fixed bug in set bad sections and added example

Author: PolarBean

DeepSlice/coord_post_processing/angle_methods.py

@@ -5,7 +5,7 @@
 
 def calculate_brain_center_coordinate(section, atlas_shape, axis):
     """
-    Calculates the coordinate closest to the middle of the section as defined by the two 
+    Calculates the coordinate closest to the middle of the section as defined by the two
     dimensions not orthogonal to the dimension along which the series is being aligned.
     for example, if the series is being aligned coronally then this will return the midpoint
     in X and Z as coronal series are spaced along the Y coordinate in the CCF.
@@ -39,7 +39,7 @@ def calculate_brain_center_coordinate(section, atlas_shape, axis):
 def calculate_angles(df):
     """
     Calculates the Mediolateral and Dorsoventral angles for a series of predictions
-    
+
     :param df: The dataframe containing the predictions
     :type df: pandas.DataFrame
     :return: a list of calculated ML and DV angles
@@ -107,7 +107,7 @@ def propagate_angles(df, method, species):
     # get the angles for each section in the dataset
     temp = df.copy()
     if "bad_section" in temp:
-        temp = temp[~temp["bad_section"]]
+        temp = temp[~temp["bad_section"].astype(bool)]
     DV_angle_list, ML_angle_list = calculate_angles(temp)
 
 

DeepSlice/main.py

@@ -19,7 +19,7 @@ def __init__(self, species):
         xception_weights =   metadata_loader.get_data_path(self.config["weight_file_paths"]["xception_imagenet"], self.metadata_path)
         weights =    metadata_loader.get_data_path(self.config["weight_file_paths"][self.species]["primary"], self.metadata_path)
         self.model = neural_network.initialise_network(xception_weights, weights, self.species)
-        
+
 
 
     def predict(
@@ -62,7 +62,7 @@ def predict(
                 image_directory
             )
         primary_weights = metadata_loader.get_data_path(self.config["weight_file_paths"][self.species]["primary"], self.metadata_path)
- 
+
         secondary_weights = metadata_loader.get_data_path(self.config["weight_file_paths"][self.species]["secondary"], self.metadata_path)
 
         if secondary_weights == "None":
@@ -105,7 +105,7 @@ def set_bad_sections(self, bad_sections: list, auto = False):
         """
         sets the bad sections for a given brain. Must be run after predict()
 
-        :param bad_sections: A list of bad sections to ignore when calculating angles and spacing
+        :param bad_sections: A list of bad sections to ignore when calculating angles and spacing, the list just needs to contain a unique string for each section, for instance the section number. do not use a string which appears in multiple filenames
         :type bad_sections: list
         """
         self.predictions = spacing_and_indexing.set_bad_sections_util(
@@ -154,7 +154,7 @@ def propagate_angles(self, method="weighted_mean"):
             self.predictions = angle_methods.propagate_angles(
                 self.predictions, method, self.species
             )
-            
+
 
     def load_QUINT(self, filename):
         """
@@ -178,9 +178,9 @@ def load_QUINT(self, filename):
         self.predictions = predictions
         xception_weights =   metadata_loader.get_data_path(self.config["weight_file_paths"]["xception_imagenet"], self.metadata_path)
         weights =    metadata_loader.get_data_path(self.config["weight_file_paths"][self.species]["primary"], self.metadata_path)
-        self.model = neural_network.initialise_network(xception_weights, weights, self.species) 
+        self.model = neural_network.initialise_network(xception_weights, weights, self.species)
+
 
-        
 
     def save_predictions(self, filename):
         """

README.md

@@ -94,6 +94,10 @@ folderpath = 'examples/example_brain/GLTa/'
 #try with and without ensemble to find the model which best works for you
 #if you have section numbers included in the filename as _sXXX specify this :)
 Model.predict(folderpath, ensemble=True, section_numbers=True)    
+#This is an optional stage if you have damaged sections, or hemibrains they may negatively effect the propagation for the entire dataset
+#simply set the bad sections here using a string which is unique to those each section you would like to label as bad. DeepSlice will
+#not include it in the propagation and instead it will infer its position based on neighbouring sections.
+Model.set_bad_sections(bad_sections=["_s094", "s199"])
 #If you would like to normalise the angles (you should)
 Model.propagate_angles()                     
 #To reorder your sections according to the section numbers 

ds.py

@@ -0,0 +1,22 @@
+from DeepSlice import DSModel
+species = 'mouse' #available species are 'mouse' and 'rat'
+Model = DSModel(species)
+folderpath = 'examples/example_brain/GLTa/'
+#here you run the model on your folder
+#try with and without ensemble to find the model which best works for you
+#if you have section numbers included in the filename as _sXXX specify this :)
+Model.predict(folderpath, ensemble=True, section_numbers=True)
+#This is an optional stage if you have damaged sections, or hemibrains they may negatively effect the propagation for the entire dataset
+#simply set the bad sections here using a string which is unique to those each section you would like to label as bad. DeepSlice will
+#not include it in the propagation and instead it will infer its position based on neighbouring sections.
+Model.set_bad_sections(bad_sections=["_s094", "s199"])
+#If you would like to normalise the angles (you should)
+Model.propagate_angles()
+#To reorder your sections according to the section numbers
+Model.enforce_index_order()
+#alternatively if you know the precise spacing (ie; 1, 2, 4, indicates that section 3 has been left out of the series) Then you can use
+#Furthermore if you know the exact section thickness in microns this can be included instead of None
+#if your sections are numbered rostral to caudal you will need to specify a negative section_thickness
+Model.enforce_index_spacing(section_thickness = None)
+#now we save which will produce a json file which can be placed in the same directory as your images and then opened with QuickNII.
+Model.save_predictions(folderpath + 'MyResults')