S1820
Physics - Dose prediction/calculation, optimisation and applications for photon and electron planning
ESTRO 2026
between closely spaced metastases were reduced compared with static collimator angles (Figure 1). Similarly, in the VMAT plan, dynamic rotation improved dose conformity, decreased dose bridges, and yielded a lower objective function value (Figure 2).
Digital Poster 1145 Optimizing dynamic collimator rotation to improve dose conformity in photon arc therapy Paulina Treskog, Johan Sundström, Albin Fredriksson Research, RaySearch Laboratories AB, Stockholm, Sweden Purpose/Objective: Treatment planning for multiple brain metastases and irregularly shaped targets is challenging, as standard arc treatments often expose normal tissue between and around targets. This study evaluated whether dynamic collimator rotation, optimized to minimize island blocking, provides dosimetric advantages over static collimator angles in two scenarios: a multiple brain metastases case planned with dynamic conformal arc (DCA) and an irregularly shaped lung tumor case using volumetric modulated arc therapy (VMAT). Material/Methods: The collimator rotation trajectory was determined by minimizing the total aperture area of a conformal MLC throughout the arc, thereby reducing unnecessary open-field regions (so-called “whitespace”). A dynamic programming algorithm was used to identify the globally optimal collimator trajectory while respecting machine motion constraints and using geometric information from target contours. In addition to constraints such as gantry and collimator rotation speed, a restriction was applied on how often the collimator angle was allowed near 0˚ and 180˚ to reduce potential leakage buildup within the patient. For the multiple metastases case, a patient with five brain targets of varying sizes was selected. Three DCA plans were generated: one using an optimized static collimator angle, and two using dynamic collimator rotation under different constraints. Under a soft constraint, up to 10% of control points could have collimator angles within ±3˚ of 0˚ or 180˚, while the hard constraint excluded these intervals, reducing the allowed angular range by half.For the irregular lung case, two VMAT plans were created, one with a static collimator and one using dynamic collimator rotation, to assess whether the latter could improve dose conformity and reduce dose bridges. Dosimetric parameters were compared for each case: for the brain metastases, the gradient index (GI) and healthy brain dose were evaluated, while for the lung case, conformity indices and doses to surrounding tissue were assessed. Results: The dynamically rotating collimator substantially reduced whitespace between targets, with the soft constraint outperforming the hard constraint in both geometric and dosimetric metrics. For the five-target case, GI improved from 3.68 to 3.22, and dose bridges
Conclusion:
Made with FlippingBook - Share PDF online