Sylvia Guillory¹, Hiu Mei Chow², Beatrix Culligan², Sierra Nastasi^1,3, Chad Peltier^1,3, Jeffrey Bolkhovsky¹; ¹Naval Submarine Medical Research Laboratory, ²St Thomas University, ³Leidos, Inc

Cognitive demands and uncertainty in decision-making can influence task-evoked changes in pupil size (e.g., increase with working memory load). Uncertainty can arise from internal or external sources, such as lapses in attention or distractions. These factors unrelated to task instruction can influence uncertainty and lead to performance errors. A goal of this experiment was to examine the relationship between changes in pupil size and uncertainty in attentional selection. We hypothesized that reduced pupil size would be associated with greater uncertainty/task-disengagement, as indexed by poorer task performance. Observers (n=10) participated in a visual search task in which a target letter “T,” presented at a 10% prevalence rate, appeared among 25 non-target offset “Ls.” Eye positions and pupil diameter were sampled using a video-based eye-tracker. For each trial, observers made a response within seven seconds of stimulus onset to report whether the target was present or absent. No feedback was given. Trial type was categorized based on target presentation and observer response (hit, miss, correct rejection, false alarm). Results showed there was a significant difference in pupil diameter based on trial type (p<.05) suggesting that task-evoked changes in pupil diameter may reflect uncertainty (i.e., errors). Specifically, the average pupil diameter across the stimulus presentation period was 1.1 times larger in trials with hit vs. miss responses, and 1.05 times larger in trials with correct rejection vs. miss responses. This pupillary pattern corresponded with task performance (e.g., shorter reaction time in trials with hit vs. miss responses (p<.013). Lapses in attention can be detrimental in professions requiring extended periods of watch, such as the military operations or long-distance driving. These findings indicate that pupil diameter may serve as an index of uncertainty during attentional tasks, providing the information needed to optimize performance by identifying operational safety thresholds in terms of lapses in attention.

Acknowledgements: This work is supported by the Defense Health Program Research, Development, Test, and Evaluation Restore Core Funds (restoral) under funding work unit number F1710 awarded to JB/SG, as well as a Natural Sciences and Engineering Research Council of Canada Discovery Grant (RGPIN-2024-06028) to HMC.