Network characteristics in benign epilepsy with centro-temporal spikes patients indicating defective connectivity during spindle sleep: A partial directed coherence study of EEG signals

Abstract

Objective: To investigate the changes in EEG connectivity in children with the typical presentation of benign epilepsy with centro-temporal spikes (BECTS).

Methods: We compared awake and spindle-sleep EEG recordings obtained by a standard electrode array in patients with lateralised (10 Right, 9 Left-BECTS) or bilateral spikes (10 MF-BECTS) and in 17 age-matched controls. We analysed EEG activity using partial directed coherence, an estimator of connectivity based on the multivariate autoregressive models and calculated in- and out-degrees, strength, clustering coefficient and betweenness centrality.

Results: In comparison with the controls, the awake EEG recordings of the patients with lateralised BECTS showed a minimal increase in out-degrees on F4 and F3. The greater differences, found during sleep, included significant reductions in both in- and out-degrees and strength in all of the patient groups, but in T4 or T3 showing increased out-degrees and strength in Right and Left-BECTS. Betweenness centrality was significantly reduced on C3 and C4 in the patients with MF-BECTS.

Conclusions: Our observations suggest that the main finding in BECTS patients is widely reduced local connectivity.

Significance: The network changes in BECTS can be interpreted as a permissive condition occurring in a developmental window that predisposes to seizure generation during spindle-sleep.

Keywords: Benign childhood epilepsy; Effective connectivity; Partial directed coherence.