Abstracts

Rationale: To determine if distinguishable spatial patterns are present in EEG power, temporal correlation (LRTC) of phase, synchronization index (SI), and in phase clustering activity, derived from 256-channel high density (dEEG) EEG data, and if these patterns correspond to the proven epileptogenic zone, as documented from invasive ECoG grid recordings. The analyses were performed in EEG frequency bands of 80-150 Hz and 150-250 HzMethods: We studied one patient with refractory epilepsy who underwent intracranial EEG to establish the localization of ictal onsets, which proved to be in left parietal-midline cortical regions. Prior to invasive EEG studies the subject underwent dense array, 256 channel EEG recordings, with a sampling rate of 1000 Hz. Four minutes of interictal dEEG segments were randomly selected for analysis. The selected segments were separated in time at least two hours away from an epileptic seizure and, based on visual analysis, were free of interictal epileptiform patterns. Data were imported into MATLAB for analysis. The EEG data was filtered in the appropriate EEG band. Hilbert transform was applied to compute the analytic power and phase, and phase was unwrapped. The SI between a pair of channels was computed with Shannon entropy function. A detrended fluctuation analysis was performed to find LRTC of SI. Spatial plots of LRTC of SI were constructed using a montage layout of 256 electrode positions. Spatiotemporal contour plots of the unwrapped phase with 1.0 ms intervals were also constructed which exhibited dynamic formation of phase cones which are similar to bubbles in boiling water. Several criterions were applied to select stable phase cone patterns. These included: (1) phase frequency was within the temporal band, e.g., 80-150 Hz, (2) sign of spatial gradient did not change for at least 3 time samples, and (3) The frame velocity should be within the range of propagation velocities of cortical axons, 1-10 meters/sec. The stable cluster of frames in each second long dEEG data were computed and averaged over a 4 min period. Spatial plots of averaged rate of phase clusters were constructedResults: LRTC of SI and stable phase cluster patterns were higher in the seizure-onset zone, as compared with the nearby and surrounding brain areas. Refer to Fig. 1 for 80-150 Hz results. Similar patterns were observed in 150-250 Hz band also. However, plots of analytic power alone did not show any reliable and discernible patterns in the seizure area for both of EEG ripple bands.Conclusions: These results show that LRTC of SI and/or phase clustering patterns in EEG ripple bands from 256-channel scalp EEG data closely correspond to the proven seizure-onset zone. The data support our hypothesize that, given adequate temporal and spatial resolution, useful information can be derived from scalp EEG in the 80-250 Hz range, and that this information has the potential to assist in localizing the seizure onset zone noninvasely, even in the absence of the classical markers of epileptiform activity. Additional studies with more patients are underway.