At a second center, high gamma envelope correlation at two contact steps was lower than alpha in 7 of 8 SEEG patients

This replication tested whether high gamma (70–150 Hz) envelopes of adjacent-contact bipolar SEEG pairs on one shaft are less correlated than alpha (8–13 Hz) envelopes when the pairs are separated by two contacts. Using the same analysis code as the prior result, we analyzed one 60 s middle segment from each of 8 awake patients in OpenNeuro ds004473 v1.0.1 (OHSU; native sampling rate 999.41 Hz; task phase unknown). Median Pearson correlation was 0.062 for high gamma and 0.180 for alpha; high gamma was lower in 7 of 8 patients (two-sided Wilcoxon W = 1, p = 0.016). The prior study reported 0.071 and 0.212, respectively, in 24 of 24 patients. Across all five bands, correlation at six contact steps was lower than at two steps in all 8 patients (p = 0.0078 each). The median held-out R² for predicting a skipped high gamma pair from pairs two steps away was −0.002. Two secondary comparisons had the same direction as the prior result but did not reach p < 0.05 (excess over between-shaft correlation: p = 0.055; skipped-pair R²: p = 0.20; n = 8). Two patients had high correlations in all bands; a shared non-neural source is possible but was not tested. These results support low shared 100 ms log band-power signal between non-overlapping high gamma bipolar pairs on a shaft in this sample. Limits include one 60 s segment per patient, unknown task phase, no temporal artifact rejection, contact-step rather than millimeter distance, and reuse of the prior analysis code, so shared code errors were not tested. Envelope correlation does not measure task-evoked signal correlation.

stereo-EEG · SEEG · bipolar channels · high gamma · alpha · envelope correlation · brain-computer interface · replication