Image registration in intensity-modulated, image-guided and stereotactic body radiation therapy

Many recent advances in the technology of radiotherapy have greatly increased the amount of image data that must be rapidly processed. With the increasing use of multimodality imaging for target definition in treatment planning, and daily image guidance in treatment delivery, the importance of image registration emerges as key to improving the radiotherapy planning and delivery process at every step. Both clinicians and nonclinicians are affected in their work efficiency. Image registration can improve the correspondence of information in multimodality imaging, allowing more information to be obtained for tumor and normal tissue definition. Image registration at treatment delivery can improve the accuracy of therapy by taking greater advantage of images available prior to treatment. Technical advances have enhanced the accuracy and efficiency of registration through several approaches to automation, and by beginning to address the tissue deformation that occurs during the planning and therapy period. When using an automated registration technique, the user must understand the components of the registration process and the accuracy and limitations of the algorithm involved. This review presents the fundamental components of image registration, compares the benefits and limitations of different algorithms, demonstrates methods of visualizing registration results, and identifies methods to optimize registration for image-guided radiation therapy.