Endoscopic video images analysis for surgical training and image-guided surgery

Abstract

Minimally Invasive Surgery (MIS) has drawn forth two main necessities, training and assessment of surgical skills, and new tools and systems to extent surgeon's capabilities. Video endoscopic images can be a rich and always available source of information. Its analysis can be used for objective assessment of motor skills, 3D reconstruction and imageto-physical intraoperative registration. This paper presents a method for tools' tracking and an algorithm for 3D reconstruction of the surgical scene. This methodological approach does not require any additional hardware or any additional burden for the surgeon. Preliminary results support the adequacy of using endoscopic video information for achieving these goals. Future works will be focused on the development of more robust, feasible and real-time video analysis methods and their integration on the development of new surgical navigation systems and evaluation methodologies, which will lead to enhanced surgical training programs and safer surgeries.