Optimization of radiation therapy

Researchers: Michelle Böck, Lovisa Engberg and Anders Forsgren, in cooperation with Björn Hårdemark and Johan Löf (RaySearch Laboratories AB).

Sponsor: RaySearch Laboratories AB

This is an industrial graduate student project, carried out jointly between RaySearch Laboratories AB and KTH. The specific goal of the project is to develop tailored optimization models and tools that are suitable for optimization problems that arise within radiation therapy.

The current students are the third generation industrial graduate student. The first project was aimed at better understanding the
fundamental mathematical aspects of the basic optimization problems that arise within radiation therapy. The second generation projects were two. One was devoted to handing conflicting treatment goals by multiobjective optimization. The other one was focused on robust optimization methods for handling geometric uncertainties.

The current generation projects are also two. One subproject is directed toward methods for adaptive radiation therapy. This gives a demand for dynamic models. This gives rise to a number of new challenges, both in terms of modeling and in terms of treatment set up. The other subproject is focused on automated radiation therapy. The goal is to enable exploration of trade-offs between clinical goals in a more direct way than what is currently done, something that would further simplify and improve the treatment planning process.