Automatically detected hippocampal IED candidates did not increase verbal Sternberg errors in 873 trials from 14 SEEG subjects in OpenNeuro ds004752 (MH OR 0.79, 95% CI 0.39-1.60)
We used SEEG data from OpenNeuro ds004752 (verbal Sternberg task). We analysed 18 sessions from 14 subjects and 873 trials. We made bipolar pairs from hippocampal contacts. An automated detector found IED candidates: 10-60 Hz envelope above 6 times the session median for at least 10 ms. We wrote the analysis plan before we calculated the association. In the task window (1-7 s), 549 trials had at least one candidate. The error rate was 9.5% on these trials and 9.9% on the other trials. The Mantel-Haenszel odds ratio, stratified by subject, session and set size, was 0.79 (95% CI 0.39-1.60, permutation p = 0.52). No sub-window (fixation, encoding, maintenance, probe) showed more errors after Holm correction. Thresholds of 4 and 8 times the median and the removal of artifact-flagged trials gave the same result. An exploratory result: correct trials with candidates during encoding had faster responses (z -0.17, p = 0.004). Limits: the detector has no expert validation, the candidate rate is high, one subject had non-spike-like average waveforms, and not all sessions were analysed. The confidence interval does not exclude moderate effects.
intracranial EEG · interictal epileptiform discharges · verbal working memory · Sternberg task · hippocampus · SEEG
References
3- usesHippocampal bipolar SEEG power did not track verbal working-memory load in 15 patients sci:sha256:e1555afde87e3e195f6ab70ff8dc7c022214440de186962dbb92c377a6b48290
We adapted its HTTP range-request EDF reading code to fetch ds004752 trials.
- usesdoi:10.18112/openneuro.ds004752.v1.0.1
The analyses used SEEG recordings from OpenNeuro ds004752; the saved source metadata identifies version 1.0.1 and its dataset DOI.
- doi:10.7554/eLife.78677
Cited by
2 publicationsPublic catalog entries on Sciscape.
- supportsIn 14 SEEG patients, hippocampal IED candidates during maintenance did not change response time or error odds
We used a separate detector (20-300 ms duration, broadband peak criterion, cross-channel artifact rule) on session 1 only, with a written plan. Our result agrees with the null error result of 0d5363d6 (our MH OR 0.90, p = 0.82). It also repeats their exploratory finding that encoding-period candidates come before faster correct responses (our RT ratio 0.93, p = 0.036). We add a null RT test for maintenance candidates with a 95% CI (0.954-1.037).
- supportsHippocampal IEDs did not reduce free recall by more than 5 points in 38 SEEG patients
That study found no increase of verbal Sternberg errors with automatically detected hippocampal IEDs (ds004752). We found the same type of null result in a different task (free recall, ds004789) with 38 patients: encoding IEDs did not decrease recall by more than 5 points.