fast fMRI - object dataset

# Data of main experiment for the fast fMRI project This BIDS dataset contains data collected at 7T in 4 subjects, over three sessions each. Each subject was shown naturalistic image stimuli from the THINGS dataset at three different presentation rates (0.5s, 1s, 4s), with a fixed (0.35s) or relative (75% of SOA) stimulus duration. Each run used one SOA and one duration type. Runs included a 6s ramp-up and 10s cool-down. Participants performed a fixation color-change task. TR=0.5s. We used a total of 642 stimuli, with 55 test stimuli that were shown 6 times in the 4s condition, 24 times in the 1s condition and 48 times in the 0.5s condition. The remaining 587 stimuli were shown 1 time, 4 times and 8 times in the 4s, 1s and 0.5s conditions, respectively. ## Notes Session 1 additionally contains a structural T1w image and a whole-brain single-volume reference EPI (in `func`). Sessions 2 and 3 contain two fLoc localizer runs each (in `func`). After each functional run, short PA EPI fieldmap data were collected for distortion correction. JSON sidecars include the presentation rate per run (key: `SOA`). `participants.tsv` contains which subjects used fixed vs relative durations; this can also be inferred from the stimulus durations in the `events.tsv` files. ## Acquisition summary - TR: 0.5s, TE: 22ms, FA: 40°, multiband: 4, GRAPPA: 2 - 40 interleaved slices; partial FoV covering occipital and temporal cortex - Fixation color-change task (0.5s red flashes, random 2–7s intervals) - No null trials ## Derivatives - `derivatives/basic_preproc_v2` custom preprocessing (motion correction, distortion correction with PA EPIs, alignment to T1w and MNI). Outputs include space-T1w and space-MNI preprocessed BOLD, masks, and transforms. - `derivatives/floc_preprocessed` preprocessing outputs and events for fLoc runs. - `derivatives/single_trial_betas` single-trial beta estimates (LSS) and analysis outputs. - `derivatives/decoding` pairwise decoding accuracy maps per SOA. - `derivatives/spose_rsa_sl` searchlight RSA (SPoSE/object-dimensions) results. - `derivatives/localizer` localizer-based maps/ROIs derived from the main experiment. - `derivatives/atlases` auxiliary atlases and masks (Harvard-Oxford, Neurosynth LOC, object parcels).