MR-related brain activity depends on cognitive traits. The brain activation clusters resulting from the whole-brain analysis are projected onto a 3D brain (A, E, I) and axial slices (B, F, J). (A and B) Brain regions where the activity was significantly stronger in object- than spatial-visualizers during MR of geometric objects (orange) and human bodies (purple). (C) Spatial-visualizers were more accurate than object-visualizers. (D) MR in spatial-visualizers was faster than in object-visualizers. (E and F) Brain regions more active in high- than low-scoring object-visualizers during MR of geometric than real objects (red), and in low- than high-scoring spatial-visualizers during MR of geometric objects than human bodies (blue). (G) The mean MR accuracy was worse in high- than low-scoring object-visualizers. (H) The mean MR accuracy was better in high- than low-scoring spatial-visualizers. (I and J) brain regions where the activity during MR of human bodies versus geometric objects was stronger in participants with low than high scores in perspective-taking. (K) The accuracy for MR was higher for high-scoring perspective-taking than for low-scoring perspective-taking. All the statistical maps of the figure were assessed with a cluster-based threshold of Z > 2.3 corrected for family-wise error (FWE) at P = 0.05.