Learning to stop looking at the task in hand

Abstract

Many surgeons have difficulties transitioning from open to laparoscopic surgery. A common report is that surgeons (particularly novices) look at their hands when performing laparoscopic operations, rather than fixating on the screen (Law et al., 2004). Skilful movements require feedback information (typically visual) to make online corrections, but if there is a mismatch between the direction of hand motion and the ensuing action then the system needs to adapt. By adulthood, individuals have become proficient in visually guiding their hand to a target, but it seems that they are also adept at guiding their hand when visual feedback is absent or distorted (e.g. using a mouse to control the movement of a visual cursor on a computer screen). We hypothesised that if manual control is dominated by visual-feedback about hand-position then distorted visual-feedback would be most problematic when the hand is visible (‘visual capture’). We tested our hypothesis by studying the rate of adaptation to distorted visual-feedback in two separate groups (n = 44). One group (indirect) were prevented from seeing their hand and viewed the display on an external monitor. The other group (direct) viewed their hand and the visual display directly. The task involved carrying out a series of aiming movements using a handheld stylus on a tablet-laptop. Both groups completed baseline trials and then two distorted visual-feedback blocks (30° and 60° rotation, counterbalanced) before being tested at 45° the following day. We used movement duration as an index of learning. The results revealed group differences, with the indirect group adapting more readily to the distorted feedback. These results show that additional information about hand-position actually impairs adaptation to distorted visual feedback, which suggests that surgeons need to learn to fixate the screen rather than their hands during laparoscopic operations.