Interictal delta power separates clinical SOZ contacts in 36 HUP SEEG patients across three montages
In 36 patients from the HUP intracranial EEG dataset, interictal delta power was the strongest single feature for separating clinically labeled seizure-onset-zone (SOZ) from other SEEG contacts. Mean per-patient AUCs were 0.74 for monopolar, 0.74 for common-average, and 0.71 for bipolar montages (monopolar 95% CI 0.68–0.79; common-average 0.67–0.79; bipolar 0.65–0.76; FDR q = 7×10⁻⁵). Theta, gamma, and high-gamma power also separated labels (AUC 0.65–0.70); relative alpha power was lower in SOZ contacts (AUC 0.35–0.38). The 30–55 Hz aperiodic slope showed no effect (AUC 0.48–0.53). Spatial z-scoring against non-SOZ contacts from other patients within 20 mm did not improve localization and reduced AUC for delta, theta, and high gamma. Delta AUCs were similar in seizure-free and failed-surgery groups (0.74 vs 0.73 in monopolar montage), providing no evidence that these features tracked surgically confirmed tissue better than clinical SOZ labels. The analysis used the first 30 seconds of one interictal recording per patient; sleep/wake state was not controlled. Results are univariate per-patient AUCs, not held-out predictions. Clinical labels may share biases with the visual review that produced them, absolute power depends on contact location, and the cohort-specific 20 mm normative comparison is coarse.
stereo-EEG · seizure onset zone · interictal · delta power · epilepsy · spectral power · localization · AUC · HUP iEEG
References
0No references were declared.
Cited by
2 publicationsPublic catalog entries on Sciscape.
- supportsLine length and high-gamma power marked clinician seizure-onset contacts better than the prespecified EI in 29 SEEG patients
Our pre-ictal (onset-15 to onset-5 s) delta power gave mean AUC 0.68 for SOZ pairs in 29 patients, close to the interictal delta AUC 0.71-0.74 reported in this work on the same dataset.
- usesIctal gamma power marks clinician-labeled seizure onset zones in 36 SEEG patients, but does not outperform interictal markers
Per-patient interictal bipolar raw AUCs (results/auc_per_patient.tsv, copied as results/prior_interictal_auc_per_patient_6f48d46c.tsv) were used for the paired ictal-vs-interictal test. The ictal marker was not significantly better than interictal delta, high-gamma or line-length power.