The University of Southampton
Engineering and the Environment

Research project: Objective measures of hearing aid benefit

Currently Active: 

We would like to develop a set of objective test methods that can be used to compare the various features of modern hearing aids and to predict which features will best improve patient benefit from hearing aids.

Project Overview

Adaptive feedback cancellation is an effective solution to the problem of acoustic feedback in hearing aids. However, adaptive cancellers are inherently prone to entrainment, a non-linear distortion artefact produced by tonal or highly periodic signal inputs such as music. Owing to the expense and variability of subjective tests (using human volunteers), a fully objective method of measuring the sound quality of signals processed by adaptive feedback cancellers is required. This project aims to produce an objective metric able to predict subjective quality ratings of normal and hearing impaired listeners for audio signals processed by adaptive feedback cancellers in hearing aids. Signal-processing algorithms for predicting subjective sound quality ratings are commonly 'trained' using a learning database of distorted sound samples, paired with examples of quality ratings obtained from actual human subjects. Generalised predictions of the subjective quality of arbitrary sound samples are made with reference to this database. Development of a metric that is 'sensitive' to the nonlinear distortion artefacts produced by adaptive feedback cancellers will require a training database of audio samples containing feedback canceller artefacts, and a corresponding array of subjective quality ratings. The training array of sound samples has been created by processing audio files of orchestral instruments though a MATLAB model of a linear hearing aid amplifier using adaptive feedback cancellation, coupled to a simulated acoustic feedback path. Collection of subjective ratings data for these audio samples is due to begin this month. The structural design of the quality prediction metric will incorporate a psychoacoustic model, and will borrow techniques used in pre-existing existing objective quality evaluation metrics designed for telephony and hearing aid signal processing applications. These techniques will be either adapted, or optimised for application to feedback-canceller-processed audio signals using the training database described. - Recordings of music samples processed by commercial hearing aids using adaptive feedback cancellation have been obtained using a KEMAR mannequin. Similarities were observed between these signals and the output of the MATLAB canceller model used to create the training database. - Informal listening tests indicate that measurements of SNR (signal to noise ratio) provide a poor indication of the subjectively perceived level of distortion at the canceller model output. The similarities between the model and actual hearing aid outputs indicate that the MATLAB model provides a reasonable representation of the characteristics of entrainment artefacts in 'real' hearing aids. The poor association between SNR and the level of perceived signal distortion from non-linear systems is well known, and highlights the importance of using a 'perceptual' modelling approach that is considered in this project. Normal hearing individuals will be used to obtain an initial training database. At a later stage, I intend to obtain data from hearing impaired subjects, which will require alterations to the psychoacoustic model. (see 2.) To develop an objective measure of subjective audio quality ratings of sound processed by adaptive feedback cancellers in digital hearing aids, that is robust for both normal, and hearing impaired listeners.

Related research groups

Human Sciences Group


Share this research project Share this on Facebook Share this on Google+ Share this on Twitter Share this on Weibo

We use cookies to ensure that we give you the best experience on our website. If you continue without changing your settings, we will assume that you are happy to receive cookies on the University of Southampton website.