Background
The natural hearing process is binaural. Normal-hearing (NH) listeners integrate their right and left inputs at the level of the brainstem allowing accurate sound localization and good speech recognition scores in noisy environments. If the brain has two different or competing signals, it will probably have a certain component of interference while trying to process them. If the hearing process involves competing or interfering signals, the listening will request effort and attention to reach a meaningful auditory percept.
Objective
We aimed to address binaural interference by measuring listening effort during speech recognition with the pupil dilation as its proxy.
Method
Speech in (babble) noise will be measured at different signal-to-noise (SNR) ratios with bilateral input and unilaterally. The SNR for 50 % word recognition will be measured and then, in steps of 5 dB, varied to higher and lower ratios looking for easy and difficult situations. While participants perform the speech-in-noise test, pupil dilation is recorded as an indicator of the listening effort. A novel pupillometry set-up will be implemented with virtual-reality (VR) glasses, which control the visual field more precisely. Initially, we will simulate CI processing in NH listeners. Later, and using the NH data as controls, SSD-CI users will be tested.
Results
We expect to see an increased listening effort when the bilateral input is less correlated, what we consider an indication of binaural interference. Preliminary data will be presented at the conference.