d

RobinNet/13unet_mrt_threshgt.py

@@ -179,7 +179,7 @@ def main():
             #number_patients = number_patients * 5
             img_path = "data/npy_thresh/"
 
-    specificmodels = [6, 7]
+    specificmodels = [2]
     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))
+        weights = init_weights(manual_seed=np.sum(self.test_patients)*10)
 
         filter_multiplier = int(self.newSize[0] / 128)
 

RobinNet/rerunModels.py

@@ -24,6 +24,7 @@ def main():
     #modeltype = "Class_"
     modeltype = "Cluster_class_"
     #modeltype = "Unet_"
+    all3dhdds = []
 
     ##################### U-Net #####################
     for fold in models:
@@ -179,15 +180,16 @@ def main():
             test_loss_hdd = []
             test_loss_hdd2 = []
             y_pred3d = []
+            y_true3d = []
             for features in test_dataset:
                 image, y_true = features
                 y_true = onehotencode(y_true)
                 y_pred = Unet(image, weights, filter_multiplier, training=False)
                 
-                if y_pred3d == []:
+                '''if y_pred3d == []:
                         y_pred3d = y_pred[:,:,:,0].numpy()
                 else:
-                    y_pred3d =  np.append(y_pred[:,:,:,0].numpy(), y_pred3d, axis=0)
+                    y_pred3d =  np.append(y_pred[:,:,:,0].numpy(), y_pred3d, axis=0)'''
 
                 loss = dice_loss(y_pred, y_true)
                 loss = tf.make_ndarray(tf.make_tensor_proto(loss))
@@ -205,6 +207,15 @@ def main():
                     hausdorff_distance_filter2 = sitk.HausdorffDistanceImageFilter()
                     hausdorff_distance_filter2.Execute(sitk.GetImageFromArray(pred_np), sitk.GetImageFromArray(y_true_np))
                     test_loss_hdd2.append(hausdorff_distance_filter2.GetHausdorffDistance())
+                    
+                    if y_pred3d ==[]:
+                        y_pred3d = pred_np
+                        y_true3d = y_true_np
+                    else:
+                        y_pred3d = np.append(pred_np, y_pred3d, axis=0)
+                        y_true3d = np.append(y_true_np, y_true3d, axis=0)
+                        
+                    
                 except:
                     pass
 
@@ -223,18 +234,23 @@ def main():
 
             print(tf.shape(ytrue))
             y_true3d = np.squeeze(ytrue[:,:,:,0].numpy() > 0.5)
-            y_true3d = y_true.astype(np.float_)
+            y_true3d = np.float_(y_true)
             print(np.shape(y_true3d))
             print(type(y_true3d))
             print(type(y_pred3d))
             pred3d = np.squeeze(y_pred3d[:,:,:].numpy() > 0.5)
-            pred3d = pred3d.astype(np.float_)
+            pred3d = np.float(pred3d)
             print(np.shape(pred3d))
-            print(type(pred3d))
+            print(type(pred3d))'''
             hausdorff_distance_filter = sitk.HausdorffDistanceImageFilter()
-            hausdorff_distance_filter.Execute(sitk.GetImageFromArray(y_true3d), sitk.GetImageFromArray(pred3d))
+            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)'''
+            print("3D HDD for patient", test_patients[j], ":", hdd3d)
+            all3dhdds.append(hdd3d)
+    print("All 3d-HDDs: ", all3dhdds)
+    print(folder_path)
+    print(modeltype)
+           
       #####
 
 if __name__ == "__main__":

RobinNet/saveImgs_run_trained_model.py

@@ -274,9 +274,8 @@ def main():
                         matplotlib.image.imsave(img_savepath + "P" + str(TP_num).zfill(2) + "_" + str(counter).zfill(2) + "_unetpred"+ str(loss)[0:6] +".png",
                                             y_pred[0, :, :, 0], cmap=plt.cm.bone)
 
-            if npys3d:
+            #if npys3d:
                 #np.save(img_savepath + "P" + str(TP_num).zfill(2) + "_pred3D", y_pred3d)
-                print("ypred shape: ", np.shape(y_pred3d))
             #plt.show()
             #print(test_loss)
             print("TestLoss Mean for P", test_patients[j], ": ", np.mean(test_loss))

RobinNet/threshunetpred_gts.py

@@ -180,7 +180,7 @@ def main():
             if npys3d:
                 filename = "P" + str(TP_num).zfill(2)
                 y_pred3d = cut_mr_ctunetpred(filename, y_pred3d)
-                ythresh3d = np.flip(cut_mr_ctunetpred(filename, GT_test))
+                ythresh3d = cut_mr_ctunetpred(filename, np.flip(GT_test))
                 #np.save(img_savepath + "P" + str(TP_num).zfill(2) + "_pred3D", y_pred3d)
                 print(np.shape(ythresh3d[:,:,:,0]))
                 np.save("data/npy_thresh/" + "P" + str(TP_num).zfill(2) + "_mr_ctunetpred.gipl", ythresh3d[:,:,:,0])

RobinNet/tryOut.py

@@ -9,16 +9,17 @@ import matplotlib.pyplot as plt
 from scipy import ndimage
 from datetime import datetime
 
+pat = "06"
 
-inputimg = np.load("data/npy_thresh/P01_mr_T1.gipl.npy")
-gt = np.load("data/npy_thresh/P01_mr_ctunetpred.gipl.npy")
-ytrue = np.load("data/npy_thresh/P01_segmr.gipl.npy")
+inputimg = np.load("data/npy_thresh/P"+pat+"_mr_T1.gipl.npy")
+gt = np.load("data/npy_thresh/P"+pat+"_mr_ctunetpred.gipl.npy")
+ytrue = np.load("data/npy_thresh/P"+pat+"_segmr.gipl.npy")
 
 print(np.shape(inputimg))
 print(np.shape(gt))
 print(np.shape(ytrue))
 
-for i in range(25):
+for i in range(np.shape(inputimg)[0]):
     fig = plt.figure()
     fig.add_subplot(2, 3, 1)
     plt.imshow(inputimg[i, :, :], cmap=plt.cm.bone)

RobinNet/utils/other_functions.py

@@ -368,4 +368,4 @@ def slicenumber_to_class(self, x):
         if (x[1] >= 27 and x[1] < 29): sliceclass = 10
     else: sliceclass = 0
     #print('Sliceclass:', sliceclass)
-    return sliceclass
\ No newline at end of file
+    return sliceclass