Fig. 3: Item stability in lateral temporal cortex, amygdala and across the extinction network.
From: Representational dynamics during extinction of fear memories in the human brain
a, Item stability, defined as the average similarity of neural patterns representing individual items across repeated presentations, was assessed separately for CS+ and CS− trials during acquisition and extinction. b, Analysis of item stability in the AMY revealed no significant differences between CS+ and CS− trials during acquisition (left). During extinction (right), item stability was significantly higher for CS+ than for CS− items in a late time period during cue presentation (1.2–1.5 s after cue onset; pcorr = 0.018). c, Item stability during extinction in the significant time period shown in b (right) for CS++ and CS+− (left; p = 0.028), CS++ and CS−− (middle; p = 0.023), and CS+− and CS−− trials (right; p = 0.2) (N = 32 in all analyses). d, Analysis of item stability in the TMP revealed no significant differences between CS+ and CS− trials during acquisition (left). During extinction (right), item stability was significantly higher for CS+ than for CS− items in the time period from 0.65–1 s after cue onset (pcorr = 0.005). e, Item stability during extinction in the time period shown in d (right) for CS++ and CS+− (left; p = 0.018), CS++ and CS−− (middle; p = 0.11) and CS+− and CS−− trials (right; p = 0.61) (N = 41 in all analyses). f, Top left: a single-trial metric of item stability was computed by assessing the similarity of each trial across repeated presentations (average of the comparisons highlighted with a red rectangle in the RSA matrix). Bottom left: fluctuations in item stability were coordinated across trials between the TMP and the AMY during the first second of cue presentation and towards the end of the cue presentation period. Right: item stability was computed in each ROI and correlated across trials with the values observed in TMP (top row) and AMY (bottom row). In b and d, thick black horizontal lines depict significant time periods after correction for multiple comparisons using cluster-based permutation statistics. Time zero indicates the onset of the CS, and shaded lines depict group average rho values ± s.e.m. Two-sided paired t-tests across conditions were applied at every time point (b,d) or every time-by-time point (f) to assess statistical significance. In c and e, two-sided t-tests against zero were conducted. In f, significant regions surviving multiple comparisons correction using cluster-based permutation statistics are outlined in black. Time zero indicates the onset of the CS in both time axes. AMY, amygdala; HPC, hippocampus; lPFC, lateral prefrontal cortex; OFC, orbitofrontal cortex.
