Interactive guidance by image overlay in robot assisted coronary artery bypass

Abstract

We present an original visual guidance system in the especially difficult context of robot assisted coronary artery bypass graft. The overlay of a preoperative coronary tree model on the endoscopic images is initialized to help the surgeon to locate himself. Then the surgeon points some landmarks observed in the operating field during the motion of the endoscope. The overlay is corrected in real-time and the landmarks are automatically identified through a multi-model estimation method until the precision necessary to the localization of the surgical targets is reached. The first in vivo experiment, presented here, has shown the efficiency of the system by identifying the main arteries in a few clicks. © Springer-Verlag Berlin Heidelberg 2003.