Merge pull request #12 from qurit/10-error-with-unreferenced-contour-…

…images

feat: throw exception if rtstruct references images outside of series data

rt_utils/rtstruct_builder.py

@@ -1,3 +1,5 @@
+from typing import List
+from pydicom.dataset import Dataset
 from pydicom.filereader import dcmread
 
 from rt_utils.utils import SOPClassUID
@@ -29,14 +31,46 @@ def create_from(dicom_series_path: str, rt_struct_path: str) -> RTStruct:
         series_data = image_helper.load_sorted_image_series(dicom_series_path)
         ds = dcmread(rt_struct_path)
         RTStructBuilder.validate_rtstruct(ds)
+        RTStructBuilder.validate_rtstruct_series_references(ds, series_data)
 
-        # TODO create new frame of reference?
+        # TODO create new frame of reference? Right now we assume the last frame of reference created is suitable 
         return RTStruct(series_data, ds)
 
     @staticmethod
-    def validate_rtstruct(ds):
+    def validate_rtstruct(ds: Dataset):
+        """
+        Method to validate a dataset is a valid RTStruct containing the required fields
+        """
+
         if ds.SOPClassUID != SOPClassUID.RTSTRUCT or \
                 not hasattr(ds, 'ROIContourSequence') or \
                 not hasattr(ds, 'StructureSetROISequence') or \
                 not hasattr(ds, 'RTROIObservationsSequence'):
             raise Exception("Please check that the existing RTStruct is valid")
+
+    @staticmethod
+    def validate_rtstruct_series_references(ds: Dataset, series_data: List[Dataset]):
+        """
+        Method to validate RTStruct only references dicom images found within the input series_data
+        """
+        for refd_frame_of_ref in ds.ReferencedFrameOfReferenceSequence:
+            for rt_refd_study in refd_frame_of_ref.RTReferencedStudySequence:
+                for rt_refd_series in rt_refd_study.RTReferencedSeriesSequence:
+                    for contour_image in rt_refd_series.ContourImageSequence:
+                        RTStructBuilder.validate_contour_image_in_series_data(
+                            contour_image, series_data)
+
+    @staticmethod
+    def validate_contour_image_in_series_data(contour_image: Dataset, series_data: List[Dataset]):
+        """
+        Method to validate that the ReferencedSOPInstanceUID of a given contour image exists within the series data
+        """
+        for series in series_data:
+            if contour_image.ReferencedSOPInstanceUID == series.file_meta.MediaStorageSOPInstanceUID:
+                return
+
+        # ReferencedSOPInstanceUID is NOT available
+        raise Exception(
+            f'Loaded RTStruct references image(s) that are not contained in input series data. ' + 
+            f'Problematic image has SOP Instance Id: {contour_image.ReferencedSOPInstanceUID}'
+        )

setup.py

@@ -1,6 +1,6 @@
 import setuptools
 
-VERSION = '1.0.4'
+VERSION = '1.0.5'
 with open("README.md", "r", encoding="utf-8") as fh:
     long_description = fh.read()
 with open('requirements.txt') as f:

tests/test_rtstruct_builder.py

@@ -17,7 +17,7 @@ def test_create_from_empty_series_dir():
 def test_only_images_loaded_into_series_data(new_rtstruct: RTStruct):
     assert len(new_rtstruct.series_data) > 0
     for ds in new_rtstruct.series_data:
-        assert ds.SOPClassUID == SOPClassUID.CT_IMAGE_STORAGE
+        assert hasattr(ds, 'pixel_array')
 
 
 def test_valid_filemeta(new_rtstruct: RTStruct):
@@ -82,6 +82,14 @@ def test_loading_invalid_rt_struct(series_path):
         RTStructBuilder.create_from(series_path, invalid_rt_struct_path)
 
 
+def test_loading_invalid_reference_rt_struct(series_path):
+    # This RTStruct references images not found within the series path
+    invalid_reference_rt_struct_path = os.path.join(series_path, 'invalid_reference_rt.dcm')
+    assert os.path.exists(invalid_reference_rt_struct_path)
+    with pytest.raises(Exception):
+        RTStructBuilder.create_from(series_path, invalid_reference_rt_struct_path)
+
+
 def test_loading_valid_rt_struct(series_path):
     valid_rt_struct_path = os.path.join(series_path, 'rt.dcm')
     assert os.path.exists(valid_rt_struct_path)
@@ -91,20 +99,20 @@ def test_loading_valid_rt_struct(series_path):
     assert hasattr(rtstruct.ds, 'ROIContourSequence')
     assert hasattr(rtstruct.ds, 'StructureSetROISequence')
     assert hasattr(rtstruct.ds, 'RTROIObservationsSequence')
-    assert len(rtstruct.ds.ROIContourSequence) == 4
-    assert len(rtstruct.ds.StructureSetROISequence) == 4
-    assert len(rtstruct.ds.RTROIObservationsSequence) == 4
+    assert len(rtstruct.ds.ROIContourSequence) == 1
+    assert len(rtstruct.ds.StructureSetROISequence) == 1
+    assert len(rtstruct.ds.RTROIObservationsSequence) == 1
 
     # Test adding a new ROI
     mask = get_empty_mask(rtstruct)
     mask[50:100,50:100,0] = 1
     rtstruct.add_roi(mask)
 
-    assert len(rtstruct.ds.ROIContourSequence) == 5 # 1 should be added
-    assert len(rtstruct.ds.StructureSetROISequence) == 5 # 1 should be added
-    assert len(rtstruct.ds.RTROIObservationsSequence) == 5 # 1 should be added
+    assert len(rtstruct.ds.ROIContourSequence) == 2 # 1 should be added
+    assert len(rtstruct.ds.StructureSetROISequence) == 2 # 1 should be added
+    assert len(rtstruct.ds.RTROIObservationsSequence) == 2 # 1 should be added
     new_roi = rtstruct.ds.StructureSetROISequence[-1]
-    assert new_roi.ROIName == 'ROI-5'
+    assert new_roi.ROIName == 'ROI-2'
 
 
 def test_loaded_mask_iou(new_rtstruct: RTStruct):
@@ -144,13 +152,12 @@ def run_mask_iou_test(rtstruct:RTStruct, mask, IOU_threshold, use_pin_hole=False
     loaded_mask = rtstruct.get_roi_mask_by_name(mask_name)
 
     # Use IOU to test accuracy of loaded mask
-    print(np.sum(mask))
-    print(np.sum(loaded_mask))
     numerator = np.logical_and(mask, loaded_mask)
     denominator = np.logical_or(mask, loaded_mask)
     IOU =  np.sum(numerator) / np.sum(denominator) 
     assert IOU >= IOU_threshold
 
+
 def get_empty_mask(rtstruct) -> np.ndarray:
     ref_dicom_image = rtstruct.series_data[0]
     mask_dims = (int(ref_dicom_image.Columns), int(ref_dicom_image.Rows), len(rtstruct.series_data))