Oscillations over 500Hz are Potentially More Specific for SOZ Localization
Abstract number :
1.104
Submission category :
2. Translational Research / 2C. Biomarkers
Year :
2021
Submission ID :
1826429
Source :
www.aesnet.org
Presentation date :
12/4/2021 12:00:00 PM
Published date :
Nov 22, 2021, 06:54 AM
Authors :
Vojtech Travnicek, MSc - The International Clinical Research Center of St. Anne's University Hospital in Brno, Czech republic; Pavel Jurak - Institute of Scientific Instruments, The Czech Academy of Sciences, Brno, Czech Republic; Petr Klimes - Institute of Scientific Instruments, The Czech Academy of Sciences, Brno, Czech Republic; Jan Cimbalnik - International Clinical Research Center, St. Anne’s University Hospital, Brno, Czech Republic; Martin Pail - Brno Epilepsy Center, Department of Neurology, St. Anne’s University Hospital, Faculty of Medicine, Masaryk University, Brno, Czech Republic; Milan Brazdil - Behavioral and Social Neuroscience Research Group, CEITEC Central European Institute of Technology, Masaryk University, Brno, Czech Republic
Rationale: Spikes, Ripples and Fast ripples were intensively studied for the last decades as the most promising biomarkers of the epileptogenic zone. With the increasing sampling frequency of EEG recordings, it is possible to analyze oscillations on higher frequencies. The aim of this study is to compare oscillations in different frequency bands (5-2000Hz) in terms of their ability to localize the epileptogenic zone.
Methods: We analyzed pre-surgical Stereo-EEG (SEEG) recordings from 37 patients with drug-resistant epilepsy, recorded at the Brno Epilepsy Center (St Anne’s Hospital) between March 2012 and May 2015 which underwent resective surgery. SEEG data were recorded with a sampling rate of 25kHz further downsampled to 5kHz and 30-minute resting recordings were analyzed for each patient. For the purpose of this analysis, we use bipolar montages for its capability to reduce artifacts. We computed power envelopes in 5 consecutive frequency bands (5-80Hz, 80-200Hz, 200-500Hz, 500-1000Hz and 1000-2000Hz) and manually inspected them for artifacts; artificial channels and artificial time segments were omitted from further analysis. With our custom peak detection algorithm based on peak prominence, we detected significant peaks reflecting oscillations in a given frequency band. Ten oscillations and more per 30-minutes record was the condition for claiming the channel for Oscillatory.
Patients were divided into two groups, one with a good outcome (Engel 1A), the second one with a bad outcome (worse than 1A). For each group and every frequency band, the percentage of resected Oscillatory channels was calculated. We used the Wilcoxon sign-rank test to evaluate the difference between groups.
Results: Each frequency band showed a significant difference (p< 0.05) between good outcome patients and bad outcome patients. While removing only 47±20% of oscillatory channels in the Ripple frequency band (80-200Hz) provided a good outcome, in frequency band 500-1000Hz it had to be 76±31%, which makes them more specific markers in the scale of this analysis.
Translational Research