Finished | Technology

Diagnostic test kit for clients with hearing in noise difficulties and normal audiograms

Project Goals 

Many people attend hearing clinics complaining of difficulty hearing in noise, even though their hearing thresholds, as measured by conventional tests, is normal or nearly normal. There are currently no tests or treatments available to address this problem, and both clients and audiologists are frustrated by the lack of solutions. This project aimed to create a diagnostic tool to identify individuals who have normal or nearly normal hearing, but still struggle to hear in noise. 

In this study, two groups of participants were recruited: a test group who reported having difficulty hearing speech in noise and a control group who did not have this problem. Both groups had normal or nearly normal hearing. The test group underwent a series of tests to determine whether there were any abnormalities in their auditory system (including the eardrum, auditory nerve, and brain) that might be causing their difficulty hearing in noise. These tests included questionnaires, hearing tests, tests of the acoustic reflex (a reflexive response of the middle ear muscles to loud sounds), a speech-in-noise test, cognitive tests, and electrophysiological tests (tests that measure the electrical activity of the auditory system). The results of these tests were compared between the test group and the control group to see if there were any differences that could help explain the test group’s difficulty hearing in noise. 


The results of the study showed that the group who reported having difficulty hearing speech in noise had a higher rate of noise exposure and performed worse on a task measuring sustained attention compared to the control group. The biggest differences between the two groups were seen in the scores on the questionnaires. Using machine learning, the researchers identified a subset of five questions from the two questionnaires that could accurately identify individuals with difficulty hearing in noise with 95% accuracy. These questions were found to be significantly correlated with at least one electrophysiological measure associated with hearing damage that is not detectable using a standard hearing test (audiogram). This suggests that the questionnaires can detect types of hearing damage that may not be evident on an audiogram. 

Related Projects & news