cuCIM Transform (#2932)

* Implement CuCIM wrapper transfrom

Signed-off-by: Behrooz <3968947+drbeh@users.noreply.github.com>

Author: bhashemian

docs/source/transforms.rst

@@ -733,6 +733,18 @@ Utility
      :members:
      :special-members: __call__
 
+`CuCIM`
+"""""""
+.. autoclass:: CuCIM
+    :members:
+    :special-members: __call__
+
+`RandCuCIM`
+"""""""""""
+.. autoclass:: RandCuCIM
+    :members:
+    :special-members: __call__
+
 
 Dictionary Transforms
 ---------------------
@@ -1374,6 +1386,17 @@ Utility (Dict)
      :members:
      :special-members: __call__
 
+`CuCIMd`
+""""""""
+.. autoclass:: CuCIMd
+    :members:
+    :special-members: __call__
+
+`RandCuCIMd`
+""""""""""""
+.. autoclass:: RandCuCIMd
+    :members:
+    :special-members: __call__
 
 Transform Adaptors
 ------------------

monai/transforms/__init__.py

@@ -359,6 +359,7 @@
     CastToType,
     ClassesToIndices,
     ConvertToMultiChannelBasedOnBratsClasses,
+    CuCIM,
     DataStats,
     EnsureChannelFirst,
     EnsureType,
@@ -368,6 +369,7 @@
     LabelToMask,
     Lambda,
     MapLabelValue,
+    RandCuCIM,
     RandLambda,
     RemoveRepeatedChannel,
     RepeatChannel,
@@ -410,6 +412,9 @@
     CopyItemsd,
     CopyItemsD,
     CopyItemsDict,
+    CuCIMd,
+    CuCIMD,
+    CuCIMDict,
     DataStatsd,
     DataStatsD,
     DataStatsDict,
@@ -440,6 +445,9 @@
     MapLabelValued,
     MapLabelValueD,
     MapLabelValueDict,
+    RandCuCIMd,
+    RandCuCIMD,
+    RandCuCIMDict,
     RandLambdad,
     RandLambdaD,
     RandLambdaDict,

monai/transforms/utility/array.py

@@ -49,7 +49,7 @@
 PILImageImage, has_pil = optional_import("PIL.Image", name="Image")
 pil_image_fromarray, _ = optional_import("PIL.Image", name="fromarray")
 cp, has_cp = optional_import("cupy")
-cp_ndarray, _ = optional_import("cupy", name="ndarray")
+
 
 __all__ = [
     "Identity",
@@ -1148,3 +1148,78 @@ def __call__(self, img: torch.Tensor):
             raise ValueError("img must be PyTorch Tensor, consider converting img by `EnsureType` transform first.")
 
         return img.to(self.device, **self.kwargs)
+
+
+class CuCIM(Transform):
+    """
+    Wrap a non-randomized cuCIM transform, defined based on the transform name and args.
+    For randomized transforms (or randomly applying a transform) use :py:class:`monai.transforms.RandCuCIM`.
+
+    Args:
+        name: the transform name in CuCIM package
+        args: parameters for the CuCIM transform
+        kwargs: parameters for the CuCIM transform
+
+    Note:
+        CuCIM transform only work with CuPy arrays, so this transform expects input data to be `cupy.ndarray`.
+        Users can call `ToCuPy` transform to convert a numpy array or torch tensor to cupy array.
+    """
+
+    def __init__(self, name: str, *args, **kwargs) -> None:
+        super().__init__()
+        self.transform, _ = optional_import("cucim.core.operations.expose.transform", name=name)
+        self.args = args
+        self.kwargs = kwargs
+
+    def __call__(self, data):
+        """
+        Args:
+            data: a CuPy array (`cupy.ndarray`) for the cuCIM transform
+
+        Returns:
+            `cupy.ndarray`
+
+        """
+        return self.transform(data, *self.args, **self.kwargs)
+
+
+class RandCuCIM(CuCIM, RandomizableTransform):
+    """
+    Wrap a randomized cuCIM transform, defined based on the transform name and args,
+    or randomly apply a non-randomized transform.
+    For deterministic non-randomized transforms use :py:class:`monai.transforms.CuCIM`.
+
+    Args:
+        name: the transform name in CuCIM package.
+        apply_prob: the probability to apply the transform (default=1.0)
+        args: parameters for the CuCIM transform.
+        kwargs: parameters for the CuCIM transform.
+
+    Note:
+        - CuCIM transform only work with CuPy arrays, so this transform expects input data to be `cupy.ndarray`.
+          Users can call `ToCuPy` transform to convert a numpy array or torch tensor to cupy array.
+        - If the cuCIM transform is already randomized the `apply_prob` argument has nothing to do with
+          the randomness of the underlying cuCIM transform. `apply_prob` defines if the transform (either randomized
+          or non-randomized) being applied randomly, so it can apply non-randomized tranforms randomly but be careful
+          with setting `apply_prob` to anything than 1.0 when using along with cuCIM's randomized transforms.
+        - If the random factor of the underlying cuCIM transform is not derived from `self.R`,
+          the results may not be deterministic. See Also: :py:class:`monai.transforms.Randomizable`.
+    """
+
+    def __init__(self, name: str, apply_prob: float = 1.0, *args, **kwargs) -> None:
+        CuCIM.__init__(self, name, *args, **kwargs)
+        RandomizableTransform.__init__(self, prob=apply_prob)
+
+    def __call__(self, data):
+        """
+        Args:
+            data: a CuPy array (`cupy.ndarray`) for the cuCIM transform
+
+        Returns:
+            `cupy.ndarray`
+
+        """
+        self.randomize(data)
+        if not self._do_transform:
+            return data
+        return super().__call__(data)

monai/transforms/utility/dictionary.py

@@ -36,6 +36,7 @@
     CastToType,
     ClassesToIndices,
     ConvertToMultiChannelBasedOnBratsClasses,
+    CuCIM,
     DataStats,
     EnsureChannelFirst,
     EnsureType,
@@ -87,6 +88,9 @@
     "CopyItemsD",
     "CopyItemsDict",
     "CopyItemsd",
+    "CuCIMd",
+    "CuCIMD",
+    "CuCIMDict",
     "DataStatsD",
     "DataStatsDict",
     "DataStatsd",
@@ -117,6 +121,9 @@
     "MapLabelValueD",
     "MapLabelValueDict",
     "MapLabelValued",
+    "RandCuCIMd",
+    "RandCuCIMD",
+    "RandCuCIMDict",
     "RandLambdaD",
     "RandLambdaDict",
     "RandLambdad",
@@ -1481,6 +1488,99 @@ def __call__(self, data: Mapping[Hashable, torch.Tensor]) -> Dict[Hashable, torc
         return d
 
 
+class CuCIMd(MapTransform):
+    """
+    Dictionary-based wrapper of :py:class:`monai.transforms.CuCIM` for non-randomized transforms.
+    For randomized transforms of CuCIM use :py:class:`monai.transforms.RandCuCIMd`.
+
+    Args:
+        keys: keys of the corresponding items to be transformed.
+            See also: :py:class:`monai.transforms.compose.MapTransform`
+        name: The transform name in CuCIM package.
+        allow_missing_keys: don't raise exception if key is missing.
+        args: parameters for the CuCIM transform.
+        kwargs: parameters for the CuCIM transform.
+
+    Note:
+        CuCIM transforms only work with CuPy arrays, this transform expects input data to be `cupy.ndarray`.
+        Users can call `ToCuPy` transform to convert a numpy array or torch tensor to cupy array.
+    """
+
+    def __init__(
+        self,
+        keys: KeysCollection,
+        name: str,
+        allow_missing_keys: bool = False,
+        *args,
+        **kwargs,
+    ) -> None:
+        super().__init__(keys=keys, allow_missing_keys=allow_missing_keys)
+        self.trans = CuCIM(name, *args, **kwargs)
+
+    def __call__(self, data):
+        """
+        Args:
+            data: Dict[Hashable, `cupy.ndarray`]
+
+        Returns:
+            Dict[Hashable, `cupy.ndarray`]
+
+        """
+        d = dict(data)
+        for key in self.key_iterator(d):
+            d[key] = self.trans(d[key])
+        return d
+
+
+class RandCuCIMd(CuCIMd, RandomizableTransform):
+    """
+    Dictionary-based wrapper of :py:class:`monai.transforms.CuCIM` for randomized transforms.
+    For deterministic non-randomized transforms of CuCIM use :py:class:`monai.transforms.CuCIMd`.
+
+    Args:
+        keys: keys of the corresponding items to be transformed.
+            See also: :py:class:`monai.transforms.compose.MapTransform`
+        name: The transform name in CuCIM package.
+        apply_prob: the probability to apply the transform (default=1.0)
+        allow_missing_keys: don't raise exception if key is missing.
+        args: parameters for the CuCIM transform.
+        kwargs: parameters for the CuCIM transform.
+
+    Note:
+        - CuCIM transform only work with CuPy arrays, so this transform expects input data to be `cupy.ndarray`.
+          Users can call `ToCuPy` transform to convert a numpy array or torch tensor to cupy array.
+        - If the cuCIM transform is already randomized the `apply_prob` argument has nothing to do with
+          the randomness of the underlying cuCIM transform. `apply_prob` defines if the transform (either randomized
+          or non-randomized) being applied randomly, so it can apply non-randomized tranforms randomly but be careful
+          with setting `apply_prob` to anything than 1.0 when using along with cuCIM's randomized transforms.
+        - If the random factor of the underlying cuCIM transform is not derived from `self.R`,
+          the results may not be deterministic. See Also: :py:class:`monai.transforms.Randomizable`.
+    """
+
+    def __init__(
+        self,
+        apply_prob: float = 1.0,
+        *args,
+        **kwargs,
+    ) -> None:
+        CuCIMd.__init__(self, *args, **kwargs)
+        RandomizableTransform.__init__(self, prob=apply_prob)
+
+    def __call__(self, data):
+        """
+        Args:
+            data: Dict[Hashable, `cupy.ndarray`]
+
+        Returns:
+            Dict[Hashable, `cupy.ndarray`]
+
+        """
+        self.randomize(data)
+        if not self._do_transform:
+            return dict(data)
+        return super().__call__(data)
+
+
 IdentityD = IdentityDict = Identityd
 AsChannelFirstD = AsChannelFirstDict = AsChannelFirstd
 AsChannelLastD = AsChannelLastDict = AsChannelLastd
@@ -1517,3 +1617,5 @@ def __call__(self, data: Mapping[Hashable, torch.Tensor]) -> Dict[Hashable, torc
 MapLabelValueD = MapLabelValueDict = MapLabelValued
 IntensityStatsD = IntensityStatsDict = IntensityStatsd
 ToDeviceD = ToDeviceDict = ToDeviced
+CuCIMD = CuCIMDict = CuCIMd
+RandCuCIMD = RandCuCIMDict = RandCuCIMd