2D-3D Correspondence in Mammography

The fusion of various images obtained from non-rigid deformed objects is discussed. We present a framework for the fusion of multi-modal medical images and we show its application to registration and correspondence of mammographic MR (3D data) and X-ray (2D data) images. The robustness of the presented work relies on the development of a novel method to establish non-linear correspondence between modalities of different dimensionality, which also represent different physical tissue aspects. The correspondence is based on a matching process, which takes into account features from internal linear structures from both images. Initial quantitative and qualitative evaluation results are given which indicate the validity of our approach.