ReDeform is an efficiency-enhancement for the current procedure of radiation therapy planning on CT data within the ENT region.
In the present approach the therapy is planned on the CT scan performed at the beginning of the treatment and is kept over the full progress. This planning does not take into account possible changes of the patient (e.g. weight loss or reduction of the tumor region) due to the physical exertion and influence of the radiation exposure. This can negatively affect the success of therapy since it is possible that those changes can cause damages to healthy cells during treatment while cancerogenous cells might be missed.
Since the construction of a new treatment plan (manual delineation of relevant structures such as the tumor or several organs at risk) is quite time consuming it is crucial to determine if a replanning is required at all. In addition, a suggestion for the adaption of the treatment plan saves additional resources.
With the help of ReDeform it is possible to tackle these challenges by tracking changes between CT datasets of different dates.
Actually, the relevant datasets are registered and the deformation is presented such that for an expert it is possible to distinguish between real volume changes and simple changes in the patient position. The deformation itself is color coded for clear visibility and small as well as large changes can be viewed distinctively. In addition a statistical evaluation of the volume changes of the given structures is available. Furthermore, the originally planned dose can also be transformed to the new data representing the already received radiation in individual cells. This gives an additional cue to the expert for the judgment of the suitability of the current treatment plan.
The original delineated structures can be transformed to the current CT data, thus giving a suggestion for the changes necessary to be applied to the treatment plan.
Despite the fact that additional CT scans are necessary the impact of the additional radiation exposure is significantly smaller than the risk of the irradiation of the wrong type of tissue.
ReDeform was created by JOANNEUM RESEARCH Forschungsgesellschaft mbH by the use of the „Insight Segmentation and Registration Toolkit” (ITK ), the „Visualization Toolkit” (VTK ), the “Medical Imaging Interaction Toolkit” (MITK ), the “Grass roots DiCoM2” (GDCM2) and, Qt 4.
- Add a new DICOM message exchange server for data transmission and send data to that server.
- Load data via a PACS.
- Remove patient information from the data (anonymization).
- Use segmentation tools in order to automatically detect certain structures, manually mark areas and create a new struct object from a ROI for future reference.
- Register two images in order to generate new transformation matrix objects for doses, structs and studies.
a) Simultaneous visualization of both datasets.
b) Manual selection of volume of interest
c) Suggestion for replanning
d) Summation of already applied dose
e) Statistics of volume changes
- Visualize the registration result and use it for interpretation.
f) Illustration of rigid and non-rigid deformation field (different possibilities)
g) Compare new and old struct objects
h) View new dose object
i) Deformation Visualization
- Apply the deformation field in order to generate a transformation video.
- Manually select points in the CT data and store it in a struct.
- Conformity with ISO 13485 / ISO 62304
- Conformity with DICOM Standard
- Easy/Intuitive handling
- Integrated database for user management
- EU-Patent pending
- Project partner during product development and responsible for the clinical trial: Peter Winkler (contact person), Medical University Graz, Department of Therapeutic Radiology and Oncology
- Study with 20 patients
- Duration: January 2012 – July 2012
- Evaluation of
a) The benefit due to replanning (especially regarding the dosimetric data)
b) The usability/integration in clinical routines.
c) The functionalities as defined in the requirements
Start & End of Project
09/2009 – 12/2011
Due to different needs, experts bring in new ideas for several extensions of the current functionality.
In this context cooperation discussions with interested companies in France and Sweden have already been set up.