Ictal patient data.

Objective Approximately 50 million people worldwide have epilepsy and 8–17% of the deaths in patients with epilepsy are attributed to sudden unexpected death in epilepsy (SUDEP). The goal of the present work was to establish a biomarker for SUDEP so that preventive treatment can be instituted. Approach Seizure activity in patients with SUDEP and non-SUDEP was analyzed, specifically, the scalp EEG extracted muscle activity (SMA) and the average wavelet phase coherence (WPC) during seizures was computed for two frequency ranges (1–12 Hz, 13–30 Hz) to identify differences between the two groups. Main results Ictal SMA in SUDEP patients showed a statistically higher average WPC value when compared to non-SUDEP patients for both frequency ranges. Area under curve for a cross-validated logistic classifier was 81%. Significance Average WPC of ictal SMA is a candidate biomarker for early detection of SUDEP.