CORE-BASED PORTAL IMAGE REGISTRATION FOR AUTOMATIC RADIOTHERAPY TREATMENT VERIFICATION

Purpose: Portal imaging is the most important quality assurance proced ure for monitoring the reproducibility of setup geometry in radiation therapy. The role of portal imaging has become even more critical in r ecent years due to the migration of three-dimensional (3D) treatment p lanning technology, including high precision conformal therapy, from t he research setting to routine clinical practice. Unfortunately, tradi tional methods for acquiring and interpreting portal images suffer fro m a number of deficiencies that contribute to the well-documented obse rvation that many setup errors go undetected, and some persist for a c linically significant portion of the prescribed dose. Significant impr ovements in both accuracy and efficiency of detecting setup errors can , in principle, be achieved by using automatic image registration for on-line screening of images obtained from electronic portal imaging de vices (EPIDs). Methods and Materials: This article presents recent dev elopments in a method called core-based image analysis that shows grea t promise for achieving the desired improvements in error detection. C ore-based image analysis is a fundamental computer vision method that is capable of exploiting the full power of EPIDs by providing for on-l ine detection of setup errors via automatic registration of user-selec ted anatomical structures. We describe a robust method for automatic p ortal image registration based on core analysis and demonstrate an app roach for assessing both accuracy and precision of registration method s using realistic, digitally reconstructed portal radiographs (DRPRs) where truth is known. Results: Automatic core-based analysis of a set of 20 DRPRs containing known, random field positioning errors was perf ormed for a patient undergoing treatment for prostate cancer. ln all c ases,the translation was within 1 mm of the actual translation with me an absolute errors of 0.3 mm and standard deviations of 0.3 mm. In all cases, the reported rotation was within 0.6 degrees of the actual rot ation with a mean absolute error of 0.18 degrees and a standard deviat ion of 0.23 degrees. Conclusion: Our results, using digitally reconstr ucted portal radiographs that closely resemble clinical portal images, suggest that automatic core-based registration is suitable as an on-l ine screening tool for detecting and quantifying patient setup errors.