cleaned up reading of dicom metadata

extras/fileformats/extras/application/medical.py

@@ -1,6 +1,6 @@
 import typing as ty
 from pathlib import Path
-import pydicom
+import pydicom.tag
 from fileformats.core import FileSet, extra_implementation
 from fileformats.application import Dicom
 import medimages4tests.dummy.dicom.mri.t1w.siemens.skyra.syngo_d13c
@@ -10,22 +10,11 @@
 @extra_implementation(FileSet.read_metadata)
 def dicom_read_metadata(
     dicom: Dicom,
-    specific_tags: ty.Optional[ty.Collection[str]] = None,
+    specific_tags: ty.Optional[pydicom.tag.TagListType] = None,
     **kwargs: ty.Any,
 ) -> ty.Mapping[str, ty.Any]:
-    dcm = pydicom.dcmread(
-        dicom.fspath,
-        specific_tags=list(specific_tags if specific_tags is not None else []),
-    )
-    [getattr(dcm, a, None) for a in dir(dcm)]  # Ensure all keywords are set
-    metadata = {
-        e.keyword: e.value
-        for e in dcm.elements()
-        if isinstance(e, pydicom.DataElement)
-        and getattr(e, "keyword", False)
-        and e.keyword != "PixelData"
-    }
-    return metadata
+    dcm = pydicom.dcmread(dicom.fspath, specific_tags=specific_tags)
+    return pydicom_to_dict(dcm)
 
 
 @extra_implementation(FileSet.generate_sample_data)
@@ -38,3 +27,29 @@ def dicom_generate_sample_data(
             out_dir=generator.dest_dir
         ).iterdir()
     )
+
+
+def pydicom_to_dict(
+    dcm: pydicom.Dataset, omit: ty.Collection[str] = ("PixelData",)
+) -> ty.Dict[str, ty.Any]:
+    """Convert a pydicom Dataset to a dictionary.
+
+    Parameters
+    ----------
+    dcm : pydicom.Dataset
+        The pydicom Dataset to convert.
+    omit : Collection[str], optional
+        A collection of keys to omit from the dictionary, by default ("PixelData",)
+    """
+    # Ensure that all keys are loaded before creating dictionary otherwise the keywords
+    # will not be set in the elem
+    [getattr(dcm, attr, None) for attr in dir(dcm)]
+    dct: ty.Dict[str, ty.Any] = {}
+    for elem in dcm.values():
+        if isinstance(elem, pydicom.DataElement) and elem.keyword:
+            key = elem.keyword
+        else:
+            key = elem.tag.json_key
+        if key not in omit:
+            dct[key] = elem.value
+    return dct

extras/fileformats/extras/application/tests/test_application_medical.py

@@ -1,3 +1,4 @@
+import pytest
 from fileformats.application import Dicom
 
 
@@ -6,3 +7,12 @@ def test_dicom_metadata():
     dicom = Dicom.sample()
 
     assert dicom.metadata["EchoTime"] == "2.07"
+
+
+def test_dicom_metadata_with_specific_tags():
+
+    dicom = Dicom(Dicom.sample(), specific_tags=["EchoTime"])
+
+    assert dicom.metadata["EchoTime"] == "2.07"
+    with pytest.raises(KeyError):
+        dicom.metadata["PatientName"]