Abstract
Computerized tomography, or CT, produces a series of cross sectional images that provide detailed three dimensional representations of the body's internal structures. According to Yu, et al., 2011, each 10 minute CT scan has a 59% chance of being interrupted by a phone call. The current study examines how these interruptions affect the accuracy and efficiency of volumetric visual search. In addition, eye-tracking data was recorded to determine how faithfully search is resumed after an interruption. Participants searched through 20 CT scans for artificial lung nodules. Lung nodules are identified as circles that pop in and out of view as the participant scrolls through the depth of the lung. Between 30 to 60 seconds after search onset, half of the trials were interrupted by a series of true or false math equations. The number of missed nodules did not differ between interruption and non-interruption trials. However, search duration was approximately 13% longer (22s) in interruption trials. Eye-tracking data was used to better understand the source of this time cost. Interruption trials were compared to non-interruption trials by randomly sampling eye position between 30 to 60 seconds into non-interruption trials. The first fixation after interruption was in a different quadrant of the lung in approximately 75% of trials, compared to a 35% change during equivalent time periods in non-interruption trials. These results suggest that interruption impairs memory for which regions of the lung have been examined. Previous evidence has suggested that memory for where we have searched is quite poor (Horowitz & Wolfe, 1998; Wolfe, Drew, Vo, 2014). The current study suggests that this limitation may be exacerbated by interruption. In future work, we will explore the factors that predict the costs of interruption with the eventual goal of helping mitigate these costs in diagnostic radiology.
Meeting abstract presented at VSS 2016