Enhanced visuomotor learning and generalization in expert surgeons

Although human motor learning has been intensively studied for many decades, it remains unknown whether group differences are present in expert cohorts that must routinely cope with and learn new visuomotor mappings such as minimally invasive surgeons. Here we show that expert surgeons exhibit greater adaptation and generalization compared to naive controls in a standard visuomotor adaptation task. These findings run counter to a widespread background assumption in the field of motor learning that visuomotor adaptation performance should be largely uniform across the adult human population. Our findings also indicate that differences in basic visuomotor learning capacities, either innate or acquired, might be an important source of difficulty in learning and performing minimally invasive surgery. This information holds potential to guide surgical candidate selection or optimize training programs to address individual needs.


33
Human motor learning has been intensively studied for many decades [1, 2]. However, 34 insufficient attention has been given to individual-or group-level differences in visuomotor 35 learning. For example, it remains unknown whether group differences are present in 36 expert cohorts that must routinely cope with and learn new visuomotor mappings such as 37 minimally invasive surgeons. indirectly via a laparoscopic camera that is in turn projected to a video display, rather than 48 through direct observation, they must also contend with a range of visualization problems 49 including absent depth information, variable magnification, and a restricted and frequently 50 distorted (e.g., rotated) field of view. These factors, which are often subsumed under the 51 general rubric of "challenges for hand-eye coordination" [7], also impose heavy 52 computational demands on the brain and likely contribute to the significant increase in 53 time to achieve proficiency in MIS    Subjects were instructed to perform fast and accurate reaching movements with the 107 dominant (right) arm using cursor feedback, whenever it was available. Subjects 108 performed reaches from a start target located at the center of the workspace to 11 109 different target locations 9 cm away from the start target and spaced 30° apart (Fig 1) The 6 110 start target was a solid red circle (5 mm diameter), and each reach target was a solid 111 green circle (5 mm diameter). The appearance of the reach target served as the go cue.

112
Subjects were positioned so that the starting target was directly in front of their torso.

113
Although these data are not reported here, a second workspace was also tested. In this 114 experiment, the target array was translated rightward, which required subjects to extend 115 their shoulder joint clockwise 45° to acquire the start target [23].

217
The within-subject factor of BIN and the between-subject factor of GROUP were 218 compared via repeated-measures ANOVAs across 11 bins of 10 trials per bin. There was 11 219 a significant within-subject effect of BIN (F 1,7.465 = 59.6, p<.001, 2 = 0.231; Greenhouse-220 Geisser corrected) as well as a significant between subject effect of GROUP (F 1,7.465 = 221 269, p<.001, 2 = 0.573; Greenhouse-Geisser corrected). These results indicate that 222 while both groups adapted rapidly, experts did in fact adapt significantly faster than 223 controls (Fig 3). In addition, averaged over the last two bins of trials during the adaptation  Reaction time and duration decreased over the course of the adaptation phase while peak 227 velocity increased for both experts and controls (Fig 3). Although not our primary focus, 228 there were significant within-subject effects of BIN as well as significant between subject 229 effects of GROUP for all remaining movement parameters (Table 1). Finally, we investigated whether generalization of learning differed between the groups.

236
Local generalization to new, untrained target directions decreased as a function of 237 distance from the trained direction (up to ±90° from the trained target direction) for both 238 experts and controls (Fig 4). The within-subject factor of TARGET and the between-  surprisingly little attention has been given to surgeons as an informative expert cohort.

303
The current study begins to close this important gap by showing that expert minimally 304 invasive surgeons exhibit enhanced visuomotor learning in every measure we tested.

305
Specifically, experts adapted to the imposed visuomotor rotation more rapidly and more 306 completely than controls. They also generalized their learning to new target directions 307 more broadly and to a considerably greater extent. 15 311 Similar to our study, they reported lower movement variability (higher consistency) and 312 greater accuracy in their expert group prior to learning, both widely considered to be 313 hallmarks of expert performance [41]. But interestingly, they observed no learning rate 314 differences in one experiment and a slower rate of adaptation in experts compared to 315 novices in another experiment. One plausible explanation for the large discrepancy 316 between their findings and ours is that handball players do not have to contend with or 317 achieve mastery over changes in visuomotor mappings as do expert surgeons with 318 extensive training and experience performing MIS.