Abstracts

Rationale: Better understanding of neurocircuitry that underlies the bidirectional relationship between emotion disorders and epilepsies extends far beyond the treatment of each condition alone. As a critical extension to this literature, sEEG allows for capturing finer spatio-temporal transitions within networks compared to other modalities (e.g., fMRI). Thus, it provides potential for new hypothesis testing and therapy development for both disorders. Our objective was to establish an initial approach to assessing amygdala fear responses and potential relationships to emotion disorder symptoms. We hypothesized the responses to fearful faces would be positively correlated to anxiety measures.

Methods: A total of 20 patients undergoing sEEG monitoring (2048Hz) were tested. Patients completed Perceived Stress Scale (PSS-10) and Profile of Mood States (POMS). In the Emotional Faces Task, 120 total static face images were presented (2s each) on a laptop monitor counterbalanced for race, sex, and emotional expression (Happy, Sad, Fearful, Neutral). Participants were asked to indicate male or female faces using a keyboard. During a post-test rating task, participants indicated perceived emotion of each face (Happy, Sad, Fearful, Neutral, Unknown). All task events were presented and recorded using E-prime (v2) and Arduino microcontroller.

Results: Electrodes were localized in native space by identifying the center of the visible artefact on CT within T1-weighted MRI using iElectrodes (Figure 1a). Location of the electrodes within bilateral amygdalae was confirmed by co-registering the MRI and CT to MNI atlas (Destrieux, et al., 2010). Participants that failed to meet inclusionary criteria for amygdala electrodes (n=2), poor task accuracy (n=3), or unusable data (n=8) were excluded from analyses. We used R Analysis and Visualization of iEEG (RAVE) to process sEEG recordings using a standard pipeline. A notch filter at 60Hz was applied to raw sEEG data before downsampling to 200Hz. Faulty electrodes and trial outliers were inspected and removed if present. The trial analysis window (0-1s) was established within the gamma band (50-150Hz) based on visual inspection of trial progressions (Figure 1b) and prior findings (Sato et al., 2011). Comparisons collapsed across left and right amygdala revealed expected increased Fearful versus Neutral responses (t[65]=2.18, p< .05; Figure 1c). Subject mean differential amygdala responses (Fearful >Neutral), collapsed across electrodes, were compared to PSS-10 and POMS Anxiety subscale measures (Figure1d). Pearson correlations revealed a significant positive relationship between amygdala responses and POMS Anxiety (r=0.68, p< 0.05) but not PSS-10 (r=0.59, p=0.08).