Pupil Dilation and Microsaccades Provide Complementary Insights into the Dynamics of Arousal and Instantaneous Attention during Effortful Listening

Abstract

Listening in noisy environments requires effort – the active engagement of attention and other cognitive abilities, as well as increased arousal. The ability to separately quantify the contribution of these components is key to understanding the dynamics of effort and how it may change across listening situations and in certain populations. We concurrently measured two types of ocular data in young participants (both sexes) - Pupil dilation (PD) (thought to index arousal aspects of effort) and Microsaccades (MS) (hypothesized to reflect automatic visual exploratory sampling) whilst they performed a speech-in-noise task under high- (HL) and low- (LL) listening load conditions. Sentences were manipulated so that the behaviourally relevant information (keywords) appeared at the end (Exp1) or beginning (Exp2) of the sentence, resulting in different temporal demands on focused attention. In line with previous reports, PD effects were associated with increased dilation under load. We observed a sustained difference between HL and LL conditions, consistent with increased phasic and tonic arousal. Importantly we show that MS rate was also modulated by listening load. This was manifested as a reduced MS rate in HL relative to LL. Critically, in contrast to the sustained difference seen for PD, MS effects were localised in time, specifically during periods when demands on auditory attention were greatest. These results demonstrate that auditory selective attention interfaces with the mechanisms controlling MS-generation, establishing MS as an informative measure, complementary to PD, with which to quantify the temporal dynamics of auditory attentional processing under effortful listening conditions.SIGNIFICANCE STATEMENTListening effort, reflecting the “cognitive bandwidth” deployed to effectively process sound in adverse environments, contributes critically to listening success. Understanding listening effort and the processes involved in its allocation is a major challenge in auditory neuroscience. Here we demonstrate that microsaccade rate can be used to index a specific sub-component of listening effort - the allocation of instantaneous auditory attention - that is distinct from the modulation of arousal indexed by pupil dilation (currently the dominant measure of listening effort). These results reveal the push-pull process through which auditory attention interfaces with the (visual) attention network that controls microsaccades, establishing microsaccades as a powerful tool for measuring auditory attention and its deficits.

Pupil size and eye movements differently index effort in both younger and older adults

“…The pattern of gaze dispersion for the difficult SNR is consistent with the idea that eye movements are responsive to changes in effort. Gaze dispersion was reduced in the difficult condition, relative to the easy condition (Contadini-Wright et al, 2023; Cui & Herrmann, 2023), and then increased to similar levels as in the easy condition after sentence offset (Figures 2, 3, and 5). Yet, the continued reduction in gaze dispersion for the impossible condition suggests that eye movements may also be responsive to other aspects of cognition, such as monitoring.…”

“…Recent works suggest that reductions in eye movements may index effort. Eye movements decrease when individuals perform a high-load compared to a low-load memory task (Hutton & Tegally, 2005; Kosch et al, 2018; Lipton, Frost, & Holzman, 1980; Walter & Bex, 2021) and when speech masking increases (Contadini-Wright et al, 2023; Cui & Herrmann, 2023). We also found that gaze dispersion decreased with decreasing SNR (Figures 2, 4, and 5), and extend this previous work by showing that the sensitivity of eye movements to changes in speech masking does not differ between younger and older adults.…”

“…Gaze dispersion is a broad measure that captures any eye movements. Analyses of saccades (including microsaccades) require setting a velocity threshold to determine what counts as a saccade (Contadini-Wright et al, 2023; Cui & Herrmann, 2023; Engbert, 2006; Engbert & Kliegl, 2003; Kadem et al, 2020; Widmann, Engbert, & Schröger, 2014), but the threshold can influence whether or not effects of cognition on saccades can be observed (Cui & Herrmann, 2023). Nonetheless, we explored the sensitivity of saccades to changes in SNR and results were very similar to the results for gaze dispersion.…”

“…Eye movements can be automatically triggered or volitionally initiated (Pierce et al, 2019). Eyemovement reductions with increasing task load may include involuntary eye movements, as suggested by observations of task-difficulty related eye-movement reductions during fixation on a point on a computer monitor (Contadini-Wright et al, 2023;Cui & Herrmann, 2023). However, individuals may also voluntarily reduce eye movements under high cognitive demands, for example, to free resources associated with eye movements, avoid processing irrelevant information, and direct attention internally during listening (stereo listening via headphones is devoid of localizable external sound sources).…”

“…Recent work suggests that the quantification of eye movements may provide an alternative effort metric. Eye movements decrease during periods of high compared to low memory load (Kosch, Hassib, Woźniak, Buschek, & Alt, 2018; Walter & Bex, 2021) and with increasing speech masking (Contadini-Wright, Magami, Mehta, & Chait, 2023; Cui & Herrmann, 2023). This research has focused mainly on increases in effort as task demands increase.…”