In five bilateral-amygdala SEEG patients, theta coherence did not change between aversive and neutral film blocks; a small high-gamma increase did not pass correction

We calculated magnitude-squared coherence (MSC) between one bipolar channel in the left amygdala and one in the right amygdala in five patients with bilateral implants from OpenNeuro ds003374 v1.1.1. Each patient contributed eight aversive and nine neutral 24-s blocks. The primary theta (4–8 Hz) test, fixed before analysis, showed no effect: mean Fisher-z difference (aversive minus neutral) −0.021, one of five patients positive, group permutation p = 0.14, and exact sign-flip p = 0.125. No patient had a within-patient permutation p below 0.23. In the secondary high-gamma band (70–145 Hz, with 95–105 Hz removed), the difference was positive in all five patients and remained positive after centering for the unequal block counts (mean centered difference 0.005). The uncorrected group permutation p was 0.019; the Holm-adjusted p was 0.075. Imaginary coherency in high gamma was also positive in all five patients (sign-flip p = 0.0625). All high-gamma MSC values were low (0.002–0.004), near the estimated floor of about 1/K with K about 380 windows per condition. With five patients, the sign-flip test cannot produce p < 0.0625. These results support no detected theta change and treat the high-gamma pattern as a hypothesis, not a finding. The study is limited to one dataset and session; alternating block order, uncontrolled slow drift and film content, and no anatomical contact-position check beyond dataset labels limit interpretation.

amygdala · intracranial EEG · coherence · theta · high gamma · aversive stimuli · film task · OpenNeuro