Abstracts

HIGH FREQUENCY OSCILLATIONS DEVIATE FROM NEURONAL AVALANCHE DYNAMICS IN THE ICTAL ONSET ZONE

Abstract number : 3.064
Submission category : 1. Translational Research: 1E. Biomarkers
Year : 2014
Submission ID : 1868512
Source : www.aesnet.org
Presentation date : 12/6/2014 12:00:00 AM
Published date : Sep 29, 2014, 05:33 AM

Authors :
Elakkat Gireesh, Christophe Jouny and Nathan Crone

Rationale: High frequency oscillations (HFOs) are considered to be a marker for the ictal onset zone. At the same time HFOs play a role in normal neurophysiological processes. In this study we attempted to differentiate HFO activity arising from normal neuronal networks, from that arising from the ictal onset zone. We used a previously described approach for analyzing spatio-temporal clusters in the form of neuronal avalanches for this purpose. One of the benchmarks of critical state dynamics as described by neuronal avalanches is a power law distribution of event sizes, with an exponent of -1.5. We evaluated these dynamics in the current study using spatial-temporal clusters of HFOs recorded with subdural electrocorticography (ECoG). Methods: We recorded ECoG (10 kHz sampling) in 3 patients on the third to fourth day of invasive monitoring with 8 x 8 grids. The data was analyzed using custom Matlab code. Signals from a 1000-sec interval were band pass filtered (100-500 Hz), and high frequency power was estimated as the absolute value of the analytical signal (Hilbert transform). Peaks in power (> 2.5 S.D., power distribution of the entire interval) were then used to calculate neuronal avalanche dynamics. Spatial-temporal clusters were identified as contiguous 100-msec time bins with power peaks in at least one electrode, and cluster sizes were estimated by summing the power of peaks in all electrodes in the cluster. Results: The initial size distribution of spatiotemporal clusters in all patients showed a power law, with slope between -1.4 and -1.7. For two patients in whom the 8 x8 grid was only minimally involved at the onset of ictal activity, a mild increase was seen in the probability of large sized spatial-temporal clusters. In a different patient in whom there was significant involvement of the grid at ictal onset, there was a clear increase in the probability of large sized spatiotemporal clusters. Conclusions: Our preliminary analyses suggest that the spatial-temporal patterns of HFOs arising in cortical networks involved at ictal onset demonstrate super critical dynamics. In contrast, these patterns in cortical regions less involved at ictal onset exhibit network dynamics that are closer to those of neuronal avalanche dynamics.
Translational Research