Measuring cochlear implant microphone performance: the accuracy of current methods and implications of inaccuracy

Abstract

Cochlear Implant (CI) microphones are frequently replaced in clinic, but how many of these are actually faulty? How many of these are replaced unnecessarily? How accurate and consistent are current methods used by CI clinics for measuring microphone functionality? This project investigates these questions and also investigates alternative methods for measuring CI microphone performance. First, the accuracy of current methods for assessing CI microphones are investigated using a survey, measuring microphones that have been reported as broken and testing the accuracy of subjective microphone checks. 57% of the microphones that werereported as broken were within ± 3dB of reference values (i.e. actually working). Furthermore, 90% of clinicians surveyed said they had replaced a microphone that theythought was working. The accuracy of subjective microphone checks was also investigated by presenting 10 microphones that were either “Working”, “Partially working” or “Not working”. Both the control group (n=10) and CI clinicians (n=4) were both able to identify the “Not working” microphones with 100% accuracy. Both groups were significantly less accurate when differentiating the other two groups of microphones. Previous testing showed that 17% of times that “Partially working” microphones were presented to clinicians they were classified as working. Impact of microphones withundetected defects on speech perception is analysed. Partial microphone failures were simulated during a speech perception test. This may mean that partial failures go unreported, and that users are struggling to hear, ascribing it to the situation rather than the microphone. Objective tests on a CI saw which of the user customizable settings affect the output of the Listening Check adaptor that was used extensively during the experiments. While the current CI microphone selection affects the adaptors’ output, none of the other settings were shown to have a significant effect on the output. The consistency CI frequency responses could be recorded was also assessed and over 56 recordings there was a maximum deviation of ± 2dB and an average deviation of ± 0.8dB. We need to develop better software based systems to be integrated into clinical processes around the testing of microphones both to reduce waste, avoid partially working microphones being overlooked, and to increase patient and clinician confidence in their equipment. These testing processes could be embedded into home monitoring systems to enable more regular and effective testing of implant microphones.

More information

Identifiers

ORCID for Mark Weal: orcid.org/0000-0001-6251-8786

Catalogue record

Export record