| This thesis tries to answer the question: how to carry out a three-dimensional vascular reconstruction from a single view only? Traditionally, 3D reconstructions are realized from biplane imagery apparatus, but these apparatus are complicated to operate and are not always available in many clinical centres. Although depth information is non-existent on a two-dimensional image, by using a priori knowledge of the subject under study our approach makes it possible to recover it. Two different contexts are addressed for 3D reconstruction.;Firstly, we consider the context where 3D reconstruction is applied to coronary arteries from single-plane cineangiography. In this case, the a priori information comes from a model of the coronary tree created from biometric measurements carried out by Dodge et al. (J. T. Dodge et al., 1986, 1992). By adding a temporal component to the spatial reconstruction, it becomes possible to follow and quantify the amplitude of cardiac contractions from the coronary tree movements.;Secondly, single-plane 3D reconstruction is carried out in the fusion of two medical imaging modalities context, namely IVUS (IntraVascular UltraSound) and angiography imaging. The shape of the IVUS transducer, which is known a priori, is used to do the 3D reconstruction of the IVUS catheter pullback path from single-plane angiography. Once geometrically aligned perpendicularly to the catheter path, precise measures on atherosclerotic plaque extents can be made on the IVUS images. |
