Abstracts

Rationale: Epileptic seizures can result from imbalances in neural activity, either through excessive excitation or insufficient inhibition. The neurophysiological theory suggests that a crucial imbalance, particularly reduced neuronal inhibition, disrupts the equilibrium between cortical excitation and inhibition (E/I balance) in patients with epilepsy. Previous research shows that features of fast oscillations (FOs) in the gamma frequency band ( >30 Hz), measured with electroencephalography (EEG) and magnetoencephalography (MEG), are related to the GABAergic inhibitory system. In vivo GABA concentrations have been previously correlated positively with gamma frequency in the visual and motor cortices. This suggests that such alterations might reflect an E/I imbalance, indicating neurophysiological changes and a potential biomarker for epilepsy. Here, we aim to record induced and evoked FOs with simultaneous high-density EEG and MEG in response to visual stimuli from children with epilepsy and healthy controls. We hypothesize that epilepsy patients will present with altered visually induced or evoked FOs compared to healthy controls and is related to cortical E/I imbalance.


Methods: We recruited 49 children with epilepsy (mean age: 14.61 years ± 2.97; 26 females) and 47 healthy controls (mean age: 13.46 years ± 2.66; 20 females), with no significant age differences. Participants had simultaneous high-density EEG and MEG recordings with child-friendly visual stimuli (cartoons on flashing checkerboards; 340 trials) (Fig. 1A & B). Data were initially filtered to remove DC and processed with band-pass Butterworth filter (1-100 Hz and 100-300 Hz), followed by artifact inspection. Time windows from -200 to 500 ms post-stimulus were averaged across trials (Fig. 1C&D). Cortical activity in the primary visual cortex (V1) was localized using dynamic Statistical Parametric Mapping and virtual sensors were reconstructed at V1 using Desikan Killiany atlas (Fig. 1E&F). We then computed Time-frequency (TF) analysis maps from evoked and induced activity (Fig. 1G). We employed cluster-based permutation test to compare TF maps between children with epilepsy and healthy controls (p< 0.05).


Results: Statistical analysis of evoked responses derived from MEG showed suppressed gamma and low-ripple band power from 35 Hz to 87 Hz in children with epilepsy from ~20 to 170 ms post-stimulus (p< 0.05) (Fig. 2A). Similarly, EEG revealed suppressed beta and gamma band power from 45 Hz to 61 Hz in the epilepsy group from ~20 to 70 ms post-stimulus (p< 0.05) (Fig. 2B). No significant differences were found in the ripple band (100-300 Hz) or induced responses. Grand-averaged MEG source waveforms (filtered 1-100 Hz) showed higher amplitudes in healthy controls from 50 to 275 ms post-stimulus (p< 0.05, corrected) (Fig. 2C).


Conclusions: Our study enhances understanding of evoked and induced FOs' pathophysiological mechanism. Identifying markers and patterns in these FOs may differentiate normal brain function from epilepsy-related pathological activity.


Funding: RO1NS104116-01A1 by NINDS.