Characterization of Epileptic Networks Through the Mapping of Interictal Epileptiform Discharges
Abstract number :
3.155
Submission category :
3. Neurophysiology / 3G. Computational Analysis & Modeling of EEG
Year :
2023
Submission ID :
777
Source :
www.aesnet.org
Presentation date :
12/4/2023 12:00:00 AM
Published date :
Authors :
Presenting Author: Samuel Tomlinson, MD – Hospital of the University of Pennsylvania
Benjamin Kennedy, MD – Division of Neurosurgery – Children's Hospital of Philadelphia; Eric Marsh, MD, PhD – Division of Child Neurology – Children's Hospital of Philadelphia
Rationale: Interictal epileptiform discharges (IEDs) are transient electrographic abnormalities observed in the EEG of patients with epilepsy. IEDs have been shown to propagate between brain regions with millisecond-scale latencies, suggesting that IEDs can activate distributed cortical networks of pathologically inter-related neuronal populations. Conceptualizing IEDs as an output of pathologic networks is novel, and characterizing IED activity utilizing the tools of network science may lend insights into the behavior and organization of the epileptic brain. This study aimed to examine the dynamic characteristics of irritative networks in children with medically-intractable epilepsy in the pursuit of improved models of epilepsy pathogenesis and surgical planning.
Methods: Long-term invasive EEG recordings were acquired using subdural electrodes in seventeen children with medically-refractory epilepsy. Full-duration recordings were segmented into consecutive, non-overlapping 30-minute segments for analysis. A validated, automated detector was used to identify IEDs in each segment. Networks were constructed to represent the co-activation of brain regions (i.e., electrodes) during IEDs. Networks were segregated into cohesive sub-regions (i.e., modules) with high rates of IED co-activation. Patterns of local and global network activity (e.g., activation frequency, propagation trajectories, and flow of IEDs between modules) were examined over time in relation to variables such as proximity to seizure onset, overlap with seizure onset zone, and sleep-wake state.
Results: IED networks segregated into cohesive modules with highly stable activation rates and propagation trajectories. Modules overlapping the seizure-onset zone acted disproportionately as spike “distributors” by transmitting IEDs to other modules. The global topology of IED networks fluctuated with factors such as sleep-wake state and temporal proximity to seizures.
Conclusions: Automated IED analysis yields new, clinically-relevant insights into the complex relationship between IEDs and seizures at the level of pathologic meso-scale cortical networks.
Funding: N/A
Neurophysiology