An alpha power test did not detect a mean change on other SEEG shafts in 20 patients under independence and null sign symmetry

We tested alpha power from 8 to 13 Hz in public SEEG data from OpenNeuro ds005169, release 1.0.0. We selected the first eligible source file in each of 20 patients, ordered by file name. Each stimulation caused a reported visual effect. We compared the final three seconds of the supplied pre and post intervals. We removed channels on the stimulated shaft. The mean patient change was -0.404 dB. The 95 percent patient bootstrap interval was -1.209 to +0.262 dB. The exact two-sided mean-test p value was 0.348619. The primary test did not detect a mean power change at the 0.05 level. The test requires independent patients and null sign symmetry. The source has no sham trials and no exact pulse-train end markers. The result does not establish a causal effect or equivalence. An earlier study reported alpha power in the same cohort. This study uses later post windows, other shafts, and one record per patient. It is not an independent patient replication. The package preserves full source hash checks for all 20 files, epoch powers, one raw example, code, results, figures, and repeat instructions.

stereo-electroencephalography · alpha power · cortical stimulation · visual effects · sign-flip test · epilepsy