Nima’s Ph.D.

PhD Candidate

Nima Enayati (Completed in May 2017)

Advisors

Prof. Giancarlo Ferrigno, Elena De Momi

Title

Adaptive Shared-Control in Surgical Robotics

Description

Nima_project_2PhDNima_Enayati

More operating rooms are equipped with robots every year under the premise that these new tools can enhance surgical procedures. Whether robotic agents have proven to produce measurable and all-aspects-considered outcomes in the operating room or not is a matter of ongoing debate. There is, however, little doubt that the current state of surgical robotics, is far from its potential. Unlike in the industries where robots have drastically outperformed humans while acting autonomously, the role given to robotic systems in the operating room has been broadly limited to the reproduction of surgeon’s commands. This is of course due to numerous technical difficulties in perception, cognition, control and actuation faced in soft tissue interaction and the complexity of surgical procedures. However, given the exponential rate of technological advancement, researchers have started preliminary works on granting a more active involvement to surgical robotic systems in recent years. The so-called shared-control approach aims at uniting the advantages of mechatronic systems with the superior cognitive abilities of humans. This work is an effort on investigating and implementing a shared-control method for surgical tele-operation where the robotic agent collaborates with a surgeon through suggestive haptic ques to enhance the operations in terms of safety, accuracy, execution time, cognitive load and surgeon fatigue.
A shared-control tele-operation framework is proposed that adapts its assistive properties based on the skill-level of users estimated through a machine-learning based tele-operation skill assessment. It is shown that different operators can demonstrate disparate behaviors in using a tele-operation system and therefore, developing a one-size-fits-all assistive method may not achieve optimal results when both explicit clinical outcomes and surgeon’s subjective experience are considered. It is shown here that tailoring the assistive characteristics of the system according to operator’s needs can improve the human-robot interaction, which in turn can lead to better surgical outcomes. The introduced guidance method can be employed in various surgical applications where precision is of value such as instrument insertion and accurate targeting, or in operations where safety is of interest such removal of tissue in the vicinity of delicate organs. Haptic assistance can also be exploited in motor skill training to improve the learning curve of motor tasks through the reduction of cognitive load. In training of surgical tasks with complex kinematics, where it has been shown that residents can face difficulties due to poor hand-eye coordination, some form of haptic guidance can speed up the initial motor learning phase. However, there are concerns about creating assistance dependency in the trainees, which can decrease performance during the actual task. The adaptive haptic assistance introduced in this work can address such concerns by monitoring the skill level of the trainee and gradually lower the intensity and frequency of the assistance to reduce the risk of creating dependency. This work includes experimental validation of the proposed methods. Experiments using implemented virtual environments show that the described haptic guidance enforcement method can enhance the outcome of a simplified surgical task while causing lower amount of distraction to the human operator. The skill estimation capabilities of the method is validated via experiments on a real-time teleoperation setup, where the feasibility of the adaptive guidance to is demonstrated for path following tasks.

PhD Period

XXIX cycle (November 2013 – November 2016)

This site uses cookies to enhance your experience. By continuing to the site you accept their use. More info in our cookies policy.     ACCEPT