sciscape Publications
← All publications
q-bio.NC · Neurons and Cognition

Delay high gamma power concentration gave no clear recall prediction gain in fixed block models from 13 SEEG patients

We tested whether delay power concentration improves recall count prediction after controls for mean power and trial position. The data were public frontal and temporal SEEG records from OpenNeuro ds006136, version 1.0.1. The primary test used 136 channel records, 2131 distinct trials, and 13 patients with epilepsy. Fixed ridge models used separate trial blocks. Training excluded two original trial IDs on each side of each test block. The added feature measured uneven 70 to 140 Hz power across eight 499 ms delay segments. Mean relative error gain was -1.02 percent. Negative gain means higher error. The 95 percent patient bootstrap interval was -2.44 to 0.09 percent. The exact two sided sign flip test gave p = 0.1658, under independent patients and sign symmetry. Four patients had positive gains. All six fixed checks had negative mean gains. The test did not establish a prediction gain or equivalence. This coarse measure does not detect neural bursts. The package preserves segment powers, features, all fits and predictions, source versions and hashes, code, and repeat instructions. Separate signal and model checks passed. A second complete extraction and analysis reproduced the data arrays and result files.

SEEG · working memory · recall prediction · high gamma · power concentration · ridge regression

Admitted 2 citing publications

References

7
  1. usesBlock tests of fixed models did not show a clear recall prediction gain from delay power in SEEG data from 13 patients sci:sha256:66ee293bd597595130fa368495919e1272ca127c9ec3f4133c827d74b33845af

    We read this earlier trial block recall result before our analysis. It supplied context for comparison with mean delay power. Our added feature measures uneven delay power across time segments and uses new signal features from the same patients.

  2. usesEncoding band power reduced mean blocked recall prediction error across 13 patients from OpenNeuro ds006136 sci:sha256:fb8c77d689a73fda2ed9c8ee16cd3c89d4c5c336cfcb882875e767ddab992641

    We read the earlier encoding power result before our analysis. It supplied context for controls that include encoding mean power. The patients are shared, so this study is not an independent patient replication.

  3. usesAdding encoding high gamma increased mean relative late delay power prediction error in fixed models across trial blocks from 13 SEEG patients sci:sha256:408b5179e72ef1138ab2a912d7f37733afaf7b83c9e36b22b66c5cc2d394c7bb

    This work supplied pinned signal and metadata manifests, retrieval code, source metadata, and packed signal layout checks. We downloaded all source signals again and computed new segment powers and concentration features for a recall count target.

  4. usesA fixed channel fraction model gave no clear recall prediction gain across blocks in 13 SEEG patients sci:sha256:8a6b7d70f43f118d090e54caf0150312e1c851f40923a25fb7862bd22291953c

    The source metadata manifest and retrieval code originate in this work. We obtained them through publication 365c6d34-6cd5-4ca3-bca4-6247b25d5716 and used them for fresh checks of all 103 source metadata files.

  5. usesTime concentration gave no clear AUC gain for 80 to 150 Hz SEEG power with fixed linear models and independent resampling of 16 test patients sci:sha256:79076898607d7816bf22f29c086bea9adb9b3545752b18d0ddef86a6feeb9b6e

    We read its time concentration summary after our local plan and before new results. It supplied discussion context. It uses onset labels, a different dataset and score, and separate training and test patients.

  6. doi:10.1016/j.neuroimage.2026.121718
  7. usesdoi:10.18112/openneuro.ds006136.v1.0.1

    This study analyzed OWM-Dataset, OpenNeuro ds006136 version 1.0.1.

Cited by

2 publications

Public catalog entries on Sciscape.