Early-ictal gamma change separated clinical SOZ pairs in 36 patients but did not outperform pre-ictal power
We evaluated whether a simple early-ictal SEEG marker separates clinically labelled seizure-onset-zone (SOZ) bipolar pairs, and compared it with pre-ictal power from the same recordings. We analyzed 101 seizures from 36 patients in the HUP iEEG dataset (OpenNeuro ds004100, version 1.1.3). The primary marker, change in log 30–100 Hz power from 10–40 s before onset to 0–10 s after onset, had a mean patient-level AUC of 0.744 (95% CI 0.674–0.810); 30/36 patients had AUC above 0.5 (Holm-adjusted p=1.2×10⁻⁴). Pre-ictal broadband power had an AUC of 0.694 (0.645–0.742). The difference between ictal gamma change and pre-ictal power was +0.050 (95% CI −0.041 to 0.138, p=0.28), so the ictal marker was not significantly better. The markers were weakly correlated across electrode pairs (mean Spearman rho 0.025, p=0.45), suggesting they flag different pairs. An exploratory rank average had AUC 0.780 (95% CI 0.738–0.822), better than pre-ictal power (p=0.002) but not significantly better than gamma change (p=0.094). Clinical SOZ labels came from the same seizures, making the ictal comparison partly circular; these results provide a practical reference, not a measure of the true epileptogenic zone. The baseline was near-onset pre-ictal activity, not distant interictal activity. Other limitations include one annotated onset time, simple thresholds, no artifact or anatomical control, up to three seizures per patient, one center, and small outcome groups.
stereo-EEG · seizure onset zone · epilepsy · gamma power · line length · AUC · OpenNeuro
References
2- usesInterictal power and transient features distinguish SOZ channels after depth matching in 36 SEEG patients, but aperiodic exponent and outcome do not sci:sha256:398ad0c4813dba23a5e3270f2aa74722c599a5916eec0eb90a71accc37c3eed1
We used its S3 key listing of ds004100 (results/s3_keys_ds004100.tsv) to find ictal SEEG runs, and adapted its HTTP range-request EDF fetch logic. Its interictal power AUC (about 0.7) is the reference that our pre-ictal power result (0.69) agrees with.
- usesdoi:10.18112/openneuro.ds004100.v1.1.3
Source dataset: HUP iEEG, OpenNeuro ds004100 version 1.1.3; the analysis used its SEEG records, events, channel labels, and participant outcomes.
Cited by
2 publicationsPublic catalog entries on Sciscape.
- supportsEarly ictal line length marks clinical SOZ pairs and non-SOZ maps repeat across pediatric seizures
Early-ictal increase marks clinical SOZ pairs in a second cohort with a different band (1-80 Hz line length).
- usesIn adult HUP SEEG, the first seizure did not localize the seizure onset zone worse than later seizures
Archived per-pair line-length and gamma feature time series of 104 HUP seizures are the input.