Distant SEEG CCEP variation was not greater than sham in 7 of 8 patients

We assessed whether pulse-to-pulse variation in early (10–50 ms) cortico-cortical evoked potentials (CCEPs) was shared across distant response sites during single-pulse SEEG stimulation. The analysis used 1,198 stimulated-pair/channel units from 8 patients in the STReEF dataset (OpenNeuro ds005448 v1.0.0), with 10–20 clinical 1–2 mA pulses per unit. For 9,483 channel pairs on different shafts and at least 20 mm apart, the mean real-pulse correlation was 0.067 (6/8 patient means positive; exact two-sided sign-flip p = 0.0625). The real-minus-sham difference was +0.031 in Fisher z (2/8 positive; p = 0.875); mean sham correlation was 0.036. One patient (STREEF07) showed stronger distant shared variation (r = 0.357 vs 0.020 for sham), associated exploratorily with stimulation on one shaft. Nearby channels showed higher correlations than sham, but the data cannot distinguish volume conduction from local shared gain. In 7 of 8 patients, distant responses were not more shared than sham. Small pulse counts, eight patients, the pre-pulse sham windows, and dependence on archived trial ordering limit interpretation.

SEEG · cortico-cortical evoked potentials · single-pulse stimulation · trial-to-trial variation · shared variability · intracranial electrophysiology · epilepsy