Automated WBRT Treatment Planning via Deep Learning Auto-Contouring and Customizable Landmark-Based Field Aperture Design
In this work, we developed and evaluated a novel pipeline consisting of two landmark-based field aperture generation approaches for WBRT treatment planning; they are fully automated and customizable. The automation pipeline is beneficial for both clinicians and patients, where we can reduce clinician workload and reduce treatment planning time. The customizability of the field aperture design addresses different clinical requirements and allows the personalized design to become feasible. The performance results regarding quantitative and qualitative evaluations demonstrated that our plans were comparable with the original clinical plans. This technique has been deployed as part of a fully automated treatment planning tool for whole-brain cancer and could be translated to other treatment sites in the future.