An individualised acoustically transparent earpiece for hearing devices

Abstract

Objective: An important and often still unresolved problem of hearing devices such as assistive listening devices and hearing aids is limited user acceptance – a primary reason is poor conservation quality of the acoustic environment. Approaching a possible solution to this problem, an earpiece prototype is presented and evaluated. The prototype is individually and automatically calibrated in situ to provide acoustical transparency, i.e., achieving an audio perception alike to the open ear. Design: A comprehensive evaluation was performed, comprising technical measurements on an advanced dummy head and listening tests, in which listeners directly compared sound perception through the prototype and a simulated open ear canal reference. Study sample: Ten normal hearing subjects, including five expert listeners, participated in the listening test. Results: The technical evaluation verified good achievement of acoustical transparency. The psychoacoustic results showed that a reliable distinction between the two conditions presented was not possible for relevant communication sounds. Conclusion: The prototype can be described as an initial realisation of an acoustically transparent hearing system, i.e. a device that does not disturb the perception of external sounds. In further developments, the device can be considered as the basis for systems integrating high sound quality, hearing support and other desired modifications.

Key Words:

• Assistive technology
• hearing aids
• hearing aid satisfaction
• psychoacoustics/hearing science

Acknowledgements

This work was supported by BMBF 13EZ1127D and DFG FOR1732 (Research Unit “Individualized Hearing Acoustics”). We thank Tobias Herzke from Hoertech gGmbH, Oldenburg, Germany for continuous support on the Master Hearing Aid (MHA). Furthermore, we thank Volker Hohmann and Giso Grimm for fruitful discussions.

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

Notes

1. Future work will include the reconstruction of individual ear canal resonances, which may be feasible through electro-acoustic modelling of the prototype (Vogl et al, 2016), impedance measurements of the residual ear canal and eardrum, and knowledge of the otoplastic geometry.

2. The pressure at the eardrum can be estimated using the in-ear microphone, see the “Pressure at the Eardrum” section.

3. The subject’s responses were modelled as independent guesses, with a probability of correct assignment P. Consequently, the total number of correct trials follows the cumulative Bernoulli distribution. The data model delivers confidence intervals for hypotheses on the assignment accuracy P, i.e., the number of correct assignments contradicting the assumption with a chosen probability, here 99%. Also, an upper boundary for P can be estimated by comparing the observed correct assignments with distributions in dependence of P.