Establishment of Workflow:

Cancer treatment with charges particle like protons and carbon ions has the potential of delivering highly conformal dose distributions at a significantly reduced dose level in the area surrounding the actual treatment target. This proves to be particular beneficial for indications located close to radiosensitive healthy structures and deep-seated tumors. The special characteristics of charged-particle energy deposition in matter is at the same time associated with a high sensitivity of the dose distribution to beam range uncertainties.  Although those uncertainties can be mitigated by an appropriate choice of safety margins in clinical practice, in-vivo range verification is demanded to fully exploit the physical superiority of charged particles.

At HIT, a state-of-the-art combined PET/CT scanner is used to measure the tissue activation in the treated volume, induced by the energetic beam particles. This measured PET image is compared to an activity prediction, derived under the assumption of correct treatment delivery [Bauer2013]. Discrepancies in range can thus be detected, evaluated and eventually trigger intervention to ensure safe beam delivery. 

More than 200 patients have received post-irradiation PET imaging for treatment verification at HIT since 2010. This patient cohort covers the wide spectrum of indications treated at HIT, including static lesions like tumors of the central nervous system [Nischwitz2015], head and neck diseases or lesions in the pelvis, as well as moving targets like carcinomas in the liver, pancreas or prostate.

Range comparison analysis a) at a certain position within the beam’s-eye-view (BEV) and b) for the whole BEV

Software package for PET-based beam range verification:

A fully automated workflow has been implemented in MeVisLab covering the wide chain of required functionality:

  • Efficient data management
  • Connection to local data base and central picture archiving systems (PACS)
  • Preparation of input data for Monte Carlo (MC)-based dose and activity forward calculations
  • Interface to and communication with the MC software running on an external cluster system
  • Mono-modal rigid image registration
  • Multi-modal visualization of dose distributions, CT images and simulated as well as measured PET images
  • Range comparison analysis for arbitrary modality images [Chen2015]