miccai2006

Telemanipulation of Snake-Like Robots forMinimally Invasive Surgery of the Upper Airway: 

Telemanipulation of Snake-Like Robots for Minimally Invasive Surgery of the Upper Airway Ankur Kapoor1, Kai Xu2, Wei Wei2 Nabil Simaan2 and Russell H. Taylor1 1 ERC-CISST Department of Computer Science Johns Hopkins University 2 ARMA: Advanced Robotics & Mechanism Applications Department of Mechanical Engineering Columbia University

Slide2: 

*Courtesy of Paul Flint M.D. Johns Hopkins School of Medicine Surgical Setup in Throat MIS*

Limitations of the Surgical Setup: 

Limitations of the Surgical Setup motion constraint No distal dexterity Hand-eye coordination Long rigid instruments Predetermined entry port No suturing or functional tissue reconstruction capability

Human-machine cooperative manipulation in surgery: 

Display atlases Manipulation enhancement Online references & decision support Cooperative control and “macros” atlases libraries Human-machine cooperative manipulation in surgery Situation assessment Task strategy & decisions Sensory-motor coordination Sensors HMCS System

Related Works:Surgical Dexterity Enhancement : 

Related Works: Surgical Dexterity Enhancement Commercial Systems Zeus Intuitive Surgical Da-Vinci (Endo-Wrist) Research Works Dario (3 mm SMA for arthroscopy visualization) Ikuta (15 mm SMA, colonoscopy) Ikuta, Yamamoto, Sasaki (Deep surgical field) Fujie (Dexterity for Brain Surgery) Asai & Mituishi (5mm snake like device for microsurgery) Salisbury & Intuitive Surgical (Endo-Wrist, 5 mm wire actuated snake) Sastry & Cavusoglu (2-3 DoF ~8mm wrists) Jan Peirs (5 mm wire actuated snake) ….. And many other works

Related Works: Virtual Fixtures: 

Related Works: Virtual Fixtures Virtual fixtures: perceptual overlays designed to enhance performance Active Compliance Rosenberg Stanisic et al. Davies et al. Park et al. Vision Based Marayong et al. Dewan et al. Based on Constrained Control Funda et al. Li et al.

Snake-Like Units (SLU): 

Snake-Like Units (SLU) Uses push-pull superelastic backbones & actuation redundancy Eliminates dependency on precision joints & backlash Simple to manufacture Easily downs-scalable to smaller diameters Enhanced force application capability Simaan N. et al, MICCAI 2004, ICRA 2004, ICRA 2005

High Level Constrained ControlSteady Hand Robot: 

High Level Constrained Control Steady Hand Robot Handle Force Kv Joint Velocities Registered Model Current State Constraint Generation Optimization Framework Low Level Controller

5 Basic Geometric Constraints(Virtual fixture library): 

5 Basic Geometric Constraints (Virtual fixture library) Kapoor, A. Li, M., Taylor, R.H. Constrained Control for Surgical Assistant Robots, ICRA 2006 Optimization Framework

Snake Like RobotSystem Architecture: 

x2 (Left & Right) x2 (Left & Right) Snake Like Robot System Architecture Low Level Controller High Level Controller Low Level Controller High Level Controller

Master Side Low-Level Controller: 

Master Side Low-Level Controller The low-level is a PD Joint Controller The force applied by user is treated as disturbance Under quasi-static approximation, position error is proportional to user force Set Point User Force Encoder Position

Master Side High-Level Controller: 

Master Side High-Level Controller A constrained least squares problem is solved for joint velocities. Objective function determines the desired outcome. Constraints modify the behavior of the robot to a given input. Joint Velocities Slave Frame Master Frames

Master Side High-Level Controller: 

Master Side High-Level Controller Objectives: Minimize error between desired motion and actual motion Oppose motion that increases master-slave tracking error Minimize the extraneous motion of the joints, and Avoid large incremental joint motions that could occur near singularities

Master Side High-Level Controller: 

Master Side High-Level Controller Objectives: Minimize error between desired motion and actual motion Oppose motion that increases master-slave tracking error Minimize the extraneous motion of the joints, and Avoid large incremental joint motions that could occur near singularities

Master Side High-Level Controller: 

Master Side High-Level Controller Objectives: Minimize error between desired motion and actual motion Oppose motion that increases master-slave tracking error Minimize the extraneous motion of the joints, and Avoid large incremental joint motions that could occur near singularities

Master Side High-Level Controller: 

Master Side High-Level Controller Constraints: General form: Hm,j¢qm ¸ hm,j Not allow motion outside joint range Not allow motion that exceeds joint velocity limits Additional constraints can be added from the VF Library

Slave Side Low-Level Controller: 

Slave Side Low-Level Controller The low-level is a PID Joint Controller The two DOF of each snake are parameterized by two angles The bending angle of primary backbone The orientation of bending plane with respect to base XZ plane Set Point Slave Frame

Slave Side High-Level Controller: 

Slave Side High-Level Controller Objectives: Minimize error between desired motion and actual motion Minimize the extraneous motion of the joints, and Avoid large incremental joint motions that could occur near singularities Constraints: Not allow motion outside joint range Not allow motion that exceeds joint velocity limits More constraints can be added from the VF Library

Experimental SetupMaster: 

Experimental Setup Master

Experimental SetupSlave: 

Experimental Setup Slave

Experimental SetupRoll motion: 

Experimental Setup Roll motion

Experimental SetupS-bend motion: 

Experimental Setup S-bend motion

Experimental SetupSurgeon’s view: 

Experimental Setup Surgeon’s view

Experimental SetupFish Hook: 

Experimental Setup Fish Hook

Conclusion: 

Conclusion A novel system designed considering special requirements for MIS of throat High-level control of a telesurgical system Efficient use of dexterity avoids motion of proximal joints Validation experiments using suturing phantom

Acknowledgements: 

Acknowledgements Dr. Paul Flint, School of Medicine, Johns Hopkins University. NSF Research Grant #EEC9731478 and #IIS9801684 NIH #R21 – B004457-01 Johns Hopkins and Columbia University Internal Funds