Premium
SU‐GG‐T‐160: Multi‐Objective Optimization of Beam Orientation and Apertures in Intensity Modulated Radiation Therapy (IMRT)
Author(s) -
Cao R,
Pei X,
Wu Y
Publication year - 2010
Publication title -
medical physics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.473
H-Index - 180
eISSN - 2473-4209
pISSN - 0094-2405
DOI - 10.1118/1.3468550
Subject(s) - multi objective optimization , orientation (vector space) , sorting , collimator , mathematical optimization , dosimetry , optimization problem , set (abstract data type) , pareto principle , beam (structure) , computer science , radiation treatment planning , medical imaging , volume (thermodynamics) , mathematics , algorithm , nuclear medicine , radiation therapy , optics , physics , artificial intelligence , medicine , surgery , geometry , quantum mechanics , programming language
Purpose: The intensity modulated radiation therapy requires to determinate the beam orientation and its apertures (the leave positions of the multi‐leaf collimator (MLC)). Inverse planning optimization is a multi‐objective optimization problem whose solution is known as Pareto solution set. According to the multi‐objective character of inverse planning in IMRT, the multi‐objective optimization of beam orientation and its apertures based on Pareto solution set was studied. Method and Materials: The clinical requirements for a treatment plan were transformed into a multi‐objective optimization problem with multiple constraints, in which the parameters are beam orientation and its apertures. And then the fast and elitist multi‐objective Non‐dominated Sorting Genetic Algorithm (NSGA‐II) was introduced to optimize the problem. For each region of interest ‐ target volume or organ at risk, this study used a “physical” objective function in which the dose delivered to each region in the patient's body was compared directly with a dose distribution prescribed by the physician, or a dose‐volume (DV) constraint which typically require that no more than/no less than a specified fraction of volume of a given region receives a dose of higher/Lower than a certain specified level. The aim of NSGA‐II based optimization algorithm was to provide a representative set of non‐dominated solutions for problems where many conflicting objectives and many constraints need to be considered simultaneously instead of a single solution. Results: A clinical example was tested with this method. The results showed that a set of non‐dominated solutions that were obtained distributed uniformly, and the corresponding dose distribution of each solution not only approached to the expected dose distribution but also met the dose‐volume constraints. Conclusions: It was indicated that the clinical requirements were better satisfied by the method and planner could select the optimal treatment plan from the non‐dominated solution set.