Abstracts

Rationale: Knowing where sources of electroencephalography (EEG) signals are located in the brain can help with diagnosis. High frequency oscillations (HFOs) have been studied as precise biomarkers of the seizure onset zone (SOZ) in epilepsy, and the resection of cortex based on HFO activity has been associated with a positive therapeutic outcome. However, since this activity stems from polyphasic oscillations of small, focal subsets of neurons, it is difficult to detect in conventional, noninvasive EEG. Tri-polar EEG (tEEG) senses high frequency activity (HFA) better than conventional noninvasive methods (1). Its increased spatial resolution allows for noninvasive recording of weak HFA, which is blurred by the spatial filtering properties of the skull and soft tissue between the cortex and scalp. In the present study, tEEG was used to non-invasively detect and localize HFA in epilepsy patients. Methods: Five subjects diagnosed with refractory epilepsy were monitored concurrently with conventional disc EEG electrodes and tri-polar concentric ring electrodes (TCREs) during periods of ictal and inter-ictal activity.  The disc electrodes were placed at standard 10-20 locations with TCREs directly behind them. A band-pass filter that was determined during post-processing to be specific to each subject’s HFA range was used to isolate their narrow-band, high-power activity. Surface potentials of HFA were localized on a case-by-case basis using distributed source methods on subject-specific, realistic head models, reconstructed from T1 weighted magnetic resonance imaging (MRI) data using FreeSurfer (n=4). Where MRI data was not available (n=1), the ICBM152 template head model [2], derived from a non-linear average of MRI scans of the 152 subjects in the MNI152 database, was used. In the open-source data analysis application Brainstorm [3], a linear L2-minimum norm estimates algorithm was used to localize sources to a source space constrained normal to the cortical surface.  Results: HFA was detected in the tEEG of each subject, but not in the EEG. The source regions of this activity in tEEG of the four subjects for whom head models were reconstructed with MRI fell within regions previously defined as seizure onset zones (SOZs) or irritative zones (IZs) independently by three neurologists. The HFA localization on the patient with the template head model was located on the posterior edge of the SOZ diagnosis. While close, this slight discrepancy can be due to the head model differences between template and subject anatomies or that there are strong paths of communication between the prefrontal and parietal regions in this patient. SOZ/IZ diagnosis was performed on video EEG alone; tEEG was not made available to the clinicians. Conclusions: HFA is difficult to detect with conventional EEG and cannot be localized. Tri-polar EEG allows for the detection and localization of HFA, using a noninvasive measurement of brain activity. The localization appears to be useful for precise identification of the SOZ or IZ in epilepsy patients.References:Besio W., Martinez I., et al., “High-Frequency Oscillations Recorded on the Scalp of Patients with Epilepsy Using Tripolar Concentric Ring Electrodes“, IEEE J Trans Eng in Health and Med, 2013.Fonov VS, et al. Unbiased average age-appropriate atlases for pediatric studies, NeuroImage, 54, 1, 2011.Tadel F, et al. Brainstorm: A User-Friendly Application for MEG/EEG Analysis Computational Intelligence and Neuroscience, vol. 2011 Funding: Science, Mathematics, and Research for Transformation (SMART) scholarship from American Society for Engineering Education (ASEE) funds C. Toole's education. Part of this work was supported by the National Science Foundation grants 1539068 and 1430833 to W. Besio.