Correct the place to save input image file as well as correct application class name

examples/apps/ai_livertumor_seg_app/__main__.py

@@ -1,4 +1,4 @@
-from app import AIUnetrSegApp
+from app import AILiverTumorApp
 
 if __name__ == "__main__":
-    AIUnetrSegApp(do_run=True)
+    AILiverTumorApp(do_run=True)

examples/apps/ai_livertumor_seg_app/app.py

@@ -24,7 +24,7 @@
 @resource(cpu=1, gpu=1, memory="7Gi")
 # pip_packages can be a string that is a path(str) to requirements.txt file or a list of packages.
 # The MONAI pkg is not required by this class, instead by the included operators.
-class AIUnetrSegApp(Application):
+class AILiverTumorApp(Application):
     def __init__(self, *args, **kwargs):
         """Creates an application instance."""
 
@@ -63,14 +63,14 @@ def compose(self):
         self.add_flow(study_loader_op, series_selector_op, {"dicom_study_list": "dicom_study_list"})
         self.add_flow(series_selector_op, series_to_vol_op, {"dicom_series": "dicom_series"})
         self.add_flow(series_to_vol_op, unetr_seg_op, {"image": "image"})
+        # Add the publishing operator to save the input and seg images for Render Server.
+        # Note the PublisherOperator has temp impl till a proper rendering module is created.
+        self.add_flow(unetr_seg_op, publisher_op, {"saved_images_folder": "saved_images_folder"})
         # Note below the dicom_seg_writer requires two inputs, each coming from a upstream operator.
         # Also note that the DICOMSegmentationWriterOperator may throw exception with some inputs.
         #   Bug has been created to track the issue.
         self.add_flow(series_selector_op, dicom_seg_writer, {"dicom_series": "dicom_series"})
         self.add_flow(unetr_seg_op, dicom_seg_writer, {"seg_image": "seg_image"})
-        # Add the publishing operator to save the input and seg images for Render Server.
-        # Note the PublisherOperator has temp impl till a proper rendering module is created.
-        self.add_flow(unetr_seg_op, publisher_op, {"saved_images_folder": "saved_images_folder"})
 
         self._logger.debug(f"End {self.compose.__name__}")
 
@@ -84,5 +84,5 @@ def compose(self):
     #     python3 app.py -i input -m model/model.ts
     #
     logging.basicConfig(level=logging.DEBUG)
-    app_instance = AIUnetrSegApp()  # Optional params' defaults are fine.
+    app_instance = AILiverTumorApp()  # Optional params' defaults are fine.
     app_instance.run()

examples/apps/ai_livertumor_seg_app/livertumor_seg_operator.py

@@ -76,7 +76,12 @@ def compute(self, op_input: InputContext, op_output: OutputContext, context: Exe
 
         # This operator gets an in-memory Image object, so a specialized ImageReader is needed.
         _reader = InMemImageReader(input_image)
-        pre_transforms = self.pre_process(_reader)
+        # In this example, the input image, once loaded at the beginning of the pre-transforms, is
+        # saved on disk, so is the segmentation prediction image at the end of the post-transform.
+        # They are both saved in the same subfolder of the application output folder, with names
+        # distinguished by postfix. They can also be save in different subfolder if need be.
+        # These images files can then be packaged for rendering.
+        pre_transforms = self.pre_process(_reader, op_output_folder_path)
         post_transforms = self.post_process(pre_transforms, op_output_folder_path)
 
         # Delegates inference and saving output to the built-in operator.
@@ -98,14 +103,20 @@ def compute(self, op_input: InputContext, op_output: OutputContext, context: Exe
         # Now let the built-in operator handles the work with the I/O spec and execution context.
         infer_operator.compute(op_input, op_output, context)
 
-    def pre_process(self, img_reader) -> Compose:
+    def pre_process(self, img_reader, out_dir: str = "./input_images") -> Compose:
         """Composes transforms for preprocessing input before predicting on a model."""
 
         my_key = self._input_dataset_key
         return Compose(
             [
                 LoadImaged(keys=my_key, reader=img_reader),
                 EnsureChannelFirstd(keys=my_key),
+                SaveImaged(
+                    keys=my_key,
+                    output_dir=out_dir,
+                    output_postfix="",
+                    resample=False,
+                ),
                 Spacingd(keys=my_key, pixdim=(1.0, 1.0, 1.0), mode=("bilinear"), align_corners=True),
                 ScaleIntensityRanged(my_key, a_min=-21, a_max=189, b_min=0.0, b_max=1.0, clip=True),
                 CropForegroundd(my_key, source_key=my_key),
@@ -125,12 +136,5 @@ def post_process(self, pre_transforms: Compose, out_dir: str = "./prediction_out
                     keys=pred_key, transform=pre_transforms, orig_keys=self._input_dataset_key, nearest_interp=True
                 ),
                 SaveImaged(keys=pred_key, output_dir=out_dir, output_postfix="seg", output_dtype=uint8, resample=False),
-                SaveImaged(
-                    keys=self._input_dataset_key,
-                    output_dir=out_dir,
-                    output_postfix="",
-                    output_dtype=uint8,
-                    resample=False,
-                ),
             ]
         )

examples/apps/ai_unetr_seg_app/app.py

@@ -73,14 +73,14 @@ def compose(self):
         self.add_flow(study_loader_op, series_selector_op, {"dicom_study_list": "dicom_study_list"})
         self.add_flow(series_selector_op, series_to_vol_op, {"dicom_series": "dicom_series"})
         self.add_flow(series_to_vol_op, unetr_seg_op, {"image": "image"})
+        # Add the publishing operator to save the input and seg images for Render Server.
+        # Note the PublisherOperator has temp impl till a proper rendering module is created.
+        self.add_flow(unetr_seg_op, publisher_op, {"saved_images_folder": "saved_images_folder"})
         # Note below the dicom_seg_writer requires two inputs, each coming from a upstream operator.
         # Also note that the DICOMSegmentationWriterOperator may throw exception with some inputs.
         #   Bug has been created to track the issue.
         self.add_flow(series_selector_op, dicom_seg_writer, {"dicom_series": "dicom_series"})
         self.add_flow(unetr_seg_op, dicom_seg_writer, {"seg_image": "seg_image"})
-        # Add the publishing operator to save the input and seg images for Render Server.
-        # Note the PublisherOperator has temp impl till a proper rendering module is created.
-        self.add_flow(unetr_seg_op, publisher_op, {"saved_images_folder": "saved_images_folder"})
 
         self._logger.debug(f"End {self.compose.__name__}")
 

examples/apps/ai_unetr_seg_app/unetr_seg_operator.py

@@ -70,7 +70,12 @@ def compute(self, op_input: InputContext, op_output: OutputContext, context: Exe
 
         # This operator gets an in-memory Image object, so a specialized ImageReader is needed.
         _reader = InMemImageReader(input_image)
-        pre_transforms = self.pre_process(_reader)
+        # In this example, the input image, once loaded at the beginning of the pre-transforms, is
+        # saved on disk, so is the segmentation prediction image at the end of the post-transform.
+        # They are both saved in the same subfolder of the application output folder, with names
+        # distinguished by postfix. They can also be save in different subfolder if need be.
+        # These images files can then be packaged for rendering.
+        pre_transforms = self.pre_process(_reader, op_output_folder_path)
         post_transforms = self.post_process(pre_transforms, op_output_folder_path)
 
         # Delegates inference and saving output to the built-in operator.
@@ -91,14 +96,20 @@ def compute(self, op_input: InputContext, op_output: OutputContext, context: Exe
         # Now let the built-in operator handles the work with the I/O spec and execution context.
         infer_operator.compute(op_input, op_output, context)
 
-    def pre_process(self, img_reader) -> Compose:
+    def pre_process(self, img_reader, out_dir: str = "./input_images") -> Compose:
         """Composes transforms for preprocessing input before predicting on a model."""
 
         my_key = self._input_dataset_key
         return Compose(
             [
                 LoadImaged(keys=my_key, reader=img_reader),
                 AddChanneld(keys=my_key),
+                SaveImaged(
+                    keys=my_key,
+                    output_dir=out_dir,
+                    output_postfix="",
+                    resample=False,
+                ),
                 Spacingd(keys=my_key, pixdim=(1.5, 1.5, 2.0), mode=("bilinear")),
                 Orientationd(keys=my_key, axcodes="RAS"),
                 ScaleIntensityRanged(my_key, a_min=-175, a_max=250, b_min=0.0, b_max=1.0, clip=True),
@@ -119,12 +130,5 @@ def post_process(self, pre_transforms: Compose, out_dir: str = "./prediction_out
                     keys=pred_key, transform=pre_transforms, orig_keys=self._input_dataset_key, nearest_interp=True
                 ),
                 SaveImaged(keys=pred_key, output_dir=out_dir, output_postfix="seg", output_dtype=uint8, resample=False),
-                SaveImaged(
-                    keys=self._input_dataset_key,
-                    output_dir=out_dir,
-                    output_postfix="",
-                    output_dtype=uint8,
-                    resample=False,
-                ),
             ]
         )