VbalFluc: A BIDS dataset of fMRI during verbal fluency and color-shape switching in 26 Chinese adults

This dataset contains task-based fMRI and structural MRI data from 26 healthy native Mandarin Chinese adult volunteers (age 18-27, 14 F / 12 M) performing three cognitive tasks: semantic verbal fluency (SemFlu), phonemic verbal fluency (PheFlu), and a color-shape switching task (CSS). Raw DICOM data were converted to NIfTI using dcm2niix and heudiconv, then preprocessed with fMRIPrep 24.0.1 on the T1w anatomical and BOLD functional images (no FreeSurfer recon-all; no field maps). First-level general linear models and second-level group analyses are provided in the derivatives/ folder, together with cortical gradient, ROI, ROI-based MVPA, generalized PPI searchlight, network correspondence toolbox (NCT) and brain-behavior correlation analyses. The dataset follows the Brain Imaging Data Structure (BIDS) v1.4.1 specification. The original DICOM data are not released, for participant privacy reasons. Note: the ProtocolName field in the imaging sidecars was relabelled from the acquisition-time protocol label to a neutral study code, and a ReferencedSOPInstanceUID plus raw DICOM group-length elements were removed from the embedded dcm2niix metadata block. All other acquisition parameters are unmodified. See CHANGES in the bundle for the exact mapping.