Authors :
Presenting Author: Trisha Mendoza, PhD – University of California, Irvine
Marco Pinto-Orellana, PhD – University of California Irvine
Joffre Oyala, MD – Childrens Hospital of Orange County
Hernando Ombao, PhD – King Abdullah University of science and technology
Daniel Shrey, MD – Childrens Hospital of Orange County
Beth A. Lopour, PhD – Department of Biomedical Engineering, University of California, Irvine, Irvine, CA, United States
Rationale:
Seizures are often characterized by what appear to be highly synchronous epileptiform discharges, yet closer analysis reveals small time delays indicative of fast traveling waves. One hypothesized source of these waves is the ictal wavefront: a slowly advancing front observed at the microscale that is associated with seizure onset and spread. Attempts to extend this analysis to the macroscale have produced conflicting theories, with studies also suggesting static or moving radial sources (Schlafly et al., 2022). Each study used different methods to characterize these sources, likely contributing to the lack of consensus. To address this, we evaluated how different methodological choices affect wave direction estimates, aiming to identify the most robust approach. We applied the resulting framework to human intracranial EEG (iEEG) to establish evidence supporting one or more source theories.
Methods:
We first simulated iEEG data using a second-order autoregressive model. Coherence was modeled as the weighted sum of all signals, with weights decreasing as the distance between electrodes increased. We created templates of ictal wave patterns seen in human data, e.g., planar, linear, and radial waves. The associated time delays were applied to the simulated data with various wave speeds. We then tested our ability to accurately measure the wave pattern while varying the referencing scheme (common average, distant single electrode, and corner electrode), sampling frequency, and spatial resolution. For each condition, we quantified the discrepancy between the template and the estimated wave pattern. We repeated this analysis on human iEEG recordings to validate the methodological recommendations and to robustly characterize ictal traveling waves.
Results:
Simulations revealed that wave patterns could be measured most accurately for wave speeds < 1000 mm/s and average iEEG coherence levels >0.4. Under these conditions, a corner electrode reference outperformed other referencing schemes. Waves could be measured accurately for all sampling frequencies tested, but electrode spacing significantly influenced wave direction estimates. Accurate results were obtained at 3 mm spacing, but not at 9 mm spacing, which approximates standard clinical subdural grids (p < 0.01). When applied to human iEEG, the results were congruent with the simulations (Figure 1). Very few wave patterns were detected using a common average reference (Figure 1A). In contrast, the corner and single electrode references (Figures 1B,C) successfully detected complex patterns during periods of low to moderate (≤800 mm/s) wave speeds and coherence levels >0.5 (Figure 1D,E). In these conditions, patterns such as sources, sinks, and spirals were observed, whereas high wave speeds (≥ 900 mm/s) were associated predominantly with planar waves.
Conclusions:
Detailed analyses of both simulated and human iEEG data highlight critical methodological choices that must be considered when characterizing propagating seizure waves. Our findings provide a validated framework to increase the rigor of future studies.
Funding: NIH Diversity Supplement