ABSTRACT
Automated planning can be used for the generation of plans with consistent high quality. Automation in planning may relieve or resolve some of the issues mentioned in the previous list, resulting in more consistent and higher plan quality with reduced operator dependence, and substantially reduced manual labour. The automated plan generation is iterative, evaluating after each iteration the defined criteria and automatically adding regions of interest and associated cost functions to iteratively improve the plan. Sophisticated, high quality dose response models are needed to fully exploit automated planning to the benefit of radiotherapy patients. The cost functions most commonly applied in both manual and automated planning are based on a series of dose-volume histogram (DVH) points. DVH points combined with quadratic cost functions favour a binary approach in generating and evaluating dose distributions, while actual dose effect relations are fairly continuous in nature.
