Deadline has passed. No further submissions will be considered.

MRI Data Details 

The data set is available both in raw and pre-processed format. The raw data set (prefix: VSD.Head*) consists of defaced T1-weighted images (T1w), diffusion-weighted images (DWI) and corresponding files with information for b-vectors (bvecs) and b-values (bvals).

The pre-processed data (prefix: VSD.Brain*) include T1-weighted images (T1w), fractional anisotropy (FA) and mean diffusivity (MD) maps, as well as probability maps for grey matter (GM) and white matter (WM).

All MRI data are provided in Neuroimaging Informatics Technology Initiative (NIfTI) format (size ~1.8GB).


T1w, grey & white matter probability maps, fractional anisotropy and mean diffusivity (l.t.r)



First, T1w images were bias field corrected and skull-stripped. Then they were rigidly registered to a cohort template in isotropic MNI152 space (1mm³). Grey and white matter probability maps were extracted from the registered images. Diffusion data were denoised, corrected for head motion and eddy current artefacts as well as bias-field corrected. Extracted FA and MD maps were finally transferred into isotropic MNI space (2mm³) by rigid registration to the corresponding T1w image (inter-subject registration).


Data Example

Raw data files start with VSD.Head*:
VSD.Head.<age in years>Y.<gender>.<modality>.000000.nii
T1w of 23 year old male:  VSD.Head.023Y.M.MR_T1.110001.nii
DWI of 64 year old female:  VSD.Head.064Y.F.MR_DWI.110001.nii

Pre-processed data files start with VSD.Brain*.
VSD.Brain.<age in years>Y.<gender>.<modality>.000000.nii
T1w of 19 year old male:  VSD.Brain.019Y.M.MR_T1.110001.nii
GM of 68 year old male:  VSD.Brain.068Y.M.MR_GM_prob.102003.nii
WM of 23 year old female:  VSD.Brain.023Y.F.MR_T1.110001.nii
FA of 35 year old female:  VSD.Brain.035Y.F.MR_DT_FA.115062.nii
MD of 27 year old male:  VSD.Brain.027Y.M.MR_DT_MD.116041.nii


Leak of Labels

Over the course of July, labels for individual subjects (3 at a time) will be revealed to allow semi-supervised refinement of the algorithms. In order to receive the labels you will need to submit the current label prediction of your algorithm and a very brief description of how the labels were computed (2-3 sentences are enough, more is welcome) via e-mail.



Privacy and Data Copyright

By registering and/or downloading the data, each participant/team agrees to use the provided data only in the scope of this challenge (mTOP)  and neither pass it on to a third party nor use it for any other publication. No copyright transfer of any kind will take place.