Binaural audio frontend processing for cochlear implants inspired by the medial olivocochlear reflex

Enrique A Lopez-Poveda^1,2,3*Almudena Eustaquio-Martín^1,2Milagros J Fumero^1,2Reinhold Schatzer⁴Joshua Stohl⁵Christian Wirtz⁴Peter Nopp⁴

• ¹Instituto de Neurociencias de Castilla y León, University of Salamanca, Salamanca, Spain
• ²Instituto de Investigacion Biomedica de Salamanca, Salamanca, Spain
• ³Universidad de Salamanca, Departamento de Cirugía, Facultad de Medicina, Salamanca, Spain
• ⁴MED-EL Elektromedizinische Gerate GmbH, Innsbruck, Austria
• ⁵North American Research Laboratory, MED-EL Corporation, Durham, NC, United States

Objectives. The "MOC strategy" is a binaural audio processing method for cochlear implants (CI) inspired by the medial olivocochlear (MOC) reflex. Although the strategy can improve hearing for CI users, it requires a bilateral channel pairing for the contralateral control of the backend compression and, therefore, the same number of frequency channels and electrodes on both sides. This may limit its clinical applicability. As a step to overcome this difficulty, here, we present and test a version of the MOC strategy designed to operate at the front end (FE) rather than the back end of processing, referred to as the MOCFE. Design. The MOCFE strategy was implemented and tested in combination with a pair of functionally independent bilateral MED-EL FS4 audio processors. The hearing of bilateral CI users was compared for the MOCFE as well as for two reference strategies: the standard FS4 strategy and a backend MOC strategy. The MOCFE and MOC strategies were implemented with fast contralateral inhibition. Measures included (1) speech reception thresholds for sentences in fluctuating and stationary noise, in unilateral and bilateral listening modes, and for three different speech levels; and (2) sound source localization in quiet and in noise. Five bilateral users of MED-EL CIs participated in the evaluations. Results. (1) For speech intelligibility in fluctuating noise, the MOCFE was as beneficial (re FS4) as the backend MOC strategy, except for one specific spatial configuration (S-60N60), where the MOCFE produced no benefits while the MOC strategy did. (2) Neither the MOCFE nor the MOC strategies improved intelligibility in a stationary noise, possibly because they involved fast rather than slow contralateral inhibition. (3) The binaural MOCFE and MOC strategies tended to improve sound source localization slightly relative to the FS4 strategy. Conclusion. The MOCFE can become a successful alternative to the MOC strategy, as it may be more easily implemented in clinical devices and applied to a broader range of clinical map configurations.