Abstract
Purpose :
Traditional fundus imaging devices display static device guides with a dynamic live view of the eye to aid with alignment. This feedback mechanism presumes a mental model that the eye is moving relative to the device. If this presumption is not consistent with the mental models of users, then successful use requires computations from users in order to interpret the alignment feedback. Matching the feedback display mechanism to the mental model of the user should decrease cognitive load during alignment, decreasing learning curve and alignment time. The interaction of initial mental model and alignment feedback mechanism is examined to determine the effect on ease-of-use and alignment times.
Methods :
Twenty novice imaging technicians captured images using the CLARUS™ 500 (ZEISS, Dublin, CA) with a modified alignment user interface (UI). Participants moved the device to align graphic representations of pupils that were offset along the horizontal (x) or vertical (y) axes. Movements along each plane represented users initial mental models. Next, participants captured images from a cohort. They practiced alignment before capturing the images. In condition A, the feedback mechanism matched the movement of the device. In condition B, the movement was inverted. The two conditions were counterbalanced. Users then rated the ease-of-use of the two conditions.
Results :
84% of participants aligned with a mental model that they were moving the device relative to the eye. Average alignment time for condition A was 43 seconds (SD=13.5). Average alignment time for condition B was 76 seconds (SD=53.1). Condition A, consistent with the mental model of most of the participants was on average 33 seconds faster. Condition A also had significantly less variability (V=182.567) than condition B (V=2821.60), with a ratio of variances of 15.455 (P<0.05). Ease-of-Use ratings between the two conditions did not differ significantly.
Conclusions :
These results show that the initial mental model of novice imaging technicians is consistent with the feedback mechanism used on many fundus imaging devices, including the CLARUS 500. Acquisition times, though faster for condition A, were not significantly different. However, variability was significantly greater in condition B, suggesting that matching the mental model of users provides a more consistent experience across imaging technicians.
This is a 2020 Imaging in the Eye Conference abstract.