Session-level re-analysis finds no significant theta-synchronized increase in hippocampal late SEPs

An independent session-level re-analysis of OpenNeuro ds006065 v1.0.0 examined whether theta-synchronized (TS) lateral temporal stimulation increased hippocampal late single-pulse evoked potentials (SEPs, 50–250 ms) more than phase-blind (PB) stimulation. We analyzed pre and post stimulation blocks from 9 sessions (4 TS, 5 PB), selected a contact inside an MNI hippocampal box by its pre-block late SEP, and compared post/pre log amplitude ratios with an exact permutation test. In the primary monopolar analysis, the mean ratios were +0.22 for TS (×1.25) and −0.10 for PB (×0.90; p=0.10). Excluding p16 reduced the difference (p=0.27). With bipolar referencing, the means were −0.06 and −0.11 (p=0.82). Early SEPs did not differ significantly (monopolar p=0.44; bipolar p=0.15), and the median over responsive contacts also did not differ significantly (monopolar p=0.11; bipolar p=0.13). In p16, pre/post waveforms inverted on most responsive contacts; p19 monopolar contacts had slow step artifacts, and PB-only patients had small late SEPs. The result depends on montage and one session with waveform inversion. This session-level analysis does not refute the published trial-level mixed-model result; contact selection and statistical unit differed, and only 4 versus 5 sessions were available.

intracranial EEG · single-pulse evoked potential · theta-synchronized stimulation · hippocampus · session-level analysis · reproducibility