Quantitative EEG Detects REM Sleep to Enhance Epileptogenic Zone Localization
Abstract number :
1.117
Submission category :
3. Neurophysiology / 3C. Other Clinical EEG
Year :
2016
Submission ID :
188198
Source :
www.aesnet.org
Presentation date :
12/3/2016 12:00:00 AM
Published date :
Nov 21, 2016, 18:00 PM
Authors :
Marcus C. Ng, University of Manitoba, Winnipeg, Canada
Rationale: REM sleep has been shown to localize the epileptogenic zone in challenging multifocal epilepsies in which interictal discharges are diffuse. However, REM sleep may be rare and easily overlooked in the Epilepsy Monitoring Unit (EMU). This study sought to determine whether quantitative EEG (QEEG) software enhances REM sleep detection, and whether these detections contribute to localization of the epileptogenic zone. Methods: Automated artifact recognition in QEEG software was retrospectively applied to 581 nights of EMU recording from 100 patients over 12 months. QEEG-based REM sleep detection was defined as a contiguous run of eye artifact without muscle artifact. All QEEG-based REM sleep detections were individually confirmed by a board-certified epileptologist. QEEG-based detection was compared to manual REM sleep detection, which had already been performed at the time of original recording. As part of routine quality improvement, the index of suspicion for manual REM sleep detection had been raised 6 months into the 12 month study period. Finally, the impact of unique QEEG-based REM sleep detections on localization was assessed. Results: REM sleep occurred in 77% of 581 nights of EMU recording (n=100). 36 patients achieved REM sleep each night and 62 patients achieved REM sleep on at least one night. The mean EMU admission length was 5.83 days. The mean duration of REM sleep was 5.92 minutes over 1.88 mean nightly bouts of REM sleep. In the 102 nights of EMU recording with seizures, there were significantly less bouts of REM sleep (1.65 vs. 1.92, p=0.038). Raising the level of suspicion increased manual detection rates from 22.6% to 40.5%; however, the QEEG-based detection rate remained steady at 96%. QEEG-based REM sleep detections uniquely provided additional localizing information in 10% of epilepsy patients (n=7). Conclusions: Although REM sleep is common in the EMU, bouts of REM sleep are few and they are brief. Automated artifact recognition in QEEG software maximizes the capture of these REM sleep episodes to enhance localization of the epileptogenic zone. Funding: University of Manitoba Vice President Research Office, University Medical Group Department of Medicine Grant
Neurophysiology