Hippocampal SEEG did not reliably distinguish in-set from out-of-set probes in 14 patients

We examined whether hippocampal activity during the first second after a probe letter distinguished in-set (Match = IN) from out-of-set (Match = OUT) probes in the verbal Sternberg task in OpenNeuro ds004752. We selected one session per subject (the smallest EDF) for 14 patients with atlas-labeled hippocampal contacts and analyzed 28–47 correct, artifact-free trials per patient. The planned primary measure was 70–150 Hz bipolar high-frequency activity (HFA) from 0.1–0.6 s after probe onset. HFA was nominally higher for IN than OUT probes (+0.096 z; 95% bootstrap CI 0.014 to 0.182; 11/14 positive; sign-flip p = 0.044), but a within-subject label-permutation test gave p = 0.065. Three of ten pre-probe 100-ms bins also showed differences of up to 0.16 z, indicating that effects of this size occurred by chance in this sample. Theta, evoked potential, non-hippocampal HFA, and late HFA did not differ significantly after Holm correction. With about 40 trials per patient, the measured hippocampal responses did not reliably separate IN from OUT probes; effects smaller than about 0.2 z are not excluded. The calibration check was added after the planned analysis. Atlas labels were not verified on MRI, and no interictal spike rejection was applied beyond the dataset artifact flag.

stereo-EEG · hippocampus · verbal working memory · Sternberg task · high-frequency activity · probe response · permutation test