Performance metrics for guidance active constraints in surgical robotics

Active constraint (AC)/virtual fixture (VF) is among the most popular approaches towards the shared execution of subtasks by the surgeon and robotic systems. As more possibilities appear for the implementation of ACs in surgical scenarios, the need to introduce methods that guarantee a safe and intuitive user-interaction increases. The presence of the human in the loop adds a layer of interactivity and adaptability that renders the assessment of such methods non-trivial. In most works, guidance ACs have been evaluated mainly in terms of enhancement of accuracy and completion time with little regard for other aspects such as human factors, even though the continuous engagement of these methods can considerably degrade the user experience. This paper proposes a set of performance metrics and considerations that can help evaluate guidance ACs with reference to accuracy enhancement, force characteristics and subjective aspects. The use of these metrics is demonstrated through two sets of experiments on 12 surgeons and 6 inexperienced users.