A high quality imaging database for adaptive radiation therapy

At the university hospital of the Ludwig Maximilians University of MunichKlinikum der Universität München (LMU), campus Grosshadern, a comprehensive imaging database of head and neck cancer patients undergoing state-of-the-art image-guided intensity modulated radiation therapy has been set up. The anonymized high quality adaptive radiation therapy database contains weekly cone-beam computed tomography images (CBCT), as well as conventional computed tomography images acquired mid treatment (re-planning CT, rpCT), in addition to the standard treatment planning imaging data (pCT).

 

World's first validated proton therapy dose calculations on cone-beam computed tomography images

In the course of the SPARTA project, researchers at the Chair of Medical Physics of the Ludwig Maximilians University of Munich have developed and validated algorithms allowing the calculation of proton therapy dose distributions on weekly cone-beam computed tomography (CBCT) images of the patient acquired at the treatment couch. In a series of articles relying on the Klinikum der Universität München (LMU) high quality imaging database for adaptive radiation therapy, the research team was the first to show that adequately corrected CBCT images can serve as basis for adaptive radiation therapy when using protons [Landry et al. Phys. Med. Biol. (2015)Landry et al. Med. Phys. (2015), Kurz et al. Acta Oncol. (2015)]. The novelty and importance of the research was acknowledged by the Medical Physics journal which chose the story for its cover page.

Dose-guided patient positioning using multi-criteria optimization (MCO)

The availability of up-to-date intensity-corrected CBCT images of the patient in treatment position enables daily and in future possibly on-line treatment optimization. This can, e.g.,  be achieved by treatment plan adaptation, which might, however, be time-consuming and require further quality assurance. In close collaboration with ITWM Kaiserslautern, we are currently evaluating an alternative approach, using MCO-based dose-guided patient positioning. The aim is not to position the patient on the basis of bony anatomy or markers, but on the basis of an optimal dose distribution, applying the original treatment plan. The tool developed at ITWM Kaiserslautern enables interactive navigation through different patient positions, taking only Pareto-optimal solution into account, and offers on-line visualization of the corresponding dose distribution and clinically relevant DVH parameter. The planner can thus interactively choose a patient position that optimally fulfills the given clinical goals. The images below shows the effect of a 3mm axial shift (right figure) with respect to the initial bony anatomy based alignment (left figure) for an exemplary head and neck tumor patient treated with intensity-modulated radiotherapy (IMRT): In the replanning CT based dose-volume histograms (DVH, dashed lines), dose to the left parotid gland is reduced to a level close to the initial planning situation (solid lines) at the expense of only a slightly decreased coverage of the low dose (LD) planning target volume (PTV).