Interictal spike-rate maps agreed with early ictal gamma maps but did not predict Engel I outcome in 36 HUP SEEG patients

This pre-registered analysis tested whether label-free agreement between interictal pair maps and early ictal 30–100 Hz power-rise maps was higher in patients who became seizure-free (Engel I) after resection or laser ablation. It used archived feature tables for 36 SEEG patients (20 Engel I, 16 Engel II–IV), 103 seizures, and a mean of 100 bipolar pairs per seizure. The interictal spike-rate map showed positive concordance with the ictal map (mean per-patient Spearman rho +0.110, 95% CI +0.043 to +0.179; 26/36 positive; Holm p = 0.015). High-gamma power, delta power, and line length did not show corrected evidence of concordance. The primary outcome comparison was not significant: mean concordance was +0.129 in Engel I and +0.085 in Engel II–IV patients (difference +0.044, bootstrap 95% CI −0.090 to +0.176, AUC 0.55, therapy-stratified permutation p = 0.54; Holm p = 0.86). The second interictal clip gave a difference of +0.089 (95% CI −0.030 to +0.210, AUC 0.64, p = 0.18); the non-SOZ control gave −0.018 (AUC 0.47, p = 0.75). All eight outcome differences across four features and two clips were positive, but no test passed correction. The sample had low power for moderate effects and could detect only large effects (d near 1); a small positive relation (AUC below 0.65) is not excluded. The analysis does not establish prediction of surgical outcome.

stereo-EEG · epilepsy · interictal spikes · ictal gamma · surgical outcome · map concordance · Spearman correlation · Engel outcome