#11/5/2024

This data was used to train the APRL model published in Lou et al. (2021).

Each subject folder contains 'raw' and 'preprocessed' subfolders. 

The 'raw' subfolder contains 1 file:
'phase_unwrapped_ubmask.nii.gz - central line of PRLs

The 'preprocessed' subfolder contains 4 files:
'flair_pp.nii.gz' - preprocessed FLAIR image
'phase_pp.nii.gz' - preprocessed T2*-PHASE image
't1_pp.nii.gz' - preprocessed T1 image
'seg2phase.nii.gz' - segmentation mask of all lesions

Preprocessing included N4 inhomogeneity correction, registration of T1 and FLAIR iamges to phase space, skullstripping, and WhiteStripe normalization. See Lou et al. (2021) for more details.