Computational Audiology Network
22 September 2024
WCA Related Society session
•
S113
•
10:00
>
11:00
•
Computational Audiology Network
•
Dragonnier 4 – 236
WCA Related Society session
The mission of the Computational Audiology Network (CAN) is to act as a global network dedicated to advancing hearing healthcare through the application of data science, computational methods, and artificial intelligence (AI) in hearing research and audiological technologies. CAN brings together academics, clinicians, industry partners, policymakers, and patients to foster innovation, promote accessible hearing healthcare solutions, and improve the quality of life for individuals with hearing loss.
This special session will highlight current research on machine learning for improving speech and music perception, personalized neural network algorithms for hearing aids, automatic audiometry and AI chatbots for hearing healthcare. The presentations will be followed by a short panel discussion with the speakers
•
INT58
•
Improved sound quality in personalized neural-network-based hearing-aid algorithms
>
S.
Sarah
VERHULST (Belgium)
Content : With the rise of neural-network solutions for hearing-aid compensation comes a challenge of how to tailor these algorithms to the hearing damage profile of the individual. We present a method to individualize the algorithms using outer-hair-cell-damage and cochlear synaptopathy estimates from the patient. Secondly, we focus on how to optimize the sound quality of the resulting algorithms for use with patients and in the next generation of devices
•
INT59
•
The Cadenza and Clarity projects: Using machine learning challenges to improve music and speech for those with hearing loss
>
S.
Simone
GRAETZER (Uk)
Content : In the Clarity and Cadenza projects, we are seeking to improve hearing aid processing of speech and music via open machine learning challenges. These projects involve collaboration between researchers at the Universities of Salford, Sheffield, Cardiff, Leeds and Nottingham (U.K.). Clarity focuses on enhancing and predicting speech intelligibility, while Cadenza focuses on understanding what people with hearing loss want from music and improving their music listening by means of signal processing and remixing approaches. The second Cadenza Challenge will be open to entries during the WCA meeting, offering an opportunity to those interested in participating
•
INT57
•
Automating air- and bone-conduction audiometry with masking
>
N.
Nihaad
PARAOUTY (Reims)
Content : Can we use machine-learning to improve audiometry tests today?
Can we use those machine-learning audiometry tests in clinical practice?
Can those machine-learning audiometry tests improve the clinical diagnosis and patient management?
•
INT178
•
AI chatbots delivering Hearing Healthcare?
>
D.
De Wet
SWANEPOEL (South Africa)
Content : This presentation explores the transformative potential of AI chatbots in delivering hearing healthcare, particularly in low- and middle-income countries (LMICs). By leveraging rapidly advancing AI technology and widespread mobile penetration, these digital tools can bridge the gap in healthcare accessibility, offering early detection, education, and support in hearing health. The talk will highlight the role of AI chatbots in addressing the global shortage of skilled hearing healthcare professionals and discuss the implications for broader adoption in regions where traditional healthcare services are limited
|