Project Summary / Abstract Advances in radiotherapy (RT) delivery seek to improve dose conformity to increase the therapeutic ratio. A key step in this process is the delineation of treatment targets and normal tissue avoidance regions. As treatment margins around delineated volumes shrink, if delineation variability is not accounted for, disparate volumes identified by different delineators, different imaging modalities, and different structure mapping algorithms may dominate the efficacy of both an individual patients' treatment and the collective results of clinical trials. While efficacy should be enhanced with improved delineation consistency, improved baseline imaging (including functional imaging) to more clearly see the tissues, and improved deformable registration algorithms to better align delineations from different images, benefits from such improvements will be jeopardized until methods to quantify and communicate the clinical effects of the patient-specific residual uncertainties in these processes exist. Uncertainty quantification enables mitigation as well as determination of when further uncertainty reductions provide no benefit. This project aims to (1) identify clinical situations in which delineation variability negatively impacts treatment efficacy in light of other inherent treatment uncertainties, (2) model delineation variability so that it may be estimated on a per-patient basis and communicated to the treatment team, and (3) develop treatment planning and delivery methods that mitigate the joint effect of delineation and other uncertainties in radiation therapy, thereby creating efficacious radiation therapy treatment plans.