Abstracts

Rationale: In a recent study, we analyzed correlations in slow amplitude modulations of the intracranial EEG (icEEG) band power time-series (i.e., the second spectrum), for evidence of relationships between and within brain regions and their correspondence with fMRI-defined resting state networks. Although we reported a lack of support for the fMRI-based default mode network (DMN) in the icEEG second spectrum, our results and other studies suggest these envelope correlations may form the basis for distant spatial coupling in the brain. In our prior analysis of resting data, we found that second spectrum relationship decreased with greater intercontact distance, and that relationships were highest in the delta band and decreased with increasing frequency [1]. In this study, we aimed to determine how these second spectrum relationships are affected by antiepileptic drug (AED) taper, day/night cycles, and seizures. Methods: We studied icEEG data collected from 13 medically refractory adult epilepsy patients who underwent monitoring and seizure localization at Yale-New Haven Hospital. We estimated magnitude-squared coherence (MSC) below 0.15 Hz of the power time-series in the delta, theta, alpha, beta, and gamma bands for all electrode contact pairs to quantify slow envelope correlations between them. We computed this on hour-long background icEEG epochs before and after AED taper when patients appeared to be resting quietly with eyes open based on video-icEEG. We also calculated second spectrum MSC over the entire record for each patient and examined variations related to time of day. Finally, for each patient, we reviewed the clinical record to find any clinical seizures that were 6 hours removed from any other seizures, and analyzed the icEEG before and after these seizures using our second spectrum measures. Results: There was a small, but significant increase in second spectrum MSC with AED taper. The second spectrum MSC increased on average during the night, and decreased during the day. In our seizure-related analyses, we studied 61 seizures across all patients, of which 28 occurred during the day and 33 occurred at night. Interestingly, second spectrum MSC was significantly increased in all frequency bands except theta for hours before and after seizure, as compared to background icEEG epochs. These changes occurred for both daytime and nighttime seizures. Conclusions: Our analysis indicates that the second spectrum MSC faithfully captures essential changes in the icEEG during monitoring, and we expect these metrics could aid in understanding processes underlying seizure generation and in developing robust seizure prediction algorithms.[1] RB Joshi, N Gaspard, II Goncharova, RB Duckrow, D Duncan, JL Gerrard, DD Spencer, LJ Hirsch, HP Zaveri. “Regional and network relationship in the intracranial EEG second spectrum.” Clinical Neurophysiology, 2016. Funding: We wish to acknowledge the support of the C.G. Swebilius Trust.