The polarity-invariant SEEG component decreased during 43.2 Hz trains, but did not differ significantly from the artifact component

We tested whether the polarity-invariant component S decreased more than the polarity-flipping component D during 43.2 Hz alternating-polarity cortical stimulation. The analysis used 8 trains from 5 patients in OpenNeuro ds006519 v1.0.0. S was about 16% lower in the last than the first second, while D was about 4% lower. The primary difference, c = log(S_last/S_first) − log(D_last/D_first), was negative in all 5 patients (mean −0.149; exact two-sided p = 0.0625; Holm-adjusted p = 0.1875), so the pre-specified p < 0.05 rule was not met. Four of 12 trains were excluded because pulse polarity did not reliably alternate; this exclusion was set after inspecting polarity diagnostics. A sensitivity analysis including those trains gave negative c in 7/7 patients (p = 0.0156), but the S/D split was invalid in those trains. S and D changes were equal in the four trains from one patient, and S amplitude correlated with D across channels (mean rho = 0.61), so residual artifact in S cannot be excluded. With only five patients, the smallest possible exact two-sided p was 0.0625. Amplifier recovery may affect the 2–15 ms measurement window; four patients contributed one train each.

intracranial EEG · SEEG · cortical stimulation · alternating polarity · stimulation artifact · neural adaptation