use myutil as the external package for some common medical image oper…

…ations.

seg_metrics/seg_metrics.py

@@ -9,57 +9,11 @@
 import PySimpleGUI as gui
 import matplotlib.pyplot as plt
 import glob
-# %%
-def load_itk(filename):
-    '''
-
-    :param filename: absolute file path
-    :return: ct, origin, spacing, all of them has coordinate (z,y,x) if filename exists. Otherwise, 3 empty list.
-    '''
-    #     print('start load data')
-    # Reads the image using SimpleITK
-    if (os.path.isfile(filename)):
-        itkimage = sitk.ReadImage(filename)
-
-    else:
-        print('nonfound:', filename)
-        return [], [], []
-
-    # Convert the image to a  numpy array first ands then shuffle the dimensions to get axis in the order z,y,x
-    ct_scan = sitk.GetArrayFromImage(itkimage)
-
-    # ct_scan[ct_scan>4] = 0 #filter trachea (label 5)
-    # Read the origin of the ct_scan, will be used to convert the coordinates from world to voxel and vice versa.
-    origin = np.array(list(reversed(itkimage.GetOrigin())))  # note: after reverseing, origin=(z,y,x)
-
-    # Read the spacing along each dimension
-    spacing = np.array(list(reversed(itkimage.GetSpacing())))  # note: after reverseing,  spacing =(z,y,x)
-    orientation = itkimage.GetDirection()
-    if (orientation[-1] == -1):
-        ct_scan = ct_scan[::-1]
-
-    return ct_scan, origin, spacing
-
-
-def get_gdth_pred_names(gdth_path, pred_path):
-    gdth_files = sorted(glob.glob(gdth_path + '/*' + '.nrrd'))
-    gdth_files.extend(sorted(glob.glob(gdth_path + '/*' + '.mhd')))
-    gdth_files.extend(sorted(glob.glob(gdth_path + '/*' + '.mha')))
+import sys
+from myutil.myutil import load_itk, get_gdth_pred_names, one_hot_encode_3d
 
-    pred_files = sorted(glob.glob(pred_path + '/*' + '.nrrd'))
-    pred_files.extend(sorted(glob.glob(pred_path + '/*' + '.mhd')))
-    pred_files.extend(sorted(glob.glob(pred_path + '/*' + '.mha')))
-
-    if len(gdth_files) == 0:
-        raise Exception('ground truth files  are None, Please check the directories', gdth_path)
-    if len(pred_files) == 0:
-        raise Exception(' predicted files are None, Please check the directories', pred_path)
-
-    if len(pred_files) < len(gdth_files):  # only predict several ct
-        gdth_files = gdth_files[:len(pred_files)]
-
-    return gdth_files, pred_files
 
+# %%
 def show_itk(itk, idx):
     ref_surface_array = sitk.GetArrayViewFromImage(itk)
     plt.figure()
@@ -257,20 +211,6 @@ def get_metrics_dict_all_labels(labels, gdth, pred, spacing, metrics_type):
 
     return metrics_dict
 
-def one_hot_encode_3D(patch, labels):
-
-    labels = np.array(labels)  # i.e. [0,4,5,6,7,8]
-    patches = []
-    for i, l in enumerate(labels):
-        a = np.where(patch != l, 0, 1)
-        patches.append(a)
-
-    patches = np.array(patches)
-    patches = np.rollaxis(patches, 0, len(patches.shape))  # from [6, 64, 128, 128] to [64, 128, 128, 6]?
-
-    return np.float64(patches)
-
-
 
 def write_metrics(labels, gdth_path, pred_path, csv_file, metrics=None):
     """