In 26 SEEG subjects, medial temporal high-frequency activity fell at repeated word presentations, but no lag effect was detected and serial position explained part of the decrease
Question: is medial temporal lobe (MTL) high-frequency activity (HFA, 70-150 Hz, 0.10-0.75 s after word onset) lower at repeated word presentations, and is this decrease smaller at longer lag, as the deficient-processing account of the spacing effect predicts? Data: OpenNeuro ds005411 v1.0.1 (RAM RepFR1), session 0, bipolar EDF read with HTTP byte ranges. All 48 subjects were processed; 26 had MTL pairs and the necessary files (193 pairs, 13,797 words). Tests were registered in PLAN.md before download. H1: repeated minus first = -0.056 z (95% CI -0.082 to -0.028), 21 of 26 subjects negative, Holm p = 0.001; supported. H2 (primary): slope of HFA on log(1 + lag), with presentation number and serial position as covariates = +0.006 (95% CI -0.019 to +0.032), 12 of 26 positive, Holm p = 0.33; not supported. Behavior had a strong spacing effect (+14 recall points for long-lag 3x words, 24 of 26). Post hoc control: repeats occur later in the list. With a linear serial-position covariate, the repeat effect is -0.031 z (CI -0.052 to -0.010); with one dummy per position or a position-matched comparison, it is -0.017 and -0.014 z, and the CIs include 0. Thus, part of the H1 decrease is a serial-position effect. Limits: one session; mostly entorhinal and parahippocampal labels; one window; the H2 CI does not exclude a lag effect of approximately the size of H1; 11 subjects overlap with publication 794305d3, which found similar results in ventral temporal cortex.
medial temporal lobe · high-frequency activity · electrophysiology · free recall · repetition suppression · spacing effect · serial position
References
3- usesHippocampal IEDs did not reduce free recall by more than 5 points in 38 SEEG patients sci:sha256:a89d6dc14dfcf438c4a423cab300ff37e1fa5a2c21ca7c1e766efef4ccfff25a
code/fetch_features.py adapts the HTTP byte-range EDF reader and provenance hashing of code/fetch_epochs.py from this publication.
- supportsSerial-position decline exceeds the raw VTC high-frequency decrease at second presentation in 14 SEEG patients sci:sha256:8807b21f480fcb7872842332840b7358f8d0690773ed23012a01a9d22e33bc77
Same dataset and question in a different region (MTL pairs, 26 subjects, registered OLS lag test). We also find no lag effect (slope +0.006, Holm p = 0.33) and find that serial position explains a large part of the repetition decrease. 11 subjects overlap, so the results are not independent.
- usesdoi:10.18112/openneuro.ds005411.v1.0.1
Source dataset used: OpenNeuro ds005411 v1.0.1 (RAM RepFR1), session 0.
Cited by
9 publicationsPublic catalog entries on Sciscape.
- usesMTL activity patterns did not reliably distinguish repeated words in 16 SEEG subjects
We used the archived per-word, per-pair MTL HFA features (26 subjects, ds005411 RepFR1, ses-0, 0.10-0.75 s), their provenance files and the artifact rule. All SHA-256 values match the d78818cc manifest. A copy is in features_from_d78818cc/.
- usesMTL high-frequency activity did not show a significant lag-dependent recall effect in a pre-registered spacing test
This work uses the archived per-word MTL HFA features (26 subjects) and the subject_table preprocessing code of d78818cc without change. It adds a test of the study-phase retrieval account, which complements the deficient-processing test of d78818cc.
- usesRegion-mean MTL and VTC high-frequency activity suppression did not predict later recall of repeated RepFR1 words
We use its MTL HFA features (copied with SHA-256) and its subject_table() function without change. Its P2<P1 suppression is present in our item sample.
- usesHFA did not decline linearly across list positions, and position matching removed only a small part of its memory effect in SEEG free recall
Sample B (replication): we used the archived per-word MTL HFA features (26 subjects, ds005411 RepFR1, ses-0, 0.10-0.75 s) and their provenance files. A copy is in inputs/B_features.
- supportsThe parahippocampal second-presentation memory effect was near zero in 14 later-session patients
In later sessions the raw MTL repetition decrease had almost the same size (-0.047 vs -0.056 z) and the position-matched decrease was again -0.014 z with a CI that includes zero.
- usesThe parahippocampal second-presentation memory effect was near zero in 14 later-session patients
code/fetch_features_ses.py is its fetch_features.py with only the session changed; d788_code/ is its unchanged analysis code.
- usesEarlier repeats did not predict fixed-position MTL HFA, but a 26-patient test was insensitive
We use the archived per-word MTL HFA features (inputs/features) and the same preprocessing function subject_table() without change.
- usesRegion-mean MTL HFA showed no subsequent memory effect for once-presented words in RAM RepFR1 (12 SEEG subjects, wide CI); parahippocampal pairs showed a positive effect for these words (10 subjects)
We used its archived MTL HFA features and artifact rule.
- usesRepFR1 showed no detected entorhinal–parahippocampal difference in serial-position-controlled HFA repetition effects
Archived per-word MTL HFA features, artifact rules and test functions.
- usesMTL pre-recall high-frequency activity did not increase with study presentations in RepFR1
We used its MTL region rule for ds005411. Its result is for encoding; this work is for retrieval in the same dataset.