AOMIC

The AOMIC dataset includes two subsamples, PIOP1 and PIOP2, comprising data of healthy university students scanned on a Philips 3T scanner. Participants were instructed to keep their gaze fixated on a fixation cross on the screen and let their thoughts run freely (^{Snoek et al., 2021}). Both samples were acquired with a voxel size of 3 × 3 × 3 mm³ and a matrix size of 80 × 80. While PIOP1 was acquired for 360 seconds with multi-slice acceleration, 480 volumes and a 0.75 TR, PIOP2 was acquired for 480 seconds without multi-slice acceleration, 240 volumes and a 2s TR (further details in https://www.nature.com/articles/s41597-021-00870-6/tables/10). Data were preprocessed using Fmriprep version 1.4.1 (^{Esteban et al., 2019; Esteban et al., 2020}), a Nipype based tool for reproducible preprocessing in neuroimaging data (^{Gorgolewski et al., 2011}). Data were motion corrected using mcflirt (FSLv5.0.9, (^{Jenkinson et al. 2002})) followed by distortion correction by co-registering the functional image to the respective T1 weighted image with inverted intensity (^{Huntenburg, 2014; Wang et al., 2017}) with 6 degrees of freedom, using bbregister (FreeSurfer v6.0.1). In a following step, motion correction transformations, field distortion correction warp, BOLD-to-T1- weighted transformation and the warp from T1-weighted to MNI were concatenated and applied using antsApplyTransforms (ANTs v2.1.0.) using Lanczos interpolation (^{Snoek et al., 2021}).