Memory Guided Visual Search (MGVS)

# Memory-guided visual search (MGVS) See dataset_description.json for authors, funding, and citation information. ## Overview This dataset contains fMRI, eyetracking, and behavioral data from a study of memory-guided visual search. Participants first learned scene-location associations (ses-train), then performed a visual search task in the scanner in which memory for these associations could be used to predict the location of an upcoming search target (ses-01). A separate population receptive field (pRF) retinotopic mapping session is also included under ses-01. Tasks: - retinotopy: pRF bar-sweep mapping (ses-01, func) - search: memory-guided visual search (ses-01, func) - expose: passive scene re-exposure (ses-train, beh) - learn: scene-location association study/test (ses-train, beh) ses-train is a behavior-only session (no MRI) and is stored under the "beh" datatype rather than "func". Participant demographics (age, sex, handedness, race, ethnicity, years of education) are in participants.tsv/participants.json. ## Eyetracking Raw EyeLink eyetracking data (.asc text format) are included under sourcedata/eyetracking/, named with BIDS-style session/task/run labels for traceability. These are not BIDS Eye-Tracking (BEP020) formatted and are placed in sourcedata/ (exempt from BIDS validation) accordingly. Each task's trial-phase (xDAT) and gaze-target (TarX/TarY) message codes, embedded as EyeLink MSG lines in the .asc files, are documented in the accompanying task-<label>_eyetrack.json files in the same directory. ## Fieldmaps Subjects sub-01 and sub-03 were acquired with reversed phase-encoding (PEPOLAR) spin-echo EPI fieldmaps (dir-AP/dir-PA), while all other subjects were acquired with a gre_field_mapping (phase-difference) fieldmap. ## Stimuli stimuli/ contains the base (unwarped) scene images (stim/), the warped versions used during the search task (stim_warped/), and the retinotopic mapping bar-aperture frames (stim_retinotopy/). Some base images in stim/ are only ever presented in warped form and are therefore not directly referenced by any events.tsv; a small number of warp-variant combinations in stim_warped/ likewise happen not to have been drawn for any of the 32 released subjects. ## Derivatives derivatives/fmriprep/ includes only the T1w (native) and fsnative output spaces; MNI152NLin2009cAsym and fsaverage outputs are not included to reduce dataset size. The per-subject fMRIPrep visual reports (sub-XX.html), however, are fMRIPrep's standard, unmodified output and still show QC figures for all four output spaces fMRIPrep computed internally, and for a task-localizer scan. This localizer scan was collected opportunistically when scan time allowed, was not used in the associated manuscript, and is not included anywhere else in this dataset or its derivatives. Aside from appearing in these reports, it can be disregarded. derivatives/mriqc/ includes per-subject/session image quality metrics and reports; group-level summary files are not included.