Added DICOM text SR assets and updated all operator test functions. (#201)

* Added DICOM text SR assets and updated all operator test functions.

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

* Correcr style checker errors.

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

* More style checker fix

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

* More style checker fix to add Union[<type>, None]

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

* Why the checker is so dumb, None for the typed object should be fine, and is handled.

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

* Correct SR class name

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

* Integrated DICOM SR Writer Operator to MedNIST classifier example

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

* Correct typos in the comments

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

* Add release note

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

* Added updates to notebooks and fixed issues.

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

* Cleaned up comment lines per review comments

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

* Updated release_notes index.md, as well as release note itself

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

Author: MMelQin

docs/source/release_notes/index.md

@@ -6,6 +6,12 @@
 
 ```
 
+## Version 0.2
+
+```{toctree}
+:maxdepth: 1
+
+v0.2.0
 ## Version 0.1
 
 ```{toctree}

docs/source/release_notes/v0.2.0.md

@@ -0,0 +1,33 @@
+# Version 0.2.0 (November 23, 2021)
+
+This is a new and enhanced version of MONAI Deploy App SDK, just in time for Thanksgiving and RSNA 2021!🎉
+
+Please visit [GETTING STARTED](/getting_started/index) guide and follow tutorials.
+
+You can learn more about SDK usage through [DEVELOPING WITH SDK](/developing_with_sdk/index).
+
+Please let us know how you like it and what could be improved by [submitting an issue](https://github.com/Project-MONAI/monai-deploy-app-sdk/issues/new/choose) or [asking questions](https://github.com/Project-MONAI/monai-deploy-app-sdk/discussions)😀
+
+## What's new in this version 0.2.0
+### Series Selection Operator
+This is to support the use case where whole DICOM studies are used as input to an AI inference application even though only specific series are applicable.
+
+The selection rules are defined in JSON, allowing multiple selections, each with a set of matching conditions. The rules processing engine is implemented in the `DICOMSeriesSelectorOperator`, which itself is regarded as a base class with a default implementation. More advanced rules and processing engines can be implemented in the derived classes.
+
+Multiple instances of the series selection operators, each having its own rules, can be chained in a MONAI Deploy application. In part this is made possible by the new App SDK Domain classes which encapsulate the selected series in a DICOM study, and are used as the output of each series selection operator.
+
+### DICOM Comprehensive Structured Report Writer
+This is introduced to support generating DICOM SR SOP instances for AI classification results, and as such, the DICOM SR writer is limited to supporting textual results only.
+
+The DICOM SR writer is implemented in `DICOMTextSRWriterOperator`, it
+- loads the AI result from a in-memory object as well as from a file path, with the in-memory object taking precedence
+- copies applicable DICOM tags from the original DICOM series used as input for the inference application, as well as generating tags anew when there is no DICOM series provided.
+- supports assigning DICOM tags via a dictionary with DICOM keywords and their respective values, so that an application can customize the tags in the DICOM SR instance
+- provides classes for an application to encapsulate the AI model information as well as DICOM equipment information, per [IHE Radiology Technical Framework Supplement AI Results (AIR)](https://www.ihe.net/uploadedFiles/Documents/Radiology/IHE_RAD_Suppl_AIR.pdf)
+
+### Updated example applications
+- [The AI Spleen Segmentation](https://github.com/Project-MONAI/monai-deploy-app-sdk/tree/main/examples/apps/ai_spleen_seg_app) application updated to demonstrate the use of series selection rules
+- [The MedNIST Classifier application](https://github.com/Project-MONAI/monai-deploy-app-sdk/tree/main/examples/apps/mednist_classifier_monaideploy) updated to demonstrate the use of DCIOM SR writing (without initial DICOM input)
+- Updated are the main functions of the built-in [operators](https://github.com/Project-MONAI/monai-deploy-app-sdk/tree/main/monai/deploy/operators), which serve as examples on how to parse the output objects
+- Updated also are [Jupyter notebook tutorials](https://github.com/Project-MONAI/monai-deploy-app-sdk/tree/main/notebooks/tutorials)
+- Multiple [issues](https://github.com/Project-MONAI/monai-deploy-app-sdk/issues?q=is%3Aissue+is%3Aclosed) were fixed and closed

examples/apps/mednist_classifier_monaideploy/mednist_classifier_monaideploy.py

@@ -9,6 +9,8 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
+from typing import Text
+
 import monai.deploy.core as md
 from monai.deploy.core import (
     Application,
@@ -20,6 +22,7 @@
     Operator,
     OutputContext,
 )
+from monai.deploy.operators.dicom_text_sr_writer_operator import DICOMTextSRWriterOperator, EquipmentInfo, ModelInfo
 from monai.transforms import AddChannel, Compose, EnsureType, ScaleIntensity
 
 MEDNIST_CLASSES = ["AbdomenCT", "BreastMRI", "CXR", "ChestCT", "Hand", "HeadCT"]
@@ -48,7 +51,7 @@ def compute(self, op_input: InputContext, op_output: OutputContext, context: Exe
 
 
 @md.input("image", Image, IOType.IN_MEMORY)
-@md.output("output", DataPath, IOType.DISK)
+@md.output("result_text", Text, IOType.IN_MEMORY)
 @md.env(pip_packages=["monai"])
 class MedNISTClassifierOperator(Operator):
     """Classifies the given image and returns the class name."""
@@ -80,9 +83,11 @@ def compute(self, op_input: InputContext, op_output: OutputContext, context: Exe
 
         result = MEDNIST_CLASSES[output_classes[0]]  # get the class name
         print(result)
+        op_output.set(result, "result_text")
 
         # Get output (folder) path and create the folder if not exists
-        output_folder = op_output.get().path
+        # The following gets the App context's output path, instead the operator's.
+        output_folder = context.output.get().path
         output_folder.mkdir(parents=True, exist_ok=True)
 
         # Write result to "output.json"
@@ -99,7 +104,15 @@ def compose(self):
         load_pil_op = LoadPILOperator()
         classifier_op = MedNISTClassifierOperator()
 
+        my_model_info = ModelInfo("MONAI WG Trainer", "MEDNIST Classifier", "0.1", "xyz")
+        my_equipment = EquipmentInfo(manufacturer="MOANI Deploy App SDK", manufacturer_model="DICOM SR Writer")
+        my_special_tags = {"SeriesDescription": "Not for clinical use. The result is for research use only."}
+        dicom_sr_operator = DICOMTextSRWriterOperator(
+            copy_tags=False, model_info=my_model_info, equipment_info=my_equipment, custom_tags=my_special_tags
+        )
+
         self.add_flow(load_pil_op, classifier_op)
+        self.add_flow(classifier_op, dicom_sr_operator, {"result_text": "classification_result"})
 
 
 if __name__ == "__main__":

monai/deploy/core/domain/dicom_sop_instance.py

@@ -21,9 +21,9 @@
 Dataset_, dataset_ok_ = optional_import("pydicom", name="Dataset")
 # Dynamic class is not handled so make it Any for now: https://github.com/python/mypy/issues/2477
 Dataset: Any = Dataset_ if dataset_ok_ else Any
-TagType_, tagtype_ok_ = optional_import("pydicom.tag", name="TagType")
+Tag_, tag_ok_ = optional_import("pydicom.tag", name="Tag")
 # Dynamic class is not handled so make it Any for now: https://github.com/python/mypy/issues/2477
-TagType: Any = TagType_ if tagtype_ok_ else Any
+Tag: Any = Tag_ if tag_ok_ else Any
 
 
 class DICOMSOPInstance(Domain):
@@ -39,7 +39,7 @@ def __init__(self, native_sop):
     def get_native_sop_instance(self):
         return self._sop
 
-    def __getitem__(self, key: Union[int, slice, TagType]) -> Union[Dataset, DataElement]:
+    def __getitem__(self, key: Union[int, slice, Tag]) -> Union[Dataset, DataElement]:
         return self._sop.__getitem__(key)
 
     def get_pixel_array(self):

monai/deploy/operators/__init__.py

@@ -26,6 +26,7 @@
 from .dicom_seg_writer_operator import DICOMSegmentationWriterOperator
 from .dicom_series_selector_operator import DICOMSeriesSelectorOperator
 from .dicom_series_to_volume_operator import DICOMSeriesToVolumeOperator
+from .dicom_text_sr_writer_operator import DICOMTextSRWriterOperator, EquipmentInfo, ModelInfo
 from .inference_operator import InferenceOperator
 from .monai_seg_inference_operator import MonaiSegInferenceOperator
 from .png_converter_operator import PNGConverterOperator

monai/deploy/operators/dicom_data_loader_operator.py

@@ -245,16 +245,50 @@ def populate_series_attributes(self, series, sop_instance):
 
 
 def test():
-    data_path = "../../../examples/ai_spleen_seg_data/dcm"
+    current_file_dir = Path(__file__).parent.resolve()
+    data_path = current_file_dir.joinpath("../../../examples/ai_spleen_seg_data/dcm")
 
     loader = DICOMDataLoaderOperator()
-    study_list = loader.load_data_to_studies(Path(data_path).absolute())
+    study_list = loader.load_data_to_studies(data_path.absolute())
 
     for study in study_list:
         print("###############################")
         print(study)
         for series in study.get_all_series():
             print(series)
+        for series in study.get_all_series():
+            for sop in series.get_sop_instances():
+                print("Demonstrating ways to access DICOM attributes in a SOP instance.")
+                # No need to get the native_ds = sop.get_native_sop_instance()
+                # sop = sop.get_native_sop_instance()
+                print(f"   'StudyInstanceUID': {sop['StudyInstanceUID'].repval}")
+                print(f"     (0x0020, 0x000D): {sop[0x0020, 0x000D].repval}")
+                print(f"  'SeriesInstanceUID': {sop['SeriesInstanceUID'].value.name}")
+                print(f"     (0x0020, 0x000E): {sop[0x0020, 0x000E].value.name}")
+                print(f"     'SOPInstanceUID': {sop['SOPInstanceUID'].value.name}")
+                print(f"          (0008,0018): {sop[0x0008, 0x0018].value.name}")
+                try:
+                    print(f"     'InstanceNumber': {sop['InstanceNumber'].repval}")
+                    print(f"         (0020, 0013): {sop[0x0020, 0x0013].repval}")
+                except KeyError:
+                    pass
+                # Need to get pydicom dataset to use properties and get method of a dict.
+                ds = sop.get_native_sop_instance()
+                print(f"   'StudyInstanceUID': {ds.StudyInstanceUID if ds.StudyInstanceUID else ''}")
+                print(f"  'SeriesDescription': {ds.SeriesDescription if ds.SeriesDescription else ''}")
+                print(
+                    f"  'IssuerOfPatientID': {ds.get('IssuerOfPatientID', '').repval if ds.get('IssuerOfPatientID', '') else '' }"
+                )
+                try:
+                    print(f"  'IssuerOfPatientID': {ds.IssuerOfPatientID if ds.IssuerOfPatientID else '' }")
+                except AttributeError:
+                    print(
+                        "    If the IssuerOfPatientID does not exist, ds.IssuerOfPatientID would throw AttributeError."
+                    )
+                    print("    Use ds.get('IssuerOfPatientID', '') instead.")
+
+                break
+            break
 
 
 if __name__ == "__main__":

monai/deploy/operators/dicom_seg_writer_operator.py

@@ -660,17 +660,17 @@ def test():
     from monai.deploy.operators.dicom_series_selector_operator import DICOMSeriesSelectorOperator
     from monai.deploy.operators.dicom_series_to_volume_operator import DICOMSeriesToVolumeOperator
 
-    data_path = "../../../examples/ai_spleen_seg_data/dcm"
-    out_path = "../../../examples/output_seg_op/dcm_seg_test.dcm"
+    current_file_dir = Path(__file__).parent.resolve()
+    data_path = current_file_dir.joinpath("../../../examples/ai_spleen_seg_data/dcm")
+    out_path = current_file_dir.joinpath("../../../examples/output_seg_op/dcm_seg_test.dcm")
 
-    files = []
     loader = DICOMDataLoaderOperator()
     series_selector = DICOMSeriesSelectorOperator()
     dcm_to_volume_op = DICOMSeriesToVolumeOperator()
     seg_writer = DICOMSegmentationWriterOperator()
 
     # Testing with more granular functions
-    study_list = loader.load_data_to_studies(Path(data_path).absolute())
+    study_list = loader.load_data_to_studies(data_path.absolute())
     series = study_list[0].get_all_series()[0]
 
     dcm_to_volume_op.prepare_series(series)
@@ -682,10 +682,10 @@ def test():
     seg_writer.create_dicom_seg(image_numpy, series, Path(out_path).absolute())
 
     # Testing with the main entry functions
-    study_list = loader.load_data_to_studies(Path(data_path).absolute())
-    _, study_selected_series_list = series_selector.filter(None, study_list)
+    study_list = loader.load_data_to_studies(data_path.absolute())
+    study_selected_series_list = series_selector.filter(None, study_list)
     image = dcm_to_volume_op.convert_to_image(study_selected_series_list)
-    seg_writer.process_images(image, study_selected_series_list, Path(out_path).parent.absolute())
+    seg_writer.process_images(image, study_selected_series_list, out_path.parent.absolute())
 
 
 if __name__ == "__main__":

monai/deploy/operators/dicom_series_selector_operator.py

@@ -277,14 +277,15 @@ def test():
 
     from monai.deploy.operators.dicom_data_loader_operator import DICOMDataLoaderOperator
 
-    data_path = "../../../examples/ai_spleen_seg_data/dcm-multi"
+    current_file_dir = Path(__file__).parent.resolve()
+    data_path = current_file_dir.joinpath("../../../examples/ai_spleen_seg_data/dcm-multi")
 
     loader = DICOMDataLoaderOperator()
-    study_list = loader.load_data_to_studies(Path(data_path).absolute())
+    study_list = loader.load_data_to_studies(data_path.absolute())
     selector = DICOMSeriesSelectorOperator()
     sample_selection_rule = json_loads(Sample_Rules_Text)
     print(f"Selection rules in JSON:\n{sample_selection_rule}")
-    series_list, study_selected_seriee_list = selector.filter(sample_selection_rule, study_list)
+    study_selected_seriee_list = selector.filter(sample_selection_rule, study_list)
 
     for sss_obj in study_selected_seriee_list:
         _print_instance_properties(sss_obj, pre_fix="", print_val=False)
@@ -314,7 +315,7 @@ def test():
                 _print_instance_properties(ss_obj, pre_fix)
                 print(f"{pre_fix}===============================")
 
-    print(f"Total # of series selected: {len(series_list)}")
+        print(f"  A total of {len(sss_obj.selected_series)} series selected for study {study.StudyInstanceUID}")
 
 
 # Sample rule used for testing

monai/deploy/operators/dicom_series_to_volume_operator.py

@@ -346,12 +346,13 @@ def test():
     from monai.deploy.operators.dicom_data_loader_operator import DICOMDataLoaderOperator
     from monai.deploy.operators.dicom_series_selector_operator import DICOMSeriesSelectorOperator
 
-    data_path = "../../../examples/ai_spleen_seg_data/dcm"
+    current_file_dir = Path(__file__).parent.resolve()
+    data_path = current_file_dir.joinpath("../../../examples/ai_spleen_seg_data/dcm")
     loader = DICOMDataLoaderOperator()
     study_list = loader.load_data_to_studies(Path(data_path).absolute())
 
     series_selector = DICOMSeriesSelectorOperator()
-    _, study_selected_series_list = series_selector.filter(None, study_list)
+    study_selected_series_list = series_selector.filter(None, study_list)
 
     op = DICOMSeriesToVolumeOperator()
     image = op.convert_to_image(study_selected_series_list)