Parameters

Parameter List

Below is a list of the parameters, along with a brief description, which can be input into the DicomReaderWriter

1.  description: string, Default (‘’), anonymizes new .nii arrays as Overall_Data_{description}_{iteration}.nii.gz in write_parallel()
   

2.  Contour_Names: Default (None) provides a list of contour names, such as [‘Liver’], for the ROIs you would like to read from RT structures present within the images.
    * Note if using arg_max: order is important in overlapping structures. Use ['Disease', 'Liver'], not ['Liver', 'Disease'] as the second will overwrite disease in the liver
   

3.  associations: Default (Empty Dictionary) allows for the creation of a dictionary to account for variations in ROI names ({variant_name: true_name}).
    For example, {‘Liver_BMA’: ‘Liver’} means that an ROI called ‘Liver_BMA’ should be treated as if it is called ‘Liver’.
   

4.  arg_max: Default (True) takes the arg_max of the mask created from the ROI contours.
    Note: Using this argument can cause problems if structures overlap. If contour names is [‘Liver’, ‘Disease’], arg_max results in ‘Liver’ overriding ‘Disease’ in the liver.
    In this case, writing [‘Disease’, ‘Liver’] is better, as this will make ‘Liver’ be equal to 2 and ‘Disease’ equal to 1
   

5.  verbose: Default (True), boolean for printing certain statements throughout the loading process, can be useful to keep track of everything going on
   

6.  create_new_RT: Default (True) saves memory by creating a new RT with only predicted structures; when set to False, predictions will be added to 
    the existing RT structure (see delete_previous_rois)
   

7.  template_dir: Default (None) allows the user to specify an RT template other than the one provided; If None, defaults to one provided
   

8.  delete_previous_rois: Default (True) saves memory by deleting the previous ROIs when writing a new RT structure;
    *Only valid if create_new_RT is False
   

9.  require_all_contours: Default (True), require all contours present when making nifti files?
   

10.  iteration: Default (0) sets the iteration number as part of the anonymization process, see description above
    

11.  get_dose_output: Default (False) is a Boolean that collects dose information from dose structures; and
    

12.  flip_axes: Default (False, False, False), a Tuple of size 3, allows the user to flip axes manually if data have been corrupted in the creation 
     of the DICOM files.
    

13.  series_instances_dictionary: A dictionary which houses series instance UIDs and paths for image and mask creation
    

14.  index: An index to reference the series_instance_dictionary, use set_index() to change
    

Functions list

Below is a list of functions provided by the DicomReaderWriter after creation

     Dicom_reader.walk_through_folders(Dicom_Path) This instructs the model to walk through all files, nested folders, and subfolders at a particular
     location. This step is what builds the series_instances_dictionary and should be your first step