Minimizing Movement Time in Surgical Telerobotic Tasks

Surgical robotic devices are rapidly achieving widespread use and acceptance. Despite the many benefits of robotic-assisted surgery, it is typical for robot-assisted procedures to take longer. This study investigates the effect of control:response ratio on simple movement time in surgical telerobotic tasks. Robot control interfaces offer "motion scaling" settings but, in practice, this feature is often not used effectively, and is confounded by factors such as the fulcrum point of the laparoscopic instrument. Results using a simple aimed movement task indicate that control:response ratio indeed has a large effect on movement time, and further that it interacts with task difficulty. These results can guide medical device designers and the developers of surgical training protocols. INTRODUCTION The benefits of surgical robotics such as motion scaling, tremor filtration and ambidexterity have been heralded by others in numerous surgical fields. Through its minimally invasive nature, patients experience less pain and discomfort, reduced trauma, shorter hospital stays, faster recovery and less scarring. However, it is also well documented that robotic surgery has a major drawback. That is, robotic surgery takes longer than traditional manual techniques. In order to better understand the humanrobot interface, and ultimately to make suggestions for minimizing movement time in robotic tasks, researchers have begun to investigate parameters such as control:response ratio (C:R), also known as control gain (CG). Cassily (2004) examined motion scaling and magnification in robotic surgery. Their task involved piercing archery targets with a needle. Their range of scaling and magnification consisted of 1:1, 5:1, 10:1 and 3.5X, 6.5X, 9.5X respectively. It was not clear, though, whether the scaling referred to the computerized settings on the robot or the actual control:response ratio as no reference was made to either. Their results show the 5:1, 10:1 settings had higher accuracy with the 5:1 scaling having a shorter task completion time than the 10:1 scaling. The 9.5X magnification had the fewest errors. Prasad (2004) also used a simple aimed movement task piercing an archery target to examine how motion scaling and tremor filtration benefits robotic surgery as compared to traditional laparoscopy. Their range of motion scaling consisted of 1:1, 2.5, 7:1 (hand:robot). Their results showed that a motion scaling of 2.5:1 and 7:1 had a higher rate of accuracy than a scaling of 1:1 or manual laparoscopy. Ellis, et al. (2004) used a simple aimed movement task at multiple settings of C:R ratio and optical zoom. They found that across several zoom settings (25X, 16X, 10X, 6.4X) there was a significant effect of C:R on response time, and further, that the optimal control gain setting appeared to vary according to zoom level. For example, the best gain for 16X zoom was 5.2:1, while the best gain for 6.4X was 2.8:1. The current study replicates and extends these results. Our intent was to map out the effect of control gain on movement time across the entire range of possible control gains encountered in using the device. For the present study, we were interested in the effect of the natural range of C:R, but also in the extreme values that could be encountered as a PROCEEDINGS of the HUMAN FACTORS AND ERGONOMICS SOCIETY 49th ANNUAL MEETING—2005 1099