Gaze, posture and gesture recognition to minimize focus shifts for intelligent operating rooms in a collaborative support system

This paper describes the design of intelligent, collaborative operating rooms based on highly intuitive, natural and multimodal interaction. Intelligent operating rooms minimize surgeon's focus shifts by minimizing both the focus spatial offset (distance moved by surgeon's head or gaze to the new target) and the movement spatial offset (distance surgeon covers physically). These spatio-temporal measures have an impact on the surgeon's performance in the operating room. I describe how machine vision techniques are used to extract spatio-temporal measures and to interact with the system, and how computer graphics techniques can be used to display visual medical information effectively and rapidly. Design considerations are discussed and examples showing the feasibility of the different approaches are presented.