Exclusion of 55 to 65 Hz changed gamma onset AUC by at least 0.05 in 14 of 35 SEEG patients when higher power was the onset score

We tested a fixed gamma power score in public SEEG from OpenNeuro ds004100, version 1.1.3. Each of 35 patients had two 60-second interictal excerpts. The cohort had 4,149 retained contacts. Higher absolute power was the fixed score for a clinical seizure onset contact. Exclusion of 55 to 65 Hz from the 30 to 80 Hz sum changed patient AUC by at least 0.05 in 14 patients. Changes ranged from -0.238 to +0.405. The median paired change was +0.0195. Its 95 percent patient bootstrap interval was 0.0000 to +0.0625. The exact two-sided sign test gave p = 0.0801433. This did not meet the fixed p < 0.05 threshold. The primary test did not establish a consistent improvement. All 72 source excerpts passed fresh MNE and SciPy spectrum checks. The package preserves input spectra, code, source metadata, scores, patient results, bootstrap draws, hashes, and retrieval instructions. The interval removes biological bandwidth as well as possible electrical noise. The result concerns this reused cohort and these short files. It does not show clinical benefit or a general noise removal rule.

stereoelectroencephalography · SEEG · epilepsy · gamma power · spectral power · seizure onset zone · AUC · contact ranking · 60 Hz · frequency bands · Welch spectrum · OpenNeuro