Ictal DC Shift Recording on Scalp EEG: Proof of Principle
Aybuke Acar1, Yael Pinero Colon2, James Dolbow1, Ankita Prasad1, Neel Fotedar1
1University Hospitals Cleveland Medical Center Case Western Reserve University, Cleveland, OH, 2Wellspan Health, Lebanon, PA
Objective:
To demonstrate the principle of recording ictal DC shift on scalp EEG with a prolonged time-constant (TC).
Background:
DC shift is well-described as an ictal onset pattern in epilepsies studied with intracranial EEG. There are only few studies demonstrating ictal DC shift on scalp EEG. Conventional EEG recordings are not DC recordings but with a prolonged TC, infraslow activity (<0.1Hz) can be identified.
Design/Methods:

We recorded multiple seizures of a 66-year old woman with intractable generalized epilepsy and intellectual disability. The EEG was recorded using the standard 10-20 international system.

Seizures were first analyzed with the settings of TC=0.1 sec and HFF=70 Hz. Subsequently, settings were adjusted to TC of 5-10 seconds, sensitivity of 15 μV/mm. O1 electrode was used as reference.

We analyzed the ictal pattern with standard EEG settings and then with the prolonged TC.

Results:
Interictal EEG showed generalized and multifocal spikes located at right & left temporal, and left parietal regions.

60 seizures were captured. All seizures were characterized by a generalized electrodecrement at onset. 32 seizures were excluded from further analysis because of movement and/or muscle artifact obscuring the EEG seizure onset.

Out of the remaining 28 seizures, 13 (46%) had a clear DC shift at onset, which was maximum in the bifrontal regions. An underlying paroxysmal fast activity was not always visible because of significant muscle artifact.
Conclusions:

Ictal DC shift can be demonstrated on scalp EEG by increasing the time constant to ≥2 seconds. In generalized epilepsies with electrodecrement at seizure onset, the DC shift is localized to bifrontal regions and can be a useful marker of ictal onset when the underlying paroxysmal fast activity is obscured by muscle artifact.

The presence of DC shift with electrodecrement also provides indirect evidence that electrodecrement is produced by a low voltage paroxysmal fast activity rather than suppression.

10.1212/WNL.0000000000205645