Introduction d'un modèle de respiration dans une méthode de recalage à partir de points d'intérêt d'images TEP et TDM du poumon Using a breathing model for landmark-based CT-PET lung registration

This paper deals with the problem of 3D registration of Computed Tomography (CT) images (at two different instants of the breathing cycle) and Positon Emission Tomography (PET) images of thoracic regions. In order to guarantee physiologically plausible deformations, we presen t a novel method to incorporate a breathing model in a nonlinear registration procedure. Our registration method is based on an automatic selection of landmark points based on the curvature of the lung surface. The rigidity of the potential tumors is preserved during the registration, while guaranteeing a continuous deformation. Initial results on one normal case and four pathological cases demonstrate the interest of this method to significantly improve the accuracy of multi-modal volume registration.