This App currently outputs neuro/dtiinit, but I believe we could instead output neuro/dwi with datatype tags such as "dtiinit" or "aligned" etc.. We can then update existing App that uses neuro/dtiinit to use 'neuro/dwi` datatype instead to make it compatible. I believe doing this will expand the number of Apps that can interoperate.
Related to this.. I believe we should start storing alignment related information as part of dataset metadata, as it is a critical attribute of image processing that determines compatibility between the Apps. We basically store alignment information in a "graph" between all datasets that are aligned to each other - within and across various datatypes.
If we want a less engineered approach, we could simply make sure to set the sform with proper affine and coordinate labels for all nifti output.. but that won't tell us if a given dwi is aligned to which t1, for example.
This App currently outputs
neuro/dtiinit
, but I believe we could instead outputneuro/dwi
with datatype tags such as "dtiinit" or "aligned" etc.. We can then update existing App that usesneuro/dtiinit
to use 'neuro/dwi` datatype instead to make it compatible. I believe doing this will expand the number of Apps that can interoperate.Related to this.. I believe we should start storing alignment related information as part of dataset metadata, as it is a critical attribute of image processing that determines compatibility between the Apps. We basically store alignment information in a "graph" between all datasets that are aligned to each other - within and across various datatypes.
If we want a less engineered approach, we could simply make sure to set the sform with proper affine and coordinate labels for all nifti output.. but that won't tell us if a given dwi is aligned to which t1, for example.