Five later pulses did not establish higher SEEG size correlation for shared contacts than for nearby separate contacts at 20 to 500 ms in eight sessions

We tested whether shared contacts had higher SEEG response-size correlation than nearby separate contacts. The public data were OpenNeuro ds004696 version 1.0.1. We used five later pulses at 76 sites in eight patient sessions. Size was log RMS of centered bipolar voltage at 20 to 500 ms. The mean shared correlation was 0.356. The mean closest-separate correlation was 0.294. The mean patient difference was 0.062. The descriptive 95% patient bootstrap interval was -0.007 to 0.137. The exact two-sided patient sign-flip p value was 0.180, under a sign-symmetric null. Thus, the primary test did not establish higher size correlation for shared contacts. It did not prove equality or independence. The all-separate control gave a difference of 0.077 with interval 0.012 to 0.147. This result shows sensitivity to control choice. The package includes 24,360 features, all 6,280 combination scores, three figures, source hashes, and a complete retrieval recipe. GPL analysis and extraction code is excluded, so exact source-only reconstruction of compact inputs is unavailable from the public package. Independent arithmetic passed. Numeric outputs repeated byte for byte. Two full public source files rebuilt exactly. Only five pulses and eight sessions limit the result. Contact spacing remains a confound. No online BCI was tested.

stereoelectroencephalography · bipolar recordings · electrical stimulation · response size · correlation · contact overlap · brain-computer interface