A binaural loudness model for evaluating round-window stimulation based on a physiological peripheral auditory front-end module.

Some active middle-ear implants use round-window stimulation to extend their application to mixed hearing loss. However, current loudness models are unable to predict loudness under round-window stimulation, which prevents the development of fitting procedures. This paper presents a new loudness model for round-window stimulation. The model consists of a peripheral auditory front-end module, which translates the signal picked up by the microphone to round-window stimulation and to effective cochlear response, and a loudness perception back-end module for converting the cochlear response into loudness. The accuracy of the front-end module was validated by comparing the predicted results with experimental data on cochlear responses to a variety of stimuli. The accuracy of the back-end module was validated by comparing the predicted results with experimental data on the loudness of pure tones and noise. The increase in loudness with increasing drive voltage amplitude under round-window stimulation was consistent with that for acoustic stimulation, but the shapes of the equal-loudness contours were different from those for acoustic stimulation, reflecting the different transfer functions and the characteristics of the transducer. When a frequency-dependent correction was applied to allow for this, the predicted equal-loudness contours were similar to those for acoustic stimulation. Additionally, the model-predicted results for diotic and dichotic noise exhibited slight deviations from those for acoustic stimulation, whereas the model-predicted results for pure tones aligned more consistently with those for acoustic stimulation. This research forms the foundation for the development of fitting procedures for round-window stimulating active middle ear implants.