Return to search

Development of techniques for optimization and verification of radiation treatments

Algorithms for optimization and verification of radiation treatments have been developed. The first one, an active set algorithm for inverse treatment planning employs a conjugate gradient routine for subspace minimization in order to achieve a higher rate of convergence than the widely used constrained steepest descent method at the expense of a negligeable amount of overhead calculations and storage. The active set algorithm is found to be superior to the constrained steepest descent in terms of both its convergence properties and the residual value of the cost functions at termination. The active set approach can significantly accelerate the process of inverse treatment planning by decreasing the number of time consuming dose calculations. / The second algorithm employs a continuous penalty function method to solve approximately a large-scale constrained minimization problem which reflects the goal of sparing healthy tissues as much as possible while delivering the necessary tumorcidal dose. The performance of the continuous penalty function method is optimized by a numerical investigation of a few integration schemes and a pair of weighting functions which influence the performance of the method. Clinical examples are presented that illustrate possible applications of the techniques in the context of multi-objective optimization. / An image correlation based algorithm for automatic registration of pairs of portal images has also been developed. Accounting for both in-plane translations and rotations, the algorithm uses fast-Fourier-transforms and a sequential approach to speed up the registration without degrading the accuracy of the match. The technique has also been applied to the automatic registration of portal images to digitally reconstructed radiographs (DRRs) which have been modified to resemble megavoltage images. The results indicate the feasibility of this approach as a tool for treatment setup verification.

Identiferoai:union.ndltd.org:LACETR/oai:collectionscanada.gc.ca:QMM.34975
Date January 1998
CreatorsHristov, Dimitre H.
ContributorsFallone, B. G. (advisor)
PublisherMcGill University
Source SetsLibrary and Archives Canada ETDs Repository / Centre d'archives des thèses électroniques de Bibliothèque et Archives Canada
LanguageEnglish
Detected LanguageEnglish
TypeElectronic Thesis or Dissertation
Formatapplication/pdf
CoverageDoctor of Philosophy (Department of Physics.)
RightsAll items in eScholarship@McGill are protected by copyright with all rights reserved unless otherwise indicated.
Relationalephsysno: 001610465, proquestno: NQ44454, Theses scanned by UMI/ProQuest.

Page generated in 0.0115 seconds