d

RobinNet/13unet_mrt_threshgt.py

@@ -179,7 +179,7 @@ def main():
             #number_patients = number_patients * 5
             img_path = "data/npy_thresh/"
 
-    specificmodels = [2]
+    specificmodels = [7]
     if cross_val:
         log = open("logs" + modelname + ".txt", "w+")
         log.write(modelname + "\r")

RobinNet/nets/BaseNetwork.py

@@ -442,7 +442,7 @@ class BaseNetwork:
             self.log.write('Loading Training Data ...'+"\r")
 
         ### initializing weights ###
-        weights = init_weights(manual_seed=np.sum(self.test_patients)*10)
+        weights = init_weights(manual_seed=np.sum(self.test_patients))
 
         filter_multiplier = int(self.newSize[0] / 128)
 

RobinNet/nets/Unet.py

@@ -135,8 +135,8 @@ def train_unet(model, inputs, gt, weights, optimizer, filter_multiplier):
 
 def init_weights(manual_seed):
     ### initializing weights ###
-    #initializer = tf.initializers.glorot_uniform(seed=manual_seed)
-    initializer = tf.keras.initializers.TruncatedNormal(seed=manual_seed)
+    initializer = tf.initializers.glorot_uniform(seed=manual_seed)
+    #initializer = tf.keras.initializers.TruncatedNormal(seed=manual_seed)
     shapes = [  # filter_height, filter_width, in_channels, out_channels
                 # for conv2d_transpose: filter_height, filter_width, out_channels, in_channels
         [3, 3, 1, 16],

RobinNet/rerunModels.py

@@ -19,11 +19,11 @@ from nets.Unet import *
 
 def main():
     models = ["12", "34", "56", "78", "910", "1112", "1314", "1516"]
-    folder_path = "finalResults/complete_seg/ae_class_cv_seg/"
+    folder_path = "finalResults/complete_segmr/mr_unet_cv_ctthresh/"
     #modeltype = "Cluster_"
     #modeltype = "Class_"
-    modeltype = "Cluster_class_"
-    #modeltype = "Unet_"
+    #modeltype = "Cluster_class_"
+    modeltype = "Unet_"
     all3dhdds = []
 
     ##################### U-Net #####################
@@ -62,7 +62,7 @@ def main():
         print('GT Type:', gt_type)
 
         ### Load test patient
-        if gt_type == "thresh" or gt_type == "ctthresh_gt":
+        if gt_type == "thresh" or gt_type == "ctthresh_gt" or gt_type == "_mr_ctunetpred":
             img_path = "data/npy_thresh/"
         else:
             img_path = "data/npy/"
@@ -106,6 +106,8 @@ def main():
                     if gt_type == "thresh":
                         ytrue_img_list.append(elem)
                         ytrue_img_list.append(elem)
+        #if gt_type =="_mr_ctunetpred":
+            #X_img_list.remove("P17_mr_T1.gipl.npy")
         list.sort(X_img_list)
         list.sort(GT_img_list)
         list.sort(ytrue_img_list)
@@ -242,11 +244,14 @@ def main():
             pred3d = np.float(pred3d)
             print(np.shape(pred3d))
             print(type(pred3d))'''
-            hausdorff_distance_filter = sitk.HausdorffDistanceImageFilter()
-            hausdorff_distance_filter.Execute(sitk.GetImageFromArray(y_true3d), sitk.GetImageFromArray(y_pred3d))
-            hdd3d = hausdorff_distance_filter.GetHausdorffDistance()
-            print("3D HDD for patient", test_patients[j], ":", hdd3d)
-            all3dhdds.append(hdd3d)
+            try:
+                hausdorff_distance_filter = sitk.HausdorffDistanceImageFilter()
+                hausdorff_distance_filter.Execute(sitk.GetImageFromArray(y_true3d), sitk.GetImageFromArray(y_pred3d))
+                hdd3d = hausdorff_distance_filter.GetHausdorffDistance()
+                print("3D HDD for patient", test_patients[j], ":", hdd3d)
+                all3dhdds.append(hdd3d)
+            except:
+                pass
     print("All 3d-HDDs: ", all3dhdds)
     print(folder_path)
     print(modeltype)