Recruitment latency increased modestly with distance, and recruitment order repeated across seizures in 35 patients

We tested whether stereo-electroencephalography (SEEG) recruitment latency increases with distance from the first-recruited site and whether recruitment order repeats across seizures. The analysis used a previously derived per-pair feature table and fsaverage contact coordinates for 94 seizures in 35 patients. Recruitment was defined by a 1-s line-length threshold above the pre-ictal mean plus 5 SD, evaluated in 0.5-s steps from -5 to 19 s relative to clinician-marked seizure onset. Mean patient-level Spearman correlation between distance and latency was 0.161 (95% CI 0.111–0.213; 28/35 positive; Holm-adjusted p = 2e-5). Across repeated seizures, recruitment latencies were correlated (mean rho 0.403, 95% CI 0.337–0.466; all 31 patients positive; Holm-adjusted p = 2e-5), more than a geometry-only control (mean rho 0.172; exploratory difference 0.231, p = 3e-5). The Theil-Sen latency-distance slope had a median of 0 s/mm (95% CI 0–0.028), so these data do not identify a spread speed; the upper slope confidence bound gives a lower speed limit of about 36 mm/s for the median patient. Results are limited by template coordinates, Euclidean distances, clinician onset marks, and a coarse threshold detector. The exact OpenNeuro release of the mirrored electrode files was not recorded.

stereo-electroencephalography · SEEG · epilepsy · seizure · recruitment latency · ictal spread · electrode distance · stereotypy · line length · OpenNeuro ds004100