Ictal-onset SEEG map overlap with treated zones did not differ by surgical outcome in one cohort

We tested whether the overlap between an ictal high-frequency SEEG onset map and the treated zone was greater in seizure-free than failed-surgery patients. In one center, 33 patients from OpenNeuro ds004100 (20 seizure-free, 13 failed; 127 seizures; 22 ablation, 11 resection) had at least one treated and untreated bipolar pair. The primary per-patient measure was the AUC of the 20–100 Hz onset map for treated versus untreated pairs. Median AUC was 0.78 in both groups; the difference was +0.00 (95% CI −0.17 to +0.26; one-sided p = 0.53). Line length, an early onset window, top-pair measures, clinical seizure-onset-zone labels, and the ablation-only subgroup also showed no seizure-free advantage. Across the cohort, the ictal map ranked treated pairs above untreated pairs (median AUC 0.78; 27/33 patients above 0.5; p < 0.001), while a sham pre-ictal map did not (median AUC 0.50; p = 0.35). The ictal map was closer to the treated zone than an interictal 70–115 Hz marker in the same patients (mean AUC difference +0.12; 22/32 higher; p = 0.042). These results show that, in this cohort, pair-level overlap with the sampled ictal map did not explain surgical failure. The study was limited to one center, included 13 failed patients with a wide confidence interval, used onset and treatment decisions informed by the recorded seizures, and used contact labels that do not measure treated volume.

stereo-EEG · seizure onset · epilepsy surgery · high-frequency activity · treated zone · surgical outcome