Under an 8–40 Hz spectral line model, a test did not detect a 6–18 s amygdala delta contrast in seven SEEG patients

We tested whether an amygdala delta decrease remained after a fixed spectral correction. We used public SEEG from OpenNeuro ds003374, version 1.1.1. The main sample had seven patients with a left amygdala pair outside the labeled seizure onset zone. We fitted a line to log power from 8 to 40 Hz in each video block. We measured delta power relative to the line prediction from 1 to 4 Hz. We used the 6–18 second window and compared each aversive block with its two adjacent neutral blocks. The corrected patient mean was −0.688 dB. Its 95% t interval was −1.893 to +0.516 dB. The exact two-sided sign-flip p value was 0.21875. Four of seven corrected contrasts were negative. Measured delta power also fell relative to 20–40 Hz power. These results do not prove zero delta change or show that a common power change explains the measured decrease. A curved background without a delta peak could also produce a corrected change. We preserve both source export derivatives, spectra, code, the fixed internal plan, source versions and hashes, all planned results, and reproduction instructions. All 18 input files rebuilt exactly from fresh source downloads. Independent calculations and full replay passed. This is an informed analysis of a reused epilepsy sample. Fixed video order, different visual content, source export differences, and the line assumption limit the result.

amygdala · delta power · intracranial EEG · spectral analysis · epilepsy · sign-flip test