Implementation of Inverse Planning Optimization in Intensity Modulated Radiotherapy Using Genetic Algorithms

Background: During the therapy of tumor with IMRT, the main problem is how to select models of dose calculation and how to select optimization methods of inverse planning to deliver an accurate prescribe dose to the planning target volume while causing the least possible damage to surrounding organ at risk and normal tissue. Objective: To study the methods of beam parameters optimization to achieve automatic beam parameters selection in IMRT. Design: To construct dose calculation model based on pencil beams and optimize the beam parameters of inverse planning with genetic algorithms. Setting: School of Biomedical Engineering, Southern Medical University. JiNan Military Center for Disease Control and Prevention. Method: Fitness function based on dose constrain was constructed to calculate the fitness of individuals. Genetic algorithms were used to the optimization of beam weights, Iso-dose curves, three-dimensional envelope surface and dose-volume histogram are used to evaluate the outcome of treatment plan. Main outcome measure: To analyze the distribution of Iso-dose curves and dose-volume histogram, the comparability between tumor and three-dimensional envelope surface. Compare the difference between 3D-CRT and IMRT. Results: Beam weights optimization with genetic algorithms in IMRT can produce highly conformal dose distributions within a clinically acceptable computation time. Beam weights optimization with genetic algorithms in IMRT is valid and efficient, which provides a new method for inverse planning optimization in IMRT.