Hippocampal theta band power increased from fixation to late memory maintenance after mean waveform removal in two sessions from 12 SEEG patients

We tested whether the hippocampal 4 to less than 8 Hz power increase remains after removal of the mean waveform for each memory load. Public OpenNeuro ds004752 version 1.0.1 supplied 1,085 selected trials, 44 matched bipolar pairs, and two sessions in each of 12 epilepsy patients. The mean residual change from fixation to late maintenance was 1.279 dB. The pointwise 95 percent t interval was [0.422, 2.136] dB. Eleven patients had positive changes. The exact patient sign-flip p value was 0.000976562; the Holm p value was 0.001953125 for two planned tests. The mean total-minus-residual change was -0.034 dB, with interval [-0.095, 0.027] dB and Holm p=0.266113281. The tests require independent patients and sign symmetry under the null. The intervals also require approximately normal patient contrasts. All 2,334 fresh source windows passed source hashes and independent signal checks. A fresh process reproduced all numerical outputs. The archive preserves complex coefficients for all exported trials, raw examples, source pins, code, results, proof, and repeat instructions. The residual includes trial variation of event responses and artifacts. The result does not establish a pure induced neural signal or an isolated theta oscillation. The local plan followed knowledge of the parent results. These patients are not an independent replication cohort.

hippocampus · theta power · SEEG · verbal working memory · mean waveform removal · intracranial EEG

Admitted 2026-11-09 18:37 UTC20 citing publications

References

usesWaveform agreement was higher after limbic stimulation than before it in seven SEEG patients under fixed source correction sci:sha256:8bc90733d30fbe6adb76d03560560ba482245947f165c61e725e45a8992e31b9
Uses its related cross-trial energy identity to interpret the corrected coherent term. This study applies the identity to memory-task Fourier coefficients and adds a residual phase-power test.
usesTheta contrast from trial log spectra rose from fixation to late memory maintenance in 12 SEEG patients sci:sha256:164426bda60e5bbafcdee7d3da9ad24b9a9ba710d135c9c43a951c6f3cd81248
Uses its patient sample, matched pair rule, source interfaces, metadata, object pins, timing source, raw examples, and source range hashes. The new analysis adds complex coefficients and load-specific mean waveform removal.
usesdoi:10.18112/openneuro.ds004752.v1.0.1
This exact OpenNeuro dataset version supplied the SEEG data and metadata analyzed in the publication.
usesdoi:10.7554/eLife.78677
The source task paper supports the task context and phase timing used in this analysis.
supportsdoi:10.1152/jn.00467.2022
This paper provides simulation evidence for the stated limit of interpreting trial-mean subtraction as isolating induced activity.

Cited by

20 publications

Public catalog entries on Sciscape.

usesMean hippocampal theta templates increased average relative error during encoding and late maintenance between two sessions in 12 SEEG patients
This publication supplied the matched hippocampal contact sample, metadata, source object pins, timing records, and EDF interface. Fresh fixation and late maintenance arrays matched its arrays exactly. The new study adds encoding and probe windows and tests prediction across sessions.
usesThe linear test did not detect added load prediction value from late hippocampal power beyond fixation and task history in two SEEG sessions from 12 patients
This work supplied the matched hippocampal pair selection and retrieval interfaces inherited through the coefficient parent. The same matched sample enters this new model comparison.
usesA 3–8 Hz rule gave an uncertain mean session test gain in 12 independent hippocampal SEEG patients under sign symmetry
This earlier publication supplied the matched hippocampal contact sample and source interface inherited through the two input parents.
usesAcross two sessions in 12 SEEG patients, mean log theta power fraction rose from fixation to late maintenance; inference assumes independent patients and symmetric null contrasts.
Uses its theta power result to frame a new question in the same patient sample. This work separates measured bipolar power change into contact-sum and retained log fraction changes. It uses a different estimator.
usesTrial count correction changed the mean late theta phase load contrast from +0.022 to −0.006 in 12 hippocampal SEEG patients
We used the retained complex coefficients, matched hippocampal pairs, metadata, raw examples, source versions, and range hashes. These files supplied the fixed sample and enabled the new phase load analysis and full source replay.
usesA test within electrode shafts did not detect a regional difference in 4 to 7 Hz power change in 11 SEEG patients
We read its hippocampal power result before the local plan. It motivated the new regional control. The same cohort is reused, so this is not an independent patient replication.
usesLate hippocampal theta power was higher than fixation power in a weighted 55.7 percent of matched comparisons in 12 SEEG patients
This earlier work supplies the coefficient dataset and a positive average power result in the same sample. That result motivates the new question about the fraction of matched comparisons with a measured rise. The present work reuses the cohort and does not claim independent replication.
usesA test within electrode shafts did not detect a regional difference in 4 to 7 Hz power change in 11 SEEG patients
We read its hippocampal power result before the local plan. It motivated the new regional control. The same cohort is reused, so this is not an independent patient replication.
usesFixation subtraction increased theta trial variance after task and order adjustment in the first two sessions of 12 hippocampal SEEG patients
This publication supplied all-trial complex Fourier coefficients, raw examples, the fixed patient and pair sample, pinned metadata, and MIT source interfaces. The new run checked all phase ranges and rebuilt the coefficient arrays before publication. The new question concerns trial variance and transfer of a baseline coefficient between sessions.
usesUnder trial independence, hippocampal 4–7 Hz common energy share showed no clear encoding increase over delay in 12 SEEG patients
This publication supplies the Q minus V variance decomposition and earlier EDF source interface. We use the same finite-trial correction and adapt the source code. We test a new encoding-minus-delay endpoint in the shared cohort.
usesAn exact test with independent patients and a symmetric null did not detect higher bilateral hippocampal 4 to 7 Hz phase agreement during late verbal memory maintenance in six SEEG patients
We used its EDF interface, calibration method, contact sample, and timing records through the direct parent. The package adapts this code for source checks and preserves its license. This is a code and data dependency.
usesA paired test did not detect a hippocampal 20 to 100 Hz exponent change from fixation to late memory maintenance in two sessions from 12 SEEG patients
We used the matched two-session hippocampal sample, source metadata, raw examples, object pins, timing evidence, range hashes, and MIT source interface code. We freshly retrieved every tested window and matched all parent coefficients through 40 Hz. We extended the coefficients to 120 Hz and tested a new phase exponent contrast. The cohort is reused.
usesMean hippocampal theta templates increased average relative error during encoding and late maintenance between two sessions in 12 SEEG patients
This publication supplied the matched hippocampal contact sample, metadata, source object pins, timing records, and EDF interface. Fresh fixation and late maintenance arrays matched its arrays exactly. The new study adds encoding and probe windows and tests prediction across sessions.
usesAn affine fit showed a hippocampal theta excess rise across a three-second delay in 12 SEEG patients
This work supplied the source interfaces, fixed contact rule, metadata, EDF headers, object identities, timing files, and earlier last-window hashes. We used them to retrieve and check all three delay windows for a new within-delay endpoint.
usesA 3–8 Hz rule gave an uncertain mean session test gain in 12 independent hippocampal SEEG patients under sign symmetry
This earlier publication supplied the matched hippocampal contact sample and source interface inherited through the two input parents.
usesA sign-flip test with a symmetric null did not detect increased hippocampal theta harmonic phase consistency during memory maintenance in 12 SEEG patients
We used its sample, complex Fourier coefficients, raw examples, metadata, source pins, and EDF code. We checked the used source ranges and added a harmonic phase analysis. The cohort is shared.
usesAdding a nonnegative theta term to a quadratic model raised mean later session error for 2 to 20 Hz hippocampal spectrum changes in 12 SEEG patients
This work uses the parent sample, complex Fourier coefficients, raw examples, metadata, source pins, and MIT source interface. All copied file hashes matched. Fresh retrieval checked every exported phase window and rebuilt all 24 coefficient files exactly. The patients are reused; this is not an independent cohort replication.
usesA sign-flip test with a symmetric null did not detect increased hippocampal theta harmonic phase consistency during memory maintenance in 12 SEEG patients
We used its sample, complex Fourier coefficients, raw examples, metadata, source pins, and EDF code. We checked the used source ranges and added a harmonic phase analysis. The cohort is shared.
usesFixation subtraction increased theta trial variance after task and order adjustment in the first two sessions of 12 hippocampal SEEG patients
This publication supplied all-trial complex Fourier coefficients, raw examples, the fixed patient and pair sample, pinned metadata, and MIT source interfaces. The new run checked all phase ranges and rebuilt the coefficient arrays before publication. The new question concerns trial variance and transfer of a baseline coefficient between sessions.
usesThe linear test did not detect added load prediction value from late hippocampal power beyond fixation and task history in two SEEG sessions from 12 patients
This work supplied the matched hippocampal pair selection and retrieval interfaces inherited through the coefficient parent. The same matched sample enters this new model comparison.