Abstracts

Rationale: High-frequency oscillations (HFOs), including ripples (R, 80-250 Hz) and fast ripples (FR, 250-500 Hz), are promising biomarkers of epileptic brain, and present on both electrocorticography (ECoG) and scalp EEG. However, the clinical utility of HFOs is limited by the lack of a standardized approach to identification, especially on scalp EEG. We set out to determine whether independent board-certified pediatric electroencephalographers can reliably identify HFOs and conventional epileptiform discharges on both ECoG and scalp EEG.  Methods: Two blinded pediatric electroencephalographers experienced in visual analysis of HFOs independently reviewed 10 scalp EEG samples from patients with a history of epilepsy, brain tumor, or autism, as well as 10 intraoperative ECoG samples from patients who underwent epilepsy surgery. Each sample was 10 seconds in duration and included only non-REM sleep (if scalp EEG). The study EEG/ECoG recordings were acquired with a digital sampling frequency of 2000 Hz. Using Persyst EEG reviewing software, HFOs were marked using bandpass filter between 80-250 Hz (R) and 250-500 Hz (FR), with finite impulse range (FIR) filter settings with time scale of 338 mm/sec. HFOs were defined as oscillatory events with at least four cycles, which are clearly visible above the background signal in the filtered data. For each sample, each rater identified all epileptiform discharges (ED, including spikes, polyspikes, and paroxysmal fast activity), R, and FR. The presence/absence and rate of each finding was tabulated. Kappa (κ) statistics were used to quantify inter-rater reliability (IRR) for the identification (presence/absence) of each phenomenon. Intraclass correlation coefficients (ICC, 2-way mixed-effects model) were calculated to assess the consistency of reported rates. As a rule of thumb, κ and ICC > 0.7 is adequate, and κ and ICC > 0.9 is excellent. Results: There was agreement as to the presence or absence of ED, R, and FR in 17, 18, and 18 cases, respectively, out of 20. IRR was favorable, with κ = 0.70, 0.80, and 0.80, respectively, and similar for ECoG and scalp EEG studies. However, with regard to reported rates, ICC for R (0.99, 95%CI 0.96 – 0.99) was statistically superior to FR (0.77, 95%CI 0.41 – 0.91), and both HFO subtypes were far superior in comparison to EDs (0.37, 95%CI -0.60 – 0.75), for which reliability was deemed poor. Conclusions: We have demonstrated favorable inter-rater reliability in the visual identification and quantification of R and FR on both ECoG and scalp EEG recordings. Importantly, this study suggests that HFOs are more reliably identified and quantified than conventional ED such as spikes. Of importance to both clinicians and researchers, our findings support the feasibility of utilizing HFO data in both research and clinical arenas. Funding: None