Distance from the stimulation site explains most of the N1 map overlap between SEEG stimulation sites; a small site-specific excess remains at 7.5-15 mm (8 patients, ds005448, binary expert N1 labels)

We measured how much the expert-checked N1 responder sets of two bipolar SEEG stimulation sites overlap (Jaccard J) as a function of site separation. Data: OpenNeuro ds005448 v1.0.0, 8 patients, 1-2 mA single pulses, from the archived table of sci 1f4c1eab. We compared each site pair with a null that shuffles labels inside 5 mm distance bins, which keeps the response-against-distance curve of each site. Median J was 0.35 at <=7.5 mm, 0.28 at 7.5-15 mm, 0.12 at 15-30 mm and near 0 at more than 30 mm. In all patients, mean J was below 0.5 at 5-10 mm. The null predicted most of the overlap. At 7.5-15 mm (primary test), the observed J was more than the null J in 8/8 patients (exact sign-flip p = 0.0078), but the excess was only +0.03 J. With a J < 0.5 limit, 175 of 281 active sites (62%) were mutually distinct. Limits: one center, clinically selected pairs, approximate separation, OR merge of runs and polarities, binary labels.

SEEG · single-pulse electrical stimulation · N1 responses · Jaccard similarity · brain-computer interfaces · stimulation-site separation · distance-matched null