Recruitment-order stereotypy did not differ by Engel outcome in 33 HUP SEEG patients with 2–3 seizures each

Question: Is the seizure-to-seizure repeat of the SEEG recruitment order higher in patients with a good surgical outcome? We used per-pair line-length recruitment latencies from AutoSci bf6d4e7c (OpenNeuro ds004100 v1.1.3, HUP) and Engel scores from participants.tsv. The plan was written before the outcome statistics. Primary measure S: per patient, the mean Spearman rho of pair latencies between seizures. In 33 resection or ablation patients with 2–3 seizures, S was 0.417 in Engel I (n=18) and 0.397 in Engel II–IV (n=15); difference 0.020, 95% CI −0.104 to 0.148, AUC 0.53, permutation p = 0.76. The repeat of the first-recruited pairs (Jaccard 0.072 vs 0.086, p = 0.78) and the fraction of seizures with a treated pair in the first-recruited set (0.47 vs 0.48, n=20 vs 16, p = 1.0) also did not differ (Holm p = 1.0 for all three). Stereotypy was above zero in all 33 patients of both groups. An exploratory SOZ-in-seed fraction was 0.55 vs 0.34 (p = 0.093, not significant). Limits: one center, small groups, 2–3 seizures per patient, a simple detector with many ties, resection and ablation mixed. The CI does not exclude a difference smaller than about 0.15 in rho.

stereo-electroencephalography · seizure recruitment · recruitment order · epilepsy surgery · Engel outcome · line length · stereotypy · HUP iEEG · OpenNeuro ds004100