SEEG responses fell during 43.2 Hz alternating-polarity trains, but the S–D contrast was not significant in five patients

We examined whether the polarity-invariant component S = (pos + neg)/2 of stereo-EEG (SEEG) responses declined from the first to the last second of 5-s, 43.2 Hz alternating-polarity stimulation trains more than the polarity-flipping component D = (pos - neg)/2. S was treated as neural and D as artifact under a linear, polarity-symmetric artifact assumption. We analyzed OpenNeuro ds006519 (version 1.0.0); 10 trains in five patients passed pulse quality control. For each train, we calculated the median over selected channels of the log last-second/first-second RMS ratio, then averaged across trains within each patient. The mean S-minus-D contrast was -0.112 (S/D ratio 0.89); four of five patients had a negative contrast, with an exact two-sided sign-flip p = 0.125 (minimum possible p = 0.0625). S fell to 0.88 of its first-second value (three of five patients negative, p = 0.25); D also fell in seconds 2–4. With a 4–15 ms response window, the contrast was -0.058 (p = 0.31). The sample is too small to reach p < 0.05 with fewer than six patients, so these results do not show that there is no effect. The assumed separation also depends on linear, polarity-symmetric artifact. A planned ds005169 replication was not performed.

stereo-EEG · intracranial EEG · alternating-polarity stimulation · artifact separation · neural response · sign-flip test · OpenNeuro ds006519