Vision-based retinal membrane peeling with a handheld robot

Abstract

Peeling delicate retinal membranes, which are often less than five microns thick, is one of the most challenging retinal surgeries. Preventing rips and tears caused by tremor and excessive force can decrease injury and reduce the need for follow up surgeries. We propose the use of a fully handheld microsurgical robot and vision-based virtual fixtures to enforce helpful constraints on the motion of the tool. Our key contribution is using only visual information to reduce and limit forces during vitreoretinal surgery: no force feedback is used in the control system. Utilizing stereo vision and tracking algorithms, the robot activates motion-scaled behavior as the tip nears the surface, providing finer control during the critical step of engaging the membrane edge. A hard virtual fixture just below the surface bounds the total downward force that can be applied. Furthermore, velocity limiting during the peeling helps the surgeon maintain a smooth, constant force while lifting and delaminating the membrane. On a retinal phantom consisting of plastic wrap stretched on top a rubber slide, we demonstrate a reduction of maximum force by 40-70%.