that HD-EEG may be of clinical utility in epilepsy surgery work-up.

Highlights

Epilepsy affects over 60 million individuals worldwide and carries significant risks of severe injury and sudden death[1]
Group analysis at the negative peak demonstrated 30% reduction (p = 0.005) of ictal epileptiform discharges (IEDs) spatial extent during rapid eye movement (REM) sleep compared to non-rapid eye movement (NREM) sleep (Fig. 3c), with individual effect size ranging from 8% to 75%
Similar to results obtained with Minimum Norm (MN), group analysis at the negative peak using coherent Maximum Entropy on the Mean (cMEM) demonstrated 4.3% spatial extent reduction (p = 0.04) of IED spatial extent (Fig. 3d) during REM sleep compared to NREM sleep, with individual effect size ranging from 0.3% to 13.7%

Summary

Introduction

Epilepsy affects over 60 million individuals worldwide and carries significant risks of severe injury and sudden death[1]. The study of inter-ictal epileptiform discharges (IEDs) is used to localize a cortical irritative zone, which often overlaps with the EZ. Invasive EEG (iEEG) could in principle provide more accurate localization for EZ12,13, but invasive studies are inherently limited by spatial sampling and cannot cover the whole brain. Their success depends on hypotheses developed from non-invasive studies. EEG source imaging (ESI) techniques are alternative methods that provide more accurate location of the cortical generators of the EEG signals compared to 10–20 clinical EEG studies, while remaining non-invasive[14,15]. Tomography; MRI: Magnetic resonance imaging; VNS: Vagus Nerve stimulator; F: female; M: male; EEG: Electroencephalogram. *The terms “Ictal onset zone’ and “Inter-ictals” refers to the location of seizure onset zone and interictal epileptiform discharges captured on clinical 10–20 EEG during the same EMU admission

Objectives

Methods

Findings

Conclusion