As part of the ENVISION (2010-2014) European project, in collaboration with the CAS-PHABIO team from IPNL (The Nuclear Physics Institute from Lyon), we are looking to contribute to the development of on-line imaging systems during irradiation procedures like on-line PET or Gamma Prompt.

We look to simulate these techniques on our biomedical model of the respiratory system. Thus, with a dynamic dosimeter module, our researches could integrate a treatment planning software. In order to do so, we look to simulate the interaction between the hadron beam and the human tissues at every instant. A first approach would be to use Monte Carlo simulations (using the Geant4 toolkit) which have proven to reproduce quite accurately the interactions between particles and matter. The disadvantage of this approach is that this type of simulations needs a big computational time. We look to optimize the algorithms in order to accelerate the simulation.

These kinds of simulations need a 3D mass density map of the respiratory system at every time step. The acquired medical data (usually from CT-scans) provides initial information on the organs’ position and density. While the patient is breathing, these organs suffer from deformations which can alter the position and the density of the different composing tissues. Taking into account the law of the conservation of mass, the biomedical model that we have developed can determine these alterations at every step of the simulation. These modifications are very important in order to accurately simulate the interactions between hadrons and organic tissues. The final objective is to accurately position the events that occur during these interactions (energy distribution, photon-pair production, gamma ray emission) in order to build a map of the distributed dose during treatment.