Automatic Extraction and 3D Visualization of Coronary Arteries from Angiography Sequences

M. Espinosa Sandoval, L. Altamirano Robles, J.A. González Bernal, and R. D'Zatarain Rivero (Mexico)


Coronary angiography, automatic extraction, 3D reconstruction, uncalibrated stereo system.


In this work we explore the feasibility of reconstructing 3D information of coronary arteries and their lesions from coronary angiography sequences. The angiographies were taken during the examination of the vessels with an uncalibrated camera. The camera is still and we assumed that natural movement of arteries is large enough to create the change of view necessary for a stereo effect. We execute an automatic extraction algorithm for the arteries, then we aligned the consecutive pairs of the sequence and finally we used a correlation algorithm to obtain a dense disparity map to finally reconstruct the 3D model of the coronary arteries and their lesions. This preliminary study indicates that this approach could provide a simple way to extract 3D information of the coronary arteries without the use of a stereo image acquisition setup.

Important Links:

Go Back