Daily variation in an intracranial EEG feature in humans detected by a responsive neurostimulator system

Abstract

Purpose: Based on the observation that epileptic seizures can occur at specific times of the day, we looked for daily variation in an intracranial electrographic feature used by a responsive neurostimulator system to detect seizures.

Methods: A computationally efficient measure of intracranial EEG energy or complexity, the line length baseline, was calculated and reported by an external responsive neurostimulator during a clinical trial of device safety. Data were obtained from 24 consecutive patients with medically intractable epilepsy undergoing intracranial monitoring over 2 to 54 days to localize the seizure onset zone. Measurements from individual subjects made at different times of day over many days were displayed on a single 24-h cycle and fit with a cosine function to characterize the time of the maximum value. The timing of epileptic seizures was also noted.

Results: The time of the maximum line length baseline value had a bimodal distribution with relative peaks at 05:30 and 15:00 hours. The time of the maximum value did not associate with specific brain regions, except that a nocturnal peak was not measured from temporal neocortex. The temporal distribution of maximum values was similar to the timing of epileptic seizures.

Conclusion: The line length baseline feature of the intracranial EEG shows daily variation with location specific characteristics within individual subjects.