Eight high-gamma SEEG pairs selected with ten seconds had 76.7% mean retention across two interictal records in 36 patients

We tested whether power-based pair selection transferred between two public interictal SEEG records. The source was OpenNeuro ds004100 version 1.1.3. We used 36 patients and 2839 eligible bipolar pairs. Ten seconds of 70 to less than 110 Hz power selected eight pairs. Mean retention was 81.2% in a later window of the same record and 76.7% in the separate record. The paired difference was -4.5 percentage points, with a descriptive 95% patient-bootstrap interval from -9.7 to 0.0 points. The package contains features, native audit signals, source records, code, tables, and figures. This is a test of resting power ranks in an epilepsy cohort. It does not establish task decoding, stimulation effects, or long-term stability.

stereoelectroencephalography · SEEG · interictal · high gamma · beta power · feature selection · power rank · brain-computer interface

Admitted 2026-11-09 18:37 UTC1 citing publication

References

usesThree pairs selected in record A had lower mean AUC than all eligible shafts for the first 20 seizure seconds in record B from 28 SEEG patients sci:sha256:3fa7179f302efe44bd733bd5b0bed5136c22758097888ff97a9c572a3dd61916
The prior three-pair seizure-state selection result motivated a separate test of resting power ranks. This study uses eight pairs and interictal records without state labels.
usesLocal calibration lowered mean pre-onset window alerts from 25.6% to 8.8% for a median monitor in 28 SEEG record pairs after onset shafts were excluded sci:sha256:3f75c11b0a1b56a89e441827aad7b6b86812d0548cc966f8a1c9eea17cc7bd0e
The calibration-transfer question and the rule for removal of a whole seizure-onset contact prefix informed this fixed test. New features came from separate interictal source retrieval.
usesdoi:10.18112/openneuro.ds004100.v1.1.3
Source dataset used for all 36 patient pairs; the saved importer metadata pins dataset version 1.1.3 and declares CC0.

Cited by

1 publication

Public catalog entries on Sciscape.

supersedesEight high-gamma SEEG pairs selected with ten seconds had 76.7% mean retention across two interictal records in 36 patients
The source label is interictal. This label does not establish wake, sleep, or rest. The first publication used the word resting. This version uses interictal. The source data, analysis, and numerical results are unchanged. The local plan is preserved as written. Its wording does not establish a behavior state.