Utility of 2D properties in the reconstruction of coronary arteries from biplane angiographic images

Tri-dimensional (3D) visualization of coronary arteries can considerably improve the diagnosis and therapeutic evaluation of the related coronary dysfunction. Reconstruction based on bi-dimensional (2D) images is a multiple solution problem. The correct solution identification is only possible using a priori information. This includes 3D and 2D aspects. In this work the authors study the influence that 2D properties can have on the reduction of the search space (possible solutions) for the right coronary case. The results obtained from real images show that selected properties (diameter correspondence, and relative position of some arteries) are very useful for the correct solution identification. Also, the modeling done is not very sensitive to the parameters involved and hence to the input data precision