No clear distance link to excess high-gamma drift was detected in 2 s windows from 15 SEEG patients

We tested whether bipolar SEEG channels closer to stimulation contacts had more post-window high-gamma power drift than their own baseline drift. We used OpenNeuro ds006519 version 1.0.0. Source coordinates permitted a test in 22 records from 15 patients. Thirteen of 35 source records lacked a required stimulation coordinate. The test used 1,255 nonshared channel occurrences and three fixed 2 s windows. The mean patient rank correlation was -0.0457. Its descriptive 95% patient bootstrap interval was -0.1318 to 0.0511. The exact two-sided sign reference gave p=0.3781 under independent patient values and symmetric signs. Fixed checks also had intervals that included zero. These results do not support a fixed distance rule for selecting read channels in these windows. They do not prove equal reliability at all distances. All 105 raw files passed pinned hash checks. All 385 recomputed spectrum arrays matched the archived inputs exactly. The package preserves code, source metadata, spectra, 210 native audit traces, channel decisions, results, four figures, and reproduction instructions. The spectra and retrieval code came from a cited prior study of the same cohort. The new spatial test is not an independent replication. No sham condition, exact pulse stop time, or decoder outcome was available. A better online brain-computer interface was not demonstrated.

stereoelectroencephalography · high-gamma power · electrical stimulation · electrode distance · signal drift · intracranial EEG · brain-computer interface