MODELING, DEVELOPMENT AND CONTROL OF A BILATERAL MASTER-SLAVE ROBOTIC SYSTEM AIMING NATURAL ORIFICE TRANSLUMINAL ENDOSCOPIC SURGERY

Abstract

Natural orifice transluminal endoscopic surgery (NOTES) combines laparoscopic minimally invasive surgery (MIS) and endoscopy. The use of the body’s natural orifices, such as the mouth, nostril, and others, has garnered a lot of interest from the scientific community in re cent years since it lessens trauma and hospital stays. The main constraints for NOTES tech nology are degrees of freedom (DOF), workspace, resolution, platform stability, retraction force, force feedback, triangulation, and visualization. A master-slave (M-S) based teleoper ated system can provide solution to the constraints of the NOTES. The main aspects of the M-S based NOTES are to utilize the redundancy of surgical tool for improved maneuverabil ity, to control the surgical tool for better tissue interaction, providing haptic feedback to assist surgeon and precise positioning of the surgical tool tip for the application such as internal or gan’s inspection etc. In this work, modeling of M-S based teleoperated system is done using bond graph for testing the feasibility of the system and simulations are performed for spatial trajectory tracking and tissue interaction. For experimental validation, a M-S based robotic surgical system is developed. The system consists of a Geomagic touch haptic device as the master robot and an in-vivo surgical tool as a slave robot. The experimental validation using the developed surgical system is done for three tasks: (a) A dual-loop control consisting of instantaneous maximization of mechanical joint limit violations (MJLV) based on kinematic control and optimal parameter-based adaptive control was designed for better maneuverabil ity and keeping surgical tools stiff during tissue interaction. Further, an impedance control was implemented to control the slave robot tip and biological tissue interaction for prevent ing the collateral damage of healthy tissue. The interaction force between the slave robot and tissue is sensed using force sensor and is given as haptic feedback to the surgeon. (b) For the online instantaneous Cartesian space trajectory tracking inside the concise surgical workspace, a neural network (NN) based intelligent control technique that uses radial basis function network (RBFN) activation function is developed and its performance is compared with trajectory tracking method that uses back propagation network (BPN) and kinematic control based instantaneous maximization of manipulability measure. (c) To control the ve locity of surgical tip during precise motion inside concise surgical workspace, a optimal xxiii kinematic control based on single network adaptive critic (SNAC) is developed. The devel oped control techniques and their validated experimental results show that the M-S based surgical system for NOTES can be helpful for the surgeon in tissue interaction application and free environment motion of the surgical tool inside the concise surgical workspace. The developed system can further be used for the training purpose of naive surgeons.