In 35 HUP SEEG patients, SOZ pairs showed no significant change versus non-SOZ pairs before marked seizure onset

We tested if seizure-onset-zone (SOZ) channels change before the clinician-marked seizure onset. Data: OpenNeuro ds004100, 35 SEEG patients, 55 seizures, window -120 s to +10 s, read with HTTP byte ranges. We used within-shaft bipolar pairs and log power in 5 bands and line length in 5-s bins. Per seizure, the contrast C is the median change from the [-120,-60) s baseline at SOZ pairs minus that at non-SOZ pairs. We tested patient means with a sign-flip test. The plan was written before the analysis. Primary result: 30-100 Hz in [-10,0) s gave mean C +0.067 log10 (95% CI -0.000 to +0.148), p = 0.089, Holm p = 0.27. In [-40,-10) s and [-60,-40) s, C was -0.003 and -0.004 (CI within +-0.036), p > 0.8. No other feature gave a pre-onset SOZ effect at p < 0.05. The positive control [0,10) s gave C +0.46 (32/35 patients, p = 1e-5). Thus, the [-40,-10] s period is a valid SOZ-neutral baseline for ictal energy-ratio markers in this dataset. Limits: only 57 of 104 queued runs were fetched because of the run time limit; SOZ and onset labels are clinical.

stereo-EEG · seizure onset zone · epilepsy · pre-ictal activity · baseline · power spectrum · line length · OpenNeuro · sign-flip permutation test