nitransforms issues

Replacement of all ``ants.ApplyTransforms`` interfaces in fMRIPrep

Except for those with `interpolation='MultiLabel'`.

oesteban

Replacement for fMRIPrep's ``MultiApplyTransforms`` interfaces

Including tests, and considering how to operate (i.e., parallelization) with very long timeseries (say 400 timepoints).

oesteban

🔧 Loaders and util for H5 and NIfTI transforms (propagation of contributions by @sgiavasis)

1

(Propagation of @sgiavasis 's PR to @jmarabotto 's fork. Original PR message copied below) Started at the NiPreps hackathon at MIT in May 2024. Thanks again to @jmarabotto and @oesteban...

jmarabotto

ENH: Serialize 4D ``apply()`` into 3D+t

2

Re-implements serialization (i.e., splitting a 4D transformation into 3D transformations). For time-dependent transforms (e.g., when also slice-time correcting), parallelization will require a different partition (spatial and/or temporal windowing).

oesteban

Implement "low memory" mode in ``apply()``

1

We are currently propagating the ArrayProxy reported dtype unless an explicit output_dtype is set. It would be more efficient to allow the user set a low memory flag such that...

oesteban

Re-enable parallelization in ``apply()``

2

One of the things I knowingly let drop from the ``apply()`` outsourcing (#195) was the parallelization of resampling. @jmarabotto - do you think this is something you can try to...

oesteban

Feat: Implement a visualisation feature to NiTransforms

4

Addition of new file vis.py. Calls nitransform's nonlinear DenseFieldTransform class to illustrate corresponding transform objects through: - Distorted image and grid - Quiver plot coloured according to Diffusion Scalar map...

jmarabotto