Contact:	Vikranth Rao Bejjanki (bejjanki@hamilton.edu)

Introduction: This dataset contains the fMRI data used to carry out the analyses described in: Vikranth R. Bejjanki, Rava Azeredo da Silveira, Jonathan D. Cohen, & Nicholas B. Turk-Browne, "Noise correlations in the human brain and their impact on pattern classification. The data included here is a subset of data from a prior fMRI study on attentional control: Tompary A, Al-Aidroos N, Turk-Browne NB. "Attending to what and where: Background connectivity integrates categorical and spatial attention (under revision).

Task: Each participant completed two face/scene localizer runs, each of which consisted of an alternating on-off block design, with 18-s blocks of stimulation interleaved with 18-s blocks of blank passive fixation. Stimulation blocks contained 12 1-s presentations of either face or scene images (the order of face and scene blocks was counter-balanced across participants), each separated by a 500-ms inter-stimulus interval. Images were presented in grayscale, cropped using a circular mask, and subtended 6 of visual angle in radius. In one run, face and scene stimuli were presented in the left visual field, and in the other run, face and scene stimuli were presented in the right visual field. Each run began with a 9-s fixation period and included a total of 12 blocks of stimulation (6 face, 6 scene), which lasted 7m 21s. During blank periods, participants were presented only with a central, white point to fixate (radius = 0.2). Data from two rest runs were also collected for each participant. Each rest run had the same duration as the localizer runs, but with only the central fixation point. Participants were instructed to passively view the fixation point without performing any overt task.

Image Acquisition: This dataset includes data from 17 human participants, acquired with a 3T scanner (Siemens Skyra) using a 16-channel head coil. Functional images for both the localizer and rest runs were acquired with a T2* gradient-echo echo-planar imaging sequence (TR = 1.5 s). Anatomical images were acquired using a T1-weighted MPRAGE pulse sequence. Please see the Methods section of the paper for further details. 

File Descriptions: For each participant, the dataset includes raw functional data corresponding to the two localizer runs and the two rest runs, as well as high-resolution (skull-stripped) anatomical data from the MPRAGE sequence. Participant_# refers to the set of individual participant data. Since data from 17 participants is included, and there are 5 files for each participant, the dataset includes a total of 85 files.

Note from Neggin Keshavarzian, Data Curator, neggink@princeton.edu: All identifiable information in this dataset has been removed; facial  information removed via brain extraction, and other identifiable metadata removed from the NIFTI headers.