We construct a multi-criteria optimization program that automatically generates the multi-dimensional tradeoff curve of IMRT objectives. Linear convex approximations of non-convex IMRT objectives are used and the Pareto-optimality of approximations in nominal objective space is assessed. Tradeoff solutions are efficiently spread out on the tradeoff curve and an approach for navigating the multi-dimensional curve is shown. The combined effect of column and row generation in speeding up the linear programs is also examined. Our approach is applied to clinical cases and the results examined. An open-source package of our programs, LIRA, is provided.
Identifer | oai:union.ndltd.org:UMASS/oai:scholarworks.umass.edu:dissertations-4734 |
Date | 01 January 2007 |
Creators | Halabi, Tarek |
Publisher | ScholarWorks@UMass Amherst |
Source Sets | University of Massachusetts, Amherst |
Language | English |
Detected Language | English |
Type | text |
Source | Doctoral Dissertations Available from Proquest |
Page generated in 0.0018 seconds