Abstracts

Rationale: High-frequency oscillations (HFOs) have been described as interictal intracranial EEG (iEEG) biomarkers for localization of seizure-generating brain tissue. Previously reported results show that HFO rates vary significantly in different brain regions as published in Neurology 2018;90:e639-e646. Some evidence exists that oscillations in different frequency ranges (gamma, ripple, and fast ripple) may have different clinical and functional significance (ref: Progress in neurobiology 2012;98:250-264). For example, gamma oscillations may play a role in learning and memory while fast ripples may be more likely to indicate epileptogenicity. The purpose of this study is to examine the spatial distribution of HFOs with different frequency characteristics within the brain. Methods: This study reports HFO rates separated into three partially overlapping frequency bands: gamma (44 – 120 Hz), ripple (73 – 197 Hz) and fast ripple (197 – 536 Hz) measured in 70 patients (3,693 recorded channels and 140 hours of total recording) with intractable epilepsy who underwent wide-bandwidth iEEG as part of clinical presurgical assessment. Discrete HFOs were identified with an automated detector (ref: J Neurosci Methods. 2018 Jan 1;293:6-16). in two-hour segments of high quality recordings likely to represent sleep. Electrode locations were determined by coregistering the patient’s preoperative MRI with an x-ray CT image acquired immediately after electrode implantation. Electrode localization in standard MNI atlas space was performed using iELVis (ref: Journal of Neuroscience Methods 2017;281:40-48) and anatomic locations for each electrode provided by the Desikan-Killiany atlas incorporated in FreeSurfer (ref: Neuroimage 2006;31:968-980). HFO rates were measured in seizure onset zone (SOZ) and non-seizure onset zone (non-SOZ) as determined from the patient’s clinical iEEG study by one or more certified epileptologists. Statistical testing was performed using the Benjamini - Hockberg false discovery rate correction for multiple comparisons. Results: The distribution of HFO in different frequency band are shown in figures 1 (A), (B) and (C). Preliminary results showed that HFO rates to be significantly higher in SOZ than non-SOZ in aggregate, and individually in frontal, parietal, temporal, parahippocampus and occipital lobes (p<0.01) in all three frequency ranges. However, the differences between SOZ and non-SOZ HFO rates in hippocampus and amygdala structures were not statistically significant. Conclusions: Rates of HFOs vary significantly with the anatomic location of electrodes, SOZ/non-SOZ as well as different frequency bands. The variation in SOZ and non-SOZ HFO rates for different brain regions highlights the importance of assessing HFO rates in the context of the anatomic locations of electrodes as well as different bands of frequency. Funding: This study was supported by the Mayo Clinic, the National Institute of Health, and a gift from Mr. and Mrs. David Hawk.