493 integrate 3rd party transforms into MONAI

[Nic-Ma]

Fixes #493 .

Description

This PR implemented a tutorial to integrate 3rd party transforms into MONAI.
Including: BatchGenerator, TorchIO and Rising.

Status

Ready

Types of changes

• Bug fix (non-breaking change which fixes an issue)
• Breaking change (fix or new feature that would cause existing functionality to change)
• New tests added to cover the changes
• Docstrings/Documentation updated

[hjmjohnson]

@Nic-Ma EXCELLENT. I need the MRI Bias correction capabilities from TorchIO for my work in about 2 weeks from now.

Thank you.

[wyli]

@Nic-Ma EXCELLENT. I need the MRI Bias correction capabilities from TorchIO for my work in about 2 weeks from now.

Thank you.

do you mean randomised simulation of bias fields? I don't think there's a bias field correction algo in TorchIO

[hjmjohnson]

@Nic-Ma EXCELLENT. I need the MRI Bias correction capabilities from TorchIO for my work in about 2 weeks from now.
Thank you.

do you mean randomised simulation of bias fields? I don't think there's a bias field correction algo in TorchIO

Yes. I do mean simulation of bias fields. I misspoke. Much of my input data is overly pre-processed, and I need to simulate what would come off the scanner. I need to simulate the "undoing" of correction. :).

[Nic-Ma]

/black

[wyli]

have you tried the adaptor in MONAI? I put a quick example here:

from monai.transforms import Compose, adaptor


class MyTransform:
    def __call__(self, **args):
        return {"img": args["image"] + 1, "new_field": args["seg"] + 1}


class MyTransformDict:
    def __call__(self, data_dict):
        data_dict["images"] += 1
        return data_dict


xform = Compose([adaptor(MyTransform(), outputs={"img": "image", "new_field": "images"}), MyTransformDict()])


out = xform({"image": 1, "seg": 2})
print(out)

>>> {'image': 2, 'seg': 2, 'images': 4}

[Nic-Ma]

Hi @wyli ,

Thanks for your review and suggestions.
With adaptor, it can compose all the transforms now.
Could you please help review it again?
Thanks.

[Nic-Ma]

/black

[Nic-Ma]

Hi @wyli ,

Thanks for your review.
I updated the PR to address the comments.
Could you please help review it again?
Thanks.

[wyli]

thanks for looking into these!

cc @mjorgecardoso @anfeng @atbenmurray