There are large, and vitally important individual differences in aptitude for laparoscopic (‘keyhole’) surgery. Intuitively it seems likely that an important component is ability to recover the shape of 3-D structures (patients’ internal organs) from displays on a video monitor. The issue was studied empirically with PICSOR, a test based on research reported by Cowie and Byth (1995 Perception 24 118 – 119). Subjects adjust a probe, which consists of a rotating arrowhead, until it appears perpendicular to a specified surface in a simple object (a sphere or a cube). Both probe and surface are depicted on a VDU screen. Measures of apparent slope are derived and related to the theoretical slope. Eight trainee surgeons were studied. They completed a battery of simulated laparoscopic tasks and the PICSOR tests. Two measures of performance on PICSOR were derived from simple regressions relating apparent and theoretical slope, R, and the slope of the fitted line. These were used in multiple regressions to predict performance in the simulated laparoscopic tasks. Performance on the PICSOR cubes task predicted performance in simulated surgery (F2,7=29.5, p=0.0017). Both measures (R and slope) contributed significantly. Relationships are probably not linear. One subject performed very badly on both PICSOR and simulated operations. If his data are excluded, both low R and high slope would appear to be related to poor laparoscopic performance. The nature of the area makes large-scale studies difficult, but the data suggest that they ought to be conducted.
