Scalp EEG seizure warning in children under a pre-specified, locked evaluation: confound audit, structured chance levels and a negative result

A pre-specified, locked scalp-EEG seizure warning scheme in 21 monitored children detected 19% of evaluable seizures at 0.60 false alarms per hour, not above chance.

• Why it matters: Prior scalp-EEG prediction reports may be inflated by lenient evaluation and by structured monitoring factors like vigilance state, clock time, and admission trends.
• What they did: The authors built a state-conditioned, directional method on seven global EEG features, froze its configuration with a hash-recorded protocol, and tested it with nested leave-one-patient-out cross-validation and then once on two validation cohorts.
• The result: In the initial development cohort it detected 33% of seizures at 0.47 false alarms per hour, but a clock-time/vigilance/days-since-admission predictor reproduced most signal, and validation showed no significant benefit (19% vs 26% chance; p = 0.96).

AI summary of the abstract · may contain errors · not medical advice · read the paper.