Scalp ripple rates for rapid epilepsy differentiation and seizure activity assessment: Applicability and influential factors

Objective We aim to determine whether automatically detected ripple rate (ADRR) of 10‐min scalp electroencephalography (EEG) during slow‐wave sleep can be a useful tool for rapid epilepsy differentiation and seizure activity assessment, and we analyze the clinical factors that may affect the scalp ripple rates. Methods We retrospectively included 336 patients who underwent long‐term video‐EEG with a sampling rate ≥1000 Hz, and three groups were established based on their final clinical diagnosis (non‐epilepsy; non‐active epilepsy [epilepsy being seizure‐free for at least 1 year]; and active epilepsy [epilepsy with one or more seizures in the past year]). ADRRs between groups were compared and diagnostic thresholds set according to the maximum of Youden index with the receiver‐operating characteristic curve. Results The 336 patients comprised 49 non‐epilepsy and 287 epilepsy patients (95 non‐active epilepsy and 192 active epilepsy). The median ADRR of the epilepsy group was significantly greater than in the non‐epilepsy group, with a diagnostic threshold of 4.25 /min (specificity 89.8%, sensitivity 47.74%, p