In 25 patients with SEEG, hippocampal ripple rate was not higher in the second before vocal recall than during silent recall

We tested whether hippocampal ripple rate increases in the second before vocal free recall using session 0 of the RAM FR1 delayed free recall dataset (OpenNeuro ds004789). We applied a prespecified 80–115 Hz envelope detector with a mean + 3 SD threshold and an interictal epileptiform discharge (IED) mask to bipolar SEEG channels. The planned analysis included 25 patients with at least 10 qualifying recalls (902 recalls; 222 hippocampal pairs; 23 patients with 319 control pairs). The mean hippocampal pre-recall minus silent-baseline rate was −0.0004 events/s (95% CI −0.015 to +0.017; 12/25 positive; p = 0.96, Holm p = 1.0). The hippocampal log rate ratio was slightly negative, and the hippocampal versus control comparison was null. Thus the planned test did not support a higher pre-recall rate. Exploratory results depended on detector settings: a mean + 2 SD threshold gave a weak, nonsignificant increase, while removing the IED mask gave a decrease. These findings do not provide robust support for a pre-recall ripple increase with this detector, but do not exclude a small increase in low-amplitude events. The analysis covers 25 of about 100 eligible patients, session 0 only; it has additional limits from sampling rates, atlas-based channel selection, manual onset annotations, a late silent baseline, and the use of epilepsy patient recordings where physiological ripples and pathological high-frequency oscillations were not separated.

hippocampus · free recall · ripple detection · SEEG · intracranial EEG · memory · OpenNeuro · interictal epileptiform discharges