search

ecaruyer / phantomas

Create realistic digital phantoms in diffusion MRI.
Other
31 stars 16 forks source link

Added additional input arguments #25

Closed gabknight closed 10 years ago

gabknight commented 10 years ago

A small PR to add arguments to phantomas_dwi script.

I found it useful to play with lambda 1,2,3 and to specify the output file name.

Maybe it would be better to use a JSON file to describe all parameter of the model (gm/wm/csf t1-t2-rho, gm/csf diffusivity, lambdas, te-tr...) What do you think?

ecaruyer commented 10 years ago

Hey Gab,

Thanks for this one!

Maybe it would be better to use a JSON file to describe all parameter of the model (gm/wm/csf t1-t2-rho, gm/csf diffusivity, lambdas, te-tr...) What do you think?

I agree, at least for the diffusion parameters. Ultimately, I think it'd be nice to be able to set per-phantom or even per-bundle diffusion parameters, and add a couple more models (ball/stick, ...). But before we do this (quite big) work, and we agree on a syntax for the json file, I am tempted to merge this PR. This is always better to offer an API than hard-coding everything.

gabknight commented 10 years ago

Sorry for the bad commit message.

I agree let think about the json file in future work, when various model parameters will be needed.

gabknight commented 10 years ago

Let me know what you think for the metavar. Thank you for the default solution. I'm very glad to see this solution! One last thing, I rename the -o argument to --output_dir. So it is less confusing with --output for dwis. It may broke some scripts though.

ecaruyer commented 10 years ago

This sounds all good to me! Thanks @gabknight!