[DLMED] add scripts

Signed-off-by: Nic Ma <[email protected]>

Author: Nic-Ma

modules/bundles/hybrid_programming/README.md

@@ -6,14 +6,18 @@ This example mainly shows 2 typical use cases of hybrid programming with MONAI b
 Please note that this example depends on the `spleen_segmentation` bundle example and show the hybrid programming via overriding the config file of it.
 
 ## commands example
-Override the `train` config with customized `transforms` and execute training:
+Export the customized python code to `PYTHONPATH`:
+```
+export PYTHONPATH=$PYTHONPATH:"<path to 'hybrid_programming/scripts'>"
+```
 
+Override the `train` config with customized `transforms` and execute training:
 ```
-python -m monai.bundle run training --meta_file configs/metadata.json --config_file "['configs/train.json','configs/custom_train.json']" --logging_file configs/logging.conf
+python -m monai.bundle run training --meta_file <spleen_configs_path>/metadata.json --config_file "['<spleen_configs_path>/train.json','configs/custom_train.json']" --logging_file <spleen_configs_path>/logging.conf
 ```
 
 Parse the config in the python program and execute inference from the program:
 
 ```
-python -m scripts.inference.run --config_file configs/inference.json
+python -m scripts.inference run --config_file <spleen_configs_path>/inference.json --ckpt_path <path_to_checkpoint>
 ```

modules/bundles/hybrid_programming/configs/custom_train.json

@@ -0,0 +1,7 @@
+{
+    "train#preprocessing#transforms#6":
+    {
+        "_target_": "scripts.custom_transforms.PrintEnsureTyped",
+        "keys": ["image", "label"]
+    }
+}

modules/bundles/hybrid_programming/scripts/__init__.py

@@ -0,0 +1,10 @@
+# Copyright (c) MONAI Consortium
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#     http://www.apache.org/licenses/LICENSE-2.0
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.

modules/bundles/hybrid_programming/scripts/custom_transforms.py

@@ -0,0 +1,32 @@
+# Copyright (c) MONAI Consortium
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#     http://www.apache.org/licenses/LICENSE-2.0
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+from monai.config import KeysCollection
+from monai.transforms import EnsureTyped
+
+
+class PrintEnsureTyped(EnsureTyped):
+    """
+    Extend the `EnsureTyped` transform to print the image shape.
+
+     Args:
+        keys: keys of the corresponding items to be transformed.
+
+    """
+
+    def __init__(self, keys: KeysCollection, data_type: str = "tensor") -> None:
+        super().__init__(keys, data_type=data_type)
+
+    def __call__(self, data):
+        d = dict(super().__call__(data=data))
+        for key in self.key_iterator(d):
+            print(f"data shape of {key}: {d[key].shape}")
+        return d

modules/bundles/hybrid_programming/scripts/inference.py

@@ -0,0 +1,47 @@
+# Copyright (c) MONAI Consortium
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#     http://www.apache.org/licenses/LICENSE-2.0
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+import torch
+from monai.bundle import ConfigParser
+from monai.data import decollate_batch
+from monai.utils.enums import CommonKeys
+
+
+def run(config_file: str, ckpt_path: str):
+    parser = ConfigParser()
+    parser.read_config(config_file)
+    # edit the config content at runtime and lazy instantiation
+    parser["inferer"]["roi_size"] = [160, 160, 160]
+
+    device = parser.get_parsed_content("device")
+    # instantialize the components
+    model = parser.get_parsed_content("network")
+    model.load_state_dict(torch.load(ckpt_path))
+
+    dataloader = parser.get_parsed_content("dataloader")
+    inferer = parser.get_parsed_content("inferer")
+    postprocessing = parser.get_parsed_content("postprocessing")
+
+    model.eval()
+    with torch.no_grad():
+        for d in dataloader:
+            images = d[CommonKeys.IMAGE].to(device)
+            # define sliding window size and batch size for windows inference
+            d[CommonKeys.PRED] = inferer(inputs=images, network=model)
+            # decollate the batch data into a list of dictionaries, then execute postprocessing transforms
+            [postprocessing(i) for i in decollate_batch(d)]
+
+
+if __name__ == "__main__":
+    from monai.utils import optional_import
+
+    fire, _ = optional_import("fire")
+    fire.Fire()

modules/bundles/spleen_segmentation/configs/metadata.json

@@ -5,7 +5,7 @@
         "0.1.0": "complete the model package",
         "0.0.1": "initialize the model package structure"
     },
-    "monai_version": "0.8.0",
+    "monai_version": "0.9.0",
     "pytorch_version": "1.10.0",
     "numpy_version": "1.21.2",
     "optional_packages_version": {