A significant optimization problem regarding radiation therapy for cancer treatment is to deliver, at least the prescribed dose to the tumor while avoiding excessive exposure of healthy organs. Over the years, many researchers have presented algorithms to automate angle selection for adequate dose distribution. This paper presents a novel algorithm that seeks the ideal balance of angles to distribute radiation doses while respecting medical prescriptions inherent to the treatment by using Ordered Weighted Average (OWA) operator. The hybrid algorithm employs a model utilizing the OWA operator as a preference criterion for selecting the optimal solution. Hybridization entails integrating an exact method with a greedy heuristic. The algorithm produces plans with a low number of angles and requires low computational effort. Furthermore, it achieved clinically feasible results in all investigated instances.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Ordered Weighted Averaging for the Beam Angle Optimization and Intensity Problems

  • Sidemar F. Cezario,
  • Thiago S. Marques,
  • Sílvia M. D. M. Maia,
  • Marco César Goldbarg,
  • Elizabeth Ferreira Gouvêa Goldbarg

摘要

A significant optimization problem regarding radiation therapy for cancer treatment is to deliver, at least the prescribed dose to the tumor while avoiding excessive exposure of healthy organs. Over the years, many researchers have presented algorithms to automate angle selection for adequate dose distribution. This paper presents a novel algorithm that seeks the ideal balance of angles to distribute radiation doses while respecting medical prescriptions inherent to the treatment by using Ordered Weighted Average (OWA) operator. The hybrid algorithm employs a model utilizing the OWA operator as a preference criterion for selecting the optimal solution. Hybridization entails integrating an exact method with a greedy heuristic. The algorithm produces plans with a low number of angles and requires low computational effort. Furthermore, it achieved clinically feasible results in all investigated instances.