ORCID ID
Graduation Date
Summer 8-14-2026
Document Type
Dissertation
Degree Name
Doctor of Philosophy (PhD)
Programs
Neuroscience
First Advisor
Dr. Stephen V. Gliske
Abstract
Epilepsy is one of the most prevalent neurological diseases worldwide, affecting approximately 50 million people across all ages and socioeconomic groups. Nearly one-third of patients develop drug-resistant epilepsy (DRE), for whom surgical evaluation may identify therapeutic options such as resection, ablation, or neuromodulation. The success of these interventions depends on accurately localizing epileptogenic tissue while preserving functionally essential cortex. Current clinical approaches, including scalp electroencephalography (EEG) and intracranial electroencephalography (iEEG), structural and functional neuroimaging, magnetoencephalography (MEG), and electrical stimulation mapping, provide essential information for the evaluation and treatment planning of epilepsy surgery. However, postoperative seizure freedom remains variable, highlighting the need for analytical approaches that better characterize the spatial, temporal, and physiological context of epileptic and functional brain activity. Neural oscillations, reflecting rhythmic fluctuations of neuronal activity within distinct frequency ranges, provide a physiologically meaningful framework for studying both cortical function and epileptic activity. Oscillatory features such as phase-amplitude coupling (PAC), high-frequency oscillations (HFOs), and interictal spikes (IISs) have been investigated as candidate biomarkers in epilepsy. However, their clinical interpretation remains limited by an incomplete understanding of their spatial specificity, state dependence, and circadian variability. This dissertation contributes to addressing these gaps by applying computational approaches to characterize selected spatial and temporal properties of neural oscillations using noninvasive MEG and long-term iEEG recordings from patients with DRE. Chapter 1 focused on identifying eloquent cortex in patients with DRE using MEG. Specifically, a -based framework was developed and validated to noninvasively localize task-activated somatosensory cortex during median nerve stimulation. In Chapters 2 and 3, we investigated the circadian fluctuations in HFOs and IISs, respectively in long-term data. We observed statistically significant circadian patterns during non-rapid eye movement sleep and awake states in both HFOs and IISs. Together, these findings highlight that neural oscillatory biomarkers encode clinically meaningful information across spatiotemporal scales and the importance of incorporating multi-scale temporal mapping strategies into the planning of definitive surgical interventions. Future studies incorporating such approaches may improve biomarker interpretation, strengthen localization strategies, and contribute to more individualized treatment planning for patients with DRE.
Rights
The author holds the copyright to this work and any reuse or permissions must be obtained from the author directly.
Recommended Citation
Das, Srijita, "The Role of Neural Oscillations as Biomarkers of Epilepsy: A Multimodal Computational Approach" (2026). Theses & Dissertations. 1108.
https://digitalcommons.unmc.edu/etd/1108