RTK/Examples/ElektaReconstruction

Elekta Data

Elekta provides easy access to raw data. The data and projection images are stored in a single directory which is user configurable. The default location is D:\db. In this folder, there is a database in DBase format. Each table is contained in a .DBF file. RTK needs the IMAGE.DBF and FRAME.DBF tables.

Patient data are stored in invidual folders. By default, the name of each patient folder is patient_ID where ID is the patient ID. In these folders, one can access the planning CT in the CT_SET subfolder and the cone-beam projections in IMAGES/img_DICOM_UID subfolders where DICOM_UID is the DICOM UID of the acquisition. The projection images are .his files. The reconstructed images are the IMAGES/img_DICOM_UID/Reconstruction/*SCAN files.

Elekta Reconstruction

The first step before one can proceed with the reconstruction is to convert Elekta's database information into RTK geometry file using a command line tool. This can be carried out following these steps:

1. Download Elekta dataset, Elekta-data

2. Run the application to convert Elekta's geometry into RTKs (DICOM_UID is contained in the subfolder name of the his files):

 rtkelektasynergygeometry \
   --image_db IMAGE.DBF \
   --frame_db FRAME.DBF \
   --dicom_uid 1.3.46.423632.135428.1351013645.166 \
   -o elektaGeometry

Note that since XVI v5, the geometry is contained in a separate _Frames.xml file which can be used with

 rtkelektasynergygeometry \
   --xml _Frames.xml \
   -o elektaGeometry

An example of such a file is available in our test data, here.

3. Reconstruct elekta-data using RTK applications such as rtkfdk algorithm. In this case, we reconstruct just one axial slice (29.5) of the whole volume:

 rtkfdk \
   --lowmem \
   --geometry elektaGeometry \
   --path img_1.3.46.423632.135428.1351013645.166/ \
   --regexp '.*.his' \
   --output slice29.5.mha \
   --verbose \
   --spacing 0.25,0.25,0.25 \
   --dimension 1024,1,1024 \
   --origin -127.875,29.5,-127.875

4. Apply the FOV (field of view) filter, in order to mask out everything that is outside the FOV:

 rtkfieldofview \
   --geometry elektaGeometry \
   --path img_1.3.46.423632.135428.1351013645.166/ \
   --regexp '.*.his' \
   --reconstruction slice29.5.mha \
   --output slice29.5.mha \
   --verbose

5. Finally, you can visualize the result (e.g. with VV) and it should look like the image below: