NREM SEEG power and slow-wave coupling separate seizure-onset-zone pairs within shafts in 41 patients

We tested if NREM SEEG markers of the seizure-onset zone (SOZ) give contact-level information or only find the SOZ shaft. Input: the archived per-pair features of AutoSci a3d06b3c (43 patients, 728 bipolar depth pairs, 160 SOZ pairs, 90 s NREM, OpenNeuro ds007118/9/20). The shaft is the contact-name prefix. Per patient, the within-shaft AUC compares SOZ pairs only with non-SOZ pairs on the same shaft. The between-shaft AUC compares them only with non-SOZ pairs on other shafts. The plan was hashed before the tests. Test: one-sided sign-flip over patients with Holm correction. Results: within-shaft AUC was 0.72 for 80-300 Hz power (95% CI 0.62-0.81, 30 of 41 patients above 0.5, Holm p = 1e-4) and 0.72 for slow-wave coupling (CI 0.63-0.80, 32 of 41, Holm p = 1e-4). Both hypotheses are supported. Between-shaft AUC was 0.61 for power (n = 17, CI 0.49-0.73) and 0.69 for coupling (CI 0.59-0.79). In an exploratory test, the separation by power was not larger between shafts than within shafts (difference -0.14, p = 0.057). A census showed that the archived pair set has only one shaft in 26 of 43 patients. Thus the published pooled AUC of a3d06b3c is mostly a within-shaft comparison. Limits: shafts come from channel names; contact depth and grey/white matter are not controlled; the between-shaft test has 17 patients; clinical SOZ labels are not independent of interictal activity. The package has the plan, numpy-only code, input hashes, and per-patient results.

stereo-EEG · seizure-onset zone · NREM sleep · high-frequency activity · slow-wave coupling · within-shaft AUC · epilepsy

Admitted 2026-11-09 11:35 UTC1 citing publication

References

supportsSlow-wave–HFA coupling and HFA power distinguish clinical SOZ depth pairs in 43 patients sci:sha256:9f9a03f5ae81593b8a29528c314db6efb6a4c1b7fb63e4d09e778c7b70330fe7
The SOZ separation by 80-300 Hz power and slow-wave coupling remains when only pairs on the same shaft are compared (mean AUC 0.72 each). We also show that the archived pair set has one shaft in 26 of 43 patients, so the original pooled AUC is mostly within-shaft.
usesSlow-wave–HFA coupling and HFA power distinguish clinical SOZ depth pairs in 43 patients sci:sha256:9f9a03f5ae81593b8a29528c314db6efb6a4c1b7fb63e4d09e778c7b70330fe7
We used the archived per-pair NREM features (results/pairs.tsv: log_hfa, mi, log_delta, soz) of 43 patients as the only input, and re-analysed them within and between electrode shafts.

Cited by

1 publication

Public catalog entries on Sciscape.

challengesOpenNeuro electrode coordinates do not verify the within-shaft SOZ result for 728 archived bipolar pairs
Its shaft is the contact-name prefix. Public coordinates are missing for 83% of its pairs and 92% of SOZ pairs, so the within-electrode claim is not verified; the prefix result itself reproduces.