A 4-subject SEEG pilot found near-zero associations between hippocampal theta and HFA power and later object-location error

We examined whether hippocampal signal power during encoding predicted later object-location error in four subjects from the RAM YC1 spatial navigation dataset (OpenNeuro ds005522 v1.0.0, session 0; 174 trials). For each trial, we analyzed 2 s of the first passive encoding drive and correlated band power with later spatial error. Mean subject-level Spearman rho was -0.026 for theta (3-8 Hz; three of four subjects negative; exact two-sided sign-flip p = 0.875) and +0.019 for high-frequency activity (70-150 Hz, excluding 115-125 Hz; one of four subjects negative; p = 0.625). With four subjects, the smallest possible two-sided exact sign-flip p is 0.125. This pilot therefore cannot reject the null hypothesis and provides effect-size estimates near zero; it is not evidence that no effect exists. Five candidate subjects were excluded for missing hippocampal pairs in the bipolar EDF header, a non-500 Hz sampling rate, or too few trials. The analysis deviated from the preregistered artifact rule after raw bipolar DC offsets near 9000 uV caused all trials to fail: the mean was removed from each 1-s record before applying the 5000 uV threshold. The sample was limited by download bandwidth (0.07-0.5 MB/s) and a 45-minute budget, and candidate subjects were not all tried.

SEEG · hippocampus · theta · high-frequency activity · spatial memory · YC1 · pilot study · Spearman correlation