Fully automatic cardiac motion estimation from Tagged MRI using non-rigid registration techniques

Many different techniques have been proposed to compute motion fields on Tagged MR imaging. However, most of them require tags segmentation or endo- epicardial contours segmentation. Automatic estimation of the myocardial motion field was extracted using a consecutive non-rigid registration algorithm based on a semilocal Bspline parametric model. The sequence was also represented in the Bspline space using this framework to construct the multiresolution pyramid within the optimization process. Validation was performed on sequences from 5 healthy volunteers comparing the estimated trajectories with the manual tracking provided by an expert. Results were also compared to HARP analysis. Subpixel accuracy was obtained, and results were superior in all cases to HARP analysis.