EEG Analysis of Eye Movement Artifact in Parietal and Occipital Lobe Seizures: A Search for Epileptic Nystagmus
Aybuke Acar1, Anthony Zampino1, Marshall Kirsch1, Neel Fotedar1
1Neurology, Case Western Reserve University, University Hospitals Cleveland Medical Center
Objective:

To report a method of eye movement artifact analysis on EEG without EOG electrodes, and to identify a characteristic eye movement artifact in parietal and occipital lobe seizures, consistent with Epileptic nystagmus (EN).

Background:

EN is a rarely reported seizure semiology, most commonly associated with seizures originating in the posterior cortical regions. EN is often easily missed in long-term video-EEG (vEEG) recordings because of multiple reasons such as poor camera resolution or camera not focused on the eyes.

Design/Methods:

We performed a retrospective vEEG analysis of three patients with occipital and parietal lobe seizures. The eyes were not clearly visible in any of the patients. We analyzed 30 seconds of pre-ictal EEG for any unusual eye movements, followed by the ictal EEG. We identified epochs of unilaterally directed repetitive saccadic eye movement artifacts (≥5 saccades) during the ictus. We analyzed these eye movements in a longitudinal bipolar montage with increased time constant (TC) of ≥2 seconds.

Results:
In all three patients, there was no specific pattern of repetitive saccadic eye movement artifact in the pre-ictal period. During the ictus, all of the patients showed repetitive contralaterally directed saccadic eye movements. When analyzed with TC≥2 seconds, the saccades were alternating with linear slow phases directed ipsilateral to the seizure focus. In one patient, there was a clear prolonged slow drift ipsilateral to the seizure focus before the onset of nystagmus, consistent with type II EN.
Conclusions:

EN is likely under-reported because of either lack of video evidence or lack of special EOG electrodes. Our study demonstrates a specific pattern of eye movement artifact seen on EEG in parietal and occipital seizures, consistent with EN. We also report a method to identify the slow and fast phases of EN on EEG, which can also delineate the underlying mechanism of EN (type I v type II).

10.1212/WNL.0000000000202667