Abstract : A clear identification of arteries, veins and urinary tracts, together with their simultaneous visualization in 3D is required when planning laparoscopic surgery on kidneys. To this aim, multiple contrast-enhanced CT scans are acquired at different times during blood circulation, but the kidney motion, among other causes, seriously hampers their direct fusion. This study aims at developing algorithms to represent in 3D, from such images, the kidney surroundings as close as possible to the surgical reality.