Figure 4
Regions in which overlapping activation were observed between the training (n-back) and the transfer task (PASAT). (a) Conjunction analysis between n-back (contrast: 2-back + 3back > 0-back) and PASAT (contrast: add > repeat). The shared areas activated during the performance of both tasks are colored in red. A corrected FDR threshold of p-value < 0.05 (cluster correction, criterion: 45 voxels of extension) and an uncorrected threshold of p-value < 0.001 (voxel level) were used. (b) Overlap of the n-back general task activation in the trained group in S1 with the decreased activation found in the DLPFC during PASAT performance in S2 (the same as the Fig. 3a). The yellow scale represents the contrast: 2-back + 3back > 0-back, and the blue scale represents the contrast: Trained group (pre-training > post-training) > Control group (pre-training > post-training). A corrected FDR threshold of p-value < 0.05 (cluster correction, criterion: 1500 voxels of extension for n-back and 113 voxels of extension for PASAT) and an uncorrected threshold of p < 0.005 (voxel level) for n-back and a p-value < 0.001 (voxel level) for PASAT were utilized. (c) Overlap of the n-back general task activation in the trained group in S1 with the decreased activation found in the DLPFC during PASAT performance in S3 (the same as the Fig. 3b). Therefore, the yellow scale represents the contrast: 2-back + 3back > 0-back, and the blue scale represents the contrast: Trained group (pre-training > follow-up) > Control group (pre-training > follow-up). We employed a corrected FDR threshold of p-value < 0.05 (cluster correction, criterion: 1500 voxels of extension for n-back and 99 voxels of extension for PASAT) and an uncorrected threshold of p < 0.005 (voxel level) for n-back and a p-value < 0.001 (voxel level) for PASAT).
