Removing links with shared contact labels gave no clear gain in later gamma scores in 34 SEEG post intervals

We tested whether removal of coherence links with shared contact labels improved a later channel score. The test used 34 public SEEG records from 20 people in OpenNeuro ds006519 version 1.0.0. Each rule used two seconds to choose a channel. Both rules received the same score in the next two seconds. The primary test used post intervals and gamma frequencies. Channels on the stimulated shaft were excluded. Removal changed 12 of 34 choices. The mean later score difference was -0.000554 in magnitude squared coherence. The 95% person bootstrap interval was -0.004256 to +0.002951. The data gave no clear gain. The interval permits a loss or a gain and does not establish equal performance. All 64 fixed test contrasts are preserved. Four unadjusted secondary intervals excluded zero under other conditions. This is an offline score test in a reused clinical group. It did not test decoding or a stimulation benefit. The package contains four figures, exact compact inputs, complete scores, source metadata and hashes, a pinned retrieval recipe, code, and reproduction instructions. Fresh public voltage files rebuilt all 68 compact array files exactly. An independent pair calculation matched all primary choices. A full repeat gave identical main numeric output files.

stereoelectroencephalography · magnitude squared coherence · channel selection · brain-computer interface · contact labels · gamma band · OpenNeuro