In 9 epilepsy patients, amygdala SEEG high-gamma power showed no reliable difference between aversive and neutral 24-s film blocks
I used public SEEG data from 9 epilepsy patients (OpenNeuro ds003374, version 1.1.1, Zurich). Each patient watched 17 blocks of 24 s. 8 blocks showed fearful faces from films (aversive). 9 blocks showed landscapes (neutral). I wrote the analysis plan before I calculated the statistics. The primary test was the aversive minus neutral difference of log10 high-gamma power (70-145 Hz) in the amygdala macro contacts, with an exact sign-flip test over patients. The difference was -0.0069 log10 units (-1.6%), 95% CI -0.0197 to 0.0098, 3 of 9 patients positive, p = 0.43. Theta power (4-8 Hz) did not change (+3.1%, p = 0.80, Holm p = 1.0). The aversive-block HG response did not decrease across the session (slope p = 0.78, Holm p = 1.0). Bipolar contacts (local signal) gave +4.2%, p = 0.22, Holm p = 0.66. Thus, at the level of 24-s block averages, this data does not show a reliable amygdala HG or theta response to aversive films. The data cannot exclude monopolar HG increases smaller than about 2.3%, bipolar HG increases smaller than about 10%, short transient responses, or responses in single neurons. Faces and landscapes differ in many visual features, so a positive result would also not be specific to fear. The package has the code, the plan, the input hashes, the results and the steps to reproduce them.
amygdala · SEEG · intracranial EEG · high-gamma power · emotion · epilepsy · block design · OpenNeuro
References
2- usesdoi:10.18112/openneuro.ds003374.v1.1.1
The analysis used the Zurich amygdala SEEG files and task metadata from OpenNeuro ds003374 snapshot 1.1.1. The input file hashes and source URLs are preserved in the package.
- doi:10.1016/j.neuroimage.2020.116705
Cited by
7 publicationsPublic catalog entries on Sciscape.
- supportsIn 42 amygdala units from 5 patients, firing was not reliably higher in 24-s fearful-face than landscape film blocks
Same dataset and block contrast. The unit-level block-average firing is also not reliably different between aversive and neutral blocks, which agrees with the high-gamma null of that work.
- usesIn 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
EDF loader code, artifact rule and input hash list for ds003374 v1.1.1
- usesFive of 42 amygdala units fired more during aversive blocks, but patient-level change and firing–high-gamma coupling were not detected
Earlier analysis of ds003374 macro contacts. We use the same block labels and add the unit data.
- usesIn 9 epilepsy patients, amygdala alpha, beta and 3-45 Hz slope showed no group differences between fearful-face and landscape blocks; alpha and beta changed in opposite directions across patients
Extends the block-average high-gamma and theta test of this publication to alpha, beta and the aperiodic slope on the same data and contrast.
- supportsNine-patient amygdala SEEG showed no detectable block-onset high-gamma or ERP response and no aversive–neutral difference in the first 2 s
The first 2 s after block onset also show no aversive-vs-neutral difference and no onset response, so the null block-average result is not explained by a diluted onset transient.
- usesNine-patient amygdala SEEG showed no detectable block-onset high-gamma or ERP response and no aversive–neutral difference in the first 2 s
This work tests the limit that 983fd1a0 stated (short transient responses that 24-s block averages cannot show), with the same dataset, channels and condition labels.
- supportsAmygdala high-gamma power showed no reliable within-block decrease or aversive–neutral difference in 9 patients
Block-average high-gamma showed no condition difference; our within-block change also shows no condition difference.
- supportsIn 9 patients, the amygdala aversive-minus-neutral high-gamma difference did not decrease over 8 repeated film blocks
In a different window and with a bipolar pair, our descriptive aversive-minus-neutral HG mean (+0.046 log10) also did not reach p < 0.05 in an exact sign-flip test over 9 patients (p = 0.078).