- renaming

- rm config options

brainles_aurora/inferer/dataclasses.py

@@ -33,8 +33,6 @@ class AuroraInfererConfig(BaseConfig):
         sliding_window_overlap (float, optional): Overlap ratio for sliding window inference. Defaults to 0.5.
         crop_size (Tuple[int, int, int], optional): Crop size for sliding window inference. Defaults to (192, 192, 32).
         model_selection (ModelSelection, optional): Model selection strategy. Defaults to ModelSelection.BEST.
-        include_whole_network_in_numpy_output_mode (bool, optional): Whether to include the whole network in numpy output mode. Defaults to False.
-        include_metastasis_network_in_numpy_output_mode (bool, optional): Whether to include the metastasis network in numpy output mode. Defaults to False.
     """
 
     tta: bool = True
@@ -44,5 +42,3 @@ class AuroraInfererConfig(BaseConfig):
     sliding_window_overlap: float = 0.5
     crop_size: Tuple[int, int, int] = (192, 192, 32)
     model_selection: ModelSelection = ModelSelection.BEST
-    include_whole_network_in_numpy_output_mode: bool = False
-    include_metastasis_network_in_numpy_output_mode: bool = False

brainles_aurora/inferer/inferer.py

@@ -449,11 +449,6 @@ def _sliding_window_inference(self) -> None | Dict[str, np.ndarray]:
                     onehot_model_outputs_CHWD=outputs,
                 )
                 if self.config.output_mode == DataMode.NUMPY:
-                    # rm whole/ metastasus network if not requested
-                    if not self.config.include_whole_network_in_numpy_output_mode:
-                        _ = postprocessed_data.pop(Output.WHOLE_NETWORK)
-                    if not self.config.include_metastasis_network_in_numpy_output_mode:
-                        _ = postprocessed_data.pop(Output.METASTASIS_NETWORK)
                     return postprocessed_data
                 else:
                     self._save_as_nifti(postproc_data=postprocessed_data)
@@ -476,8 +471,8 @@ def infer(
         t2: str | Path | np.ndarray | None = None,
         fla: str | Path | np.ndarray | None = None,
         segmentation_file: str | Path | None = None,
-        whole_network_file: str | Path | None = None,
-        metastasis_network_file: str | Path | None = None,
+        whole_tumor_unbinarized_floats_file: str | Path | None = None,
+        metastasis_unbinarized_floats_file: str | Path | None = None,
         log_file: str | Path | None = None,
     ) -> Dict[str, np.ndarray] | None:
         """Perform inference on the provided images.
@@ -491,8 +486,8 @@ def infer(
             ### The following file paths are only required when in Nifti output mode:
 
             segmentation_file (str | Path | None, optional): Path where the segementation file should be stored. Defaults to None. Should be a nifti file. Defaults internally to a './segmentation.nii.gz'.
-            whole_network_file (str | Path | None, optional): Path. Defaults to None.
-            metastasis_network_file (str | Path | None, optional): _description_. Defaults to None.
+            whole_tumor_unbinarized_floats_file (str | Path | None, optional): Path. Defaults to None.
+            metastasis_unbinarized_floats_file (str | Path | None, optional): _description_. Defaults to None.
 
             ### Custom log file path. BY default this is set internally to the same path as segmentation_file with the extension .log or to ./{self.__class__.__name__}.log if no segmentation_file is provided
 
@@ -528,7 +523,7 @@ def infer(
             self.log.info(
                 f"Same inference mode {self.inference_mode} as previous infer call. Re-using loaded model"
             )
-
+        # self.model.eval()
         self.log.info("Setting up Dataloader")
         self.data_loader = self._get_data_loader()
 
@@ -542,8 +537,8 @@ def infer(
                 )
             self.output_file_mapping = {
                 Output.SEGMENTATION: segmentation_file or default_segmentation_path,
-                Output.WHOLE_NETWORK: whole_network_file,
-                Output.METASTASIS_NETWORK: metastasis_network_file,
+                Output.WHOLE_NETWORK: whole_tumor_unbinarized_floats_file,
+                Output.METASTASIS_NETWORK: metastasis_unbinarized_floats_file,
             }
 
         ########

segmentation_test.py

@@ -22,17 +22,36 @@ def load_np_from_nifti(path):
 
 def gpu_nifti():
     config = AuroraInfererConfig(
+        tta=False
+    )  # disable tta for faster inference in this showcase
+
+    # If you don-t have a GPU that supports CUDA use the CPU version: AuroraInferer(config=config)
+    inferer = AuroraGPUInferer(config=config)
+
+    inferer.infer(
         t1=t1,
-        t1c=t1c,
-        t2=t2,
-        fla=fla,
-        output_metastasis_network=True,
-        output_whole_network=True,
+        segmentation_file="test_output/segmentation.nii.gz",
     )
-    inferer = AuroraGPUInferer(
-        config=config,
+
+
+def gpu_nifti_2():
+    config = AuroraInfererConfig(
+        tta=False
+    )  # disable tta for faster inference in this showcase
+
+    # If you don-t have a GPU that supports CUDA use the CPU version: AuroraInferer(config=config)
+    inferer = AuroraGPUInferer(config=config)
+
+    inferer.infer(
+        t1=t1,
+        segmentation_file="test_output/nevergonna_seg.nii.gz",
+        whole_tumor_unbinarized_floats_file="test_output/whole_network.nii.gz",
+        metastasis_unbinarized_floats_file="test_output/metastasis_network.nii.gz",
+    )
+    inferer.infer(
+        t1=t1,
+        segmentation_file="test_output2/randomseg.nii.gz",
     )
-    inferer.infer()
 
 
 def cpu_nifti():
@@ -77,4 +96,4 @@ def gpu_output_np():
 
 
 if __name__ == "__main__":
-    gpu_nifti()
+    gpu_nifti_2()