Backend requirements update and support for the cavity class

.github/workflows/build_macos.yml

@@ -45,7 +45,7 @@ jobs:
     - name: Install dependencies
       run: |
         python -m pip install --upgrade pip
-        pip install git+https://github.com/dbouget/raidionics_rads_lib.git@software
+        pip install git+https://github.com/dbouget/raidionics_rads_lib.git@master
         pip install -r assets/requirements.txt
         pip install matplotlib==3.3.4
         pip install --force-reinstall --no-cache-dir pyside6

gui/SinglePatientComponent/CentralAreaExecutionWidget.py

@@ -146,7 +146,7 @@ def on_pipeline_execution(self, pipeline_code: str) -> None:
 
         """
         self.model_name = ""
-        if ("Classification" not in pipeline_code) and ("Brain" not in pipeline_code) and ("postop" not in pipeline_code):
+        if ("Classification" not in pipeline_code) and ("Brain" not in pipeline_code) and ("postop" not in pipeline_code) and ("Edema" not in pipeline_code) and ("Cavity" not in pipeline_code):
             diag = TumorTypeSelectionQDialog(self)
             code = diag.exec_()
             if code == 0:  # Operation cancelled
@@ -165,10 +165,23 @@ def on_pipeline_execution(self, pipeline_code: str) -> None:
                 self.model_name = self.model_name + '_multiclass'
                 if diag.tumor_type == 'Low-Grade Glioma':
                     self.model_name = "MRI_GBM_multiclass"
+        elif "postop" in pipeline_code:
+            diag = TumorTypeSelectionQDialog(self)
+            code = diag.exec_()
+            if code == 0:  # Operation cancelled
+                return
+            if diag.tumor_type == 'Glioblastoma':
+                self.model_name = "MRI_GBM_Postop_FV_4p"
+                pipeline_code = pipeline_code + '_GBM'
+            elif diag.tumor_type == 'Low-Grade Glioma':
+                self.model_name = "MRI_LGGlioma_Postop"
+                pipeline_code = pipeline_code + '_LGGlioma'
         elif "Brain" in pipeline_code:
             self.model_name = "MRI_Brain"
-        elif "postop" in pipeline_code:
-            self.model_name = "MRI_GBM_Postop_FV_4p"
+        elif "Edema" in pipeline_code:
+            self.model_name = "MRI_Edema"
+        elif "Cavity" in pipeline_code:
+            self.model_name = "MRI_Cavity"
 
         self.process_started.emit()
         self.pipeline_main_wrapper(pipeline_code)

utils/data_structures/AnnotationStructure.py

@@ -25,6 +25,7 @@ class AnnotationClassType(Enum):
     Tumor = 1, 'Tumor'
     Necrosis = 2, 'Necrosis'
     Edema = 3, 'Edema'
+    Cavity = 4, 'Cavity'
 
     Lungs = 100, 'Lungs'
     Airways = 101, 'Airways'

utils/logic/PipelineCreationHandler.py

@@ -63,8 +63,10 @@ def create_pipeline(model_name: str, patient_parameters, task: str) -> dict:
         return __create_folders_classification_pipeline()
     elif task == 'preop_segmentation':
         return __create_segmentation_pipeline(model_name, patient_parameters)
-    elif task == 'postop_segmentation':
-        model_name = select_appropriate_postop_model(patient_parameters)
+    elif 'postop_segmentation' in task:
+        # @TODO. Will have to clean up all this for dealing with the new use-cases...
+        if "GBM" in task:
+            model_name = select_appropriate_postop_model(patient_parameters)
         download_model(model_name=model_name)
         return __create_postop_segmentation_pipeline(model_name, patient_parameters)
     elif task == 'other_segmentation':
@@ -392,7 +394,7 @@ def __create_custom_pipeline(task, tumor_type, patient_parameters):
             pip[pip_num]["description"] = "Lungs segmentation in T1CE (T{})".format(str(timestamp_order))
             download_model(model_name="CT_Lungs")
 
-        if split_task[1] == 'Tumor':
+        if split_task[1] == 'Tumor' or split_task[1] == 'Edema' or split_task[1] == 'Cavity':
             infile = open(os.path.join(SoftwareConfigResources.getInstance().models_path, tumor_type, 'pipeline.json'),
                           'rb')
             raw_pip = json.load(infile)
@@ -401,7 +403,7 @@ def __create_custom_pipeline(task, tumor_type, patient_parameters):
                 pip_num_int = pip_num_int + 1
                 pip_num = str(pip_num_int)
                 pip[pip_num] = raw_pip[steps]
-        if split_task[1] == 'Brain':
+        elif split_task[1] == 'Brain':
             infile = open(os.path.join(SoftwareConfigResources.getInstance().models_path, tumor_type, 'pipeline.json'),
                           'rb')
             raw_pip = json.load(infile)