The feasibility of using pupillometry to measure cognitive effort in aphasia: Evidence from a working memory span task

• ¹ University of Alberta, Department of Communication Sciences and Disorders, Canada
• ² University of Alberta, Faculty of Rehabilitation Medicine, Canada

BACKGROUND Cognitive effort, defined as the amount of processing resources devoted to performing a cognitive task (Tyler et al., 1979), is a clinically important construct that is often overlooked in the assessment and treatment of aphasia. Typically, aphasic language performance is assessed using measures of speed or accuracy of a verbal response. Such responses can be confounded by concomitant motor speech impairments, and give no indication of amount of cognitive effort being expended. For decades, pupillometry (i.e., the measurement of pupil dilation/constriction during a task) has been used to index cognitive effort (Beatty, 1982; Kahneman & Beatty, 1966), but has not yet been used in studies including people with aphasia (PWA). The purpose of this study was to examine the feasibility of using pupillometry to examine cognitive effort in the context of a working memory (WM) span task. We hypothesized that pupil dilation would increase as WM demands increased. METHODS Participants. Three individuals with aphasia participated in this study: P1 = 57 year-old male, 12 years post-onset, mild fluent aphasia; P2 = 61 year-old male, 9 months post-onset, mild fluent aphasia; P3 = 74 year-old female, 2 years post-onset, moderate fluent aphasia. Working memory span task - behavioural. Participants were administered a picture span task, adapted from DeDe et al. (2014). A sequence of pictures (2-5) of concrete, mono-syllabic nouns were presented at a rate of approximately 2 seconds/picture, followed by a 3 x 3 array containing target and distractor pictures. Participants pointed to the pictures in the order they were presented. Five trials of each sequence length were presented; span capacity was defined as the level where 3/5 trials were correctly identified. Working memory span task – pupillometry. An EyeLink 1000+ (SR Research, Ltd.) eye-tracking system, sampling at a rate of 250 Hz was used to measure pupil dilation while participants completed a computerized version of the picture span task described above. Procedures were identical to the behavioural version, except participants were instructed to gaze at the pictures in the 3 x 3 array to indicate their selections. Average pupil dilation during the presentation of single pictures was calculated by subtracting baseline pupil size (minimum pupil size for each trial) from every sample of pupil size during this ‘loading’ (presentation) phase. Average change in pupil size across all trials of the same span length was then calculated and subjected to a one-way ANOVA for each participant. RESULTS Pupil dilation results for all three participants are presented in Figure 1. Span capacity was determined by behavioural results; average pupil dilation is plotted up to each individual’s span capacity. All 3 participants demonstrated a significant effect of span length on average pupil size: P1 (F (2, 4524) = 265.631, p < 0.0001); P2 (F (3, 7036) = 983.115, p < 0.0001); P3 (F (1, 2510) = 1570.317, p <0.0001). CONCLUSION In this study, three PWA completed a computerized picture span task while an eye-tracker measured pupil dilation. As short-term memory demands (i.e., span size) increased, average pupil size significantly increased in all three PWA. These data provide preliminary support for the use of pupillometry to gauge cognitive effort in PWA. A larger study of PWA and demographically-matched control participants is currently underway, allowing for analysis of change in pupil size within and between groups. Examination of cognitive effort will provide a more comprehensive understanding of the nature of linguistic and cognitive functioning in aphasia.