In two SEEG patients, 2 mA stimulation was associated with more N1 responses than 1 mA after distance adjustment

Expert-checked N1 labels from OpenNeuro ds005448 v1.0.0 were used to examine whether the greater N1 response rate at 2 mA previously observed in STREEF08 could be explained by run order. Among five patients with two checked runs, only STREEF08 repeated the same stimulation sites (22 of 22); the other patients shared zero or one site, so the dataset could not provide a same-current, same-site control for change over time. A distance-adjusted logistic model in STREEF08 and STREEF12 estimated higher odds of N1 responses at 2 mA (pooled odds ratio 4.4, 95% cluster-bootstrap CI 2.6–9.1; 99 stimulation pairs and 8,599 contact rows). The estimate was larger in STREEF08, where 1 mA was first (OR 8.9, CI 4.1–19.9), and smaller and not significant in STREEF12, where 2 mA was first (OR 1.6, CI 0.7–5.2; only seven 1 mA pairs). In STREEF08 run 031053, with both currents in one run but at different sites, mean pair N1 rate below 30 mm was 0.05 at 1 mA (12 pairs) and 0.34 at 2 mA (22 pairs; two-sided permutation p=1e-4). This within-run result shows that run order alone does not explain the observed current difference, but it remains a between-site comparison. The analysis includes only two patients for the current model; stimulation-site differences were large, and bootstrap intervals do not include uncertainty in expert labels. The findings therefore do not isolate a causal current effect.

SEEG · single-pulse electrical stimulation · N1 response · stimulation current · run order · distance-adjusted model · epilepsy