Optimization Methods for Volumetric Modulated Arc Therapy and Radiation Therapy under Uncertainty.

Radiation Therapy;Treatment Plan Optimization;Volumetric Modulated Arc Therapy;Optimization Under Uncertainty;Column Generation;Stochastic Programming;Industrial and Operations Engineering;Engineering;Industrial & Operations Engineering

Peng, FeiCohn, Amy Ellen ;

University of Michigan

关键词: Radiation Therapy;    Treatment Plan Optimization;    Volumetric Modulated Arc Therapy;    Optimization Under Uncertainty;    Column Generation;    Stochastic Programming;    Industrial and Operations Engineering;    Engineering;    Industrial & Operations Engineering;   

Others  :  https://deepblue.lib.umich.edu/bitstream/handle/2027.42/99813/feipeng_1.pdf?sequence=1&isAllowed=y

瑞士|英语

来源: The Illinois Digital Environment for Access to Learning and Scholarship

【 摘 要 】

Treatment plan optimization is crucial in the success of radiation therapy treatments. In order to deliver the prescribed radiation dose to the tumor without damaging the healthy organs, plans must be carefully designed so thatcollectively the dose delivered from different angles achieves the desired treatment outcome.The development of Volumetric Modulated Arc Therapy (VMAT) enables planners to explore additional benefits compared to traditional Intensity Modulated Radiation Therapy (IMRT). By rotating the gantry and attached source continuously, VMAT treatments can be delivered in a short period of time. While clinics can benefit from improved equipment utilization and less patient discomfort, treatment planning for VMAT is challenging because of the restrictions associated with the continuous gantry motion. We propose a new column generation based algorithm that explicitly considers the physical constraints, and constructs the treatment plan in an iterative process that searches for the maximum marginal improvement in each iteration. Implemented with GPU-based parallel computing, our algorithm is very efficient in generating high quality plans compared to idealized 177-angle IMRT plans.While treatments can benefit from VMAT in many ways, the capital expenditure in upgrading to a dedicated VMAT system is an important factor for clinics. Conventional IMRT machines can deliver VMAT treatments with constant rate of radiation output and gantry speed (VMATC). The absence of the ability to dynamically change the dose rate and gantry speed makes VMATC different in nature from VMAT. We propose two optimization frameworks for optimizing themachine parameters in the treatment, and recommend one configuration that consistently produces high quality plans compared to VMAT treatments.Finally, we consider uncertainties in radiation therapy treatments associated with errors in the daily setup process. We propose a stochastic programming based model that explicitly incorporates the range of uncertain outcomes in both the daily and cumulative dose distributions. While the problem is difficult to solve directly, we use a dynamic sampling procedure that can guarantee close to optimal solutions by establishing bounds on the optimal objective. Experiments with clinical cases show that the stochastic plans outperform the conventional approach, and reveal important information for planning adaptive treatments.