Effects of analysis window on 40-Hz auditory steady-state responses in cochlear implant users

Wouter David,Elise Verwaerde,Robin Gransier,Jan Wouters

DOI: https://doi.org/10.1016/j.heares.2023.108882

Abstract:Auditory steady-state responses (ASSRs) are phase-locked responses of the auditory system to the envelope of a stimulus. These responses can be used as an objective proxy to assess temporal envelope processing and its related functional outcomes such as hearing thresholds and speech perception, in normal-hearing listeners, in persons with hearing impairment, as well as in cochlear-implant (CI) users. While ASSRs are traditionally measured using a continuous stimulation paradigm, an alternative is the intermittent stimulation paradigm, whereby stimuli are presented with silence intervals in between. This paradigm could be more useful in a clinical setting as it allows for other neural responses to be analysed concurrently. One clinical use case of the intermittent paradigm is to objectively program CIs during an automatic fitting session whereby electrically evoked ASSRs (eASSRs) as well as other evoked potentials are used to predict behavioural thresholds. However, there is no consensus yet about the optimal analysis parameters for an intermittent paradigm in order to detect and measure eASSRs reliably. In this study, we used the intermittent paradigm to evoke eASSRs in adult CI users and investigated whether the early response buildup affects the response measurement outcomes. To this end, we varied the starting timepoint and length of the analysis window within which the responses were analysed. We used the amplitude, signal-to-noise ratio (SNR), phase, and pairwise phase consistency (PPC) to characterize the responses. Moreover, we set out to find the optimal stimulus duration for efficient and reliable eASSR measurements. These analyses were performed at two stimulation levels, i.e., 100% and 50% of the dynamic range of each participant. Results revealed that inclusion of the first 300 ms in the analysis window leads to overestimation of response amplitude and underestimation of response phase. Additionally, the response SNR and PPC were not affected by the inclusion of the first 300 ms in the analysis window. However, the latter two metrics were highly dependent on the stimulus duration which complicates comparisons across studies. Finally, the optimal stimulus duration for quick and reliable characterization of eASSRs was found to be around 800 ms for the stimulation level of 100% DR. These findings suggest that inclusion of the early onset period of eASSR recordings negatively influences the response measurement outcomes and that efficient and reliable eASSR measurements are possible using stimuli of around 800 ms long. This will pave the path for the development of a clinically feasible eASSR measurement in CI users.