A System to Support Laparoscopic Surgery by Augmented Reality Visualization Stijn De Buck 1 , Johan Van Cleynenbreugel 1 , Indra Geys 1 , Thomas Koninckx 1 , Philippe R. Koninck 2 , and Paul Suetens 1 1 Faculties of Medicine and Engineering Medical Image Computing (ESAT and Radiology) 2 Department of Gyneacology University Hospital Gasthuisberg Herestraat 49 B-3000 Leuven stijn.debuck@uz.kuleuven.ac.be Abstract. This paper describes the development of an augmented re- ality system for intra-operative laparoscopic surgery support. The goal of this system is to reveal structures, otherwise hidden within the laparoscope view. To allow flexible movement of the laparoscope we use optical tracking to track both patient and laparoscope. The necessary calibration and registration procedures were developed and bundled where possible in order to facilitate integration in a cur- rent laparoscopic procedure. Care was taken to achieve high accuracy by including radial distortion components without compromising real time speed. Finally a visual error assessment is performed, the usefulness is demon- strated within a test setup and some preliminary quantitative evaluation is done. 1 Introduction Alaparoscopicsurgeryconsistsofmakingtinyholesintotheperitoneumthrough which subsequently a camera and surgical tools are inserted. The camera view is in fact the only view on the surgery scene. Because this viewpoint is constantly changing to meet the surgeons need and because it is very different from the exoscopic view of the surgeon, the latter has to be very well trained and be able to interpret the laparoscopic images well. Due to its minimal invasive approach, laparoscopic surgery is able to reduce both the time of surgery and the recovery time of a patient. The laparoscopic view though does not reveal all the structures the surgeon needs to see in order to complete the surgical procedure with success. These structures can for instance be hidden behind the peritoneal wall, like the ureter. This limitation can not only lead to a less efficiently performed surgery but also to further complications for the patient. Often such structures can be extracted from pre-operative CT/MR-images. However the surgeon needs to interpret and to fuse these images mentally with W. Niessen and M. Viergever (Eds.): MICCAI 2001, LNCS 2208, pp. 691–698, 2001. c  Springer-Verlag Berlin Heidelberg 2001