3D Image Matching Using a Finite Element Based Elastic Deformation Model

摘要: We present a new approach for the computation of deformation field between three dimensional (3D) images. The minimizes sum squared differences images to be matched and is constrained by physical properties different objects represented image. are modeled as elastic bodies. Compared optical flow methods, this distinguishes itself main characteristics: it can account actual deformed, provide us with deformed (i.e. stress tensors), computes global solution instead set local solutions. This latter characteristic achieved through finite-element based scheme. finite element requires in meshed. Therefore, tetrahedral mesh generator using pre-computed case table specifically suited segmented has been developed. Preliminary experiments on simulated data well medical have carried out successfully. Tested applications included muscle exercise imaging ventricular multiple sclerosis.