Add b=0 scan from the DWI sequence as a new contrast for training model

[jcohenadad]

The b=0 volume from the DWI series looks like this:

It would be nice to add it for training the contrast-agnostic model. We could use the same GT as the one for the mean DWI scan.

[naga-karthik]

Apologies for not following up on this! Seems that this suggestion has indeed been incorporated and our final dataset does include DWI images for training.

[sandrinebedard]

B0 is not the same as the DWI data. It is without the diffusion gradients, we also have the data. But we did not follow up on this, could be done for the active learning and second super model!

[naga-karthik]

oh my bad! reopening then