Low-Cost Pupillometric Measurement of Listening Effort: A Validation Study

Background: Pupillometry has recently shown great promise for the non-invasive measurement of listening effort in psychoacoustic experiments. Current pupillometric methods yet rely on costly, high-end eye-tracking equipment, which prevents the generalized adoption of the technique by researchers and clinicians.

Objectives: We aim to validate the low-cost, wearable PupilLabs^TM pupillometry system against a high-end, gold-standard solution for listening effort assessment (SMI iView X RED)

Methods: 7 normal-hearing participants took the Hearing In Noise Test with 4-talker babble noise in 3 different signal-to-noise ratios (70 dB SPL -4/+0/+4 dB speaker vs. masker). Baseline-normalized Peak Pupil Dilation (PPD) was considered as a proxy to listening effort. PPD was assessed in all 3 conditions using sequentially the SMI and the PupilLabs^TM wearable system. Data post-processing was carried out consistently across both eye-tracking modalities. PPD values were computed on a 25-sentence average basis. We then carried out statistical testing for determining constant bias, proportional bias and to assess the degree of similarity between measurements.

Results: PPD values recovered with the SMI and PupilLabs^TM systems did not show significant constant (p=0.58) or proportional bias (p=0.65). The 95% CI for the difference between PPD values assessed by PupilLabs and SMI was [-0.045 0.033] normalized arbitrary units (df = 16).

Conclusions: The PupilLabs^TM system offers a valid low-cost alternative to high-end pupillometry equipment for assessing listening effort during the Hearing in Noise Test.