Abstracts

The Seizure Onset Zone Connectivity Analysis

Abstract number : 2.146
Submission category : 3. Neurophysiology / 3G. Computational Analysis & Modeling of EEG
Year : 2017
Submission ID : 349644
Source : www.aesnet.org
Presentation date : 12/3/2017 3:07:12 PM
Published date : Nov 20, 2017, 11:02 AM

Authors :
Petr Klimes, International Clinical Research Center, St. Anne’s University Hospital, Brno, Czech Republic; Mayo Systems Electrophysiology Laboratory, Department of Neurology, Mayo Clinic, Rochester MN, USA; Jan Cimbalnik, Mayo Clinic ;International

Rationale: The exact localization of the seizure onset zone (SOZ) is crucial for invasive treatment of focal epilepsy. Localization of the SOZ in intracranial EEG recordings can be done by detection and spatial mapping of abnormal interictal epileptiform discharges and pathological high-frequency oscillations. More recent studies using bivariate measures to map SOZ connectivity show its functional isolation from non-SOZ tissue. Furthermore, bivariate measures allow us to analyse the edges of the SOZ. Here we present further analysis of these phenomena on extended number of patients, which may prove useful for precise SOZ localization. Methods: 19 patients, implanted with intracranial depth electrodes (10-15 contacts per electrode, inter-contact distance 1.5 mm), sampled at 5 kHz. 30 minutes of relax were analysed. Localization of electrodes was achieved by co-registration of pre- and post-implant MRI. The SOZ electrodes were determined from the clinical report.All channels were re-montaged to an average signal from white matter, and filtered into 8 frequency bands (delta 1-4 Hz, theta 4-8 Hz, alpha 8-12 Hz, beta 12-20 Hz, low gamma 20-45, high gamma 55-80, ripples 80-250 Hz, fast ripples 250-600 Hz). Subsequently, we analysed linear correlation, relative entropy and phase synchronization in all adjacent, i.e. near neighbor, contacts. Contact pairs were divided into 3 groups: SOZ (both contacts inside SOZ), non-SOZ (both outside SOZ) and BRIDGE (one inside, one outside SOZ). Differences between groups were computed by Wilcoxon rank-sum statistics. Results with p < 0.01 were considered statistically significant. Results: Linear correlation detected statistically significant differences between SOZ and non-SOZ in all frequency bands, except fast ripples. Differences between SOZ and BRIDGE were detected in fast ripples. Significant differences for non-SOZ and BRIDGE were found in low and high gamma bands.Relative entropy exhibited differences between SOZ and non-SOZ in all bands. Significant differences between SOZ and BRIDGE were detected in fast ripples. Significant differences for non-SOZ and BRIDGE were found in theta and beta band.Phase synchrony detected differences between SOZ and non-SOZ in all bands, except fast ripples. Differences between SOZ and BRIDGE were detected in fast ripples only. For non-SOZ and BRIDGE were significant differences found in beta, low gamma and high gamma. Conclusions: The results show clear evidence for reduced functional connection between SOZ and surrounding brain. Furthermore, we can speculate about increased activity between SOZ and non-SOZ regions in fast ripple band, which might play a role in seizure spread and may be a clinically useful signature of ictogenesis in epileptic brain. Further analysis of these principles may prove effective for SOZ localization and understanding seizure generation. Funding: Supported by Ministry of Education, Youth and Sports of the Czech Republic project no. LH15047 (KONTAKT II), International Clinical Research Center (FNUSA-ICRC: Project no. LQ1605 (MEYS CR, NPU II)). NIH R01-NS92882 and NIH UH2NS095495-01.
Neurophysiology