Seriesgrouping

heudiconv/dicoms.py

@@ -12,6 +12,7 @@
 from typing import TYPE_CHECKING, Any, Dict, List, NamedTuple, Optional, Union, overload
 from unittest.mock import patch
 import warnings
+import hashlib
 
 import pydicom as dcm
 
@@ -114,7 +115,7 @@ def create_seqinfo(mw: dw.Wrapper, series_files: list[str], series_id: str) -> S
 
 def validate_dicom(
     fl: str, dcmfilter: Optional[Callable[[dcm.dataset.Dataset], Any]]
-) -> Optional[tuple[dw.Wrapper, tuple[int, str], Optional[str]]]:
+) -> Optional[tuple[dw.Wrapper, tuple[int, str, str], Optional[str]]]:
     """
     Parse DICOM attributes. Returns None if not valid.
     """
@@ -135,7 +136,7 @@ def validate_dicom(
     try:
         protocol_name = mw.dcm_data.ProtocolName
         assert isinstance(protocol_name, str)
-        series_id = (int(mw.dcm_data.SeriesNumber), protocol_name)
+        series_id = (int(mw.dcm_data.SeriesNumber), protocol_name, mw.dcm_data.SeriesInstanceUID)
     except AttributeError as e:
         lgr.warning('Ignoring %s since not quite a "normal" DICOM: %s', fl, e)
         return None
@@ -159,10 +160,12 @@ def validate_dicom(
 class SeriesID(NamedTuple):
     series_number: int
     protocol_name: str
+    series_uid: str
     file_studyUID: Optional[str] = None
 
     def __str__(self) -> str:
-        s = f"{self.series_number}-{self.protocol_name}"
+        suid_hex = hashlib.md5(self.series_uid.encode('utf-8')).hexdigest()[:16]
+        s = f"{self.series_number}-{self.protocol_name}-{suid_hex}"
         if self.file_studyUID is not None:
             s += f"-{self.file_studyUID}"
         return s
@@ -285,7 +288,7 @@ def group_dicoms_into_seqinfos(
             removeidx.append(idx)
             continue
         mw, series_id_, file_studyUID = mwinfo
-        series_id = SeriesID(series_id_[0], series_id_[1])
+        series_id = SeriesID(series_id_[0], series_id_[1], series_id_[2])
         if per_studyUID:
             series_id = series_id._replace(file_studyUID=file_studyUID)
 
@@ -320,6 +323,7 @@ def group_dicoms_into_seqinfos(
                 series_id = SeriesID(
                     mwgroup[idx].dcm_data.SeriesNumber,
                     mwgroup[idx].dcm_data.ProtocolName,
+                    mwgroup[idx].dcm_data.SeriesInstanceUID,
                 )
                 if per_studyUID:
                     series_id = series_id._replace(file_studyUID=file_studyUID)

heudiconv/parser.py

@@ -278,15 +278,9 @@ def infotoids(
             )
             lgr.info("Study session for %r", study_session_info)
 
-            if study_session_info in study_sessions:
-                if grouping != "all":
-                    # MG - should this blow up to mimic -d invocation?
-                    lgr.warning(
-                        "Existing study session with the same values (%r)."
-                        " Skipping DICOMS %s",
-                        study_session_info,
-                        seqinfo.values(),
-                    )
-                    continue
+            if grouping != "all":
+                assert (study_session_info not in study_sessions), (
+                    f"Existing study session {study_session_info} "
+                    f"already in analyzed sessions {study_sessions.keys()}")
             study_sessions[study_session_info] = seqinfo
     return study_sessions