Surgical robotic systems and computer assisted surgery have revolutionized the delivery of surgical care by providing precision, accuracy and miniaturization of instrumentation unachievable with manual equipment and techniques. Despite advancements generated by these technologies, unmet challenges exist when working in the unstructured flexible environment of a surgical site. Current surgical robots are limited to preoperative surgical plans and minimal intraoperative sensory information due to inherent design limitations of these systems and lack of frameworks for utilization of real-time data. These limitations place restrictive burdens on surgeons for control with incomplete information on the robot, the surgical environment and their interaction.

Acquisition and intelligent use of real-time intraoperative sensory information to augment current systems will result in a new generation of “smart” surgical robots that will enable surgeons to make strides in the complexity of techniques, precision, and overall capabilities of surgical procedures. These improvements will be made by adapting to the environment during teleoperation in order to make surgical slaves behave more reliably and safely. Improved intraoperative data integration will also allow autonomous performance of low level tasks, such as retraction, exploration of tissue margins, biopsy and suturing.

This doctoral study seeks to advance methods and systems for addressing fundamental limitations of existing surgical robots by adding intraoperative intelligence based on mechanical sensory information. Interaction force data on the surgical slave unit can be acquired either through direct measurement by a dedicated force/torque sensor at the interaction site or indirect measurement of robot actuation forces. With interaction data, models of the environmental interaction can be constructed and novel control laws applied to modify the behavior of a robotic system to improve safety and surgical performance.

In the first section of this dissertation, general algorithms for exploration and control in flexible environments are investigated for surgical robots with force sensing capabilities. Hybrid forcemotion control and redundant coverage paths are described for exploration of the shape of a flexible environment. Based on localized excitation of tissue coupled with simultaneous force measurements, an algorithm for discrete tissue impedance estimates is presented and evaluated for adaptive exploration and segmentation of embedded features. The application of these algorithms for autonomous exploration and estimation are shown in flexible tissue models.

In the second section of the dissertation, a framework for compliant motion control is presented for continuum surgical robots subject to unknown interactions with the environment. Through a mapping of unknown environmental interaction forces to a generalized description, joint level actuation force measurements serve as an input to a compliant motion controller that allows surgical slaves to actively comply with environmental forces. Un-modeled effects on the joint level forces are corrected via a feed-forward online estimate. Linear and non-linear regression techniques are evaluated for estimation and compensation of these model uncertainties. Conditions for the stability of the controller are defined and experimentally validated for complex multi-point robot-environmental interaction.

Finally, the design and analysis of a novel telerobotic system is presented for minimally invasive surgery in deep surgical sites. The clinical requirements for a benchmark application in transurethral resection of bladder tumor and the design considerations for this system are described. Kinematic analysis of the dexterity at the surgical site and experimental evaluation of the manipulation capabilities of the system are presented in the context of representative clinical tasks.

In summary, the algorithms and analysis presented in this dissertation constitute a methodology for collection and integration of sensory information toward the development and deployment of surgical robots with improved capabilities. The fundamental discoveries introduced will contribute to the development of a next generation of smart surgical robots that intelligently interact with the surgical environment leading to safer, faster, less invasive procedures with improved surgical outcomes.