TY - GEN
T1 - Non-rigid registration techniques and applications in high-precision radiotherapy
AU - Hoogeman, M. S.
AU - Vásquez Osorio, E.M.
AU - Bondar, M.
AU - Han, X.
AU - Levendag, P. C.
AU - Heijmen, B. J.M.
PY - 2009
Y1 - 2009
N2 - Radiotherapy is a major treatment modality in the management of cancer. In the last decade, advances in planning and dose delivery, and in image guidance and tracking have substantially widened the therapeutic window. However, changes in patient anatomy during the treatment limit the productive use of these new technologies. To further improve radiation treatments, anatomical changes need to be modeled and accounted for. Non-rigid registration can be used for this purpose. In this paper we describe a novel flexible framework for non-rigid registration and present two applications of the framework. The first application quantifies anatomy changes in head and neck patients due to irradiation. The changes were quantified on a local scale and were correlated to the dose received. The second application uses non-rigid registration to obtain tissue correspondence between cervix-uterus surfaces obtained for different bladder volumes. This correspondence was used to build prediction models to predict the shape and position of the cervix-uterus for a given bladder volume. The latter can be used to assist in a daily online selection of the best treatment plan that is based on for example an ultra sound bladder volume measurement. Finally, we discuss the role of non-rigid registration to automate contouring. Atlas-based auto-segmentation is a promising application of non-rigid registration which dramatically reduces contouring time. Two approaches are evaluated that use atlas-based autosegmentation: a single subject approach and a multiple-subject approach. It was shown that the multiple-atlas approach significantly improved the automatically generated contours for head and neck patients. Overall, we conclude that non-rigid registration techniques are playing an increasingly important in modern radiotherapy techniques. However, it is important to realize that before non-rigid registration can be used safely and effectively the registration methods should be validated thoroughly.
AB - Radiotherapy is a major treatment modality in the management of cancer. In the last decade, advances in planning and dose delivery, and in image guidance and tracking have substantially widened the therapeutic window. However, changes in patient anatomy during the treatment limit the productive use of these new technologies. To further improve radiation treatments, anatomical changes need to be modeled and accounted for. Non-rigid registration can be used for this purpose. In this paper we describe a novel flexible framework for non-rigid registration and present two applications of the framework. The first application quantifies anatomy changes in head and neck patients due to irradiation. The changes were quantified on a local scale and were correlated to the dose received. The second application uses non-rigid registration to obtain tissue correspondence between cervix-uterus surfaces obtained for different bladder volumes. This correspondence was used to build prediction models to predict the shape and position of the cervix-uterus for a given bladder volume. The latter can be used to assist in a daily online selection of the best treatment plan that is based on for example an ultra sound bladder volume measurement. Finally, we discuss the role of non-rigid registration to automate contouring. Atlas-based auto-segmentation is a promising application of non-rigid registration which dramatically reduces contouring time. Two approaches are evaluated that use atlas-based autosegmentation: a single subject approach and a multiple-subject approach. It was shown that the multiple-atlas approach significantly improved the automatically generated contours for head and neck patients. Overall, we conclude that non-rigid registration techniques are playing an increasingly important in modern radiotherapy techniques. However, it is important to realize that before non-rigid registration can be used safely and effectively the registration methods should be validated thoroughly.
UR - https://www.scopus.com/pages/publications/77950177633
U2 - 10.1007/978-3-642-03882-2_366
DO - 10.1007/978-3-642-03882-2_366
M3 - Conference proceeding
AN - SCOPUS:77950177633
SN - 9783642038815
T3 - IFMBE Proceedings
SP - 1381
EP - 1384
BT - World Congress on Medical Physics and Biomedical Engineering
T2 - World Congress on Medical Physics and Biomedical Engineering: Image Processing, Biosignal Processing, Modelling and Simulation, Biomechanics
Y2 - 7 September 2009 through 12 September 2009
ER -
