Virtual Stimulation of Interictal Stereo-eeg to Localize the Epileptogenic Network
Abstract number :
2.073
Submission category :
3. Neurophysiology / 3G. Computational Analysis & Modeling of EEG
Year :
2022
Submission ID :
2204987
Source :
www.aesnet.org
Presentation date :
12/4/2022 12:00:00 PM
Published date :
Nov 22, 2022, 05:27 AM
Authors :
Rachel June Smith, PhD – University of Alabama at Birmingham; Sophia Zhai, - – Johns Hopkins University; Kristin Gunnarsdottir, PhD – Johns Hopkins University; Daniel Ehrens, PhD – Johns Hopkins University; Adam Li, PhD – Columbia University; Jorge Gonzalez-martinez, MD, PhD – University of Pittsburgh; Sridevi Sarma, PhD – Johns Hopkins University
Rationale: For patients with drug-resistant epilepsy, localizing and surgically treating the epileptogenic network can bring seizure freedom. However, surgical success rates vary from 30% to 70% because no clinical biomarker of epileptogenic nodes currently exist. We hypothesize that epileptogenic nodes act as powerful sources of pathological activity during seizures but are actively inhibited during interictal periods. Thus, although stimuli of various forms enter these regions during interictal times, the network response to these stimuli are subdued. We tested this hypothesis in silico by performing virtual stimulation of stereo EEG (sEEG) channels during interictal periods and used the magnitude of the network response to localize the epileptogenic network.
Methods: We performed virtual stimulation in sEEG data gathered from 34 epilepsy patients that were assessed for clinical outcome after one year. The data were first divided into non-overlapping 500 ms windows, and a linear time-invariant model of the form x(t+1) = Ax(t) was constructed for each window. Then, we added an exogenous perturbation for each channel: x_stim(t+1) = Ax_stim(t)+B(t) where B is a unit vector with the 1 corresponding to the index of the channel being virtually stimulated. Next, we measured the simulated network response, x_stim(t), for 500 time steps, and calculated the L-2 norm (“size”) of this network response. This process was repeated for every channel in the window and then for every window in the dataset, creating a heat map of network responses as a function of virtually stimulating each electrode over time.
Results: We found that in surgical success patients, stimulating regions within the hypothesized epileptogenic network (EN) evoked a smaller network response than stimulating regions outside of the EN in 9/14 success cases, with 5/14 being statistically significant differences in response magnitudes. In failed surgery cases, stimulating the EN resulted in a statistically significantly reduced response in only one case (1/20), but followed a trend of lower EN-stimulated network responses in 9/20 cases. These results support our hypothesis that nodes of the epileptogenic network are being inhibited during interictal periods, producing smaller network responses when stimulated than non-EN regions.
Conclusions: We believe that a better understanding of how the epileptogenic network interacts with surrounding brain regions in both ictal and interictal periods may provide valuable insight into the mechanisms that give rise to seizures. The development of an interictal biomarker of epileptogenic tissue would add information to ictal data gathered from capturing seizures in the hospital while simultaneously improving surgical outcomes.
Funding: Funding came from NIH IRACDA through JHU’s ASPIRE Program (RJS) and R01 NS125897-01 (SVS).
Neurophysiology