Deformable 4DCT Lung Registration with Vessel Bifurcations
Anna Hilsmann, Vik, T., Kaus, M., Kevin M. Franks, Bissonette, J.-P., Purdie, T., Beziak, A., Aach, T. · Publikationsdatenbank der Fraunhofer-Gesellschaft (Fraunhofer-Gesellschaft) · 2007
In radiotherapy planning of lung cancer, breathing motion causes uncertainty in the determination of the target volume. Image r egistration makes it possible to get information about the deformation of the lung and t he tumor movement in the respiratory cycle from a few images. A dedicated, a utomatic, landmark-based technique was developed that finds corresponding vessel bifur cations. Hereby, we developed criteria to characterize pronounced bifurcations for which correspondence finding was more stable and accurate. The bifurcations were ext racted from automatically segmented vessel trees in maximum inhale and maximum exhale CT thorax data sets. To find corresponding bifurcations in both data set s we used the shape context approach of Belongie et al . Finally, a volumetric lung deformation was obtain ed using thin-plate spline interpolation and affine registration. The m ethod is evaluated on 10 4D-CT data sets of patients with lung cancer.