Editorial changes to correct typos, rename parameters, adding comments. Retested examples.

Signed-off-by: M Q <mingmelvinq@nvidia.com>

Author: MMelQin

examples/apps/ai_unetr_seg_app/app.py

@@ -80,7 +80,7 @@ def compose(self):
         _algorithm_family = codes.DCM.ArtificialIntelligence
         _algorithm_version = "0.1.0"
 
-        # List of (Segment name, [Code menaing str]), not including background which is value of 0.
+        # List of (Segment name, [Code meaning str]), not including background which is value of 0.
         # User must provide correct codes, which can be looked at, e.g.
         # https://bioportal.bioontology.org/ontologies/SNOMEDCT
         # Alternatively, consult the concept and code dictionaries in PyDicom

examples/apps/ai_unetr_seg_app/unetr_seg_operator.py

@@ -56,7 +56,7 @@ class UnetrSegOperator(Operator):
 
     def __init__(
         self,
-        frament: Fragment,
+        fragment: Fragment,
         *args,
         app_context: AppContext,
         model_path: Path,
@@ -71,13 +71,13 @@ def __init__(
         self.model_path = model_path
         self.output_folder = output_folder
         self.output_folder.mkdir(parents=True, exist_ok=True)
-        self.fragement = frament  # Cache and later pass the Fragment/Application to contained operator(s)
+        self.app_fragment = fragment  # Cache and later pass the Fragment/Application to contained operator(s)
         self.app_context = app_context
         self.input_name_image = "image"
         self.output_name_seg = "seg_image"
         self.output_name_saved_images_folder = "saved_images_folder"
 
-        super().__init__(frament, *args, **kwargs)
+        super().__init__(fragment, *args, **kwargs)
 
     def setup(self, spec: OperatorSpec):
         spec.input(self.input_name_image)
@@ -102,7 +102,7 @@ def compute(self, op_input, op_output, context):
 
         # Delegates inference and saving output to the built-in operator.
         infer_operator = MonaiSegInferenceOperator(
-            self.fragement,
+            self.app_fragment,
             roi_size=(
                 96,
                 96,

examples/apps/breast_density_classifier_app/README.md

@@ -9,7 +9,7 @@ Sample data and a torchscript model can be downloaded from https://drive.google.
 python app.py -i <input_dir> -o <out_dir> -m <breast_density_model>
 ```
 
-## Package the application as a MONAI Application Package (contianer image)
+## Package the application as a MONAI Application Package (container image)
 In order to build the MONAI App Package, go a level up and execute the following command.
 ```
 monai-deploy package breast_density_classification_app -m <breast_density_model> -c breast_density_classifer_app/app.yaml --tag breast_density:0.1.0 --platform x64-workstation -l DEBUG
@@ -20,4 +20,4 @@ monai-deploy package breast_density_classification_app -m <breast_density_model>
 monai-deploy run breast_density-x64-workstation-dgpu-linux-amd64:0.1.0 -i <input_dir> -o <output_dir>
 ```
 
-Once the container exits successfully, check the results in the output directory. There should be a newly creeated DICOM instance file and a `output.json` file containing the classification results.
+Once the container exits successfully, check the results in the output directory. There should be a newly created DICOM instance file and a `output.json` file containing the classification results.

examples/apps/breast_density_classifier_app/breast_density_classifier_operator.py

@@ -40,7 +40,7 @@ class ClassifierOperator(Operator):
 
     def __init__(
         self,
-        frament: Fragment,
+        fragment: Fragment,
         *args,
         model_name: Optional[str] = "",
         app_context: AppContext,
@@ -67,7 +67,7 @@ def __init__(
         # The name of the optional input port for passing data to override the output folder path.
         self.input_name_output_folder = "output_folder"
 
-        # The output folder set on the object can be overriden at each compute by data in the optional named input
+        # The output folder set on the object can be overridden at each compute by data in the optional named input
         self.output_folder = output_folder
 
         # Need the name when there are multiple models loaded
@@ -80,7 +80,7 @@ def __init__(
 
         self.model = self._get_model(self.app_context, self.model_path, self._model_name)
 
-        super().__init__(frament, *args, **kwargs)
+        super().__init__(fragment, *args, **kwargs)
 
     def _get_model(self, app_context: AppContext, model_path: Path, model_name: str):
         """Load the model with the given name from context or model path
@@ -116,7 +116,7 @@ def _convert_dicom_metadata_datatype(self, metadata: Dict):
         if not metadata:
             return metadata
 
-        # Try to convert data type for the well knowned attributes. Add more as needed.
+        # Try to convert data type for the well known attributes. Add more as needed.
         if metadata.get("SeriesInstanceUID", None):
             try:
                 metadata["SeriesInstanceUID"] = str(metadata["SeriesInstanceUID"])

examples/apps/hugging_face_integration_app/med_image_generation/app.py

@@ -1,39 +1,38 @@
-import logging 
+import argparse
+import logging
 from pathlib import Path
-import torch 
+
+import numpy as np
+import torch
 from diffusers import StableDiffusionPipeline
-from monai.deploy.core import AppContext, Application
 from PIL import Image
-import numpy as np
-import argparse
 
+from monai.deploy.core import AppContext, Application
 
 
 class App(Application):
-	name = "Diffusion Image App"
-	description = "Simple application showing diffusion to generate Images"
-	def compose(self):
-		model_id = "Nihirc/Prompt2MedImage"
-		device = "cuda"
-		parser = argparse.ArgumentParser()
-		parser.add_argument("--input_prompt", type=str, default="Generate a X-ray")
-		parser.add_argument("--output", type=str, default="./out.jpg")
-		args = parser.parse_args()
-
-		input_prompt = args.input_prompt
-		output_path = args.output
-		print("Input Prompt: ", input_prompt)
-		pipe = StableDiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.float16)
-		pipe = pipe.to(device)
-		prompt = "Show me an X ray pevic fracture"
-		image = pipe(prompt).images[0]
-		image.save(output_path)
+    name = "Diffusion Image App"
+    description = "Simple application showing diffusion to generate Images"
+
+    def compose(self):
+        model_id = "Nihirc/Prompt2MedImage"
+        device = "cuda"
+        parser = argparse.ArgumentParser()
+        parser.add_argument("--input_prompt", type=str, default="Generate a X-ray")
+        parser.add_argument("--output", type=str, default="./out.jpg")
+        args = parser.parse_args()
+
+        input_prompt = args.input_prompt
+        output_path = args.output
+        print("Input Prompt: ", input_prompt)
+        pipe = StableDiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.float16)
+        pipe = pipe.to(device)
+        prompt = "Show me an X ray pevic fracture"
+        image = pipe(prompt).images[0]
+        image.save(output_path)
 
 
 if __name__ == "__main__":
-	logging.info(f"Begin {__name__}")
-	App().run()
-	logging.info(f"End {__name__}")
-
-	
-
+    logging.info(f"Begin {__name__}")
+    App().run()
+    logging.info(f"End {__name__}")

examples/apps/mednist_classifier_monaideploy/mednist_classifier_monaideploy.py

@@ -100,7 +100,7 @@ class MedNISTClassifierOperator(Operator):
 
     def __init__(
         self,
-        frament: Fragment,
+        fragment: Fragment,
         *args,
         app_context: AppContext,
         model_name: Optional[str] = "",
@@ -127,7 +127,7 @@ def __init__(
         # The name of the optional input port for passing data to override the output folder path.
         self.input_name_output_folder = "output_folder"
 
-        # The output folder set on the object can be overriden at each compute by data in the optional named input
+        # The output folder set on the object can be overridden at each compute by data in the optional named input
         self.output_folder = output_folder
 
         # Need the name when there are multiple models loaded
@@ -138,7 +138,7 @@ def __init__(
         self.model = self._get_model(self.app_context, self.model_path, self._model_name)
 
         # This needs to be at the end of the constructor.
-        super().__init__(frament, *args, **kwargs)
+        super().__init__(fragment, *args, **kwargs)
 
     def _get_model(self, app_context: AppContext, model_path: Path, model_name: str):
         """Load the model with the given name from context or model path

examples/apps/simple_imaging_app/_version.py

@@ -307,7 +307,7 @@ def git_pieces_from_vcs(tag_prefix, root, verbose, runner=run_command):
         # TAG-NUM-gHEX
         mo = re.search(r'^(.+)-(\d+)-g([0-9a-f]+)$', git_describe)
         if not mo:
-            # unparseable. Maybe git-describe is misbehaving?
+            # unparsable. Maybe git-describe is misbehaving?
             pieces["error"] = ("unable to parse git-describe output: '%s'"
                                % describe_out)
             return pieces

examples/apps/simple_imaging_app/app.py

@@ -49,10 +49,10 @@ def compose(self):
         output_data_path = Path(app_context.output_path)
         logging.info(f"sample_data_path: {sample_data_path}")
 
-        # Please note that the Application object, self, is passed as the first positonal argument
+        # Please note that the Application object, self, is passed as the first positional argument
         # and the others as kwargs.
         # Also note the CountCondition of 1 on the first operator, indicating to the application executor
-        # to invoke this operator, hence the pipleline, only once.
+        # to invoke this operator, hence the pipeline, only once.
         sobel_op = SobelOperator(self, CountCondition(self, 1), input_path=sample_data_path, name="sobel_op")
         median_op = MedianOperator(self, name="median_op")
         gaussian_op = GaussianOperator(self, output_folder=output_data_path, name="gaussian_op")

examples/apps/simple_imaging_app/gaussian_operator.py

@@ -24,7 +24,7 @@ class GaussianOperator(Operator):
         single input:
           an image array object
         single output:
-          an image arrary object, without enforcing a downsteam receiver
+          an image array object, without enforcing a downstream receiver
 
     Besides, this operator also saves the image file in the given output folder.
     """

monai/deploy/operators/clara_viz_operator.py

@@ -33,7 +33,7 @@ class ClaraVizOperator(Operator):
         seg_image: Image object of the segmentation image derived from the input image.
     """
 
-    def __init__(self, fragement: Fragment, *args, **kwargs):
+    def __init__(self, fragment: Fragment, *args, **kwargs):
         """Constructor of the operator.
 
         Args:
@@ -43,7 +43,7 @@ def __init__(self, fragement: Fragment, *args, **kwargs):
         self.input_name_image = "image"
         self.input_name_seg_image = "seg_image"
 
-        super().__init__(fragement, *args, **kwargs)
+        super().__init__(fragment, *args, **kwargs)
 
     def setup(self, spec: OperatorSpec):
         spec.input(self.input_name_image)

monai/deploy/operators/dicom_encapsulated_pdf_writer_operator.py

@@ -45,12 +45,12 @@ class DICOMEncapsulatedPDFWriterOperator(Operator):
         None
 
     File output:
-        Generaed DICOM instance file in the provided output folder.
+        Generated DICOM instance file in the provided output folder.
     """
 
     # File extension for the generated DICOM Part 10 file.
     DCM_EXTENSION = ".dcm"
-    # The default output folder for saveing the generated DICOM instance file.
+    # The default output folder for saving the generated DICOM instance file.
     DEFAULT_OUTPUT_FOLDER = Path(os.getcwd()) / "output"
 
     def __init__(
@@ -249,7 +249,7 @@ def _is_pdf_bytes(self, content: bytes):
         return True
 
 
-# Commenting out the following as pttype complains about the contructor for no reason
+# Commenting out the following as pttype complains about the constructor for no reason
 # def test(test_copy_tags: bool = True):
 #     from monai.deploy.operators.dicom_data_loader_operator import DICOMDataLoaderOperator
 #     from monai.deploy.operators.dicom_series_selector_operator import DICOMSeriesSelectorOperator

monai/deploy/operators/dicom_seg_writer_operator.py

@@ -206,7 +206,7 @@ def __init__(
                 Object encapsulating the description of each segment present in the segmentation.
             output_folder: Folder for file output, overridden by named input on compute.
                            Defaults to current working dir's child folder, output.
-            custom_tags: Optonal[Dict[str, str]], optional
+            custom_tags: Optional[Dict[str, str]], optional
                 Dictionary for setting custom DICOM tags using Keywords and str values only
             omit_empty_frames: bool, optional
                 Whether to omit frames that contain no segmented pixels from the output segmentation.

monai/deploy/operators/dicom_text_sr_writer_operator.py

@@ -42,12 +42,12 @@ class DICOMTextSRWriterOperator(Operator):
         None
 
     File output:
-        Generaed DICOM instance file in the provided output folder.
+        Generated DICOM instance file in the provided output folder.
     """
 
     # File extension for the generated DICOM Part 10 file.
     DCM_EXTENSION = ".dcm"
-    # The default output folder for saveing the generated DICOM instance file.
+    # The default output folder for saving the generated DICOM instance file.
     # DEFAULT_OUTPUT_FOLDER = Path(os.path.join(os.path.dirname(__file__))) / "output"
     DEFAULT_OUTPUT_FOLDER = Path.cwd() / "output"
 
@@ -259,7 +259,7 @@ def write(self, content_text, dicom_series: Optional[DICOMSeries], output_dir: P
         self._logger.info(f"DICOM SOP instance saved in {file_path}")
 
 
-# Commenting out the following as pttype complains about the contructor for no reason
+# Commenting out the following as pttype complains about the constructor for no reason
 # def test(test_copy_tags: bool = True):
 #     from monai.deploy.operators.dicom_data_loader_operator import DICOMDataLoaderOperator
 #     from monai.deploy.operators.dicom_series_selector_operator import DICOMSeriesSelectorOperator

monai/deploy/operators/inference_operator.py

@@ -17,7 +17,7 @@
 class InferenceOperator(Operator):
     """The base operator for operators that perform AI inference.
 
-    This operator preforms pre-transforms on a input image, inference with
+    This operator performs pre-transforms on a input image, inference with
     a given model, post-transforms, and final results generation.
     """
 

monai/deploy/operators/monai_bundle_inference_operator.py

@@ -546,7 +546,7 @@ def compute(self, op_input, op_output, context):
         # `context.models.get(model_name)` returns a model instance if exists.
         # If model_name is not specified and only one model exists, it returns that model.
 
-        # The models are loaded on contruction via the AppContext object in turn the model factory.
+        # The models are loaded on construction via the AppContext object in turn the model factory.
         self._model_network = self.app_context.models.get(self._model_name) if self.app_context.models else None
 
         if self._model_network:

monai/deploy/operators/monai_seg_inference_operator.py

@@ -58,7 +58,7 @@ class InfererType(StrEnum):
 class MonaiSegInferenceOperator(InferenceOperator):
     """This segmentation operator uses MONAI transforms and Sliding Window Inference.
 
-    This operator preforms pre-transforms on a input image, inference
+    This operator performs pre-transforms on a input image, inference
     using a given model, and post-transforms. The segmentation image is saved
     as a named Image object in memory.
 
@@ -241,7 +241,7 @@ def _convert_dicom_metadata_datatype(self, metadata: Dict):
         if not metadata:
             return metadata
 
-        # Try to convert data type for the well knowned attributes. Add more as needed.
+        # Try to convert data type for the well known attributes. Add more as needed.
         if metadata.get("SeriesInstanceUID", None):
             try:
                 metadata["SeriesInstanceUID"] = str(metadata["SeriesInstanceUID"])
@@ -313,6 +313,7 @@ def compute_impl(self, input_image, context):
         with torch.no_grad():
             for d in dataloader:
                 images = d[self._input_dataset_key].to(device)
+                self._logger.info(f"Input of {type(images)} shape: {images.shape}")
                 if self._inferer == InfererType.SLIDING_WINDOW:
                     d[self._pred_dataset_key] = sliding_window_inference(
                         inputs=images,
@@ -331,7 +332,14 @@ def compute_impl(self, input_image, context):
                     )
 
                 d = [post_transforms(i) for i in decollate_batch(d)]
-                out_ndarray = d[0][self._pred_dataset_key].cpu().numpy()
+                self._logger.info(f"Post transform length/batch size of output: {len(d)}")
+                self._logger.info(
+                    f"Post transform pixel spacings of '{self._pred_dataset_key}' in the first output: {d[0][self._pred_dataset_key].pixdim}"
+                )
+                out_ndarray = d[0][self._pred_dataset_key].cpu().numpy()  # Single output to numpy on CPU
+                self._logger.info(
+                    f"Post transform '{self._pred_dataset_key}' of {type(out_ndarray)} shape: {out_ndarray.shape}"
+                )
                 # Need to squeeze out the channel dim fist
                 out_ndarray = np.squeeze(out_ndarray, 0)
                 # NOTE: The domain Image object simply contains a Arraylike obj as image as of now.
@@ -343,7 +351,9 @@ def compute_impl(self, input_image, context):
                 #       the resultant ndarray for the prediction image needs to be transposed back, so the
                 #       array index order is back to DHW, the same order as the in-memory input Image obj.
                 out_ndarray = out_ndarray.T.astype(np.uint8)
-                self._logger.info(f"Output Seg image numpy array shaped: {out_ndarray.shape}")
+                self._logger.info(
+                    f"Output Seg image numpy array of type '{type(out_ndarray)}' shape: {out_ndarray.shape}"
+                )
                 self._logger.info(f"Output Seg image pixel max value: {np.amax(out_ndarray)}")
 
                 return Image(out_ndarray, input_img_metadata)

monai/deploy/operators/publisher_operator.py

@@ -29,10 +29,10 @@ class PublisherOperator(Operator):
     generates the render config file and the meta data file, then save all in the `publish` folder of the app.
     """
 
-    # The default input folder for saveing the generated DICOM instance file.
+    # The default input folder for saving the generated DICOM instance file.
     DEFAULT_INPUT_FOLDER = Path(getcwd()) / "input"
 
-    # The default output folder for saveing the generated DICOM instance file.
+    # The default output folder for saving the generated DICOM instance file.
     DEFAULT_OUTPUT_FOLDER = Path(getcwd()) / "output"
 
     def __init__(

monai/deploy/operators/stl_conversion_operator.py

@@ -41,7 +41,7 @@ class STLConversionOperator(Operator):
     """Converts volumetric image to surface mesh in STL format.
 
     If a file path is provided, the STL binary will be saved in the said output folder.
-    This operator also save the STL file as bytes in memory, idenfied by the named output. Being optional,
+    This operator also save the STL file as bytes in memory, identified by the named output. Being optional,
     this output does not require any downstream receiver.
 
     Named inputs: