Pupil Dilation Indicates Individual Effort and Success During Speech Audiometry in Cochlear Implant Users - OMAI 2018- The 10th International Symposium on Objective Measures in Auditory Implants

Michel Hoen ⁴ Russo Francesca Yoshie ^1,2,3 Thomas Demarcy ⁴ Marine Ardoint ⁴ Caroline Desbrosses ² Maria-Pia Tuset ² Daniele De Seta ^1,2,3 Olivier Sterkers ^1,2 Ghizlène Lahlou ^2,5 Isabelle Mosnier ^1,2

¹Unité de Réhabilitation chirurgicale mini-invasive robotisée de l’audition, Sorbonne Université, Inserm, Paris, France
²Service ORL, Otologie, Implants auditifs et chirurgie de la base du crâne, AP-HP, GHU Pitié-Salpêtrière, Paris, France
³Department of Sense Organs, Sapienza University of Rome, Roma, Italy
⁴Clinical and Scientific research group cochlear implants, Oticon Medical, Vallauris, France
⁵Unité de génétique et de physiologie de l’audition, Sorbonne Université, Institut Pasteur, Inserm, Paris, France

Background: Cochlear implant (CI) outcomes are mostly evaluated through psychophysical testing, including auditory thresholds and speech perception in quiet. For a given performance level however, different individuals might show very different levels of motivation, cognitive engagement and listening effort. As a result, laboratory obtained speech intelligibility scores may only poorly reflect the real-life listening experience of CI users.

Objective: To assess the clinical relevance of pupillometry as an objective measure to evaluate speech perception abilities and associated listening effort in CI users.

Methods: The present experiment combined pupillometric measures to speech audiometry using word recognition either with a +10 dB SNR noise or in quiet.

Results: We found that different parts of the evoked pupil response were showing different behavior depending on the listening condition. Sensitivity to noise was reflected during the background period, before the occurrence of the target word. Then we found significant correlations between average relative peak dilation and average peak latency and speech recognition scores. Finally, the relaxation period, the relative return to baseline following event-related peak dilation, was modulated by success or failure with a surprisingly high precision. The average pupil dilation during the relaxation phase directly indicated how many phonemes patients had understood.

Conclusion: These results will be discussed in the context of current models of listening effort and in regards of the potential applications of pupillometry for the objective measure of performance of CI users as well as for other potential applications such as screening, diagnostic or improvement of mapping parameters.