Remove bone lesion

Author: heyufan1995

configs/train/hyper_parameters_stage1.yaml

@@ -44,7 +44,7 @@ skip_iter_prob: 0
 train_datasets: [CTPelvic1K-CLINIC, AbdomenCT-1K, AeroPath, AMOS22,
   BTCV-Abdomen, BTCV-Cervix, CT-ORG, FLARE22, Multi-organ-Abdominal-CT-btcv,
   Multi-organ-Abdominal-CT-tcia, Pancreas-CT, Task03, Task06, Task07,
-  Task08, Task09, Task10, VerSe, Bone-NIH, CRLM-CT, TotalSegmentatorV2, NLST, LIDC, StonyBrook-CT, TCIA_Colon, Covid19]
+  Task08, Task09, Task10, VerSe, CRLM-CT, TotalSegmentatorV2, NLST, LIDC, StonyBrook-CT, TCIA_Colon, Covid19]
 val_datasets: [TotalSegmentatorV2]
 transforms_train:
   _target_: Compose

configs/train/hyper_parameters_stage2.yaml

@@ -42,9 +42,9 @@ resample_to_spacing: [1.5, 1.5, 1.5]
 skip_iter_prob: 0
 train_datasets: [CTPelvic1K-CLINIC, AbdomenCT-1K, AeroPath, AMOS22, BTCV-Abdomen,
   BTCV-Cervix, CT-ORG, FLARE22, Multi-organ-Abdominal-CT-btcv, Multi-organ-Abdominal-CT-tcia,
-  Pancreas-CT, Task03, Task06, Task07, Task08, Task09, Task10, VerSe, Bone-NIH, CRLM-CT,
+  Pancreas-CT, Task03, Task06, Task07, Task08, Task09, Task10, VerSe, CRLM-CT,
   TotalSegmentatorV2]
-val_datasets: ['CRLM-CT', 'AeroPath', 'Task03','Task06','Task07','Task08','Task10','Bone-NIH']
+val_datasets: ['CRLM-CT', 'AeroPath', 'Task03','Task06','Task07','Task08','Task10']
 transforms_train:
   _target_: Compose
   transforms:

configs/train/hyper_parameters_stage3.yaml

@@ -43,7 +43,7 @@ resample_to_spacing: [1.5, 1.5, 1.5]
 skip_iter_prob: 1
 train_datasets: [CTPelvic1K-CLINIC, AbdomenCT-1K, AeroPath, AMOS22, BTCV-Abdomen,
   BTCV-Cervix, CT-ORG, FLARE22, Multi-organ-Abdominal-CT-btcv, Multi-organ-Abdominal-CT-tcia,
-  Pancreas-CT, Task03, Task06, Task07, Task08, Task09, Task10, VerSe, Bone-NIH, CRLM-CT,
+  Pancreas-CT, Task03, Task06, Task07, Task08, Task09, Task10, VerSe, CRLM-CT,
   TotalSegmentatorV2, NLST, LIDC, StonyBrook-CT, TCIA_Colon]
 val_datasets: ['TotalSegmentatorV2']
 transforms_train:

configs/train/hyper_parameters_stage4.yaml

@@ -43,7 +43,7 @@ resample_to_spacing: [1.5, 1.5, 1.5]
 skip_iter_prob: 1
 train_datasets: [CTPelvic1K-CLINIC, AbdomenCT-1K, AeroPath, AMOS22, BTCV-Abdomen,
   BTCV-Cervix, CT-ORG, FLARE22, Multi-organ-Abdominal-CT-btcv, Multi-organ-Abdominal-CT-tcia,
-  Pancreas-CT, Task03, Task06, Task07, Task08, Task09, Task10, VerSe, Bone-NIH, CRLM-CT,
+  Pancreas-CT, Task03, Task06, Task07, Task08, Task09, Task10, VerSe, CRLM-CT,
   TotalSegmentatorV2]
 val_datasets: ['CRLM-CT', 'AeroPath', 'Task03','Task06','Task07','Task08','Task10','Bone-NIH']
 transforms_train:

scripts/utils/sample_utils.py

@@ -289,6 +289,7 @@ def organ_add(self, id, Np=1, Nn=0):
                 self.label[self.label == id] = id + self.map_shift
                 self.shifted[id] = id + self.map_shift
                 self.read_only_id.append(id + self.map_shift)
+            print("organ add")
             return _point, _point_label
         return self.regular(id, Np, Nn)