SOZ AUC did not differ by shaft count in 59 SEEG patients across two cohorts

We compared interictal seizure-onset-zone (SOZ) separation between patients whose clinical SOZ labels occupied one shaft (focal, F) and patients whose labels occupied two or more shafts (spread, S). We used archived bipolar-pair feature tables from 59 SEEG patients in two cohorts, calculated each patient's SOZ-versus-non-SOZ AUC, and compared groups with a pooled Mann-Whitney test. The median AUCs for F versus S were 0.777 versus 0.667 for 80-150 Hz power (Holm p = 0.592), 0.791 versus 0.688 for 1-40 Hz power (Holm p = 0.463), and 0.557 versus 0.720 for spike rate (Holm p = 0.592). No primary feature passed the decision rule, and we did not detect a group difference. Confidence intervals include potentially meaningful shifts, and only 18 patients were focal (6 in one cohort). An uncorrected adjusted secondary analysis found a focal advantage for 1-40 Hz power only in ds003876, with the opposite direction in ds004100; we do not claim this effect. Among spread patients, high-gamma power separated SOZ pairs from same-shaft non-SOZ pairs better than from other-shaft pairs, while spike rate showed the opposite pattern (uncorrected). Shaft labels may not represent distinct brain regions, cohorts may differ in clinical labeling, and the analysis used one 120-s clip per patient without surgical outcomes or electrode coordinates.

stereo-EEG · epilepsy · seizure onset zone · interictal · AUC · high-gamma power · spike rate