Speaker
Description
The Heidelberg Ion Beam Therapy Centre (HIT) plans to implement multi-energy extraction, which allows faster cancer treatment by re-accelerating unextracted beam in the synchrotron. Thus an increase of the number of particles injected into the synchrotron would directly benefit patient treatment time and consequently also the number of patients treated. However, optimization of beam line segments by operators is often time consuming and it is difficult to align with a global optimum for the entire beamline. To address this problem we investigate the implementation of a machine learning algorithm based on Bayesian optimization using gaussian process models for online accelerator tuning. We specifically focus on sample-efficient transmission optimization for larger beamlines in medical accelerators and the effects of input space and objective selection.
| In which format do you inted to submit your paper? | LaTeX |
|---|