ABSTRACT
Inverse treatment planning is a patient-specific, time-consuming, and resource-intensive task, with plan quality heavily dependent on the institution and the planners' skill and experience. A significant effort has been dedicated to automate or facilitate the planning process using either prior knowledge or some advanced optimization techniques. In compressed sensing (CS) inverse planning is introduced; it could be considered an intermediate approach that incorporates the interplay between delivery and planning and enjoys the convexity of Beamlet-based optimization (BBO). CS inverse planning enjoys the convexity and simplicity of BBO on one hand, and it takes into account the interplay between deliverability and planning on the other hand, and it can be employed for either implementation of intensity-modulated radiotherapy or Station parameter optimized radiation therapy. In the traditional beamlet-based algorithms (BBO), beamlet intensity is optimized as an independent and continuous variable.
