The HFA memory effect did not decrease across FR1 session 0, while theta power increased
We examined time on task in session 0 of a free-recall task using archived 0–1.6 s per-word bipolar band-power features from 83 subjects. The analysis included 72 pure-SEEG subjects with at least 12 lists. For the two primary tests, HFA (70–150 Hz) did not drift from the first to last list (−0.027 z, 95% CI [−0.088, +0.038], Holm p = 0.41), and its subsequent memory effect (SME) did not decrease in the late half relative to the early half (+0.026 z, 95% CI [−0.001, +0.053], Holm p = 0.12). The confidence interval excludes a decrease larger than 0.001 z. The HFA SME was present in the late half (+0.029 z, p = 0.0012) but not the early half (+0.003 z); this increase was not significant after Holm correction. In secondary uncorrected tests, theta power (3–8 Hz) increased over lists (+0.145 z, 95% CI [+0.086, +0.208], p = 0.00005), while the theta SME did not change. Recall probability did not decrease. These findings apply to session 0, one 0–1.6 s window, and averages across all bipolar pairs; list rank and time on task are confounded, breaks were not modeled, and the analysis cannot distinguish state changes from recording changes as causes of theta drift.
free recall · subsequent memory effect · intracranial EEG · high-frequency activity · theta power · time on task · FR1
References
2- usesIn 83 pure-SEEG patients, mean HFA and theta slopes were not significant, and HFA did not explain primacy sci:sha256:f677a6fd27b2fd81b0e22487cc5ecd4440063dedc4bacf3999b8bdd88c1974c0
We use the archived 83-subject FR1 per-word band-power features of this publication as the only input.
- usesdoi:10.18112/openneuro.ds004789.v3.1.0
The archived band-power features analyzed here were derived from session 0 recordings in OpenNeuro ds004789.
Cited by
3 publicationsPublic catalog entries on Sciscape.
- supportsIn 69 catFR1 patients, low-frequency SEEG power increased over one session without predicting recall change
The 3-8 Hz power increase over the session that this work found in FR1 also occurs in an independent catFR1 cohort (+0.036 log10, 52/69 patients).
- supportsSample-based timing leaves the primary e9c2d073 null results unchanged in 24 RAM FR1 SEEG subjects, while theta drift replicates
With sample/fs epoch timing, the e9c2d073 primary tests (HFA drift and HFA SME late-minus-early change) stay non-significant (Holm p=0.71 for both, n=24), and the theta drift increase replicates (+0.138 z, p=0.008). The e9c2d073 descriptive pattern of a near-zero early-half HFA SME does not appear with sample timing (early-half SME +0.028 z, p=0.009), but the paired timing difference for it is not significant (p=0.19).
- usesSample-based timing leaves the primary e9c2d073 null results unchanged in 24 RAM FR1 SEEG subjects, while theta drift replicates
We ran the verbatim e9c2d073 analysis code and used its archived per-subject results for the T0 reproduction check and the sensitivity comparison.
- supportsCorrectly timed epochs show stable HFA memory effects and rising theta, with an early-half effect also present
Re-test with sample/fs-timed epochs (n = 24) supports the e9c2d073 main results: no decrease of the HFA SME from the early to the late session half (+0.016 z, CI -0.015 to +0.049), no HFA drift, and a theta increase (+0.138 z, p = 0.0076). It does not support the e9c2d073 observation that the HFA SME was absent in the early half; with correct timing the early-half SME is +0.028 z (p = 0.0085). The verbatim e9c2d073 code reproduced its archived per-subject values on onset-timed features (20/20 subjects).