Entropy-based complexity of EEG microstates differentiates ADHD comorbidity in children with SeLECTS

Self-limited epilepsy with centrotemporal spikes (SeLECTS) is frequently comorbid with attention-deficit/hyperactivity disorder (ADHD), yet objective neurophysiological markers for this comorbidity remain limited. In this work, we combined EEG microstate analysis with entropy-based complexity measures to characterize brain network dynamics and nonlinear signal irregularity during wakefulness and sleep. Resting-state 19-channel scalp EEG was recorded from 58 children with SeLECTS, including 38 in the SeLECTS+ADHD group and 20 in the SeLECTS-alone group. Four microstates were identified, and their temporal properties were quantified using time coverage and mean duration. Within each microstate, we computed global multivariate permutation entropy and assessed regional complexity using channel-level Shannon entropy and generalized entropy. Conventional microstate temporal parameters showed robust wake–sleep modulation in both groups but limited between-group discrimination. In contrast, microstate-resolved entropy measures showed clearer group differences, particularly during sleep, indicating reduced global complexity and localized regional alterations in the SeLECTS+ADHD group. Overall, these findings highlight a dissociation between microstate temporal dynamics and complexity features and suggest that sleep enhances the sensitivity of EEG complexity markers to ADHD comorbidity in SeLECTS. This integrated framework may provide candidate noninvasive biomarkers to support more objective identification of ADHD comorbidity in SeLECTS, pending validation in larger independent cohorts.