Pre-ictal thalamus–SOZ coherence was not reliably higher than coherence with non-onset cortex in ds007445

We tested a pre-registered question using OpenNeuro ds007445 v1.0.4: in the 60 s from 70 s to 10 s before seizure onset, was coherence between ipsilateral thalamic bipolar pairs and seizure-onset-zone (SOZ) pairs higher than coherence between the same thalamic pairs and ipsilateral non-SOZ grey-matter pairs on other shafts? For each patient we analyzed the first eligible seizure and, separately, the second eligible seizure as a replication. We calculated magnitude-squared coherence (MSC) and absolute imaginary coherency (|iCoh|) in delta, theta, alpha, beta, and gamma bands. In the first seizures (n=18), the mean patient-level MSC difference (SOZ minus control) was positive in all five bands, but no band passed Holm correction (Wilcoxon raw p=0.048–0.42; Holm p≥0.24). For |iCoh|, the lowest Holm p was 0.080 in theta. In second seizures (n=16), no band passed correction (all Holm p=1.0) and median differences were near zero. AN-only and CM-only analyses also had no band passing Holm correction. In this cohort and window, pre-ictal thalamus–SOZ coupling did not identify the SOZ better than coupling with other ipsilateral cortex. The common analysis rate was 200 Hz because 9 of 19 first-seizure EDFs were recorded at 200 Hz; the planned 500 Hz rate was not used. Electrode distance was not controlled. The cohort included paediatric and young-adult patients, and recordings at 200 Hz and 2000 Hz were pooled. The package retains code, pair-level results, and byte ranges and SHA-256 hashes for the data read.

epilepsy · stereo-EEG · thalamus · seizure-onset zone · pre-ictal · coherence · OpenNeuro · ds007445