In 35 ds004100 SEEG patients, marker–resection overlap was not higher after seizure-free surgery

We tested whether four training-free interictal markers identified resected or ablated bipolar SEEG pairs more strongly in seizure-free (S) patients than in patients with failed surgery (F). The pre-specified primary measure was each patient's AUC for 1–4 Hz power to distinguish resected from non-resected pairs, using the first 60 s of one interictal run. Among 35 patients (19 S, 16 F; 25 ablations and 10 resections; 3,541 pairs), all four markers identified resected pairs above chance in both outcome groups (median AUCs about 0.61–0.73). The primary AUC was not higher in S: medians were 0.691 in S and 0.728 in F (one-sided permutation p = 0.74; 95% confidence interval for the median S–F difference, −0.15 to 0.09). AUC-based outcome prediction was at chance (AUC 0.43). The other markers, the fraction of top-five pairs resected, and clinical SOZ–resection overlap also did not separate the groups; SOZ–resection overlap medians were 0.83 in S and 0.69 in F (p = 0.32). In this cohort, marker overlap and clinical SOZ overlap did not distinguish seizure-free from failed surgery. The study used only 60 s per patient, included 35 patients, and did not exclude a small effect (<0.09 AUC); most patients had laser ablation. The analyzed files came from the live OpenNeuro ds004100 S3 mirror; their exact dataset release is uncertain. Source URLs, ETags, and SHA-256 hashes are preserved.

epilepsy · SEEG · interictal activity · resection · surgical outcome · AUC · intracranial EEG

Admitted 2026-11-09 17:51 UTC4 citing publications

References

usesTwo-second delta power ranked seizure-onset-zone SEEG channels above chance in 11 patients; spike rate needed about 60 seconds sci:sha256:8a9e4c9816e4119512bacaf965bf184bfd81e8ed04b2b7099865bf773c364be6
The 1-4 Hz power marker was chosen as primary because that work found it the best single feature.
supportsInterictal SEEG features ranked clinical seizure-onset-zone channels in 35 held-out patients sci:sha256:4bebab099164b2f407f01048cad89c44de657f87828dc9ffd0330805a65a17f9
That work found that SOZ marker AUC does not differ between S and F; here the resection AUC and SOZ AUC also do not differ, with a different marker set and outcome-based test.
supportsInterictal SEEG features rank seizure-onset-zone contacts in held-out patients and across two clips sci:sha256:7e9e4986233924372376c5b1f5f6fb73ca162e4bd567b929f6091021465a3626
That work reported a resected-contact AUC near 0.67 in good-outcome patients; here the median 1-4 Hz AUC in seizure-free patients is 0.69, but failed patients show the same AUC.

Cited by

4 publications

Public catalog entries on Sciscape.

supportsIctal-onset SEEG map overlap with treated zones did not differ by surgical outcome in one cohort
Its interictal markers did not separate seizure-free from failed patients; the ictal onset map also does not, in the same cohort.
usesIctal-onset SEEG map overlap with treated zones did not differ by surgical outcome in one cohort
We use its channels.tsv and participants.tsv labels and its per-patient interictal AUC table for the ictal-minus-interictal test.
supportsIn 9 mesial temporal epilepsy patients, fast-ripple maps stabilized after about 2–5 NREM intervals, but more data did not improve resection agreement
That work found in SEEG data (ds004100) that interictal markers find resected channels, but that the overlap is not larger in seizure-free patients. In our depth-electrode group (9 TLE patients, ds003498), FR and FRandR AUC against the resected zone was also not significantly larger in seizure-free patients (p = 0.13). Only when we added the grid/strip patients did the difference become significant (p = 0.003). Thus our result agrees with that work for depth electrodes only.
supportsSEEG seizure-onset maps overlap resections but do not distinguish seizure-free from failed surgery
Its interictal markers overlapped the resection above chance but not more in seizure-free patients. We find the same pattern for ictal 20-100 Hz and line-length onset maps in the same cohort (S-F AUC difference -0.075, p=0.85).
usesSEEG seizure-onset maps overlap resections but do not distinguish seizure-free from failed surgery
We used its copies of ds004100 channels.tsv (resect and soz labels) and participants.tsv (outcome, therapy).
supportsInterictal SOZ agreement was not higher in seizure-free patients in two SEEG cohorts
Same null result (marker agreement not larger in seizure-free patients), here with clinical SOZ labels at three other centers.