fMRI replicability during emotional scene viewing: Functional regions and sample size

Area of Science:

• Neuroimaging
• Cognitive Neuroscience
• Affective Science

Background:

• Recent studies question the replicability of functional magnetic resonance imaging (fMRI) in affective processing research.
• Low replicability may stem from insufficient emotional engagement in tasks like face matching.

Purpose of the Study:

• To assess the replicability of emotional enhancement in fMRI during tasks with reliable emotional engagement.
• To investigate the influence of sample size on fMRI replicability in emotion studies.

Main Methods:

• Tested emotional enhancement during pleasant, neutral, and unpleasant picture viewing in a large sample (N=160).
• Employed a subsampling technique with varying sample sizes (N=20, 40, 60, 80) and 500 permutations to compute replicability.
• Analyzed replicability at peak, cluster, and voxel levels.

Main Results:

• fMRI replicability increased with larger sample sizes, consistent with prior research.
• Good to excellent fMRI replicability was observed during emotional scene viewing, even with small samples.
• Excellent replicability (cluster and peak levels) was found in the amygdala and visual cortex with N=40 at p < .001.

Conclusions:

• The degree of fMRI replicability in emotion research is contingent upon successful emotional engagement.
• Findings challenge general sample size recommendations, emphasizing task design for robust results.
• High emotional engagement ensures good fMRI replicability, particularly in key brain regions like the amygdala.