A FEATURE MATCHING-BASED APPROACH TO DEFORMATION FIELDS MEASUREMENT FROM MR IMAGES OF NON-RIGID OBJECT

Though a variety of difierent algorithms have been implemented for estima- tion of the deformation flelds of biological tissue from magnetic resonance (MR) images, few attempts in feature tracking areas have been reported. In this study, we propose a method to measure deformation flelds of biological tissues based on local feature track- ing. First, we use correlation score (cs) based method to obtain a candidate matches set. Next, relaxation technique is used to disambiguate matches. Third, the dense deforma- tion flelds is calculated using linear interpolation approach within Delaunay triangles net. To test the validity of our approach, we applied the proposed approach to MR images of a volunteer's calf. Moreover, the reverse movement of selected check points was used to evaluate the reliability and accuracy of the result. Preliminary experiment results of this