In 9 patients, the amygdala aversive-minus-neutral high-gamma difference did not decrease over 8 repeated film blocks

We tested for habituation of the amygdala high-gamma (HG, 70-145 Hz) response to aversive film blocks in OpenNeuro ds003374 v1.1.1 (9 patients, 1 session each). We wrote and hashed the plan before we read any signal. For each side, we used the bipolar macro pair mA1-mA2 and one window per block (onset +1 s to +23 s), so that the window has no zero-filled gap samples. HG was the log10 mean Welch PSD over 70-145 Hz, without 95-105 Hz. D_k was the HG of aversive block k minus the HG of the neutral block just before it (k = 1..8). We calculated the slope of D_k against k for each side and averaged the slopes for each patient. The primary test (H1: mean slope below zero) gave a mean of -0.0039 log10 per block (bootstrap 95% CI -0.020 to 0.010), 3 of 9 patients negative, exact one-sided sign-flip p = 0.35. Thus the data gave no evidence of habituation. The CI excludes a mean decrease larger than about 0.020 log10 (about 4.7% power) per block. Secondary slopes (aversive HG, neutral HG, theta 3-8 Hz difference) were all non-significant after Holm correction (Holm p 0.66 to 1.0). The second-half-minus-first-half test agreed (p = 0.44). The descriptive mean D_HG was +0.046 log10 (CI 0.010 to 0.086, two-sided sign-flip p = 0.078). Limits: n = 9 gives low power; the fixed block order confounds repetition with time in session; there was no anatomy check of contacts; macro HG can include broadband and muscle components. One naming deviation: 7 of 9 EDF headers label contacts 'mmA..', and the code maps them to the 'mA..' names in channels.tsv. The package has the plan and its hash, the numpy-only code, S3 listing with ETags, sha256 of all inputs, per-block features, results and reproduction steps.

amygdala · high-gamma · habituation · intracranial EEG · stereo-EEG · aversive film · OpenNeuro ds003374 · sign-flip test