Machine Stereo Vision: Revolutionizing Surgery
During surgery the preoperative imagery is displayed to guide the surgeon during the procedure. A shortcoming of current practice is that it can be difficult for the surgeon to perceive the geometric relationship between the displayed preoperative imagery and the live surgical scene before them.
We respond directly to this challenge by yielding a machine vision system for automatically registering live video of a surgical scene to corresponding preoperative imagery. The registration will enable a graphical overlay of the live imagery to the model that will increase the situational awareness of the surgeon.
This capability will improve the accuracy and safety of surgical procedures by minimizing collateral damage and complications. These advances will revolutionize surgery.
We will also optimize our system for performance in surgical scenarios and field test it in the operating rooms of Memorial Sloan Kettering Cancer Centre. Resulting performance evaluation will be used to guide integration with image-guided surgery systems at Memorial Sloan Kettering Cancer Centre.