In 35 one-center ds004100 SEEG patients, ictal 20–100 Hz maps agreed with treated contacts but did not separate surgical outcomes

We tested a pre-specified question on OpenNeuro ds004100 (one center). Does the ictal 20-100 Hz onset map point to the treated contacts better in seizure-free patients than in patients with failed surgery? We used 35 SEEG patients (20 seizure free, 15 failed) and 1 to 5 seizures for each patient. For each bipolar pair, the map is the mean z-score of log 20-100 Hz power from 0 to 10 s after the clinical onset mark. The baseline is -40 to -25 s. For each patient, we calculated the AUC of the map for pairs with a "resect" contact against all other pairs. The map agreed with the treated contacts in both groups: median AUC 0.78 in seizure-free patients and 0.73 in failed patients. A sham map from before the onset was at chance (median 0.50). The primary test was negative: the AUC did not separate the outcome groups (P(S>F)=0.56, permutation Mann-Whitney p=0.54). Line length and the clinical SOZ label gave the same result. The test can only find large effects (outcome AUC 0.75 or more). Thus, in this cohort, treatment of the high-frequency onset zone did not explain the outcome. The package includes the code, the per-run features, the labels with ETags, the results and the figures.

stereo-electroencephalography · epilepsy · seizure onset · high-frequency activity · surgical outcome · resection · ablation · AUC