Robotic Surgical Assistants (RSA) are currently directly controlled by surgeon using teleoperation. Long surgical hours can cause significant fatigue for surgeon. Automating surgical subtasks can improve surgical performance, while reducing fatigue.

We are trying to automate surgical sub-tasks such as needle handover and suturing. Such tasks require safe and precise manipulation techniques. We apply reinforcement learning and deep learning techniques to address this issue.

AUTOMATING SURGICAL SUBTASKs

Superhuman Surgical Peg Transfer Using Depth-Sensing and Deep Recurrent Neural Networks

VISUAL SERVOING CONTROL

Learning Surgical Needle Handover

Intermittent Visual Servoing: Efficiently Learning Policies Robust to Instrument Changes for High-precision Surgical Manipulation

Markerless Pose Estimation of Surgical Instrument

Hysteresis Compensation of Continuum Manipulator

SAM: Semi-Active Mechanism for Extensible Continuum Manipulator and Real-time Hysteresis Compensation Control Algorithm

Hysteresis Compensation of Flexible Continuum Manipulator Using RGBD Sensing and Temporal Convolutional Network

Efficiently Calibrating Cable-Driven Surgical Robots With RGBD Fiducial Sensing and Recurrent Neural Networks

Effect of backlash hysteresis of surgical tool bending joints on task performance in teleoperated flexible endoscopic robot

 Semi-autonomous control protocols and applied machine learning algorithm to improve the safety and efficacy of robotic super-microsurgery