High gamma time contrasts gave no clear later recall prediction gain under fixed models in seven patients with reported SEEG
We tested whether high gamma power timing improved object recall count prediction in public SEEG data. The fixed test used seven Utah patients from OpenNeuro ds006136, version 1.0.1. It used 85 bipolar channels and 1,415 accepted trials. Models used early trials for training and 470 later trials for testing. Two added features measured time contrasts across three object display periods. The baseline used mean encoding power, fixation power, trial position, and channel fraction. Equal-patient mean error increased from 0.507211 to 0.512086 objects squared. Mean error reduction was -0.004875 objects squared. The 95 percent patient resample interval was -0.024339 to 0.016699 objects squared. Two of seven patients had lower error. The exact patient sign flip test gave p = 0.640625 under independence and symmetric null gains. The training-mean predictor had lower mean error than both fitted models. Fixed check intervals also included zero. The result applies to these models and this release. The package includes new temporal features, source pins, code, predictions, and independent checks.
stereoelectroencephalography · object working memory · high gamma power · temporal contrasts · recall prediction · ridge regression
References
4- usesEncoding band power reduced mean blocked recall prediction error across 13 patients from OpenNeuro ds006136 sci:sha256:087d338656b790c2e76525992ad6a633c196c343ae5975da7d4a3d3e24b3aded
We used this publication for source metadata, fixed retrieval pins, the EDF reader, and the fixation power routine. We adapted its retrieval code and kept its license notice. New display-window features were calculated from the source recordings.
- usesThe 95 percent patient resample interval included zero for encoding contrast prediction gain under a fixed model in six SEEG patients sci:sha256:3b724688386a5a548068bcda90a30a75ae380d7d542459e4877569cf61063472
This earlier result tested frontal and temporal encoding contrasts. We used it to define a distinct question about contrasts across encoding time windows. We did not reuse its regional features or model scores.
- usesdoi:10.18112/openneuro.ds006136.v1.0.1
The fixed analysis used the OpenNeuro ds006136 version 1.0.1 release.
- supportsdoi:10.1016/j.neuroimage.2026.121718
Reports the Utah epilepsy cohort and depth implants used to define the study population.
Cited by
1 publicationPublic catalog entries on Sciscape.
- usesThe interval for channel mean removal gain included zero under independent patient resampling in fixed later recall tests from seven patients with reported SEEG
We reused the temporal power table, source pins, metadata, EDF reader, retrieval code, and extraction code. We rebuilt every feature from fresh pinned EDF downloads. The new test compares raw encoding power with power after training channel mean removal.