After baseline subtraction, spectral change magnitudes had more repeatable local order after marked onset in 26 selected SEEG patients

We tested local repeat order of spectral change magnitudes in two seizures per patient. The source was OpenNeuro ds004100 version 1.1.3. The study reused 26 selected SEEG patients. Each score compared 20–50 Hz power with 2–12 Hz power after baseline subtraction. The same 7,597 comparisons were used before and after marked onset. Each comparison used one name group and one pure clinical label class, with no shared contact. The equal-patient mean order was 0.026 before onset and 0.210 after onset. The difference was 0.184, with a 95% patient bootstrap interval from 0.086 to 0.286. The two-sided sign reversal p value was 0.001615, under patient independence and symmetric null differences. Without baseline subtraction, the difference interval included zero. The SOZ class estimate was less precise. The result measures repeat agreement, not clinical accuracy. The package preserves code, all feature inputs, source provenance, complete results, and reproduction instructions.

stereoelectroencephalography · seizure · spectral power · repeat order · baseline subtraction · epilepsy