Shadow-Based Depth Perception with Adaptive Motion Scaling for Safe Teleoperated Retinal Surgery

Ahnaf Naheen, Korab Hoxha, Angelo Henriques, Junjie Yang, Mahdi Tavakoli, M. Ali Nasseri

Keywords: Robotics, Teleoperation, Microsurgery

TL;DR: We demonstrate that combining shadow-based depth estimation with adaptive motion scaling significantly improves safety in long-distance teleoperated retinal surgery.

Abstract: Retinal microsurgery demands extreme precision, with tool positioning tolerances often within tens of micrometres to avoid damaging delicate tissue. In teleoperated retinal surgery, network latency and the absence of stereoscopic cues can compromise safety. Surgeons traditionally rely on naturally occurring shadows of instruments on the retina to infer depth when using a microscope. We leverage these shadows to quantitatively estimate tool-retina distance in real time; integrating this with latency-aware motion scaling to enhance safety and intuitiveness in long-distance teleoperation.

Submission Number: 18