Date: Friday, Nov. 16, 2018
Time: 1 - 1:50 pm
Location: D3 W122
Abstract: One of the challenges in radiation therapy treatment planning is to identify which planning objectives to use in the treatment optimization algorithm and how to balance the potential conflict between the objectives. Currently treatment planners typically rely on their own experience and trial and error to come up with final treatment planning settings such as planning objectives and associated priority parameters, which is often time-consuming and may lead to suboptimal solutions. We propose a novel optimization/machine learning model that learns from past treatment data and infers a personalized set of planning objectives that are most critical in treatment planning. Specifically, we propose an inverse-optimization-based approach with a cardinality constraint to determine such objectives. We apply this methodology to prostate radiation therapy treatment planning.
Biography: Dr. Taewoo Lee is an assistant professor of Industrial Engineering at the University of Houston. His research focuses on data-driven optimization and learning in healthcare applications including radiation therapy treatment planning, medical decision making in organ transplantation and disease screening, and public health policy-making. Dr. Lee’s work has appeared in journals such as Operations Research, Management Science, and Medical Physics, and is supported by the National Science Foundation. For his research, Dr. Lee has partnered with various health organizations such as Princess Margaret Cancer Centre, MD Anderson Cancer Center, Baylor College of Medicine, Ben Taub Hospital, and Harris Health System.