Between-session SEEG band-power shifts exceeded within-session shifts in 7 of 8 patients, with unknown session dates

We compared bipolar stereo-electroencephalography (SEEG) band-power shifts between two recording sessions and between two segments within one session in the OpenNeuro ds005624 1.0.0 cohort. Eight of 24 patients had both sessions. For each session, we analyzed fixed 30-second segments at 60–90 seconds and 150–180 seconds, using adjacent bipolar pairs of good, non-flat contacts and log power in five bands from 4 to 150 Hz. The between-session shift exceeded the within-session shift in 7 of 8 patients (median 0.90 versus 0.60 standardized units; median ratio 1.13, range 0.98–3.00; exact one-sided sign-flip p = 0.0078). Spatial-profile similarity was lower between sessions in 6 of 8 patients (median Pearson r 0.976 versus 0.981; p = 0.031). High-gamma shift was greater between sessions in all 8 patients (p = 0.0039; secondary, uncorrected band test). The direction matches a prior analysis on ds004752, while the effect was smaller. Session dates were unavailable, so between-session comparisons cannot be called between-day comparisons. The study is limited by n = 8, one center, a 90-second within-session gap, and uncontrolled behavioral state. A post-hoc rule removed flat contacts marked good in one patient; the primary result was unchanged without that rule.

stereo-electroencephalography · SEEG · band power · session variability · intracranial EEG · OpenNeuro · epilepsy