SEGMENTATION FOR RADIOTHERAPY PLANNING VIA M-REPS

University Of North Carolina Chapel Hill, Chapel Hill NC

Investigators

Linked publications & trials

Abstract

DESCRIPTION (provided by applicant): A large number of patients undergo 3D radiation therapy treatment planning (3D RTP) daily in the U.S. and other countries with modern radiotherapy facilities. Medical image segmentation is required for 3D RTP and is thus, a performed clinical task that guides critical treatment decisions. It is almost certain that image segmentation is performed more often as a clinical procedure in radiation oncology than for all other medical specialties combined. Most methods in routine clinical practice are user-guided and require highly skilled, well-trained users to produce results acceptable for 3D RTP. Current segmentation practice is thus an inherently inefficient and expensive task. Other flaws of current methods that tend toward suboptimal treatment planning include segmentation variabilities, the lack of clinically practical approaches that fully consider all three spatial dimensions, and the inability to deal with all sources of surface localization uncertainties. The development of new clinically useful 3D methods is compelled by needs to improve efficiency and contain costs, and to improve accuracy and reproducibility to steer user-guide planning decisions and inverse treatment planning algorithms consistently in the right direction. Moreover efficient, automatic segmentation can facilitate new technologies such as target volume localization, including methods to account for mechanical deformation or displacement of anatomical structures, for accurate patient positioning immediately prior to each treatment fraction, and for on- line analysis of treatment geometry during dose delivery. The overall aim of this project is to significantly improve the efficiency, reproducibility and accuracy of defining normal anatomical objects for 3D RTP via m-rep deformable models.