gms | German Medical Science

Introduction: Loudness constitutes a core aspect of auditory perception. Sensorineural hearing loss results in a reduced dynamic range and, therefore, a more rapid increase in loudness than normal. Patients with cochlear implants (CIs) also have a reduced dynamic range, leading to difficulties in loudness perception. However, the primary aim of a hearing aid (HA) or CI provision is to restore speech recognition (SR), not normal loudness perception. Nonetheless, close-to-normal loudness perception can increase performance in SR tasks and lead to higher acceptance of HA/CI provision. For a normalization of loudness in CI users, a sensible target is essential. This ongoing study investigates the loudness perception with special focus on binaural broadband loudness summation of bilateral CI and bimodal CI users (CIs with a contralateral HA), using a novel loudness validation method (revoloud) [1]. Revoloud proposes using natural sounds for categorical loudness scaling, aiming for a high ecological validity. Furthermore, revoloud considers binaural loudness summation effects of broadband (BB) signals and provides reference values of normal-hearing (NH) listeners, that can be used as a fitting target.

Methods: So far, four adult CI users (two bilateral, two bimodal) have participated in the measurements. Loudness perception is measured using revoloud, which uses natural signals and considers narrowband and BB conditions at presentation levels of 50, 65 and 80 dB SPL. Furthermore, an adaptive revoloud procedure is used, that adjusts a gain table (G50, G65 and G80) for low, middle and high frequency band towards reference loudness ratings of NH people. This gain table is applied to the test signals using a three-band compressor. Both conditions are measured in free field for each ear separately (monaural conditions) and in binaural listening mode.

Results: Participants are generally able to perform the revoloud measurements in a meaningful manner. Deviations from the NH reference can be observed for low-frequency and BB signals, with a tendency to perceive loud signals as excessively loud. Across most conditions, the ratings appear to be dominated by the better ear, except for the 65 dB and 80 dB BB signals. The loudness ratings used for the compressor-setting calculation closely matched the NH reference. First results indicate a tendency towards convergence to the NH reference using the adaptive revoloud procedure.

Conclusions: Performing the revoloud measurement is feasible for both bilateral and bimodal CI users and yields meaningful results. The adaptive procedure approximates towards the NH reference. It will further be investigated how the procedure can be integrated to optimize CI fitting in bilateral and bimodal CI recipients.

Research funded by Deutsche Forschungsgemeinschaft (DFG, German Research Foundation), Project ID 352015383, SFB 1330 Project C4.