Fig. 5: Amygdala theta power correlates with item stability and context specificity across the extinction network.
From: Representational dynamics during extinction of fear memories in the human brain
a, Left: AMY theta power was averaged within the time–frequency cluster showing significant differences between CS+ and CS− items during extinction (highlighted in yellow). This average theta power was correlated across trials with two measures: item stability (middle) and context specificity (right). These metrics were either averaged during the same time period of significant AMY theta power effects (b,c), or in sliding windows across all time periods in each trial (d,e). Group-level rho values were contrasted against zero using two-sided t-tests in all analyses. b, AMY theta power was correlated across trials with item stability in both AMY (left; pcorr < 0.001; N = 32) and HPC (right; pcorr < 0.001; N = 24). c, AMY theta power was also correlated with lPFC context specificity across trials (pcorr = 0.005; N = 14). d, Time-resolved analysis. Left: AMY theta power and item stability were correlated throughout the entire cue presentation period (pcorr = 0.005). Right: correlations between AMY theta power and HPC item stability were observed during a later time window of cue presentation (600 ms–1.65 s; pcorr = 0.005). e, AMY theta power was correlated with lPFC context specificity during a time period that overlapped with the significant condition differences in theta power in AMY (900 ms–1.5 s; pcorr = 0.005). In d and e, shaded lines depict group average rho values ± s.e.m., and time zero marks the onset of cue presentation during the extinction phase. Blue horizontal lines indicate time periods of statistical significance after correction for multiple comparisons using cluster-based permutation statistics. In a, the dotted vertical line represents cue onset (time zero).
