Landmark based correlation of medical images

Abstract

In recent years, correlation of medical images has been of increasing importance for the definition of regions of interest for diagnostics, and for the definition of target volumes and target points for therapy. The method described only requires the definition of three corresponding landmarks in each of the correlated image data sets. Based on the Cartesian coordinates of the landmarks, it makes it possible to map additional points and contours defined as a set of points from one image data set into another. The transformation works in both directions. As additional information, the expected accuracy of the resulting point coordinates is calculated.