gms | German Medical Science

To enable earlier diagnosis and treatment of hearing loss, and address the average delay of 8.9 years between hearing-aid candidacy and hearing aid adoption [1], it is crucial to provide the general population with an easily accessible assessment tool to self-monitor their hearing capabilities. A smartphone based self-monitoring approach, i.e., a virtual hearing clinic, could provide listeners with a characterization and classification of their hearing loss, as well as a simulation of a personalized hearing aid to demonstrate the benefit a hearing aid could offer. To provide an accurate characterization of hearing loss without extensive testing, auditory profiles aim to represent distinct groups of audiological patients from different clinical databases. Using the information obtained during the mobile assessment, listeners will be matched into the auditory profiles. Within each profile, listeners are maximally similar to each other regarding measurement ranges of audiological tests, while retaining distinctive features across profiles.

Model-based clustering resulted in a first set of auditory profiles, which separate listeners within certain ranges across audiogram data, loudness scaling, speech tests, and anamnesis questions. Further, listeners can be classified into the auditory profiles with high precision. The generated profiles, thus, support the idea to summarize information across databases and may allow for an accurate diagnostic classification and provision of simulated hearing aids within the virtual hearing clinic, even if fewer tests are performed by listeners. In subsequent studies, additional databases will be incorporated and the profiles will be expanded with mobile and aided measurements, fitting parameters and further information from databases.