Skip to main content
. 2014 Nov 26;5:5547. doi: 10.1038/ncomms6547

Table 2. Subsequent memory (‘DM’)-related group activity after ROI-based alignment.

Cluster size Cluster Puncorr Alpha (simulated) Peak T Peak location Template x, y, z {mm} Side
Group activity for 1.5 mm smoothing
21 0.007 <0.10 4.60 pyramidal CA1 (body) 31.1, 22.1,21.1 R
33 0.001 <0.02 4.58 EC & sub (head) −14.6, 41.2, −26.0 L
16 0.017 <0.25 4.31 CA1/presub.(head) 15.1, 38.1,23.4 R
17 0.014 <0.20 4.20 CA2/3-DG (posterior body) 23.9, 14.8,12.7 R
Group activity for 2.4 mm smoothing
96 0.001 <0.01 5.94 CA1/presub. (head) 15.1, 38.1, −23.4 R
72 0.003 <0.01 5.51 EC −13.8, 41.6, −25.3 L
25 0.053 <0.20 4.90 CA2/3-DG (posterior body) 25.5, 13.7,13.1 R
44 0.014 <0.05 4.49 pyramidal CA1 (body) 31.1, 22.5, −22.2 R
37 0.022 <0.10 4.31 Sub/CA1 (body) 22.3, 23.5,21.9 R
57 0.006 <0.02 4.05 EC 13.5, 34.5, −26.9 R

DG, dentate gyrus; EC, entorhinal cortex; R, right; ROI, region of interest L, left.

Functional data were smoothed with two different kernels to allow for higher specificity and higher sensitivity, respectively. Hippocampal and entorhinal subregions showing group-level activation (at Pvoxel level<0.005; k1.5 mm≥15 voxels and k2.4 mm≥25 voxels, N=19) for the DM contrast (difference due to memory: remembered>forgotten). Alphas were simulated with 3dClustSim (AFNI). Note that clusters that would not survive correction at low smoothing (highlighted in italics; for example, pyramidal layers in CA1) become significant with higher smoothing.