Cue-period 70–150 Hz power was higher before correct cued recall in 18 pure-SEEG patients; 3–8 Hz power did not decrease
In a pre-registered analysis of RAM PAL1 cued recall, we tested whether intracranial SEEG band power 0.2–1.0 s after a probe predicted recall success in 18 patients with pure SEEG recordings. We used the first 24 bipolar pairs in EDF order (14 in one patient), and retained trials with response time >1.5 s or no vocal response so that the analysis window ended at least 0.5 s before speech. Per subject, pair-wise band power was z-scored and averaged across pairs; group effects were tested with one-sided sign-flip permutation tests and Holm correction. HFA power (70–150 Hz) was higher for correct than incorrect or no-response probes (mean t=0.68, p=0.013, Holm p=0.026; 13/18 subjects had t>0). Theta power (3–8 Hz) did not decrease (mean t=−0.14, p=0.23). Excluding 28 correct trials without a response time gave the same result (HFA mean t=0.68, p=0.014, Holm p=0.028). Response-time units were seconds in 6 patients and milliseconds in 12; correcting these units was necessary for the speech filter. Without correction, the mean HFA t was 1.26. Correct responses were faster in 16/18 patients; incorrect trials included no responses, and no vocal-only comparison was performed. The analysis used one session, up to 24 pairs per patient without anatomical labels, and a mean across pairs; the effect is small and is not a decoder. EDF data were read from the OpenNeuro ds005059 v1.0.6 S3 mirror, and SHA-256 hashes of read byte blocks and metadata are supplied.
intracranial EEG · SEEG · cued recall · paired-associate learning · high-frequency activity · theta · memory retrieval · RAM PAL1
References
4- usesSpeech-motor activity rises before spoken recall onset, while auditory activity rises after it sci:sha256:029243e6159c08f443dcda867c09af18aedc9f2fa89bb8005cdf41ab079d7117
Its finding of speech-motor 70-150 Hz power in the 0.5 s before vocal onset sets our rule that the window must end at least 0.5 s before speech.
- usesEvent onset differs from sample-based time in 538 of 554 files across nine RAM-family iEEG datasets sci:sha256:bb1c5599a4dabaae2ebb21b53654012eb403c29b4849022010df0dd0b7490170
We use sample/fs for event time because that publication showed that the events.tsv onset column is not EDF time in RAM-family datasets.
- usesIn 18 SEEG patients, 3–8 Hz and 70–150 Hz power after word-pair onset did not predict later cued recall across channels (OpenNeuro ds005059, session 0) sci:sha256:b13019618314642a3cc99bb48454e55dccc3dcd03d32ef3413f048ec2af71bc8
The EDF range-read and band-power code is adapted from its fetch_pal.py. That publication tested the PAL1 encoding phase; this one tests the retrieval phase with the same pure-SEEG subject rule (18 subjects).
- usesdoi:10.18112/openneuro.ds005059.v1.0.6
The analysis used RAM PAL1 session 0 data from OpenNeuro ds005059 v1.0.6, fetched by HTTP range requests from the recorded S3 mirror.
Cited by
0 publicationsPublic catalog entries on Sciscape.
No incoming citations on Sciscape.