Friday September 24 at 11.00 Lindstedtsvägen 25, room 3721

Albin Fredriksson 
RaySearch Laboratories and
Division of Optimization and Systems Theory, Department of Mathematics, KTH.

Minimax optimization for handling range and setup uncertainties in proton therapy

Intensity modulated proton therapy (IMPT) for cancer treatment allows for dose distributions that conform closely to the target volume. Resulting treatment plans are however sensitive to errors and conventional margins around the regions of interest are often insufficient to ensure robustness. To obtain robust plans, we perform minimax optimization, taking range and setup uncertainties into account. This method is related to stochastic programming by the more general minimax stochastic programming formulation. We apply the minimax method to a lung case, a paraspinal case, and a prostate case. Resulting plans are compared to plans for which conventional methods are used to handle the uncertainties. For all cases, the minimax method provides more robust target coverage and better sparing of healthy tissues than the other methods. By incorporating information about the uncertainties into the optimization, the modality can be better utilized and some of the problems associated with margins in IMPT can be eliminated.


Calendar of seminars Last update: September 16, 2010.