Technical Note: Collimator angle optimization for multiple brain metastases in dynamic conformal arc treatment planning

To reduce the exposed area by the multileaf collimator between lesions for single-isocenter dynamic conformal arc (DCA) therapy for stereotactic radiosurgery treatment of multiple brain metastases by optimizing the collimator angle orientation. In particular, this is achieved by the avoidance of collimator angles where multiple lesions are exposed by the same leaf pairs.

An algorithm that estimates the quality of an arc by considering the target projections onto the plane perpendicular to the central axis of the arc beam. A penalty proportional to the exposure of healthy tissue between metastases is assigned to each control point and each feasible collimator angle from a discretized set of angles. The algorithm can generate two outputs: the fixed optimal collimator angle over all the control points, or the optimal collimator angle trajectory through all the control points considering the rotation speed of the collimator. The first output is based on explicit enumeration of all collimator angles, and the second one generates the optimal trajectory using dynamic programming to find the globally optimal solution with respect to the objective function cost. The algorithm was validated on eight clinical cases having a different number of cranial metastases: two metastases (n = 1), three metastases (n = 5), four metastases (n = 1), and five metastases (n = 1). Plans with optimized fixed collimator angles and plans with optimized dynamic collimator trajectories were compared between each other.

When comparing optimal dynamic trajectories to fixed optimal collimator trajectories, the resulting plans demonstrated a total reduction of the exposed area between lesions over the entire beam configuration from 21.7% up to 71.3%; similarly, beam-wise reductions ranging from 5.83% to over 90% have been registered.

Collimator angle optimization has the potential to reduce the magnitude of the exposed area between lesions in an efficient way for non-isocentric treatments where multiple lesions are treated simultaneously. Dynamic trajectories are capable of limiting the island blocking problem more than optimal fixed trajectories by exploiting the extra degree of freedom of rotating the multileaf collimator. The algorithm can also lead to time saving during the treatment planning process.