Abstract:

Radiotherapy is an important component of treatment for about 40% of cancer patients. In current clinical practice, expert judgment plays an important role in determining the parameters of the increasingly complex treatment plans needed to deliver emerging therapies such as IMRT (Intensity Modulated Radiation Therapy), IMAT (Intensity Modulated Arc Therapy), and Tomotherapy. For example, commercial radiation treatment planning systems generally require a set of beam angles as part of the user input to the planning process. While expert judgment may produce good solutions to these complex problems, optimization techniques implemented on advanced computing platforms offer the ability to automate the evaluation of huge numbers of potential beam and couch angle sets. By thus automating the determination of good angle sets and presenting the results to the treatment planner along with a comparison of the relative effectiveness of simpler plans such as equally-spaced beams, important insights may be obtained into effective treatment planning alternatives. This presentation will consider several areas of radiation treatment planning (RTP) in which advanced computing on clusters, supercomputers, or the “grid” can significantly advance the state-of-the-art and therefore the quality of radiation treatments delivered to patients. These RTP research areas include data generation (which requires the determination of radiation doses delivered from thousands of positions into thousands of voxels), beam/couch angle selection, dose optimization, and intensity map segmentation (to determine the final set of radiation delivery parameters).

Biosketch:

Robert R. Meyer received a B.S. in Mathematics from Caltech and an M.S. and a Ph.D. in Computer Sciences from the University of Wisconsin-Madison. After several years as a researcher with Shell Oil, he moved to academia and is currently Professor of Computer Sciences at the University of Wisconsin-Madison, where he has been on the faculty since 1973. He has co-edited six volumes of optimization conference proceedings and written more than 80 papers dealing with nonlinear network optimization, parallel algorithms for large-scale optimization, genetic algorithms, theory and applications of discrete optimization, machine learning, and radiation treatment planning in brachytherapy and teletherapy. His current research, funded by the National Science Foundation, focuses on optimization methods, advanced computing based on the Condor system, and cyberinfrastructure utilization for radiation treatment planning.