Improving intensity-based lung CT registration accuracy utilizing vascular information

摘要: Accurate pulmonary image registration is a challenging problem when the lungs have deformation with large distance. In this work, we present nonrigid volumetric algorithm to track lung motion between pair of intrasubject CT images acquired at different inflation levels and introduce new vesselness similarity cost that improves intensity-only registration. Volumetric datasets from six human subjects were used in study. The performance four algorithms was compared without adding function. Matching accuracy evaluated using landmarks, vessel tree, fissure planes. Jacobian determinant transformation reveal pattern local parenchymal tissue. average matching error for methods on order 1mm landmarks 1.5mm After preserving function, landmark positioning errors decreased approximately by 25% 30%, respectively. function effectively helped improve regions near thoracic cage diaphragm all tested also produce more consistent reliable patterns regional tissue deformation.