Co-registered MS MRI scans for Machine Learning


I am a graduate student doing a thesis on MS Lesion segmentation using Dictionary Learning and Sparse Representation.

I am looking to use brainweb as a means of coding up my experiment before using patient data.

I am wondering if the mild/moderate/severe MSlesion scans are consistently the same simulation through each weighting(T1,T2,PD)?

And if possible, can I get the scans with a lesion mask for identifying the voxels that are in the MS lesions?

Thank you for any asistance
Signed - Grey