Torch: map_binary_to_indices, map_classes_to_indices, correct_crop_centers, generate_pos_neg_label_crop_centers, generate_label_classes_crop_centers

[rijobro]

Description

Torch: map_binary_to_indices, map_classes_to_indices, correct_crop_centers, generate_pos_neg_label_crop_centers, generate_label_classes_crop_centers

Status

Ready

Types of changes

• Non-breaking change (fix or new feature that would not break existing functionality).
• New tests added to cover the changes.
• Integration tests passed locally by running ./runtests.sh -f -u --net --coverage.
• Quick tests passed locally by running ./runtests.sh --quick --unittests.
• In-line docstrings updated.
• Documentation updated, tested make html command in the docs/ folder.

[wyli]

looks good to me, @Nic-Ma this might change the behaviour of the spleen seg. notebook...please confirm

[Nic-Ma]

Looks good to me.
Could you please help double confirm by executing this example?
https://github.com/Project-MONAI/tutorials/blob/master/3d_segmentation/torch/unet_training_dict.py
Just also need to add EnsureTyped() after LoadImaged() to test with Tensor data.

Thanks in advance.

[wyli]

/build

[wyli]

/build

[Nic-Ma]

Hi @rijobro ,

Would you like to add Tensor support for RandCropByPosNegLabel and RandCropByLabelClasses with a high priority? As you already updated the help utilities.
And I think this PosNegCrop transform is popular in our training tasks, if we can support Tensor in v0.7, that would be great.

Thanks in advance.