got rid of file_name_convention stuff

examples/autotest.py

@@ -45,7 +45,8 @@ def __init__(self,
                  is_nnunet=False,
                  train_size=1.0,
                  random_state=42,
-                 label_names=None):
+                 label_names=None,
+                 ignore_missing_regex=False):
         """Initialize the pipeline.
 
         Parameters
@@ -76,6 +77,10 @@ def __init__(self,
             Proportion of the dataset to use for training, as a decimal
         random_state: int, default=42
             Random state for train_test_split
+        label_names: dict of str:str, default=None
+            Dictionary representing the label that regexes are mapped to. For example, "GTV": "GTV.*" will combine all regexes that match "GTV.*" into "GTV"
+        ignore_missing_regex: bool, default=False
+            Whether to ignore missing regexes. Will raise an error if none of the regexes in label_names are found for a patient.
         """
         super().__init__(
             n_jobs=n_jobs,
@@ -96,6 +101,7 @@ def __init__(self,
         self.train_size = train_size
         self.random_state = random_state
         self.label_names = label_names
+        self.ignore_missing_regex = ignore_missing_regex
 
         if self.train_size == 1.0:
             warnings.warn("Train size is 1, all data will be used for training")
@@ -246,12 +252,15 @@ def process_one_subject(self, subject_id):
 
                 # make_binary_mask relative to ct/pet
                 if conn_to == "CT" or conn_to == "MR":
-                    mask = self.make_binary_mask(structure_set, image, self.existing_roi_names)
+                    mask = self.make_binary_mask(structure_set, image, self.existing_roi_names, self.ignore_missing_regex)
                 elif conn_to == "PT":
-                    mask = self.make_binary_mask(structure_set, pet, self.existing_roi_names)
+                    mask = self.make_binary_mask(structure_set, pet, self.existing_roi_names, self.ignore_missing_regex)
                 else:
                     raise ValueError("You need to pass a reference CT or PT/PET image to map contours to.")
 
+                if mask is None: #ignored the missing regex
+                    return
+
                 for name in mask.roi_names.keys():
                     if name not in self.existing_roi_names.keys():
                         self.existing_roi_names[name] = len(self.existing_roi_names)
@@ -312,7 +321,7 @@ def process_one_subject(self, subject_id):
 
                 print(subject_id, " SAVED PET")
 
-            metadata[f"output_folder_{colname}"] = pathlib.Path(subject_id, file_name_convention()[colname]).as_posix()
+            metadata[f"output_folder_{colname}"] = pathlib.Path(subject_id, colname).as_posix()
         #Saving all the metadata in multiple text files
         metadata["Modalities"] = str(list(subject_modalities))
         metadata["numRTSTRUCTs"] = num_rtstructs
@@ -390,9 +399,10 @@ def run(self):
                             modalities="CT,RTSTRUCT",
                             visualize=False,
                             overwrite=True,
-                            is_nnunet=True,
+                            # is_nnunet=True,
                             train_size=0.5,
-                            label_names={"GTV":"GTV.*", "Brainstem": "Brainstem.*"})
+                            # label_names={"GTV":"GTV.*", "Brainstem": "Brainstem.*"},
+                            ignore_missing_regex=True)
 
     # pipeline = AutoPipeline(input_directory="C:/Users/qukev/BHKLAB/dataset/manifest-1598890146597/NSCLC-Radiomics-Interobserver1",
     #                         output_directory="C:/Users/qukev/BHKLAB/autopipelineoutput",

imgtools/modules/structureset.py

@@ -89,7 +89,8 @@ def to_segmentation(self, reference_image: sitk.Image,
                         roi_names: Dict[str, str] = None,
                         force_missing: bool = False,
                         continuous: bool = True,
-                        existing_roi_names: Dict[str, int] = None) -> Segmentation:
+                        existing_roi_names: Dict[str, int] = None,
+                        ignore_missing_regex: bool = False) -> Segmentation:
         """Convert the structure set to a Segmentation object.
 
         Parameters
@@ -142,7 +143,10 @@ def to_segmentation(self, reference_image: sitk.Image,
             labels = self._assign_labels(roi_names, force_missing)
         print("labels:", labels)
         if not labels:
-            raise ValueError(f"No ROIs matching {roi_names} found in {self.roi_names}.")
+            if ignore_missing_regex:
+                raise ValueError(f"No ROIs matching {roi_names} found in {self.roi_names}.")
+            else:
+                return None
 
         # size = reference_image.GetSize()[::-1] + (max(labels.values()) + 1,)
         size = reference_image.GetSize()[::-1] + (len(labels),)

imgtools/ops/ops.py

@@ -325,18 +325,16 @@ def __init__(self,
                  nnunet_info: Dict = None):
 
         # File types
-        self.file_name = file_name_convention()
         # print(self.file_name)
 
         self.output = {}
         print(output_streams)
         for colname in output_streams:
             # Not considering colnames ending with alphanumeric
             colname_process = ("_").join([item for item in colname.split("_") if item.isnumeric()==False])
-            extension = self.file_name[colname_process]
             if not nnunet_info:
-                self.output[colname_process] = ImageSubjectFileOutput(pathlib.Path(root_directory,"{subject_id}",extension.split(".")[0]).as_posix(),
-                                                                    filename_format=colname_process+"{}.nii.gz".format(extension))
+                self.output[colname_process] = ImageSubjectFileOutput(pathlib.Path(root_directory,"{subject_id}",colname_process).as_posix(),
+                                                                    filename_format=colname_process+".nii.gz")
             else:
                 self.output[colname_process] = ImageSubjectFileOutput(pathlib.Path(root_directory,"{label_or_image}{train_or_test}").as_posix(),
                                                                     filename_format="{subject_id}_{modality_index}.nii.gz")
@@ -1469,7 +1467,7 @@ def __init__(self,
         self.force_missing = force_missing
         self.continuous = continuous
 
-    def __call__(self, structure_set: StructureSet, reference_image: sitk.Image, existing_roi_names: Dict[str, int]) -> Segmentation:
+    def __call__(self, structure_set: StructureSet, reference_image: sitk.Image, existing_roi_names: Dict[str, int], ignore_missing_regex: bool) -> Segmentation:
         """Convert the structure set to a Segmentation object.
 
         Parameters
@@ -1488,7 +1486,8 @@ def __call__(self, structure_set: StructureSet, reference_image: sitk.Image, exi
                                              roi_names=self.roi_names,
                                              force_missing=self.force_missing,
                                              continuous=self.continuous,
-                                             existing_roi_names=existing_roi_names)
+                                             existing_roi_names=existing_roi_names,
+                                             ignore_missing_regex=ignore_missing_regex)
 
 class MapOverLabels(BaseOp):
     """MapOverLabels operation class:

imgtools/utils/args.py

@@ -34,4 +34,7 @@ def parser():
     parser.add_argument("--random_state", type=int, default=42,
                         help="The random state to be used for the train-test-split.")
 
+    parser.add_argument("--ignore_missing_roi_regex", default=False, action="store_true",
+                        help="Whether to ignore patients with no ROI regex's that match the given ones. Will throw an error on patients without matches if this is not set.")
+
     return parser.parse_known_args()[0]