An algorithm for optimization of numerous modulated beam weights has been developed. This algorithm employs a penalty function theorem and a simulated annealing (SA) routine to model a large-scale constrained optimization problem incorporating dose and dose volume constraints in reflecting the goal of inverse treatment planning by sparing sufficient healthy tissues while delivering a necessary tumorcidal dose. The convergence property of the dose-volume SA algorithm is investigated for validation. Its performance is also evaluated by comparing the algorithm with a gradient technique minimizing the same dose-volume objective function that incorporates the target dose objectives and organ dose-volume constraints by the penalty functions. The comparison shows that the objective function exhibits a global valley in which multiple local minima with similar outcomes in terms of the function values, the dose-volume histograms, and the dose distributions exist. Thus, the gradient algorithm is preferred for this optimization approach due to its fast efficiency.
| Identifer | oai:union.ndltd.org:LACETR/oai:collectionscanada.gc.ca:QMM.33837 |
| Date | January 2001 |
| Creators | Sham, Edwin O. H. |
| Contributors | Hristov, Dimitre (advisor) |
| Publisher | McGill University |
| Source Sets | Library and Archives Canada ETDs Repository / Centre d'archives des thèses électroniques de Bibliothèque et Archives Canada |
| Language | English |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Format | application/pdf |
| Coverage | Master of Science (Department of Medical Radiation Physics.) |
| Rights | All items in eScholarship@McGill are protected by copyright with all rights reserved unless otherwise indicated. |
| Relation | alephsysno: 001863629, proquestno: MQ78955, Theses scanned by UMI/ProQuest. |
Page generated in 0.0019 seconds