Audiology applications of conversational Artificial Intelligence: remote, natural and automated testing of hearing and device prescription

Funder:

Medical Research Council – Doctoral Training Partnership.

Staff involved:

Mohsen Fatehifar, Josef Schlittenlacher, David Wong, Timothy Cootes, Kevin Munro.

Project overview:

This project aims to propose a new way of conducting hearing tests by using text-to-speech and automatic speech recognition models, without the need to involve a hearing professional. The project enables remote, natural, and automated testing of hearing capability.

These technologies can also be used to evaluate various configurations of hearing devices without any human interaction. For example, the technology can be used to prescribe amplification for someone with a hearing loss, in a way that provides the best speech intangibility to the patient.

Extended high-frequency hearing

Funder:

NIHR Manchester Biomedical Research Centre in Hearing Health, National Institutes of Health.

Staff involved:

Melanie Lough, Garreth Prendergast, David Moore.

Project overview:

Extended high-frequency (EHF) hearing, i.e. beyond the range of hearing thresholds that are currently measured in clinical audiometry, may be a sensitive predictor of age-related hearing loss that can identify individuals at risk of hearing loss much earlier in life. EHF hearing may be used to predict which individual young adults, adolescents or even children are likely to develop hearing loss in the conventional range of frequencies (0.25 – 8 kHz) later in life.

This could lead to early prevention on an individualised basis. Preventative measures might include advice on protecting hearing or, in the future, taking certain drugs found to prevent hearing loss. We have also found that preventing access to EHF results in poorer performance (SRT) on speech-in-noise, as tested with the DIN.

The effects on auditory function of radiotherapy and chemotherapy treatments for head and neck tumours (EARAD)

Funder:

Marston Scholarship, BRC Advanced Radiotherapy and Hearing Health themes.

Staff involved:

Jenna Littlejohn, Chris Plack, Catharine West, James Price, Marianne Aznar, Lip Wai Lee. 

Project overview:

The EARAD project is designed to evaluate the effects of radiotherapy and combined chemoradiotherapy treatments (CCRTs), both known to cause damage to the cochlea and auditory brainstem (ototoxicity).

The overall aim is to provide data that will define the amount of radiotherapy needed to damage individual substructures of the auditory pathway so that a better balance can be made between ototoxic hearing loss and tumour control during radiotherapy for head and neck cancer.

The feasibility of estimating the prevalence of hearing loss in people with dementia

Funder:

NIHR Manchester Biomedical Research Centre.

Staff involved:

Hannah Cross, Piers Dawes, Laura Brown, Chris Armitage, Iracema Leroi, Rebecca Millman.

Project overview:

The risk of developing dementia and hearing loss increases with age, leading to a significant overlap of these conditions. Detecting and managing hearing difficulties in individuals with dementia is crucial, as untreated hearing loss can exacerbate confusion, depression, social isolation, and dementia-related symptoms.

However, the prevalence rate of hearing loss in people with dementia remains unknown in the UK. Prior to conducting a large-scale study to estimate this prevalence, we will assess the reliability and acceptability of hearing tests for this population and explore participant recruitment and retention strategies, both for people living in the community and in care homes.

We will also use qualitative interviews to understand the reasons behind unidentified hearing loss in people with dementia, aiming to identify barriers to engaging with audiology services. The results will be used to propose recommendations for improved hearing assessments and audiology pathways tailored to individuals with dementia.

Behavioural audiometry measures in babies: Innovation, novelty, and optimisation (BAMBINO)

BAMBINO logoFunder:

William Demant Foundation.

Staff involved:

Anisa Visram, Iain Jackson, Michael Stone, Tim Cootes, Kevin Munro, Helen Whiston, Josef Schlittenlacher.

Project overview:

The BAMBINO study works with families and children under two years to improve hearing tests for babies. Our goal is to increase the amount of information audiologists obtain during hearing assessment by making tests more fun and engaging, and by exploring new ways of measuring whether or not babies respond to sounds.

Our work examines how sounds and images that are meaningful to children, like the theme from a favourite TV show, can increase engagement during hearing tests, and how machine learning can help to interpret subtle changes in babies’ faces and behaviour in response to sounds.