A low frequency SEEG score exceeded a control with fixed shaft onset counts by 0.058 mean AUC in two 120 s interictal records from each of 36 patients

We tested whether a fixed low frequency SEEG score gave contact ordering above a control with fixed onset counts in each named electrode shaft. We used archived public measurements from OpenNeuro ds004100 version 1.1.3. The cohort had 36 patients, two 120 s interictal records per patient, and 4,252 matched contacts. The score was log10 of median 2 to 4 Hz density divided by median 0.5 to 1 Hz density. Higher scores predicted clinical onset labels. The mean observed AUC was 0.647. The exact label-control mean was 0.589. The mean patient excess was 0.058, with a 95 percent patient bootstrap interval of 0.015 to 0.104. The excess was positive in 24 patients, negative in 10, and zero in two. Average-reference and second-record results were weaker. An exact rank identity gives the control mean and variance, with a shared label draw across records. Complete enumeration and independent pair counts passed. The package preserves the proof, reusable code, all source features, clinical metadata, pinned raw retrieval records, results, figures, and repeat instructions. This is a secondary analysis of an earlier AutoSci measurement package. Named shafts do not control anatomy, and artificial label placement does not establish spatial exchangeability. No clinical decision or treatment outcome was tested.

stereoelectroencephalography · SEEG · epilepsy · AUC · rank benchmark · interictal EEG · electrode shaft