Two control sessions show large pre-to-post decreases in hippocampal evoked potentials, with pulse timing confounding one session

This analysis measures pre-to-post change in single-pulse evoked-potential (EP) magnitude in two control sessions from OpenNeuro ds006065; the planned closed-loop session did not download, so there are no closed-loop data. EP magnitude was the RMS of the trial-average waveform from 50 to 300 ms. The median log2 post/pre change was -1.56 (p = 0.001; 13 responsive pairs) in UC004 and -1.91 (p = 0.012; 3 pairs) in UC005. Within-file drift was small (|D| <= 0.43) and not significant. UC004 had different pulse timing in pre and post files; restricting both files to pulses after a 0.5 s interval reduced the change to -0.39 (p = 0.27). UC005 had the same timing in both files and the restricted change remained -2.45 (p = 0.002), but only 3 pairs were responsive, so this estimate is not stable. With n = 2 control subjects and no closed-loop session, no group statistics are possible. These results show that pre-to-post EP change alone cannot establish a closed-loop effect; comparisons require control sessions and matched pulse timing. Tests were per session over trials, responsive pairs were fixed and not independent, pulse times came from artifact detection, and the analysis plan changed after data inspection.

single-pulse stimulation · evoked potentials · intracranial EEG · hippocampus · pulse timing · closed-loop stimulation · control sessions