Abstracts

Rationale: Historically, the frequencies believed to be relevant EEG have been limited to less than 100Hz. Recently, multiple studies have implicated higher frequency activities (>100Hz) in both normal brain function and epilepsy. In particular, high frequency oscillations (HFO) in the range of 100-250 Hz and very-high frequency oscillations (VHFO) in the range of 250-500 Hz have been found and are thought to be a marker for epileptogenicity. The objective of this pilot study was to determine the feasibility of recording HFOs and VHFOs in neocortical epilepsy using subdural macroelectrodes. Methods: Three neocortical epilepsy patients were examined with invasive EEG using subdural macroelectrodes at a sampling rate of 2000 Hz without filters. One additional patient was recorded at 1000 Hz sampling rate. 63 macroelectrodes were used in each patient and seizures were recorded, ranging from 1-4 per patient. EEG was manually examined from one minute before ictal onset until one minute after the end of the EEG seizure, and divided into “pre-ictal”, “early ictal”, and “ictal propagation”. A one-minute sample of interictal EEG was also examined to assess the presence of interictal activity. High-pass filters of 100 Hz and 200 Hz were used to identify HFOs, which were then compared to EEG findings with conventional filter settings (LF 1.6, HF 70) to determine their relationship and timing. Results: High frequency oscillations > 100 Hz were found in the "early ictal" and "ictal propagation" periods of all four epilepsy patients. Additionally, discrete HFOs were seen interictally in the same regions of all four patients. Conclusions: In this pilot study, HFOs were found in all four invasive EEGs examined when using a high-sampling rate. All four EEGs displayed HFOs in the early-ictal and ictal-propagation periods, as well as in the interictal periods. Our findings, however, differed in many aspects compared to the findings from other institutions, such as the Montreal Neurological Institute. These differences may be due to recording technique, electrode construction methods, filters, and use of microelectrodes instead of macroelectrodes.