Transforms in CapsDataset #687
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
camillebrianceau
approved these changes
Dec 19, 2024
| transforms: Transforms, | ||
| data: Optional[Union[pd.DataFrame, Path]] = None, | ||
| caps_directory: Union[str, Path], | ||
| preprocessing: Preprocessing, |
There was a problem hiding this comment.
maybe we can give a default preprocessing like PreprocessingT1() WDYT ?
There was a problem hiding this comment.
Yes, I think T1 is the most common (at least in our lab).
| will look for masks with that values as a suffix. For example, if the image of the subject | ||
| 'sub-001' for the session 'ses-M000' is in 'sub-001/ses-M000/sub-001_ses-M000_T1w.nii.gz' | ||
| and `masks=["brain", "hippocampus"]`, it will look for the masks in | ||
| 'sub-001/ses-M000/sub-001_ses-M000_brain.nii.gz' and 'sub-001/ses-M000/sub-001_ses-M000_hippocampus.nii.gz'. | ||
| """ | ||
|
|
There was a problem hiding this comment.
Suggested change
| caps_directory = Path(caps_directory) # maybe to put in the CapsReader |
There was a problem hiding this comment.
It is done in the CapsReader yes:
Co-authored-by: camillebrianceau <[email protected]>
…dvx/clinicadl into caps_dataset_transforms
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
PR on
CapsDataset,PreprocessingandTransforms.For
CapsDataset:labelandmasksarguments inCapsDataset:labelcan be eitherNone(reconstruction), a mask (segmentation) or a column of the tsvdata(classification, regression).masksare masks that are potentially necessary to compute some transforms. Consequently, I added methods_get_masksand_get_labelthat are called in the__getitem__.CapsDataset's__getitem__to handle TorchIO'sSubject(following Continuing refactoring of the extraction objects #686).CapsDatasetclass was growing pretty big, I tookcheck_preprocessingandprepare_datamethods out of it.check_preprocessingis now inCapsReader; andprepare_datais now a function that takes aCapsDatasetas an input.CapsDatasetSampleasCapsDataset's__getitem__now returns aSampleobject (following Continuing refactoring of the extraction objects #686).For
Preprocessing:BasePreprocessingtoPreprocessingand the enumPreprocessingtoPreprocessingMethod. The motivation is to havePreprocessingin type hints.t2-linearandt1-extensiveas they no longer exist.use_uncropped_imageforPreprocessingDTIandPreprocessingCustomas it is only an argument for linear pipelines.For
Transforms:size_reduction,size_reduction_factorandnormalizeinTransforms, as equivalent transforms can easily be passed.Transformsobject.Sorry for the huge PR. Besides, these are only suggestions, feel free to question them!