This is the MICCAI-challenge data structure of VerSe2020. The data is published under the licence CC BY-SA 4.0 (see licence.txt, https://osf.io/amh4f/). When using the data you must agree to this license and cite the papers mentioned in readme.txt. **Data structure** The dataset has four files corresponding to one data sample, structured as follows: verse000_CT-orient.nii.gz - Image verse000_CT-orient_seg.nii.gz - Segmentation Mask verse000_CT-orient_iso-ctd.json - Centroid annotations verse000_CT-orient.png - A PNG overview of the annotations. Minor variiations of the naming can apply. All verse IDs smaller than 450 were already present in Verse19 (https://osf.io/nqjyw/). Please refer to this part of the repository to access this data. All images are in the original orientation and voxel spacing as they have been reconstructed by the different scanners. In contrast to Verse19, they have not been re-sampled. Thus, they are NOT in the same orientation. Their spacing is NOT the same. However, an image and its corresponding mask will be in the same orientation and spacing. Centroid coordinates (.json file) are given in 1mm resolution with coordinate origin at 'X'= anterior, 'Y'= superior, and 'Z'= left (ASL). 'label' corresponds to the vertebral label: 1-7: cervical spine: C1-C7 8-19: thoracic spine: T1-T12 20-25: lumbar spine: L1-L6 26,17: sacrum, cocygis - not labeled in this dataset 28: additional 13th thoracic vertebra, T13