Associative Prediction of Visual Shape in the Hippocampus

This dataset consists of fMRI data from 24 participants that were exposed to an experiment consisting of auditory cues and abstract visual shapes. The experiment consisted of 6 functional runs (~13 minutes per run), described below.

In two out of six runs (run 1 and run 6), one of five shapes was presented on each trial, in a pseudorandom manner. In the other four runs, auditory cues (falling or rising tones, respectively) predicted which shape (either shape 2 or shape 4) would be presented on that trial, with 75% validity. In two of the four prediction runs, an unexpected shape was presented on the 25% invalid trials (e.g. shape 2 was predicted, but shape 4 presented). In the other two runs, no shape at all was presented on the 25% invalid trials (omission trials).

All imaging data were acquired using a 3T Siemens Prisma MR scanner, with a 64-channel head coil. Functional images were obtained with a multiband EPI sequence: TR = 1000 ms; TE = 32.6 ms; 60 transversal slices; voxel size = 1.5 x 1.5 x 1.5mm; 55 degree flip angle; multiband factor = 6. Additionally, anatomical T1 (TR = 2300 ms;
TE = 2.27 ms; voxel size= 1 x 1 x 1 mm; 192 transversal slices; 8 degree flip angle; GRAPPA acceleration factor 3) and T2 (TR = 11,390 ms; TE = 90 ms; voxel size = 0.44 x 0.44 x 1.5mm; 54 coronal slices, perpendicular to the long axis of the hippocampus; distance factor = 20%; 150 degree flip angle) weighted images were acquired. To correct for susceptibility-induced distortions in the echoplanar images, a pair of spinecho volumes was acquired in opposing phase-encode directions (anterior/posterior and posterior/anterior) with matching slice prescription, voxel size, field of view, bandwidth, and echo spacing (TR = 8000 ms; TE = 66 ms).