LMU - Ludwig-Maximilians-Universität München

Advanced Monte Carlo and imaging methods 

Trainee: Liheng Tian
Supervisor: Katia Parodi

Compared to traditional Treatment Planning System (TPS) for Proton Radiotherapy, Monte Carlo (MC) techniques show great potential for improving the accuracy of dose calculation. On the other hand, in-vivo dose monitoring or proton range monitoring is needed to ensure the safety of treatment delivery considering the uncertainty of proton range. Detecting the Prompt Gamma (PG) emission following inelastic nuclear reaction of incident protons and their secondaries with tissues is currently one of the most promising solutions. However, due to the inhomogeneous density of tissue, the PG profile is not always corresponding to the dose profile, which would lead to misevaluation for proton range.

The focus of this project is to investigate dose and PG profiles, and quantify the PGdose correlation for each pencil beam. In this way, pencil beams with poor PG-dose correlation could be excluded in treatment planning ensuring the precision of dose delivery. First, proton treatment planning will be performed based on a research TPS environment based on the CERR framework coupled to the Geant4 Monte Carlo calculation engine, which is able to simulate both dose delivery and PG profiles on the same patient anatomy (from the planning CT or updated patient information). Then, the dose and PG profile will be investigated with different methods. Pencil beams with a large difference between the fall off position of dose and PG profile suggest poor dose-PG correlation and thus will receive a lower weight in a new optimization. This method will be optimized, tested for different anatomical locations and finally introduced in a research TPS environment.

Image courtesy of LMU