Joseph O. Deasy, Ph.D.

DEPARTMENT OF Radiation Oncology
Keywords: optimization, outcomes, radiation therapy, treatment planning

Before any patient receives a radiation therapy cancer treatment, careful treatment planning is undertaken. This includes detailed imaging, radiation dose calculations, and graphical review of the treatments. To achieve tumor control, a dose distribution is planned which has a good chance of sterilizing all cancer cells without causing unacceptable normal tissue complications. Dr. Deasy's research focuses on improving the treatment planning process. More particularly, the focus is on two main areas: optimized selection of dose distribution characteristics and modeling the probability of treatment success and normal tissue complications as the dose distribution varies. Treatment planning optimization utilizes revolutionary radiation therapy machine developments to deliver very many small beams with varying shapes and intensities to create high-dose regions that conform closely to the target region, while avoiding nearby sensitive structures such as the spinal cord. To understand the probabilities of competing outcomes, dose distributions are collected as given to many patients, along with the observed rates of tumor control and normal tissue complication rates, and we create mathematical models which can, with some uncertainty, predict the probability of an outcome or tumor control for any dose distribution. Because this research involves treatment planning data at many different stages, the researchers have constructed a Computational Environment for Radiotherapy Research (CERR, pronounced 'sir'), which is an extensive graphical treatment planning software tool written in Matlab. CERR has tools for reviewing treatment planning CTs, anatomical structures, dose distributions, etc., and can extract treatment plans from commercial treatment planning systems using widely-available protocols. CERR provides a flexible and powerful tool for gathering and analyzing dose-volume models and also optimized treatment planning, along with sharing data and results with other investigators.