Image data used in this challenge were acquired on a 3T scanner at the UMC Utrecht (the Netherlands). For each of the 30 subjects, fully annotated multi-sequence (T1-weighted, T1-weighted inversion recovery and T2-FLAIR) scans are available. The 30 subjects include patients with diabetes, dementia and Alzheimers, and matched controls (with increased cardiovascular risk) with varying degrees of atrophy and white matter lesions (age > 50).
For each patient, the following MRI scans are provided:
- T1: 3D T1-weighted sequence (TR 7.9 ms, TE 4.5 ms)
- T1-IR: Multi-slice T1-weighted inversion recovery sequence (TR 4416 ms, TE 15 ms, TI 400 ms)
- T2-FLAIR: Multi-slice T2 FLAIR sequence (TR 11000 ms, TE 125 ms, TI 2800 ms)
All scans have a voxel size of 0.958mm × 0.958mm × 3.0mm and all are aligned. The scans are bias field corrected using the N4ITK algorithm ('pre' folder), but the original data is also available ('orig' folder).
Some of the T1-IR images contain artifacts at the bottom of the scan. These artifacts occur often in clinical scans, and it is therefore interesting to know how automatic segmentation methods that use these scans perform under these circumstances.
All scans are accompanied with a file containing the manual reference standard (segm.nii.gz), containing the following 11 labels:
- Background
- Cortical gray matter
- Basal ganglia
- White matter
- White matter lesions
- Cerebrospinal fluid in the extracerebral space
- Ventricles
- Cerebellum
- Brain stem
- Infarction
- Other
Notes on the manual segmentations:
- White matter lesions were segmented on the FLAIR scan
- The outer border of the CSF was segmented using both the T1-weighted scan and the T1-weighted inversion recovery scan.
- All other structures were segmented on the T1-weighted scan (0.958mm × 0.958mm × 3.0mm).
- Vessels were not segmented separately. The CSF segmentation therefore also includes the superior sagittal sinus and transverse sinuses.
- The cerebral falx is also included in the CSF segmentation.