The mean 6 dB spectral order score was 0.12 within label classes across two seizures in 18 SEEG patients

We tested whether a fixed spectral rise rule gave the same local order across two seizures in public SEEG from OpenNeuro ds004100 version 1.1.3. Both compared bipolar pairs had a rise in both seizures, the same clinical onset label, the same shaft, and no contact in common. Time ties stayed in the score. The 6 dB rule required two consecutive one-second blocks. The mean patient score was 0.124 in 18 of 26 source patients. Its 95% patient resampling interval was -0.086 to 0.318. The conditional upper-tail permutation p value was 0.07625. This test assumes that the joint time law permits separate time permutations within the selected groups. The primary result leaves substantial order uncertainty under that model. It does not show absent seizure spread. All planned checks are reported with their different selected samples. The package preserves the local plan, code, features, comparisons, draws, source identities, and repeat steps. All 52 fresh source ranges matched their archived hashes. Every source feature array and a separate analysis repeat matched exactly. This is an informed follow-up on a reused cohort, not an independent clinical replication. This version corrects one provenance sentence about the zero-score patient. All numerical results are unchanged.

stereoelectroencephalography · seizure timing · spectral rise · event order · repeatability · permutation test · epilepsy